Professional Documents
Culture Documents
Chapter 21
Introduction
All crude oils are composed primarily of hydrocarbons, which are made by the combination of the elements carbon and hydrogen. In addition, most crudes contain sulfur compounds and trace quantities of oxygen, nitrogen, and heavy metals. The difference in crude oils is caused by the amount of sulfur compounds and by the types and molecular weights of the hydrocarbons making up the oil. The hydrocarbons found in crude oil range in size from the smallest molecule, methane, which contains 1 atom of carbon, to the largest ones, which contain nearly 100 atoms of carbon. The types of hydrocarbon compounds are paraffin, naphthene, and aromatic, found in raw crude, and olefin and diolefin, which are sometimes found in refined products after thermal treatment. Since any crude oil will have several thousand different compounds in it, it has been impossible so far to develop exact analyses of the actual compounds present. Three methods of reporting analyses are available-ultimate analysis, chemical analysis, and evaluation analysis. Ultimate analysis lists the composition in percentages of the elements carbon, hydrogen, nitrogen, oxygen, and sulfur. This tells very little about the type of compounds present or the physical characteristics of the oil. It is useful, however, in determining the amount of sulfur that must be removed. Table 21.1 shows the ultimate analysis of several crude oils. Chemical analysis gives composition in percentage of paraffin, naphthene, and aromatic-type compounds present in the crude. This type of analysis can be determined with fair accuracy by means of chemical reaction and solvency tests. An analysis of this sort gives an idea of the usefulness of refined products but does not give any
This author also wrote the tiginal chapter on this topic in the 1962 edation.
means of predicting the amount of various refined products. Table 2 1.2 gives the chemical analysis of several fractions of four crude oils. The crude-oil evaluation consists primarily of a fractional distillation of the oil followed by physicalproperty tests (for parameters such as gravity, viscosity, and pour point) on the distillation products. Since the primary means of separating products in the refinery is fractionation, this analysis makes it possible to predict yields of refined products and physical properties studied in the evaluation. The evaluation curves shown in Fig. 2 1.1 make it possible to predict the physical properties of the refined products. As an example of the use of evaluation curves, Table 2 1.3 shows product yields and properties when a refinery is operated for maximum gasoline yield, and Table 2 1.4 shows product yields and properties when the objective is to produce lubricating oils and diesel fuel. Since the early 1970s, much research has-been performed on the use of the gas chromatograph to generate simulated distillations. This has the advantage of producing crude-oil evaluation curves with very small samples of crude and in a period of about an hour, compared with about a gallon of crude for a fractional distillation column and about 2 days for the analysis. The simulated distillation is called ASTM Test Method D2887. I
21-2
PETROLEUM ENGINEERING
HANDBOOK
TABLE Pl.l-ULTIMATE Specific Gravity -r 0.862 0.897 0.912 0.951 0.91 0.97 0.897 0.948
CHEMICAL ANALYSES
OF PETROLEUM Component WI N 0 3.6 0.37 0.76 0.60 0.70 4.30 1.5 0.54
Petroleum Pennsylvania pipeline Mecook, WV Humbolt, KS Healdton, OK Coalinga, CA Beaumont, TX Mexico Baku, USSR Colombia, South America
Combined mtrogen and oxygen.
Temperature PC) 15 0
15 15 20
--
naphthene naphthene
TABLE 21.2-CHEMICAL
Grozny (High Paraffin) 45.3% at 572OF Aromatic 3 z 14 18
17
ANALYSES OF PETROLEUM, %
Oklahoma California (Huntmgton Beach), 34.2% at 57ZF Paraffin 73 65 55 55 52 51
-A
Fraction (0 140 to 203 to 252 to 302 to 392 to 482 to 203 252 302 392 482 572
Grozny (ParaffinFree Upper Level), 40.9% at 572OF Paraffin 72 65 56 57 59 61 Aromatic 4 8 13 21 26 35 Naphthene 31 40 52 55 63 57 Paraffin 65 52 35 24 11 8 Aromatlc 5 7 12 16 17 17
(Davenport),
64% at 572OF Naphthene 21 28 33 29 31 32
Naphthene 25 30 35 29 23 22
Aromatic i
11
Naohthene 31 46 64 61 45 40
Paraffin 65 46 25 22 30 31
17 25 29
TABLE 21.3-EVALUATION
WHEN OPERATING PRIMARILY FOR GASOLINE Percent Distilled Gravity (OAPI) 56 28.8 6.4$ 32.0
Material Gas loss Straight-run gasoline (untreated) Catalytic charge V&breaker Crude oil charge or asphalt
Other Properties 390DF ASTM endpoint7 165OF aniline point or 47.5 diesel index 110 penetration 11.65 characterization factor
Topping follwed by YaWUrn flashing to produce a gas 011for catalflic cracking. breaker chargestock.
The Cycle stcck IrOm catalytic cracking is thermally cracked along wtth the asphalt or vis-
Average gravity from instantaneouscurve of API gravity. ?At about 400aF endpoint the truebOiling.pCint cut point is about 2PF higher than the ASTM end point
*By a material balance.
TABLE 21.4-EVALUATION
STOCK0
Midpoint
Other Properties
Light gasoline (untreated) Reforming naphtha Diesel fuel Light lube or cracking stock Lube stock (untreated) Asphalt Crude oil
remainder
100 Ws viscosity at 2lOOF
100 penetration
-13 1.3 21.0 10.5 38.5 29.7 56.5 47.5 74.9 65.7 80.9 77.9 100.0
63.8 octane numberd 0.16% sulfur 41 (estimated) 50 diesel Index; 0.82% sulfur 145 at 100F 1.49% sulfur 100 at 210F
21-3
TABLE 21.5-BASES
OF CRUDE OILS Approximate LowBoiling 12.2+ 12.2+ 12.2 + 11.5 to 12.0 11.4 to 12.1 11.4 to 12.1 11.511.511.4UOP* * HighBoiling 12.2+ 11.4 to 12.0 11.412.2+ 11.4 to 12.1 11.411.4 to 12.1 12.2+ 11.4-
Characterization Factor
Part
paraffin intermediate naphthene paraffin intermediate naphthene intermediate paraffin naphthene
1
40+ 40+ 40+ 33 to 40 33 to 40 33 to 40 333333-
its light materials and another set for the heavy-lube fractions, the USBM has developed a more useful method of classifying oils. Two fractions (called key fractions) are obtained in the standard Hempel distillation procedure. Key Fraction 1 is the material that boils between 482 and 527F at atmospheric pressure. Key Fraction 2 is the material that boils between 527 and 572F at 40 mm absolute pressure. Both fractions are tested for API gravity, and Key Fraction 2 is tested for cloud point. In naming the type of oil, the base of light material (Key Fraction 1) is named first, and the base of the heavy material (Key Fraction 2) is named second. If the cloud point of Key Fraction 2 is above 5F, the term wax-bearing is added. If the pour point is below 5F, it is termed waxfree. would mean Thus, paraffin-intermediate-wax-free a crude that has paraffinic characteristics in the gasoline portion and intermediate characteristics in the lube portion and has very little wax. Table 21.5 shows the criteria used in establishing bases of oil by the USBM method. Several attempts have been made to establish an index to give a numerical correlation for the base of a crude oil. The most useful of these is the characterization factor K developed in Ref. 2,
TABLE 21.6-TYPICAL CHARACTERIZATION FACTOR VALUES Characterization Factor 12.1 11.8 11 .o 11.5 10.5 10.0 9.8 to to to to to to to 12.5 12.0 11.6 11.8 11.5 11.0 11 .O
Product Pennsylvania stocks (paraffin base) Mid-Continent stocks (intermediate) Gulf Coast stocks (naphthene base) Cracked gasoline Cracking-plant combined feeds Recycle stocks Cracked residuum
number of worldwide crudes and products and typical hydrocarbon compounds that have the same characterization factor as the oil in question.
Physical Properties
Fig. 21.2 shows the relationship of carbon-tohydrogen ratio, average molecular weight, and mean average boiling point as a function of API gravity and characterization factor. The API Technical DataBook3 has published a number of correlations for physical properties of petroleum. For the most accurate data, this reference should be consulted. When oil is heated or cooled in a processing operation, the amount of heat required is best obtained by the use of the specific heat. Fig. 21.3 shows the specific heat of liquid petroleum oils as a function of API gravity and temperature. This chart is based on a characterization factor of 11.8, and if the oil being studied is other than that, there is a correction shown at the lower right side of the chart. The number obtained for the specific heat should be multiplied by this correction factor. Certain paraffin hydrocarbons are also shown on the chart. No correction need be applied to these. If vaporization or condensation occurs in a processing operation, the heat requirements are most easily handled by the use of total heats. Fig. 2 1.4 gives total heats of petroleum liquid and vapor, with liquid at 0F as a reference or zero point. This eliminates the necessity of selecting a latent heat, specific heats of both vapor and liquid, and deciding at what temperature to apply the latent heat. Certain corrections must be applied for characterization factor and for pressure.
K=- 3% Y
in which TB is the molal average boiling point (degrees Rankine) and y is the specific gravity at 60F. This has been used successfully in correlating not only crude oils, but refinery products both cracked and straight-run. Typical numerical values for characterization factors are listed in Table 2 1.6. In addition to the relationship between the characterization factor and the specific gravity and boiling point defined above, a number of other physical properties have been shown to be related to the chamcterization factor. Among these properties are viscosity, molecular weight, critical temperature and pressure, specific heats, and percent hydrogen. Table 21.7 shows characterization factors for a
21-4
PETROLEUM
ENGINEERING
HANDBOOK
TABLE 21.7-CHARACTERIZATION
Characterization Factor 14.7 14.2 13.85 13.5 to 13.6 13.0 to 13.2 12.8 12.7 12.6 12.55 12.5 12.1 to 12.5 12.2 to 12.44 12.0 to 12.2 11.9 to 12.2
PETROLEUMS,
Hydrocarbons propane propylene isobutane butane butane-l and isopentane hexane and tetradecened P-methylheptane and tetradecane pentene-1, hexene-1, and cetene 2,2,4-trimethylpentane hexene-2 and 1.3-butadiene 2,2,3,3tetramethyl butane 2,l l-dimethyl dodecadiene
Miscellaneous Products
94.5 API adsorption gasoline Four Venezuelan paraffin waxes paraffin wax*: MC. 82.2 API natural gasoline CA 81.9 API natural gasoline Cotton Valley (LA) lubes Pennsylvania-Rodessa Big Lake (TX) Lance Creek (WY) Mid-Continent (MC.) Oklahoma City (OK) (LA) debutanized E. TX natural gasoline San Joaquin (Venezuela) wax distillate Panhandle (TX) lubes Six Venezuelan wax distillates
paraffin-base gasolines
Middle East light products cracked gasoline from paraffinic feeds E. TX gas oil and lubes light cycloversion gasoline from M.C. feeds Middle East gas oil and lubes cracked gasoline from intermediate feeds E. TX and IA white products cracked gas oil from paraffinic feeds catalytic cycle stocks from paraffinic feeds cracked gasoline from naphthene feeds Tia Juana (Venezuela) gas oil and lubes naphthenic gasoline: catalytic (cracked) gasoline catalytic cycle stocks from MC. feeds cracked gasoline from hrghly naphthenrc feeds high-conversion catalytic cycle stocks from parafbnic feeds typical catalytic cycle stocks liaht-ail coil thermal feeds catalytic cycle stocks from 11.7~characterization-factor feeds gasoline from catalytic re-forming
11.9 11.8 to 12.1 11.85 11.7 to 12 11.75 11.7 11.6 11.5 to 11.8 11.5
hexylcyclohexane Fullerton (W. TX) Illinois; Midway (AR) W. TX; Jusepin (Venezuela) Cowden (W TX) Santa Fe Springs (CA) Slaughter (W. TX); Hobbs (NM) Colombian Hendrick and Yates (W. TX) Elk Basin, heavy (WY) Kettleman Hills (CA) Smackover (AR) Lagunillas (Venezuela) Gulf Coast light distillates
hexadiene
12.66 (range 12.1 to 13.65) calculated lrom factors of raw and dewaxed lube stocks
(YIELD1
*.-
IO
20
30 40 50 60 i-0 PERCENTAGEDISTILLED
80
90 I
100 I
Ftg. 21 .l-Evaluation curves of a 32.0API intermediate-base crude oil of characterization factor 11.65.
CRUDE-OIL 8 CONDENSATE
21-5
9.0
8.0
7-o
1100
6.0
1000
900
800
700
600
500
400
300
200
100 IO 20 30 40 50
Fig. 21.2-Petroleum properties as a function of API gravity and characterization factor. Note: the parameters in the curves refer to the characterization factor.
21-6
PETROLEUM ENGINEERING
HANDBOOK
III
200
000
Fig. 21.3-Specific heats of Mid-Continent liquid oils with a correction factor for other bases of oils.
7 / L o-
o-
o-
o9 3-
1,.,!,,,,,,.,
K=CHARACTEklZATION FACTOR = 3MOLAL AVG. BOILINGPOINT,R / SPEClF(CG.,ilTYf~ /
1,100 I I 000 , OF I 900 I 1,000 I 1,200
34 0
Fig. 21.4-Heat
CRUDE-OIL
& CONDENSATE
PROPERTIES
& CORRELATIONS
21-7
TABLE 21.8-TRUE-BOILING-POINT
Smackover, AR Gravity, API Sulfur, % Viscosity, SUS at lOOoF Date Characterization factor At 25OF At 450F At 550F At 750DF Average Base Loss, % Gasoline % at 300F Octane number, clear Octane number, 3 cc TEL % to 400F Octane number, clear Octane number, 3 cc TEL % to 450F Quality Jet stock % to 550F API gravity Qualitv Kerosene distillate %, 375 to 500F API gravity Smoke point Sulfur, % Quality Distillate or diesel fuel %, 400 to 700F Diesel index Pour point Sulfur, O/O Quality Cracking stock (distilled) %, 400 to 900F Octane number (thermal) API gravity Quality Cracking stock (residual) % above 550F API gravity API cracked fuel % gasoline (on stock) % gasoline (on crude oil) Lube distillate (undewaxed) % 700 to 900Fc Pour point Viscosity index Sulfur, % Quality Residue, % over 900F Asphalt quality 20.5 2.30 270 413139 11.62 11.48 11.47 11.55 11.53 I 0 6.0 73.2a a9.0a 11 .o 66.0b 14.4 good b 24.1 41.9 good 9.5 38.0 16.0b 0.29b
Santa Maria, CA 15.4 4.63 368 812154 11.90 11.42 11.29 11.11 11.48 IN 0 7.0
11.5 11.53 11.59 11.72 11.58 I 1.1 21 .6d 72.ob 31 .6d 66.7b 35.6d excellent b 46.2d 46.0d good Il.Od 37.0 17.0b 0.06b
39.2d
1.2d
56.3d 57.4
6.1 d 29.5b
29.2
43.0b Ob 0.82 b
0.8b
45.6d 70.4b 28.0 good 52.P 18.2 5.0 42.2 22.2 17.6d 56.0b 0.43b 21.7d good
75.9
42.2d
22.2d
16.0
a Simply aviation gasoline, not always 300-F cut point Esbmated from general cotrelat~ons. Sour oils (1.e.. oils containing more than 0.5 cu ft hydrogen sulfide per 100 gal before stabilization.) dApproximat.+d from data on other fractions of same oil. Research method Octane number
21-8
PETROLEUM
ENGINEERING
HANDBOOK
TABLE 21.9-ANALYSIS
Chapel Hill Palusy Zone Sampling pressure Sampling temperature Total fluid mol wt Liquid/gas ratio, bbl per million scf Gas mol WI Gas analysis, mol% Carbon dioxide Nitrogen Methane Ethane Propane i-butane n-Butane i-pentane n-Pentane Hexanes Heptane plus Total Liquid gravity, OAPI Llquld mol wt Liquid analysts Light gasoline Naphtha Kerosene dtstlllate Gas oil Nonviscous lube Residuum and loss 645 82 25.03 88.74 20.18 0.794 1.375 76.432 7.923 4.301 1.198 1.862 0.937 0.781 1.415 2.992 100.00 71.8 68.64 Vol % ---__ 55.1 37.2 21.1
Carthage Lower Pettile Zone 63267 20.19 29.28 18.25 0.646 1.967 88.799 3.363 1.536 0.335 0.583 0.302 0.254 0.574 1.641 100.00 64.8 81.55 Vol % 40.7 47.0 79 API 76.6 59.3 47.6
Old Ocean Chenault Zone 752 85 20.76 29.33 18.70 0.448 0.370 87.584 5.312 2.302 0.584 0.630 0.416 0.207 0.505 1.642 100.00 54 0 85.93 Vol % 21.2 55.3 15.0 3.8 4.7
Old Ocean Larson Zone 702 85 20.51 28.71 18.17 0.468 0.414 90.162 4.067 1.616 0.464 0.390 0.274 0.123 0.418 1.604 100 00 47.6 110.07 API 70.9 52.2 42.1 36.6 29.8
Seellgson 21 D Zone 810 80 20 64 29.88 18.42 0.130 0 075 89.498 4 555 1 909 0 465 0 493 0.286 0209 0 385 2015 100.00 52.7 94.49 Vol % 22.6 47.7 15.9 7.3 6.5
Seeligson 21 A Zone 410 85 20.63 24.48 18.69 0.200 0.253 88.731 5.224 1.795 0.488 0.452 0.172 0.241 0.414 2.032 100.00 52.1 103.22 API 68.4 53.1 43.0 37.0
Saxet 1087 88 21.34 41.33 18.89 0.299 0.281 86.733 4.816 2.873 0.836 0.788 0.583 0.256 0.633 2.102 10000 60.0 68.73 Vol % 35.7 47.6 10.0 2.4 4.3 OAPI 73.6 55.9 44.9 38.2
5.6
4.3
4.4
=APl Vol % --71.2 14.7 52.9 36.9 42.6 17.4 37.8 21.3 74 2.3
OAPI Vol % 70.1 53.4 43.8 37.4 20.7 49.5 16.1 7.2 6.5
An important physical property of petroleum necessary in studying flow characteristics is viscosity. Viscosity of petroleum is often reported in Saybolt Universal Seconds (SUS), derived from one of the common routine tests for oils. For engineering calculation, however, the viscosity should be obtained in centipoise. The relation between these two systems, according to the U .S Bureau of Standards, is 5 =0.219ts Yo --, 149.7 tsu
where FL0 = viscosity, cp Yo = specific gravity of oil at measured temperature, and tSU = Universal Saybolt viscosity, seconds. An accurate correlation for viscosity is difficult, especially for viscous oils, but an estimate of viscosity may be obtained from Fig. 21.5. Four characterization factors are given, and interpolation must be made for other factors.
True-Boiling-Point
Crude-Oil Analyses
tempera-
A number of true-boiling-point crude-oil analyses are included in Table 21.8. In addition to the gravity, viscosity, sulfur content, and characterization factor, there is a breakdown of typical products made from each crude. This table may be used either to estimate the value of the products listed or to plot and evaluate any set of products obtained (see Fig. 21.1). The table is separated first according to state, and within each group according to gravity.
CRUDE-OIL
& CONDENSATE
PROPERTIES
& CORRELATIONS
21-9
When the quality of a product is indicated as good or excellent, it means not only that the quality is good but that it is present in normal amounts and that a salable product can be made without excessive treatment. Table 21.9 shows the analysis of the gas and liquid phases after a stage separation of several condensates. Nelson4 gives a compilation of 164 crudes and lists the gravity, characterization factor, sulfur content, and viscosity of each. Those tables include yields of typical refined products, along with their physical properties and an indication of their quality. A true-boiling-point curve can be generated by plotting the end points of these products against the cumulative volume percent yield. If the characterization factor is plotted on the same graph, the characterization factor at any instantaneous boiling point can be calculated. When instantaneous temperatures and characterization factors at different percents are known, specific gravity, API gravity, and viscosity curves may be estimated. Thus, evaluation curves such as those in Fig. 21 .l may be produced for any of the 164 crudes listed. A typical page of these data is shown in Table 21.8. More recently, a series on evaluations of non-U.S. crude oils was published. 5 The format is similar to those in Nelsons compilation, 4 but the physical properties are usually more complete. An example of an analysis from this series is shown in Table 21.10. The USBM in Bartlesville, OK, began making distillation analyses before 1920. This laboratory [U.S. DOE Bartlesville Energy Technology Center (BETC)] has continued to evaluate crude oil up to the present time and has two publications6,7 that show the distillation data along with gravity and viscosity of the distilled fractions. They also show the percentage composition of the fractions in terms of paraffins, naphthenes, and aromatics. This set of tables uses the correlation index rather than characterization factor as a correlating number. In general, low correlation index (1,) numbers indicate highly paraffinic (pure paraffin hydrocarbons, I, =O). High numbers indicate a high degree of aromaticity (benzene, I,. = 100). The correlation index is defined as follows.
1,=413.7 y-456.8+87552/T~,
TABLE Pl.lO-TYPICAL CRUDE OIL EVALUATION, EKOFISK, NORWAY Crude Gravity, API Basic sediment and water, vol% Sulfur, wt% Pour test, OC Viscosity, SUS at lOOoF Reid vapor pressure, psi at 1OOF Salt, lbm/l,OOO bbl Nitrogen compounds and lighter, ~01% Gasoline Range, OF Yield, VOWI Gravity, OAPI Sulfur, wt% Research octane number, clear Research octane number, - 3 mL tetraethyl lead per gallon Gasoline Range, OF Yield, ~01% Gravity, OAPI Paraffins, ~01% Naphthenes. vol% Aromatics, ~01% (0 + A) Sulfur, wt% Research octane number, clear Research octane number, + 3 mL tetraethyl lead per gallon Kerosene Range, OF Yield, ~01% Gravity, OAPI Viscosity, SUS at lOOoF Freezing point, OF Aromatics, VOW (0 + A) Sulfur, wt% Aniline point, OF Smoke point, mm Liaht Gas Oil Range, OF Yield, ~01% Gravity, OAPI Viscosity, SUS at lOOoF Pour point, OF Sulfur, wt% Aniline point, OF Carbon residue, Ramsbottom, wt% Cetane index TopPed Crude Range, OF Yield, ~01% Gravity, OAPI Viscosity, SUS at 122OF Pour point, OF Sulfur, wt% Carbon residue, Ramsbottom, wt% Nickel, vanadium, ppm 650 + 38.8 21.5 80.25 -85 0.39 4.0 5.04, 1.95 500 400 to 500 13.5 40.2 32.33 -38 13.1 <0.05 146.2 21 60 to 400 31.0 60.1 56.52 29.52 13.96 0.0024 52.0 76.0 60 36.3 1 .o 0.21 +20 42.40 5.1 14.5 1.0
to 200
10.7 77.2 0.003 74.4 90.0
to 650
15.7 33.7 43.83 -25 0.11 164.3 0.08 56.5
where y is the specific gravity of the fraction at 60F and T, is the average normal boiling point in degrees Rankine . All U.S. DOE analysis data have been built into the BETC Crude Oil Analysis Data Bank.8 The data retrieval system, Crude Oil Analysis System (COASYS), is available by telephone hookup, and customers may search, sort, and retrieve analyses from the file. More than 30 keywords are available for searching; for example, YEAR, APIG, LOC, GEOL and SULF, allow a search on year analyzed, API gravity, location by state and country, geological formation, and percent sulfur in the oil, respectively. Table 21.11 shows the type of information obtained in a typical analysis retrieved from a computer search by COASYS.
21-10
PETROLEUM ENGINEERING
HANDBOOK
TABLE 21.11-ADAPTATION
Crude Petroleum Analysis: BETC Sample-B75008 lndentification Webb W Field, Grant County, OK Red Fork, Des Moines, Middle Pennsylvanian-4,464 General Characteristics Gravity, specific [OAPI] Sulfur, wt% Viscosity, SUS at 77OF at 1 OOF Pour point, OF Nitrogen, wt% Color Distillation. USBM Method (First droo at 79OF) Stage l-Distillation Fraction Cut Number (OF) Vol% -1221
2 3 4 5 6 7 8 9 10 167 212 257 302 347 392 437 482 527 0.820[41.1] 0.24 42
to 4,482 ft
39
<5 0.054
brownish-black
Gravity at 6OOF
Specific API
Correlation
Index 7 17 21 22 23 22 23 24 25
~~-
(OF)
-1.5 1.5
2.2 5.5 7.4 5.8 6.7 6.0 59 6.8 5.1
0.639
0.670 0.712 0.738 0.757 0.773 0.785 0 798 0.812 0.823
89.9 79.7
67.2 60.2 55.4 51.6 48.8 45.8 42.8 40.4
Stage 2-Distillation
11 12 13 14 15 392 437 482 527 572 7.2 6.2 5.6 4.8 5.1
continued at 40 mm Hg
60.1 66.3 72.1 76.9 82.0 99.0 0.842 0.851 0.863 0.874 0.887 0.934 38.6 34.8 32.5 30.4 28.0 20.0 7.1 0.67 0.235 1.4 30 30 33 35 38 1.46481 1.47017 1.47736 141.2 148.4 145.6 40 i: 96 179 14 ii 76 98
17.0
Approximate Summary
9.2 35.1
17.8
0.690
0.743 0.311 0.845 0.854 to 0.875
11.6 10.3
6.0 1.3 17.0 1.0
73.6 58.9 43.1 35.9 34.3 to 30.3 5oto 100 loot0 200 >200
0.875 to 0.890 30.3 to 27.4 0.890 to 0.894 27.4 to 26.8 0.934 20.0
vapor, or two-phase state. With crude oils, the fluid may be subcooled liquid, but with some dissolved gas. Upon reduction in reservoir pressure, a point where the gas starts to come out of solution, called bubblepoint pressure, is reached. At this point the flow characteristics change. Some of the earliest work in this field was done by Lacey , Sage, and Kircher. 9 Several empirical correlations have been developed to predict the bubblepoint pressure, and some of these arc presented later.
Dewpoint-Pressure Correlations
The dewpoint, like the bubblepoint, is characterized by a substantial amount of one phase in equilibrium with an infinitesimal amount of the other phase. At the dew-
point, the liquid phase is at its minimum. In general, petroleum-reservoir systems that involve dewpoint behavior at reservoir conditions are characterized by (1) surface gas/oil ratios (GORs) greater than 6,000 cu ft/bbl in most instances; (2) lightly colored tank oils, usually straw-colored to light orange for reservoir systems at 3,000 to 5,000 psi but grading to brown for systems at 7,C00 psi and greater; (3) tank oil gravity usually greater than 45API; and (4) methane content usually greater than 65 mol% . Few dewpoint-pressure correlations of reservoir systems have been published. Sage and Olds published a very general correlation of the behavior of several San Joaquin Valley, CA, systems. A correlation developed by Organick and Golding I1 is discussed in
21-11
TABLE 21.11-ADAPTATION
(continued)
Hvdrocarbon-tvpe Analvsis for Crude Petroleum Analvsis 875008 Fraction Number 1 2 3 4 5 a 7 a 9 10 11 12 vow0 of Crude 1.5 2.2 5.5 7.4 5.8 6.7 6.0 5.9 6.8 5.1 72 6.2 Specific Gravity 0.639 0.670 0.712 0.738 0.757 0.773 0.785 0.798 0.812 0.823 0.842 0.851 Correlation Index Aromatics (VOW0 of Fraction) 0.0 2.4 5.9 7.5 9.1 10.2 10.6 10.7 11.5 10.4 13.0 16.2 P-N * (vol% of Fraction) 100.0 97.6 94.1 92.5 90.9 89.8 89.4 89.3 88.5 89.6 87.0 83.8 Correlation index of P-N Gravity of P-N 0.639 0.665 0.702 0.727 0.746 0.761 0.772 0.784 0.796 0.809 0.825 0.831
7 17 21 22 23 22 23 24 25 30 30
5 13 16 17 17 16 16 17 18 22 21
Analvsis of Naotha Fractions Fraction Number 2 3 4 5 6 7 Vol% of P-N ANaphtha 7.1 23.7 38.6 44.0 43.6 43.6 Paraffin 92.9 76.3 61.4 56.0 56.4 56.4 Vol% of Fraction Naphtha 6.9 22.3 35.7 40.0 39.2 39.0 Paraffin 90.7 71.8 56.8 50.9 50.7 50.4 Fraction Number 12 14 Number of Total 1.4 1.7 Aromatic Rings 0.3 0.6 Naphthenes per mol 1.1 1.1
Summarv Data for Blends Naphtha Blend (Fractions 1 through 7) VOW of Crude in Blend Aromatic, VOW of Blend Paraffin-Naphthene, vol% of Blend Naphthene, ~01% of Blend Paraffin, ~01% of Blend Naphthene, ~01% of P-N in Blend Paraffin, vol% of P-N in Blend Naphthene Ring, wt% of P-N in Blend Paraffin + Side Chains, ~1% of P-N in Blend
Parafbn-Naphtha
28.3 71.7
detail. Calculation of the dewpoint pressure by means of the composition and equilibrium ratios is discussed in Chap. 23.
primarily conditions that pertain to dewpoints, and it is in this capacity that they will be discussed. The reader should be aware, however, that the charts also may be used to estimate critical pressure and temperature of the more volatile systems. The correlation has limited usefulness as a bubblepoint-pressure correlation because it covers primarily high-volatility systems. system. The short-cut method suffices for most calculations. Calculation of Ts. The molal average boiling point of the system is defined as TB=CyxTa, . . . . . . . . . . . . . . . . . . . . . . . . . . ...(l)
where y is the mole fraction and T, the atmospheric boiling point. Boiling points of the pure compounds (methane, ethane, nitrogen, carbon dioxide, etc.) are listed in Chap. 20. The boiling point of the CT + fraction is taken as the Smith and Watson I2 mean average boiling point (MABP). The MABP can be calculated from the ASTM distillation curve, the procedure being first to calculate the ASTM volumetric average boiling point (VABP, F) and then to apply a correction factor to obtain the
21-12
PETROLEUM ENGINEERING
HANDBOOK
TABLE Pl.lZ--RELATION OF DEWPOINT PRESSURE OF CALIFORNIA CONDENSATE SYSTEMS Tank-Oil Gravity (API) lOOoF 52 54 56 58 60 62 64 16OOF 52 54 56 58 60 62 64 220F 54 56 58 60 62 64 4,440 4,190 3,970 3,720 3,460 3,290 3,080 4,760 4,400 4,090 3,840 3,610 3,390 3,190 4,410 3,990 3,700 3,430 3,150 2,900 4,140 3,920 3,730 3,540 3,340 3,190 3,010 4,530 4,170 3,890 3,650 3,430 3,240 3,060 4,230 3,780 3,480 3,210 2,970 2,740 3,000 3,710 3,540 3,380 3,220 3,070 2,920 4,270 3,950 3,690 3,470 3,280 3,100 2,930 4,050 3,600 3,280 3,030 2,800 2,590 3,680 3,540 3,390 3.250 3,100 2,970 2,840 4,060 3,760 3,520 3,320 3,150 2,990 2,820 3,890 3,440 3,110 2,880 2,670 2,470 3,530 3,410 3.280 3,140 3,010 2,880 2,770 3,890 3,610 3,380 3,200 3,040 2,090 2,740 3,750 3,300 2,970 2,760 2,570 2,380 3,420 3,310 3,180 3,060 2,930 2,800 2,700 3,650 3,490 3,270 3,110 2,960 2,810 2,670 3,620 3,180 2,850 2,660 2,480 2,300 GOR (cu ft/bbl) 15,000 20,000 25,000 35,000 30,000 -~ 40,000
9 2
hi 4500 5
z E ; z x i x GAS-OIL RATIO, CU FT/EEiL 3000 3500 4000
Fig. 21.6-Influence of gas/oil ratio and tank-oil gravity on retrograde dewpoint pressure at 160F.
TABLE 21.13-CORRECTION TO ADD TO ASTM VOLUMETRIC AVERAGE BOILING POINT TO OBTAIN MEAN AVERAGE BOILING POINT Slope of ASTM Curve (OF/%) 10 to 90% points 2.0 2.5 3.0 3.5 4.0 4.5 ASTM VABP (OF) 200 -13-17 -22 -27 -33 300 -11.5 - 15.5 -20 -25 - 30.5 400 - 10.5 - 14 - 18.5 -23 - 28.5 - 34.5 500 -9.5 _._ -13 -17 -21.5 -26.5 -32.5
TABLE 21.14-VALUES OF EQUIVALENT MOLECULAR WEIGHTS FOR NATURALGAS CONSTITUENTS Methane Ethane Propane i-butane n-Butane i-pentane n-Pentane Hexanes Ethylene Nitrogen Carbon dioxide Hydrogen sulfide 16.0 30.1 44.1 54.5 58.1 69.0 72.2 85 26.2 28.0 44.0 34.1
TABLE 21.15-CORRECTION TO ADD TO ASTM VOLUMETRIC AVERAGE BOILING POINT TO OBTAIN CUBIC AVERAGE BOILING POINT Slope of ASTM Curve (oF/%) 10 to 906/o points 2.0 2.5 3.0 3.5 4.0 4.5 ASTM VABP (OF) 200 - 5.0 - 6.5 -8.0 - 10.0 - 12.5 - 15.0 400 -4.0 - 5.5 -7.0 -8.5 -10.0 -12.5 600 -3.5 -4.5 - 5.5 - 7.0 - 8.5
10.0
MABP. The VABP is the average of the temperatures at which the distillate plus loss equals 10, 30, 50, 70, and 90% by volume of the ASTM charge, that is,
y, = TlOW + T30% + Tsox + T70% + T90% 5 ) . . . . (2)
where TI/ is the ASTM volumetric average boiling point. The correction to add to TV to obtain the mean average boiling point is given in Table 2 1.13 as a function of TV and the slope of the ASTM curve between the 10 and 90% distilled points. In the correlation, r, is in degrees Rankine (i.e., F+460). Calculation of W,. The modified weight average equivalent molecular weight, W,, is a more complex function to evaluate. It is defined as the equivalent molecular weight multiplied by the summation of the weight fractions. The equivalent molecular weight of a paraffin hydrocarbon compound is its true molecular
weight. For other than straight-chain paraffin compounds (isoparaffns and olefins), the equivalent molecular weight is defined as the molecular weight that an n-paraffin would have if it boiled at the same temperature as the isopamftin or olefin in question. Values of the equivalent molecular weights for naturalgas constituents are given in Table 21.14. The equivalent molecular weight of the C 7 + fraction is determined by calculating the Watson characterization factor, Kw , and using Fig. 21.7. Use of the characterization factor permits some account to be taken of the paraffinicity of the heavy-end material.
Kw=
........................(3)
where Tc is the cubic average boiling point, R. The cubic average boiling point (Fc) is obtained by adding the corrections in Table 21.15 to the ASTM TV, F.
21-13
SLOPE
OF
ASTM lo%-90%,
DISTILLATION OF/%
CURVE TEMPERATURE.F
at W, =80.
POOC
TEMPERATURE, TEMPERATURE.F
at W,
= 100.
at W, = 70.
v 3
E a
6 ;: 2 2 TEMPERATURE, F
5000\
400030002000-
TEMPERATURE,F
at W, = 90.
at W, =60.
21-14
PETROLEUM ENGINEERING
HANDBOOK
8000 $ g 4 g a 5 F i 3 & m 7000 6000 5000 4000 3000 2000 0 100 TEMPERATURE.F 200 300 TEMPERATURE.F
at W, = 55.
at W,
=40.
L O-50 TEMPERATURE.F
I 0
I 100
I 200 OF
I 300
TEMPERATURE,
at W,
= 50
at W,
=35.
I
O-50 TEMPERATURE.F
I
0 100 TEMPERATURE,F 200 10 C
at W, = 45.
21-15
TEMPERATURE,
OF
TEMPERATURE,
at W, =30.
Example Problem 1. The dewpoint pressure at 200F for a well effluent having the composition shown in Table 21.16 is predicted as follows. 1. Calculating first the properties of the separator liquid CT+, we have TV= and 497 -232 lo-90% slope= 80 =3.31. 232+260+313+383+497 5 =337F
nw 2 4000
9 g 6 c 2 2 :: 1000 0 130 TEMPERATURE,OF 200 300 3000 2000
at W, =25.
From Table 21.13, MABP is 337-22.5=315F or 775R. From Table 21.15, CABP is 337-8.3=329F or 789R, giving
Kw=
3vTiG
0.7535
= 12.3.
Accuracy of Organick-Golding Correlation. About 50% of the 2 14 points that form the basis for the correlation were in error less than 5 % and 82 % were in error less than 10%. Standard deviation of all points is about 7.0%.
From Fig. 21.7 the W, for the CT + material from the separator liquid is estimated to be 142. Properties of the C 7 + material from the separator gas are assumed to be equal to those of n-octane (i.e., Tg=718R, W, = 114). 2. Calculating values of TB and W, for the well effluent, we obtain the results shown in Table 2 I. 17. 3. Having calculated Ts and I@, for the well effluent, we can now determine the desired dewpoint pressure at 200F by interpolation between Figs. 21.14 and 21.15. At TB =240F, the dewpoint pressure is w, =50 w, =45 4,850 4,ooO
and the calculated dew point (at W, =49)is 4,680 psia. It will be noticed that at 4,680 psia and 200F the material is about 200F and 900 psi above the critical temperature and pressure of the system. (From Figs. 21.14 and 2 1.15, the locus of critical states line gives Tc=OF and pc=3,800 psia.)
21-16
PETROLEUM ENGINEERING
HANDBOOK
TABLE 21.16-WELL
heavier components. These values can be obtained by the chart shown in Fig. 21.22. The following example illustrates the calculation of M, and v, . Example Problem 2. The specific volume of a gascondensate system at reservoir conditions given the system molal analysis shown in Table 21.18 is calculated as follows, assuming 1 pound mole of system. 460 + 199 Tpr = 370.7 =1.78,
Component co*
N2
Separator Gas 0.0060 0.0217 0.8986 0.0461 0.0131 0.0043 0.0043 0.0019 0.0017 0.0019 0.0004
-
Separator Liquid 0.0988 0.0350 0.0201 0.0382 0.0495 0.0313 0.1284 0.5606 1 .oOOo
Effluent 0.0056 0.0204 0.8498 0.0454 0.0146 0.0053 0.0064 0.0048 0.0035 0.0096 0.0004 0.0342 1 .oooo
0.0381
2,500 -=3.75, Ppr= 666.0 z=O.885 (from Fig. 20.2) and at 2,500 psia and 199F
1 .oooo
Properties of C, + *separator gas C, + mOfeclar we,gtlt= 114 Separator liquid C, + Molecular weight = 139 Density= 0.7535 g/cc=56.3API ASTM distillation BP (%) 21WF 10 232 20 245 30 260 40 269 50 313 60 349 70 363 60 416 90 497 95 Endpoint tEffluen1 composition calculated on the basis of separator liquid/gas ratio 3.0 gal/lo3 cu H.
=O. 129,
system volume. Usually, at reservoir pressures and temperatures and for systems whose composition can be expressed as having a surface GOR greater than 10,000 cu ft/bbl, the presence of 10 ~01% liquid phase will not cause errors greater than 2 or 3% when the two-phase mixture density is calculated as though the mixture existed in only a single phase. This comes about because the partial volumes of components in the liquid phase are substantially the same as the partial volumes of the same components in the vapor phase. Calculations from Composition of the Condensate System. As outlined previously, the formation volume (total or single phase) can be calculated from the relation
B=Mm vro L M,,v,, .,,.,..................~ . . .
where Tpr is the pseudoreduced temperature, ppr the pseudoreduced pressure, z the compressibility factor, and v, the specific volume (cu ftilbm) at reservoir conditions. In the above solution, two phases ale present at 2,500 psia, as the dewpoint pressure calculated by the method of Organick and Golding is 2,690 psia at 199F. Probably no correlation will indicate directly the amount of liquid present at pressures less than the dewpoint pressure, although it can be calculated by use of suitable equilibrium-ratio and density data. Calculations from GOR and Produced Fluid Properties. A second method of calculating specific volume or formation volume on the basis of the gas-law equation was developed by Standing. I4 This method uses a correlation (Fig. 21.23) to obtain the gravity of the well effluent (or reservoir system) from the condensate liquid/gas ratio, gas gravity, and the stock-tank-oil gravity of the surface products. The effluent gravity is then used to obtain values of pseudocritical temperatures and pressures and, by means of these, to evaluate compressibility factors for the entire effluent. The condensate curve of Fig. 2 1.24 should be used when employing this method. Example Problem 3. The total formation volume of a gas-condensate system at reservoir conditions given the parameters in Table 21.19 is calculated as follows, assuming 1 bbl of stock-tank condensate. 3,700x0.65+170x1.20 Yg 3,700+170
where
=0.675
L can be calculated by use of equilibrium ratios and the methods outlined in Chap. 23. To use the pseudoreduced-temperatuatureipseudoreduced-pressure/compressibility chart in the calculation of vrO, it is necessary to determine suitable pseudocritical temperature and pressure values for the heptanes and
and 1 bbl condensate per million cubic feet is 325 3.70+0.17 where yp is the gravity of total surface gas. =84,
21-17
TABLE 21 .17-CALCULATED
VALUES OF 7, and W, Weight Fraction Times Equivalent Molecular Weight 0.47 0.68 9.33 1.76 1.21 0.72 0.92 1.03 0.77 3.03 0.22 28.90 w, = 49.04
Component
co2 F2 c: C3
Fraction 0.0056 0.0204 0.8498 0.0454 0.0146 0.0053 0.0084 0.0048 0.0035 0.0096 0.0004 0.0342 1 .oooo
Boiling Point (W 350 139 201 332 416 471 491 542 557 600 718
Fraction Times Boiling Point (W 2.0 2.8 170.8 15.1 6.1 2.5 3.1 2.6 1.9 5.8 0.3 26.9 7s = 239.9
Fraction 0.0107 0.0244 0.5831 0.0586 0.0274 0.0133 0.0158 0.0150 0.0107 0.0356 0.0019 0.2035 1.0009
Equivalent Molecular Weight 44 28 16.0 30.1 44.1 54.5 58.1 69.0 72.2 85 114 142
TABLE 21.18-CALCULATION
OF SPECIFIC VOLUME OF GAS-CONDENSATE Critical Temperature of Components, 7, vu 548 227 344 550 666 733 766 830 847 915 1,090.
SYSTEM
Component
Mole Fraction, y 0.0059 0.0218 0.8860 0.0460 0.0134 0.0045 0.0048 0.0026 0.0021 0.0037 0.0084
Molecular Weight, M 44.0 28.0 16.0 30.1 44.1 58.1 58.1 72.1 72.1 86.2 138
Weight, ybf Ubm) 0.26 0.61 14.18 1.39 0.59 0.26 0.28 0.19 0.15 0.32 1.16 19.39
yT, (OR) 3.2 4.9 304.8 25.3 8.9 3.3 3.7 2.2 1.8 3.4 9.2 370.7
Critical Pressure of Components, pc (wia) 1,072 492 673 709 618 530 551 482 485 434 343
YPC 6.3 10.7 596.3 32.8 8.3 2.4 2.6 1.3 1 .o 1.6 2.9 666.0
co2 N2 Cl c,
i-Cd n-C, i-C 5 n-C, C6 c,+
C,
1.5
GAS
060 GR
1.4
ri $
F d
Id0
120
220
240 1
BOOM.&+--+
60&O
-j
hw 1
iz k
lo3 120
220
J
240 CFB 20 40 60 El0 ICC
and pressures
for
Fig. 21.23-Effect of condensate volume on the ratio of surface-gas gravity to well-fluid gravity.
21-18
PETROLEUM ENGINEERING
HANDBOOK
TABLE 21 .l g--DATA FOR CALCULATING TOTAL FORMATION VOLUME OF A GAS-CONDENSATE SYSTEM Reservoir pressure, psia Reservoir temperature, OF Stock-tank-condensate production, B/D Stock-tank condensate gravity, OAPl Tank vapor rate, IO3 cu ft/D Tank vapor gravity (air = 1) Trap gas rate, lo3 cu ft/D Trap gas gravity, (air = 1 .O)
0as1s 1 bbl of stock-tank condensate
TABLE 21.21-DATA FOR CALCULATING TOTAL FORMATION VOLUME OF THE GAS-CONDENSATE SYSTEM DESCRIBED IN EXAMPLE PROBLEM 4 Reservoir pressure, psia Reservoir temperature, OF GOR (condensate total), cu ft/bbl Gas gravity (total) Tank-oil gravity, OAPI 3,000 250 11,900 0.675 45
TABLE 21.20-DATA FOR CORRELATION FOR OBTAINING TOTAL FORMATION VOLUME FACTORS OF DISSOLVED GAS AND GAS-CONDENSATE SYSTEMS SHOWN IN FIG. 21.25 Pressure, psia GOR, cu ft/bbl Temperature, OF Gas gravity Tank-oil gravity, OAPI 400 to 5,000 75 to 37,000 100 to 258 0.59 to 0.95 16.5 to 63.8
TABLE 21.22-DATA USED TO CALCULATE TOTAL FORMATION VOLUME FACTOR IN EXAMPLE PROBLEM 5 Reservoir pressure, psia Reservoir temperature, OF GOR, cu ftlbbl Separator Tank Total Gas gravity Tank-oil gravity, OAPI 1,329 145 566 37 603 0.674 36.4
From Fig. 21.23, at 45API ~&~~=1.367 and yl,,=1.367x0.675=0.923, where Ylw = well fluid gravity, ysr = trap gas gravity, and Ylwr = well fluid reservoir gravity. From Fig. 21.24,
From Fig. 21.23 the molecular weight of stock-tank condensate, M, , is 140, moles of stock-tank condensate per barrel is 281/140=2.00, moles of surface gas per barrel of stock-tank condensate is l/325 x (3,870x 103) x l/379=31.4, and total moles per barrel of stock-tank condensate is 2.00+31.4=33.4. From gas law,
y=-=
n*T P
33.4~0.845~ 10.73~710
=71.7 3,000
Tpc =432
and
12.8,
ppc=647.
At reservoir conditions of 3,000 psia and 250F, 460+250 Tpr = 432 PPr and from Fig. 20.2 z=O.845. By using 350 lbm/bbl for water, the weight of stocktank condensate per barrel is 350x 141.5 131.5+API =281. 3,000 =------4.64, 647 =1.64,
where the first value of V is in cubic feet and the second in barrels, giving a formation volume B, of 12.8 bbllbbl of stock-tank condensate. Total Formation Volume Factors of Dissolved Gas Systems A suitable correlation for obtaining total formation volume factors of both dissolved gas and gas-condensate systems was developed by Standing. t5 This correlation is shown in Fig. 21.25, and the graphical chart for simplified use of the correlation is given by Fig. 2 1.26. The correlation contains 387 experimental points, 92% of which are within 5% of the correlation. Range of the data comprising the correlation is given in Table 2 1.20. Example Problem 4. The total formation volume of the gas-condensate system described in Example Problem 3 is calculated as follows, given the data in Table 2 1.2 1.
21-19
675 a 4 650
F. 475 E
: 2 f 450 425
: - 400 2 z 8 2 I a 375 350 325 3oo 060 I I 080 ~ ~ 100 j I / 120 I 1 I / 140 I 1 I I 160 I I 160
Fig. 21.25-Formation volume of gas plus liquid phases from GOR, total gas gravity, tank-oil gravity, temperature, and pressure.
volume by
21-20
PETROLEUM
ENGINEERING
HANDBOOK
Tc = Tm = Tm =
TV = vro = -vst = w, = Y=
and
=11,90()
(250)o5
x~~~~~~~~.~X~-~ooo27x~wo
z = Ye =
ygt =
Ylw Ylwr Yo p
= = = =
cubic average boiling point, K (R) mean average boiling point, K (R) mean average boiling point, K (R) volumetric average boiling point, F specific volume of reservoir system specific volume of stock-tank oil modified weight average equivalent molecular weight mole fraction compressibility factor gas specific gravity trap gas gravity well fluid gravity well fluid reservoir gravity tank-oil specific gravity viscosity, Pa. s (cp)
where y0 is the tank-oil specific gravity. From Fig. 21.25, B,=13+bbl/bbl oftank oil. From Fig. 21.26, B, = 13.7 bbl/bbl of tank oil. Example Problem 5. The total formation volume of well production at reservoir conditions given the data in Table 21.22 is calculated as follows. From Fig. 21.26, B,= 1.72 bblibbl of tank oil. Experimental value calculated from PVT test results is 1.745 bbl/bbl of tank oil.
References
1. ASTM Standards on Petroleum Products and Lubricants, Part 24, ASTM, Philadelphia (1975) 796. 2. Watson, K.M., Nelson, E.F., and Murphy, G.B.: Charactetization of Petroleum Factions, Ind. and Eng. Chem. (Dec. 1935) 1460-64. 3. Technical Data Book-Petroleum Refining, API, Washington, D.C. (1970) 2-11. 4. Nelson, W.L.: Petroleum Refinery Engineering, fourth edition, McGraw-Hill Book Co. Inc., New York City (19X3), 910-37. 5. A Guide to World Export Crudes, Oil and Gas J, (1976). 6. Ferrem, E.P. and Nichols, D.T.: Analyses of 169 Crude Oils fmm 122 Foreign Oil Fields, U.S. Dept. of the Interior, Bureau of Mines, Bartlesville, OK (1972). 7. Coleman, H.J. et a[.: Analyses of 800 Crude Oils from United States Oil Fields, U.S. DOE, Bartlesville, OK (1978). 8. Woodward, P.J.: Crude Oil Analysis Data Bank, Bartlesville Energy Technology Center, U.S. DOE, Bartlesville, OK (Oct. 1980) 1-29. 9. Lacey, W.N., Sage, B.H., and Kircher, C.E. Jr.: Phase Equilibrja in Hydrocarbon Systems III, Solubility of a Dry Natural Gas in Crude Oil, Ind. and Eng. Chem. (June 1934) 652-54. Gas 10. Sage, B.H. andOlds, R.H.: VolumetricBehaviorofOiland from Several San Joaquin Valley Fields, Trans., AIME (1947) 170, 156-62. 11. Organick, E.I. and Golding, B.H.: Prediction of Saturation Pressures for Condensate-gas and Volatile-oil Mixtures, Trans., AIME (1952), 195, 135-48. 12. Smith, R.L. and Watson, K.M.: Boiling Points and Critical Pmperties of Hydrocarbon Mixtures. Ind. and Enn. Chem. (19j7) 1408. 13. Vink, D.J. er al.: Multiple-phase Hydrocarbon Systems, Oil and Gas J. (Nov. 1940) 34-38. 14. Standing, M.B.: Volumetric and Phase Behavior of Oil Field Hydrocarbon Systems, Reinhold Publishing Corp., New York City (1952). 15. Standing, M.B.: A Pressure-Volume-Temperature Correlation for Mixtures of California Oils and Gases, Drill. and Prod. Prac., API (1947), 275.
Nomenclature
B= formation volume, m3 (bbl) I, = correlation index K= characterization factor L, = moles of stock-tank condensate per barrel L, = moles of stock-tank oil per 1 mole of reservoir system, kmol/m3 (lbm moligal) M= molecular weight Mm = molecular weight of reservoir system Mst = molecular weight of stock-tank oil n= total moles PC = critical pressure, psia (lbflsq in.) Ppr = pseudoreduced pressure R= universal gas constant tsu = Universal Saybolt viscosity, seconds T= temperature, F T, = critical temperature, C (F) Tpr = pseudoreduced temperature
Ta = atmospheric boiling point, K (R) TB = molal average boiling point, K (R)
Chapter 22
Introduction
Knowledge of petroleum fluids physical properties is required by petroleum engineers for both reservoir and production system calculations. These properties must be evaluated at reservoir temperature and various pressures for reservoir performance studies, and at conditions of both changing pressure and temperature for wellbore hydraulics calculations. If reservoir fluid samples are available, the fluid properties of interest can be measured with a pressure-volumetemperature (PVT) analysis. However, these analyses usually are conducted at reservoir temperature only and the variation of the properties with temperature is not available for production system calculations. Also, in many cases a PVT analysis may not be available early in the life of the reservoir or may never be available because of economic reasons. To overcome these obstacles, empirical correlations have been developed for predicting various fluid physical properties from limited data. The development and application of several of these empirical correlations are presented in this chapter. Methods for estimating physical properties for both saturated and undersaturated oils as functions of pressure, temperature, stock-tank oil gravity, and separator gas gravity are given. Fluid properties are calculated here only for oil systems with and without fluid composition known. Methods for calculating physical properties of gas-condensate systems are presented in Chaps. 2 I, 23, and 30. Therefore, no correlations for dewpoint pressure are presented, as the dewpoint pressure can be calculated with the procedures outlined in Chap. 2 1 if the composition of the fluid is known. Many of the older correlations were presented in graphical form only and are therefore not suitable for use in computers or programmable calculators. These graphs are converted to equation form where possible. The generally accepted definitions of the fluid properties correlated in this chapter are as follows.* Oil density, p,, , is the ratio of the mass of the oil plus its dissolved or solution gas per unit volume, which varies with temperature and pressure. Bubblepointpressure, P/), is the pressure at which the first bubble of gas evolves as the pressure on the oil is decreased. It also is frequently called saturation pressure, as the oil will absorb no more gas below that pressure. The bubblepoint pressure varies with temperature for a particular oil system. Solution gas/oil ratio (GOR), R,, , is the amount of gas that will evolve from the oil as the pressure is reduced to atmospheric from some higher pressure. It is usually expressed in units of scf/STB. The gas is frequently referred to as dissolved gas. Oilformation volume factor (FVF), B,, , is the volume occupied by 1 STB oil plus its solution gas at some elevated pressure and temperature. It is usually expressed as bbl/STB. It is a measure of the shrinkage of the oil as it is brought to stock-tank conditions. Total FW, B,, means the volume occupied at some elevated pressure and temperature by 1 STB oil, its remaining solution gas, and the free gas (R,i -R,) that has evolved from the oil. It is also expressed as bbl/STB. Oil viscosi@, po. measures the oils resistance to flow, defined as the ratio of the shearing stress to the rate of shear induced in the oil by the stress. It is usually measured in centipoise and is required for both reservoir and piping system calculations. Inferfacial tension (IFT), co, is the force per unit length existing at the interface between two immiscible fluids. This property is not required in most reservoir calculations but is a parameter in some correlations for piping system calculations. It is usually expressed in units of dyne/cm.
General terms are deftned I the Glossary at the end of this chapter
PETROLEUM ENGINEERING
HANDBOOK
increased pressure will merely compress the liquid and increase its density. For the case of p> p,, . the oil density is calculated from p. =poh exp[c,(p--ph)], where PO poh p pb .. . .(2)
oil density at p, T, oil density at ph, T, pressure, psia, bubblepoint pressure at T, psia, and co = oil isothermal compressibility at T, psi - .
= = = =
Correlations for calculating R,T, B,, c, and ~b at various conditions are presented later. In the petroleum industry, it is common to express gravity in terms of the API gravity of the oil, or: 141.5 Yo = 131,5+YAP,,
..
where y. is oil specific gravity, and YAPI is oil gravity, API. Density From Ideal Solution PrinciplesComposition Known The principle of ideal solutions states that the volume of the total solution is the sum of the individual component volumes. The principle applies at atmospheric pressure for fluids in which the components are closely related chemically, such as petroleum. If the composition of the fluid is known, the density at standard conditions (14.7 psia and 60F) may be calculated from
of
systems
containing
where
350y,+O.O7647,R, 5.6158, PO = m; , I.. . . . . .
= mass of the ith component, = volume of the ith component, PI = density of the ith component at standard conditions, and C = number of components.
Vi
oil density, lbmicu ft, oil specific gravity, gas specific gravity, solution or dissolved gas, scf/STB, oil FVF, bbl/STB, density of water at standard conditions, lbm/STB, 0.0764 = density of air at standard conditions, lbmlscf, and 5.615 = conversion factor, cu ft/bbl.
If the pressure and temperature conditions are such that all of the available gas is in solution-i.e., the pressure is above the bubblepoint at the temperature of interest-
Once the density at standard conditions is calculated, it must be corrected for compressibility and thermal expansion if the density at other conditions is required. This can be accomplished by use of charts presented by Standing. When the ideal solution principle is applied to reservoir ,oils that contain large amounts of dissolved gas, it is obvious that the fluid cannot be brought to standard or stock-tank conditions and still remain in the liquid phase. This limitation is overcome by calculating a pseudoliquid density, the value of which depends on the mass or weight fractions of methane and ethane in the fluid. The pseudoliquid density correlation was presented by Standing and is illustrated in Fig. 22. I,
22-3
10 b: G
E5 a F 4I b3 mu =\
9
B 7 6 5
25
30
45 50 55 35 40 DENSITY AT 60F
60
65
expansion
of
The procedure for calculating oil density at any pressure and temperature when the composition is known is as follows. 1. Calculate the mass or weight of the ethane and heavier components in the mixture. 2. Calculate the density of the propane and heavier components with Eq. 4. 3. Calculate the weight or mass percent of ethane in the ethane and heavier mixture. 4. Calculate the weight percent methane in the total mixture. 5. Determine the pseudoliquid density from Fig. 22.1. 6. Correct for compressibility with Fig. 22.2 7. Correct for thermal expansion with Fig. 22.3. Example Problem 1. Using the known composition of a reservoir fluid as given in Table 22.1, calculate the den-
1. Weight of ethane plus=130.69-7.046=123.46 lbm. 2. Density of propane plus equals (weight of propane plus) divided by (volume of propane plus): 130.69-7.046-1.296 2.227 3. Weight percent ethane in ethane plus: 1.296(100) 123.46
= 1.05.
TABLE 22.1-
Component Cl C* C3 C4 C5 C6 C T&l
Mole Fraction. Y, 0.4404 0.0432 0.0405 0.0284 0.0174 0.0290 0.4011 1 .oooo
Mole Weight of Components, M, 16.0 30.1 44.1 58.1 72.2 86.2 297
Weight of Components mi =Y,M, (Ibm) 7.046 1.296 1.766 1.650 1.256 2.500 115.1 130.69
values
22-4
PETROLEUM ENGINEERING
HANDBOOK
is 27.4API and the quantities and gravities of the produced gas are given in Table 22.2.
Solution.
1. Average gas gravity, yr. =CI?iy,~iICRj, (414)(0.640)+90(0.897)+25(1.540) r, = and 141.5 I= 131.5+27.4 =0.89. 414+90+25 =. 726 . .
Fig. 22.4-Apparent
3. Mass of dissolved gas, m,, is given by 529 scf/STB 379,5 scf,mol (21.03 lbm/mol)=29.32 4. Weight percent methane in methane plus: 7.046(100) 130.69 5. From Fig. 22.1, psr=50.8 lbm/cu ft at 60F and 14.7 psia. 6. From Fig. 22.2, the correction for pressure is 0.89 lbmicu ft. Therefore, the density at 3,280 psia and 60F is 50.8+0.89=51.7 lbmicu ft. 7. From Fig. 22.3, the correction for temperature is -3.57 Ibm/cu ft. Therefore, the density at 3,280 psia and 218F is 51.7-3.57=48.1 Ibm/cu ft. 4. Mass of stock-tank oil, m,,, is given by =5.39. 350 lbm/STB(O. 89) = 3 11.50 lbm/STB. Fig. 22.4 shows that the apparent liquid density of the dissolved gas is about 24.9 lbmicu ft at 60F and 14.7 psia. This is used to calculate the volume of the dissolved gas. 5. Volume of dissolved gas, I,, is given by
lbm/STB.
mR -=
29.32 lbm/STB
= 1.178 cu ft/STB.
PK 24.9 lbmicu ft
6. Volume of stock-tank 5.615 cu ft/STB.
04 V,, is given by
Density From Ideal Solution PrinciplesComposition Unknown The procedure for estimating oil density outlined in the preceding section used charts for determing the apparent gas density, which required knowledge of the total fluid composition. Katz* extended the apparent density concept to apply to natural gases in general. This results in a method that can be used when solution GOR stock-tankoil gravity, and gas gravity are known. The fluid composition is not required. The correlation for the apparent density of the dissolved gas as a function of oil and gas gravity is shown in Fig. 22.4. The gravity of the produced gas is calculated as a volume-weighted average of the gas evolved at the separator and the stock tank. Application of Fig. 22.4 in estimating the oil density from limited data is illustrated in Example Problem 2. In this example, the fluid passed through two separators between the wellhead and the stock tank. Example Problem 2. Calculate the density and specific volume of the oil system at the bubblepoint conditions of pb =3,280 psia at T=218F. The stock-tank oil gravity 7. Pseudoliquid density, zz-m, fm,
PAL
0 - ,
- 311.50 lbm/STB+29 32 1bmiSTB - 5.615 cu ft/STB+1.178 cu ft/STB =50.17 Ibm/cu ft.
GAS
R (scf/STB) First-stage separator Second-stage separator Stock tank Total 414 90 25 529
22-5
Correction of the density for compression and thermal expansion is accomplished with Figs. 22.2 and 22.3. Fig. 22.2 shows that the pressure correction to 3,280 psia is 0.90 lbmicu ft. Therefore, p3Z80,60=50.17+0.90=51.07 lbmicu ft. correction to
characterize the tank oils other than by the API gravity. The value for gas gravity to be used is apparently the volume-weighted average of the gas from all stages of separation. The correlation should apply to other oil systems as long as the compositional makeup of the gases and crudes is similar to those used in developing the method. The equation for estimating bubblepoint pressure is
0.83 x IO!v ,
Fig, 22.3 shows that the temperature 218F is -3.63 Ibm/cu ft. Therefore, ~32~0,218=51.07-3.63=47.44
(5)
lbmicu ft where Y<?= = PJ, = R ,,b = 7,s = TR = YAPI = mole fraction gas, 0.00091(TR)-0.0125yA~I, bubblepoint pressure, psia, solution GOR at PLP~, scf/STB, gas gravity (air= 1.O), reservoir temperature, F, and stock-tank oil gravity, API.
The specific volume of the oil is defined as the reciprocal of the density. Therefore, v,, = i I = -----0.021 47.44 P 0 cu ft/lbm.
Bubblepoint-Pressure
Correlations
Reservoir performance calculations require that the reservoir bubblepoint pressure be known. This is determined from a PVT analysis of a reservoir fluid sample or calculated by flash calculation procedures if the composition of the reservoir fluid is known. However, since this information is frequently unavailable. empirical correlations for estimating P/, from limited data were developed. These correlations may be used to estimate bubblepoint or saturation pressure as a function of reservoir temperature, stock-tank oil gravity. dissolved-gas gravity, and solution GOR at initial reservoir pressure. That is,
A nomograph developed from Eq. 5 is shown in Fig. 22.5. The example bubblepoint determination shown in the nomograph is calculated with Eq. 5 in the following example.
Example Problem 3. Estimate pl, where R ,,, = 350 scfi STB. TR =200F, -yX=0.75, and ?,+#I =30APl.
Solution.
y!: =0.0009i(200)-0.0125(30)=
0.83
-0.193.
10 -0.1Y3
A value for R.,h = R,,i can be obtained from the initial solution GOR (produced) if the reservoir pressure is above !I/>, where R,,b is the solution GOR at bubblepoint pressure and R,, is the solution GOR at initial reservoir pressure. Three methods for estimating bubblepoint pressure are presented. The correlations were developed by use of experimentally measured bubblepoint pressures obtained from PVT analyses on reservoir fluid samples. Other correlations were developed for application in specific reservoirs, but the methods presented here gave good results over a wide range of oil systems.
Lasater Correlation A correlation by Lasater4 was developed in 1958 from 158 experimental data points, which included the ranges of variables shown in Table 22.4. The correlation was presented graphically in the form of two charts. Equations were fitted to these graphical correlations to enhance the use of this method with computers or calculators. The graphical correlations are shown in Figs. 22.6 and 22.7. The following procedure is used to estimate ph using Figs. 22.6 and 22.7.
Standing Correlations Standing presented an equation and nomograph to estimate bubblepoint pressures greater than 1,000 psia. The correlation was based on 105 experimentally determined bubblepoint pressures of California oil systems. The average error of the correlation when applied to the data used to develop the method was 4.8% and 106 psi. The ranges of data used to develop the method are given in Table 22.3 The gases evolved from the systems used to develop the correlation contained essentially no nitrogen or hydrogen sulfide. Some of the gases contained CO,. but in quantities less than 1 mol%. No attempt was made to
prb. via R, F
R sb, scf/STB
22-6
PETROLEUM ENGINEERING
HANDBOOK
BUBBLE-POINT
PRESSURE,
Fig. 22.5-Chart
for calculating
bubblepotnt
pressure
by Standings correlation.
1. Find the effective molecular weight of the stock-tank oil from the API gravity using Fig. 22.6. 2. Calculate the mol fraction of gas in the system from
R,yh1379.3 ?I: = R,h1379.3+350y,,lM, ' '.""""..'
TABLE
22.4-VARIABLE
RANGES
(6)
R sb, scf/STB
3. Find the bubblepoint pressure factor, phyRITR, Fig. 22.7 4. Calculate the bubblepoint pressure ph = [(ph~~)lT] TR/~,~ where TR is in R. The following equations can be used to replace Figs. 22.6 and 22.7.
from
-=8.26y;.56+1.95.
...
.(lO)
A nomograph that combines Figs. 22.6 and 22.7 is presented in Fig. 22.8. The example given in Fig. 22.8 is worked with the equations in the following example.
For API > 40: hi,, =73,1 10 (-yAPI) -.56. Equations for Fig. 22.7 For ys 5 0.60: .(8)
Example Problem 4. Given the following data, use the Lasater method to estimate P,,
Rch =500 scf/STB, TR=200F=660R, YAP[=30, and yg =0.876.
yI: =0.80,
Solution.
P bY,q -=0.679
TR
ex~(2.786y,~)-0.323.
.(9)
M,,=630-10(30)=330.
22-7
5.2 4.8
II II
EFFECTIVE
WEIGHT
3.6 3.2
Fig. 22.6-Effective
gravity.
molecular
related
to tank-oil
2.8
=0.587
Phh -=3.161.
TR
3.161(660)
Ph=
=2,608 psia.
0.80
1~~~~
0 0.2 0.4 GAS MOLE 0.6 0.8 FRACTION 1.0
Fig. 22.7-Lasaters
correlation of bubblepoint-pressure tor with gas-mole fraction. fac-
Vasquez and Beggs Correlation Vasquez and Beggs used results from more than 600 oil systems to develop empirical correlations for several oil properties including bubblepoint pressure. The data encompassed very wide ranges of pressure, temperature, oil gravity, and gas gravity and inciuded approximately 6,000 measured data points for R,, , B, and pLoat various pressures and temperatures. The ranges of the pertinent parameters are given in Table 22.5. It was found that the gas gravity was a strong correlating parameter and, unfortunately, usually is one of the variables of most questionable accuracy. The gravity of the evolved gas depends on the pressure and temperature of the separators, which may not be known in many cases. The gas gravity used to develop all the correlations reported by Vasquez and Beggs was that which would result from a two-stage separation. The first-stage pressure was chosen as 100 psig and the second stage was the stock tank. If the known gas gravity resulted from a first-stage separation at a pressure other than 100 psig, the corrected gas gravity to be used in the correlations can be obtained from Eq. 11. If separator conditions are unknown,
the uncorrected gas gravity may be used in the correlations for ph, R,v, B,, and c,. y,,.=y,[1.0+5.912x 10P5yAP,Ts log(p,Y/l 14.7)]
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . (11)
where ync = corrected gas gravity, 7,: = gas gravity resulting from a separation at
P .57 Ts
= separator temperature, F, p,, = separator pressure, psia, and -yAPI = oil gravity, API.
T,,
22-8
PETROLEUM ENGINEERING
HANDBOOK
Fig. 22.8-Chart
for calculating
bubblepoint
pressure
by
Lasaters correlation
The correlations are presented in equation form only. The bubblepoint pressure is calculated from
Example Problem 5. Calculate the bubblepoint pressure for the oil system given in Example Problem 4 using the Vasquez and Beggs correlation and the following data. Use the uncorrected gas gravity. R.rb =500 scf/STB, TR = 220F, yfi =0.80, and YAPI=30API. Solution. Eq. 12 and the correct C values from Table 22.1 give:
. . . . . . . . . . . . . . . . . . . . . . . . . . . (12)
500
where pb = R sb = = -,,,: = TR =
Ph=
bubblepoint pressure, psia, solution GOR at pb, scf/STB, gas gravity, oil gravity, API, and temperature, F.
0.0362(0.80)
exp[25.724(30)/680]
l-
I.0937
pb =2,562 psia. This compares well with the value of 2,608 obtained in Example Problem 4 with Lasaters correlation. With Standings Eq. 5, a value of 2,415 psia is obtained. Accuracy of Bubblepoint Correlations Comparison of the accuracy with which the measured bubblepoint pressures used in each correlation agreed with
The accuracy of the correlation was greater if the samples were divided into ranges of oil API gravity. A dividing point of 30API was chosen. The values of the constants in Eq. 12 depend on API gravity of the stocktank oil and are given in Table 22.6.
TABLE 22.8-CONSTANTS FOR BUBBLEPOINT EQUATION OAPI I 30 C, C2 C3 0.0362 1.0937 25.7240 OAPI > 30 0.0178 1.1870 23.9310
22-9
TABLE
22.7-COMPARISON
OF ACCURACY CORRELATIONS
OF BUBBLEPOINT-PRESSURE
Standing Number of points in correlations Data pointswlthin 10% of correlation, % Data points more than 200 PSI in error, % Mean.error, % 105 87 27 4.0
Vasquez-Beggs
5.008 85 - 0.7
values determined from the final correlation shows that the Vasquez and Beggs correlation is the most accurate, followed by Lasaters and then by Standings This is shown in Table 22.7.
For 7
ys =0.359 In
+0.476
(15)
For &?3
T
29:
O.l21py, 0.!81
ys =
-0.236
>
(16)
R,~=CI~,P~~
exp
T+460
(17)
The nomographs presented in Figs. 22.5 and 22.8 can be used to determine R, by entering the bubblepoint axis at the pressure of interest and proceeding backward through the graph to determine R,. Standing Correlation
where R,, = yh = p = YAP] = T = Cl, Cl. C3 = gas in solution at p and T, scf/STB. gas gravity, pressure of interest, psia, stock-tank oi] gravity, API temperature of interest. F, and are obtained from Table 22.6.
(13)
Example Problem 6. Estimate the solution GOR of the following oil system using the correlations of Standing, Lasater, and Vasquez and Beges and the data: n=765 psia, T= 137F, +A~1=22AF[ and ys ~0.65. Standing
765
R,, =0.6.5
I204 1
= 90 scf/STB.
(14) ms
T-
765(0.65) 137+460
=0.833.
where M,, is obtained from Eq. 7 or 8 and ys is calculated by either Eq. 9 or IO, depending on whether the value of the pressure factor is less than or greater than 3.29.
.~=0.359 M,,=630-
(Eq. IS).
22-10
PETROLEUM ENGINEERING
HANDBOOK
=0.922
R = 132755(0.922)(0.191) \ 410(1-0.191)
Standing Correlation. Standing used the same oil systems described in the bubblepoint correlation section to develop a correlation for B,, at pressures less than ph. The method was presented in both equation and nomograph form. In equation form, B~,=0.972+0.000147F~, . . . (18)
If yh is read from Fig. 22.7 rather than calculated from Eq. IS, a value of approximately 0.25 is obtained. This gives a value of R,, = 100 scf/STB. In this example, the graphical value is closer to those calculated from the other two correlations. The accuracy of the Lasater equation is much better at higher R,, values. Vasquez and Beggs
R, =0.0362(0.65)(765)~
exp[ 2~3~4@~]
where F is a correlating function and is determined from the equation F = R,y(y,ly,j)o.5 + 1.25T, B, = oil FVF, bbl/STB, R,, = solution GOR, scf/STB, Yn = gas gravity, YO = oil specific gravity= 141.5/(131.5 +yAPt), and T = temperature of interest, F. A nomograph that solves Eq. 18 graphically is presented in Fig. 22.9. Example Problem 7. Use both Standings equation and nomograph to estimate the oil FVF for the oil system described by the data T=2009F, R,v =350 scf/STB, Y,~= 0.75, and YAP]=30API.
Solution. Yo = 141.5/(131.5+30)=0.876. F = 350(0.75/0.876).5 + 1.25(200) =574.
R,, = 87 scf/STB.
B,, = 0.972+0.000147(574)~75; B,, = 1.228 bbl/STB. A value of 1.22 bbl/STB is obtained from the nomograph, Fig. 22.9. Vasquez and Beggs Correlation. In conjunction with development of the bubblepoint and solution GOR correlations, Vasquez and Beggs also presented an equation for oil FVF for saturated oils. To improve the accuracy of the correlation. the 600 oil systems were divided into two groups, those having API gravities 5 30 and those having gravities > 30. The gas gravity used in the equation should be the corrected gravity calculated by Eq. 11 if the separator pressure is known. If separator conditions are unknown, the uncorrected gas gravity may be used. The equation is
. . . . . . . (19)
oil FVF at p and T, bbl/STB solution GOR at p, T, scf/STB temperature of interest, F pressure of interest, psia stock-tank oil gravity, API. and gas gravity, corrected, air= I
Z-11
EXAMPLE
FORMATION
VOLUME
OF
BUBBLE-POINT
LIQUID,
Fig. 22.9-Chart
Example Problem 8. Use the Vasquez and Beggs equation to determine the oil FVF at bubblepoint pressure for the oil system described by ph ~2,652 psia, Rsb =500 scf/STB. y,,-=O.SO. YAP]=30API, and 7=220F.
Solution. B ,,h =
c2 c3
TABLE 22.8-CONSTANTS
OIL FVF OAPI 530 Cl 4.677 x 10 -4
1.751x10-5 -1.811x10-8
FOR
API>30 4.670~10-~
1.100x10-~
1.337x1o-g
1+4.677x1O-4(5OO)+1.751x1O-5(16O) ~(3010.80)-1.811x10-s(500)(160) .(30/0.80). Oil Isothermal Compressibility-Trube Method. The Trube method6 makes use of the following relationships.
B,,, = 1.285 bbl/STB. Undersaturated Systems The oil FVF decreases with pressure increase at pressures above the bubblepoint. In this case, B, is calculated from B, =Boh exp[c,,(ph -p)], where
B nh
(20)
and
= ph = p = cc1 =
oil FVF at pb, bubblepoint pressure, psia, pressure of interest, psia, and oil isothermal compressibility, psi - . where cpr- = pseudoreduced compressibility. CO = oil isothermal compressibility. P,x = pseudocritical pressure,
T,,, = pseudocritical temperature,
Values for B,,h can be calculated with the Standing or Vasquez and Beggs correlation. The oil compressibility can be determined from a PVT analysis or estimated from
empirical presented. correlations. Two correlations for c,, will be
PETROLEUM ENGINEERING
HANDBOOK
a C-J CL
,.,..
\ .i
2ooo\ I.84
Fig. 22.11-Pseudocritical
pressure variation with pseudocritical temperature and 60F specific gravity of reservoir oil; Trubes correlation.
PP = pseudoreduced pressure,
T,,,.
when used to predict the 2,000 measured values of c, from which the correlation was developed. No comparison of the two methods is available using independent data. The equation for c, is 5R,b+17.2T-1,180y,+12.61yAP,-1,433 co pxlo
The pseudoreduced compressibility is a function ofp,, and T,,,. Once cpr is obtained, c,) is calculated from
c,=cp/pP. . . . I. . . . .. . . (21)
Three graphs are required to obtain the necessary parameters to calculate cO. The pseudocritical temperature Tp,. is obtained from Fig. 22.10 as a function of the spec~fic gravity of the undersaturated liquid at the bubblepoint pressure and 60F. It also depends on the bubblepoint pressure of the oil at 60F. Values for these parameters may be estimated using the correlations for density and bubblepoint pressure presented previously. A value for ppc. is obtained from Fig. 22.11 using the liquid specific gravity at 60F and the value of T,, obtained from Fig. 22.10. Once ppi. and T,,< are known, p,,,. and T,,, at the pressure and temperature of interest can be calculated. A value of c,,,. is obtained from Fig. 22.12 using ppr and T,,,. Then c, is calculated by Eq. 2 1. Application of the Trube method using a computer or calculator would require digitization of Figs. 22.10, 22.11, and 22.12. Because of its complexity, no example will be given illustrating the application of this method. An example problem may be found in Ref. 6. Oil Isothermal Compressibility-Vasquez and Beggs Method. Vasquez and Beggs5 used approximately 2,000 experimentally measured values of c,, from more than 600 oil systems to develop a correlation for c, as a function of R ,h. T, ys, y,4pI, and p. This method is much simpler to use than the Trube method and is more accurate
oil isothermal compressibility, solution GOR, scf/STB temperature of interest, F pressure of interest, psia gas gravity, and stock-tank oil gravity, API.
psi -
Example Problem 9. Calculate the oil FVF for the oil system described in Example Problem 8 at a pressure of 3,000 psia. Use Eq. 22 to determine a value for c, where pb =2,652 psia, Rsb =500 scf/STB, y,? =0.80, YApI = 30API, T=220F, and Bob = 1.285 bbl/STB.
Solution.
co =
5(500)+17.2(200)-1,180(0.80)+12.61(30)-1,433
3.000x
105
c,,=1.43X10-5
psi-.
22-13
B, = 1.285(0.995)=
Total FVFs
When material-balance calculations are made for reservoir engineering, it is often convenient to calculate the volume at reservoir conditions that is occupied by all the material associated with a stock-tank barrel of oil-i.e., the volume of the saturated oil and the volume of the evolved or liberated gas. This can be expressed as a total FVF, B,. The total FVF can be calculated if B, and the amount of liberated gas is known. That is, B, equals volume of oil plus dissolved gas/STB plus volume of liberated gas/STB. In equation form,
B, =B, +B,(R,Th -R,y), . . (23)
10
20
50 p,
100
PSEUDOREDUCED
PRESSURE.
where total FVF, bbl/STB, oil FVF, bbl/STB, Rsh = solution GOR at Ph. scf/STB, Rs = solution GOR at pressure of interest, scf/STB, and B, = gas FVF at pressure and temperature of interest, bbl/scf.
B, = = B, Fig. 22.12-Variation of pseudoreduced compressibility pseudoreduced pressure and temperature. with
The gas FVF requires a value for the gas compressibility or zR factor that may be obtained from Chap. 17.
0.005042, T
B,= 0
P
) .,.,,.,.................
where
B,
= gas FVF, bbllscf, ZK = gas compressibility factor at p, T, p = pressure of interest, psia, and T = temperature of interest, R.
Values of oil viscosity are required at various pressures and temperatures for both reservoir and production engineering calculations. If a PVT analysis is available, measured values of oil viscosity will be reported at reservoir temperature and at various pressures. However, as the fluid flows through the production system, the temperature also changes. This necessitates correcting the viscosity for temperature changes, which is usually accomplished by empirical correlations. The absolute viscosity, which is usually referred to merely as the viscosity, can be expressed in various units. The so-called oilfield unit is the centipoise or poise. A relationship among various systems of units is given as 1 cp = 0.01 poise = 0.001 Pa*s = 6.72~ 10m4 lbm/(ft-set). The kinematic viscosity of a fluid is the absolute viscosity divided by the density, or CL Y=-. P The most commonly used unit of kinematic viscosity is the centistoke (cSt), where the conversion to SI units is 1 cSt=10-6 m*/s. In addition to absolute and kinematic viscosity units, the units of Saybolt seconds universal (SSU) and Saybolt
A nomogra h for estimating total FVF was presented by Standing. P The correlation was developed from 387 experimental data points that included the ranges given in Table 22.9. The nomograph, on which an example calculation is worked, is shown in Fig. 22.13.
TABLE
22.9-DATA
RANGES
FOR
STANDING CORRELATION Pressure, psia GOR, scf/STB Temperature, OF Gas gravity Oil gravity, OAPl 400 to 5,000 75 to 37,000 100 to 258 0.59 to 0.95 16.5 to 63.6
22-14
PETROLEUM
ENGINEERING
HANDBOOK
seconds furol (SSF) are commonly used. An approximate conversion between centistokes and the time units can be made with the following equations.
v=0.220tsu180/rsU,
and
v=2.12tsF-139/tp, .. . . . (25)
rect this value for dissolved gas. The dead-oil viscosity depends on API gravity of the stock-tank oil and the temperature of interest. The dead-oil viscosity can be obtained from empirical correlations or, if measured values are available at two temperatures, the viscosity at any other temperature can be calculated from the equation
log[log(v+0.8)]=A+B log 7-, . . .(26)
where v = kinematic viscosity in cSt, tsu = Saybolt seconds universal, and tSF = Saybolt seconds furol. Factors Affecting Oil Viscosity The principal factors of interest to the petroleum engineer that affect viscosity are composition, temperature, dissolved gas, and pressure. Oil viscosity increases with a decrease in API gravity and also increases with a decrease in temperature. The effect of dissolved gas is to lighten the oil and thus decrease its viscosity, while an increase in pressure on an undersaturated oil compresses the oil and causes the viscosity to increase. Oil Viscosity Correlations-Saturated Systems
where v = kinematic viscosity at T, T = temperature of interest, and A,B = constants for a particular oil that can be determined if two measured values of u and T are available.
Beals Correlation for Dead Oils. Beal presented a graphical correlation showing the effects of both oil gravity and temperature on dead-oil viscosity. The correlation was developed from measurements made on 6.55 oil samples. The relationship among viscosity, API gravity, and temperature is shown in Fig. 22.14. Effect of Dissolved Gas-Chew and Connally Method. The decrease in the dead-oil viscosity as gas goes into solution can be estimated with Fig. 22.15, which was published by Chew and Connally 8 in 1959. The procedure for using the chart and an example problem are shown
The most common method for obtaining the viscosity of a crude oil that contains dissolved gas is first to estimate the viscosity of the gas-free or dead oil and then to cor-
22-15
-_ 1.
50 pressure
I 60
Oo304
i II I 06 Ill 08,
456
I 810
20 I 30 40
60 ml00 1
ond atmospheric
Fig. 22.14-Gas-free crude viscosity as a function of reservoir temperature and stock-tank crude gravity.
Fig. 22.15-Viscosity
and gas-saturated crude oils at reservoir temperature and pressure; Chew and Connallys correlation.
in the figure. They also proposed an equation for correcting for the dissolved gas: h /.lr,,,=upo(j ,...............,..........,..
(27)
where pLos = saturated-oil viscosity P,~~, = dead-oil viscosity, A,B = functions of R,,,
A = 10.715(R. ., + 100~~s, (30)
where PO = saturated oil viscosity, p,,d = gas-free or dead-oil viscosity, and a,h = functions of R,, , shown in Fig. 22.16. Beggs and Robinson Correlation. A method for calculating both dead-oil and saturated-oil viscosity was presented by Beggs and Robinson in 1975. The correlation was developed from more than 2,000 measured data points using 600 oil systems. The range of variables of the data is given in Table 22.10. The equation developed reproduced the measured data with an average error of - 1.83% and a standard deviation of 27%. The equation for dead-oil viscosity is j~~>~=lO-~-l.O. where x = pod = T = YAPI = T-.16 exp(6.9824-0.04658TpiI), dead-oil viscosity, cp, temperature of interest, F, and stock-tank oil gravity, API. .. ... .. . . . (28)
B = 5.44(R,+l50)p"8.
. . . . . . . . . . . . ..(31)
where R,v is the solution GOR in scf/STB. No graphs are required to apply this method. Example Problem 10. Calculate the viscosity of the saturated oil system described next using the Beal and Chew and Connally correlations and the Beggs and Robinson correlation where T= 137F, -yApI =22API, and R,, =90 scf/STB. Solution-Beal with Chew and Connally. From Fig. 22.14, p,d=20 cp. From Fig. 22.16, a=0.82 and h=
0.9. po,s=0.82(20)09=12.15 cp.
Interpolation was necessary on both Figs. 22.14 and 22.16 to obtain values for pl,d. a, and b. Use of Fig. 22.15 to correct for R,y gives a value of approximately 12 cp for p,,. Solution-Beggs and Robinson. x = (137) -1.63 exp [6.9824-0.04658(22)]:
x = P Od = A = 1.2658.
To correct for the effect of dissolved gas, an equation similar to Eq. 27 was developed.
CL 0,
=Ap<,f,
. . . . . . . . . . . . . . . . . . . . . . . . . . . ..(29)
10.*6* - l.O= 17.44 cp. 10.715(90+ 100) -o.s5 =0.719. B = 5.44(90+ 150) -o.338 =0.853. p,,,? = 0.719(17.44)".8"3 =8.24 cp.
22-16
PETROLEUM ENGINEERING
HANDBOOK
Fig. 22.16-a
TABLE 22.10-BEGGS AND ROBINSON CORRELATION DATA R,) SCflSTB YAW3,API 7: rg 20 to 2,070 16 to 58 0 70 to to 5,250 295
>E
Systems
Solution.
The effect of increasing the pressure on an oil system above the bubblepoint is to compress the liquid and to increase the viscosity. This effect was measured by Beal and presented graphically in Fig. 22.17. The graph gives the viscosity increase in cpi 1,OOO-psiincrease in pressure above pb as a function of the saturated or bubblepoint viscosity, pLob. Vasquez and Beggs5 extended the Beggs and Robinson correlation for undersaturated oils with the equation p. =/.~,(p/p~)~, where PO = @ob = p = pb = ... . .. ... . . . (32)
532 L 0.0178(0.745)
ph
exp 23.931(31)/(240+460)
111.187
=3,093 psia. From Eq. 28: exp[6.9824-0.04658(31)]; x = 0.4336, pod = 104336-1.0=1.71 Cp. From Eqs. 29, 30, and 31: A = 10.715(532+100-~55=0.387, B = 5.44(532+ 150) -.338 =0.599,
n = (240)-I.163
p >ph,
exp(Cj +Cdp),
. .
.(33)
where
p
C, C2 C3 C4
= = = = =
m = 2.6(4,750),87
=0.63 cp.
Example Problem 11. Calculate the viscosity of the oil system described at a pressure of 4,750 psia, with T=240F, YApI =31API, yR =0.745, and Rsb =532 scf/STB.
Gas/Oil IFT
The interfacial or surface tension existing between a liquid and gas is required for estimating capillary-pressure
22-17
forces in reservoir engineering and is a parameter in some correlations used in wellbore hydraulics calculations. The surface tension between natural gas and crude oil ranges from zero to about 35 dyne/cm. It is a function of pressure, temperature, and compositions of the phases. The surface tension of a hydrocarbon mixture can be calculated if the composition of the mixture at the pressure and temperature of interest is known. The parachor of each component must also be known. The parachor is the molecular weight of a liquid times the fourth root of its surface tension, divided by the difference between the density of the liquid and the density of the vapor in equilibrium with it. It is essentially constant over wide ranges of temperature. An equation for estimating surface tension is
oo.25=
9oc
80C
7oc
600
(34)
0 500 Jz u 0 i 400
where
Pchi=
u = surface tension, dyne/cm. parachor of ith component, xi = mole fraction of ith component in the liquid phase, v; = mole fraction of ith component in the vapor phase, pi = density of the liquid phase, g/cm, P,, = density of the vapor phase, g/cm 3, ML = molecular weight of liquid phase, M,. = molecular weight of vapor phase, and C = number of components.
300
200
fO0
i
0 400
200 300 weight
0,
Parachors for some hydrocarbons, nitrogen, and carbon dioxide are given in Table 22.11. A correlation of the parachor with molecular weight was presented by Katz and is shown in Fig. 22.18. Empirical correlations in the form of graphs were presented by Baker and Swerdloff where surface tension is correlated with temperature, API gravity, and pressure. The correlations are shown in Figs. 22.19 and 22.20. Equations that approximate Figs. 22.19 and 22.20 are: aes=39-0.2571yAP1, and u,oo=37.5-0.2571y/,~1, where
068 = IFT at 68F, dyne/cm, alOO = IFT at lOOF, dyne/cm, and yAPI = gravity of stock-tank oil, API.
400
Molecular
Flg. 22.18-Parachors for hydrocarbons. 0, n-paraffins; heptanes plus; A, gasolines; A, crude oil.
TABLE
FOR
..
. .
. ..
. (35)
77.0 108.0 150.3 181.5 190.0 225 232 271 311 352 41.0 78.0
. . . . _. .
.(36) lsopentane
n-Pentane n-Hexane n-Heptane n-Octane Nitrogen (in n-heptane) Carbon dioxide
It has been suggested that if the temperature is greater than lOOF, the value at 100F should be used. Also, if T<68F, use the value calculated at T=68F. For intermediate temperatures, use linear interpolation between the values obtained at 68 and 100F. That is:
(T-68"F)(u68-LJloo) , ,..........
The effect of gas going into solution as pressure is increased on the gas/oil mixture is to reduce the IFT. The dead-end oil IFT can be corrected by multiplying it by the following correction factor. F,.= 1.0-0.024p~45, where p is in psia . .. . . . . . . (38)
(37)
22-18
PETROLEUM ENGINEERING
HANDBOOK
dead oil viscosity saturated oil viscosity kinematic viscosity fluid density surface tension, interfacial
The IFT becomes zero at miscibility pressure, and for most systems, this will be at any pressure greater than about 5,000 psia. Most of the correlations presented in this chapter are included in a calculator program for the HP-41C. The program is available from Hewlett-Packard.
PO =
(1)
Nomenclature
CI = A = h = B = B,? = B,, = function of R,, in Eq. 27 constant in Eq. 26 function of R,, in Eq. 27 constant in Eq. 26 gas formation volume factor oil formation volume factor oil isothermal compressibility pseudo-reduced compressibility number of components
)I# =
(6)
= = = = = =
=
tn ; =
RI; R,
= =
T,,r = TR = , = v; =
constants correlating function in Eq. 18 correcting factor for dead oil interfacial tension. Eq. 38 pressure-dependent exponent in Eq. 32 mass of ith component molecular weight of stock-tank oil pressure of interest bubblepoint pressure pseudo-critical pressure pseudo-reduced pressure parachor of ith component free gas/oil ratio solution gas/oil ratio solution gas/oil ratio at the bubblepoint pressure Saybolt seconds universal Saybolt seconds furol temperature of interest pseudo-critical temperature pseudo-reduced temperature reservoir temperature specific volume volume of ith component symbol for various groups of terms, Eq. 28 symbol for various groups of terms, Eqs. 5, 6, 9. IO, and 13-16 gas compressibility factor fluid specific gravity oil API gravity corrected gas gravity oil specific gravity fluid viscosity oil viscosity at bubblepoint pressure
MO =
For ys 50.60:
PhY,g
=8.427 exp(2.78~,Y)-4.01.
TR
102.51~ -R .6+24.20.
O.O15lT, Y,p Yg YO
3.848 Yn
-0.OlJT,,3.576)
log (G)
22-19
42 40. 38.
1
SURFACE TENSION OF CRUDE OILS AT ATMOSPHERICPRESSURE SURFACETENSIONOF CRUDEOILS
EFFECT OF DISSOLVED GAS IN SOLUTION AT VARIOUS SATURATION PRESSURES ANO ATMOSPHERIC TEMPERATURE
$J36. 5 34. E
u z ; ; 32. 30. x8-
5 26. 24.
y. <0.876 Cl c2 C3
c4
B 0 =0.972
-0.000147F~75,
..
. . (18)
where
R,, =yg
P 519.7x10?F
+2.25T-5.75.
>
3 .....
1.769/r,.
where
254.7T -236.7T-Yo Yo
(14)
FZ-....--
.7.. Cl C2
ys =0.359 In
+0.476
>
..
c3
For 3
T
~40.7: 28.1RS+30.6T-1180r,+co =
1784 - 10 910
Yo
O.O098py,
?',q =
0.281
px105 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (22)
-0.236
T
>
c, exp($-+), R.S=ClYg
where the following constants apply
. .. .. . (17)
...
. (28)
exp(l3.108-6.591/y.).
y. > 0.876 7.803x10- 1.0937 2022.19 1879.28 A=10.715(5.615R.+100)-0~5. A B=5.44(5.615Rs+150) -.338. . . . . (30) . (31)
22-20
PETROLEUM
ENGINEERING
HANDBOOK
in in in in in in in in
Differential process is a term used primarily in PVT work to indicate that, as a system is caused to move through a bubblepoint or dewpoint and as a result forms two phases, the minor phase is removed from further contact with the major phase. Thus, during a differential process, the system is continuously changing in quantity and composition. Dissolved gas (solution gas) identifies tnaterial ordinarily gaseous at atmospheric conditions but which is part of a liquid phase at elevated pressure and temperature. Dissolved-gas drive (solution-gas drive) is a primary recovery process whereby liquid (oil) is displaced from the reservoir rock by energy of expansion of gas components originally dissolved in the liquid. Flash gas liberation means that a gas forms from a liquid (usually on reduction of pressure) under conditions wherein the total composition of the system remains the same in time. An example of flash gas liberation occurs during the steady-state flow of oil and gas through a gas-oil separator. Flash process is one in which the composition of the system remains constant but the proportion of gas and liquid phases that comprise the system changes as pressure or some other independent variable is changed. For example. the determination of the PV relations of a reservoir fluid sample involves a flash process. Formation volume factor (FVF) is a means of expressing a volumetric relation of a system. The volume of fluid, at formation pressure and temperature. that results in 1 bbl stock-tank oil is, by definition, the formation volutne B. If the system is in a single condensed state in the formation, the term oil-FVF B,, is used. If two phases exist, the term used is total-FVF B, or two-phase formation volume. Oil-formation volumes normally are between 1.1 and 2.0. Total formation volumes may be as great as 200. depending on the system composition and the pressure involved. Gas-formation volume B,q refers to the volume in the reservoir (usually expressed as barrels) per 1,000 std surface cu ft of the gas. Water-formation volume B,,. designates the volume in the fortnation occupied by I surface bbl water and its attendant gases. Water-formation volumes are usually in the range of 0.99 to 1.07. In all instances, the reference state is standard surface conditions of 14.7 psia and 60F unless specified otherwise. Gas gravity is a simple means of expressing the molecular weight of a gas. The unit of gas gravity is dry air of molecular weight 28.97. Thus, methane (molecular weight= 16.04) has a gas gravity of 16.04/28.97=0.55. Gas-oil ratio (GOR) is a loose expression of system composition. Normally, the units involved are cubic feet of gas per I bbl liquid. both measured at 14.7 psia and 60F. However. in some instances. local usage calls for tneasuring the gas at some pressure other than 14.7 psia. Further, the barrel of liquid may refer to some pressure other than the usual stock-tank oil. The units of reservoir oil and residual oil are encountered quite often in PVT work. Lastly. several kinds of GOR are used in rescr-
Glossary
These are generally accepted definitions, peculiar to reservoir engineering phase-behavior work. Apparent liquid density is the ratio of mass to volume of a gas when dissolved m a liquid. It normally is calculated for conditions of 14.7 psia and 60F by correcting the observed density of the system to that state and subtracting the mass and volume of the liquid component of the system. Bubblepoint of a system is the state characterized by the coexistence of a substantial amount of liquid phase and an infinitesimal amount of gas phase in equilibrium. Compressibility factor (gas-deviation factor, supercompressibility factor) is a multiplying factor introduced into the ideal-gas law to account for the departure of true gases from ideal behavior (pV=&T: z is the compressibility factor). Condensate (distillate) liquids are liquids formed by condensation of a vapor or gas. The term usually identifies a light-colored liquid, usually ofSOAPI gravity or higher, obtained from systems that exist in the gaseous phase in the reservoir. Critical state is the term used to identify the unique condition of pressure. temperature, and composition wherein all properties of coexisting vapor and liquid become identical. Critical temperature T, and/or pressure pC is the tenperature or pressure at the critical state. Dewpoint of a system is analogous to the bubblepoint in that a large volume of gas and an infinitesimal amount of liquid coexist in equilibrium. Dewpoint pressure p(, is the gas pressure in a system at its dewpoint. Differential-gas liberation indicates removal of a gas phase from the system as the gas forms at bubblepoint conditions. It is generally believed that the differentialgas-liberation process is the one that predominates in solution-gas-drive reservoirs during the greater part of their producing life.
OIL SYSTEM
CORRELATIONS
22-21
R,
voir engineering, such as solution (pas solubility in oil) producing R, and cumulative R,, Pressure and tcmperature of separation and the number of stages used affect the GOR number obtained for a given system.
ttirential gas-removal process as stock-tank oil is to a flashgas process-the end liquid product. Saturated liquid is a liquid in equilibrium with vapor at the saturation pressure. Likewise, saturated vapor dcnotcs equilibrium with liquid. These terms are often used aynonymously with the term bubblepoint liquid (dewpoint vapor) at the bubblepoint (dewpoint) pressure. Note that the terms bubblepoint and dewpoint identify, the spccial case where the minor phase is present only III an infinitcsimal amount. whereas the term saturated liquid dots not involve the relative amounts of phases present. Shrinkage refers to the decrease in volume of a liquid phase caused by release of solution gas and/or by the thermal contraction of the liquid. Shrinkage may be expressed (1) as a percentage of the final resulting stocktank oil or (2) as a percentage of the original volume of the liquid. Shrinkage factor is the reciprocal of FVF expressed as barrels of stock-tank oil per barrel of rcscrvoir oil. A reservoir oil that rcsultcd in 0.75 bbl of stock-tank oil per I bbl of reservoir oil would have a shrinkage of 0.25/0.75=33%~ under Definition I. a shrinkage of 0.25/1.00=25% under Definition 2, a shrinkage factor of0.75. and an FVF of I .00/0.75= 1.33. Solution gas-oil ratio, R,, . expresses the amount of gas in solution. or dissolved, in a liquid. The reference oil may be stock-tank oil or residual oil. On occasion. reservoir saturated oil is used as the rcferencc. Standard conditions (surface) arc 14.7 psia and 60F. Gas volumes may be specified on occasion at prcssurcs slightly removed from 14.7 psia. Stock-tank oil is the liquid that results from production of reservoir material through surface equipment that separates normally gaseous components. Stock-tank oil may be caused to vary in composition and properties by varying the conditions of gas/liquid separation. Stock-tank oil is normally reported at 60F and 14.7 psia but may be measured under other conditions and corrected to the standard condition. System refers to a body or to a composition of matter that represents the material under consideration. The term system may be defined further as a homogeneous system or a heterogeneous system. In a homogeneous systern, the intensive propertles vary only in a continuous manner with respect to the extent of the system. A hctcrogeneous system is made up of a number of homopcneous parts, and abrupt changes in the intensive proportics occur at the interface between the homogeneous parts. Undersaturated fluid (liquid or vapor) is material capablc of holding additional gaseous or liquid components in solution at the specified state.
Mole is one molecular weight unit of any substance. For example. 16.04 Ibm methane is a lbm-mol. Likewise. 16.04 p methane constitute a gmol. The Ibm-mole is used for petroleum engineering work. A Ibm-mol of gas (perfect) occupies 370.4 cu ft at 14.7 psia and 60F. Phase is a portion of a system that differs in its intensive properties from adjacent portions of the system. An interface exists between phases because of this difference in properties. Systems involved in petroleum production normally occur in not more than two phases. gas and liquid. On occasion and for hydrocarbon systems of unusual composition. two liquid phases or a solid phase may occur. Properties, Extensive and Intensive. Properties dircctIy proportional to the amount of material making up the system are termed extensive properties. Examples arc area, mass, inertia. and volume. An intensive property is one that is independent of the amount of material. Density. pressure. temperature, viscosity, and surface tension are intensive properties. Energy is the product of an intensive property and an extensive property: for example. pressure times volume is mechanical energy. Pseudocritical and Pseudoreduced Properties (Temperature, Pressure). Properties of pure hydrocarbons are often the same when expressed in terms of their reduced properties. The same reduced-state relationships often apply to multicomponent systems if pseudo critical tcmperatures and pressures are used rather than the true criticals of the systems. Calculation of the pseudocritical values from the composition of the system varies depending on the correlation being used. The ratio of the properly to the pseudocritical property is called the pseudoreduced property, e.g., pseudoreduced pressure Ppr =Ph,x Reduced properties (temperature, pressure, volume) are the ratio of the property to the critical property: for example, the reduced pressure p I =p/~),. . Relative oil volume is analogous to formation volume except that the reference state is other than at standard surface conditions and the oil is other than stock-tank oil. For example, the term is used often on the basis of I bhl bubblepoint oil, or saturated oil, as the reference volume. Relative oil volumes must specify pressure, tcmpcraturc. and some composition parameter, e.g., relative oil volume =0.7 (2.520 psia, 185F. bubblepoint oil = 1.O). Residual oil is a term common in PVT work to identify the liquid remaining in a PVT cell at the completion of a differential process carried out at or near the reservoir temperature. By analogy. it refers also to the liquid that remains in an oil reservoir at depletion. General usage is that residual oil volumes and gravities are reported in PVT work at 60F and 14.7 psia. Residual OII is to a dif-
u/id Pm/.
Pmr
. API (1942).
22-22
PETROLEUM
ENGINEERING
HANDBOOK
4. Lahater, J.A : Bubble Point Prcwrc Correlation. Twf\.. AIME (1958) 213. 379-81. 5. Vasquer. M. and Beep\, H.D.: Correlat~on$ for Fluid Phy\al Property Prediction, J. PH. Twh. (June 1980) 96X-70. 6. Trube. -A.S.: Compresaihility of Undersaturated Hydrocarbon Reservoir Fluids. Trw~s.. AIME (1957) 210. X-57 7. Beal. C.: The Viscosity of Air. Water. Natural Gas. Crude Oil and Its Associated Gases at 011 Field Temperatures and Pre\wre~. T/-w.\. , AIME (1946) 165. 94- Il5. X. Chew. J. and Connally, C.A : A Viscwity Correlation for (;a\Saturated Crude Oils. Trcmx.. AIME (1959) 216. 23-25 9. Beg, H.D. and Robinson. J.R.: Estimating the Viwwty of Crude Oil Sysrems. J. Pe. Twh. (Sept. 19751 1140-41. IO. Katz. D.L . Monroe, R R.. and Tramer. R.R : Surface Tension (11Crude Oils Contaimng Dioolved Case\, PH. KY/I. (Sept. 19431 I-IO. Surface Tcnswn of. I I. Baher. 0. and Swerdloff. W.. Findmg Hydrocarbon Liquids. Uii oni/ Gay J. (Jan. 2. 1956). IZ. HP-41C Petroleum Fluid!, Pac. Hewlett-Packard. 1000 N.E. Clrcle Blvd.. Corvalhs. OR 97330.
Lacey. W N.. Sage. B.H.. and Kircher. C.E. Jr.: Phnsc Equllihrla in Hydrocarbon System\,. 111~ Solubility of a Dry Natural Ciao 111 Crude Oil. /!I(/. N)/(l Orx. Clwrr. (IY3.l) 26. hS?. Lcwts. W.K. and Squires, L.: Mechanism Cm J. (1934) 33, No. 26. 92. ofOil Viwo~ity. Oi/tr,~t/
Nelson. W.L.: How to Handle Viscous Crude Oil. (1954) 53. No. 28. 269. Norton. A.E.: L~thnccoirv~. McGmwHill Cay (1942). Perry. J .H.: Chrrnic~trl Eqitwers Inc.. New York City (19.50). Schilthuis. R.I.: Active (1936) 118. 33-52.
Htrwliwok.
McGraw-Hil
I Bwh Co.
~rwn.\. . AIME
General References
Borden. G. Jr. and Raw, M.J.: Correlation of Bottom Hole Sample AlME (1950) 189. 345-48. Data. ~wrcr,~.s.. Dodson. C.R., Goodwill. D.. and Mayer. E.H.: Application of Lahoratory PVT Data to Reservoir Engineerins Problems. frrras.. AIME (19.53) 198. 287-98. Gosline. J.E. and Dodson. C.R.: Solubility Relations and Volumetric Behavior of Three Gravitler of Crudes and Associated Gases. Drill. crml Pd. Prtrc. . API (1942) 137.
Standing. M.B. and Katz. D.L.: Densq ofCrude 011s Saturated With Ndtural Gas. Twms., AIME (1941) 136, ISYQS van Wijk. W R., devries. D.A., and Thijscn. H.A.C.: Study nf PVT Relations of Reservoir Flui&. Pro<,. , Fourth World Petroleum Congress. II, 313. Vink. D.J. cf a/.: Multiple-Phase Cm J. (1940) 39. No. 28. 34. Hydrocarbon System\. Ofi cr,rd
Petroleum
Chapter 23
Phase Diagrams
F.M. OTKJr., New Mexico Inst. of Mining and Technology J.J. Taber, New Mexico Inst. of Mining and Technology*
Introduction
Petroleum reservoir fluids are complex mixtures containing many hydrocarbon components that range in size from light gases such as methane, ethane, and so on to very large hydrocarbon molecules containing 40 or more carbon atoms. Nonhydrocarbon components such as nitrogen, hydrogen sulfide (HzS), or CO2 also may be present. Water, of course, is usually present in large quantities in all reservoirs. At a given temperature and pressure, the components distribute between whatever solid, liquid, and vapor phases are present. A phase may be defined as that portion of a system that is homogeneous, bounded by a surface, and physically separable from other phases present. Equilibrium phase diagrams offer convenient representations of the ranges of temperature, pressure, and composition within which various combinations of phases coexist. Phase behavior plays an important role in a variety of reservoir engineering applications ranging from pressure maintenance to separator design to enhanced oil recovery (EOR) processes. This chapter reviews the fundamentals of phase diagrams used in such applications. terminates at the critical point. At temperatures above the critical temperature, T,, a single phase forms over the entire range of pressures. Thus, for a single component, the critical temperature is the maximum temperature at which two phases can exist. Critical temperatures of hydrocarbons vary widely. Small hydrocarbon molecules have low critical temperatures, while those of large ones are much higher. Critical pressures generally decline as the molecular size increases. For instance, the critical temperature and pressure of methane are - 117F and 668 psia. For decane, the values are 652F and 304 psia. (Fig. 23.9 shows additional vapor pressure curves and critical points for several other light hydrocarbon molecules.) For many reservoir engineering applications liquid/vapor equilibrium is of greatest interest, though liquid/liquid equilibria are important in some EOR processes. Solid/liquid phase changes, such as asphaltene or paraffin precipitation, occasionally occur in petroleum production operations. Fig. 23.2 shows typical volumetric behavior of a single component in the range of temperatures and pressures near the vapor-pressure curve in Fig. 23.1. If the substance under consideration is placed in a pressure cell at constant temperature T1 below T, and at a low pressure (Point A, for instance), it forms a vapor phase of high volume (low density). If the volume of the sample is decreased with the temperature held constant, the pressure rises. When the pressure reaches p,, (T1 ), the sample begins to condense. The pressure remains constant at the vapor pressure until the sample volume is reduced from the saturated vapor volume (V,.) to that of the saturated liquid (V,). With further reductions in volume, the pressure rises again as the liquid phase is compressed. Note that small decreases in volume give rise to large pressure increases in the liquid phase because of the low compressibility of liquids. At a temperature Tz above the critical temperature, no phase change is observed. Instead, the sample can be
PETROLEUM
ENGINEERING
HANDBOOK
MELTING CURVE
i:
SOLID
CRITICAL POINT
3 z Pb
VAPORIZATION
k a
Pd
SUBLIMATION CJRVE
VAPOR
PV2
TC
A MOLE
B FRACTION
XE COMPONENT I
1.0
T Fig. 2X1-Phase
MPERATURE
Fig. 23.3-Pressure-composition phase diagram for a binary mixture at a temperature below the critical temperature of both components.
Tc TC Tc
I I L I Vc VOLUME 1 Vv
of freedom (F=O). Thus, the maximum number of phases possible is three. Therefore, if three phases coexist in equilibrium (possible only at the triple point), the pressure and temperature are fixed. If only two phases are present for a pure component, then either the temperature or the pressure can be chosen. Once one is chosen, the other is determined. If the two phases are vapor and liquid, for example, choice of the temperature determines the pressure to be the vapor pressure at that temperature. These permitted pressure-temperature values lie on the vapor pressure curve of Fig. 23.1. In a binary system, two phases can exist over a range of temperatures and pressures. The number of degrees of freedom is
F=2-nc, . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...(2)
compressed from high volume (low density) and low pressure to low volume (high density) and high pressure with only one phase present.
and hence both the temperature and pressure can be chosen, though there is no guarantee that two phases will occur at a specific choice of T and p. For multicomponent systems, the phase rule provides little guidance, since the number of phases is always far less than the maximum number that can occur. Hence, as the number of components increases, more component concentrations must be known to determine the system.
Phase Rule
The maximum number of phases that can coexist at fixed temperature and pressure is determined by the number of components present. The phase rule states that
F=2+C-P-n,, ...... .... .... . . . . . . (1)
Types of Diagrams
Binary Phase Diagrams Fig. 23.3 shows typical vapor/liquid phase behavior for a binary system at a fixed temperature below the critical temperature of both components. Such a diagram is known as a pressure-composition phase diagram. At pressures below the vapor pressure of Component 2, ~2, any mixture of the two components forms a single vapor phase. At pressures between p 1 and p,,~ , two phases can coexist for some compositions. For instance, at pressure pb two phases will occur if the mole fraction of Component 1 lies between xE and xE. If the mixture
where F is the number of degrees of freedom, C is the number of components, P is the number of phases, and n c is the number of constraints. For a single-component system, the maximum number of phases occurs when there are no constraints (n c =0) and there are no degrees
PHASE DIAGRAMS
23-3
--pi pd
--PZPl
--P3P2 PPd2
PUMP
Fig. 23.6-Volumetric behavior of a binary mixture at a constant temperature showing retrograde vaporization.
Pd2
Pdl
PVZ VAPOR
xI
MOLE FRACTION COMPONENT
K2
I
TEMPERATURE
Fig. 23.5-Pressure-composition phase diagram for a binary mixture at a temperature above the critical temperature of Component 1.
Fig. 23.7--Regions of temperature, pressure, and composition for which two phases occur in a binary liquid/vapor system.
composition is xB, it will be all liquid, if it is xE, all vapor. For 1 mol of mixture of overall composition, z, between -xE and xE, the number of moles of liquid phase is XE --z
XE -xB
L=-.
. . . . . . . . . . . . . . . . . . . . . . . . . . ...(3)
Eq. 3 is known as an inverse lever rule because it is equivalent to a statement concerning the distances along a tie line from the overall composition to the liquid and vapor compositions; L=ZE/BE, where ZE and BE are lengths on the tie line shown in Fig. 23.3. Thus, the amount of liquid is proportional to the distance from the overall composition to the vapor composition divided by the length of the tie line. The line connecting the compositions of phases in equilibrium is known as a tie line. In binary phase diagrams such as Fig. 23.3, the tie lines are always horizontal.
Phase diagrams such as Fig. 23.3 can be determined experimentally by placing a mixture of fixed overall composition in a high-pressure cell and measuring the pressures at which phases appear and disappear. For example, a mixture of composition xa would show the behavior indicated qualitatively in Fig. 23.4. At a pressure less thanpd (Fig. 23.3), the mixture is a vapor. If the mixture is compressed by injecting mercury into the cell, the first liquid, which has composition xA, appears at the dewpoint pressure, pd. As the pressure is further increased, the volume of liquid grows as more and more of the vapor phase condenses. The last vapor, of composition X,R, disappears at the bubblepoint pRssut& pb. If the system temperature is above the critical temperature of one of the components, the phase diagram is similar to that shown in Fig. 23.5. (See Fig. 23.13 for additional examples of this type of phase diagram.) At the higher temperature, the two-phase region no longer extends to the pure Component 1 side of
23-4
PETROLEUM ENGINEERING
HANDBOOK
TEMPERATURE
Fig. 23.8-Projection of the vapor pressure (p,, and pa) curves and locus of critical points for binary mixtures. Points C, and C2 are the critical points of the pure components. Fig. 23.10-Properties of ternary diagrams
Fig. 23.9-Vapor pressure curves for light hydrocarbons and critical loci for selected hydrocarbon pairs.
the diagram. Instead, there is a critical point, C, at which liquid and vapor phases are identical. The critical point occurs at the maximum pressure of the two-phase region. The volumetric behavior of mixtures containing less Component 1 than the critical mixture is like that shown in Fig. 23.4. Fig. 23.6 shows the volumetric behavior of mixtures containing more Component 1. Compression of mixture of composition x2 (in Fig. 23.5) leads to the appearance of liquid phase of composition xt when pressure pDl is reached. The volume of liquid first grows and then declines with increasing pressure. The liquid phase disappears again when pressure pm is reached. Such behavior is called retrograde vaporization (or retrograde condensation if the pressure is decreasing). If the system temperature is exactly equal to the critical temperature of Component 1, the critical point on the binary pressure-composition phase diagram is located at a Component 1 mole fraction of 1.O. Fig. 23.7 shows the behavior of the two-phase regions as the temperature rises. As the temperature increases, the critical point moves to lower concentrations of Component 1. As the critical temperature of Component 2 is approached, the two-phase region shrinks, disappearing altogether when the critical temperature is reached. A typical locus of critical temperatures and pressures for a pair of hydrocarbons is shown in Fig. 23.8. The critical locus shown in Fig. 23.8 is the projection of the critical curve in Fig. 23.7 onto the p-T plane. Thus, each point on the critical locus represents a critical mixture of different composition, though composition information is not shown on this diagram. For temperatures between the critical temperature of Component 1 and Component 2, the critical pressure of the mixtures can be much
PHASE DIAGRAMS
23-5
Fig. 23.11-Ternary phase diagram at a constant temperature and pressure for a system that forms a liquid and a vapor.
Fig. 23.12-Ternary phase diagram for the methane/ butane/decane system at 160DF 171Cl.
higher than the critical pressure of either component. Thus, two phases can coexist at pressures much greater than the critical pressure of either component. If the difference in molecular weight of the two components is large, the critical locus may reach very high pressures. Fig. 23.9 gives critical loci for some hydrocarbon pairs. The binary phase diagrams reviewed here are those most commonly encountered. More complex phase diagrams involving liquid/liquid and liquid/liquid/vapor equilibria do occur, however, in hydrocarbon systems at very low temperatures (well outside the range of conditions encountered in reservoirs or surface separators) and in COz/crude oil systems at temperatures below about 50C. For reviews of such phase behavior see Refs. 2 and 3. Ternary Phase Diagrams Phase behavior of mixtures containing three components is conveniently represented on a triangular diagram such as that shown in Fig. 23.10a. Such diagrams are based on the property of equilateral triangles that the sum of the perpendicular distances from any point to each side of the diagram is a constant equal to length of any of the sides. Thus, the composition of a point in the interior of the triangle is
x, =-)
(Fig. 23.10b). In addition, mixtures lying on any line connecting a comer with the opposite side contain a constant ratio of the components at the ends of the side (Fig. 23.1Oc). Finally, mixtures of any two compositions, such as A and B in Fig. 23. lOd, lie on a straight line connecting the two points on the ternary diagram. Compositions represented on a ternary diagram can be expressed in volume, mass, or mole fractions. For vapor/liquid equilibrium diagrams, mole fractions are most commonly used. Typical features of a ternary phase diagram for a system that forms a liquid and a vapor at fixed temperature and pressure are shown in Fig. 23.11. Mixtures with overall compositions that lie inside the binodal curve will split into liquid and vapor. Tie lines connect compositions of liquid and vapor phases in equilibrium. Any mixture on one tie line gives the same liquid and vapor compositions. Only the amounts of liquid and vapor change as the overall composition changes from the liquid side of the binodal curve to the vapor side. If the mole fractions of component i in the liquid, vapor, and overall mixture are xi, yi, and zi, the fraction of the total moles in the mixture in the liquid phase is given by
L= Yi -zi yi-xi ........ . . . . . . . ......
. . (5)
Ll
LT
x2 =-)
L2 LT
x3 =-)
L3 LT
Several other useful propetties of triangular diagrams are also a consequence of this fact. For mixtures along any line parallel to a side of the diagram, the fraction of the component of the comer opposite to that side is constant
Fq. 5 is another lever rule similar to that described for binary diagrams. The liquid and vapor portions of the binodal curve meet at the plait point, a critical point where the liquid and vapor phases are identical. Thus, the plait point mixture has a critical temperature and pressure equal to the conditions for which the diagram is plotted. Depending on the pressure and temperature and components, a plait point may or may not be present. Any one tematy diagram is given for fixed temperature and pressure. As either the temperature or pressure is varied, the location of the binodal curve and
23-6
PETROLEUM ENGINEERING
HANDBOOK
1
P A , 3 z i; 1 A TEMPERATURE
1000
0.2 MOLE
0.4 FRACTION
0.6 METHANE
0.S
: I 0
Fig. 23.13-Pressure-composition phase diagrams for methane/butane and methaneldecane binary systems at 160F [71C].
slopes of the tie lines may change. Fig. 23.12 shows the effect of increasing pressure on ternary phase diagrams for mixtures of methane (C 1), butane (C,), and decane (COO) at 160F. 4J The sides of the ternary diagram represent a binary system, so the ternary diagram includes whatever binary tie lines exist at the temperature and pressure of the diagram. Fig. 23.13 shows the corresponding binary phase diagrams for the C l-C4 and C 1-C 1o pairs. The Cd-C 10 pair is not shown because it forms two phases only below the vapor pressure of C4, about 120 psia at 160F (see Fig. 23.9). As shown in Fig. 23.12, at 1,000 psia the two-phase region is a hand that stretches from the C 1-C 10 side of the diagram to the tie line on the C I -Cd side. If the pressure is increased above 1,000 psia, the liquid composition line shifts to higher methane concentrations; methane is more soluble in both C4 and Cl0 at the higher pressure (see Fig. 23.13). The two-phase region detaches from the C ,-C4 side of the diagram at the critical pressure of the C I -C4 pair (about 1,800 psia). As the pressure increases above that critical pressure, the plait point moves into the interior of the diagram (Fig. 23.12, lower diagrams). With further increases in pressure, the two-phase region continues to shrink. It would disappear completely from the diagram if the pressure reached the critical pressure of the Cl-C 10 system at 160F (nearly 5,200 psia). Reservoir Fluid Systems Real reservoir fluids contain many more than the two or three components, so phase composition data can no longer be represented with two or three coordinates. In-
stead, phase diagrams, which give more limited information, are used. Fig. 23.14 shows one such diagram for a multicomponent mixture. Fig. 23.14 gives the region of temperatures and pressures at which the mixture forms two phases. The analog of Fig. 23.14 for a binary system can be obtained by taking a slice at constant mole fraction of Component 1 through the diagram in Fig. 23.7. Also given arc contours of liquid volume fractions, which indicate the fraction of total sample volume occupied by the liquid phase. Fig. 23.14 does not give any compositional information, however. In general, the compositions of coexisting liquid and vapor will be different at each temperature and pressure. At temperatures below the critical temperature (Point C), a sample of the mixture described in Fig. 23.14 splits into two phases at the bubblepoint pressure (Fig. 23.4) when the pressure is reduced from a high level. At temperatures above the critical temperature, dewpoints are observed (Fig. 23.6). In this multicomponent system, the critical temperature is no longer the maximum temperature at which two phases can exist. Instead, the critical point is the temperature and pressure at which the phase compositions and all phase properties are identical. The bubblepoint, dewpoint, and single-phase regions shown in Fig. 23.14 are sometimes used to classify reservoirs. At temperatures above the cricondentherm, the maximum temperature for the formation of two phases, only one phase occurs at any pressure. For instance, if the hydrocarbon mixture of Fig. 23.14 were to occur in a reservoir at temperature TA and pressure PA (Point A), a decline in pressure at approximately constant temperature caused by removal of fluid from the reservoir would not cause the formation of a second phase. While the fluid in the reservoir remains a single phase, the produced gas splits into two phases as it cools and expands to surface temperature and pressure at Point A. Thus, some condensate would be collected at the
PHASE DIAGRAMS
23-7
surface even though only one phase is present in the formation. The amount of condensate collected depends on the operating conditions of the separator (or separators). The lower the temperature at a given pressure, the larger the volume of condensate collected (Fig. 23.14). Dewpoint reservoirs are those for which the reservoir temperature lies between the critical temperature and the cricondentherm for the reservoir fluid. Production of fluid from a reservoir starting at Point B in Fig. 23.14 causes liquid to appear in the reservoir when the dewpoint pressure is reached, and as the pressure declines further, the saturation of liquid increases because of retrograde condensation. Because the saturation of liquid is low, only the vapor phase flows to producing wells. Thus, the overall composition of the fluid remaining in the reservoir changes continuously. However, the phase diagram shown in Fig. 23.14 is for the original composition only. The preferential removal of light hydrocarbon components in the vapor phase generates new hydrocarbon mixtures which have a greater fraction of the heavier hydrocarbons. Differential liberation experiments, in which a sample of the reservoir fluid initially at high pressure is expanded through a sequence of pressures, can be used to investigate the magnitude of the effect of pressure reduction on the vapor composition. At each pressure, a portion of the vapor is removed and analyzed. Such an experiment simulates what happens when condensate is left behind in the reservoir as the pressure declines. As the reservoir fluid becomes heavier, the boundary of the two-phase region in a diagram like Fig. 23.14 shifts to higher temperatures. Thus, the composition change also acts to drive the system toward higher liquid condensation. Such reservoirs are candidates for pressure maintenance by lean gas injection to limit the retrograde loss of condensate or for gas cycling to vaporize and recover some of the liquid hydrocarbons. Bubblepoint reservoirs are those in which the temperature is less than the critical temperature of the reservoir fluid (Point D in Fig. 23.14). These reservoirs are sometimes called undersaturated because there is insufficient gas for a gas phase at that temperature and pressure. Isothermal pressure reduction causes the appearance of a vapor phase at the bubblepoint pressure. Because the compressibility of the liquid phase is much lower than that of a vapor, the pressure in the reservoir declines rapidly during production in the single-phase region. The appearance of the much more compressible vapor phase reduces the rate of pressure decline. The volume of vapor present in the reservoir grows rapidly with reduction of reservoir pressure below the bubblepoint. Because the vapor viscosity is much lower than the liquid viscosity, and the gas relative permeability goes up markedly with increasing gas saturation, the vapor phase flows more easily. Hence, the produced GOR climbs rapidly. Again, pressure maintenance by water drive, water injection, or gas injection can substantially improve oil recovery over the 10 to 20% recovery typical of pressure depletion in these solutiongas-drive reservoirs. As in dewpoint reservoirs, the composition of the reservoir fluid changes continuously once the two-phase region is reached. There is, of course, no reason why initial reservoir temperatures and pressures cannot lie within the twophase region. Oil reservoirs with gas caps and gas reser-
r-----Fig. 23.15~Increase in OAPI gravity with depth: (a) Ordovician Ellenberger reservoirs in Delaware Val Verde basin; (b) Pennsylvanian Tensleep reservoirs in Wyoming.
voirs with some liquids present are common. There also can be considerable variation in the initial composition of the reservoir fluid. The discussion of single-phase, dewpoint, and bubblepoint reservoirs was based on a phase diagram for one fluid composition. Even for one fluid, all the types of behavior occur over a range of temperatures. In actual reservoir settings, the composition of the reservoir fluid correlates with depth and temperature. Deeper reservoirs usually contain lighter oils.j Fig. 23.15 shows the relationships between oil gravity and depth for two basins. The higher temperatures of deeper reservoirs alter the original hydrocarbon mixtures to produce lighter hydrocarbons over geologic time. 6 Low oil gravity, low temperature, and relatively small amounts of dissolved gas all combine to produce bubblepoint reservoirs. High oil gravity, high temperatures, and high GORs produce dewpoint or condensate systems. Phase Diagrams for EOR Processes Phase behavior plays an important role in a variety of EOR processes. Such processes are designed to overcome, in one way or another, the capillary forces that act to trap oil during waterflooding. In surfactant/polymer processes, the effects of capillary forces are reduced by injection of surfactant solutions that contain molecules with oil- and water-soluble portions. Such molecules migrate to the oil/water interface and reduce the interfacial tension, thereby reducing the magnitude of the capillary forces that resist movement of trapped oil. Miscible displacement processes are designed to eliminate interfaces between the oil and the displacing phase, thereby removing the effects of capillary forces between the injected fluid and the oil. Unfortunately, fluids that are strictly miscible with oil are too expensive for general use. Instead, fluids such as methane or methane enriched with intermediate hydrocarbons, CO*, or nitrogen are injected, and the required miscible displacing fluid is generated by mixing of the injected fluid with oil in the reservoir.
23-8
PETROLEUM
ENGINEERING
HANDBOOK
TYPE b1a.b
II(
DEW
POINTS
O-0-b
SURfACTANl b. TYPE III
SURfACTANl
/ & BllHE
Fig. 23.16-Ternary
c. TYPE
II( + )
\ 75% OIL
VOLUME
50%
PERCENT LIQUID
25%
PHASE
5% I
MOLE
PERCENT
CO2
Phase diagrams typical of those used to explain the behavior of surfactant systems are shown in Fig. 23.16. In those ternary diagrams, the components shown are no longer true thermodynamic components since they are mixtures. A crude oil contains hundreds of components, and the brine and surfactant pseudocomponents may also be complex mixtures. The simplified representation, however, has obvious advantages for describing phase behavior, and it is reasonably accurate as long as each pseudocomponent has approximately the same composition in each phase. In Fig. 23.16a for instance, the oil pseudocomponent can appear in an oil-rich phase or in a phase containing mostly surfactant and brine. If the oil solubilized into the surfactantlbrine phase is nearly the same mixture of hydrocarbons as the original oil, then the representation in terms of pseudocomponents is reasonable. The compositions shown in Fig. 23.16 are in volume fractions. An inverse lever rule similar to Eqs. 3 or 5 gives the relationship between the volumes of the two phases for a given overall composition, as illustrated in Fig. 23.16. Fig. 23.16a is a phase diagram for the liquid/liquid equilibrium behavior typical of mixtures of brines of low salinity with oil. If there is no surfactant present, the oil and brine are immiscible; mixture compositions on the base of the diagram split into essentially pure brine in equilibrium with pure oil. Addition of surfactant causes some oil to be solubilized into a microemulsion rich in brine. That phase is in equilibrium with a phase containing nearly pure oil. Thus in the low-salinity brine, the surfactant partitions into the brine phase,
Fig. 23.17-Typical pressure-composition phase diagram for a binary mixture of CO, with a crude oil at temperatures above 120F [49%].
solubilizing some oil. The plait point in Fig. 23.16a lies close to the oil comer of the diagram. Because only two phases occur and the tie lines all have negative slope, such phase is often called Type II( -). Phase diagrams for high-salinity brines are often similar to Fig. 23.16~. In the high-salinity systems the surfactant partitions into the oil phase and solubilizes water into an oil-external microemulsion. In this case the plait point is close to the brine apex on the ternary diagram. For intermediate salinities, the phase behavior can be more complex, as shown in Fig. 23.16b. According to the phase rule, if the temperature and pressure are set, then up to three phases can coexist for a three component system. If three phases do occur, then the compositions of the phases are fixed at a given temperature and pressure. The three-phase region on a ternary diagram is represented as a triangle (Fig. 23.16b). Any overall composition lying within the three-phase region splits into the same three phases. Only the amounts of each phase change as the overall composition varies in the three-phase region. The edges of the three-phase region are tie lines for the associated two-phase regions. Thus, there is a two-phase region adjacent to each of the sides of the three-phase triangle. In Fig. 23.16b, the two-phase region at low surfac.tant concentrations is too small to show on the diagram. It must be present, however, since oil and brine form only two phases in the absence of surfactant.
PHASE DIAGRAMS
23-9
Fig. 23.18-Pressure composition diagram-Gas 1 system for Rangely oil: 95% CO, and 5% methane gas system at 160aF 171Cl.
Phase behavior of COzlcrude oil systems is often summarized in pressure-composition (p-x) diagrams such as those shown in Fig. 23.17. Fig. 23.18 is an example of a p-x diagram for mixtures of CO1 (containing a small amount of methane contamination) with crude oil from the Rangely field. 8 In such diagrams, the behavior of binary mixtures of COZ with a particular oil is reported for a fixed temperature. Thus, the oil is represented as a single pseudocomponent on such a diagram. Such diagrams indicate bubble- and dewpoint pressures, the regions of pressure and composition for which two or more phases exist, and information about the volume fractions of the phases. However, they provide no information about the compositions of the phases in equilibrium. The reason for the absence of composition data is illustrated in Fig. 23. 19,3 which gives data reported by Metcalfe and Yarborough9 for a ternary system of CO*, Cd, and C 10. Binary phase data for the CO*-Cd (Ref. 10) and COz-C tc (Ref. 11) systems also are included. Fig. 23.19 shows a triangular solid within which all possible compositions (mole fractions) of COz-Cd-C to mixtures for pressures between 400 and 2,000 psia are contained. The two-phase region is bounded by a surface that connects the binary phase envelope for the COz-C ra binary pair to that on the CO+4 side ofthe diagram. That surface is divided into two parts-liquid compositions and vapor compositions. Tie lines (heavy dashed lines in Fig. 23.19) connect the compositions of liquid and vapor phases in equilibrium at a fixed pressure. Thus, the ternary phase diagram for CO2-C4-C tc mixtures at any pressure is just a constant
mixtures at 16OOF
pressure (horizontal) slice through the triangular prism. Several such slices at different pressures are shown in Fig. 23.19. At pressures below the critical pressure of CO 2-C 4 mixtures ( 1,184 psia) , both CO 2-C , c mixtures and CO2-C4 mixtures form two phases for some range of COz concentrations. At 400 and 800 psia, the twophase region is a band across the diagram. Above the critical pressure of COz-Cd mixtures, CO2 is miscible with Cd. and ternary slices at higher pressures show a continuous binodal curve on which the locus of liquid compositions meets that of vapor compositions at a plait point. The locus of plait points (labeled P in Fig. 23.19) connects the critical points of the two binary pairs. To see the effect of representing the phase behavior of a ternary system on a pseudobinary diagram, consider a p-x diagram for an oil composed of 70 mol% C ,c and 30 mol% CJ. At any fixed pressure, the mixtures of CO2 and oil which would be investigated in an experiment to determine ap-x diagram lie on a straight line (the dilution line), which connects the original oil composition with the CO2 apex. Thus, a P-X diagram for this
23-10
PETROLEUM ENGINEERING
HANDBOOK
LIQUID
LIQUID
J I
-,I / / /
Fig. 23.20-p-x diagrams for mixtures of CO, with Wasson oil, where L, is Liquid Phase 1 (oil-rich phase), L, is Liquid Phase 2 (COP-rich phase), and V is the vapor phase. Dashed lines indicate constant volume fraction of L, phase.
system is a vertical slice through the triangular prism shown in Fig. 23.19. The saturation pressures on a p-x diagram are those at which the dilution plane intersects the surface which bounds the two-phase region. Bubblepoint pressures (B) occur where the dilution plane intersects the liquid composition side of the two-phase surface, while dewpoint pressures (D) occur at the intersection with vapor compositions. Comparison of the phase envelope on the resulting p-x diagram with binary phase diagrams yields the following observations. 1. Tie lines do not, in general, lie in the dilution plane. Instead, they pierce that plane. This means that the composition of vapor in equilibrium with a bubblepoint mixture on the p-x diagram is not the same as that of the dewpoint mixture at the same pressure. 2. The critical point on the p-x diagram occurs where the locus of plait points pierces the dilution plane. It is not, in general, at the maximum saturation pressure on the p-x diagram. The maximum pressure occurs where the binodal curve is tangent to the dilution plane. The critical point on the p-x diagram can lie on either side of the maximum pressure, depending on the position of locus of plait points on the two-phase surface. It is apparent from Fig. 23.19 that the composition of the original oil has a strong influence on the shape of the saturation pressure curve, and on the location of the critical point on the p-x diagram. If the oil had been richer in C4, the critical pressure and maximum pressure both would have been lower. Thus, it should be anticipated that the appearance of p-x diagrams for C02icrude oil systems should depend on the composition of the oil. Figs. 23.18 and 23.20 illustrate the complexity of
phase behavior observed for COz/crude oil systems. Fig. 23.18 gives the behavior of mixtures of CO2 (with about 5 % methane as a contaminant) with Rangely crude oil at 16OF. The oil itself has a bubblepoint pressure of about 350 psia. Mixtures containing up to about 80 mol% CO;? (+C 1) show bubblepoints, while those containing more CO2 show dewpoints. At the relatively high temperature of the Rangely field, only two phases, a liquid and a vapor, form. At lower temperatures, more complex phase behavior can occur. Figs. 23.20a, b and c show the behavior of mixtures of a dead oil from the Wasson field3 with CO2. At 90F and 105F, the mixtures form a liquid and a vapor at low pressures and two liquid phases at high pressures and high CO2 concentmtions. They form three phases, two liquids and a vapor, for a small range of pressures at high CO* concentrations. The liquid/liquid and liquid/liquid/vapor behavior disappears if the temperature is high enough. At 120F (Fig. 23.2Oc), the three-phase region had disappeared. For the systems studied to date, 120F appears to be a reasonable estimate of the maximum temperature for liquid/liquid/vapor separations. For detailed discussions of such phase behavior, see Refs. 2 and 3.
PHASE DIAGRAMS
23-l 1
The use of K-values, also called equilibrium ratios or equilibrium constants, is based on the behavior of mixtures of gases at relatively low pressures and temperatures. According to Raoults law, the partial vapor pressure phi of component i in a liquid mixture is equal to the product of the mole fraction of component i in the liquid and its pure component vapor pressure p~i=Xjpvj. ... . . f.. . ......... . . (6)
In addition, Daltons law states that the partial pressure of component i in the vapor is p;j =yjpr, .... ......... . . . . . . . . . *. . . . (7)
where yi is the mole fraction of component i in the vapor andp, is the total pressure. Rearrangement of Eqs. 6 and 7 gives the definition of K-value for an ideal (low pressure) system:
K.-Y; I -pvi xi Pr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (8)
PRESSURE,PSIA Fig. 23.21-Typical equilibrium ratios at 22CPF [104%] (dashed lines are the ideal ratios).
Thus, for a multicomponent mixture at low pressure, the equilibrium value can be estimated from the vapor pressure, which is a function of temperature only, and the total pressure. The assumption of an ideal gas in Raoults and Daltons laws is reasonable only if the pressure is below about 50 to 100 psia. * At higher pressures, equilibrium ratios are functions of pressure, temperature, and composition. Fig. 23.21 shows a typical set of equilibrium ratios for a hydrocarbon system containing some CO2 at 220F. I3 Also shown (as dashed lines) are the ideal equilibrium ratios. At high pressures, the K-values for ethane and heavier hydrocarbons pass through a minimum and appear to converge to a value of one, in this case at 4,200 psia. This observation is the basis for a widely used empirical correlation for K-values. K-value charts for a variety of convergence pressures and a recommended technique for estimating the convergence pressure are given in GPSAs Engineering Data Book. If K-values ate known or can be estimated, then amounts of liquid and vapor and phase compositions can be calculated easily. Consider 1 mol of a mixture in which the overall mole fraction of the ith component is zi . If the mole fractions of component i in the liquid and vapor are Xi and yi , and the fraction of the mole of mixture that is liquid is L, then a material balance gives Zi=Xil+y~(l-L). ...... . . . . . . . . . . . (9) into Eq. 9
which
f(L)=
Ky;;;;)L 1
=o. 1
.. .... .. .
. (12)
Eq. 12 can be solved for L by application of a NewtonRaphson iteration. If Lk is the kth estimate of the solution, an improved estimate is given by Lk+l=Lkpi z where Lk ) .....,,,...,........(13)
The iterative calculation is complete when AL) as given by Eq. 12 and AL=Lk+l -Lk are both smaller than some preset tolerances. Once the liquid mole fraction has been determined, the Xi and yi are obtained from Eqs. 10 and 11. If the mixture is at its bubblepoint pressure, then L= 1 and Czi=l, and Eq. 12 reduces to C(ziKi)=l. ...........,.......... . . . . . ..(15)
. . . . . . . . . . . . . . . . . . . . . (10)
Thus, if Kis are known as a function of pressure, then the bubblepoint pressure can be obtained as the pressure at which Eq. 15 is satisfied. Bubblepoint pressures are generally most sensitive to the K-values of the lightest components, which axe the largest. If the mixture is at its dewpoint pressure, then L=O, CZi = 1, and c =l. . . . . . . . . . . . . . . . . . . . . . . . . (16)
.. ..,...... . ....
. -(II)
23-12
PETROLEUM ENGINEERING
HANDBOOK
volume is small enough to be close to the constant h, then the pressure increases rapidly as the volume is reduced. Thus, the EOS is qualitatively consistent with liquid behavior when the pressure is high. The calculation of phase compositions is based on the fact that, at thermodynamic equilibrium, the fugacity of each component must be the same in each phase. The fugacity of a component in a phase can be calculated if the volumetric behavior of the phase is known. It can be shown I4 that the fugacity of component i, fi , in a phase is given by
RTlnfi=Sm[ Vt
(2) t
TV -$]dV, 3 l, t
-RTln
Fig. X3.22-Comparison of calculated and measured phase compositions for ternary mixtures of CO,. methane (C,), and decane (C,,), at 160F [71C] and 1,250 psia.
vt
niRT
.... ...
.. .. .
Dewpoint pressures are most sensitive to the smallest Kvalues, those of the heavy components, which often are least accurately known. Thus, there is often more uncertainty in calculated values of dewpoint pressure. The sums of Eqs. 15 and 16 also are useful for determining whether the mixture forms one or two phases. If C(K;z;) < 1, the mixture is all liquid. If C(ziIKi) < 1, the mixture is all vapor. If both C(K,zi)> 1 and C(K;Iz,) > 1, the mixture forms two phases. In recent years, EOSs also have been used extensively for phase equilibrium calculations. Most of the widely used EOSs are refinements of the equation proposed by van der Waals:
RT p= ---b a v2, ,,........................
where T is the temperature, p the pressure, V, the total volume, ni the number of moles of component i, and R the gas constant. If the relationship between pressure, composition, and total volume is known from an EOS, then taPlaniJT,V ,n can be obtained and the integral evaluated. A var!e<y of EOSs have been suggested for hydrocarbon mixtures. For example, the original Redlich-Kwong equation has the form
n,RT V,-0, 2 ntam
P=
T v (v +nb ) t t ,,,
. . . . .
(20)
b,
where n, =Cni is the total number of moles and a,,, and depend on the mixture composition and the critical properties of the components as follows.
(EyiA; 1/2)2R2T52 p )
am=
......... ....
. . (21)
(17) where
where p is the pressure, R the gas constant, T the temperature, and V the molar volume. The constants a and b can be determined for a particular component from thermodynamic constraints at the critical point, which requires that
(ap/w)~, =(ap%v*)T, =o,
A, = Qa(PIpci) , (T,T,i)5,2
.. ............
. . . (22)
EO.4275.
..................
which gives
a=-
= @Y;BiPT
P
21 R2Tc2
64 PC
, . . . . . . . . . . . . . . . . . . . . . . . (24)
where
B, = nb(P/Pci) (T,T,i) . . . . . . . . . . . . . .
and
b=F, c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (18)
(25)
where T, is the critical temperature and pc is the critical pressure. van der Waals equation reduces to the ideal gas law if the molar volume is large (low pressure). If the molar
PHASE DIAGRAMS
23-13
In Eqs. 21 through 26, y, is the mole fraction of component i in the mixture, and pci and T,; are the critical pressure and temperature of component i. The constants Q, and fib arise from the thermodynamic constraints, =0, at the critical point. From cwa v) Tr =(a2piavl)T, Eqs. 19 through 26, an expression for the fugacity of each component in a phase can be obtained. To calculate phase compositions, the following procedure is used. 1. Estimate compositions of liquid and vapor. 2. Calculate fugacities of each component in each phase. 3. Iff,,, =fiL stop. Otherwise, obtain improved phase compositions and return to Step 2. Similar calculations can be performed for liquid/liquid and even liquid/liquid/vapor systems. Because the equations for fugacities are complex and nonlinear, computer implementation of this iterative scheme to find phase compositions is required. The Redlich-Kwong EOS given as an example here is by no means the only equation available. Many modifications to the Redlich-Kwong equation have been proposed to improve the accuracy of the predictions of and equations with different phase compositions, analytical forms are also in use. The Soave modification of the Redlich-Kwong equation and the Peng-Robinson equation are among the most widely used. Ref. 14 gives details of a variety of EOSs, and Ref. 15 is a useful collection of papers relevant to phase equilibrium calculations for hydrocarbon systems. Computer programs for such calculations are available. * EOSs currently in use are quite accurate for mixtures of light hydrocarbons for which critical properties are known and extensive phase behavior data are available. For instance, Fig. 23.22 shows a comparison of phase compositions calculated with the Peng-Robinson EOS with measured values for mixtures of CO*, C 1, and C 10 at 1,250 psia and 160F. For this well-characterized system, the calculated values agreed well with the measured compositions. For crude oil systems, phase behavior predictions are less reliable because the chamcterization of heavy components is less certain. For such systems some experimental data are required to tune the EOS to represent the particular hydrocarbon system. Improvement of the predictive power of EOSs for complex hydrocarbon systems is an area of active current research.
p = PJ,, = pd = pt = PV = pVi =
p:i = PVZ =
pyl
pv2
T,. = T, = T2 = VL = V, = Zi = Q, = Qb = Subscripts
pressure bubblepoint pressure dewpoint pressure total pressure vapor pressure vapor pressure of Component i in liquid mixture partial vapor pressure of Component i in liquid mixture any mixture of two components which form a single vapor phase pressure below vapor pressure of Component z which may form a single vapor phase pressure above vapor pressure of Component z which may form a single vapor phase critical temperature constant temperature below T, constant temperature above T,. saturated liquid volume saturated vapor volume overall mole fraction of the ith component defined in Eq. 23 defined in Eq. 26
References
1. Engineering Data Book, Gas Processors Suppliers Assn., ninth edition, Tulsa (1972). 2. Stalkup, F.I. Jr.: Miscible Displacement, Monograph Series, SPE, Dallas (1983) 8. 3. Orr, F.M. Jr. and Jensen, C.M.: Interpretation of PressureComposition Phase Diagrams for CO*-Cmde Oil Systems, Sot.
Pet. Eng. J. (Oct. 1984) 485-97. 4. Reamer, H.H., Fiskin, J.M., and Sage, B.H.: Phase Equilibria in Hydrocarbon Systems, Ind. Eng. Chem. 41 (Dec. 1949) 2871. 5. Sage, B.H. and Lacey, W.N.: Thermodynamic Properties of the Lighter Parafin Hydrocarbons and Nitrogen, Monograph on API
6. 7. 8. 9.
Ai = defined by Eq. 22 6, = defined by Eq. 24 Bi = defined by Eq. 25 C = critical point when liquid and vapor phases are identical fir. = liquid fugacity of Component i fiV = vapor fugacity of Component i Ki = K-value of Component i L = total moles of liquid-phase in mixture Lk = kth estimate of L by Newton-Raphson iteration no = number of constraints
Tulsa, OK
10. 11.
12. 13.
Research Project 37, American Petmleum Inst., New York City (1950). Hunt, J.M.: Petroleum Geochemistry and Geology, W.H. Freeman and Co., San Francisco (1979). Nelson, R.C. and Pope, GA.: Phase Relationships in Chemical Flooding, Sot. Pet. Enr. 1. (Oct. 1978) 325-38. Graue, D.J. and Zana, E.T.: Study of a Possible CO, Flood in the Rangely Field, Colorado, J. Per. Tech. (July 1981) 1312-18. Metcalfe, R.S. and Yarbomugh, L.: The Effect of Phase Equilibria on the CO* Displacement Mechanism, Sot. Pet. Eng. J. (Aug. 1979) 242-52; Trans., AIME, 267. Olds, R.H. el al.: Phase Equilibria in Hydmcarbon Systems, Ind. Eng. Chem. 41 (March 1949) 475-82. Reamer, H.H. and Sage, B.H.: Phase Equilibria in Hydrocarbon Systems. Volumetric and Phase Behavior of the n-Decane-CO2 System, J. Chem. Eng. Data 8, No. 4 (1963) 508-13. Standing, M.B.: Volumetric and Phase Behavior of Oil Field . Hydrocarbon Sysrems, SPE, Dallas (1977). Allen, F.H. and Roe, R.P.: Performance Characteristics of a Volumetric Condensate Reservoir, Trans., AIME (1950) 189,
83-90.
14. Reid, R.C., Prausnitz, J.M., and Sherwood, T.K.: The Properties of Gases and Liquids, third edition, McGraw-Hill Book Co. Inc., New York City (1977). 15. Phase Behavior. Reprint Series, SPE, Dallas (1981) 15.
Chapter 24
were rewarded by finding unwanted increased amounts of oil and gas associated with the saline waters. In the Appalachian Mts.. many saline along the crests of anticlines. water springs occurred
ty, PA, that encountered higher than average oil saturation in the lower part of the Bradford sand. This high oil saturation resulted from the action of an unsuspected flood. the existence of which was not known when the location for the test well had been selected. Toward The upper part of the sand was not cored. the end of the
In 1855 it was found that distillation of petroleum produced a light oil that was similar to coal oil and better than whale oil as an illuminant. waters This knowledge oil. Edward PA. in spurred the search for saline that contained
Using the methods of the salt producers, Col. Drake drilled a well on Oil Creek, near Titusville. 1859, He struck well produced about 3.5 B/D.
cutting of the first core with a cable tool core barrel. oil began to come into the hole so fast that it was not necessary to add water for the cutting oil in a hole of the second secwater. Two tion of the sand. Therefore, sand was cut with the lower 3 ft of the Bradford free from
The early oil producers did not realize the significance of the oil and saline waters occurring together. In fact, it was not until 1938 that the existence of interstitial water in oil reservoirs was generally recognized. 4 Torrey was convinced as early as 1928 that dispersed interstitial water existed in oil reservoirs, but his belief was rejected by his colleagues did not produce because any most of the producing upon completion. wells Occurwater
samples from this section were preserved in sealed containers for saturation tests, and both of them, when analyzed, had a water content of about 2O%PV. This well made about IO BOPD and no water after being shot with nitroglycerine. Thus, the evidence developed by the core analysis and the productivity test after completion provided a satisfactory indication of the existence of immobile water, indigenous to the Bradford sand oil reservoir, which was held in its pore system and could not be produced Fettke by conventional pumping methods. was the first to report the presence of water in
oil and gas with water were and Munn,6 but they believed separation of the oil and water, did not occur in the
sand before a well tapped the reservoir. It was not until 1928 that the first commercial laboratory for the analysis of rock cores was established, and the first core tested was from the Bradford (from
Now
an oil-producing sand. However, he thought that it might have been introduced by the drilling process. Munn* recognized might be the primary perimental that moving underground water cause of migration and accumulathis theory Mills had little exconducted several
the Bradford
field.
McKcan
and Energy
County.
Research
The
Eartlesv~lle
was J Wade
laboratory experiments on the effect of moving water and gas on water/oil/gas/sand and water/oil/sand systems. Mills concluded that the up-dip migration of
24-2
PETROLEUM ENGINEERING
HANDBOOK
in water, as well as the migration of oil, either up or down dip, caused by hydraulic currents, are among the primary factors influencing both the accumulation and the recovery of oil and gas. This theory was seriously questioned and contemporaries. buoyancy, completely rejected by many of his
well from a leaky casing or faulty completion and this can be repaired. Enormous quantities of water are produced with the oil in some fields, and it is necessary to separate the oil from the water. Most of the oil can be removed by settling. Often, however, an oil-in-water emulsion forms, which is very difficult and various separation. to break. In such cases, the oil is heated chemicals are added to induce surface-active
Rich I assumed
that hydraulic
currents,
are effective
in causing accumulation
acits retention. He did not believe that the hydraulic cumulation and flushing of oil required a rapid movement of water but rather that the oil was an integral stituent of the rock fluids and that it could along with them whether relatively rapid. The effect of water the movement displacing conbe carried
days, the water was dumped on the it seeped below the land surface. Until waters were disposed into local killing fish and even surface vegeta-
oil during
was not recognized in the early days of the petroleum in dustty in Pennsylvania. Laws were passed, however, to prevent operators from injecting water into the oil reservoir sands through unplugged wells. In spite of these laws, some operators at Bradford secretly opened the well casing opposite shallow groundwater sands to start a watertlood in the oil sands. Effect of artificial watertloods were noted in the Bradford field in 1907, and became evident about 5 years later in the nearby oil calculations of the fields of New York. Volumetric oil-reservoir volume that were made for engineering studies of these waterflood operations proved that inwater was generally present in the oil sands. and Schilthuis the distribution information gave detailed information of water and oil in porous and occurrence concerning of connate of the relationship
producing sand or another deep aquifer. The primary concern in such disposal practice is to remove all oil and basic sediment from the waters before pumping them into injection wells to prevent clogging of the pore spaces in the formation receiving the waste water. Chemical compatibility of waste water and host aquifer water also must be ensured. Waters portance. produced In years with past, petroleum are growing were in imthese waters considered
terstitial
waste and had to be disposed of in some manner. Injection of these waters into the petroleum reservoir rock serves three purposes: it produces additional petroleum (secondary recovery), it utilizes a potential pollutant. and in some areas it controls land subsidence. The volume of water produced with petroleum in the U.S. is large. In 1981 domestic oil production was about 8.6~ IO B/D and the amount of water produced with the oil ranges from 4 to 5 times the oil production. Therefore, 4.5, would the water production, assuming be about 38.7~ IO BID. a factor of
water saturation to formation permeability. The word connate was first used by Lane and Gordon to mean interstitial water that was deposited with and large the sediments. The processes of rock compaction mineral diagenesis result in the expulsion of
amounts of water from sediments and movement out of the deposit through the more permeable rocks. It is therefore highly unlikely that the water now in any pore was there when the particles is the same as that which
Secondary and tertiary oil recovery processes that use water injection usually result in the production of even more water along with the oil. To inject these waters into reservoir moved formations. rocks, suspended injection solids systems and oil plugging require must be refrom the waters to prevent Water of the porous xepardtors,
that surround it were deposited. White redefined connate water as fossil water because it has been out of contact with the atmosphere for an appreciable gcnlogic time period. Thus. connate water part of a is distin-
filters, and, in some areas, deoxygenating and bacteria control equipment with chemical and physical methods to minimize corrosion and plugging in the injection system. In waterflooding volume of produced most petroleum reservoirs, water is not sufficient to rccovcr the the
water, which has entered the cruished from meteoric 2 rocks in geologically recent times, and from juvenile water. which has come from deep in the earths crust and has never been in contact with the atmosphere. Meanwhile. petroleum engineers and geologists had learned that waters associated with petroleum could be identified with regard to the reservoir in which they occurred by a knowledge of their chemical characteristics. Commonlv, the waters from different strata differ considerably In their dissolved chemical constituents. making the identification easy. Howjcvcr. dissolved constituents not ditfcr significantly, waters is difficult The amount of water of a water from a particular in some areas in waters from different and the identification produced with stratutn of strata do of such inthe concentrations
additional petroleum efficiently. Therefore, supplemental water must be added to the petroleum reservoir. The use of waters from the other sources requires that the blending of produced water with supplemental water must yield a chemically stable mixture so that plugging solids will not be formed. For example, a produced water containing considerable calcium should not be mixed with a water containing considerable carbonate because calcium carbonate may precipitate and prevent injection of the tloodwater. The design and successful operation of a secondary or tertiary recovery operation requires a thorough knowledge of the composition of the waters used. are the Chemical analyses of waters produced with oil useful in oil production problems. such as identifying
creases as the amount of oil produced decreases. lfthis is edge water. nothing can be done about it. If it is botton-
24.3
of
Intrusive
water,
watcrfood to prevent
and corro-
tained.
A test that
ilows
water
will sample.
give
even
higher
disposal
projects.
assurance of an uncontaminated
problems
recovery. knowlc$Fc
in primary, secondary, and tertiary rcquircs a Electrical well-lo g interpretation of the dissolved solids concentration and
water sample is taken for analysis. it should bc taken just above the tool. since this is the last water to enter the tool and is least likely to show contamination. Analyses of water obtained from a DST of Smackovcr limestone water in Rains County. TX. show how errors can be caused Analyses by improper sampling of DST water. of top, middle. and bottom samples taken from
composltton of the interstitial water. Such information also is useful in correlation of stratigraphic units and of the aquifers within these movcmcnt of xubsurfacc understand other waters. the processes without minerals units. and in studies of the waters. It is impossible to that accumulate insight into petroleum or the nature of these
show an increase in salinity with indicating that the first water wa\ I) Thus. the bottom samof Smackovcr water.
by mud filtrate.
Sampling
The composition of subsurface water commonly changes aquifer. of other to the sari--
Sample Procedure
No single procedure is universally applicable for obtaining a sample of oilfield water. For cxamplc. inthrmation may bc desired concerning the dissolved gas or hydrocarbons in the water or the reduced species present. such as ferrous or manganous compounds. Sampling procedures applicable to the desired infomlation must be used. Some of the special information and sample or references location cited for
with depth and also laterally in the same Changes may be brought about by the intrusion waters. aquifer. and by discharge from to obtain and recharge It is thus difficult
a representative
pie of a given subsurface body of water. Any one sample is a very small part of the total mass. which may vary MJidely samples solution in composition. Therefore. it is generally Also. the necessary to obtain and analyze may change with time and supersaturated sites should many samples.
be selected, network to
of ccrused in
There is a tendency for some oilfield waters to become more diluted as the oil reservoir is produced. Such dilution may result from the movement of dilute water from adjacent compacting clay beds into the petroleum rescrvoir as pressure declines with the continued removal of oil and brine. The composition tion within of oilfield water varies with the posifrom which it is ob-
method
of obtaining such
gases is to place a
a device. The device is connected to the flowline. and water is allowed to flow into and through the container. which volumes is held above the flowlinc. through. until 10 or more valve of water have flowed The lower
the geologic
structure
tained. In some cases the salinity will increase upstructure to a maximum at the point of oil/water contact. Few of the samples arc truly representative collected by drillstem test (DST) During formation-water samples.
drilling. the pressure in the wellbore is intentionally maintained higher than that in the formations. Filtrate from the drilling mud seeps into the permeable formation-water strata. sample and this filtrate The most truly is the first liquid representative to enter the test tool.
usually is obtained after the oil well has produced for a period of time and all extraneous fluids adjacent to the wellhore mediately taminated tion fluids. have been flushed out. Samples taken imafter the well is completed may be conwith drilling fluids and/or with well complcsuch as filtrate from cement, tracing fluids,
can be used to transfer the sample from the sample valve into the container (usually plastic). The source and sample container should be flushed to remove any foreign material system. before a sample is taken. After flushing the of the the end of the tube is placed in the bottom
container, and several volumes of fluid are displaced bcforc the tube is removed slowly from the container and the container obtalnmg is sealed. Fig. 24.2 illustrates a method of a sample at the wellhead. An extension of this
and acids. which contain many different chemicals. Sampling methods are discussed in publications of the American Petroleum Inst. (API), American Sot. for Testing Corrosion and Materials Engineers (ASTM), (NACE). I8 and the Natl. Assn. of
method is to place the sample container in a larger container. insert the tube to the bottom of the sample container. allow the brine to overflow both containers. and withdraw the tube and cap the sample under the fluid. At pumping wellheads the brine will surge out in heads and be mixed with oil. In such situations a larger con-
Drillstem Test
The DST, if properly made, can provide a reliable formation water sample. it is best to sample the water after each stand of pipe is removed. Normally, the total dissolved solids (TDS) content will increase downward and become constant when pure formation water is ob-
tainer equipped
with a valve
at the bottom
can be used as
device, place the sample tube in the bottom of the large container, open the wcllhead valve, rinse the large container with the well fluid, allow the large container to
24-4
PETROLEUM ENGINEERING
HANDBOOK
system is simple
It convalves.
sists of a SO-mL disposable and an inline-disk-filter size 47-mm diameter, 74 Sample container I-+ option of a prefilter
syringe,
holder. The filter holder takes 0.45pm pore size filters, with the and depth prefilter.
After the oilfield brine is separated from the oil, the brine is drawn from the separator into the syringe. With the syringe, tion bottle. it is forced through the filter into the collecto be used The check valves allow the syringe
as a pump for filling the collection bottle. If the filter becomes clogged, it can be replaced in a few minutes. About 2 minutes are required to collect 250 mL of sample. Usually two samples are taken, with the one being acidified to pH 3 or less with concentrated HCI or HN03. The system can be cleaned easily or flushed with brine to prevent contamination.
Analysis.
research
brines. 22-24
Unstable
Properties
or
was designed
to measure
Oz, resistivity, S=, HCOT, water at the wellhead. When in the field and of the The
are collected
transported to the laboratory for analysis, many unstable constituents change in concentration. amount of change depends on the sampling sample storage, the constituents
method,
ambient conditions, and the amounts of in the original sample. Therefore an at the wellhead is necessary to ob-
Sample Containers.
Containers
that are used include rubber, adsorb metal cans, various ions
polyethylene, other plastics, hard and borosilicate glass. Glass will fill, and withdraw a sample through the valve will at the botserve to ob-
such as iron and manganese, and may contribute boron or silica to the aqueous sample. Plastic and hard rubber containers are not suitable if the sample is to be analyzed to determine its organic content. A metal container is used by some laboratories if the sample is to be analyzed for dissolved hydrocarbons such as benzene. The type of container selected depends on the planned use of the analytical data. Probably the more satisfactory container, if the sample is to be stored for some time
tom of the large container. This method tain samples that are relatively oil-free.
Fig. 24.2-Example of the method used for obtaining a sample at the wellhead.
Fig. 24.3-Example
24-5
TABLE 24.1-DESCRIPTION
NEEDED FOR EACH OILFIELD WATER SAMPLE Field Well No. ~ Range Operator Date Representing in the
Sample Number Farm or lease ~ Townshlp of Section State County Operators address (main office) Sample obtained by Address Sample obtained from (lead line, separatory flow tank, etc.)
Completion date of well Name of productive zone from which sample is produced Shale Lime Sand Other Names of formations Name of productive well passes through formation Bottom of formation Depths: Top of formation Top of producing zone Bottom of producing zone Present depth Top of depth drilled Bottomhole pressure and date of pressure Bottomhole temperature Are any chemicals added to treat well Date of last workover If yes, what? Initial Present Well production Casing service record, Oil, BID Water, B/D Gas, cu ft/D Method of production (primary, secondary, or tertiary) Remarks: (such as casing leaks, communication or other pay in same well, lease or field)
before
analysis.
is the
polyethylene
bottle.
Not
all
TABLE 24.2-GEOCHEMICAL
Produced Water PH Eh Speciilc resistwty Speciitc gravity Bacteria Barium Bicarbonate Boron Bromide X 0 X X 0 X
WATER ANALYSES*
Steam Injection Generation Water Water X Disposal Water
polyethylenes are satisfactory because some contain relatively high amounts of metal contributed by catalysts in their manufacture. The approximate metal content of the plastic can bc determined by a qualitative emission
spectrographic technique. If the sample is transported during freezing temperatures, the plastic container is less likely to break than is glass.
X :: 0 X :: X 0 0 X :: 0 0 X 0 0 X solIds X X ; X
Calcium
Carbonate Carbon Chlonde Hydrogen Iodine IlOll Magnesium Manganese Oxygen Potassium Residual hydrocarbons Sodium Silica Strontium sulfide dioxide
X :: X 0
X :: 0
Sulfate
Suspended Total dissolved solids X = usually requesm O=somellmes requested
the analyses
for various
proper-
ties or constituents of oilfield water. Methods to determine most of these properties or constltucnts can bc found in Refs. 16, 17, and 25 through 30.
24-6
PETROLEUM ENGINEERING
HANDBOOK
TABLE 24.3-CHARACTERISTICS
FIELDS
Constituents (mg/L) Ca Mg Na K
System
Formation
Devonian-Silurian Mississippian -
8 7 8 10 12
Blue Lick Sub Trenton Second Water Big Lime First Water Big Lime Berea
1,843 to 3,263 3,820 to 5,815 2,175 to 3,270 5,175 to 5,300 401 to 1,592
1,390 9,230 11,000 44,000 32,300 51,200 25,900 29,600 4,600 11,900
650 2,900 2,700 6,600 5,180 10,200 4.100 1O;OOO 1,500 3,000
10,500 33,600 39,500 58,600 36,000 60,700 21.600 861400 25,000 43,900
West Virginia39 29 6 21 44 43 Mtssisstppian Mississippian Mississippian Pennsylvanian Devonian Big lnjun Squaw Maxton Salt Sand Oriskany 1,390 to 3,215 1,908 to 2,019 1,287 to 3,259 450 to 1,960 3,036 to 8,089 30 1,730 630 8,920 100 15,300 400 20,600 2,500 33,600 300 3,910 200 2,250 40 2,740 340 2,650 480 3,800 50 52,200 6,300 38,100 3,800 35,100 2,500 50,900 34,000 98,300 IO 750 290 340 30 660 3,6:: 200 6,900
However. few of
usually
in only
properties.
is understandable
basin, a large area not generally otherwise identifiable. This division has been made arbitrarily for convenience and because of the lack of a uniform system and is not intended as a precedent for any system of classification. The were states or provinces are listed from which reliable analyses tables sources available alphabetically in the indicated information.
because oil producers wish to spend the least amount of money possible. Therefore, they will look at only the properties that are necessary to evaluate any treatment of the for rcinjection to recover oilfield waters. more oil or to dispose
to the original
Composition
The composition
of Oilfield Waters
, ,, ot otltteld waters varies from relatively Several thousand oilfield files. of pro brines.
water analyses are available on computerized Tables 24.3 through 24. I4 show characteristics
area was the first was produced commerand studied geologic gives from the Apfrom to the of the
duced waters. and much of the text was taken from the 1962 edition of this book. The tabulated data on water analyses lihted alphabetically in order of general areas of the U.S.. Canada. and Venezuela. the smaller subdivisions An of basins. exception or gcosynclines. to this following arc oil-productive rather than by provinces. Illinois is the
features of North America. Table 24.3 characteristics of some waters produced palachian fields. F--~, and associated water Petroleum Permian. although
are produced
geologic
of the productive
are productive.
PROPERTIESOF
PRODUCED WATERS
24-7
TABLE24.3-CHARACTERISTICSOF
SOMEWATERSPRODUCEDFROMAPPALACHIANFIELDS Specific Gravity 60/600 1.022 1.120 1.036 1.070 1.020 1.039
(continued)
(mg/L) Cl 19,600 93,900 31,700 61,000 14,000 26.000 I Trace 10 ND ND ND ND Br 120 820 ND ND ND ND
18,200 77,600 113,500 189,400 113,000 216,300 114,200 193.100 52,700 93,400
0 IO 10 30 ND ND ND ND ND ND
1.025 1.089 1.150 1.224 1.151 1.240 1.125 1.211 1.063 1.115
31,030 125,180 167,030 304,020 189,100 344,110 167,540 324,350 84,260 154,820
30 560 0 0 0 40
1.001 1.149 1.010 1.101 1.007 1.115 1.004 1.159 1.059 1.219
475 191,580 18,832 132,110 9,825 148,090 5,810 206,430 51,552 318,630
sandstones
are nonuniform
and discontinuous.
although
the Big In.jun and Berea sands have been traced across wide areas. The oil-producing states included in the Appalachian area from which analyses were available arc Kentucky, Ohio. Pennsylvania. and West Virginia. The Gth than concentrations ofdissolved salts in waters produced petroleum range from a few hundred to more
U.S. Gulf Coast. For many years since the Spindletop dome was discovered in 1901, copious quantities of oil have been produced mations on the flanks, from Tertiary and Quaternary forin the caprock, and in structures
abovle the capmck of massive salt domes. usually considered intrusive in nature. During recent years. offshore drilling has focused attention on drilling oft the coasts of Louisiana and Texas. Some waters produced from gulf coast fields are quite fresh; others have concentrations of dissolved salts as high as 170.000 nngit, (Table 24.5). 4L44
300,000 IllgiL.
California. In different
duced principal from many Cretaceous
fields of California. oil is proreservoirs, ranging in age from Sandstones Many and sands are the arc of the formations
and much folding and faulting are mineralized water produced with reservoirs is by no means as
California
is produced Pennsylvanian
concentrated as that from reservoirs in many other areas. especially the midcontincnt. Table 24.4 gives the characteristics fields, JO.41 of some water produced from California
principally
Mississippian sandstones and. to a smaller extent. limestones. TDS in the produced waters range about I.000 to more than 160,000 mg/L (Table
24.6).j5
24-a
PETROLEUM ENGINEERING
HANDBOOK
TABLE 24.4-CHARACTERISTICS
FIELDS
Number of Analyses 17 10 5 4 2 26 18
Formation Coalinga Midway Sunset Kern River Lost Hills Maricopa Zone A
TDS Ca --z390 20 2,890 60 1,280 20 10 200 220 200 2,900 10 80 Mg 10 340 IO 690 20 570 20 10 140 230 10 1,300 3 140 Na 40 3,290 910 13,250 3,650 11,650 1,550 50 4,770 7,640 1,300 15,015 2,050 7,090 a 50 480 0 360 0 90 390 0 150 460 0 510 0 340 HCO, 180 360 180 360 50 4,270 7: 0 0 4 1,020 1,700 3,900
so,
Cl 90 2,520 1,010 23,550 4,360 21,420 10 60 7,740 11,950 1,170 27,100 1,300 9,560
FwU
580 14,640 2,140 42,120 8,145 39,320 80 2,130 13,020 21,120 2,686 47,995 5,064 21,200
400 to 3,000 -
'Upper ligure in each column IS m~n~murn value and lower figure IS maximum
Midcontinent
associated mian,
waters
from
Jurassic, rocks.
PerProare of the
Pennsylvanian,
Mississippian
of this section,
sidered to include Arkansas, Kansas. northern Louisiana, Missouri, Nebraska, Oklahoma. and all of Texas except the gulf coast fields. Oil and associated brines are produced from many sandstones and limestones, as well as from other types of formations, in geologic systems ranging from the Cambrian through the Upper Crctaceous. Waters produced with petroleum from midcontinent fields have a wide range of concentration of dissolved salts, from little more than 1,000 to more than 350.000 mg/L. Tables 24.7 through 24.9 present the characteristics of some produced waters from the midcontinent Oklahoma, and Texas.36- fields of Kansas,
low concentrations of dissolved salts and often characterized by comparatively high concentrations bicarbonate. Tables 24.10 and 24.11 give
characteristics of some waters produced from Rocky Mt. fields of Colorado, Montana, and Wyoming. 55m5y
oil-productive areas in Canada are the lower Ontario Peninsula, where oil is produced from rocks ranging from Ordovician to Devonian age, the western provinces, principally Alberta, Saskatchewan, and the Northwest Territories. Reservoir rocks in western Canada range in age from Devonian to
and
Cretaceous. Although many of the waters produced with petroleum have quite low concentrations of dissolved in Colorado, salts, others are quite concentrated. Tables 24.12 and 24. I3 present the characteristics of some waters from Canadian fields in Alberta, Manitoba, and Saskatchewan, hObh5
is produced
Montana. New Mexico, Utah, and Wyoming from many fields in the Rocky Mt. area. The principal production is from rocks of the Cretaceous system, although oil and
TABLE 24.5-CHARACTERISTICS
Constituents
Ca
(mg/L)
Number 42 5 6 6
Analyses'
FormatIon or field
Mg
50
1,000 10 30
Na
HCO,
30 990 230 770 70 400
so, 0
110 3 160 120 210 Trace 1.750 270 3.010
Cl 3.180 69.100
Coastal Creek
40610
1 100
2.240 40.600 60
1.330 3,800 18,200 3.600 61.000 6.700 40.800 340 30 4.460 12.730
Goose Humble
20 2.130 6.300
33.700 6.100 105000 11 300 63400 110 610 6.700 21 600
353
4 500 10,860 54480 10.470 171.300 18.900 109.990
5
4 6
Oligocene Pliocene-Miocene
70 550 30 240
16 230 10 210
570 1.710
11.490 36400
24-9
TABLE 24.6-CHARACTERISTICS
Number 12 18 57 17 134
Depth (N 1.994 2 437 1.125 2,596 1.045 2.960 672 to 4,000 1.104 to 3.519
Analyses
so, 0
1 620 0 980 10
Cl 38 300 56 700 700 76 000 25 800 83 200 200 82 400 14000 95 400
i 28 590
31 140 143.940
3.840 30
1.350 30 2.990
680
135870 25 600 167.940
8.740
47.660 Indicated
Upper
figure
in each
column
IS minimum
value
IS ma~~rnum
value
for number
The
principal
productive
formations
in
oilfield
waters
are sodium,
calcium,
and magnesium.
are Tertiary sandstones and Cretaceous In general. the various waters produced with concentrations of dissolved salts
The concentrations
of these ions can range from less than and from less than I .OOO mg/L me/L for calcium and/or
Inorganic Constituents
Petroleum determine The major companies their major constituents often analyze oilfield waters constituents. calcium, The idento dissolved usually inorganic
Other cations that often are present in oilfield waters in concentrations greater than 10 mg/L are potassium, strontium, contain aluminum,
and
lithium,
and
barium. in iron,
Some of
oilfield IO
waters of silicon,
concentrations ammonium,
excess
mg/L
lead, manganese,
are sodium,
zinc, 26.70.71
magnesium, chloride, bicarbonate, analytical data are used in studies tification. log evaluation, tal impact, geochemical valuable Cations The presence of various cations minerals. 26
Anions
The major anion chloride in most oilfield waters is chloride. less than The concentration can range from 10,000
and anions
in oilfield
to more than 200,000 mg/L. There are exceptions to some Venezuelan oilfield waters contain this-e.g., more bicarbonate than chloride. Most oilfield waters contain bromide and iodide. The concentrations of these anions range from less than 50 to for bromide and from less than IO
waters can cause solubility, acidity, and redox (Eh) potential changes as well as the precipitation and adsotption of some constituents. The major cations in most
TABLE 24.7-CHARACTERISTICS
FIELDS (KANSAS)
Speclflc ~ I 2. 15 Trace 3 0 Trace 0 10 2 10 1 60 0 Trace 2 3 0 2 0 0 0 0 NO ND Br -~ 30 400 80 x50Trace 60 5 90 20 200 2 3 200 400 10 420 10 70 5 10 0 0 ND ND Gravely (60160~) 1040 1 159 1015 1140 ,014 1 091 1012 1116 1016 1 141 ,017 1140 1023 1 105 1 088 1143 1007 1075 1019 ,045 1033 1 139 ,021 1088 TDS (mg,L) 53.959 256.830 28.120 369.180 20.180 145.060 6,455 160.740 20.782 224.870 24.363 201.153 33.850 116.660 70,902 222,383 9.410 113.460 19.265 58,940 45.350 221.900 26.810 126.930
System Pennsylvanian Ordovlclan Ordowaan Ordoviaan Pennsylvania Mississippian Basal Pennsylvaman Kansas WllCOX
Depth tft) 1.228 lo 3.409 3.500 2.750 2091 to 3 800 lo 3 770 lo 4 14,
Pennsylanlan Cambraan
Upper
fqure
I each
column
IS mlnimum
hgure
IS maximum
of analyses
Ind~caled
24-10
PETROLEUM
ENGINEERING
HANDBOOK
TABLE 24.8-CHARACTERISTICS
FIELDS (OKLAHOMA)
Speclflc _~
Number 75 94 25 28 I9 12 22 18 I7 10 22 22 22 27 12 20 13 15 24 15 14
of System PennsylvanEln Ordovlclan Pennsyfvanlan Ordovua" Pennsylvanian Pennsylvanian M~ss~ss~pp~an Mlssisslpplan Pennsylvaman Ordowan Pennsylvania" Formaflon Bartlesvllle WllCOX Layton Atbuckle Cromwell Burgess Mississippi Mlsener Pennsylvanian Slmpso" Skinner Booth Hunton Red Fork VIola Prue Healdion Tonkawa Burbank Dutcher Bromide
Depth ut1 4.489 to 5,524 3,436 to 7.233 1,240 10 4.800 542 to 6.094 1.48070 5430 1,800 to 2,490 1,837 lo 4.872 3.927 10 5.977 1.258 to 6.025 1.213 lo 6.495 1,030 to 4.567 1.876 to 2 300 3.197 lo 5,021 2.403 to 4.650 3.458 to 5.004 2.267 to 3.587 982 to 3.163 2 417 to 3,254 790 to 5.000 1.882 to 3.218 2.173 to 7,569
Analyses
Ba ~--~ HCO,
0 730 0 130 0 380 0 110 1 20 1 200 2 30 0 10 10 280 0 2 0 IO 0 240 0 170 5 220 0 10 5 20 0 50 0 5 2 40 0 10 10 450 0
SO,
0 I390 0 720 0 510 0 1.880 0 1.130 0
Cl
24 100 144.000 91.300 163.000 34,900 160.000 33,000 127.000 65,000 113.500 81.600 115.000 84.200 157.000 55,400 156,000 29.800 121.000 50.900 140.000 64,100 139.000 90,000 117,000 8,200 142,000 101.000 149,000 4,400 122,000 86.300 142.000 18.600 63,600 132,800 156.200 99,300 149.000 45,500 108.000 19,500 167,000
Grady (60/600) 1031 1 175 1 103 1170 1075 1179 1.034 1 147 1073 1 130 1.091 I 129 1095 1 173 1.066 1 173 1.039 1 134 1059 1.159 1075 1157 1.103 1.131 1012 1155 1115 1 164 1005 1137 1 110 1 158 1022 1076 1 160 1 171 1 109 1 163 1073 1 122 1024 1 183
TDS (mg/L) 39.010 251 460 147.820 266.010 73.310 263.170 50,100 214.140 105 701 182,660 132,016 189,120 136.212 254440 90,690 250,640 49.730 204.320 82.100 233.042 103,540 228.890 141.050 189.760 11.995 258.948 155.208 243.660 7.600 204.330 139.885 230,320 30.392 102,170 172.930 253,525 155,235 241.930 86.900 179,500 32,110 275,270
300 20
160 10 80 0 850 0 310 15 120 10 80 30 110 20 90 0 110 IO 130 5 140 15 660 3 170 40 940 50 120 20 380 0 50 20 133 50 130
200
24 430 60 1.920 0 2,750 0 440 30 450 0 680 0 7.010 0 370 0 980 30 480 2 40 130 370 15 260 40 760 0 920
TABLE 24.9-CHARACTERISTICS
FIELDS (TEXAS)
Specific Grawty (60~160~)
ND ND ND ND ND ND ND ND
_ so4
1
Formation
Mg
5 2.450 2,440 2.860 2,030 2,440 370 600
Na
530 48.200 58,300 66,800 52,500 61,000 32.100 62,900
HCO,
Cl
460 97,900 122,200 139,800 106,000 119,000 57.500 112,500
TDS (m9Q
1,017 160,550 197,640 226.680 171,360 196,990 93,450 184.680
Upper Cretaceous
Pennsylvanian Pennsylvanian
Cambro-Ordovcian
160 3,000 640 2,300 350 2,850 810 3,500 310 7,900 200 3,700
6.030 60,400 15,700 57,100 12,000 55.700 25,500 74,300 4.400 67.000 210 122,500
0
180 10 650 4 1,840 2 710 210 1.840 0 0
10.000 134,000 31,400 109,500 25,300 130,500 82,900 161,800 19,000 140,500 890 212,000
1.015 1.157 1044 1 145 1.035 1 173 1.105 1212 1.033 1 154
16,700 221,080 50,010 188,390 39,374 214,330 112,414 262,320 25,010 241,940 1,710 356.600
ND ND
PROPERTIES
OF
PRODUCED
WATERS
24-11
TABLE 24.10-CHARACTERISTICS
OF SOME WATERS PRODUCED FROM ROCKY MOUNTAIN FIELDS (COLORADO AND MONTANA)
Subsurface Constrtuents (mg/L) TDS NC3 Cl
Number
of System
Depth (ft)
Analyses Colorado5ss6 7 6 6 4 3
Ca 0
1,180 0 190 30 900 0 80 0 380
Mg
0
0 70 40 410 0 30 0 80
co3
0 160 0 240 0 150 0 120 0 0
HCO3
210 3,600 340 4.900 120 2,000 540 3,350 200 3,030
so,
40 890 0 90 20 870 160 980 0 1,040
ml/L)
2.819 to 5.830 1,230 IO 3.464 2,230 to 5.283 3,020 to 4.395 4,564 lo 6,263
310 13,000 820 8,200 1,800 10,600 1,400 3,600 1,070 5,250
560 41,220 1,980 26.490 3.990 33,830 2.360 13,160 1,530 17.840
Montana5s-s7
9
10 11 55 22 25
3.900
220 710 6,200 260 4.670 1.110 3,140 30 1,390 20 2,330
0 0 0 0 0 0 0 0 0 0 0 0
140 2,000 260 1,400 500 4,900 1,670 4,040 150 400 220 4,830
4,050 9,770 1,250 16,900 790 16.010 3,150 11,060 1.560 8,470 250 12,990
2,790
TABLE 24.11-CHARACTERISTICS
Number 24 35 45 50 14 22 24 5 60 20 50 19 20
of System Cretaceous Cretaceous Cretaceous Cretaceous Jurassrc Jurasstc Jurasstc Jurassrc Jurassic Permian Pennsylvaman Mississippian Triassic Formatron Shannon Frontier First Wall Creek Second Cleverly Dakota Dakota Greybull Sundance Embar Tensleep Madtson Mmnelusa Wall Creek
Analvses
Mg
10
330 trace 130 trace 100 trace 10 trace 20 trace 160 trace 60 trace trace 0 60 30 220 10 250 trace 180 50 60
CO, trace 230 trace 1,050 trace 320 trace 590 trace 300 trace 280 trace 380 0 60 0 330 0 0 0 10 trace 20 0 0
HCO, 280 1,900 1,270 7,800 1,000 5.460 890 6.950 110 2,290 230 6,900 1.000 3.680 480 980 410 6.850 210 1.690 30 1,000 20 1,080 190 550
so, 0
3,710 trace 240 trace 60 trace 880 0 110 20 980 trace 60 60 820 40 5,880 190 5,790 10 2,500 50 1,940 1,930 3,870 20
Cl 7,670 70 27,900 220 5,940 1,170 6,600 150 7,590 20 19,200 110 1,930 40 90 140 7,700 10 3,930 3 1,080 4 1,070 250 610
@WLl ~ 730 19.650 1,890 57.340 1,420 17.230 3,800 22.070 1,300 16,630 450 40,750 1,740 11.730 760 2.420 1.110 28,020 620 17,430 98 6,350 114 5.740 3,300 7,210
410 5,560 550 20,000 200 5,320 1,740 7,000 1,040 6,210 180 13,000 630 5.560 180 430 520 6,800 140 5,170 5 790 20 580 630 1,670
24-12
PETROLEUM
ENGINEERING
HANDBOOK
TABLE 24.12-CHARACTERISTICS
Number of AllalySeS
FIELDS
Specllic GLWy TDS ~60/60l (mg/L)
System
Formalton
Br
10 50 70 620 5 1.250 -
10 10 20 230 60 190 -
0 80 0 60 0 410
1.570 to 3 323 2.200 3.000 to 2 942 lo 3 422 980 67 340 240 2.000 550 1 400 850 44.900 4.900 81.400 21.300 72.800 0 40 0 30 0 80 100 2.140 110 360 80 780 1 000 4,600 900 4.900 3900 4.300 850 94,900 7.000 149.600 34 900 120 700 740 31 200 530 8.800 7.900 173.500 14.300 154 900
ND NO 10 40 ND ND 0 20 2 20 3 10 2 20
~ ~ -
1010 ,089 1016 1157 1031 1136 1 004 1 033 1 002 1 025 1 025 1180 1 026 1176
3.940 150 380 14.150 248.990 62 930 203 880 3 150 25 680 1 840 25 780 16.120 290.070 26 760 264,300
to
more
in determining and
constituent.
as
_. _. (la)
Other anions found borate, carbonate, salts, and phosphates. 100 mg/L
arsenate, acid
I (',,. = -This sechon. except
organic 26
av
-
in excess of
v ( ap >
for the pH and Eh, was writlen by Howard B Bradley
TABLE 24.13-CHARACTERISTICS
Number 27 of System Cretaceous
Constituents (mg/L) Na
Analyses
CO,
80 110 mdlcafed
HCO, 190 1,300 190 300 140 350 60 440 200 2,350 100 860 120 850 480 600 70 2,600 40 1,580 290 1,320 65
trace ~ trace 2,200x---870 100 230 220 370 130 3,100 190 7.150 170 900 trace 1,600 90 610 trace 100 0 1,200 trace 160 20,300 8,800 t2,400 11,700 t2.900 760 73.700 1,000 73,000 940 17.700 4,300 71,000 1,400 27,000 1,100 9,300 0 69,100 4,300 10,900 of analyses 170 1,850 470 620 100 7.100 740
5
5 25 12 9 11 4 11 Devonian Devonian MissIssippIan Mississippian Lower Cretaceous
38,900 890
t 3,800 14,500 20,100 280 155,000 640 142,800 700 31,100 5,700 123,800 580 45,700 2,100 12,700 4,800 111.000 2,800 15,400
14,100 680 9,000 trace 5,600 730 2.800 Viking Devonian Jurassfc shale 2,395 to 3,026 3,356 to 6,605 3,105 to 4,325 trace 190 Devonian Jurassic 0 1,100 trace 8,100 Upper ftgure tn each column IS mmmum value and lower figure is maxmum
8 a
PROPERTIES
OF
PRODUCED
WATERS
24-l 3
TABLE 24.14-CHARACTERISTICS
Number of System Tertiary Tertiary Tertiary Cretaceous Tertiary Tertiary Tertiary Cretaceous Formation Zeta (Quiriquire) Eta (Quiriquire) Cabtmas La Rosa 7 a 8 7 8 field, formation
FIELDS
TDS SO, 4 10 5 30 0 0 5 0 0 0 0 0 0 140 60 130 0 150 10 trace 20 0 100 0 Cl 1,910 5,420 710 11 ,170 1,780 90 3,700 690 6,250 8,550 3,450 1,260 9,000 640 560 4,230 5,500 10,500 12,090 9,260 10,900 11,600 13,050 19,800 (mglL) 7,190 16,260 6,900 36,500 5,643 5,260 14,657 6.210 12,955 15.911 7,320 5,460 20,640 4,424 4,830 8,520 11,030 22,260 22,690 21,240 20,860 20,590 23,860 33,820
Analyses 5 7 6
50
300 60 60 60
Lagunillas field, lceota formatlon Bachaquero Pueblo Mene Grande field, Viejo main field, Mason-Trujillo range field. sands
formation field, ftelds 50 40 40 40 140 70 160 110 140 330 940 and lower figure IS maximum value for number 30 150
Du and Eu sands F, sand H sand L, sand M sand P sand S sand U sand Upper tlgure I each column 1s minimum value
ot analyses
or 1 T,=v ( PI--P2
01
. . . . . . . (lb)
solution on compressibility of water with NaCl concentrations up to 200 g/cm3 is essentially negligible. Osifs results show no effect at gas/water ratios (GWRs) of 13 scf/bbl, certainly at GWRs no more of 35 scf/bbl than a 5% probably increase no effect, in the but com-
74 of water
compressibility of compressibili-
resulted
ty vs. pressure.
m r , and intercepts
where CkV = water compressibility and temperature, at the given bbl/bbl-psi, pressure
The plots of l/c, vs. p have a slope of linear in salinity and temperature. tested containing no gas in
within the -cw = average water compressibility given pressure and temperature interval, bbl/bbl-psi, V = water volume water bbl, at conditions 1 and 2 with p r >pz, at the given bbl, volume within p and T interpressure and temperature, V = average vals,
l/c~,=m~p+m~C+m~T+m4,
where cw = water compressibility, p = pressure, psi, C = salinity, = 7.033, and g/L psi - ,
(2)
of solution, F,
T = temperature,
ml m2 = 541.5, lfl3 = -531,
B,,, and B4 = water FVF p I and ~2, bbl/bbl, and B,. = average water FVF corresponding to V,
bbhbbl. In an oil reservoir, water compressibility also depends on the salinity. In contrast to the literature, laboratory measurements by Osif 74 show that the effect of gas in
m4 = 403.3 X 103.
Eq. 2 was fit for pressures psi, salinities between 1,000 and 20,000
of 0 to 200 g NaClIL,
and temperatures
from 200 to 270F. Compressibilities were independent of dissolved gas. Where conditions overlap, the agreement with the results reported by both Dorsey 75 and Dotson and Standing 76 is very good. Results from the Rowe and Chou
24-14
PETROLEUM
ENGINEERING
HANDBOOK
NaCl
solutions
at
14.7
psia
vs.
equation agree well up to 5.000 psi (their upper pressure limit) but result in larger deviations with increasing pressure. In almost all cases, the Rowe and Chou compressibilities are less than that of Eq. 2.
be estimated
from
the
formation
resistivity
(obtained
from electric log measurements) by use of Fig. 49.3 (see Chap. 49). The density of formation water at reservoir conditions can be calculated in four steps. at atmospheric percent NaCl I. Using the temperature and density pressure, obtain the equivalent weight from Fig. 24.5. 2. Assuming mains constant. voir temperature 3. Knowing
Density
The density temperature. most of formation water is a function of pressure, and dissolved constituents. It is determined
in the laboratory on a representative water densample of formation water. I7 The formation sity is defined as the mass of the formation water per unit volume posts, equal to specific ing calculations interchangeable. of the formation density in metric water. For engineering purunits (g/cm) is considered
accurately
the equivalent weight percent NaCl reextrapolate the weight percent to reserand read the new density. the density at atmospheric pressure and
reservoir temperature, use Fig. 24.6 to find the increase in specific gravity (density) when compressed to reservoir pressure. blepoint, the Note that for oil reservoirs saturated-with-gas below the bubbe curves should
When laboratory data are not available, the density of fomration water at reservoir conditions can be estimated (usually to within & 10%) from correlations (Figs. 24.4 through 24.6). The only field data necessary are the denconditions, which can be obtained The salt content from can by use of Fig. 24.4. sity at standard the salt content
used; for water considered to have no solution gas, the no-gas-in-solution curves should be used. These curves were computed from data given by Ashby and Hawkins. 4. The density conditions and 24.6. They of formation water (g/cm) at reservoir Figs. 24.5 is the sum of the values read frotn can be added directly
24-15
to the common
density
base of water (1
g/cm3). The metric units can be changed to customary units (1 bmicu ft) by multiplying by 62.37. Also the specific gravity of formation water can be estimated if the dissolved solids are known. The equation is y,*>=1+c,~xo.695x10-6, where Csd is the concentration . . . . . I.. of . .(3) solids
dissolved
calculations,
the reader is
referred to a recent paper by Rogers and Pitzer. 79 They tabulated a large number of values of compressibility, expansivity and specific volume vs. molality , temperature, and pressure. A semiempirical equation of the same type found to be effective in describing thermal properties of NaCl (0.1 to 5 molality) was used to reproduce the volumetric data from 0 to 300C and I to 1,000 bars.
is reduced.
Fig. 24.7 is
a typical plot of water FVF as a function of pressure. As the pressure is decreased to the bubblepoint, ph. the FVF increases as the liquid expands. At pressures below the bubblepoint. gas is liberated, but in most cases the FVF still will increase from because the shrinkage is insufficient of the water to counterof resulting gas liberation
balance the expansion of the liquid. This is the effect the small solubility of natural gas in water.
The most accurate method of obtaining the FVF is from laboratory data. It also can be calculated from density correlations accounted if the effects of solution gas have been equation is used in the for properly. The following
H,,=L,.. VW
where
. . P r
(4)
V,. = volume
occupied
at pb PRESSURE, PSI A
conditions in water
or standard
V,,. = volume
p,(.
prc.
occupied
The density correlations and the methods P,,~ and prc. were described previously.
of estimating
24-16
PETROLEUM
ENGINEERING
HANDBOOK
The FVF of water can be less than one if the increase in volume resulting from dissolved gas is not great enough to overcome the decrease in volume caused by increased pressure. The value of FVF is seldom higher than I .06.
Figs. 24.X through 24. IO may be used to approximate water viscosity for engineering purposes. These figures show the effects of pressure, temperature, and NaCl content on the viscosity of water. They may be used when the primary contaminant is sodium chloride. Some engineers assume that reservoir brine viscosity is equal to that of distilled water at atmospheric pressure and reservoir the viscosity encountered). temperature. In this case it is assumed that of brine is essentially independent of premise for the pressure ranges usually
Resistivity
The resistivity of formation directly water is a measure of the It resistance offered can be measured by the water to an electrical or calculated. current.
The direct-
pressure (a valid
measurement method is essentially the electrical resistance through a 1-m cross-sectional area of I m7 of formation water, The fomlation water resistivity,
The pH
The
of for-
pH of oilfield
waters
usually
is controlled
by the
mation water is used in electric log interpretation and for such use the value is adjusted to formation temperature. i (See Chap. 49 for more information).
COfibicarbonate system. Because the solubility of CO? is directly proportional to temperature and prcssurc, the pH measurement should be made in the field if a closeto-natural-conditions value is desired. The pH of the water is not used for water but it does indicate identification possible or correlation scale-forming or purposes.
of a water. The pH also may inof drilling-mud filtrate or wellbrines usually is less than 7.0.
and the pH will rise during laboratory storage. indicating that the pH of the water in the reservoir probably is appreciably lower than many published ion to sodium chloride values. solutions Addition will of raise the carbonate
Flooding, describing Low-1FT Phase Behavior and IFT Flooding section). is measured in the laboratory
the pH. If calcium is present, calcium carbonate precipitates. The reason the pH of most oilficld waters rises during storage in the laboratory is because of the fomlation of carbonate ions as a result of bicarbonate decomposition.
siometer. by the drop method, or by other methods. Descriptions of these methods arc found in most physical chemistry texts.
Viscosity
The viscosity pressure. of formation water, p,, , is a function solids. of temperature. and dissolved In gcncral,
in volts.
brine viscosity increases with increasing prcsaure, inand decreasing tempcraturc. creasing salinity. Dissolved gas in the fomlation water at reservoir conditions generally There results is in a negligible little information effect on on hater the actual viscosity.
equilibrium it is related to the proportions of oxidized and reduced species present. Standard equations of chemical thermodynamics of the redox express potential the relationships. is useful iron. in studies sulfur. depends and The on Knowledge other minerals
numerical cffcct of dissolved gas on water viscosity. Gas in solution behaves entirely differently from gas in hydrocarbons. * In water the presence of the gas actually causes the water molecules to interact with each other more strongly, thus increasing the rigidity of the water. However. this effect is very not been literature measured there is to date. an enormous amount and viscosity small and has chemistry of indirect the
such as uranium.
are transported
systems.
of some elements
the redox potential and the pH of their environment. Some water associated with petroleum is interstitial (connate) proved water, and has a negative field studies. Eh: this has been of the Eh is in various Knowledge
In the physical
evidence to support this concept. For the best estimation of the viscosity reader is referred relating equations involve 32 parameters
of water.
useful in determining how to treat a water before it is rein.jected into a subsurface formation. For example. the Eh of the water will be oxidizing if the water is open to the atmosphere, but if it is kept in a closed system in an oil-production operation the Eh should not change appreciably as it is brought to the surface and then reinjetted. In such a situation. the Eh value is useful in determining deposit how much iron in the wellbore. will stay in solution and not of
Their cor-
for calculating
the numerical effect of pressure, temperature. and conccntration of aqueous NaCl solutions on the dynamic and of water. Twenty-eight tables kinematic viscosity gcncratcd from the correlating equations cover a temperature range from 20 to 150C. a pressure range from 0. I to 35 mPa. and a concentration range from 0 to
6 molal.
oxygen which
gen from the interstitial water. Sediments laid down in a shoreline environment will differ in degree of oxidation
24-17
I191111 1000
I 10,000
PSILI
compared ment.
with
in a deepwater
environsediments
For example,
may range from -50 to 0 mV, but the Eh of deepwater sediments may range from - 150 to - 100 mV. The aerobic ly consumed; bacteria die when the free oxygen bacteria attack is totalthe anaerobic the sulfate
TEMPERATURE
.F
ion, which is the second most important anion in the seawater. During this attack. the sulfate reduces to sulfite and then to sulfide; the Eh drops to -600 mV, H 2 S is liberated, above 8.5. and CaCO 3 precipitates as the pH rises
Fig. 24.9-Viscosity of sodium chloride solutions as a function of temperature and concentration at 14.7 psia.
brines primarily
problems constituents
ficult and very time-consuming. Knowledge of the dissolved organic portant because these origin and/or migration as to the disintegration tion. The concentrations
Dissolved Gases
Large oilfield are with quantities of dissolved gases are contained in brines. Most of these gases are hydrocarbons; other gases such as CO2 , N?. The solubility water of the salinity, with increased and HzS gases often generally
constituents are related to the of an oil accumulation, as well or degradation of organic of an accumulaconstituents in
however, decreases
present. pressure.
and increases
oilfield brines vary widely. In general, the more alkaline the water, the more likely that it will contain higher concentrations organic acids: of organic consists constituents. of anions The bulk of the matter however, and salts of organic also are present.
Hundreds of drillstem samples of brine from waterbearing subsurface formations in the U.S. gulf coast area were analyzed to determine their amounts and kinds of of the dissolved hydrocarbons. 2o The chief constituent gases usually was methane, with measurable amounts of ethane, propane, and butane. The concentration of the dissolved hydrocarbons generally increased with depth in a given formation and also increased basinward with regional and local variations. In close proximity to some the waters were enriched in dissolved oiltields, hydrocarbons, and up to I4 scf dissolved gasibbl water was observed in some locations. A more detailed sion of this topic is given in Chap. 22. discus-
other compounds
im 0 0.5. ,o
; 0.611
Organic Constituents
In addition of organic form these occur to the simple constituents in oilfield However, have not hydrocarbons, ionic, been In recent a large number and molecular years, some of quanFig. 24.10-Effect
TEMPERATURE, F
organic
constituents
measured
24-18
PETROLEUM
ENGINEERING
HANDBOOK
Knowledge of the concentrations and other components in oilfield ploration. in water elevated The solubilities at ambient conditions tern eraturex and determined. x3. f:
water is the water con pores or spaces between the minute Interstitial waters are .snl,yc,,trric'
of some of these compounds and in saline waters at pressures have been of these and other oilfield brines is experimentally that
at the same time as the enclosing rocks) or cyigcrwric (originated by subsequent infiltration into
However. the actual concentrations organic constituents in subsurface another matter. It has been shown
connate implies born. together-connascent. Thereshould bc considered origin. Connate an inof water
the solubilities of some organic compounds found in crude oil increase with temperature and pressure if pressure is maintained on the system. The increased solubilitiea solubilities become decrease significant with above increasing 150C. The water salinity.
this definition is fossil water that has been out of contact with the atmosphere for at least a large part of a geologic period. The waters are those that have and that implication with that connate waters are only those born restriction. the enclosing rocks is an undesirable
Waters associated with paraffinic oils are likely to contain fatty acids. while those associated with asphaltic oils more likely contain naphthenic acids. Quantitative recovery of organic constituents from oilfield brines is difficult. Temperature and pressure changes. bacterial actions. adsorption. and the high inorganic/organic-constituents ratio in most oilfield brines are some difficult. reasons why quantitative recovery is
changed chemically and physically, before. during, after sediment consolidation. Some of the reactions occur in or to diagenetic waters include change, replacement (dolomitization). permeation, and membrane filtration.
drilling. or
A water containing
dissolved
excess of 100 g/L can be classified as a brine. Waters associated with the other 30% of petroleum reserves contain less than 100 g/L dissolved solids. Some of these waters fresher are almost fresh. However, the presence of waters usually is attributed to invasion trap. waters fields after the can be such as
Condensate
Water.
gas
sometimes is carried as vapor to the surface of the well where it condenses and precipitates because of temperature and pressure changes. More of this water occurs in the winter and in colder climates and only in gas-producing wells. This water is easy to recognize because it contains a relatively small amount of dissolved solids, mostly derived from reactions with chemicals in or on the well casing or tubing. Water analyses may be used to identify the water source. In the oil field one of the prime uses of these analyses prevent is to determine the source of extraneous water to in an oil well so that casing can be set and cemented
petroleum accumulated in the reservoir Examples of some of the low-salinity found in the Rocky Mt. areas in Wyoming
Enos Creek, South Sunshine. and Cottonwood Creek. The Douleb oil field in Tunisia is another example. The composition are the composition of dissolved of the water solids found in oilfield enwaters depends on several factors. Some of these factors in the depositional
vironment of the sedimentary rock, subsequent changes by rock/water interaction during sediment compaction. changes by rock/water interaction during water migrawith tion (if migration occurs), and changes by mixing
In some wells a leak may develop in the casing or cement, and water analyses are used to identify the waterbearing horizon so that the leaking area can be repaired. With the current emphasis on water pollution prevention. it is very important to locate the source of a polluting brine so that remedial action can be taken. Comparisons of water-analysis data are tedious and time-consuming; therefore. graphical methods are commonly used for positive, rapid identification. A number of systems have been developed. merit. all of which have some
other waters, including infiltrating younger waters such as meteoric waters. The following are definitions of some types of water.
was the
inage with
compared
the age of the enclosing rocks and is not more small part of a geologic period. I
than a
Seawater.
The
composition
of
seawater
varies
Graphic Plots
Graphic lustrate graphical plots of the reacting the relative presentation amount values can be made to ilpresent. The of of each radical
somewhat, but in general will have a composition relative to the following (in mg/L): chloride--19.375, bromide-67, - 10,760, tium-8. sulfate-2,712. magnesiumpotassium-387. sodium 1,294, calcium-4 13, and stron-
Several methods
PROPERTIES
OF PRODUCED
WATERS
24-19
diagram
with
events
plus transgres-
or star diagram.
of the ions are plotted are calculated masses (EPMs) of a given total by 100. reaction values, the actual values,
regression of oceans and seas, even these probably were subjected to marine waters by event. the petroleum, which formed from
values proton
organic matter deposited with the sediments, from what usually is called the source rock
percentage
are often
porous and permeable sedimentary rock. Petroleum (i.e., oil and gas) is less dense than water; therefore, it tends to float to the top of a water body regardless of whether the water is on the sutiace or in the subsurface. Therefore, water associated with petroleum in a subsurface reservoir is called an oilfield water. By this definition, any water associated water. with a petroleum deposit is an oilfield
ion concentrations.
The question of the origin of oilfield brines is difficult to answer in a general manner. The water involved and the constituents dissolved in the water to form the brine can involve divergent histories. Subsurface water is there either because it originally was there or because it infiltrated to the subsurface from the surface. If it was there originally, it would be endogenetic, whereas if it infiltrated from the surface and/or penetrated with sediment accumulations, it would be exogenetic. Obviously these two types of waters could meet and mix in the subsurface and thus the mixture The problem would could contain multiwater of two separate origins.
the reaction values of the ions on a system of rectangular coordinates. 87 The catto the left and the anions to the right of a The endpoints then are connected by sometimes called a constituent that usually zero line.
straight lines to form a closed diagram, a butterfly diagram. To emphasize may be a key to interpretation, by changing the denominator
ply if more than one exogenetic water were involved. The chemical composition of an oilfield brine is an end product dissolved of several variables. These variables include (I) ions, salts, gases, and organic matter, (2) reac-
in multiples of 10. However, when a group of waters is being considered, all must be plotted on the same scale. Many investigators of comparing oilfield ple. and nontechnical construct the diagrams. believe that this is the best method water analyses. The method is simpersonnel can be easily trained to
that could cause the composition of a subsurface oilfield brine to change in composition, including leaching of the rocks, ion exchange between water and rock, redox, mineral hydration, mineral formation and/or dissolution, and be ion diffusion, gravitational segregation of ions, membrane filtration, or other osmotic effects. It is rather difficult to rank the factors that might more important for general consideration. of the more important factors probably
Other
diagrams
Methods.
Several
other
water primarily
fresh waters,
and Piper diagrams, 87~88 were ada ted to automatic *B and Morgan processing by Morgan et al. McNellis. plot diagram * The Piper diagram the water shows the chemical uses a multiple and this to depict analysis.
trilinear
quaternary ap-
composition of the water and interaction of that water with the rocks. If one assumes that the original water was a marine quently original portant water and that the associated sedimentary composition factor. rocks) were sediments marine, (subsethen the
composition Angino
terms of cations
and anions.88
plied the automated Stiff and Piper diagrams oilfield brines and obtained good results.
However, even the salinity in the various oceans and seas is not constant. For example. the salinity of the waters in the major oceans ranges from 33,000 mg/L, about 40,000 mg/L in the Mediterranean to 38,000 Sea, up
to 70,000 mg/L in the Red Sea. about 18,000 to 22,000 mg/L in the Black Sea, and only about I .OOO mg/L in the Baltic Sea. Some land-locked waters. such as the Dead Sea. Great Salt Lake, etc., contain waters that are nearly saturated with dissolved solids. Studies of formation waters in the western Canada sedimentary deposited deposited freshwater that 80% basin under indicate marine that 85% conditions, of the strata were while 15% were
originally
water. Sediments laid down by lakes and streams would not contain a marine water during their initial deposition.
with oceans and seas. The original was marine in such therefore.
associated sediments.
24-20
PETROLEUM
ENGINEERING
HANDBOOK
deposited
under marine
conditions.
bromide
in
the
Smackover
brines
is
3,100
clusion that one could assume with negligible error that all sedimentary strata originally contained seawater. Further, the study important volumetric indicated that evaporites form an part of several of the stratigraphic
Therefore. the average degree of concentration brines compared to seawater is 3,100+65 =48. ing that seawater imate composition
units. Some of the stratigraphic units possibly contain bitterns subsequent to halite precipitation but preceding precipitation average of potassium salts. They calculated an mg/L formation water salinity of about 46,000
Shale Compaction
Some found water first investigators in oilfield through was
waters
TDS. which indicated a net gain of dissolved salts. Of the major and some minor components, all showed a net gain with the exception Factor analyses. major dilution of Mg and SOa. analysis was used in interpretation of the and the following factors were considered to be composition of the original membrane seawater, filtration, by freshwater recharge,
shale.
filtration Laboratory
suggested
controls:
periments indicated that semipermeable membranes. nature should cause a water of the shale membrane and the other side. This follows
natural shales can act as ys.y6 This system working in to be more saline on one side almost fresh or less saline on because the shale membrane
dolomitization, bacterial reduction of of chlorite. cation exchange on clays, and solubility matter, from organic
should filter out the dissolved ions on the upflow side. causing the water on the downflow side to contain few or no dissolved ions.
It was concluded that (I) the formation waters of western Canada are ancient seawaters in which the deuterium concentration was changed because of mixing with infiltrating fresh water, (2) oxygen-18 was exchanged with carbonates in the rocks, and (3) dissolved salts are in equilibrium with the rock matrix subsequent to their redistribution by membrane filtration and/or dilution by freshwater recharge. Equilibrium was attained by such processes as dissolution formation. cation exchange from clays and organic The majority of evaporates, on clays. and mineral new mineral of ions solubilities. are involved brines. Furthat desorption
are: Alabama, 0.3%; Alaska, 19.9%; I I .7%; Colorado, I .O%; Florida, I .4%; 13.4%; 0.2%; Montana, 1.0%; Mississippi, New Mexico, 2.3%; North
matter,
of the published
Texas, 31.2%; Utah, 0.8%: and Wyoming, 4.2%. Fig. 24.1 I indicates the crude oil and water production from wells in the 14 states. The figure indicates production of about 4.3 bbl water/bbl oil. Fig. 24.12 is a similar graph for 13 states excluding Alaska. This figure indicates production of about 5.2 bbl wateribbl oil. Further. it can be shown as cumulative the older the well, that oil wells produce increases. the WOR. more water oil production the higher In other words,
agree that all these controls and/or reactions in establishing the composition of oilfield ther. most investigators agree with
the assumption
marine water usually is a part of the original material from which formation waters evolve. Opinions, however, are not unanimous with respect to how they evolved. The major disagreements are related to the membrane ccntrating evaporation. It oilfield filtration the theory and to other solids such modes of conas seawater of some dissolved
is possible brines
to reconstruct in sedimentary
the evolution
they are genetically related to evaporites. For example. geochemical and geological studies of some very concentrated brines indicate that in deep quiescent bodies of water. strong bitterns can persist for long periods of time under a layer of near-normal seawater. As a result, carbonates can precipitate from the less saline water and fall through the bittern at the bottom, and. as compaction proceeds. the pore spaces remain filled with bitterns. very far the the (I) (4) (6) from Some fossil brines once trapped have not moved or very fast.
damage from brine, (3) to produce potable water. to conserve all valuable minerals and energy. Precipitation is the most common separation
used in separating minerals from any type of brine. Minerals recovered from brines in the U.S. include compounds of iodine, bromine. chlorine, sodium. lithium. potassium, magnesium, and calcium. Evaporation of a saline water or brine will cause the
precipitation of calcium carbonate. calcium sulfate, sodium chloride, magnesium potassium sulfate, chloride, and finally magnesium chloride. These are the major chemicals in most brines, and this is the sequence order in which these compounds will precipitate. Mere preclpltation a brine; will not produce a very pure chemical or physical from therefore, other chemical processes
A geochemical model can be built to represent origin and evolution of this type of brine by using relatively simple operations evaporation. (2) precipitation, mineral leaching. fomlation and (7) expulsion and processes of (3) sulfate reduction, (5) ion exchange. interstitial that about with fluids of
and diagenesis.
evaporites during compaction. Experimental work indicated bromide seawater in seawater evaporates. The
arc used. For example. iodine, which is in the foml of iodide in a brine. is recovered by these steps: (I) the iodide is oxidized to iodine, (2) the iodine is stripped from the brine, (3) the iodine then is reduced back iodide is oxidized again to iodine, to iodide, (4) the (5) the iodine is
precipitates average
concentration
PROPERTIES
OF PRODUCED
WATERS
24-21
13 Stclcs 30000
(ercludlng
Aiasio)
29000
D 2 6 6 2 E
2700C
27000
g
2600C o z I ZSOOC 25000
c
26000
BOO<
7000 H---L,Crude
700(
,/ --4
6000
6OOt 1975
I
76
I
77 YEAR
I
78
I
73 80
5000 I! 5 76 77 YEAR 70 79
80
Fig. 24.1 I-Crude oil and water produced ( x 1,000 B/D) from wells in 14 states including Alaska.
oil and water produced ( x 1,000 B/D) from wells in 13 states excluding Alaska.
crystallized purified
and filtered,
crystals
are
recover study
minerals
from
brines
should
make
an in-depth Some of
further.
of the current
and potential
economics.
Bromine is recovered by a similar though, of course, different process. It is present in a brine as bromide. The U.S. Bureau ofMines Yearbook9 contains information on general recovery technology, domestic production, consumption and uses, prices, stocks, etc., of minerals recovered from brines as well as minerals recovered from solely other sources. in more than 40 years built brine went on stream in to produce iodine from The first plant in the U.S. 1977 near Woodward, million lbm iodineiyr demand. Some
the factors that must be determined are (1) concentmtions of valuable minerals in a brine, (2) amount of brine available, (3) costs of gathering the brine, (4) costs of recovering the minerals from the brine, (5) present potential market demand for the recovered minerals, (6) costs of delivering the minerals to the market. 26 and and
References
1. Rogers, W.B. and Rogers, H.D.: On the Connection of Thermal Springs m Virgin& with Anticlinal Axes and Faults, Am. GPO/. Rep. (1843) X-47. 2. Howell, J.V.: Historical Development of the Structural Theory of Accumulation of Oil and Gas, W.E. Wrathcr and F.H. Lahee (edi.). AAPG, Tulsa (1934) l-23. 3. Dickey, P.A.: The First Oil Well, J. Pet. Tec~h. (Jan. 1959) 14-26. 4. Schilthuis. R.J.: Connate Water m Oil and Gas Sands, Trtrnr., AIME (1938) 127. 199-214. 5. Torrey. P.D.: The Discovq of Interstitial Water. Prtrd. Mmh/\ (1966) 30, 8-12. 6. Griswold, W.T. and Munn, M.J.: Geology of Oil and Gas Fields in Steubenville. Burg&&town and Claysvllle Quadrangles, Ohio, West Virginia and Pennsylvania, Bull., USGS (1907) No. 318. 196. 7. Fettke, C.R.: Bradford Oil Field, Pennsylvania. and New York, Bull , Pennsylvania Geologic Survey, Fourth Ser. (1938) M21, l-454. 8. Munn, M.J.: The Anticlinal and Hydraulic Theories of Oil and Gas Accumulation, Icon. Gee!. (1920) 4. 509-29.
OK. It has a design capacity of 2 or about 30% of the annual U.S. operations to recover
smaller-scale
iodine started in 1982 near Kingfisher, OK. Several plants recover bromine and other constituents from from brine near El Dorado, AR. the Smackover formation. The brine is produced
Magnesium and other chemicals are recovered from lake brines, well brines, and seawater in plants located near Ogden, UT; Ludington, MI: Freeport, TX; and Port St. Joe, FL. Clayton Valley, Lithium NV. 98 is recovered from brine near
Economic Evaluation
It is true therefore, that any market company demand fluctuates with supply; to considering an enterprise
24-22
PETROLEUM
ENGINEERING
HANDBOOK
9 IO
Mills, F. van A.: Experimental Studies of Subsurface Relationships in Oil and Gas Fields. Econ. Geol. (1920) 15. 398-421. Rich, J.L.: Futther Notes on the Hydraulic Theory ofOil Migrdtion and Accumulation. Bull AAPG (1923) 7. 213-25
41 42 43 44 4s. 46. 47 48 49. 50. Jensen. F.W and Rolshausen. F.W.: Water\ from the Frio Formation, Texas Gulf Coast. Trms.. AIME (1944) 155. 23-38. Rogen, G.S.. Some Oil-field Waters of the Gulf Coast, Buil., AAPG (1919) 3. 310-31 Minor. H.E.: Oil-field Waters of the Gulf Coast Plain, Siciner Pmers Memoriul Volume. AAPG (1934) 89 I-905. Meents. W.F. er ul. : Illinois Oil-field Brines. Illinois Geoi. Survq Div., I//. Per. (1952) 66. Schoewe. W.H.: Kansas Oil-field Brines and Their Magnesium Content, Bull., U. of Kansas Publ. (1943) 47, pt. 2. Rail. C. and Wright. J.: Analyses of Formation Brines in Kansas. RI 4974, USBM (1953). Survey of Oklahoma Formation Waters. Mid-Continent District Study Comm. and the U. of Tulsa (1957). Wright, J. efal. : Analyses of Brines from Oil-productive Formatlons in Oklahoma. RI 5326. USBM (1957). Plummer. F.B. and Sament. E.C.: Underzround Waters and Subsurface Temperature; of the Woodbme Sand in Northeast Texas. Bull.. U. of Texas Publications (1931) 3138. Plummer, F.B.: Texas Water Resources. Bull.. U. of Texas PublIcationa (1943) 4301. 301-12. Barnes, V.E.: Eanh Temperatures and O&field Waters of North-Central Texas, Bull., U. of Texas Publications (1943) 4301. 313-57. Bergcr. W.R. and Fash. R.H.: Relation of Water Analvseh to Structure and Porosity in the West Texas Permian Basin. -5. . 10 I I<\ Pon,ers Mmorial I/oiunte. AAPG (1934) 869-89. Beeler. H.S., McKinney, O.B.. and White, V.C.: The Chemical Analyses of Brines from Some Fields m North and West Texas. Petroleum Branch. AIME (1953). Coff-in. C.R. and DeFord. R.K.: Waters of the Oil and Gas Bearing Formationr of the Rocky Mountains. Sidnev Povert Memoriul Volume, AAPG (1934) 927-52. Crawford. J.G.: Waters of Producing Fields in the Rockv Mountain Region, 7rcm.x 1 AIME (1948) 179, 26486. . Crawford, J.G.: Oil-field Waters of Montann Plains. Blrll.. AAPG (1942) 26. pt. 2. 1317-74. ROSS, J.S. and Swedenborg, E.A.: Analyses of Waters of the Salt Creek Field Applied to Underground Problems. Trmx.. AIME (1929) 82. 207-20. Cmwford. J.G.: Oil-field Waters of Wyoming and Their Relation to Geological Formations. Bull., AAPG (1940) 24, pt. 2. 1214-1329. Campbell. W.P.: Oil-field Waters of Alberta and Saskatchewan. Trans. 1 Cdn. Inst. Mining Met. (1929) 32. 316-32. Elworthy, R.T.: A Field Method and Apparatus for the Determination by Means of Electrical Conductivity Measurements of the Character of Waters Leaking into Oil and Gas Wells. Cdn. Dept. Mines Summary Rept. No. 605 (1922) 58-70. Hitchon. B.: Formation Fluids, Gw/o#d H/srory of Wesrem Cnnrrda, R.G. McCrossan and R.P. Glaister (eda.). Alberta Sot. of Petroleum Geologists, Calgary, Alta. (1964) 201-17. Harris, W.E. etal.: Viking Formation Waters of Alberta, Alto. SK Per. Geol. (July 1957). Analyses of Formation Water Samples. Dept. Mming Resources, Mmes Branch, Province of Manitoba. Canada. Analyses of Formation Water Samples, Dept. Mining Resources, Mines Branch, Provmce of Saskatchewan. Canada. Borger, H.D.: Case History of Quniquire Field. Venezuela. Bull. , AAPG. 31, 229 I-2330. Oil Fields of Royal Dutch-Shell Group in Western Venezuela, Bull., AAPG (1948) 31. 517-623. Suter. H.H.: El Mene De Acosta Field, Falcon, Venezuela. Bull.. AAPG, 31, 2193-2206. Hedberg. H.D., Sass. L.C., and Funkhouser, H.J.: Oil Fields of Greater Olicma Area, Central Anaoategui. Venezuela, Euli., AAPG (1947) 31, No. 12, 2089-2169. Rittenhouse, G. et al.: Minor Elements in Oilfield Waters, Chem Gee/. (1969) 4, 189-209. Krejci-Graf, K.: Data on the Geochemistry of Oilfield Waters, Geologisches Jahrbuch Reihe KJ (1978) 3, No. 174, 21-25. Bockmeulen. H., Barker, C.. and Dickey, P.A.: Geology and Geochemistry of Crude Oils, Bolivar Coastal Field?, Venezuela. Bul/. , AAPG (1983) 242-70. Collins. A.G.: Geochemistry of Some Tertiary and Cretaceous Age Oil-Bearing Formation Waters, Environmental Science and Tec,hnologv (1967) 1, 725-30.
11 Torrey. P.D.: A Review of Secondary Recovery of 011 in the United States. Secondary Recovers of Oil in thy United Statr.r. API, New York City (l9iO) 3-29. I2 Garrison, A.D.: Selective Wetting of Reservoir Rock and Ita Relation to Oil Production, Dri(!. bnd Prod. Pruc.. New York City (1935) 130-40. 13 Lane. A.C. and Gordon. W.C.: Mine Waters and Their Field Assay, Bull., Geologx Sot. of America (1908) 19. 501-12. Connate, and Metamorphic Water, I4 White, D.E.: Magmatic, Bull., Geologic Sot. of America (1957) 68. 1659-82. IS Wallace, W.E.: Water Production from Abnormally Pressured Gas Reservoirs in South Louisiana, J. Pet. Tmb. (Aug 1969) 969-82. I6 Rrcovmended Pructice j?v Anulyi.~ oj Oilfieid Wuters. API RP-45, API, Dallas (1968). 17 Salme and Brackish Waters, Sea Waters and Brines, American Society for Testing and Materials, Annucd Book of ASTM Smr/c&~. Part 3l-Water. Section VII. Philadelphia (1982). to Oilfield Water Technology. 18 Ostroff. A.G.: Introduction NACE. Houston (1979) 394. 19 Noad. D.F.: Water Analysis Data, Interpretation and Application, J. C&z. Pet. Tech. (1962) 1, 82-89. 20 Buckley. S.E.. Hocott. C.R. and Tanean. M.S. Jr.: Distribution of Dissolved Hydrocarbons in Subsurface Waters. Hubrtrrr o/ Oil, L.C. Weeks (ed.). AAPG, Tulsa (1958) 850-82. 21 Zarrella. W.M. et ol. Analysis and Significance of Hydrocarbons in Subsurface Brines. Geochim. Cosmochirn. Arm (I 9b7) 31. 1155-66. of Crude Oils by Water and 22 Bailey. N.J.L. ef al. : Alteration Bacteria-Evidence from Geochemical and Isotope Studieh. Buli.. AAPG (1973) 57, No. 7, 1276-90. Acid Anions and Stable Carbon 23 Carothers. W W.: Aliphatic Isotopes of Oil Field Waters in the San Joaquin Valley, California. Masters thesir, San Jose (CA) State U. (1976). 24 Sunwall. M.T. and Pushkar, P.: The Isotopic Composition of Strontmm in Brines from Petroleum Fields of Southeastern Ohio, Chem. Geol. (1979) 24. 189-97. 25 Hoke, S.H. and Collins, A.C.: Mobile Wellhead Analyzer for the Determination of Unstable Constituents in Oil-Field Waters. ASTM STP 735 (1981) 34-48. 26 Collins. A.G.: Geochemisr~ of Oil&/d Wafers, Elsevier Scientific Publishing Co.. New York City (1975) 496. Inmroducrion to 27. Beestecher, E : Petroleum Microbiology-on Microbiological Engineering, Elsevier Scientific Publishing Co.. New York City (1954) 375. J.B.: Prrrdrum Microbrology, Elsevier Scientific 28. Davis. Publishing Co., New York City (1967) 604 Campbell Petroleum Series. 29. Patton, C.C.: Oilfield Water Swtems, Norman. OK (1974) 65. 30. Postgate, J.R.: The Sulfate Reducing Bacteria, Cambridge U. Press, New York City (1979) 151. Water Analysis Data Bank, DOE/EC/ 31. Bright, J.: Oil&Id 10116-2, U.S. DOE (1983). 32. Watkins, J.W.: Properties of Produced Water. Perroleum Production Handbook, T.C. Frick and R.W. Taylor (eds.). SPE, Richardson, TX (1962) II, 21.1-21.20. Report on the 33. McGrain. P. and Thomas. G.R.: Preliminary Natural Brines of Eastern Kentucky, Kenluck?/ Geol. Sur~je> Report (1951) 3. Analyses of Kentucky Brines. 34. McGrain, P.: Miscellaneous Kentucky Geol. Survey Report (1953) 7. 35. Lambom. R.E.: Additional Analvses on Brines from Ohio. Ohio Grol. Surveys Report (1952) li. 36. Stout, W., Lambom. R.E., and Schnaf, D.: Brines of Ohio. BUM., Ohio Geol. Survey, 4th Series (1932) 37. Waters of Pennsylvania. Penn. Slate. 37. Barb, C.F.: Oil-field Cdl. BUU. (1931) 8. of Oil-Field Waters of the Ap38. Torrey, P.D.: Composition Sidney Powers Memorial Volume. AAPG palachian Region, (1934) 84-53. Wrsr Virginra 39. Price, P. er al.. Salt Brines of West Virginia, Cd survqv (1937) 8. 40. Rogen. G.S.: Chemical Relation of the Oil-Field Waters in San Jaaquin Valley, Calif., Bull.. U.S. Geol. Survey (1917) 653. 5.
51 .52,
53.
54.
55.
59.
60. 61.
62.
73.
PROPERTIES
OF PRODUCED
WATERS
24-23
74
0\1f. T.L.: The Eftccth of Salt. Gar. Temperature. and Pre\>urc on the Comprewhdtty of W&x-, paper SPE I.1 174 pre\cnttd at the 1984 SPE Annual Technical Conference and Exhibition. Houston. Scpt I6- 19.
7s 246. and Standing. M.B.: Pressure. Volume, 76. Dotaon, C.R. Temperature and Soluhility Relation5 for Natural Gas-Water MIXturcs: Drill. t/d Prod. Prw.. API (1944) 173-79. A.M. Jr. and Chou. J.C.S.: Pressure-Volume71. Rowe, Temperature-Concentration Relations of Aqueous NaCl Solution\. J. Chw~. Dutcr (1970) 15, 61-66. 78 Ashhy. W.H. Jr. and Hawkins, M.F.: The Solubility of Natural Gas in Oil-Field Brines. paper presented at the 1948 SPE Annual Meetmg. Dallas. Oct. 4-6. Propcrtlc5 of 79 Roger\. P.S.Z. and Pitzer. K S.: Volumctnc Aqueous Sodium Chlonde Solutions. J. Ply.\. Clrtwr. Rcf: D~rt<r (1982) 11. No. I. 15-81. of WC,// Lo,q //lrc,rpn,ta/i/,/r, x0. Wy II ie. M. R.J.: 77w F[rn~~~rr,lrnr~r/s wcond edstnn. Academx Press [nc.. New York CQ (1954) ret. 1957. 81. Amyx. J W.. Bass. C.M. Jr.. and Whiting, R.L.: Pcrrolrwi Rccr~~vir Fqi,wcv%,q. McCrdw~Hill Book Co. Inc., New York City (1960). J.. Khahfa. H.E.. and Co&a, R.J.: Tables of the 82. K&in. Dynamic and Kinematic Viscosity of Aqueour NaCl Solutitrn\ in the Temperature Range 2Ol5OC and the Pressure Range 0 I-35 MPa.J. fhrs. Chcwr. Rrf Dmtr (1981) 10. No. I, 71-87. 83. McAuliffe. C.D.: Solubiiity in Water of Paraffin, Cycloparatiin. Olefin. Acetylene. Cycle-olefin and Aromatic Hydrocarbons, J. Phy. C/w,,,. ( 1966) 70, 1267-75. 84. Prince, L.C.: Aqueous Solubility of Petroleum as Applied IO Its Ongin and Primary Migration. &t/l.. AAPG t 1976) 2 13-44 85. Tickell, F.G.: A Method for Graphical Interpretation of Water Analy\i\. Co/,/: Sfrw Oil Gr~r Suprrv. (1921) 6. 5-I I
X6. Reistlc. C.E.: ldentlticatwn of Oilficld Water\ bp Chcnwal Analysi\. U.S. Bur. Min. Techmcal Paper, 404 (1927). x7. Stllf. H.A. Jr.: The Intcrprctntion of Chemical Water Analysis by Mean5 of Patterns. J. Pet. Twh. (1951) 192. 15-17. 88. Piper. A.M.: A Graphic Proccdurc m the Geochemical Interpretation of Water Analyw. U.S. Geol. Sun: Ground Water Note 12, (1953). 89. Morgan. C.O.. Dingman, R.J.. nnd McNelli\. J.M.: Digital Computer Methods fur Water-quahty Data. Griwrcl wow, (1966) 4, 35-42. 90. Morgan. C.O. and McNellis. J.M.: Still Diapraut~ ul Wnterquality Data Programmed for the Digital Cumputer. Kanses State Geol. Sun,. Spec. Dtstrib. Publ. 43 (lY6Y) 91. Angino. E.E. and Morgan. C.O. Appllcotlon of Pattcm Analysi\ to the Classification of Oilfield Brine\. Kanw Slale Cc01 Sure. Comput. Contrlb. 7 (1966) 53-56. 92. Hitchon. B.. Billings. G.K., and Klovan. J E.: Geochemlwy and Origin of Formation Waters in the Western Canada Sedimcn~ tary Basin-Ill. Factors Controlling Chemical Compovt~on. G<w/?i/,~ Cmr,wc~hi~,~.Am, ( I97 I ) 35. 567-98. 93. Clayton. R.N. ct r/l. : The Origin 01Snline Fomwion Waters. I. Isotopic Composition. J. G~qdi,u Rc.\. (1966) 71. 3869-112 of Oil-field Bnnes. &f/I. AAPG 44. De Sitter, L.U.: Diagene\ir ( 1947) 2030-40. 95. Kharaka. Y.K. and Berry. F.A.F.: Simultaneous Flou of Water and Solutes Through Geological Membranes 1. Experimental Investigations. Geoc~hiul. Co.\r,?oc~/iil,i. ACICI. ( 1973) 37. 2.577~260.7. 96. McKelvey. J.G. and Milne. J.H.: The Flow of Salr Solution Through Compactrd Clay. in Clays and Clay Mmen~l\. 9th Natl. Conf. Clays and Clay Mineral\ (1962) 24X-59. Minertrlc Yrwrhd. Ah/d\. Miwn~l~. mrl 97. U.S. Bwwrrc c~fh4irw.s Fwk. U.S. Bureau of Mine\. Washinztcm. D.C. (Yearlv). Resources and Requirements by the Year 9x. Vine. J.D.: Lithntm 2000. U.S. Geological Survey Professional Paper 1005, U.S. Gwemment Printini Office. iahhingtun D.C. t i976)
Chapter 25
Systems
Introduction
The occurrence of water with hydrocarbons both in the reservoir and in the produced states represents the norm. Even though streams saturated with water enter the producing tubing, subsequent cooling generally produces separate phases of low mutual solubilities. Nevertheless, the mutual solubility of water and hydrocarbons is extremely significant in the processing of produced fluids. In the reservoir, their mutual solubilities increase as the temperature increases, either as a result of reservoir depth or external heating, as in the case of a steamflooding operation. Thus, the definition of the saturated water content in the equilibrium phases is the subject of this chapter. The coexisting phases may be gas, G, hydrocarbon-rich liquid, LHC, water-rich liquid, L,, or hydrate, H, although the coexisting phases are seldom pure. The advent of EOR processes that use CO 2 gives rise to the occurrence of a CO*-rich liquid phase, which we designate Lco2 . which states that the degrees of freedom or number of independent variables required to define the system, F, equals the number of components, C, minus the number of phases, P+2, or F=C-PC2. A few examples given this important relationship are summarized in Table 25.1. The phase rule is essentially a rule of algebra applied to equations of equilibrium and as such does not specify which phases are in equilibrium or their equilibrium concentrations. Nevertheless, its value in the organization of knowledge regarding phases in equilibrium and their relationship is important and always should be respected and used. The pressure-temperature projection of the univariant coexistence lines of binary hydrocarbon/water systems represents a unique way of summarizing the phase behavior of the system. The methane/water system is one in which the critical point of the gas falls far below the ice point. With high pressures, even at moderate temperatures, solid hydrates are formed. The pressuretemperature projection of the system is given by Fig. 25.1 .2 Of particular interest to petroleum production engineers is the precise location of the L,.-H-G line or the initial hydrate formation condition. Since natural gases vary in composition, the L,-H-G line is usually compositionally dependent. Fig. 25.2 shows a constant-pressure trace of the saturation conditions of the methane/water binary system at a constant pressure exceeding the critical pressure of methane. Among the important features of this diagram are the dissolved gas concentration of the water-rich liquid phase (e.g., at Point 4), the equilibrium dewpoint locus, the initial hydrate formation temperature along the horizontal line 7-2-5-6, and the equilibrium and
25-2
PETROLEUM ENGINEERING
HANDBOOK
TABLE 25.1-SUMMARY AND MEANING OF DEGREES OF FREEDOM Number of Independent Variables Required To Represent each Locus of States
Number of Comoonents
Equilibrium Phases
metastable dewpoint locus (Line 7-8) below the initial hydrate formation temperature and below the stable dewpoint line. The definition of the equilibrium dewpoint water content for methane has been determined by Olds et al. 3 and the initial hydrate formation condition by Villard,4 Deaton and Frost, 5 Kobayashi and Katz, and Marshall et al. 6 The definition of the gas-hydrate equilibrium locus, as distinguished from the metastable equilibrium locus erroneously reported in most dewpoint charts, has been reported by Sloan et al. The measurement of the gas hydrate compositions or hydrate numbers for methane was conducted successfully by Galloway et al. 8 and found in essential agreement with the statistical
mechanical theory of van der Waals and Platteeuw, as applied by Saito et al. O as well as Panish and Prausnitz, and other subsequent workers. The second type of pressure-temperature projection presented is for the propane-water system in which the critical temperature of the hydrocarbon exceeds the ice point. Since the aqueous and hydrocarbon-rich phases exhibit low mutual solubilities, a three-phase H-L,-G condition occurs in the neighborhood of the vapor pressure of condensable hydrocarbons such as propane. Fig. 25.4 presents a constant-pressure trace of Fig. 25.3 at a pressure less than the critical pressure of propane (Kobayashi t t ), The principal qualitative difference between Figs. 25.2 and 25.4 is the existence of the LHC-G region in the latter that is absent in the former, at least at moderate temperatures. Note that in the LH~-G region of Fig. 25.4, the water concentration of the gas phase is larger than in the equilibrium liquid phase at the same temperature. On the contrary, in the G-L,,, region, the water concentration in the saturated gas is rather less than in the equilibrium water-rich liquid phase. Owing to the differences in the molecular sizes of methane and propane, the hydrate structures of methane hydrate (Structure I) and of propane hydrate (Structure II) are quite different. By using phase diagrams such as Figs. 25.2 and 25.4, it is possible to trace the phase transitions one would expect to encounter for various composition mixtures as the temperature of the system is lowered. It is important to note that hydrates can form directly from the fluid hydrocarbon phase (i.e., in the absence of free liquid
600
G - L,
_ ,;
$:I
KEY TEPPERATURE, F ------0 c. Tw C@+MENTS
-HETASTABLE LOCUS
DEW PT.
'-% I
@iE CCh'FOdEM CRITICAL LrK"S ~TVANE CRIT~L WATER CRITICAL QsLwPLE PolNl THREE PHASE CRITICAL
-300 N-H
PURE METHANE
-9
El
03)
I ERCENT
t'LIRE WATER
Fig. 25.2-Constant-pressure
trace of the methane/water system at a pressure greater than the methane critical pressure.
SYSTEMS
25-3
P -l-l
-400 0 400 ml PERCENT WATER
Fig. 25.4-Constant-pressure
I
I-lJKt WAlltft
TEWERATLUE, F
Fig. 25.3-Pressure-temperature
heterogeneous system.
trace
of
the
propane/water
system at pressure.
water at low water compositions) provided that the temperature is sufficiently low and turbulence is present or the equilibration times are large, as shown by Cady * The metastable dewpoint line shown in Fig. 25.2 (Line 7-8) rather than the true equilibrium state may be attained under certain situations to give a false or metastable indication of equilibrium (Kobayashi and Katz 13). References for aqueous/volatile-gas systems related to petroleum production are listed in the references at the end of the chapter. These are for methane/water system (Refs. 3 and 14 through 44), natural-gas/water (Refs. 13 and 45 through 52), COz/water (Refs. 31, 44, and 53 through 96). and for nitrogen/water (Refs. 24. 31( 33 through 36. 62. and 97 through 120). References for other binary and ternary hydrocarbon/water systems are listed in the general references. References for hydrate/volatile-gas systems are for methane/water (Refs. 2,4,6 through 8. and 121 through 129). natural-gas/water (Refs. 126 and 130 through 135). CO?/water (Refs. 5 and 136 through 139), and nitrogen/water (Refs. 6, 140, and 201). References for other binary and ternary hydrocarbon/hydrate systems are listed in the general references.
As shown in Figs. 25.1 through 25.4 for the methane/water and propane/water systems, the phase behavior of such systems is complicated by the existence of multiphase equilibria, including solid phases such as hydrate or ice. The former phase is particularly troublesome in petroleum production since solid hydrates can form at temperature above the ice point wherein the concentration of water in the system is quite low. Fig. 25.5 presents the univariant three-phase loci for various binary hydrocarbon/water systems and shows their critical hydrate formation loci (solid lines). The variability of the water content of hydrocarbonrich phases for a condensable system perhaps is illustrated best by Fig. 25.6, which shows the concentration of water in the propane-rich phases in the propane/water system. In the low-pressure gaseous region the water content is determined primarily by the vapor pressure of water, whereas in the condensed propane region the solubility of water in the propane-rich liquid phase is determined by the strong repulsion of water by the hydrocarbon owing to the thorough hydrogen-bondbreaking ability of the condensed hydrocarbon phase. This phenomenon can be illustrated further by Fig. 25.7, a composite plot of the activity coefficient of water at
25-4
PETROLEUM ENGINEERING
HANDBOOK
10
50
TWERAW, F for
70
93
ax
400
I%ESSUE,
lml
PSIA
2cm
4col
conditions
paraffin
hydro-
Fig. 25.6-Concentration of water in propane-rich fluid phases in the two- and three-phase regions.
moderate pressures from hydrocarbon molecules vs. temperature for molecules ranging from propane to li 7 uids as heavy as naphtha and SAE 20 lubricating oil. I4 Fig. 25.7 also shows that at higher temperatures water becomes increasingly miscible with hydrocarbons, with total miscibility conditions having been reported for intermediate-range hydrocarbons. This phenomenon, which is relevant to high-pressure steamfloods, also is illustrated by Fig. 25.8, which gives their three-phase miscibility conditions. 14* In addition to fluid condensed phases, as the pressure is increased and the temperature decreased, nonpolar molecules and weakly ionizing molecules of appropriate molecular sizes form solid gas hydrates. Davidson 43 presents the approximate molecular diameters and structure type of the hydrate formers and their theoretical hydrate chemical formulas.
\ -
40
TEJEMTIBE, 'C
Fig. 25.7-Composite
in
SYSTEMS
25-5
Estimating Initial Hydrate Formation The estimation for the stability conditions of the hydrates when liquid water is present, the L,-H-G equilibrium locus, can be estimated by various means. Sweet Natural Gas Systems Method. The initial hydrate formation condition can be estimated for sweet natural gas systems by using pressure, temperature, and gas gravity as parameters (see Katz 45. 46) as shown in Figs. 25.9 and 25.10. These figures should be confined for usage when the natural gas mixtures are similar to those used in developing them, with data taken from Deaton and Frost, Wilcox ef al., Kobayashi and Katz, 2 and Katz!45 Typical compositions of natural gases corresponding to gas gravities of Fig. 25.9 are given in Table 25.3. It should be noted that no H2S or CO1 content is tolerated by the correlations. Vapor/Solid Equilibrium Ratios Method. The second approach involves the development and application of vapor/solid equilibrium ratios for gas/liquid/hydrate equilibrium (Carson and Katz, I47 Unruh and Katz, 39 Noaker and Katz, 48 Robinson, 49 and Robinson and Ng), as shown in Figs. 25.11 through 25.16. The development of the vapor/solid K-values involved experimental hydrate formation conditions for mixtures of methane with other gas(es) and the hypothesis that hydrates could be treated as solid solutions, in hindsight a brilliant hypothesis. The vapor/solid Ki(,.,) values are used to calculate the hydrate frost point in direct analogy with dewpoint calculations:
phase at three-
was developed by Parrish and Prausnitz, lo using the Kihara potential and the method of van der Waals and Platteeuw . 9 where Ki(,,.rj is the vapor/solid equilibrium values of component i, yi is the mole fraction of component i in the vapor phase, and xi is the mole fraction of component i in the solid phase, so that Computer Method For Hydrate Dissociation Predictions. Pam sh and Prausnitz Development. lo The method of van der Waals and Platteeuw as developed by Parrish and Prausnitz is slightly more complex than the previous method, but it has two considerable advantages: (1) the equations are related to the microscopic hydrate structure, and (2) the theoretical nature of the model allows it to be extended beyond the G-L,-H region.
f;
i= I
yiiKic\>.n=l.O.
.. .. .
. .
(1)
An example calculation for a complex mixture is presented in Table 25.4. 44 As suggested, the K-value of n-butane is taken numerically as that for ethane. TABLE25.2-PHYSICAL Statistical Mechanics for Adsorption Approach. A third method of estimating initial hydration formation involves the application of statistical mechanics (van der Waals and Platteeuw 15 m well-defined hydrate cages as determined by X-ray crystallography by von 52, Stackelberi and Miiller ) 53 and others. The theory of van der Waals and Platteeuw was applied to predict the initial hydrate formation of pure gases at temperature above the ice point by Marshall et al. 6 and later to binary mixtures by Saito and Kobayashi. 54 Nagata and Kobayashi 55 used the Kihara potential to calculate the dissociation pressures of hydrates, following the earlier work of McKay and Sinanoglu. 56 A more convenient estimation method for hydrate decomposition conditions
DATA OF TWO KNOWN HYDRATE LATTICES Structure I Water molecules per unit cell Cavities per unit cell Small Large Cavity radius, rc Small Large Typical gases that form in each cavity of this structure
Small. Large.
I T
PETROLEUM ENGINEERING
HANDBOOK
Mole Fraction
; A C;H;
I-C,H ,0 n-C,H,, CsH,z + Gravity calcu-
0.603
0.704
0.803
0.906
1.023
40
60
chemical potential of pure component i, universal gas constant, absolute temperature, and activity of component i in mixture.
TDJPERAIIE, F
Fig. 25.9-Initial
Van der Waals and Platteeuw used theory to derive a similar equation for the chemical potential of water in the hydrate structure as follows.
In the late 1940 s and early 1950 s the molecular structures of hydrates, shown in Table 25.2, were studied through the use of X-ray diffraction. 52, s3 The structural determination enabled van der Waals and Platteeuw 9 to develop a model for the prediction of hydrate dissociation pressure at any temperature. Their basic equation looks complex until one considers it as being very similar to the basic equation for the chemical potential of Component I in a mixture of Components 1 and 2. This equation, which relates the chemical potential of
where p,,,H = chemical potential of water in filled hydrate, pw~r = chemical potential of water in empty hydrate, n,i = number of cavities of type i per water molecule in basic lattice, and yji = fractional occupancy of type i cavity by type j molecule.
30 43 4) 60 xl
TDQERATU~ F
Fig. 25.10-Initial
hydrate-formation
condition.
SYSTEMS
25-7
Fig. 25.11-Vapor/solid
1.0-
<OS6 i 5 0.4 Bi 2
3
0.2
0.1
Fig. 25.12-Vapor/solid
0.02
0.01 t 30 40 M TOYPERITIRE, F
Fig. 25.13-Vapor/solid
60
70
80
25-8
PETROLEUM ENGINEERING
HANDBOOK
v A
A /A /I//I//
I I
TABLE 25.4-CALCULATION OF PRESSURE FOR HYDRATE FORMATION OF A COMPLEX MIXTURE AT 50F Mole Fraction in Gas At 300 psia
K 2.04------Y/K
The second term on the right accounts for hydrocarbon filling of the lattice. The term yii is given by
cjifi I+ C Cki.fk
At 350 osia
K 1.90Y/K
Yii =
) . . . . . . . . . . . . . . . . . . . (4)
EHA
&I: I-C, - ,o n-C,H to N2
co2
Total
The linearly mterpolated anwer is 322 ps~a and the experimentally observed hydrate formatcn pressure at 5O F was 325 ps,a.
where Cji is a unique function of temperature for each guest molecule in each size cavity, fi and fk are the fugacities of j and k in the gas phase, and k is ordered from one to the number of components. The fugacity term is determined by an equation of state (EOS), such as Peng-Robinson (PREOS). Is7 The Cji functions may be calculated using the Lennard-JonesDevonshire spherical cell model. The interaction between the guest molecule and a uniform spherical surface representing the cage commonly is described by a Kihara potential function. The method is described by Parrish and Prausnitz. lo The Kihara parameters are determined from single or binary gas dissociation conditions and then may be used to predict dissociation conditions for multicomponent gases. A fit of Cji for common gases is given in Table 25.5. To predict hydrate formation conditions from liquid water and gas, the following conditions must be satisfied. p,,,L =p,,,H=pwg, TL=TH=Tg, and pL =P,T, =pg, ........... .. .. . . (k) ... .... . . .. .... W . .(5b)
......
equilibrium
constants
for hydrogen
where p is the absolute pressure and subscripts L, H, and g are liquid water, hydrate, and gas, respectively.
SYSTEMS
25-9
TABLE 25.5-PARAMETERS
DISSOCIATION
MODEL
Structure I*
~I,,,(T,,~P~),
Parameters for Calculating Langmuir Constant Between 260 K and 300 K C,, = A/T exp (B/T)atm - Small Cavities of Structure I
AxlO CH.4 B
Large Cavities of Structure I AxlO 1.4865 0.4071 0.0 0.0 0.0 5.7883 1.0242 1.3532 B 2878.0682 3820.7119 0.0 0.0 0.0 1669.2292 3172.6655 3739.3355
Small Cavities of Structure II AxlO 2.1778 0.0 0.0 0.0 14.8724 1.1620 1.8306 B 2713.4259 0.0 0.0 0.0 2002.6644 2837.3018 3671.9126
Large Cavities of Structure II AxlO 6.6777 2.9157 1.3212 0.0404 0.0788 15.6182 5.3986 6.4567
8
2.7711
2752.8047
0.0 0.0 0.0 0.0 0.0 0.0 17.7986 1931.5130 1.5227 2943.9948 2.3458 3701.3170
1. Estimate a pressure, p. 2. At the estimated pressure, for the given temperature and gas mole fraction of component j, determine the fugacity ofj from an EOS, such as PREOS. 15 7 Yji) APLW-P~'MT-P~L=-RT C i n,i ln (l3. By using Cj; values from the Kihara potential and yji values from Eq. 4, determine Ap,, from Eq. 3. 4. By using constants from Table 25.5, calculate Ap w +RT In a;, . .. . .(6) from Eq. 7 at the temperature and estimated pressure. 5. If the A,u, values for Eqs. 6 and 7 are equal, then where APT is the difference in chemical potential of the p value is correct for hydrate formation. If not, pure water and ai accounts for the normally small estimate a new value of p and return to Step 2. solubility of the guest gas in the aqueous phase. Now A computer program that performs the above progiven T, p, and&, one may calculate Ap,., in Eq. 6 by cedure has been written by Ng and Robinson and is using Eq. 3 and the Kihara potential. However, the AP,~ obtained are useless until they arc matched to the AfiLM. available commercially through the Gas Processors Assn. (GPA). A second hydrate program generated by from the following equation. Erickson and Sloan I60 at the Colorado School of Mines is also available through GPA. That the model predicts the three-phase data accurately RT = RT -T RT2 po RT is demonstrated in Fig. 25.17 for the methane/propane 0 system. The dramatic decrease of dissociation pressure ..,__ ___. __ ___ __ (7) for a 1% addition of propane is evidence that the hydrate has changed structure (from I to 11)as mentioned earlier. where In other words, as small amounts of propane are added, a T,, = reference temperature. greater temperature at constant pressure is tolerated for I 0 = reference pressure, hydrate formation. Ah,,. = specific enthalpy difference, and Fig. 25.18 also demonstrates the accuracy of the AI.,, = specific volume difference. model while indicating that hydrates denude the gas of propane. At 5 atm, while the gas phase has about 8% propane, the solid phase has 50% propane. From Figs. Eq. 7 corrects a chemical potential difference at a stan25.17 and 25.18 and typical earth temperature gradients, dard T,, (0C) and I>,, (0 atm). ApcL, (T,,.p,,) for the Davidson 16 proposed the following in-situ, onshore change in temperature and pressure of the hydrate of inhydrate conditions. terest. The constants .Ip ), (T,, ./I,,) and Al? ,,. (T,,), meaxGoing toward the center of the earth, one may enured recently by Dharmawardhana and Weiler. Is counter Structure II hydrate at relatively shallow depths and parameters for Langmuir constants arc fivcn in until the point at which the hydrate denudes the gas of Table 25.5. propane. No more hydrate is encountered until the range of Structure I stability conditions is obtained. Continuing Procedure for Determining Hydrate Formation downward from the Structure I region no hydrates are Pressure. The procedure for determming hydrate formaencountered. Finally, a Structure II region is realized for tion pressures for a given gas at a given temperature IS as Eq. 5a may be combined with Eqs. 3 and 2 to obtain
lhllow\.
25-10
PETROLEUM ENGINEERING
HANDBOOK
0 -
40 TEWEFS?~E,
(WATER-PREE BASIS)
Fig. 25.17-Hydrate formation conditions propane/water system. for the methane/ Fig. 25.18-Pressure-composition diagram for the methane/ propane/water system.
Determining the Water Content of Gas (or Hydrocarbon-Rich Liquid) in Equilibrium with Hydrates
The position of lines G-H and LHc-H in Figs. 25.2 and 25.4 determines the extent to which the hydrocarbon must be dried to prevent the formation of hydrate from the gas phase, The water content in this region is relatively small and difficult to measure. Until recently, straight lines (log water content vs. l/T from the gaslliquid region were extrapolated into the gas/hydrate region with only limited experimental justification. However, as indicated by Kobayashi and Katz, I3 thermodynamics tells us that such concentration extrapolations across phase boundaries yield severe errors. This observation, that straight-line extrapolation into the gas hydrate region represents gas in equilibrium with metastable liquid water, explains the field data anomalies observed by
Q Laboratory confirmation that the Records and Seely . water content of gas in equilibrium with hydrate should be much different from the extrapolated values has been verified by Sloan et al. 7 for methane hydrates, and Song and Kobayashi 163 for methane/propane hydrate!: When one predicts the water content of a singie fluid phase, such as a fluid, in equilibrium with hydrates, the basic equation is as follows. fw.=fwH , . . . . . . . . . ... . . . . . . (8)
SYSTEMS
25-11
Fig. 25.19-Structure
Fig. 25.20-Structure
where K,,,, is the fugacity coefftcient* of water (determined by an EOS) and y, is the mole fraction of water in the hydrocarbon. The value y,,, is the solution to this problem, stated as: given a flowing hydrocarbon at T and p, determine how much the fluid should be dried to prevent hydrate formation. The quantityfWH in Eq. 8 is determined by
fwH =fwMT
exp(-Ap,IRT),
..
........
.(lO)
In this equation, all of the ice properties are well known, the Ap is obtained from three-phase data fit to Eqs. 3 and 4; the only unknown is PMT, which was fit to a number of hydrates three-phase data below 273 K and found to be a single function of temperature; Figs. 25.19 and 25.20 show these values for empty hydrate vapor pressure as determined from a number of hydrates. Recent work at the Colorado School of Mines has shown the method represented by Eqs. 9 and 10 and Figs. 25.19 and 25.20 to represent the water-in-liquid hydrocarbon/hydrate equilibria accurately. Hydrate Formation on Expansion of Gases The simultaneous solution of the isoenthalpic (throttling) expansion of natural gases with initial hydrate formation conditions, Fig. 25.9 yields a first approximation of the prediction of permissible expansions. Katz 45 presented a useful chart for various gas gravities of natural gases.
where f,,,HT is the fu acity of water in empty hydrate. Ng and Robinson k4 give an expression for fW~r in both structures, obtained by fitting the vapor/hydrate data of Kobayashi et al. 7~12i*165 Therefore, their method may be considered a correlation of existing two-phase (vapor/hydrate) data. By equating the fugacity of hydrate to ice in threephase data, Dharmawardhanat58 showed that fwMT of Eq. 10 can be expressed as an empty hydrate vapor pressure as follows:
P y~Kg exp Vl(P-Pf) RT
=P~MTK~MT exp@WW,
Definition of the Saturated Water Content of Natural Gases in Equilibrium With Aqueous Phases
The saturated water content of natural gases in equilibrium with aqueous phases generally is presented in the familiar dewpoint chart. Fig. 25.21 presents a dewpoint chart for methane-rich gases synthesized from several sources of data. The upper limit of the dewpoint chart given in Fig. 25.21 is defined by the properties of steam. Fig. 25.22 presents a correction to Fig. 25.21 caused by the hydrate formation of methane gas, and a mixture of methane and propane gas. The water content at temperatures above the hydrate stability conditions up to 10,000 psia are based primarily on the data of Olds et al. 3 At temperatures below the initial hydrate formation
. . . . .
. . . . . . . ..I.....
(11)
where
= vapor pressure of ice, = fugacity coefficient of ice, = volume of ice, = vapor pressure of ice, pvMT = vapor pressure of empty hydrate, and KIT = fugacity coefficient of empty hydrate.
*In some prior publications, 0 was used as the symbol for fugaclty ccefhcient.
pvf Kg VI pI
25-12
PETROLEUM ENGINEERING
HANDBOOK
TEMPERATURE, IOOC
44 I 4.2 I 4.0 I 30 I
103/K
36 I 3.4 ' I f-1
10303
c
lo
&y/9
oto-C!i4-C~H8(531%
mold-
Hz0
4 --*--DEW
CALCULATED FOR
1
0.01 I I I
-50
-30
-10
14
40
70
lFJ T%m F
TEMPERATURE,
water content
conditions, the data are primarily those measured by Skinner. 66 However, below the initial hydrate formation conditions, as initially observed (in the gas fields) Q Fig. 25.21 presents metastable by Records and Seely , values. Using the hydrate theory discussed previously, and recent measurements on the water content of two gases (a methanel5.3 1-mol % propane and a pure methane in the gaseous state in equilibrium with hydrates), a corrective correlation to Fig. 25.21 in the gas/hydrate region was developed. A typical replacement chart is shown in Fig. 25.22. 163 In this figure the high-temperature vapor/liquid-water region is separated from the low-temperature vapor/hydrate region by a line representing the threephase (vapor/liquid-water/hydrate) boundary. The isobaric data in the vapor/hydrate region follow semilogarithmic straight lines when plotted against reciprocal temperature, but these lines have slopes different from the straight lines in the vapor/liquid-water region. In addition, the three-phase lines are indicated to be a function of gas composition, so that the change in the slope of an isobar from the vapor/liquid region occurs at different temperatures. With the above complexities, a comprehensive water content chart (or series of charts) for gases of differing compositions would be cumbersome. Instead, a mathematical method for determining the water content of gases in the vapor/hydrate region is used.
Steps in Determining the Water Content of Vapor in Equilibrium with Hydrates The method for determining the water content of vapor in equilibrium with hydrates, at a given temperature and pressure, is composed of six steps. 1. Calculate the metastable water content at the temperature and pressure of interest by using Eq. 12. 2. Calculate the three-phase temperature at the pressure and gas gravity of interest by using Eq. 13. Obtain the temperature difference (AT) by subtracting the temperature of interest from the calculated three-phase temperature. 3. Calculate the displacement from the metastable water content (AW) at the above A f and pressure of interest using both Eq. 14a for methane and Eq. 14b for a 94.69-mol% methane/5.31-mol% propane mixture. 4. Calculate the AW for the gas composition of interest by a linear interpolation between the AW for methane (gravity =0.552) and the A W for the mixture containing 5.31% propane (gravity=0.603), using gravity as an interpolation parameter. 5. Calculate the equilibrium water content by subtracting the AW value obtained in Step 4 from the metastable water value obtained in Step 1. 6. Consider the range of data used to tit Eqs. 12, 13, 14a, and 14b, as discussed later, to determine whether the answer obtained in Step 5 is within the bounds of the correlation.
SYSTEMS
25-13
TABLE
25.6-COEFFICIENTS
FOR ECWATIONS
Eq. 12 C,
Eq. 13 + + + + + 1 3 0 5 2 2 2.7707715E - 2.782238E - 5649288E - 1.298593E 1.407119E 1.785744E l.l30284E-3 5.9728235E - 2.3279181 E - 2.6840758E 4.6610555E-3 5.5542412E - 1.472776% 1.3938082E 1.488501 DE 3 3 4 3 3 4 4 4 5 4 5 5 6
Eq. 14a - 1.605505E + 3 8.181485E+2 9.289352E + 2 - 1.578381 E + 2 - 3.899544E + 2 - 2.009926E + 1 1.368723E + 1 5.500387E i- 1 4.068990E + 0 1.517650E + 0 - 4.524342E - 1 - 2.590273E + 0 - 2.465SSE - 1 - 7.543630E - 2 1.034443E - 1
Eq. 14b 2.59097E + 3 -1.51351E+3 - 1.16506E + 2 3.26066E + 2 6.65280E + 1 - 1.17697E + 1 - 3.05SSOE + 1 - 1.20352E + 1 2.94244E + 0 7.83747E - 1 i.O4913E+O 7.23943E - 1 - 2.94560E - 1 7.087SSE - 2 - 1.24938E - 1
C,
C3 Cd
c5
2 d
c9
C 10 C,,
Equations for Determining the Water Content of Vapor in Equilibrium with Hydrates In the following equations, the pressure is expressed in psia, the temperature in degrees Rankine, and the water content is expressed in pounds of water per million cubic feet of gas at 1 atm and 60F. A listing of the coefficients in each equation is found in Table 25.6. Fit of Methane-Rich Gas (Fig. 25.21). The following equation, which is a fit of Fig. 25.21, is in the pressure range of 200 to 2,000 psia, and in the temperature range of -40 to 120F, for metastable water content, IV,,,,, as a function of temperature, T, and pressure, p: W,,,=exp[C1 +C~lT+C3(lnp)+Cq/P _. _. .(12)
fit as a function of pressure and of the temperature difference from the three-phase condition, AT, for both methane and a mixture containing 5.3 1 mol% propane. The fit for methane was in the pressure range of 500 to 1,500 psia and in the temperature range of -28 to 26F; coefficients for methane in the following equation are labelled Eq. 14a in Table 25.6. The fit for the mixture was in the pressure range of 500 to 1,500 psia, and in the temperature range of -38 to 40F; coefficients for the mixture are labelled Eq. 14b in Table 25.6. The general regression equation for both gases is:
AW=exp[C,
+C:!(lnp)+C3(ln
AT)+Cq(lnp)*
+CS(lnp)lT+Cg(lnp)Z],. where Cl .
Fit of Three-Phase (Vapor/Liquid/Hydrate) Formation Conditions. The three-phase condition was fit in the temperature range of 34 to 62F, the pressure range of 65 to 1,500 psia, and the gas gravity, y. range from 0.552 to 0.9. Only hydrocarbons were used in the fit, and gases containing CO, and hydrogen sulfide were excluded specifically. T=lI[C, +CZ(lnp)+C3(ln +y)+Cq(lnp)*
+Ct2(lnpJ3(ln +C,4(lnp)(ln
AT)2 . .(14)
+C,(ln
+C,30np) (ln +C,s(ln where CI -y)4],. .
~)~+C~14(lnp)(ln
r13 . . .(13)
Fit of Water Content Suppression. The suppression of the water content from the metastable region, AW, was
Example Calculation of Water Content of a Vapor in Vapor/Hydrate Region. Determine the water content of a gas whose gravity is 0.575 in equilibrium with hydrate at 1,000 psia and 8.4F. 1. The metastable water content of the gas is calculated from Eq. 12 at 1,000 psia and 8.4F as 2.745 lbm/106 scf. 2. The three-phase temperature at 1,000 psia is calculated to be 60.35F from Eq. 13. The AT value is 51.95F. 3. The displacement from the metastable condition is calculated for methane to be 0.653 lbm/106 scf from Eq. 14a. The displacement from the metastable condition is calculated for the mixture containing 5.31 mol% propane as 1.55 lbm/106 scf from Eq. 14b. 4. An interpolation between the values obtained in Step 3 (based on gas gravity) is done to determine the displacement for a gas with a gravity of 0.575. The resulting displacement is 1.0575 lbm/106 scf.
25-14
PETROLEUM ENGINEERING
HANDBOOK
8 0 200 -. PSIA
UN
Fig. 25.25-Water content per cubic feet of volume occupied by various gases at system pressure and temperature.
mPER4lURE,"F
Fig. 25.23-Depression of metastable temperature. dewpoint
5. The displacement in Step 4 is subtracted from the metastable value in Step 1 to obtain a water concentration of 1.687 lbm/106 scf. The 0.575gravity gas must be dried to less than 1.687 lbm/106 scf to prevent hydrate formation at 1,000 psia and 8.4F. 6. A check of the conditions of regression of Eqs. 12 through 14 indicates that the correlation should apply for the conditions of the example. Fig. 25.23 presents the corrective correlation below the ice/gas region, Fig. 25.24 gives corrections to Fig. 25.21 from the salinity of the aqueous phase. The work of Deaton and Frost indicated that the water content of gases in equilibrium with liquid water fall in the order
DISXMD XlXE,
FM
Fig. 25.24-Correction to water content for natural gas in equilibrium with brines.
where y is the mole fraction of water in each subscripted gas at the same pressure and temperature as shown in Fig. 25.25. The water content of nitrogen up to moderate pressures has been repotted by Rigby and Prausnitz. 34 No data exist, however, on the water content of nitrogen in equilibrium with hydrates. The water content of liquid and gaseous CO2 in equilibrium with hydrates has become increasingly more important in the development of EOR technology. Fig. 25.26 presents a pressure-temperature projection of the univariant loci for the COz/water system. It defines both the univariant and two-phase regions presently being studied and the regions over which the various combinations of binary equilibrium exist. Like
SYSTEMS
25-15
TABLE 25.7-VAPOR PRESSURE OF WATER USEDTOCALCULATEENHANCEMENT OF THE WATER CONTENT OF THE CO,-RICH FLUID PHASES T
ml
(OF)
0 15 30 45 65 77 87.8 122
0 ,., (psia) 0.02215 * 0.04352 0.08162 * 0.1487+ 0.30a9+ 0.4641 + 0.6518% 1.789%
T
H-%(1 lEWFMW, F
Extrapolated lo kwer temperat re from hl9h temperatures Of ** Data from Perry, R.H. and Chtlton, C H. Chemvsal Engineer s Handbook, llfth edlbon, McGraw-HIV Book Co Inc., New York City (1973). Table 3-4 Data from Tables of Thermal Propertles of Gases, U S Dept of Commerce, Washtngton DC. (1955) Circular 564, Table 9-9/a bla from Tables of Thermal Properlles of Gases. U S Dept of Commerce, Washmgton D C. 1,955) Crcular 564, Table 3-5
propane, the location of the critical point of CO* relative to the H-L,-G line results in an invariant quadruple point in which H-L,-LCO, -G coexist, the hub from the other four univariant three-phase lines emanate. A second quadruple point involving H-I-L,-G exists at lower pressure in the neighborhood of the ice point. The pioneering work of Wiebe and Gaddy provides extensive data on the solubility of CO2 in wate& and the vapor-phase composition of CO 2 /water mixtures. s* The measurements were confined to temperatures exceeding the initial hydrate formation condition. Measurement of the water concentration of gas and liquid in equilibrium with hydrates currently is being pursued by Song and Kobayashi. 37, 38 By using the vapor pressure of water or metastable liquid water given in Table 25.7, the experimental data from several sources have been presented as enhancement of the vapor pressure resulting from the CO2 pressure on the system (Fig. 25.27). In Fig. 25.27, p is the total pressure,p?,, is the saturation pressure of water, and yw is the mole fraction of water. The water content of coexisting COI-rich liquid and gas along the three-phase region is shown as a loop terminating at its three-phase critical point in the neighborhood of the critical temperature of CO2. Fig. 25.27 shows that the enhancement of the water vapor pressure is favored by lower temperatures and high pressure, variables that also favor a substantial increase in the density of fluid CO*. It is noted that the solubility of water in liquid CO* is greater than in gaseous CO1 at the same pressure and temperature, precisely the reverse of the solubility sequence for liquid and gaseous hydrocarbons, This phenomenon is related to the excessive collection of water in the middle of a distillation column separating a COz-rich feed from its associated LPG constituents. Fig. 25.28 presents the water content of CO2 vs. inverse absolute temperature along isobars demonstrating the complexity of the system. The curves show a break downward at the initial hydrate formation conditions as methane and natural gases.
Fig. 25.27-Enhancement of the vapor pressure of water caused by CO, total pressure at various temperatures.
inverse
absolute
25-16
PETROLEUM ENGINEERING
HANDBOOK
UOll
um
B I Kn
lol 0.m
o.au
O.Dlr
0.01
0.04
0.1
Fig. 25.29-The effect of molecular weight on the water content in the vapor phase (from Ref. 167).
In recent years estimations of water in both the nonaqueous and aqueous phases in both the two- and threephase regions have been carried out successfully b? Peng and Robinson, 68 Baumgaertner et al., l6 and s Moshfeghian et al. O In these exercises various EOS were used-the PREOS (Peng-Robinson equation of state), 157 the.Schmidt-Wenzel EOS, 169 and the PFGCMES (Parameter from Group Contribution-Moshfeghian, Erbar, Shariot) EOS. Collectively, these correlation procedures are able to describe vapor/liquid equilibria in systems as diverse as CHd/H*O, C~H~/HZO through n-CgH,a/H20, Nz/HzO, C021H,0, CO/H20, 02/H20, H2/H20, C~H~/HZO, air/HzO, H2S/H20, andNH3/H20. Both the PREOS and PFGC-MES EOS require temperaturedependent interaction arameters, while the method of Baumgaertner et al. I68 describes the water structure in terms of its degree of polymerization in the liquid as a function of the temperature. Examples of the successful application of the calculational method to multicomponent systems by Moshfeghian et al. I are reported.
5.6
1.6
008
TEFfMTlJRL F
Fig. 25.30-The solubility of water in various liquid hydrocarbons at their vapor pressures.
FLg. 25.31-The
SYSTEMS
25-17
Fig. 25.31 presents the volumetric solubility of methane in water to 10,ooO psia as measured by Culberson and McKetta. t7 The solubility of methane increases monotonically with pressure but shows a relative minimum with respect to temperature as do the Henry s law constants. 14* Dodson and Standing,45 on the other hand, presented the solubility of a typical natural gas in water and gave corrections for the solubility due to brine salinity, Fig. 25.32. The relative solubility of gas in brine relative to its solubility in pure water is given in Fig. 25.32. The analytical prediction of the solubility of nonpolar or low-polarity solutes in water can be used by the application of the relationships introduced by Krichevskii and Kasamovskii I and Wiebe and Gaddy,90 which were later applied by Kobayashi and Katz 14*and Leland et al. i7* to condensing light-hydrocarbon systems. The expression is given as follows.
A
p,p,,s )
(15)
= partial fugacity of the solute in water at T and pt, KH = an empirical constant that is a function of temperature only but is equal numerically to Henry s law constant for noncondensable gases and a hypothetical Henry s law constant for condensable gases, Xl = mole fraction of the dissolved solute, ^ R = gas constant per mole, VA41= partial molal volume of components in the solution, T = absolute temperature, pt = total pressure, and P 3 = vapor pressure of the solvent for a noncondensable gas, but equal to some constant pressure slightly above the threephase pressure determined from experimental data. The usual absence of values for the partial volume of the solute P,,, precludes the application of Eq. 15. However, given a good set of experimental data to moderate pressures, Eq. 15 provides a good method of extending the solubility data to higher pressures, particularly if the solute remains essentially pure at the higher pressures. Fig. 25.33 gives an application of Eq. 15 to the aqueous phase of the propane/water system. 142 Examples of the agreement in the experimental and predicted water compositions are iven for the B methane/water38*40 and ethane/water 3s 74 and propane/water 142 systems in Figs. 25.34 through 25.36 by Peng and Robinson. 16* Both the Peng-Robinson 16* and the Baumgaertner er al. 69 methods present successful predictions of the aqueous-phase concentrations as well as in the nonaqueous phase.
where f,
6.6
6.0
0.w
a.aa
0.20
0.24
o.zB
Fig. 25.33-Solubility of prop≠i water to determine partial molar volume and Henry s law constant.
PETROLEUM ENGINEERING
HANDBOOK
/ - /
W IN LIGlJlD a ul
0 2 4
s o r / vm
0,.I
6
CALCULATION I
8
I
lo
I
I2 14
NlEFlWICNWERXld
Fig. 25.36-Experimental and predicted water content of propane liquid and vapor phases along three-phase locus.
CALmTICN
"
1,O
Fig. 25.34-Experimental and predicted vapor and liquid phase compositions for methane/water system at
25ooc.
III
Ill
l -R EF1 7 3
petroleum production and processing. From an experimental standpoint the developments have been spurred by the accumulation of experimental data typically encountered in sour water processes. 75~178 Meanwhile, basic theoretical work had been laid for the interpretation of such data. 179- 82 Interpretations of these and other theories are reviewed by Friedman and Krishnan, 179 Newman, lg3 and Knapp and Sandler. Is4 Thus, predictive methods have been developed to estimate vapor/liquid equilibria in complex mixtures containing such components as NH 3 /CO 2 /H 2 S (Wilsont75), N2/C02/H2S/CH40H (Moshfeghian et al. O), NH31H2S/H20, NHs/S02/H20, and NHs/C021H20 (Renonlg5), NH~/COI/H~S/H~O (Mauererlse), CO~INaClIKClIH20 (Chen et al. lB7), and many others. tg3 The calculations now enable the estimation of physical and chemical equilibria from purely aqueous to strong electrolyte solutions.
Oil and Gas Reservoirs That Exist in the Gas Hydrate Region
In recent years an increasing number of oil and gas reservoirs in the hydrated state have been discovered by Katz, 18 St011 er al., Is9 Makogon et al., 90 Bily and Dick,19 Verma et al., 19* Holder et a[., 193 and Trofimuk er al. 194 In fact, hydrate cores have been recovered from below the ocean floor in the MidAmerican Trench off Central America. 95 A review of geologic occurrences of hydrates is presented by Kvenvolden and McMenamin. 196 In Feb. 1982 a hydrate core was recovered by the Deep Sea Drilling Project from the Mid-American Trench off Guatemala in 1718 m of water. All evidence to date indicates that these hydrates are of biogenic origin, since they are composed mainly of products from anaerobic digestion. These hydrates are of the Structure I form. In May 1983 a second hydrate core was recovered by a Getty drilling operation from the Gulf of Mexico in 530 m of water. These hydrates are thermogenic in origin with probable prior biogenesis and have substantial amounts of ethane, propane, and isobutane. These hydrates are of the Structure II form.
FOWE IN LIQUID
m.E FRWICN Fig. 25.35-Experimental and predicted vapor and liquid phase compositions for the ethanelwater system at 220F.
SYSTEMS
25-19
,,I
u1
.u-
20
1
I I i I I -I
Fig. 25.37-Effect
of GOR on hydrate formation conditions for a 46B0API oil in the presence of excess fluid brine.
Fig. 25.39-Crystallization
of aqueous diethyl-
These natural hydrates bring many imphcatlons to mind; however, only a few will be outlined. 1. The simultaneous solution of hydrate formation for the ocurrence of hydrates and earth temperature gradients yields possible environments for the occurrence of hydrates. 2. The potential reserve of natural gas existing in nature as solid hydrates is probably very enormous. 3. Petroleum reservoirs existing in the hydrate region may be denuded of a substantial amount of the lighter constituents and, thus, will be energy deficient for production by a gas-drive mechanism. 4. Waterflooding of dissolved gas reservoirs with near-freezing water may cause hydrate formation in the reservoir. 5. A tremendous amount of technological development will be necessary to unleash hydrocarbons from hydrated reservoirs.
Hydrate Inhibition
While crude oils may not inhibit hydrate formation per se, their oils with dissolved gases affect the initial hydrate formation significantly, as shown in Fig. 25.37. 97 Under some circumstances it may be advantageous to inhibit hydrate formation rather than dehydrate oil and/or gas streams. While many ionic and hydrogen-bonding substances inhibit hydrate formation, the two compounds that are used most frequently for hydrate inhibition are methanol and ethylene glycol. Natural brines cause only a few degrees of hydrate depression. 34 Figs. 25.38 through 25.40 present the freezing points of aqueous glycol solutions showing their eutectics in the absence of gas. 198 Fig . 25.41 presents the freezin point 87 depression of diethylene/glycol/water solutions.
ul
81
100
Fig. 25.38-Crystallization
of aqueous trieihyl-
PETROLEUM ENGINEERING
HANDBOOK
Component
lo
30
50
70
lElWRMW, F
Fig. 25.41-Effect of dlethy.len@ glycol hydrate fopation s. on conditions for
The inhibition effect of methanol and ethylene glycol on the hydrate formation condition of a methane/ 12-mol% propane mixture is currently under study at Rice U. Recently, Ng and Robinson have presented the hydrate formation conditions of methane, ethane, propane, COz, hydrogen sulfide, methane/10.49-mol% ethane, 90.09-mol% methane/9.91-mol% CO2, 95.01-mol% methane/4.99-mol% propane, a synthetic natural mixture, and a synthetic natural gas mixture containing CO?, each in the presence of methanol solutions of 10 to 20 wt % methanol in water. With the increasing amount of oil and gas transmission along ocean floors prior to dehydration, such data are of increasing usefulness.
n,; = number of cavities of type i per water molecule in basic lattice PO = reference pressure pvl = vapor pressure of ice pb,~T = vapor pressure of empty hydrate P S = vapor pressure of the solvent for a noncondensable gas, but equal to some constant pressure slightly above the three-phase pressure determined from experimental data To = reference temperature Av, = specific volume difference = partial molal volume of components in the V,l solution Wms = nkastable water content Xl = mole fraction of the dissolved solute Yji = fractional occupancy of type i cavity by type j molecule pi = chemical potential of Component i pi0 = chemical potential of pure Component i &Lw = difference in chemical potential of water /J,+,H= chemical potential of water in filled hydrate p,,,,+,T = chemical potential of water in empty hydrate Subscripts L = liquid water
H = hydrate
g = gas
References
I. Gibbs, J.W.: The Collected Works of J. Willard Gibbs, Volume I, Thermodynamics, Yale U. Press, New Haven, CN (1948). 2. Kobayashi, R. and Katz, D.L.: Methane Hydrate at High Pressure, Trans., AIME (1949) 186, 66-70. 3. Olds, R.H., Sage, B.H., and Lacey, W.N.: Composition of Dew Point Gas Methane-Water System, Ind. Eng. Chum. (1942) 34, 1223-27. 4. Villard, P.: On Some Hydrates, Compr. Rend. (1888) 106, 1602-03; (1888) 107, 395-97. 5. Deaton, W.M. and Frost, E.M.: Gas Hydrares and Their Relation fo the Operation of Narural Gas Pipe Lines, Monograph 8, USBM, Washington D.C. (1946). 6. Marshall, D.R., Saito, S., and Kobayashi, R.: Hydrates at High Pressures: Part I. Methane-Water, Argon-Water, and Nitrogen-Water Systems, AIChE J. (1964) 10, No. 2, 202-05. 7. Sloan, E.D., Khoury, F., and Kobayashi, R.: Water Content of Methane Gas in Equilibrium with Hydrates, Ind. Eng. Chem. Fund. (1976) 15, No. 4, 318-22. 8. Galloway, T.J. ef al. : Experimental Measurement of Hydrate Numbers for Methane and Ethane and Comparison with Theoretical Values, Ind. Eng. Chem. Fund. (1970) 9, No. 2, 237-43. 9. van der Waals, J.H. and Platteeuw, J.C.: Clathrate Solutions, Adv. in Chemical Physics, Vol. II, 1. Prigogine (ed.), Interscience Publishers Inc., New York City (1959) l-58. 10. Parrish, W.R. and Prausnitz, J.M.: Dissociation Pressures of Gas Hydrates, hi. Eng. Chem. Proc. Des. Dev. (1972) 11, No. I, 26-35. Il. Kobayashi, R.: Vapor-Liquid Equilibria in Binary Hydrocarbon-water Systems, PhD dissertation, U. of Michigan, Ann Artxv (1951).
Nomenclature
ai = activity of component i in mixture Cj; = a unique function of temperature for each guest molecule in each size cavity ?; = fugacity of j in the gas phase fl = partial fugacity of the solute in water at T
and pf
fwj = fugacity of water in the fluid phase A/z,+ = enthalpy difference K1 = an empirical constant that is a function of temperature only but is equal numerically to Henry s law constant for noncondensable gases and a hypothetical Henry s law constant for condensable gases Kg = fugacity coefficient of ice
PHASE
BEHAVIOR
OF WATER/HYDROCARBON
SYSTEMS
25-21
Compositions of Clathrate Gas Hydrates 01 12. Cady. G.H.: CHCIF,. Ccl,, C12:CI0,F, H2S. and SF,. J. P&s. Clwm. (1981) 85. No. 22, 3225-30. Metastable Equilibrium rn the 13. Kobayashi, R. and Katz, D.L. Dew Point Determutation of Natural Gases in the Hydrate Region. Truns., AIME (1955) 204. 262-63. Hydrophobic In14. Ben-Naim. A., Wilf. J., and Yaacobi, M.: teractton tn Light and Heavy Water, J. Phys. Chrm. (1973) 77, No. I, 95-102. Solubdities and Sttucturcs 15. Claussen, W.F. and Polglaae. M.F.: tn Aqueous Aliphatic Hydrocarbon Solutions. J. Am. C/w??. SM. (1952) 74. 4817-19. Phase Equilibria in 16. Culberaon, O.L. and McKetta. J.J.: Hydrocarbon-Water Systems. IV. Vapor-Liquid Equilibrium Constants in the Methanc-Water and Ethane-Water Systems. Trcrn.\. , AIME (1951) 192. 297-300. Phase Equilibria in 17. Culberson, O.L. and McKetta. J.J.: Hydrocarbon-Water Systems, III. Solubility of Methane in Water at Pressures up to lO.tXlB psia, Tiwts.. AIME t 1951) 192, 223-x 18. Culberson. O.L... Horn. A.B., and McKetta. J.J.: Phase Equilibria in Hydrocarbon-Water System\. The Solubility of Ethane in Water at Pressures to 1,200 psi. Truns.. AIME (1950) 189. l-6. 19. Davis. J.E and McKetta. J.J.: Solubility of Methane in Water, Pr/. Rejwr (1960) 39. No. 3. 205-6. 20. Dut fy. J.R., Smith. N.A., and Nanv, B.: Solubility of Natural Gases in Aqueous Salt Solutton;~ I. Liquid Sur&cs in the System CH, -H ?0-NaCI-CaCl 2 at Room Temperatures and Pressure\ beloti 1000 Lb.!Sq.in abs.. Gcmilirtl. Co.w~r~c~/~i~t~. Acrcr (1961) 24, 23-31. Salting Out Effect\ and Ion 1-I. Eucken. A. and Henzberg. G.: Hydration. Z. P/r!\;,! Clwrrt (1950) 195. l-23. The Solubility of Heltum and 22. Feillolay. A. and Lucas. M.. Methane in Aqueous Tetrabutyl Ammonium Bromide Solu~tons at 25 and 35 : J. Phrr C/KVL 11972) 76. 3068-72. The Solubility of Methane in Water 23 Fischer, F and Zerbe. C: and Organic Solvents under Prcssurc. Bre,7ft\r,,~Chcfri. ( 1923) 4, 17-I ). Solubilitie\ of Gases in Liquids at High 24. Frolich. P.K. er cl/.: Pressure. Irrd. G~R. Cltcwt. (1931) 23. 518-50. G.. Casaie, C.. and Neuri. G.: Liquid-Vapor 25. Grew. Equiltbrium at Elevated Pressures of One Component in the Presence of Noncondensable Components, /~r~/~c.~f~~r C/rim B&y (1955) 20. Spec. No. 251-57. 76. Harder. A.H. and Holden. W.R.: Measurement of [Methane] Re.wcrn h ( 1965) 1, Gas in Ground Water. W~IW Re.\r~urc~c~.\ No. I, 75-82. The Solubtltty ot hlethane 27. Lannung. A. and Glaldbaek. J.C.: in Hydrocarbons. Alcohols, Water and Other Solvents. ,4c,r<r Chrrn. Swnd. ( 1960) 14. I 124-28. C.: Solubility in Water ol C, C, Hydrocarbons. 28. McAuliffe. /Ware (1963) 200, 1092-93. A.. Grever, I.. and Bijl. A.: The lnlluence of 2 ). M)chel\. Prcssurc on the Snlubilities of Gases. PIt~.\iw (1936) 3, 797-808. Parttal Molar Volumes of Gas at lnftnitc 30. Moore. J.C. rt (I/.: Dilution in Water at 298.15 K. J. Chrrn Eq. Lkrrcr (1982) 27, 22-24 T.J. and Billctt. F.: The Salting Out of Non31. Mornson. electrolytes. Part II. The Efl ect olVariation in Non~electrolyte. J. Chmr. SM. ( 1952) 3819-22. Determinatton o! CH, tn 32. Na\one. R. and Fenninger. W.D.: Water by Gas Chromstography. J. Am Wurer Wwk\ A!.s/x ( 1967) 59, No. 6. 757-59. Sullivan. T.D. and Smith. N.O.: Soluhility and Partial 33. O Molar Volume of Nitrogen and hlethane tn Water in Aqueous Sodium Chlortde from SO to 125 dey and I00 to 600 am. J P/IS\. Chcwr. (lY70) 74, No 7. 1460-66. Solubdity of Water in Corn34. Rtgby. M. and Prauwitz, J.M.: messed Nitrozen. Arcon. and Methane. J. Phw Clxw. (1968) 72. No. I. 330-34. c Gas Solubility tn Water, .~ 35. Schroeder. W.: ~rtlr~l~~i.~.\e,,., ih(lfiPt1 (1968) 55. No. I I, 542. 36. Schroeder. W.: Observations on Solutnms 01 Gavz\ in Ltq uds: Z. Mlntrfor,vc,l7rt,1r: (1969) B24. 500-0X. 37. Show. S.K.. Walker. R.D., and Gubbins, K.E.: SaltingOut of Nonpolar Gase\ in Aqueous Potassium Hydroxide Solutions. J. Ph\s. Clwni. (1969) 73. 3 12-17
38. Sultanov, R.G., Skripka, V.G., and Namiot, A.Y.: Moisture Content of Methane at High Temperatures and Pressutes, Gazov. Prom. (1971) 16, No. 4, 6-8. Phase 39. Sultanov, R.G., Skripka, V.G., and Namtot, A. Y.: Equilibriums and Critical Phenomena in the Water-Methane System at Increased Temperatures and Pressures, .??I. Fiz. Khim. (1972) 46, No. 8, 2160. Solubility 40. Sultanov, R.G., Skripka, V.G., and Namiot, A.Y.: of Methane in Water at High Temperatures and Pressures, Gazov. Prom. (1972) 17, No. 5, 6-7. Thermodynamics of Hydmcar41. Wen, W.-Y. and Hung, J.H.: bon Gases in Aqueous Tetraalkylammonium Salt Solutions, J. Phys. Chem. (1970) 74, 170. The Solubility of Gases in Water, Ber. Dar. 42. Winkler, L.W.: Chem. Ges. (1901) 34, 1408-22. Solubility 43. Yamamoto, S., Alcauskas, J.B., and Crazier, T.E.: of Methane in Distilled Water and Sea Water, J. Chem. Eng.
25-22
PETROLEUM
ENGINEERING
HANDBOOK
Density, Viscosity. and Carbon Dioxide Solubility and Dtffusivtty in Aqueous Ethylene Glycol Solutions, J. Chem. Eng. Data (1971) 16, No. 2, 143-46. and Ritchie, P.D.: Com66. Houghton, G., McLean, A.M., pressibility, Fugacity. and Water Solubility of Carbon Dioxide in the Region O-36 atm., O-100 , Chem. ERR. Sci. (1957) 6. 132-37. 67 Khitarov, N.1. and Malintn, S.D.: Experimental Characteristics of a Part of the System H ?O-CO? , - Gwkhimi.w (1956) 3. 1827. Khitarov, N.I. and Malinin. S.D.: Phase Equilibrium Relatmn tn the System H,O-CO2 , Geokhimi.va (1958) 7, 678-79. Krichevskti, J.1.. Zhavoronkov, N.M.. and Aepelbaum. V.A.: Measured Solubilities of Gases m Liquids under Pressure. I. Solutions of Carbon Dioxide in Water and Mixtures with Hydrogen at 20. and 30C and Pressures to 30 Kg/cm . Z. l hw. Chum. (1936) A175. 232-38. Kunenh. W.: Solubtltty of CO> and Hz0 in Certain Solvents, Phys. Rev. (1922) 19. 512-24 Loprc\t. F.J.: A Method for the Rapid Determination of the Solubillty of Gases in Liquals at Various Temperature\. J. Ph\\. Chr,,~ (1957) 61. 112830. Moss. 0. and Mcnnie. J.H. Aherrationr from the Ideal Gas Laws in System\ of One and Two Components. Prw. ~ Roy. Sot.. London (lY26) Al IO. IYX-232. Malinm. SD. and Savcl eva. N.I.: Carbon Dtoxide Solubility in Sodturn Chloride and Calctum Chloride Solutmnrat Temprrnturcs ot 25. SO. and 75 deg. and Elevated Carbon Dioxidc Pre~rum. Gn,khinliw (1072) 6. 643-53. Malmin. S.D.. The System Water-Carbon Dioxide at High (19.59) 3. 23555 Temperatures and Prewres. Gcoihi,rliw Malmin. SD : Soluhiliry of Carbon Dioxide in Water at Low Pattial Pressure\ at High Tempcraturc~. l>r~ri.\ . ?~~~w/~r~l~. L-A.\[J. Tdlr. Miiicrirl. P~W/J,(v. ( 197 I) 8. 229-34 Markham. A.E. and Kohe. K.A.: Soluhility of Carbon Diowidc and Nitrous Oxide tn Aqueous Salt Solutions. J. Am. Chrw. .Siu. (IYJI I 63. 44Y-54 Matou\. J (11rll.: Soluhtlity of Carbon Dioxtdc In Water at Preswres up to 40 atm.. C,l//CC1. C:ech chcv?l Co!n,riwi. (1969) 34. No. 12. 3982-85 McCay. R C . Scely. D.H. Jr and Gardner. F.H.: Dtstrthutam of Gaseous Solutes hetwcn Aqueou\ dnd LtqmdH!drocarhon Phases Carbon Diowtde. hid. 01g. c-hrfrr. I 1049) 41. 1377-80. Murray. C.N. and R&y. J P Soluhility of Gaw\ in Disttlled Water and Sea Water. 1V. Carbon Dwxtde. Dwp-SW RI,\. O~cw,,o,~, . 4h.\rr. 11971) 18. No. S. 5.3341 Nczdonninoga. N.A. Soluhiltty of Carbon Dtoxidc in Water. /:I, ~krrl. h trd Ann. SSSR. Ser. T&. M/rck (19681 21. No. 3. I I-17. Pollitzer. F. and Strehel. E.: The lntluence of an lndtlierent Cd\ on the Saturation Vapor Concentratwn oi a Liquid, Z. Ph~sii. Chwi. (1924) 110. 768-85 Sunder. W.: The Soluhility of CO, in Water and Other S~>lw tions at Higher Pressures. Z. Phrsik. Chrw. (19121 78. 513-4 ). Stewart. P.B. and Munjal. P.K.: Solubility of Carbon Dtoxidc m Dtstilled Water. Synthetic Sea Water. and Synthettc Sea-Water Concentrates. U. otCalif.. Scc, W~rrrr Conwr.~. ( 1969) 69. No. 2. 44. Stewart, P.R. and Munlal. P.K.: Solubility of Carbon Dioxide m Pure Water, Synthetic Sea Water Concentrates at -5 deg. to 25 deg. and IO- to 45attn. Pressure. J. Chrwr. EJI,~. Dufrr (1970) 15. No I. 67-7 I. Takenouchi. S. and Kennedy. G.C.: The Binary System HIOCO, at Hugh Temperatures and Pressure\. Aar. J. Sti. ( IYti) 262. 1055-74. Todheide. K. and Franck. E.U.. .Two Phase Range and the Critical Curve in the System Carbon Dmxtdc-Water up to 3500 bar. 2. Ph!~irC Chcv~~.(19631 NF37. 387-401 Second Vtrial Coefficient\ Vandcrzee. C E. and Haa\. N.C.. B II for rhc Gas Mixture (Carbon Dioxide + Water) from 300 to IO00 K. J. Cbwr. T/lrr7~lor/~,~~,,lir.( (10X I , 13. No 3. 203 I I Vilcu. R. and Galnar. I.: Soluhtlity of Case\ tn Llqutd Under Prcswre I. Carbon Dioxide-Water System. Rr,i, Kracar. #tini (1967) 12. No.2. 1814. Wets\. R.F Carbon Dtoxidc m Water and Sea Water. h trrim, Chrm ( 1974, 2. No. 3. 203m I S Wiehe. R and Gaddy. V L The Soluhiltty in Waterot Cnrhon
91.
92.
68 69
93.
94.
95.
70 71
72
73
99.
74 7s
100.
76
101.
102.
77
103. 104
78
70
105.
X0
106
Xl
107
82 83
I08
109.
x4
110.
85
III.
X6
112.
x7
113.
88
114.
89 00
Dioxide at 50, 75 and 100 at Pressures to 700 Atmospheres. J. Am. Chem. Sot. (1939) 61, 315-18. Wiebe. R. and Caddy. V.L.: The Soluhihty of Carbon Dtoxtde in Water at Various Temperatures from 12 to 40 and at Pressures to 500 Atmospheres, Critical Phenomena, J. Am. Chem. Sm. (1940) 62, 815-17. Wiebe, R. and Caddy, V.L.: Vapor Phase Compositton of Carbon Dioxide-Water Mixtures at Various Temperatures and at Pressures to 700 Atmospheres, 1. Amer. Chem. Sm. (194 I) 63, 475-77. Wmblewski, S.: Investigation of the Absorptton of Gases in Ltquids at High Pressure. I. Carbon Dioxide in Water, ARII. PhyJik. Chem. (1882) 17,103-28. Yeh, S.-Y. and Peterson, R.E.: Solubility of Carbon Dioxide, Krypton, and Xenon in Aqueous Solution, J. Pharm. Sri. (1964) 53, No. 7, 822-24. Zawisza, A. and Malesinska, B.: Soiubility of Carbon Dtoxtde in Ltquid Water and of Water in Gaseous Carbon Dioxide in the Range of 0.2 to 5 MPa and at Temperature up to 473 K, 1. Chem. Eng. Dam (1981) 26, No. 4, 388-91. Zel evskii, Y.D.: The Solubility of Carbon Dioxide under Pressure, J. Chem. Ind. (USSR) (1937) 14, 1250-57. Paatella, A.A., and Sagramora, G.: Solubilities of Liqutda in Gases, Ricerca Sci. (1959) 29, 2605-13. Adeney, W.E. and Becker, H.G.: The Determination of the Rate of Solution of Atmospheric Nitrogen and Oxygen in Water, ScY. Prmc. Roy. Dublin Ser. (1919) 15, 609-28. Bartlett, E.P.: The Concentration of Water Vapor in Compressed Hydrogen, Nitrogen and a Mixture of These Gases in the Presence of Condensed Water, J. Am. Chewy SW. (1927) 49, 65578. Basset, I. and Dade, M.: Solubility of Nitrogen in Water at High Pressures to 4500 Kg/cm2, Cumpr. Rend. (1936) 203, 775-77. Benson, B.B. and Krause, D.J.: Empirical Laws for Dilute Aqueous Solutions of Nonpolar Gases, J. Chem. Phys. (1976) 64: 689-709. Benson, B.B. and Parker. P.D.M.: Relations among the Solubilities of Nitrogen, Argon, and Oxygen in Dtstilled Water and Sea Water, J. Phys. Chem. (1961) 65, 1489-96. Douglas, E.: Solubilities of Oxygen, Argon and Nttrogen m Distilled Water, J. Phys, Chem. (1964) 68. 169-74. Farhi, L.E.. Edwards, A.T.W., and Homma, T.: Determination of Dissolved N, in Blood by Gas Chromatography and (a-A) N, Difference, J. Appl. Physiol. (1963) 18, 97-106. Fox, C.J.J.: On the Coefficients of Absorption of Nitrogen and Oxygen in Distilled Water and Sea-Water, and of Atmospheric Carbonic Acid in Sea-Water, Truns., Faraday Sot. (1909) 5, 68-87. Goodman, J.B. and Krase, N.W.: Solubility of Nitrogen m Water at High Pressures and Temperatures. tnd. Eng. Chem. (1931) 23. 401-4. Klots, C.E. and Benson, B.B.: Solubilities of Nitrogen, Oxygen, and Argon in Distilled Water, J. Marine Res. (1963) 21, 48-57. Maslennikova, V.Y.: Solubility of Nttrogen m Water, Tr. Gas. Nauch. -Is&d. Proekt. Insr. Azotn. Prom. Prod. Org. Sin. (1971) 12, 82-87. Maslennikova, V.Y., Vdovina, N.A., and Tsiklis, D.S.: Solubility of Water in Compressed Nitrogen, Zh. Fiz. Khim. (1971) 45, No. 9, 2384. Murray. C.N. and Riley, J.P.: Solubility of Gases in Distilled Water and Sea Water. II. Oxygen, Deep-Sea Res. Oceanography Abstr. (1969) 3, 31 l-20. Pray, H.A., Schweickent, C.E., and Minnich, B.H.: Solubility of Hydrogen, Oxygen, Nitrogen, and Helium in Water, Ind. Eng. Chem. (1952) 44, 1146-51. Saddington, A.W. and Krase, N.W.: Vapor-Liquid Equilibria m the System Nitrogen-Water, J. Am. Chem. Sot. (1934) 56. 353-61. Smith, N.O., Keleman, S., and Napy, B.: Solubility of Natural Gases in Aqueous Salt Solutions. 11. Nitrogen in Aqueous NaCI, and MgSO, at Room Temperatures and at CaCl z. Na,SO, Pressures below loo0 Ibisq. in. abs. , Geochim. Cosmochim. Acra (1962) 26, 921-26. Suciu, S.N. and Sibbitt. W.L.: Study of the Nitrogen and Water and Hydrogen and Water Systems at Elevated Temperatures and Pressures, U.S. Atomic Energy Commission, Washington, ANL (1951) 2, 4603.
SYSTEMS
25-23
I IS
116.
117.
I I8
I I9 120. 121
122.
I23
I24
Tsiklij, D.S. and Maslenntkwa, V.Y.: Limited Mutual .Solubtlity ot Gases in the Water-Nitrogen Systent. Doi/. Ahrrd. , vhh SSSR (1965) 161. 645-47. Wiebe. R.. Gaddy. V L.. and Heins. C.: Solubtltty ofNitrogen tn Water at 25C from 25 to 1000 Atmospheres. /Ed. En#. c-/wt (lY32) 24. 927. Wiebe. R.. Gad&. V.L.. and Heins, C : The Solubilitv of Nttrogen in W&r dt 50. 75 and 100 from 25 to IOOO Atmospheres. J. Am. Chem .%K. (1933) 55, 937-55. Wilcock. R.J. and Battino. R. Solubiltty of Oxygen-Nttrogen London (1974) 252, No 5484, Mtrturc in Water. /v&w, 614-15. Winkler, L W. The Solubility of Gases in Water. Be!-. I)cirr.) Chu~r. Cc.\ (1891) 24. 3602~10: 89101. Winkler. L.W.: Measurements of the Abatrrptton of Case< tn Ltqutd\. z. P/tr.v~. C/wtr. (1892) 9. 171-75. Aokagi. K., 12,II/.: Improved Me~aurements and Correlation of ths- Water Content ot Methane Ga\ m Equtlibrium wtth Hydrate. paper prewtted at the 1979 CPA Annual Conference. Denver Water Content ot Hydrater (11 Byk. S S.. and Fomina. V I : Gases at Different Temperature. Zh. Fi:. Khim. (1978) 52. No. 5. 1306-08. Experimental Determination Falabella. B J. and Vanpee. M.: of Ga\ Hydrate Equilibnunt bcluw the Ice Point. /~zd. Gig. Chcm. hrrrd (1974) 13. No. 3. X8-31. Frwt. E M and Dcoton. W M.: Gas Hydrate Compositmn and
133.
Equilibrium Data, Oi/ and Gus J (1946) 45, No. 12. 170-78. 125. Glew, D.N.: Aqueous Solubility and the Gas Hydrates. The Methane-Water S&m, .I. fhys. Chm. (1962) 66, 605-09. 126. Hammenchmtdt, E.G.: Formation of Gas Hydrates in Natural Gas Transmission Lines, Ind. ERR. Chem. (1934) 26, No. 8, 851-55 127. McLeod. H.D. Jr. and Campbell, J.M.: Natural Gas Hydrates at Pressures to 10,ooO psia, J. Pet. Tech. (1961) 13, 590-94. 128. Roberts, O.L.. Brownscombe, E.R.. and Howe. L.S.: Constitution Diagrams and Composition of Methane and Ethane Hydrates, Oil and Gas J. (1940) 39, No. 30. 37-40. 129. Snell, L.E., Otto. F.D.. and Robinson, D.G.: Hydrates in Systems Containing Methane, Ethylene, Propylene, and Water, AIChE J. (1961) 7, No. 3. 482-85. 130. Ballard, D.: How to Operate a Glycol Plant. Hydrocurbon Process. Per. Refiner (1966) 45, No. 6. 171-80. 131. Deaton, W.M. and Frost. EM.: Gas Hydrates in Natural Gas Pipe Ltnes, Oil and Gas J. (1937) 36, No. I. 75-81. 132. Hammerschmtdt. E.G.: Preventing and Removmg Hydrates in No.5.30-40. Natural Gas Pipe Lines. Gas (1939) 15, 133. Trebin, F.A. and Makogan, Y.I.: Process of Hydrate Formation tn Natural Gas. Condttions of Formatton and Decomposttion of Hydrates, Tr. Mask. Insr. Nefrekhim. i Grrz. Prom. (1943) 42, 196-208. 134. Kobayashi, R., e( ui.: Gas Hydrate Formation with Brine and Ethanol Solutions, Proc., Nat. Gas Assoc. Am. (1951) 27-31. 135. Wilcox, W.1.. Carson. D.B., and Katz, D.L : Natural Gas Hydrates, Ind. Eng. Chem. (1941) 33. 662-65. 136. Larson, S.: Phase Studies of the Two-Component Carbon Dioxide-Water System Involving the Carbon Dioxtde Hydrate, PhD thesis, U. of Illinois. Urbana (1955). 137. Song, K.Y. and Kobayarhi, R.: The Water Content of CO?-Rtch Fluids in Equilibrium with Liqutd Water and/or Hydrate, Research Report, GPA, Tulsa (Sept. 1983). 138. Takahashi. S., Sony, K.Y., and Kobayashi, R.: Availability and Deficiencies in Thermodynamic Data Needed for the Design of Glycol Dehydrators for CO.-Rich Fluids. paper 67e presenied at the 1983 AlChE Summer National Meeting, Denver, Aug. 28-3 I. 139. Unmh. C.H. and Katz, D.L.: Gas Hydrates of Carbon Dioxide-Methane Mtxtures, Trans., AIME (1949) 186, 83-86. 140. van Cleeff, A., cr ui. : Studies of the Ternary System EthyleneEthanol-Water. II Formation of Ethylene Hydrate. BwnnstoffChem. (1960) 41, 55-57. 141. Alder, S.B. and Spencer, C.F.: Case Studies of Industrial Problems, Phase Equilibria and Flutd Properties in the Chemical Industry, Pm-. , Equtltbrium Fluid Properties in the Chemtcal Industw (1980) 465-95. Vapor-Liquid Equilibna for 142 Kobayasht, R. and Katz, D.L.: Binary Hydrocarbon-Water Systems. Ind. Eng. Chrm. (1953) 45.440-51.
Phase Equilibria of Natural Gas Hydratca. paper 144. Sloan, E.D.: 67f presented at the 1983 AIChE Summer Natl. Mretine. Denver. Aug. 2X-3 1, 145. Katz. D.L.: Prediction of Conditions for Hydrate Formatton m Natural Gases. Twts., AIME (1945) 160. 140-49. 146. Katz. D.L. cttrl. : Water-Hydrocarbon Systems. Hu~z&x& o/ Nrrrnru/ GQS En,,e~neurin~. McGraw-Hill Book Co. Inc., New York City (1959) 189-221. 147. Canon, D.B. and Katz, D.L.: Natural Gas Hydrates. Trtirt.>.. AIME (1942) 146. 150-59. 148. Noaker. L.J. and Katz. D.L.: Gas Hydmtes of Hydrogen Sulphide-Methane Mixtures, Trcins., AIME (1954) 201. 237-39. 149. Wu. B.-J.. Robinson, D.B., and Ng. H.-J : Three- and FourPhase Hydrate Fotming Conditions in Methane-lsobutaneWater, J. Chrrrr. 77wrmodynomit:u (1976) 8. 461-69 150. Robinson, D.B. and Ng. H.-J.: Improve Hydrate Predictions. Htdrocar/m Proc. (Dec. 197.5) 54, No. I?. 95-98. 151. Platteeuw, J.C. and van der Waals, J.H.: Thermodynamic Properties of Gas Hydrates. II Phase Equilibnum in the System, Rpc. Tim . Chirn. (1959) 78, 126-33. 152. van Stackelberg. M.: Solid Gas Hydrates, Ndrur~r,i.~.~[,,2sc,hufrpn (1949) 36. 327-33, 359-62. 153. van Stackelberg. M. and Mullcr. H.G.: On the Structure ofGas Hydrates. J. Char. fhys. (I951 I 19. 1319-20 154. Saito, S. and Kobayashi, R.: Hydrates at High Pressurns: Part III. Methane-Argon-Water, Argon-Nitrogen-Water Systems, AIChE J. (1965) 11, No. I, 96-99. 155. Nagata, I. and Kobayashi, R.: Calculation of Dissociation Pressures of Gas Hydrates Using the Kihara Model, Ind. Eng. Chem. Fund. (1966) 5, 344-48. 156. McKay, V. and Sinanoglu, 0.: Theory of Dissociation Pressures of Some Gas Hydrates, J. Chem. fhys. (1963) 38, No. 12, 2946-56. 157. Peng, D.-Y. and Robinson, D.B.: A New Two-Constant Equation of State, Ind. Eng. Chem. (1976) 15, 59-M. 158. Dharmawardhana, P.B.: The Measurement of the Thermodynamic Parameters of the Hydrate Structure and Application of Them in the Prediction of Natural Gas Hydrates, PhD dissertation, Colorado School of Mines, Golden (1980). 159. Weiler. B.E.: Experimental Determination of the Thermodynamic Parameters of Structure II Hydrate Using Propane, MS thesis, Colorado School of Mines, Golden (1982). 160 Erickson, D.D.: Development of a Natural Gas Hydrate Prediction Computer Program, MS thesis, Colorado School of Mines, Golden (1982). 161. Davidson, D.W.: Thermodynamtc Aspects of Natural Gas Hydrates, paper presented at the ClC Conference, Ottawa (June 1980). 162. Records, J.R. and Seely, D.H. Jr.: Low Temperature Dehydration of Natural Gas, Trans. AIME (1951) 192, 61-68. 163. Song, K.Y. and Kobayashi, R.: Measurement and Interpretation of the Water Content of a Methane-Propane Mixture in the Gaseous State in Equilibrium with Hydrate, Ind. Eng. Chrm. Fund. (1982) 21, No. 4, 391-95. 164 Ng, H.-J. and Robinson, D.B.: A Method for Predicting the Equilibrium Gas Phase Water Content in Gas-Hydrate Equilibrium, Ind. Eng. Chem. Fund. (1980) 19, No.1, 33-36. 165. Aoyagi, K. and Kobayashi, R.: Repon on the Water Content Measurement of High Carbon Dioxide Content Simulated Ptudhoe Bay Gas in Equilibrium with Hydrates, Proc., 50th Annual GPA Convention, New Orleans (1978). 166. Skinner, W. Jr.: The Water Content of Natural Gas at Low Temperatures, MS thesis, U. of Oklahoma, Notman (1948). 167. McKetta, J.J. and Katz, D.L.: Phase Relationships of Hydrocarbon-Water Systems, Trans.. AIME (1947) 170. 34-43, 168. Peng. D.-Y. and Robinson, D.B.: Two- and Three-Phase Equilibrium Calculations for Coal Gasification and Related Pmces, Thermodynamics of Aqueous Systems wirh Indusirial Applica~ions, S.A. Newman (cd.) Symposium Series 133, ACS (1980) 393-414. 169. Baumgaertner, M., Mootwood, R.A.S., and Wenzel, H.: Phase Equilibrium Calculations by Equation of State for
25-24
PETROLEUM
ENGINEERING
HANDBOOK
170
171.
172
173.
174. 175.
176 177
178
179.
180.
181.
182.
183. 184.
185.
186.
187.
188. 189.
190
191
192.
Aqueous Systems with Low Mutual Solubdlty, 7&rmodynamics of Aqueous Swrwrs with lndustrral Apphcut~on.~. S.A. Newman (ed.), Symposium Series 133. ACS (1980) 415-34. Moshfeghlan, M., Shariot, A., and Erbar, J.H.: Application of the PFGC-MES Equation of State to Synthetic and Natural Gag Systems, 7Fwrmodynumic.s <d Aqueous Systems with Inductrio/ Applitutions, S.A. Newman (cd.), Symposium Series 133, ACS (1980) 333-59. Ktichevskli. 1.R. and Kasamovskii, J.S.: Thermodynarrucal Calculations of Solubihties of Nitrogen and Hydrogen in Water at High Pressures. J. Am. Chrm. Sot. (1935) 57. 2 168-7 1. Leland. T.W. Jr., McKetta. J.J.. and Kobe. K.A.: Phase Equilibrium in I-Butene-Water System and Correlation of Hydrocarbon-Water Solubility, Ind. Enx. Chern. (1955) 47, 1265-71 Culherson, O.L. and McKetta. J.J.: Phase Equilibria in Hydrocarbon-Water Systems. 11. The Solubility of Ethane in Water at Pressures to 10,000 psi. Truns., AIME (1950) 189, 319-22. Reamer. H.H. et crl. : Composttion of Dew-Point Gas in Ethane-Water System, Ind. Eng. Chrm. (1943) 35, 790-93. Wilson, G.M.: A New Correlation of NH,. CO, and H?S Volatility Data from Aqueous Sour Water Systems, API Data Eooli, API, Dallas (Feb. 1978). Otsuka, E.: Measure Vapor Liquid Equilibria of Kogyo Kogcrke Zashi (1960) 63, 1214. NH,-CO,-H,, Yasunishi, A. and Yashida, F.: Solubility of Carbon Dioxide in Aqueous Electrulyte Solution, . 3. Chem. Eng. Data(1979) 24, 11-14. Malinin, S.D. and Kumvskaya, N.A.: Soiubility of CO2 in NaCI-H,O Solutions at Elevated Temperatures and Pressures. Go. Chem. Anal. Chem. (1975) 4, 547-50. Friedman, R.L. and Krishman, C.V.: Water-A Comprehensive Treatise, Vol. S, F. Franks (ed.) Plenum Press, New York City (1973) 55-59. Pitzer. K.S.: Thermodynamics of Electrolytes 1. Theoretical Basis and General Equations, J. Phys. Chem. (1973) 77. 268-77. Pitrer. K.S. and Mayorga. G.: Thermodynamics of Electrolytes. Il. Acuvity and Osmotic Coefficients for Strong Electrolytes with One or Both Ions Univalent. J. Phys. Chem. (1973) 77, No. 19, 2300-08 Pitzer. K.S and Kim. J J. Thermodynamics of Electrolytes. IV. Activity and Osmotic Coefficxnts for Mixed Electrolytes, J. Am. Chem. So,. (1974) 96, 5701-07. Thrrmodwumics of Aqueous System5 wrth hdustn ul Appliutions, S.A. Newman (ed.), Symposium Senes 133, ACS (1980) Knapp, H. and Sandier, S.: Phase Equilibna and Fluid Propcrties in the Chemical Industry. paper presented at the 1980 Second Intl. EFCE Conf., West Berlin, March 17-21. Renon, H.: Representation of NH,-H,S-H,O, NH ,-SO,H,O, and NH,-CO?-H?O Vapor-Liquid Equilibrium, Thrrmodynamics of Aqueous Systems with Industrirrl Applications. S.A. Newman (ed.). Symposium Series 133, ACS (1980) 173-86. Mauerer. 0.: Einnchtung zur Messung des Wasserdampfanteils in Gasen. Auslegcxhrift DE 2713617 B2. Deutsches Patentamt (1981). Chen. C.-C.. Bntt, H.I., Boston, J.F.. and Evans, L.B.: The New Actlvlty Coefficient Models for the Vapor-Liqutd Equilibrium of Electrolyte Systems. Thrrmod~~mtnrc.\s of Aqueous Systems with lndustriul Applications. S.A. Newman (ed.) Symposium Series 133. ACS (1980) 61-89. Katz, D.L.: Depths to Which Frozen Gas Fields (Gas Hydrates) May be Expected. 3. Pet. Tech. (1971) 23$ 419-23. Stall, R.D., Ewing, J., and Bryan. G.M.: Anomalous Wave Velocities in Sediments Contaming Gas Hydrates. J. Geophy.\. Res. (1971) 76, 2090-94. Makogon, Y.F. PI ui. : Detection of a Pool of Natural Gas !n a wk Solid (Hydrated Gas) State. (in RussIan) D&i. AX&. A SSSR (1971) 196. 203. Blly, C. and Dick, J.W.L : Naturally Occumng Gas Hydrates m the Mackenzie Delta. N.W.T.. Bull.. Cdn Pet. Geol. (1074) 22, No. 4, 340-52. Venna. V.K et crl.: Dcnudlng Hydrocarbon Liquids of Natural Gas Conatltuents by Hydrate Formation, J. Per. Tcvh. (Feb. 1975) 223-26.
IY.3 Holder, G.D.. Katz. D.L.. and Hand. J.H.: Hydrate FormatIon in Subsurface Environments, &f/l AAPG (1476) 60. No. 6. 081-X8. 194. Trofimuk. A.A.. Cherskii. N.V . and Tsarw. V.P. Ga\ Hydrates-New Sources of Hydrocarbon\. Priodu. Moscow (1979) 1. 18-27. 105. Ceo/irnrt (1979) 24. No. 12. 18-10. 1%. Kvenvolden. K.A. and McMenamin. M.A.: Hydrate\ ot Natural Gas: A Review of Their Geologic Occurrence. Gr,,/. Sun,. Circ. (1980) 825. 14. 197. Scauzdlo. F.R.: Inhibiting Hydrate Formatlow in Hydrocarbon Gases, Chem. Eng. Progr. (1956) 52, No. 8. 324-2X. 198. Gab Conditioning Fact Book, Dow. Chemical Co., Mldland, Ml ( 1962) 69-7 I. 100. Ng. H.-J. and Robinson. D.B.: Equilibrium Phase Composilions and Hydrating Conditions m Sy\tcm\ Containmg Methanol. Light Hydrocarbons. CO? and H: S, Rese,mh Report, GPA. Tulsa (Feb. 1983). 200. Salto, S., Marshall. D.R.. and Kobaya\hi. R. Hydrates at High Preswre Part Il. Application of Swi\tical Mechanic3 to the Study o(the Hydrates of Methane. Argon. and Nitrogen. .AIC/rEJ. (1964) 10. No. 5. 734-40. A. and Diepen. G.A.M Gab Hydrate\ olNitrogen 201. \an Cl&f, and Oxygen, Rec.. True,. Chim. ( I9601 79, 5X2-86 207. Kobnya\hi. R. er cri.: Final Rcwrcd Report on the Water Content of C,lrbon Dioxtde Gas and Llquld Equlllbrlum with Liquid W&r and with Gas Hydrates. Rcpori to ARC0 Oil and Ga\ C(1.. Dallas (June 1979).
General References
Water/Volatile
Acctylenc/Water -. Billitrct-. I.: The Acid Nature of the Acetylcnc\. ~lY)oZJ 30. 5x-44. Z. P/IV\. C/wm
Gas Systems
bled. K.M. and Golynets. Y.F.: Soluhility of Acctylenc I ,Aqueou\ Solutwns ot Electrolytes in Relation to the Tcmpcraturc and Conxntration of Salt. /x l+.ct/r. U&/w. Zolr,t/. K/,r,ri. /+c~h,ir~/.( IYSY 1 2. 173. HiraoL,l. quid\. H.: The Solubilities of Compre\wd Acetylens Ga\ in Li1. Water. f?c ~. Ph!,. C&w ./[xr. (105-l) 24. Ii.
SW also Rel . 30. EthylcneiWater Br,IJhul>. E.J. ( I (I/ : Soluhility of Ethylene in Water-Ettcct ot Temperature and Prcssurc. hid. EQ Clw~l. j lY52) 44. 2 I I - 17. D,lvi\. J.E. and McKetta. J J. Solubdlty Chetu. Eu,y. Durrr (lY60) 5. 371-75. of Ethqlcnc in Water. J.
Phahc Equtlibrium of Water~Propene Sonctw. M and Lentr. H.: and Wdtcr-Ethcne Sy\~em\ at High Tcmpcraturc\ ,~nd Pw\w~-cs. Hiri/ 7iwp -Hi,qh Pre.$ww\ ( lY73) 5. No. 6, 68Y~YY. See al\o Rcl\ 30. 3 I, and 97
Dwpcn. G.A.M and Schcffer. F.E.C. The Soluhdltj (11 Water ,n Supercntul Ethane. R<,c. l>iw C/r;,,, ( IY50) 69. 6OJ~OY.
SYSTEMS
25-25
Gjaldbaek. J.C. and Niemann. H.: The Soluhllity ofh+trogen. Argon and Ethane m Alcohol! and W&r. Am Chow. SC mci. (1958) 12. 1015-23. Murzin. V.I. and Afanas eva. N.L.: Solubility of Water m L~qurf~cd Ethane near its Cntlcal Point. Zh Fi:. Khim (1968) 42. No. 8. 194245. See alsoRefs. 174. 14. 15. 16. 18. 21, 28. 30, 31.41.42, 53. 54. 173. and
PropaneiWater Chdddock. R.E. : Liquid-Vapor Equilibrium m Hydrocarbon Water Systems [Propane-Water]. PhD dissertation. U of Michigan, Ann Arbor (1442). DeLoo\, T.W., Wijen, A.J.M., and Diepen. G.A.M.: Phase Eauilibna and CritIcal Phenomena in Fluid (Pronanc +Wateri at Hi gh Pressures and Temperatures. J. Chum Ti~~rr~~ntl~ricir,~ic.\ (1980) 12. 193-204. Goldup, A. and Wcstaway. M.T.: Detemlination of Trace Quantities of Water in Hydrocarbons. Application of the Calcium CarbIde-Gas Chromatagraphic Method to Streams Containing Methanol. An& Chwm (1966) 38. No. 12. 1657-61. Hachmuth, K.H.: Dehydrating Commercial (1931) 8. No 1, 55-56. 62. 64. Propane. 1.1 (1.\1cm Grr.3
Propylene/Water Azamonsh. A. and McKetta, J.J.: Solubility J. Chrm. Eng Darrt ( 1959) 4. 2 1 I- 12. of Propylene in Water.
Kazaryan. T.S. and Ryabtsev, N.1.: Solubility of Saturated Propylene. Iaobutylene. Isobutane, and n-Butane in Water and Aqueous Snlulionr. A +. Kho:. (1969) 47, No. IO. 54-S6. Klausutis. N.A.: Phase Equilibrium in the Propane-Propylene-Water System in the Three-Phase Region, PhD dissertation. U. ofTexas. Austm (I 968). LI. C.C. and McKetta. J.J.: Vapor-Liquid Equilibriums in the Propylene-Water Systems. J. Chetn. Eng. Dutu (1963) 8, 271-7.5.
Perry, C.W.: Determining Dissolved Water in Llquetied M. Eyq. Chm And Ed. (1938) 10. 513-14.
Case+.
Kre\heck. G.C.. Schneider, H.. and Scheraga. H.A.: The Effect of D,O on the Thermal Stability of Proteins. Thermodynamtc Parameter for the Transfer of Model Compounds from H ,O to D,O. J. Phy. Chrm. (1965) 69, 3132 Pocttmann, F.H. and Dean. M.R.: Water Content of Propane. Ppr. Rejner (1946) 25. No. 12. 125-28. Sanchez. M. and Coil. R.: Propane-Water System at High Pressures and Temperatures. I. Two-Phase Region, Am. Quirn. (1 178) 74. No. 11. 1329-3.5. Wehe. A.H. and McKetta. J.J.: Method for Determining Total Hydrocarbons Dissolved in Water. Am. Chm. (1961) 33. No. 20. 291-93. Wetlaufer. D.B. rt crl.: Nonpolar Group Participatwn m thr Desaturation of Proteins by Urea and Guanidinium Salts Model Compound Studies. J. Am. Chern. Sm. (1964) 86. SO8- 14. See Propene/Weter: Azamoosh and McKetta.
McBam. J.W. and O Connor, J.J.: The Effect of Potassium Oleate Upon the Solubility of Hydrocarbon Vapors m Water. J. ,4rn. Chmt. SM. (1941) 63. 875-77. McBam. J.W. and Soldate. A.M.: The Solubility of Propylene Vapor in Water aa Affected by Typical Detergents. J. Am. Chem. Sot,. (1942) 64. 1556-57. Oleinikova. A.L. and Bogdanov. M.I.: Solubility of Propylene m Water and Aqueous Sulfuric Acid Solutions. Uch. .X$X. Yaroslav. Tekhno. Inst. (1971) 27. 28-31. Petrw. A.N.. Pankov. A.G., and Bogdanov, M.I.: Liquid-Gas Chmmatographic Determinatmn of the Solubility of Hydrocarbon Gases in Water. Uch. .%p., Yaroslav. Tekhnol. Inst. (1970) 13, 186-90 Slemikova. A.L.. Pctroc. A.N.. and Bogdanov. M.I.: Thermodynamic\ of Dwwlution of Unsaturated C,-C, Hydrocarbons in Water. Uch. Zup.. Yaroslav. Tekhnol. Inst. (1971) 26. 35-41. See Ethylene/Water: See also Ref. 30. PropyneiWater Inga, R.F. and McKetta, J.J.: Solubility Chrm. Eng. Dota (1961) 6, 337-38. of Propyne in Water, J. Sanchez and Lentz.
See also Refs. IS. 28, 30, 31, 41. and 142. 1.3.ButadieneiWater See Ref. 30.
I-Butene/Water Brooks. W.B. and McKetta. J.J.: The Solublllty of I-Butene tn Water at Pressures to loo0 pbia. Per. Refinrr (19%) 34. No. 2. 143-44. Brooks, W.B. and McKetta, J.J.: The Solubiltty I-Butene, Pet. Rejner (1955) 34. No. 4, 138. ol Water in
Simpson, L.B. and Lovell, F.P.: Solubility of Ethyl, and Vinyl Acetylene in Several Solvents, J. Chum. Eng. D&o (1962) 7, 498-500. Cyclopmpane/Water Hafemann. D.R. Cyclopropane. and Miller, S.L.: The Clathrate Hydrates J. Phw. Chem. (1969) 73. No. 5. 1392-97. of
Brooks, W.B., Haughn. J.E., and McKetta, J.J.: The I-ButeneWater System in the Vapor and Three-Phase Regions. Prr. /&finer (1955) 34, No. 8. 129-30. See also Ref. 172. lsobutaneiwater Razaryan, T.S. and Ryabtsev. N.I.: Solubility of Saturated Propylene. Isobutylene. Isobutane. and n-Butane in Water and Aqueous Solutions, Nef. Khoz. (1969) 47. No IO. 54-56. Black, C., Joris, G.G., and Taylor. H.S: The Snlubility of Water in Hydrocarbons, J. Chwn. Ph~s. (1948) 16, No. S, 537-43 Nosov. E F. and Barlyaev. E.V.: Solubillty of Hexafluoropropylene and Isobutane in Water. Zh. Otnhtl~. Khirn. (1968) 38. No. 2, 211-12.
Imal, S.: Biophysicochemical Studies on Cyclopmpane. 1. Solublhty Coefticient of Cyclopropane for Various Solutions. ~liartr. I,qclku Zmshi. (1961) 12, 973-79. Thomson. ES. and Gjaldbaek. J.C.: Solubility of Cyclopropane m Pertluoro-Heptane, n-Hexane, Benzene, Dioxane. and Water, Dan. Tidsskr. Furm. (1963) 37, 9- 17. See also Ref. 30
Cyclopropane/KCI/Water Zerpa, C. of rri.: Solubility of Cyclopropane in Aqueous Solutions of Potassium Chloride. f. Chern. Enx. Data (1979) 24. No. I. 26-28.
25-26
PETROLEUM
ENGINEERING
HANDBOOK
Reed. C.D. and McKetta. J.J.: The Solubtltty Water, Pet. Refiner (1959) 38. No. 4, 159-60. See also Ref. 28
ot Isobutane
2.4.Dtmethylpentaneiwater;
and
Jordan, D. ( I a/.: Vapor-Liquid Equilibrium of C, HydrocarbonFurfural-Water Mixtures. Experimental and Theoretical Methods for Three- and Four-Component Systems. Chem. Eq. Pro yr. ( 1950) 46. 601-13.
See also Ref. 28. IsopentanelWater Pavlova. S.P. er crl.: Mutual Solubihty of C, Water, Prom. Sin Kauch. (1966) 3. 18-20. Hydrocarbons and
n-Butane/Water Brooks, W.B., Gibbs, G.B.. and McKetta. J.J.: Mutual Solubility of Light Hydrocarbon-Water Systems. Per. Refiner (1951) 30. No 10. 118-20. LeBreton, J.G. and McKetta. J.J.: Low Pressure Solubilrty of n-Butene in Water. Hydrocarbon Proc. PEWO. Ref: (1964) 43. No. 6. 136-38. Composition of Co-Existing Phases of Reamer. H.H. et al.: n-ButaneWater System in the Three-Phase Region. I&. E~IR. C&w (1944) 36, 381-83. n-Butane-Water System in the Two-Phase Reamer, H.H. er al.: Region, Ind. Eng. Chem. (1952) 44, 609-15. Rice. P.A., Gale, R.P.. and Barduhn. A.J.: Solubility of Butane in Water and Soft Solutions at Low Temperatures. J. Chern. Eng. Dam (1976) 21, 20406. Tsikiis, D.S. and Maslennikova, V.Y.: Limited Mutual Solubility of Gases in the Water-Butane System, Pokl. Akad. Nuuk SSSR (1964) 157, 426-29. See tsobutane/Water: See also Ref 28 Black er u/
NeopentaneiWater See Refs 30 and 37 CyclohexaneiWater Farkas. E.J.. New Method for Determination of Hydrocarbon-inWater Solubilities. Am/. Cbem. (1965) 37. No. 9, 1173-75. Roddy. J.W. and Coleman, C.F.: Solubility of Water m Hydrocarbons as a Function of Water Activity. Trr/r~n/n (1968) 15, No. I I. 12X1-86. Sultanov, R.G. and Skripka. V.G.: Solubility of Water in n-Hexane. Cyclohexane, and Benzene at Elevated Temperatures and Pressures, Z/t. Fiz. Khmz. (1973) 47, No. 4, 1035. See also Ref. 28 n-HexaneiWater Gester. G.C.: Design and Operation ofa Light Hydrocarbon Distillation Drier. Chrm. Eng. Pro,yr. (1947) 43, 117-22. See Cyclohexane/Water: Skripka. See also Ref. 28 Roddy and Coleman, and Sultanov and
See also Refs. 14. 15. 28, 30, 31. 41, and 55 See Ref. 28. 2-2-Dimethylbutane/Water See Ref. 28. and Cyclop-entaneiWater n-DecaneiWater Latter, Y.G., Asymyan, K.D., and Skripka, V.G.: The Volume Properties of the Coexisting Phases of the n-Decane-Water System at 275C. Zb. Fiz. K/rim. (1976) 50. 2171. MethaneiEthaneiWater Amitijafari, B. and Campbell. J.M.: Solubility ofGaseous Hydrocarbon Mixtures in Water, So,. Per. EUK. J. (Feb. 1972) 12, No. I. 21-27. Villameal. J.F., Bissey, L.T., and Nielson, R.F.: Dew Point Water Contents of Methane-Ethane Mixture\ at a Series of Pressures and Temperatures, Prod. Month!? (1954) 18. No. 7. 15- 17. Methane/Propane/Water; EthanelPropaneiWater; and
n-PentaneiWater Connolly. Critical 13-16. J.F.: Solubility of Hydrocarbons in Water Near the Solution Temperature, J. Chem. En,q Dora (1966) 11.
Fuhner. H.: Water Solubility in Homologous Chrm. Ges. (1924) 5713. 510-15.
Liabastre. A.A.: Experimental Determination of the Solubility of Small Organic Molecules in Water and Dideuterium Oxide and the Application of the Scaled Partrcle Theory to Aqueous and NonPhD dissertation. Georgia Inst. of Tech., aqueous Solutions. Atlanta (1974). Namiot, A.Y. and Beider. S.Y.: The Water Solubility of n-Pentane and n-Hexane, Khim. Tekhnol. Top/iv. Masel (1960) 5, 52-55. See Isobutane/Water: Black et al.
MethaneiEthanelPropaneiWater See MethaneiEthanelWater: Methane/n-Butane/Water Anthony, R.G. and McKetta, Hydrocarbons, Hydrocarbon See also Refs. 52 and 167. J.J.: How to Estimate H,O in Proces.s. (1968) 47. No. 6, 131-34. Amirijafari and Campbell
SYSTEMS
25-27
Methane/n-PentaneiWater See Ref. 44 EthyleneiEthaneiWater Anthony, R.G. and McKetla. J.J.: Phase Equilibrium in the Ethylene-Ethane-Water System. J. Chem. Eq. Data (1967) 12. No: 1. 21-28. Aromatic-Hvdrocarbon/Water Eganhtruae. R.P. and Calder. J.A.: The Snlubility of Medium Molecular Weight Aromatic Hydrocarbon(s) and the Effects of Hydrocarbon Co-Solutes and Salinity. Geochim. Cmmochim Actu. (1976) 40, No. 5, 555-61. Kerosme/Water and OlliWater
Hydrate/Volatile
Methane/Brine See Ref. 134 Ethylene/Water
Gas Systems
and MethaneiEtOHIWater
(NaCl)/Water
van Cleef, A. and Diepen, G.A.M.: Ethylene Hydrate Presaurrs, Rec. Trai,. Chim. (1962) 81, 425-29.
at High
Diepen. G.A.M.
The
Ethylene-Water
System, Rec. Truv. Chim. (1950) 69, 593-603. See also Ref. 122. EthylenelEtOH/Water Reamer, H.H., Selleck, F.T.. and Sage, B.H.: Some Properties of Mixed Paraffinic and Olefinic Hydrates, Trans.. AIME (1952)
Hydrocarbon-Water Solubilities at Griswold, J. and Kasch, J.E.: Elevated Temperatures and Pressures. Ind. Eng. Chrm. (1942) 34,
804-06.
Groschuff. E.: Solubility of Water in Benzene. Petroleum and Paraf(191 I) 17. 348. fin Oil. Nckrrochern. Naphtha/Warcr Amero, R.C.. Moore, J.W., and Capell. R.G.: Design and U&e of Adaorptwe Drying Units. Chrrn. E~IR. fro$!r. (1947) 43, 349-70. See Kerosine-Water: Griswold and Kaach
195, 197-201.
See also Ref. 129 and 140.
EthaneiWater
See EthyleneiEtOHiWater: Reamer er cri.
Set also Refs. 4. 8, 123, 124, 126, and 128. CyclopropaneiWater Callahan. J.E. and Sloan, E.D.: Heat Capacity Measurements on Structure I and II Pure Hydrates at Low Pressures and Below Room Temperature. Research Repon, Gas Research Inst. (Sept. 1982) Contract No. 5081-360-0487. Dhannawardhana. P.B., Parrish, W.R. andSloan. E.D.: Experimental Thermodynamic Parameters for the Prediction of Natural Gas Hydrate Dissociation Conditions, fnd. Eng. Chrm. Fund. (1980) 19, 410-14. Hafemann. D.R. Cyclopropane. and Miller, S.L.: The Clarhrate Hydrates J. Phys. Chem. (1969) 73, No. 5. 1392-97. and CyclopropanelCaCl z /Water of
Naiur;ll-GasiGlvcollWarer Ru\wll. G.F.. Reid. L S., and Huntington. R.L : Experimental Determinanons of the Vapor-Liquid Equilibria Between Natural Gas and 95% Dlethylene Glycol up to a Pressure of 2COOpsi, Trtrns., AlChE (1945) 41, 3 15-25. GdwlineiWater Derr. R.B and Willmore, C.B.: Dehydration of Organic Liquids wth Activated Alumina, Ind. Eng. Chem. (1939) 31. 866-68. Wachter. A. and Smith, S.S.: Preventing Internal Corrosion of the Pipe Lmcs-Sodium Nmite Treatment for Gasoline Lines, Azd. hg. C~PIPI. (1943) 35. 358-67. Argon/Methane/Water and Helium/Methane/Water
CyclopropaneiKCIIWater
Menten. P.D.. Pamsh, W.R. and Sloan. E.D.: Effect of Inhibitors on Hydrate Formatmn, Id En,q Chem Proressrs Des. Dev (1981) 20. No. 2. 399-401. CyclopropaneiMethanoliWater Giussani, A.: Inhibition of Natural Gas Hydrates, . orado School of Mines. Golden (1981). See Cyclopropane/KCI/Water: PronaneiWater Miller. B. and Strong. E.R. Jr.: Possibilitiesof Strong Natural Gas in the Form of a SolId Hydrate; a Method for Delerminmg the Ratio of Hydrocarbon to Water in a Solid Hydrate of a Normally Gaseous Hydrocarbon. and ita Application to Propane Hydrate, Prot. Anz. Cu.\ Acroc. (1945) 27. No. 2. 80-94. Set: EthyleneiEtOHIWater: Reamer e/ a/. Menten et a/ MS thesis, Col-
Namiol. A.Y. and Bondareva. M.M.: Solublhty of Helium-Methane Mixture m Water at High Pressures. N(zuch. Tekhn. SB. PO
Lowering of rhe Separation Mahara.jh. D.M. and Waikley. J.: Solubility of Oxygen by Presence of Another&s, Nutuw. London (1972) 236. No 5343, 165. See NitropcnlMethanelWater: Maharajh and Walkley.
Hydrogen-Sulfide/Carbon-Dioxlde/Merhane/Wdter Froning. H.R.. Jacoby. R.H., and Richards, W.L.: Vapor-Liquid Equillhnums of rhe Methane. Carbon Dioxide, Hydrogen Sulfide. Water System and Apphcation to the Design of a Water Wash System for Removing Acid Ga\e\. Pwr., Annual Convention of the Natural Ga\ A\wc. Am. (I9631 42. 32-39. Hydrogen-SulfideiWater Selleck. F.T.. Cannichacl, L.T.. and Sage. B.H.: Phase Behavior In the Hydrogen Sulfide-Water System. /mu. E/l,q. Chrru. (19.52) 44. 2219-26.
See also Refs. 5, 122, 126, 134, and I35 lsohutaneiwater Rouher. O.S. and Barduhn. A.J.: Hydrates of Iso- and Normal Butane and Their Mixtures. D~.\c/ir/urr/on (1969) 6. 57-73. See also Refs. 126 and 149. 1.l-DioxaneiWater and I .3-DloxanelWater Ices-Recent
Davidwn, D.W. and Ripmeester. J.A.: Clathrate Results, J. Glacial. (1978) 21. No. X5. 33-49.
25-28
PETROLEUM
ENGINEERING
HANDBOOK
and MethanelEthyleneiPrcIpylene/Water
Methane/Ethane/Propane/COz/Water See Refs. 164 and 165. MethaneiEthaneiPropanelIsobutaneln-Butane/Water See Ref. 127. Methane/Propane/Condensate/Water Methane/Propane/Crude-Oil/Water and
Holder, G.D. and Gngoriou. CC.: Hydrate Dissoctation Pressures of (Methane-Ethane-Water). Existence of a Locus of Minimum Pressures, J Chem. Thrnnnclwumrcs (1980). 1093- 1104. See ~1s~~ Rcf\. 5, 127, and 128.
See Ref. 192. Methane:Propylene/Water MethaneiPropaneiHydrogenSulfideiWater Otto, F.D. and Robmson. Methane-Propylene-Water 602m05. D.B.. A Study of Hydrates in the System, AlChE J. (1960) 6. No. 4. Schrocter, J.P., Kobayashi, R., and Hildebrand. M.A.: Hydrate Decomposition Conditions in the System HzS-Methane-Propane. fnd. Eng. Chem. Fund. (1983) 22, No. 4. 361-64 laobutylenelPropane/lsobutaneiWater Carson. D.B. and Katz.. D.L.: AIME (1941) 146, 150-53. Natural Gas Hydrates. Trunc., and
Verma, V.K., Hand, J.H., and Katz, D.L.: Gas Hydrates from Liquid Hydrocarbons (Methane-Propane-Water Systems), paper presented at the GVC/AIChE Joint Meeting, Munich, Germany, Sept. 17-21. See alw Ref\ 5, 127, and 163.
Cracked-Gas/Water Byk, S.S.: Experiments on the Conditions for the Formation Hydrates of Cracked Gases, Gaz. from. (1957) 4, 33-35. Oil/Water Set Ref. 193 ol
Methane!JsobutaneiWater Ng. H.-J. and Robinson. D.B.: The Measurement and Prediction of Hydrate Formation in Liquid Hydrocarbon-Water Systems. lnd. En,q. Chmi. Fund. (1976) 15. No. 4, 293-98. See ah Refs. 5. 127. and 149.
See Ref. 154. Methane/Nitrogen/Water Jhavcri, I. and Robmson. D.B.: Hydrates in the Methane~Nitrogen System, Cdn. J. Chew. Eng. (1965) 43, No. 2, 75-78. ProoaneiNitrogeniWater Ng, H.-J., Pettunia. J.P., and Robinson. D.B.: Expenmental Measurement and Prediction of Hydrate Forming Conditions in the N,-Propane-Hz0 System. Nuid Phasr Equil. (1978) 1. 2X3-91. Methane/Carbon-Dioxide/Water See Ref. 139.
Methane/n-Butane/Water John. V.T. and Holder. G.D.: Improwd Predictions of Hydrate Equilibria, paper presented at the 1981 AIChE Annual Meeting. New Orleans The Prediction of Hydrate Formation Ng. H.-J. and Robinson. D.B.: in Condensed Systems. AlChE J. (1977) 23. No. 4. 477-82 See also Ref. 127
Propane/Carbon-Dioxide/Water Methane/Pcntane!Water and MethaneiHexanelWater Carson and Katz. Propane/Hydrogen-Sulfide/Water EthyleneiEthnneiWater Koshelev. V.S. CI (I/. : Study ot Hydrate-Gas Phase Equilihrtum m Khirn., Lemgrad (1971) the Ethane-Ethylene System. Zh. PM. 44, 2573-74. EthvlenelPmpaneiWater Fomina. V.I. and Byk. S.S.: Study of Hydrate-Gas Equtltbrium m the Ethylene-Propane System. Zh. Prilil. Leningrad (1969) 42. 2855-58. EthanelPropaneiWater and MethanelEthanelPropaneiWater Natural-Gas/Ammonium/Water Holder, G.D. and Hand, J.H.: Multiple-Phase Equilibria in Hydrates from Methane, Ethane. Propane and Water Mixtures. AlCliE J. (1982) 28, No 3. 440-47. Propane/PnlpylcnciWater Fomina. V I. and Byk. S.S.: Hydrate-Gas Phase Equilthrium m the Propane-Propylcne~Water System. Go:. Pronr. (1967) 12, Nil. 3. X-56. See Ethyletw EtOHiWater: Reamer 01 o/. Russell. J.T.: Anhydrous Ammonia to Inhibit uon, Cur (1937) 13. No. 6. 38-40. Natural-GaslNaCliWater See Ref. 134. Natur&Gas/CaCl SW Ref. 5 ,/Water Gas Hydrate FormaPhase Khiw. Set Ref. 197 See Ref. 151. MethaneiHvdroaen-Sulfide/Carbon-Dioxide/Water Robinson. D.B. and Hutton. J.M.: Hydrate Formatton m Systems Containing Methane. Hydrogen Sulfide and Carbon Dioxide. J. Ch. Per. Tech. (1967) 6. No. I, 6-9. Robinson. D.B. and Mehta, B.R.: Hydrates in the PropaneCO,-H,O System, J. Cdn. PH. Twh. (1971) 10. 33-35.
See Methane-Propane-Water:
Chapter 26
Introduction
This chapter deals with the fundamental properties of reservoir rocks. The properties discussed are (1) pomsi8-a measure of the void space in a rock: (2) pertneahi/it~-a measure of the fluid transmissivity of a measure of the gross void rock; (3) fluid saturutim-a space occupied by a fluid; (4) cupillaryprrssure rekrim--a measure of the surface forces existing between the rock and the contained fluids; and (5) electrical cmclucrivity offluid-saturated rocks-a measure of the conductivity of the rock and its contained fluids to electric current. These properties constitute a set of fundamental parameters by which the rock may be described quantitatively. Typical core-analysis data are presented to illustrate the description of porous media by these fundamental properties. characteristics than rocks in which a large part of the porosity is induced. For direct quantitative measurement of porosity, reliance must be placed on formation samples obtained by coring. Unit cells of two systematic packings of uniform spheres are shown in Fig. 26.1. The porosity for cubical packing (the least compact arrangement) is 47.6% and for rhombohedral packing (the most compact arrangement) is 25.96%. Considering cubical packing, the porosity may be calculated as follows. The unit cell is a cube with sides equal to 2r where r is the radius of the sphere. Therefore, Vh=(2r)3=8r3, where V,] is the bulk volume. Since there are 8% spheres in the unit cell. the sand-grain volume, V,,, is given by 4ar v, = 3 The porosity, $, is given by
Porosity
Porosity is defined as the ratio of the void space in a rock to the bulk volume (BV) of that rock, multiplied by 100 to express in percent. Porosity may be classified according to the mode of origin as primary and secondary. An original porosity is developed during the deposition of the material, and later compaction and cementation reduce it to the primary porosity. Secondary porosity is that developed by some geologic process subsequent to deposition of the rock. Primary porosity is typified by the intergranular porosity of sandstones and the intercrystalline and oolitic porosity of some limestones. Secondary porosity is typified by fracture development as found in some shales and limestones and the vugs or solution cavities commonly found in limestones. Rocks having primary porosity are more uniform in their
Thts author also Wrote the orlglnal chapter on this topic m the 1962 edition with coauthor James W. Amyx (deceased] D.M Bass Jr IS currently a petroleum COnSUltant
X100=(1--&00
=47.6%. Of particular interest is the fact that the radii cancel and the porosity of uniform spheres is a function of packing only.
26-2
PETROLEUM ENGINEERING
HANDBOOK
POROSITY:
47.6 %
POROSITY=25.96%
CuB~~~KXE;IDE
-.08
-.04
0 SKEWNESS
+.04
+.08
Fig. 26.1-Unit cells and groups of uniform spheres for cubic and rhombohedral packing.
Tickell et al. has presented experimental data indicating that, for packings of Ottawa sand, porosity was a function of skewness of the grain-size distribution (see Fig. 26.2). Skewness is a statistical measure of the uniformity of distribution of a group of measurements. Other investigators have measured the effects of distribution, grain size, and grain shape. In general, greater angularity tends to increase the porosity, while an increase in range of particle size tends to decrease porosity. In dealing with reservoir rocks (usually consolidated sediments), it is necessary to define total porosity and effective porosity because cementing materials may seal off a part of the PV. Totalporosity is the ratio of the total void space in the rock to the BV of the rock, while effective porosity is the ratio of the interconnected void space in the rock to the BV of the rock, each expressed in percent. From the reservoir-engineering standpoint, effective porosity is the desired quantitative value because it represents the space that is occupied by mobile fluids. For intergranular materials, poorly to moderately well cemented, the total porosity is approximately equal to the effective porosity. For more highly cemented materials and for limestones, significant differences in total-porosity and effective-porosity values may occur.
Photographs of oilwell cores are presented in Fig. 26.3. 3 The pore configuration of the sandstones shown is complex, but the pores are distributed relatively uniformly. Complex pore configurations arise from the interaction of many factors in the geologic environment of the deposit. These factors include the packing and particle-size distribution of the framework fraction, the type of interstitial material, and the type and degree of cementation. The influence of these various factors may be evaluated as statistical trends. Material having induced porosity, such as the carbonate rocks shown in Fig. 26.3, have even more complex pore configurations. In fact, two or more systems of pore openings may occur in such rocks. The basic rock material is usually finely crystalline and is called the matrix. The matrix contains uniformly small pore openings that comprise one system of pores. One or more systems of larger openings usually occur in carbonate rocks as a result of leaching, fracturing, or dolomitization of the primary rock material. Vugular pore openings are frequently as large as an ordinary lead pencil and usually are attributed to leaching of the rock subsequent to deposition. Fractures also may be quite large and contribute substantially to the volume of pore openings in the rock.
(b)
(d)
Fig. 26.3-Oilwell cores. Consolidated sandstone: (a) wireline core, Lower Frio; (b) whole core, Seven Rivers. Vugular, solution cavities, and crystalline limestone and dolomite: (c) whole core, Devonian; (d) whole core, Hermosa.
26-3
Laboratory
Measurement
of Porosity
Numerous methods have been developed for the determination of the porosity of consolidated rocks having intergranular porosity. Most of the methods developed have been designed for small samples, roughly the size of a walnut. Since the pores of intergranular material are quite small, determining the porosity of such a sample involves measuring the volume of literally thousands of pores. The porosity of larger portions of the rock is represented statistically from the results obtained on numerous small samples. In the laboratory measurement of porosity, it is necessary to determine only two of the three basic parameters (BV, PV, and grain volume). In general. all methods of BV determination are applicable to determining both total and effective porosity.
BV. Usual procedures use observation of the volume of
Sample
Im-
weight of saturated sample in air, D=22.5 g, and weight of saturated sample in water at 40F, E= 12.6 g. we can calculate that weight of water displaced=D-E=22.512.6 x9.9 g, and 9.9 volume of water displaced= =9.9 cm3. 1.0 Therefore, BV rock, If,,, =9.9 cm3.
Example Problem cury Pycnometer. 3-Dry Sample Immersed in Mer-
fluid displaced by the sample. This procedure is particularly desirable because the BV of specially shaped samples may be determined as rapidly as that of shaped samples. The fluid displaced by a sample may be observed either volumetrically or gravimetrically. In either pm cedure, it is necessary to prevent fluid penetration into the pore space of the rock. This may be accomplished by (1) coating the rock with paraffin or a similar substance, (2) saturating the rock with the fluid into which it is to be immersed, or (3) using mercury, which by virtue of its surface tension and wetting characteristics does not tend to enter the small pore spaces of most intergranular materials. Gravimetric determinations of BV may be accomplished by observing the sample s weight loss when immersed in a fluid or the difference in weight of a pycnometer when filled with a fluid and when filled with fluid and the core sample. The details of gravimetric determinations of BV are best summarized by Example Problems I through 3.
Example Problem l-Coated Water. Given that the Sample Immersed in
Given A from Example Problem I and the following values, weight of pycnometer filled with mercury at 20C. F=350.0 g and weight of pycnometer filled with mercury and sample at 20C. G=235.9 g (density of mercury= 13.546 g/cm3). we can calculate that weight of sample + weight of pycnometer filled with mercury=A+F=20+350=370 g, weight of mercury displaced =A+F-G=370235.9= 134.1 g. and 134.1 =9.9 cm3 volume of mercury displaced = 13.546 Therefore, BV of rock=9.9 cm3. Determinations of BV volumetrically use a variety of specially constructed pycnometers or volumeters. An electric pycnometer from which the BV may be read directly is shown in Fig. 26.4. The sample is immersed in the core chamber, causing a resulting rise in the level of the connecting U tube. The change in mercury level is measured by a micrometer screw connected to a lowvoltage circuit. The electric circuit is closed as long as the measuring point is in contact with the mercury. The travel of the measuring point is calibrated in volume units such that the difference in the open-circuit readings with and without the sample in the core chamber represents the BV of the sample. Either dry or saturated samples may be used in the device.
Volume (GV). The various porosity methods usually are distinguished by the means used to determine the GV or PV. Several of the oldest methods of porosity measurement arc based on the determination of GV. The GV may be determined from the dry weight of the sample and the sand-grain density. For many purposes, results of sufficient accuracy may be obtained by using the density of quartz (2.65 g/cm3) as the sand-grain density. For more rigorous determinations either the A.F. Melcher-Nutting4 or Russell method may be employed. The BV of a sample is determined: then this sample, or an adjacent sample, is reduced to grain size and the GV is determined. In the M&her-Nutting technique, all the measurements are determined Sand-Grain
weight of dry sample in air. A=20.0 g. weight of dry sample coated with paraffin, B=20.9 g (density of paraffin=0.9 gicmj), and weight of coated sample immersed in water at 40F. C= 10.0 g (density of water= 1 .OOg/cm3). we can then calculate that weight of paraffin=&A=20.9-20.0=0.9 g. 0.9 volume of paraffin = = 1 cm3 , 0.9 weight of water displaced=B-C=20.9= 10.9 g, 10.0
10.9 volume of water displaced = ~ = 10.9 cm3, 1.o volume of water displaced-volume =10.9-1.0=9.9 cm3, and BV ofrock, Vb,=9.9 cm3. of paraffin
26-4
PETROLEUM ENGINEERING
HANDBOOK
TABLE
26.1~METHODS
OF DETERMINING
POROSITY
Effective Porosity Methods Washburn-Bunting Porosimeter Type of sampling One to several pieces per increment (usually one). Solvent extraction and oven drying. Occasionally use retort samples. Functions measured PV and BV. Stevens Porosimeter One to several pieces per increment (usually one). Solvent extraction and oven drying. Occasionally use retort samples. Sand grain volume and unconnected PV and BV. Difference in volume of air evolved from a constant-volume chamber when empty and when occupied by sample. BV by Russell tube. Mercury does not become dirty. Possible leaks in system; incomplete evacuation caused by rapid operation or tight sample. Kobe Porosimeter One to several pieces per increment (usually one). Solvent extraction and oven drying. Occasionally use retort samples. Sand grain volume and unconnected PV and BV. Difference in volume of air evolved from a constant-volume chamber when empty and when occupied by sample. BV by Russell tube. Mercury does not become dirty. Possible leaks in system; incomplete evacuation caused by rapid operation or tight sample. Boyle s Law Porosimeter One to several pieces per increment (usually one). Solvent extraction and oven drying. Occasionally use retort samples. Sand grain volume and unconnected PV and BV. Difference in volume of air evolved from a constant-volume chamber when empty and when occupied by sample. BV by Russell tube. Mercury does not become dirty. Possible leaks in system; incomplete evacuation caused by rapid operation or tight sample.
Manner of measurement
Reduction of pressure on a confined sample and measurement of air evolved. BV from mercury pycnometer.
Errors
Air from dirty mercury; possible leaks in system; incomplete evacuation caused by rapid operation or tight sample.
MICROMETER
SCALE
ADJUSTING
SCREW
gravimetrically, using the principle of buoyancy (Example Problem 2). The Russell method uses a specially designed volumeter, and the BV and GV are determined volumetrically. The porosity determined is total porosity, $r. Thus,
Vb - vs vb
+r=-
x loo.
From the data of Example Problem 2 and using a sandgrain density of 2.65 g/cm3, vb =9.9 cm3, 20 2.65 OH
INDICATOR LIGHT
v,=-
=7.55 cm3,
x 100=23.8%.
Fig. 26.4-Electric
pycnometer
The Stevens6 porosimeter is a means of determining the effective GV. The porosimeter (Fig. 26.5) consists of a core chamber that may be sealed from atmospheric pressure and closed from the remaining parts of the porosimeter by a needle valve. The accurate volume of the core chamber is known. In operation, a core is placed in the core chamber; a vacuum is established in the
26-5
TABLE 26.1-METHODS
OF DETERMINING POROSITY (continued) Total Porosity Method Core Laboratories, Dry Sample One to several pieces per increment (usually one). Solvent extraction and oven drying. Occasionally use retort samples. Sand-grain volume and unconnected PV, and BV. Difference in volume of air evolved from a constant-volume chamber when empty and when occupied by sample. Sand Density Several pieces per increment.
Effective Porositv Methods Saturation* One to several pieces per increment (usually one). Solvent extraction and overy drying. Occasionally use retort samples. PV and BV. Core Laboratories, Wet Sample Several pieces for retort; one for mercury pump. None.
Solvent extraction; then, in second step, crush sample to grain size. BV of sample and volume of sand grains.
Weight of dry sample; weight of saturated sample in air: weight of saturated sample immersed in saturated fluid.
Weight of retort sample; volume of oil and water from retort sample; gas volume and BV of mercury-pump sample.
Weight of dry sample; weight of saturated sample immersed; weight and volume of sand grains.
Possible leaks in system; incomplete evacuation caused by rapid operation or tight sample.
system by manipulating the mercury reservoir; the air in the core chamber is expanded repeatedly into the evacuated system and measured at atmospheric pressure in the graduated tube. The Stevens method is an adaptation of the Washburn-Bunting7 procedure, which will be discussed with the measurement of PV.
Example Problem 4-Determination Volume by Gas Expansion (Stevens of Grain Porosimeter).
PULLEY.
MERCURY RESERVOIR
--P
GTOPCOCK ;RADUATED
TUBE
Given that volume of core chamber, H= 15 cm3 and total reading, 1=7.00 cm3 where volume of air (first reading)=6.970, volume of air (second reading)=0.03, and volume of air (third reading)=O, we can calculate effective grain volume=H-Z=8 cm3, bulk volume of sample (from pycnometer)= 10 cm3, and effective porosity, 4, =[(lO-8)110] X 100=20%. PV yield effective porosity. The methods are based on either the extraction of a fluid from the rock or the introduction of a fluid into the pore space of the rock. The Washburn-Bunting 7 porosimeter measured the volume of air extracted from the pore space by creating a partial vacuum in the porosimeter by manipulation of the attached mercury reservoir. In the process, the core is exposed to contamination by mercury and, therefore, is
PV. All methods of measuring
Fig. 26.5-Stevens
porosimeter
26-6
PETROLEUM
ENGINEERING
HANDBOOK
TABLE
26.2-CHARACTERISTICS
OF SAMPLES
USED IN POROSITY-MEASUREMENT
COMPARISONS
Porosity (O/O) Sample Number 1 2 3 4 BZE 0.x 61-A 722 1123 1141-A Approximate Gas Permeability (md) 1 2 20 1,000 0.2 0.8 1,000 3 16 45 Average From Gas Methods 17.81 26.60 14.21 31.06 4.15 4.10 28.70 16.73 33 10 19.68 Average From Saturation Methods 16.96 27.97 13.70 29.13 3.66 3.71 28.00 16.08 32.03 19.12 Value From High Observation 18.50 29.30 15 15 31.8 4.60 4.55 29.4 17.80 33.8 20.2 Value From Low Observation 16.72 27.56 13 50 26 8 3 50 3.48 27.8 16.00 31 7 100
Type of Material LImestone Fretted glass Sandstone Sandstone Semiquartzltic Semlquartzitic Alundum Alundum Chalk Sandstone
Average 17.47 28.40 14.00 30.29 3.95 3 94 28.47 16.47 32.67 19.46
sandstone sandstone
not suitable for additional tests. The previously described Stevens method is a modification of the Washburn-Bunting procedure especially designed to prevent mercury contamination of the samples. A number of other devices have been designed for measuring PV; these include the Kobe8 porosimeter, the Oilwell Research porosimeter, and the mercury-pump porosimeter. Kobe and Oilwell Research porosimeters are Boyle s-law-type porosimeters designed for use with nitrogen or helium with negligible adsorption on rock surfaces at room temperature. The mercury-pump porosimeter is designed so that the BV may be obtained as well as the PV. The saturation method of determining porosity consists of saturating a clean dry sample with a fluid of known density and determining the PV from the gain in weight of the sample. The sample usually is evacuated in a vacuum flask to which the saturation fluid may be admitted by means of a separatory funnel. If care is exercised to achieve complete saturation, this procedure is believed to be one of the best available techniques for intergranular materials. Example Problem 5 illustrates the saturation technique of measuring PV.
Example Problem Saturation Method. G-Effective Porosity by the
2.5 g =2.5 cm3, volume of water in pore space = ~ 1 g/cm3 effective PV=2.5 cm , and BV (Example Problem 2)=9.9 Therefore, 2.5 ~,=~x100=25.3%.
cm3.
From the data of Example Problems 1 and 2, we can calculate weight of water in pore space=D-A=22.5-20 =2.5 g.
EXPANSION
KNOWN VOLUME
PROPERTIESOF RESERVOIRROCKS
26-7
50
6CCO
POROSITY-PER
CENT
which high-pressure gas is equalized between two chambers. The porosity may be calculated from the measured pressures, the volume of either chamber, and the bulk-sample volume by the use of Boyle s law. Keltont reported results of whole-core analysis. a method utilizing large sections of the full-diameter core. The following table of matrix vs. whole-core data summarizes a part of Kelton s work. lo
Grollp
2 1.58 2.62
3 2.56 3.17
4 7.92 8.40
Matrix porosity is that determined from small samples; total porosity is that determined from the whole core. Whole-core analysis satisfactorily evaluates most carbonate rocks. However, no satisfactory technique is available for the analysis of extensively fractured materials because the samples cannot be put together in their natural state.
Compaction and Compressibility of Porous Rocks
The depletion of fluids from the pore space of a reservoir rock results in a change in the internal stress in the rock, thus causing the rock to be subjected to a different resultant stress. This change in stress results in changes in the GV, PV, and BV of the rock. Of principal interest to the reservoir engineer is the change in the PV of the rock. The change in rock-bulk compressibility may be of importance in areas where surface subsidence could cause appreciable property damage. Hall I3 reported PV compressibility as a function of porosity. These data are summarized in Fig. 26.8. The effective rock compressibility in Fig. 26.8 results from the change in porosity caused by grain expansion and decrease in pore space because of compaction of the matrix. Fatt I4 indicates that the pore compressibility is a function of pressure. Within the range of his data, he was unable to find a correlation with porosity.
The porosity of sedimentary rocks has been shown by Krumbein and Sloss to be a function of the degree of compaction of the rock. The compacting forces are a function of the maximum depth of burial of the rock. The effect of natural compaction on porosity is shown in Fig. 26.7. This effect is caused principally by the resulting packing arrangement after compaction; thus, sediments that have been buried deeply, even if subsequently uplifted. exhibit lower porosity values than sediments that have not been buried at great depth. Geertsma12 states that three kinds of compressibility must be distinguished in rocks: (I) rock-matrix compressibility, (2) rock-bulk compressibility, and (3) pore compressibility. Rock-bulk compressibility is a combination of pore and rock matrix compressibility. Rock-matrix compressibility is the fractional change in volume of the solid rock material (grains) with a unit change in pressure. Pore compressibility is the fractional change in PV of the rock with a unit change in pressure.
725
Gradient
(North
6 35 I 30
Formation
I 25
Compressibility(Microsip1
I 20
I 15
I IO
I 5
Fig. 26.9-Depth vs. formation compressibility in abnormally pressured segment of an abnormally pressured reservoir.
26-6
PETROLEUM ENGINEERING
HANDBOOK
-I
100
0 0 0 0 O 0 08 0 0 00 00 0 0 B 0 0 @ OQ 0 0
0 O 00
0 0
0 0
0 0 0 0
0 O oo 0 0
@ >& . l l \eet l l 0 0 0 0
0 0 0 : 0 I
J 35
Fig. 26.10-PV compressibility at 75% lithostatic pressure vs. initial sample porosity for limestones.
Fig. 26.12-PV compressibility at 75% lithostatic pressure vs. initial sample porosity for friable sandstones.
The additional significance of changes in porosity with the discovery of oil at deeper depths and new geological areas resulted in the need for a better understanding of the changes in porosity with the depletion of reservoir fluid pressure. Hammerlindl I5 developed a correlation from measured field data that indicates the change in porosity compressibility with depth of burial of an unconsolidated sand (Fig. 26.9). Similar correlations have also been presented by others in the technical literature. Considerable laboratory work has been performed recently in an attempt to understand better the effect of formation compaction on porosity. Newman I6 performed measurements on samples of limestone and consolidated, friable, and unconsolidated sandstones. He compared his results with those of Hall I3 and van der Knaap, as illustrated in Figs. 26.10 through 26.13. As noted in his data, an approximate correlation may exist between PV compressibility and porosity for limestones and consolidated sandstones. Little or no correlation exists between PV compressibility and porosity for friable and unconsolidated sandstones. By averaging the PV compressibility in 5% ranges of porosity, Newman attempted to correlate all four types of porous media. The results of this averaging technique are presented in Fig.
26.14.
Fig. 26.11-PV compressibility at 75% lithostatic pressure vs. initial sample porosity for consolidated sandstones.
The methods used to measure PV compressibility have come under discussion. PV compressibility can be measured in the laboratory by the hydrostatic (same pressure in all three directions) or the triaxial (different pressure in the z direction than in the x and y directions) techniques. The test samples also can be stress cycled
Fig. 26.13-PV
compressibility at 75% lithostatic pressure vs. initial sample porosity for unconsolidated sandstones.
Fig. 26.14-Class
sample porosity.
until the strain resulting from a supplied stress is constant, or the sample can be placed under stress and the strain measured for purposes of calculating PV compressibility. Krug I8 and Graves I9 demonstrated that when a formation sample was stress cycled to a stable strain condition, the sample gave repeatable values of PV compressibility even when the sample was left in an unstressed condition for 30 days or more. The data reported by Newman were for samples that were not stress cycled. Lachance* compared PV compressibilities obtained by the hydraulic and triaxial methods. The reported results (Fig. 26.15) indicate a large difference in the magnitude of PV compressibilities obtained by the two methods. The triaxial data indicate that PV compressibility is essentially independent of the sample porosity. Newman s I6 data were obtained by the hydrostatic method, whereas Krug s I8 and Graves I9 data were obtained by the triaxial method. In summary, rock compressibility is an important factor in reservoir evaluation. Oil reservoirs with high initial pressures and low fluid bubblepoint pressures are sensitive to the true value of PV compressibility. Gas reservoirs with initial reservoir pressures in excess of 6,000 psi are also sensitive to the value of PV compressibility. Newman, I6 Krug, I8 and Graves I9 all recommend that PV compressibility be measured on
Fig. 26.15-Calculated
26-10
PETROLEUM ENGINEERING
HANDBOOK
6 3
-_----_ a t r/--i
h,
.--------_ ;
h-h,
where v d Ap L fluid velocity, cm/s, diameter of conductor, cm, pressure loss over length L, dynes/cm2, length over which pressure loss is measured, cm, p = fluid viscosity, Pa s, p = fluid density, g/cm , and f = friction factor, dimensionless. = = = =
4=
.... ...
(3)
where r is the radius of the conduit, cm, q is the volume rate of flow, cm/s, and the other terms are as previously defined. If a porous medium is conceived to be a bundle of capillary tubes, the flow rate q1 through the medium is the sum of the flow rates through the individual tubes. Thus,
qr=s i
P I 1
lljrj4,
. .
(4)
(T/8)
njrj4
j=l
samples from the reservoir in question when the PV compressibility may be significant in reservoir evaluation.
is treated as a flow coefficient for the particular grouping of tubes the equation reduces to
Permeability
Introductory Theory
qr=c$;
where
k
. . . . . . . . . . . . . . . . . . . . . . . . . . . . ...(5)
It is the purpose of this section to discuss the ability of the formation to conduct fluids. In the introduction to API Code 27, it is stated that permeability is a property of the porous medium and a measure of the medium s capacity to transmit fluids. The measurement of permeability, then, is a measure of the fluid conductivity of the particular material. By analogy with electrical conductors, the permeability represents the reciprocal of the resistance that the porous medium offers to fluid flow. The following equations for flow of fluids in circular conduits are well known. Poiseuille s equation for viscous flow: d2Ap v=- 321.LL) . . . . . . . . . . . . . . . . ..I..........
The pore structure of rocks does not permit simple classification of the flow channels. Therefore, empirical data are required in most cases. In 1856, Darcy investigated the flow of water through sand filters for water purification. His experimental apparatus is shown schematically in Fig. 26.16. Darcy interpreted his observations to yield results essentially as given in Eq. 7. q=KA-
ht --h2
L
(1)
.......
.............
. . (7)
Fanning s equation for viscous and turbulent flow: 2dAp $=- fpL ,
..
..
(2)
q represents the volume rate of flow of water downward through the cylindrical sandpack of cross-sectional area A and length L. h 1 and h2 are the heights above the standard datum of the water in manometers located at the input and output faces and represent the hydraulic head at
PROPERTIES
OF RESERVOIR
ROCKS
26-11
Points 1 and 2. K, a constant of proportionality. was found to be characteristic of the sandpack. Darcy s investigations were confined to flow of water through sandpacks that were 100% saturated with water. Later investigators found that Darcy s law could be extended to fluids other than water and that the constant of proportionality K could be written as klp, where p is the viscosity of the fluid and k is a proportionality constant for the rock. The generalized form of Darcy s law as presented in API Code 27 is presented in Eq. 8. u,yz-; where s = distance in direction of flow, always positive, volume flux across a unit area of the porous us = medium in unit time along flow path S, z = vertical coordinate, considered positive downward, cm, p = density of the fluid, g = acceleration of gravity, = pressure gradient along s at the point to ds which u refers, p = viscosity of the fluid, k = permeability of the medium, and dt -= ds sin 8, where fl is the angle between s and the horizontal. ($-gp;). . .
Fig. 26.17-Sand
model for rectilinear flow of fluids.
Conditions of viscous flow mean that the rate of flow will be sufficiently low to be directly proportional to the potential gradient. Darcy s law holds only for conditions of viscous flow as defined. Further, for the permeability k to be a proportionality constant of the porous medium, the medium must be 100% saturated with the flowing fluid when the determination of permeability is made. In addition, the fluid and the porous medium must not react-i.e., by chemical reaction, adsorption, or absorption. If a reactive fluid flows through a porous medium, it alters the porous medium and, therefore, changes the permeability of the medium as flow continues.
Flow Systems of Simple Geometry Horizontal Flow. Horizontal rectilinear steady-state flow is common to virtually all measurements of permeability. If a rock is 100% saturated with an incompressible fluid and is horizontal (Fig. 26.17), then dz/ds=O, dplds=dpi&, and Eq. 8 reduces to 4
u, may further be defined as q/A where q is the volume rate of flow and A is the average cross-sectional area perpendicular to the lines of flow. The portion of Eq. 8 in parentheses may be interpreted as the total pressure gradient minus the gradient caused by a head of fluid. Thus, if the system is in hydrostatic equilibrium, there is no flow and the quantity inside the parentheses will be zero. Eq. 8 may be rewritten as kd u, =--@gz-p). fib . .. . .. .. .
-k dp tth
ux=-=--3 A
9=
(11)
The quantity d(pgz-p)lds may be considered to be the negative gradient of a potential function b, where a=p-pgz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(lO)
The potential function is defined such that flow will be from higher to lower values. The dimensions of penmeabilit may be established from an analysis of Eq. 8 as k=L P . In the cgs system of units, the unit of permeability would be cm*, a large unit for common usage; therefore, the petroleum industry adopted the darcy as the standard unit of permeability, which is defined as follows.
A porous medium has a permeability of one darcy when a single phase fluid of one centipoise viscosity that completely fills the voids of the medium, will flow through it under conditions of viscous flow at a rate of one cubic centimeter per second per square centimeter cross sectional area under a pressure or equivalent hydraulic gradient of one atmosphere per centimeter.
where k is the specific permeability. If a compressible fluid flows through a porous medium, Darcy s law, as expressed in Eq. 8, is still valid. However, for steady flow, the mass rate of flow rather than the volume rate of flow is constant through the system. Therefore, the integrated form of the equation differs. Considering steady rectilinear flow of compressible fluids, Eq. 8 becomes
pux=---,
b dp
ttb
.,..........,,...........
(12)
=constant.
26-12
PETROLEUM ENGINEERING
HANDBOOK
I
FREE FLOW
Thus,
tbqb -=-_b dp
P dx
where pb and qb are the density and volume rate of flow, respectively, at the base pressure, ph. Substituting for p
Pbqb k dp
-=--p-, A
.........
. (16)
cl&
model
for vertical
flow
q=G
kA P,*-Pz~ Pb
.......
_. __ (17)
The density-pressure relationship for isothermal conditions of a slightly compressible fluid may be expressed as p=poe( p
Define p as (p, +p2)12, and yP as the volume rate of flow at p such that pqp =pbqh; then
41, = and
k4PI
-P2) uL
.,....................
(18)
ap=a,
CP
..
. .
.(13)
Therefore, flow rates of ideal gases may be computed from the equations for incompressible liquids so long as the volume rate of flow is defined at the algebraic mean pressure. Vertical Flow. Fig. 26.18 shows three sandpacks in which linear flow occurs in the vertical direction. First consider Case 1 (Fig. 26. IS)--when the pressure at the inlet and outlet are equal (free flow), such that only the gravitational forces are driving the fluids. Given
Thus,
-k dp pc dx
dx
On integration,
WP PO90 = I -02)
dz
and -=O, ds
dp
cLcL
If terms of cp of second and higher order are neglected, the density can be expressed as
q&g. P
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .(19)
Next consider Case 2-the case of downward flow when the driving head (difference in hydraulic head of inlet and outlet) is h (Fig. 26.18). We know that
z=l
dz,
and z=-
-ah L
The density-pressure relationship for isothermal conditions of an ideal gas may be expressed as
P P Pb PPb Pb
-=Ph
orp=-.
,(15)
.... ..
. . . .(20)
PROPERTIES
OF RESERVOIR
ROCKS
26-13
When the flow is upward and the driving head is h, Case 3 (Fig. 26.18), and z is defined as positive downward,
and u=+then
kApgh 4= . . .._..........................
k P
(21)
system, analogous to flow into a wellbore, is idealized in Fig. 26.19. If flow is considered to occur only in the horizontal plane under steady-state conditions, an equation of flow may be derived from Darcy s law to be Fig. 26.19-Sand
2akMp, 4=
~ ln rplr,,
-P,,
model wellbore.
for
radial
flow
of fluids
to central
) .
. .
(22)
Conversion of Units in Darcy s Law
where rI, is the radius at the external boundary at which pc (pressure at the external boundary) is measured, and r,v is the radius of the wellbore at which pa, (pressure at the wellbore) is measured. All other terms are as defined for linear flow. Eq. 22 may be modified appropriately for the flow of compressible fluids. The details of modifying this equation are omitted because they are essentially the same as the ones used in the horizontal rectilinear flow systems. After modification for variations in flowing volumes with changing pressures, Eq. 22 becomes for slightly compressible fluids
h = 2xk&, -P,,,) . ,_.....................
It is convenient in many applications of Darcy s law to introduce commonly used oilfield units. The following is a summary of the more common equations with the conversion factors to convert to oilfield terminology.
Linear-Flow Liquids (or Gases with Volume at Mean Pressure). Rate, BID, is given by
q=1.1271
. . . .
. .
(26)
AL
I -~2)
.. ....
. . . (27)
In TJr,,.,
Radial-Flow Liquids (or Gases with Volume at Mean Pressure). Rate, liters per day, is given by
where q. is defined at the pressure p. where the density is po. For ideal gases, Eq. 22 becomes 7rkh(p<, -~,,.~) 4/J = Wb ln r,,r,, or 911= 2dd-0, -p w) cLln r ,,ra, , . . . . (25)
q=92.349x
103
Wp,
-PM )
. .
..
(28)
. (24) q-92.349x
(29)
Gases at Base-Pressure, p,,, and Average Flowing Temperature, Tf. Linear flow rate, cu ft/D. is given by
where q,, is the volume rate of flow at the algebraic mean pressure (p, +p,,,)/2.
3.1615T,,kA(p, Yb = TfW&PlJ
26-14
PETROLEUMENGINEERING HANDBOOK
Pe
Fig.
Fig. 26.21~Radial
flow-parallel
combination of beds.
and radial flow rate, cu ft/D, is given by* q/J= 19.88T~kh(p,2 -p,,,?) .,............. (31)
Flow Systems
of Combinations
of Beds
Tfq.4,~p~ In rplr,,,
where k is in darcies; A is in sq ft; h is in ft; p , , p2, pp, p ,,.. and Ph are in psia; p is in cp; L is in ft; and re and T,, are in consistent units. Since the previous equations describe the flow in the medium, appropriate volume factors must be introduced to account for changes in the fluids caused by any decrease in pressure and temperature from that of the medium to standard or stock-tank conditions.
Permeability Conversion Factors.
Consider the case where the flow system comprises layers of porous rock separated from each other by infinitely thin, impermeable barriers as shown in Fig. 26.20. The average permeability k may be evaluated by Eq. 32.
(32)
C j=l hj
1 darcy (d)= 1,000 millidarcies (md) = (cm3/s)cp cm (atm/cm) (cm3/s) cp =9.869x 10 - ,,* =9.869x =1.062x =7.324x lop9 cm2 IO- lop5 sq ft (cu ft/sec)cp sq ft(psi/ft) =9.697x10p4 WD)cp sq ft (psi/ft) = 1.424 x 10 - (gal/min)cp sq ft (ft water/ft) (cu ft/sec)cp cm (cm water/cm) =1.127 (dynIrm*)
Fig. 26.21 shows that the same terms appear in the radial-flow network as in the linear system. The only difference in the two systems is the manner of expressing the length over which the pressure drop occurs. Because all these terms are the same in each of the parallel layers, an evaluation of the parallel radial system yields the same solution as obtained in the linear case.
Example Problem 6-Average Permeability of Beds in Parallel. What is the equivalent linear permeability of
four parallel beds having equal widths and lengths under the following conditions?
Pay
Bed 1 2 3 4 i:
j& j=l
Thickness (ft) 20 15
10
kihj
.
ihi j=I
10)+(400x5)
md.
26-15
AP,
AP,
AP,
Fig. 26.22-Linear
flow-series
combination of beds.
Fig. 26.23-Radial
flow-series
combination of beds
Another possible combination for flow systems is to have the beds of different permeabilities to be arranged in series as shown in Fig. 26.22. In the case of linear flow, the average series permeability may be evaluated by Eq. 33. k=
J==I
Therefore, j& 250+2.50+500+ 2.50 250 500 1,000 1,000 200 =80 md.
L LJ lkj
. . .
.. .
.(33)
The same reasoning can be used in the evaluation of the average permeability for the radial system (Fig. 26.23) so as to yield
In
log 2.000/0.5
k=
log 25010.5 25
log 500/2so SO
kc = j&l
r,,/r,,.
rjlr,
=30.4 md. -1
Permeability of Channels and Fractures
In
h, .I
Only the matrix permeability has been discussed in the analysis to this point. In some sand and carbonate reservoirs, the formation frequently contains solution channels and natural or artificial fractures. These channels and fractures do not change the permeability of the matrix but do change the effective permeability of the flow network.
Circular Channel. Equating Darcy s and Poiseuille s equations for fluid flow in a tube, the permeability may be expressed as a function of the tube radius.
Example Problem 7-Average Permeability of Beds in Series. What is the equivalent permeability of four
beds in series, having equal formation thicknesses under the following conditions? Assume Bed 1 adjacent to the wellbore (1) for a linear system and (2) for a radial system if the radius of the penetrating wellbore is 6 in. and the radius of effective drainage is 2,000 ft. Length of Bed @I 250 250 500 1,000 Horizontal Permeability (md) 25 50 100 200
k=;,
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(35)
where k and r are in consistent units. If r is in centimeters, then k in darcies is given by k= r2 8(9.869x 10 -9) factor from the I 12.50X 106r ,
k=12.50~
106(2.54)*r2
=80~10~r*=20~10~d*,
26-16
PETROLEUM ENGINEERING
HANDBOOK
~~~~~~
01
03
07
09
opening
where p u L A = = = = resistivity, Q-cm, llp=conductivity, length of flow path, cm, and cross-sectional area of conductor,
Fracture. For flow throu h slots of fine clearances and unit width, Buckingham - 4 reports that
cmZ.
Ap=-
12pL h?
k=;.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(36)
the
k= 12(9.869x 10-9)
=84.4x
10 h2,
k
I*
1 and P
and when h is in inches and k is in darcies, permeability is given by k=54.4 x 106h2, so that the permeability of a fracture 0.01 in. in thickness would be 5,440 darcies or 5,440,OOO md.
Physical Analogies to Darcy s Law
-4 =- E L L The Fourier heat equation may be written as d=k,aF, where Q= A = AT = L = kh = rate of heat flow, Btulhr, cross-sectional area, sq ft, temperature drop, F, length of the conductor, ft, and thermal conductivity, Btu/(ft-hr-F).
A comparison with Darcy s law indicates that s=Q, k -=kl,, and P Ap AT -=L L
r=p-
26-17
Electrical and heat models (based on these analogies) of rock and well systems frequently are used to solve fluidflow problems involving complex geometry.
Measurement of Permeability
The permeability of a porous medium may be determined from samples extracted from the formation or by in-place testing. The procedures discussed in this section pertain to the permeability determinations on small samples of extracted media. Two methods are used to evaluate the permeability of cores. The method most used on clean, fairly uniform formations uses small cylindrical samples called perm plugs that are approximately % in. in diameter and 1 in. in length. The second method uses full-diameter core samples in lengths of 1 to 1% ft. The fluids used with either method may be gas or any nonreactive liquid.
Method. As core samples ordinarily contain residual oil, water, and gas, it is necessary that the sample be subjected to preparation before the determination of the permeability. The residual fluids normally are extracted by retorting or solvent extraction. The core is dried before permeability measurements are taken. Air commonly is used as the fluid in permeability measurements. The requirement that the permeability be determined for conditions of viscous flow is best satisfied by obtaining data at several flow rates and plotting results, as shown in Fig. 26.24, based on either Eq. 17 or 18. For conditions of viscous flow, the data should plot a straight line, passing through the origin. Turbulence is indicated by curvature of the plotted points. The slope of the straight-line portion of the curve is equal to k/p, from which the permeability may be computed. To obtain k in darcies, 4 must be in cm3/s, A in cm , p1 and p2 in atm. L in cm, and ~1in cp. A permeameter designed for the determination of the permeability of rocks with either gas or liquid is illustrated in Fig. 26.25. Data ordinarily are taken from this device at only one flow rate. To assure conditions of viscous flow, the lowest possible rate that can be measured accurately is used. Perm-Plug Example Problem S-Permeability Measurement.
Fig. 26.25-Ruska
universal permeameter.
1.63x2x0.02 2 x0.45
x 1,000
=72.5 md. Assuming that the data indicated were obtained, but water was used as the flowing medium, compute the permeability of the core. The viscosity of water at test temperature was 1.O cp. v 1,000 4 - 50040 t and
.-
The following data were obtained during a routine penneability test. Compute the permeability of this core. 1. Flow rate= 1,000 cm3 of air at 1 atm absolute and 70F in 500 seconds. 2. Pressure, downstream side of core= 1 atm absolute; flowing temperature, 70F. 3. Viscosity of air at test temperature=0.02 cp. 4. Cross-sectional area of core=2.0 cm*. 5. Length of core=2 cm. 6. Pressure, upstream side of core= 1.45 atm absolute.
PI VI fP?V2 =Pv
k=!!kx,,)oO= A AP
Whole-Core
2X1X2 ~ 2x0.45
1,000=4,450
md.
where 1 is upstream conditions and 2 is downstream conditions, and 1.45+1 P +Pz E-=1.225, 2 2
P=-
Measurement. The core must be prepared in the same manner as perm plugs. The core is then mounted in special holding devices as shown in Fig. 26.26. The measurements required are the same as the perm plugs but the calculations are slightly different.
26-18
PETROLEUMENGINEERINGHANDBOOK
1
-GASKET OPENING
AIRPRESSURE
END VIEW
TOP
VIEW
SPL T
TELOWMETER
GASKET1
b CARBON / I I I
DIOXIDE I 1 J
for large
cores.
Fig. 26.27-Permeability constant of core sample L to hydrogen, nitrogen, and COP at different pressures (permeability constant to iso-octane, 2.55 md).
In the case of the clamp-type permeameter, the geometry of the flow paths is complex, and an appropriate shape factor must be applied to the data to compute the permeability of the sample. The shape factor is a function of the core diameter and the size of the gasket opening. The shape factor affects the quantity L/A in the previous equations.
Factors Affecting Permeability Measurements
others established that the permeability of a porous medium to a nonreactive homogeneous single-phase liquid was equal to the equivalent liquid permeability. The linear relationship between the observed permeability and the reciprocal of mean pressure may be expressed as follows.
kg kL=1 +blp
=k, -ml, P
....
(37)
In the techniques of permeability measurement previously discussed, certain precautions must be exercised to obtain accurate results. When gas is being used as the measuring fluid, corrections must be made for gas slippage. When liquid is the testing fluid, care must be taken that it does not react with the solids in the core sample. Also, corrections may be applied for the change in permeability because of the reduction in confining pressure on the sample.
Effect of Gas Slippage on Permeability Measurements
Klinkenberg I4 has reported variations in permeability determined by using gases as the flowing fluid from that determined by using nonreactive liquids. These variations were ascribed to slippage, a phenomenon well known with respect to gas flow in capillary tubes. The phenomenon of gas slippage occurs when the diameter of the capillary openings approaches the mean free path of the gas. Fig. 26.27 is a plot of the permeability of a porous medium as determined at various mean pressures using hydrogen, nitrogen, and carbon dioxide as the flowing fluids. Note that for each gas a straight line is obtained for the observed permeability as a function of the reciprocal of the mean pressure of the test. The data obtained with the lowest-molecular-weight gas yield the straight line with the greatest slope, indicative of a greater slippage effect. All the lines when extrapolated to infinite mean pressure (l/j=O) intercept the permeability axis at a common point. This point is designated as the Klinkenberg and equivalent liquid permeability,
where kL = permeability of the medium to a single liquid phase completely filling the pores of the medium, k, = permeability of the medium to a gas completely filling the pores of the medium, j = mean flowing pressure of the gas at which k, was observed, b = constant for a given gas in a given medium, and m = slope of the curve.
kL.
Reactive Fluids. While water commonly is considered to be nonreactive in the ordinary sense, the occurrence of swelling clays in many reservoir rock materials results in water s being the most frequently occurring reactive liquid in connection with permeability determinations. Reactive liquids alter the internal geometry of the porous medium. This phenomenon does not vitiate Darcy s law but rather results in a new porous medium, the permeability of which is determined by the new internal geometry. While fresh water may cause the cementation material in a core to swell because of hydration, it is a reversible process. A highly saline water may be flowed through the core and return the permeability to its original value. The effect of water salinity on permeability is shown in Table 24.3. 25
26-19
TABLE 26.3-EFFECT
OF NATURAL
Field ~ S : S S s S S S S S S s S
T
-Zone 34 34 34 34 34 34 34 34 34 34 34 34 34 36 36 ;: 36 36
300
k 200
k 100
kw
T T T T T
4,080 24,800 40,100 39,700 12,000 4.850 22,800 34.800 27,000 12,500 13,600 7,640 11,100 6,500 2,630 3,340 2,640 3,360 4,020 3,090
1,445 11,800 23,000 20,400 5,450 1,910 13,600 23,600 21,000 4,750 5,160 1,788 4,250 2,380 2,180 2,820 2,040 2,500 3,180 2,080
1,380 10,600 18.600 17,600 4,550 1,430 6,150 7,800 15,400 2,800 4,640 1,840 2,520 2,080 2,140 2,730 1,920 2,400 2,900 1,900
1,290 10,000 15,300 17,300 4,600 925 4,010 5,460 13,100 1,680 4,200 2,010 1,500 1,585 2,080 2,700 1,860 2,340 2,860 1,750
1,190 9,000 13,800 17,100 4,510 736 3,490 5,220 12,900 973 4,150 2,540 866 1,230 2,150 2,690 1,860 2,340 2,820 1,630
885 7,400 8,200 14,300 3,280 326 1,970 3,860 10,900 157 2,790 2,020 180 794 2,010 2,490 1,860 2,280 2,650 1,490
17.2 147 270 1,680 167 5.0 19.5 9.9 1,030 2.4 197 119 6.2 4.1 1,960 2,460 1,550 2,060 2,460 1,040
10 500.gram/gal
Care must be taken that laboratory permeability values are corrected to liquid values obtained with water whose salinity corresponds to formation water.
Overburden Pressure. When the core is removed from the formation, all the confining forces are removed. The rock matrix is permitted to expand in all directions, partially changing the shapes of the fluid-flow paths inside the core. Compaction of the core caused by overburden pressure may cause as much as a 25 % reduction in the permeabil ity of various formations, as observed in Fig. 26.28.2h Note that some formations are much more compressible than others; thus, more data are required to develop empirical correlations that will permit the correction of surface permeability for overburden pressures. Factors in Evaluation Other Parameters of Permeability From
If a porous system is conceived to be a bundle of capillary tubes, then it can be shown that the permeability of the medium depends on the pore-size distribution and porosity. A flow network of tubes would be similar to layers of different permeability in parallel, such that the average permeability could be calculated by adapting Eq. 32 to read m
C kjAj
J=I . WI . . . . . . . ..I..............
(38)
J=l
where kj is the permeability of one capillary tube and Aj is the area of flow represented by a bundle of tubes of permeability ki.
,Permeability. like porosity, is a variable that can be measured for each rock sample. To aid in understanding fluid flow in rocks and possibly to reduce the number of measurements required on rocks, correlations among porosity, permeability, surface area, pore size, and other variables have been made. The reasoning behind some of the correlations among porosity, permeability, and surface area is presented here to enable the reader to gain some understanding of the interrelation of the physical properties of rocks. Although these relations are not quantitative, they are indicative of the interdependence of rock characteristics.
Use of Capillary Tubes for Flow
--
lCURVE
nr
SOURCE rr.n.-
PERMEABILITY . .^
POROSITY --l?rruT
derived
ti d 60 3 zt I t
:z
2
~WXIERN
CALIFORNIA BASIN
335
I10
25 22 5000
y 50 5
Fig. 26.28-Change
26-20
PETROLEUM ENGINEERING
HANDBOOK
The quantities k, and A, can be defined in terms of the radius of capillary tubes.
A,=sn Ir I
Wyllie* derived the Kozeny equation from Poiseuille s law by using a specified flow network. The resulting permeability for this flow network is given by Eq. 43. , . . . . . . . . . . . . . . . . . . . . . . (43)
and 7 k;+, where k = F,Y = L = L, = If Lo 2 where 4 is the porosity of the flow network and A, is the total cross-sectional area of the flow network. By substitution, Eq. 38 reduced to (T >
permeability of the porous medium, shape factor, length of the sample, and actual length of the flow path.
5 A,=$A,, J=l
and K, =F,Y7=Kozeny constant, then Wyllie s equation will reduce to the same form as Eq. 42.
,,1
Fluid Saturations
(39)
where li is the average permeability of the tube bundle. Note that the permeability of a bundle of tubes is a function not only of the pore size but of the arrangement or porosity of the system. Consider a system that comprises a bundle of capillary tubes of the same radii and length; k, the permeability, may be written from Eq. 39 as
7
k=$.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(40)
In the previous sections of this chapter, the storage and the conduction capacity of a porous rock were discussed. To the engineer there is yet another important factor to be determined-i.e., the fluid content of the rock. In most oil-bearing formations, it is believed that the rock was completely saturated with water before the invasion of the rock by petroleum. The oil will not displace all the water from the pore space. Therefore, to determine the quantity of hydrocarbons accumulated in a porous rock formation, it is necessary to determine the fluid saturation (oil, water, and gas) of the rock. There are two approaches to the problem of determining the fluid saturations within a reservoir rock. The direct approach is to measure. in the laboratory, the saturations of selected rock samples recovered from the parent formation. The indirect approach is to determine the fluid saturation by measuring some related physical property of the rock.
Factors Affecting Fluid Saturations
The internal surface area per unit PV, A,,, may be defined in terms of the tube radius by
of Cores
A,,=?.
,.,.....,.,,.,,_...,..........,
.(41)
Combining Eqs. 40 and 41 gives the permeability as a function of porosity and internal surface area. This function is
If l/K; is substituted for the constant l/z, the resulting expression is Kozeny s equation wherein K; is defined as the Kozeny s constant.
4 k=- K,A,
~~~~..........,......___,,,,.
(42)
The core samples delivered to the laboratory for fluidsaturation determinations are obtained from the ground by either rotary, sidewall, or cable-tool coring. In all cases, the fluid content of these samples has been altered by two processes. First, in the case of rotary drilling, the mud column exerts a greater pressure at the formation wellbore surface than the fluid in the formation. The differential pressure between the mud column and the formation fluids causes mud and mud filtrate to invade the formation, thus flushing the formation with mud and its fyltrate. The filtrate displaces some of the original fluids. This displacement process changes the original fluid contents of the in-place rock. Second. as the sample is brought to the surface, the confining pressure of the fluid column is constantly decreasing. The reduction of pressure permits the expansion of the entrapped water, ail, and gas. Gas. having the greater coefficient of expansion, expels oil and water from the core. Thus, the content of the core at the surface has been changed from that which existed in the formation. Because the invasion
26-2 1
of the filtrate precedes the core bit, it is not possible to use pressurized core barrels to obtain undisturbed samples. Drill cuttings, chips, or cores from cable-tool drilling also have undergone definite physical changes. If little or no fluid is maintained in the wellbore, the formation adjacent to the well surface is depleted because of pressure reduction. As chips are formed in the well, they may or may not be invaded, depending on the fluids in the wellbore and the physical properties of the rock. In all probability, fluid will permeate this depleted sample, resulting in flushing. Thus, even cable-tool cores undergo the same two processes as was noted in the case of rotary coring, although in reverse order. Sidewall cores from either rotary- or cable-tool-drilled holes are subjected to these same processes. In an attempt to understand better the overall effect of the physical changes that occur in the core because of flushing and fluid expansion, Kennedy et al. 28 undertook a study to simulate rotary-coring techniques. The effects of both invasion and expansion because of pressure reduction were measured. Schematics of the changes in saturation resulting from these two processes for oil- and water-based muds are shown in Fig. 26.29. For the water-based mud, the original displacing action of the water filtrate reduced the oil saturation by approximately 14%. The expansion to surface pressure displaced water and additional oil. The final water saturation was greater than the water saturation before coring. With oil-based mud, wherein the filtrate is oil, the displacing action did not alter the initial water saturations but did result in replacement of approximately 20% of the initial oil. On pressure depletion, a small fraction of the water was expelled, reducing the water saturation from 49.1 to 47.7 % The oil saturation was reduced by both processes from 50.9 to 26.7%. Thus, even when high water saturations are involved, up to approximately 50%, the water-saturation values obtained with oil-based muds may be considered to be representative of the initial water saturations in the reservoir. Hence, it is possible to obtain fairly representative values of in-place water saturations by using oil-based muds. Attempts have been made to use tracers in the drilling fluid to determine the amount of water in the core that is caused by mud-filtrate invasion. The theory was that mud filtrate displaced only oil. Thus. upon recovering the core to the surface, the salt concentration of the core water could be determined. Thus, if the salt concentration in the reservoir water and the tracer concentration in the drilling fluid were known, the volume of filtrate and reservoir water in the core could be calculated. Because a large fraction of the initial reservoir water may have been displaced by the invading filtrate, the tracer method results in incorrect values for reservoir-water saturation. To obtain realistic values of fluid saturation, it is necessary to choose the proper drilling fluid or to use indirect methods of saturation determination.
Determination Rock Samples of Fluid Saturations from
RESIDUAL SATURATION
(a)
ORIGINAL SATURATION RESIDUAL SATURATION
T---l
SATURATION AFTER MUD FLUSHING BEFORE PRESSURE REDUCTION OIL BASE MUD (b) Fig. 26.29-Typical changes in saturations of cores flushed with water-based and oil-based muds.
vaporize the water and the oil, which is condensed and collected in a small receiving vessel. The retort method has several disadvantages as far as commercial work is concerned. The water of crystallization within the rock is driven off, causing the water-recovery values to be too high. The second error that occurs from retorting samples is that the oil, when heated to high temperatures, has a tendency to crack and coke. This change of hydrocarbon molecules tends to decrease the liquid volume. The fluid wetting characteristics of the sample surface may be altered during the retorting process as a result of the two previous factors. Before retorts can be used, calibration curves must be prepared for water and oils of various gravities to correct for losses and other errors. These curves can be obtained by running blank runs (retorting known volumes of fluids of known properties). The retort is a rapid method for determining fluid saturations and, if the corrections are used, yields satisfactory results. It gives both water and oil volumes such that the oil and water saturations can be calculated from the following equations.
S&,
P S, +, P
One of the most popular means of measuring fluid saturations of cores is the retort method. This method takes a small rock sample and heats the sample to
26-22
and S,=l-S,-S,,
where s, s, s, v, VP
water saturation, oil saturation, gas saturation, water volume, cm3, pore volume, cm3, and 3 v, = oil volume, cm .
= = = = =
The other method of determining fluid saturation is by extraction with a solvent. Extraction may be accomplished by a modified ASTM distillation method or a centrifuge method. In the standard distillation test, the core is placed such that a vapor of either toluene, pentane, octane, or naphtha rises through the core. This process leaches out the oil and water in the core. The water and extracting fluid are condensed and collected in a graduated receiving tube. The water settles to the bottom of the receiving tube because of its greater density, and the extracting fluid refluxes over the core and into the main heating vessel. The process is continued until no more water is collected in the receiving tube. The water saturation may be determined directly by
S,,=v,;. VP
The oil saturation is an indirect determination. The oil saturation as a fraction of PV is given by
The gas saturation is obtained in the same manner as the retort. Another method of determining water saturation is to use a centrifuge. A solvent is injected into the centrifuge just off center. Because of centrifugal force, it is thrown to the outer radii and forced to pass through the core sample. The outflow fluid is trapped and the quantity of water in the core is determined. The use of the centrifuge provides a very rapid method because of the high forces that can be applied. In both extraction methods, at the same time that the water content is determined, the core is cleaned in preparation for the other measurements such as porosity and permeability. There is another procedure for saturation determination that is used in conjunction with either of the extraction methods. The core as received from the well is placed in a modified mercury porosimeter in which the BV and gas volume are measured. The volume of water is determined by one of the extraction methods. The fluid saturations can be calculated from these data. In connection with all procedures for determination of fluid content, a value of PV must be established in order that fluid saturations may be expressed as percent of PV. Any of the porosity procedures previously described may be used. Also, the BV and gas volume determined from the mercury porosimeter may be combined with the oil and water volumes obtained from the retort to calculate PV, porosity, and fluid saturations. Porosity, permeability, and fluid-saturation determinations are the measurements commonly reported in routine core analysis. A laboratory equipped for such determinations is shown in Fig. 26.30. Interstitial Water Saturations Essentially, three methods are available to the reservoir engineer for the determination of interstitial water saturations. These methods are (1) determination from cores cut with oil-based muds, (2) determination from capillary-pressure data, and (3) calculation from electriclog analysis (see Chap. 49). Oil-Based Mud. The obtaining of water saturations by using oil-based muds has been discussed. A correlation between water saturation and air permeability for cores obtained with oil-based muds is shown in Fig. 26.31. 29
so=
where WII = weight of wet core, g, W td = weight of dry core, g, W, = weight of water, g, 3 VP = PV, cm , and 3 PO = density of oil, g/cm .
26-23
A general trend of increasing water saturation with decreasing permeability is indicated. It is accepted from field and experimental evidence that the water content determined from cores cut with oil-based mud reflects closely the water saturation as it exists in a reservoir, except in transition zones where some of the interstitial water is replaced by filtrate or displaced by gas expansion. Fig. 26.32 shows permeability/interstitial-water relationships reported in the literature for a number of fields and areas. There is no general correlation applicable to all fields; however, an approximately linear correlation between interstitial water and the logarithm of permeability exists for each individual field. The general trend of the correlation is decreasing interstitial water with increasing permeability.
Capillary Pressure. Capillary pressure may be thought of as a force per unit area resulting from the interaction of surface forces and the geometry of the medium in which they exist. Capillary pressure for a capillary tube is defined in terms of the inter-facial tension between the fluids, a. the angle of contact of the interface of these two fluids and the tube. 0(., and the radius of the tube, VI. This relationship is expressed in Eq. 44.
6 AIR
20 PERMEABILITY,
II I
40
60 100 MILLIDARCYS
ttic: t t ttn
200 400
Fig. 26.31--Relation of air permeability to the water content of the South Coles Levee cores.
vature (these radii touch at only one point), r, and r7 _ and the interfacial tension of the fluids. This relationship is given in Eq. 4.5.
(45)
P,. =
20 cos 8,. rt
Comparing Eq. 4.5 with the equation for capillary pressure as determined by the capillary-tube method, it is found that the mean radius U is defined by I 1 I 2 cos 0 _=-+-=-.--.-------' r rt r2 rf
rir2 y=-=-.--rl +r2 rf 2cos*,.. . . . . . . . . . . . . . .
, ,,..,.....................
(44)
where the angle 0,. is measured through the more dense fluid. In a packing of spheres, the capillary pressure is expressed in terms of any two perpendicular radii of cur-
(46)
LEGEND I = HAWKINS 2: MAGNOLIA 3= WASHINGTON 4=ELK BASIN 5= RANGELY 6: CREOLE 7= SYNTHETIC ALUNDUM 8: LAKE ST. JOHN
C
fil
9= LOUISIANA
MIOCENEAGE-WELL
IO 20
30 40
50 60 70 80 90 100
26-24
PETROLEUM ENGINEERING
HANDBOOK
MEASURING APPARATUS
method
of
permeable membrane of uniform pore-size distribution containing pores of such size that the selected displacing fluid will not penetrate the diaphragm when the pressures applied to the displacing phase are below some selected maximum pressure of investigation. Various materials including fritted glass, porcelain, and cellophane have been used successfully as diaphragms. Pressure applied to the assembly is increased by small increments. The core is allowed to approach a state of static equilibrium at each pressure level. The saturation of the core is calculated at each point defining the capillary-pressure curve. Any combination of fluids may be used: gas, oil, and/or water. Although most determinations of capillary pressure by the diaphragm method are drainage tests, by suitable modifications imbibition curves similar to Leverett s may be obtained. The mercury-capillary-pressure apparatus was developed to accelerate the determination of the capillary-pressure/saturation relationship. Mercury is normally a nonwetting fluid. The core sample is inserted in the mercury chamber and evacuated. Mercury is forced into the core under pressure. The volume of merat each pressure determines the cury injected nonwetting-phase saturation. This procedure is continued until the core sample is filled with mercury or the injection pressure reaches some predetermined value. Two important advantages are gained by this method: (1) the time for determination is reduced to a few minutes, and (2) the range of pressure investigation is increased because the limitation of the diaphragm s properties is removed. Disadvantages are the difference in wetting properties and permanent loss of the core sample.
Mercury Injection. Centrifuge Method. A third method for determining capillary properties of reservoir rocks is the centrifuge method.3 The high accelerations in the centrifuge increase the field of force on the fluids, in effect subjecting the core to an increased gravitational force. By rotating the sample at various constant speeds, a complete capillary-pressure curve may be obtained. The speed of rotation is converted into force units in the center of the core sample, and the fluid saturation is read visually by the operator. The advantage of the method is the increased speed of obtaining the data. A complete curve may be established in a few hours, while the diaphragm method requires days. Dynamic Method. Brown3* reported the results of determining capillary-pressure/saturation curves by a dynamic method. Simultaneous steady-state flow of two fluids is established in the core. By the use of special wetted disks that permitted hydraulic pressure transmission of only the selected fluid phase, the difference in the resulting measured pressures of the two fluids in the core is the capillary pressure. The saturation is varied by regulating the quantity of each fluid entering the core. Thus, it is possible to obtain a complete capillarypressure curve. Comparison of Methods of Measurement
It is practically impossible to measure the values of rl and t-2; hence, they generally are referred to by the mean radius of curvature and empirically determined from other measurements on a porous medium. The distribution of the liquid in a porous system depends on the wetting characteristics. It is necessary to determine which is the wetting fluid so as to ascertain which fluid occupies the small pore spaces. From packings of spheres, the wetting-phase distribution within a porous system has been described as either funicular or pendular in nature. In funicular distribution, the wetting phase is continuous, completely covering the surface of the solid. The pendular ring is a state of saturation in which the wetting phase is not continuous and the nonwetting phase is in contact with some of the solid surface. The wetting phase occupies the smaller interstices. As the wetting-phase saturation progresses from the funicular to the pendular-ring distribution, the volume of the wetting phase decreases and the mean radius of curvature or the values of rl and r2 tend to decrease in magnitude. Referring to Eq. 46, we see that if r I and r2 decrease in size, the magnitude of the capillary pressure would have to increase in value. Since r , and r2 can be related to the wetting-phase saturation, it is possible to express the capillary pressure as a function of fluid saturation when two immiscible phases are within the porous matrix.
Laboratory Measurements of Capillary Pressure
Essentially, five methods of measuring capillary pressure on small core samples are used. These methods are (1) the desatumtion or displacement process, through a porous diaphragm or membrane (restored-state method of Welge3 ), (2) the mercury-injection method, (3) the centrifuge or centrifugal method, (4) the dynamiccapillary-pressure method, and (5) the evaporation method. *
Porous Diaphragm. The first of these, illustrated in Fig. 26.33, is the displacement or diaphragm method. The essential requirement of the diaphragm method is a
Since the method IS seldom used today, it WIII not be dwzussed The procedure conslsts ot continuously monitoring the decrease m weight caused by evaporabon of a core sample lnitlally 100% saturated wlh a wetltng fluld. See Messner, E S : Intetsbtial Water Determmation By An Evaporatmn Method. Trans, AIME (1951) 192,269-74
Intuitively, it appears that the diaphragm method (restored state) is superior in that oil and water are used; therefore, actual wetting conditions are more nearly ap-
proached. Hence, the diaphragm method is used as the standard to which all other methods are compared. The mercury-injection test data must be corrected for wetting conditions before they can be compared with results from the restored-state method. If it is assumed that the mean curvature of an interface in rock is a unique function of fluid saturation, then the ratio of mercury to water capillary-pressure data is given by
+-+f+6.57. M 0,.
. .
. .
(47)
DEPTH FEET BELOW SEA-LEVEL
where urn is the surface tension of mercury and u, is the surface tension of water. Experimentation has shown the ratio to vary between 5.8 (for limestones) and 7.5 (for sandstones). Thus, no conversion factor can be defined that will apply to all rocks. Good agreement of centrifuge data with those from the diaphragm method was reported by Slobod. 31 Unlike the mercury-injection method, there is no need of conversion factors to correct for wetting properties. The same fluids are used in the centrifugal and diaphragm methods. Excellent correlation was obtained by Brown3 between the diaphragm and dynamic methods. The dynamic data were obtained by simultaneous steady flow of oil and gas through the porous sample at a predetermined level of pressure difference between the fluids. Care was taken to maintain uniform saturations throughout the core as well as to conduct the test such that a close correspondence to drainage conditions existed. If capillary-pressure data are to be used for determining fluid saturations, the values obtained should be comparable with those of other methods. Water distributions, as determined from electric logs, and capillary-pressure data are normally in good agreement. A comparison of these methods is shown in Fig. 26.34. 33 Shown also is the approximate position of the gas/oil contact as determined from other test data. In the gas-bearing portion of the formation there is no significant variation in water saturation with depth or method of determination. However, in a thin oil zone, such as that shown in Fig. 26.34. there is a significant variation in the water saturation with depth. Variations in water saturations with depth within an oil zone must be taken into account to determine accurately average reservoir interstitial-water saturations.
Water Saturation from Capillary-Pressure Data. In
Fig. 26.34-Comparison
of water saturation
from capillary
gas/water or an oil/water system that normally does not have the same physical properties as the reservoir water, oil, and gas. Essentially two techniques, differing only in the initial assumptions, are available for correcting laboratory capillary-pressure data to reservoir conditions. uw><, cos 0,.,,
p,..R= uwg cos 0,., p,., L
or
P,..R=~P&
(JL
...
. (49)
Since the interfacial tensions enter as a ratio, pressure in any consistent units may be used together with the interfacial tension in dynes/cm.
Averaging Capillary-Pressure Data. Two methods have been proposed for correlating capillary-pressure data of similar geologic formations. The first correlating procedure is a dimensionless grouping of the physical properties of the rock and the saturating fluids. This function is called a J function34 and is expressed as
P,.=+(p,
-p2),
.......................
where h is in feet and p, and p2 are the densities of Fluids 1 and 2, respectively, in lbmicu ft at the conditions of the capillary pressure.
Converting Laboratory Data. To use laboratory capillary-pressure data, it is necessary to convert them to reservoir conditions. Laboratory data are obtained with a
water saturation, fraction of PV, capillary pressure, dyne/cm*, interfacial tension, dyne/cm, permeability, cm*, and fractional porosity.
26-26
PETROLEUM
ENGINEERING
HANDBOOK
90
81 z
72 I+ 2
63 3 g 54 ai: 3;
45 2 ,o 36 gs
732
2, 2 2 3& I8 z
3
The results of the statistical correlation previously discussed applied to the capillary-pressure data presented in Fig. 26.35 are shown in Fig. 26.36. The reader should note the linearity of the curves for each value of capillary pressure and the tendency of all capillary-pressure curves to converge at high permeability values. This behavior is what normally would be expected because of the larger capillaries associated with high permeabilities. To convert capillary-pressure saturation data to height saturation, it is necessary only to rearrange the terms in Eq. 48 so as to solve for the height instead of the capillary pressure-i.e., h,,=-, P,. x 144
PW -PO
..
. ..
. . . (52)
20
30
40 FLUID
50
60
70
80
90
IO8
WATER RESERVOIR
where hh
Fig. 26.35~Series
P II =
Some authors alter the above expression by including the cos f3(.(where 8,. is the contact angle) as follows.
= height above the free-water surface, ft, density of water at reservoir conditions, lbm/cu ft, PO = density of oil at reservoir conditions, lbm/cu ft, and P,. = capillary pressure at some particular saturation for reservoir conditions (it must be converted from laboratory data first, psi).
The J function originally was proposed as a means of converting all capillary-pressure data to a universal curve. There exist significant differences in correlation of the J function with water saturation from formation to formation such that no universal curve may be obtained, but the / function may be used to correlate the data from one formation. The second method of evaluating capillary-pressure data is to analyze a number of representative samples and treat the data statistically to derive correlations that, together with the porosity and permeability distribution data, may be used to compute the interstitial-water saturations for a field. A first approximation for the correlation of capillary-pressure data is to plot water saturation against the logarithm of permeability for constant values of capillary pressure. A straight line may be fitted to the data for each value of capillary pressure, and average-capillary-pressure curves may be computed from permeability-distribution data for the field. The resulting straight-line equation takes the general form of
P<.,L = = CJ 11 0 = P M is 8 = PO =
18 24 68 72 53
then, from Eq. 49, P(.,R = 18(24/72)= 18/3=6 psi, and h= Pc,R x 144 P II -PO 6 x 144 =-=58 68-53 ft.
Thus, a water saturation of 35 % exists at a height of 58 ft above the free-water surface. To calculate the fluid saturation in the gas zone, it is necessary to consider all three phases: oil, water, and gas. If all three phases are continuous, it can be shown that
Fluid-distribution curves are reported for several values of permeability, ranging from 10 to 900 md in Fig. 26.35. These data also may be considered to be capillary-pressure curves. The ordinate on the right reflects values of capillary pressure determined by displacing water with air in the laboratory. The ordinates on the left include the corresponding oil/water capillary pressure that would exist at reservoir conditions and the fluid distribution with height above the free-water surface.
where PC., R = capillary pressure at a given height above the free-water surface determined by using water and gas, = capillary pressure at a given height above PC,H%) the free-water surface, using water and oil, and pC.O,q= capillary pressure at a given height above the free-oil surface, using oil and gas.
26-27
If the wetting phase becomes discontinuous, then the wetting-phase saturation takes on a minimum value, and, at all heights above the point of discontinuity, the wetting-phase saturation cannot be less than this minimum value. It is then possible to calculate the fluid saturations above the free-oil surface by the following relations. 1. S,. at h is calculated using oil and water as the continuous phases. 2. SL at h is calculated using oil and gas as the continuous phases and height denoted by the free-oil surface. 3. S,=l-SL andS,=SL-S,, where SL is the total liquid saturation, oil plus water, fraction.
Example Problem lo-Calculation of Water and Oil Saturation in Gas Zone From Capillary-Pressure Data. Let oil-zone thickness, h,, equal 70 ft and
IOOC )-
700
I- IOC -
7c )- 50 -
u R = 72 dynes,
(JOf = (JI, 0= P II = Pn = PO = 50 dynes, 25 dynes, 68 lbm/cu ft 7 Ibm/cu ft, and 53 lbm/cu ft.
% 40 3c I2c I-
From Fig. 26.28 for a 900-md sample, let PI,,L = 54 psi by the method illustrated in Example Problem 9, Pc,R = 18 psi, hfi = height above free-water level= 120 ft, and S,,. = 16% at a height of 70 ft or greater (read from curve). Since the oil zone is only 70 ft thick, the height of 120 ft above the free-water surface must be at least 50 ft into the gas-saturated zone. The first step is to calculate the total fluid saturation, SL, using gas and oil as the continuous phases. h,,,=h,,-h,=120-70=50
hj, Pc,.R = IJ~(PI,
IO- 0
20
40
60
80
(%)
WATER SATURATION
Fig. 26.36-Correlation of water saturation with permeability for various capillary pressures.
and
ft;
-P,h
50/144x(53-7)=501144x46. = 15.97,
It must be understood that the relationships used in Example Problem 10 for calculating the fluid saturations in the gas zone were based on continuity of all three phases. Since this is not normally the case, it might be expected that saturations somewhat different from the calculated values exist. Because the capillary pressure for a discontinuous phase could vary from pore to pore, it is impossible to ascertain the exact relationships that should exist. Hence, the preceding method of calculating fluid distributions is not exact but is usually as accurate as the data available for making the computation.
From Fig. 26.35, for a laboratory capillary pressure of 23 psi and a permeability of 900 md, the total wetting saturation, SL, equals 18 % Therefore. S,,=SL-S,,,=18-16=2X
26-28
PETROLEUM
ENGINEERING
HANDBOOK
The second fundamental notion of electrical properties of porous rock is the resistivity index, IK, which is defined as
(55)
where R, is the true resistivity of the rock system at some particular value of water saturation and Ra is as previously defined. Three idealized representations have been introduced in the literature from which the formation resistivity factor, FR, and the resistivity index , f~, have been related to the porosity, 4, and the rock tortuosity, 7. From Wyllie s2 analysis the relations are F& 4 and
Fig.
26.37-Core-sample
resistivity
cell
IR=7(,--r-, SW
where T,, is the effective rock tortuosity at some water saturation. Cornell and Katz3 presented an analysis of a slightly different model. The relationships developed are as follows. F,=d4 and
conductor when it contains dissolved salts. Current is conducted in water by movement of ions and therefore may be termed electrolytic conduction. The resistivity of a material is the reciprocal of conductivity and commonly is used to define the ability of a material to conduct current. The resistivity of a material is defined by the following equation.
JR=&& se Wyllie and Gardner 37 later presented an analysis based on a probability theory from which the following relationships were obtained. FK=;
and
For electrolytes, p is commonly reported in Q-cm, and r is expressed in ohms, A in cm , and L in cm. In the study of the resistivity of soils and rocks. it has been found that the resistivity may be expressed more conveniently in n-m. To convert to 9-m from O-cm, divide the resistivity in Q-cm by 100. In oilfield practice. the resistivity in Q-m commonly is represented by the symbol R with an appropriate subscript to define the condition to which R applies.
Fundamental Concepts
4
and I IR=-s,,. From the analysis of the electrical properties of the foregoing idealized pore models, general relationships between electrical properties and other physical properties of the rock may be deduced. The formation resistivity factor has been shown to be some function of the porosity and the internal geometry of the rock system. In particular, the formation resistivity factor may be expressed in the following form. FK=I@ -I, (56)
The definition of electrical formation resistivity factor is perhaps the most fundamental concept in considering electrical properties of rocks. The formation resistivity factor as defined by Archie is F/$5 R \I where R. is the resistivity of the rock when saturated with water having a rcsistivity of R,, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(54)
where K is some function of the tortuosity and VI is a function of the number of reductions in pore-opening sizes or closed channels. It is suggested that K should be I or greater. The value of rn has been shown from theory to range from I to 2.
PROPERTIES
OF RESERVOIR
ROCKS
26-29
Both the formation resistivity factor, F,, and the resistivity index. IR, depend on ratios of path length or tortuosities. Therefore, to compute the formation resistivity factor or resistivity index from the relations derived from models, it is necessary to determine the electrical tortuosity. Since direct measurement of the path length is impossible, reliance has been placed primarily on empirical correlations based on laboratory measurements. Winsauer et al. 3X devised a method of determining tortuosity by transit time of ions flowing through a rock under a potential difference. The data obtained were correlated with the product FR+. The resulting correlation of formation resistivity factor, porosity, and tortuosity is given in Eq. 57.
1.67 =F,&
5 5 2 zi 5 F $ II: k-
-60 40
20
FOR
where L, is the actual length of flow path, and L is the length of the sample, or T .~? =F&. . . . . . .(57)
IO 0 6
4 1 HUMBLE RELATION F= 5 ,t-j&tt-
The deviation from the theory is believed to be an indication of the greater complexity of the actual pore system than that of the models on which the theory was based. Archie suggested that the formation resistivity factor could be correlated with porosity and have the form FR=c#- , . . . ., ._.. . . .(58) Fig.
2p-M
1, 2
CONSOLIDATED
FORMATIONS
8 IO
20
40
60 80
POROSITY, =/a
26.38-Comparison correlations. of various formation resistivity factor
where 4 is the fractional porosity and m is the cementation factor. Archie further reported that the cementation factor probably ranged from 1.8 to 2.0 for consolidated sandstones and, for clean unconsolidated sands, was about 1.3.
Measurement of Electrical Resistivity
of Rocks
Laboratory measurements of electrical properties of rocks have been made with a variety of devices. The measurements require a knowledge of the dimensions of the rock, the fluid saturation of the rock, the resistivity of the water contained in the rock, and a suitable resistivity cell in which to test the samples. A simple cell is shown in Fig. 26.37. 39 A sample cut to suitable size is placed in the cell and clamped between electrodes A. Current is then passed through the sample and the potential drop observed. The resistance of the sample is computed from Ohm s law,
r=-.
E I
R, A
L where A is the cross-sectional area of the sample and L is the length of the sample. The saturation conditions of the test may be established at known values before measurement or determined by an extraction procedure after measurement.
In discussing the theory of the formation resistivity factor, it was stated that K should be greater than 1 and that m should be 2 or less. At this time, the discrepancy between theory and experiment must be attributed to the possible effect of conducting solids. Improved correlations should result from considering other parameters, such as permeability, as variables in the relations. A comparison of suggested relationships between porosity and the formation resistivity factor is shown in Fig. 26.38.
26-30
PETROLEUM ENGINEERING
HANDBOOK
A COMPARABLE
CLEAN
SAND
water resistivity of about 0.1 Q-m. Wyllie proposed that the observed effect of clay minerals was similar to having two electric circuits in parallel-the conducting clay minerals and the water-filled pores. Thus, FRO=where = apparent formation resistivity factor, Rosh = resistivity of a shaly sand when 100% saturated with water of resistivity R,,., R,., = resistivity caused by the clay minerals, R,,. = resistivity caused by the distributed water, and FR = true formation resistivity factor of the rock (i.e., the constant value of formation factor approached when the rock contains low-resistivity water).
FRO
Ro.\/z
R,
1 and -==++
Roslr R,./
FRR,,, \
. (60)
4 !,
1.0
R,-WATER
IO
100
RESISTIVITY, OHM-METERS
Effect
of Conductive
Solids
The data presented in Fig. 26.40 represent graphically the confirmation of the relationship expressed in Eq. 60. The graphs were plotted by deWitte@ from data presented by Hill and Milburn.3 The plots are linear and are of the general form 1 -=m-+b, Rosh 1 R,i ....... . . . . . . . . . . . . (61)
Investigations by Wyllie4 Indicate that clays contribute substantially to the conductivity of a rock when the rock is saturated with a low-conductivity water. The effect of water resistivity on the formation resistivity factor for sands containing clay minerals is shown in Fig. 26.39. The formation resistivity factor for a comparable clean (clay-free) sand is a constant. The formation resistivity factor for the clayey sand increases with decreasing water resistivity and approaches a constant value at a
where m is the slope of the line and h is the intercept. Comparing Eq. 60 with Eq. 61, note that m= ~/FR and b= l/R,.[. The curve labeled Suite 1 No. 40 indicates a clean sand because the line passed through the origin, l/R,.,=O. Then thus having a zero intercept l/Rosh =m(l/R,,)=(l/FRR,,), or Ro=F,R,,.. The remaining samples are from shaly sands, which have a finite conductivity of the clay minerals, as indicated by the intercepts of the lines. The linearity of the plots indicates that l/R,, is a constant independent of R,,. This phenomenon may be explained in terms of the ions absorbed on the clay. When the clay is hydrated, the absorbed ions form an ionic conducting path, which is closely bound to the clay. The number of absorbed ions is apparently little changed by the salt concentration of the interstitial water. Eq. 60 may be rearranged to express the apparent formation resistivity factor in terms of R,., and FRR, .
b=
&I
R,, +(R,.,IF,)
As R,,-+O,
-+I Rw
Fig. 26.40-Water-saturated rock conductivity plotted against water conductivity yields these measurements: (A) Suite 1 No. 40; (X) Suite 1 No. 21; (0) Suite 1 No. 4; (0) Suite 2 No. 13: (II!) Suite 6 No. 2. lim
FRY =-
RCl
R,IIFR
=FR.
Therefore, FRY approaches FR as a limit as R, becomes small. This behavior was observed in Fig. 26.39.
26-31
Hill and Milbum43 evaluated 450 samples from both sandstone and limestone formations. The formation resistivity factor was determined at a water resistivity of 0.01 R-m, a value at which the apparent formation resistivity factor, FRa, approaches the formation resistivity factor, FR. They designated the formation resistivity factor as FR,O,oI The data were fitted by the method of least squares to yield .78. F R.0,0,=1.4+- .. . . ._ .(62)
loo
60 40 z 20 2k ; IO m
\ \
WITHOUT
CONDUCTIVE
This equation conforms to the theory previously discussed. They also fitted the data with K in Eq. 56 restricted to a value of 1. This yielded FR,P,o, =$ - 93, which corresponds closely to Archie s onginal expression In summary, Eqs. 58 (with m=2.0) and 59 have been used widely to represent the relation between formation resistivity factor and porosity. Both equations yield results satisfactory for most engineering purposes. However, we propose that Eq. 62 be considered as more valid because the data were taken to minimize the effect of clays. The selection of a particular relation should be based on independent observations on the formations or formations of interest in a given geologic province.
56 ii
4 SOLIDS
F=$ - ,
I.01
.02
.04
WATER
08
.20
40
.80
SATURATION
Fig. 26.41-Effect of conductive solids on the reslstivity-indexvs.-saturation relationship in Stevens sandstone core.
where
Resistivity of Partially Water-Saturated Rocks
A rock containing both water and hydrocarbon has a higher resistivity than when fully saturated with water. The resistivity of partially water-saturated rocks has been shown to be a function of the water saturation, S,,. . From theoretical developments, the following generalization may be drawn. IR=K S,,/ , .. . (63)
k = C#I = F, = A,v = r =
permeability, porosity, fraction, shape factor, internal surface area/unit PV, and Kozeny tortuosity.
7 has been shown to be a function of FR+i.e., 7= (FR#)- , where x ranges between 1 and 2. If the internal surface area is expressed in terms of the mean hydraulic radius, rH, by A,&
rH
where 1~ = R,/Ro, the resistivity index; K is some function of tortuosity; and n is the saturation exponent. Archie compiled and correlated experimental data from various sources from which he suggested that the data could be represented by IR =S,,. G. .. . (64)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(66)
Wyllie confirmed the suggested relationship for clean sands but found that the presence of clays [conductive solids) altered the relationship. A comparison of Archic s relationship and that for a core containing conductive solids is shown in Fig. 26.41. The change in the relationship depends on both the amount of clays and the water resistivity. Therefore. a general correlation for sands containing conductive solids is not available, although dewitte has proposed a method of using Eq. 64 for evaluation of shaly sands.
Use of Electrical Parameters in Characterizing Porous Media
or
@H) 2 =F.
~ &(FR)X
. . . . . . . . . . . . . . . . . .._. (67)
..
..~~~~....
(65)
Eq. 67 perhaps provides an improved basis for the correlation of characteristic physical data of porous media with the electrical formation resistivity factor. The electrical properties of porous rocks, as discussed in the foregoing sections, form the basis of quantitative evaluation of electrical-logging records. In particular, the Humble relation (Eq. 59) has been used widely by service companies in estimating porosity from measurements made with contact resistivity devices such as the Microlog. Eq. 64 forms the basis of interpretation of water saturations from deeper-penetration
26-32
PETROLEUM
ENGINEERING
HANDBOOK
resistivity devices such as the conventional resistivity curves of the standard electrical log or from the various focused electrical-resistivity devices. Improvements in the statistical correlations of electrical properties of rocks will improve the results of such analyses.
Nomenclature
A = cross-sectional area = area of flow represented by a bundle of tubes of permeability k, A, = internal surface area/unit PV h = constant for a given gas in a given medium c = fluid compressibility C = flow coefficient d = diameter E = voltage drop f = friction factor F, = shape factor FR = formation resistivity factor = apparent formation resistivity factor F~ri K = acceleration of gravity h = driving head I?to = height above free-oil level IT,,~ = height above free-water level /f, = thickness ofjth layer I = current IH = resistivity index J = J function, Eq. 48 k = permeability of the medium k,, = permeability to air k,y = permeability of the medium to a gas completely filling the pores of the medium kl, = thermal conductivity k, = permeability of one capillary tube kL = permeability of the medium to a single liquid phase completely filling the pores of the medium k,,. = permeability to fresh water km = permeability to 500-grain/gal chloride solution K = constant of proportionality, Eq. 7 K = some function of the tortuosity. Eq. 63 K; = Kozeny s constant f. = length of flow path or length of the sample L,, = actual length of the flow path L.; = length ofjth layer UI = slope of the curve or cementation factor ~ number of tubes of radius r, n/ p = mean flowing pressure A~YI = pressure loss over length L ~1~ = base pressure Pe = pressure at the external boundary p,\. = pressure at the wellborc P,. = capillary pressure Pc.,l, = capillary pressure. laboratory conditions P,.,,, = capillary pressure of mercury pc~.i,,y= capillary pressure at height above free-oil surface, using oil and gas
A.,
P c .R = capillary pressure, reservoir conditions P,.,,. = capillary pressure of water P c I, (,= capillary pressure at a given height above the free-water surface determined by using water and gas P,,,,,, = capillary pressure at a given height above the free-water surface, using water and oil q/, = volume rate of flow at the base pressure 40 = volume rate of flow of oil q/J = volume rate of flow at the algebraic mean pressure (p, +p,, )/2 qr = total flow rate Q = rate of heat flow r = radius or resistance Y = mean radius r, = radius at the external boundary TH = hydraulic radius r 1,= radius of the wellbore R,., = resistivity caused by the clay minerals Rn = resistivity of the rock when saturated with water having a resistivity of R,, Ro,I, = resistivity of a shaly sand when 100% saturated with water of resistivity R,, s = distance in direction of flow, always positive lTs = gas saturation SL = total wetting saturation S,, = oil saturation S,,. = water saturation T,f = average flowing temperature AT = temperature drop u., = volume flux across a unit area of the porous medium in unit time along flow path s UY = volume flux across a unit area of the porous medium in unit time along flow path x v = fluid velocity Vh = BV V,, = oil volume VP = PV V, = sand-grain volume V,,. = water volume w = mass rate of flow WC.I, = weight of dry core W,.,,. = weight of wet core W,,. = weight of water z = vertical coordinate, considered positive downward 0 = angle between s and the horizontal 19,. = angle of contact of the interface of two tluids and the capillary tube Oc.,,,p= water/gas contact angle fI,.,,i, = water/oil contact angle p = fluid viscosity p = fluid density or resistivity ~1, = fluid density at base pressure P 0 = oil density u = IFT or conductivity ~I11= surface tension of mercury
PROPERTIES
OF RESERVOIR
ROCKS
26-33
surface tension of water water/gas IFT water/oil IFT Kozeny tortuosity effective rock tortuosity fractional porosity effective porosity potential function
k4p I -I+)
FLL
(26)
16. Newman. G.H.: PowVolume Comprcshihllity ofConsol~datet1. Friable. and Unconwlidatcd Reservoir Rock\ Under Hydrostutlc Loading. J. PH. T~I. (Feb. 1973) 129-34. Nonlinear Behavior of Elastic Porou\ 17. Van der Knaap. W.: Media. Trcrns.. AIME (1959) 216. 179-87 18. Krug. J.A.: The Effect of Stress on the Petrophyhical Propcrtics of Some Sandstones. PhD dissertation (T- 1964). Colorado School of Mines. Golden. CO (1977). Biaxial Acoustic and Static Measurement ol 19. Graves, R.M.: Rock Elastic Properties. PhD dissertation (T-2596). Colorado School of Mines, Golden. CO (1982). Comparison of Uniaxial 20. Lachance. D.P. and Anderson. M.A.: Strain and Hydrostatic Stress Pore-Volume Compre%\ibilitic\ in the Nugget Sandstone. paper SPE 11971 prcscntcd at the 1083 SPE Annual Technical Conference and Exhibition. Sdn Francirco. Oct. S-8. 21.
y=92.349x
IO c1 ]n r,lr,,
MP,
-P
,,.I
(28)
qb=23.1454x104
Entrapment of Petroleum Under Hydrodynamic 22. Hubben. M.K.: Conditions. Bull., AAPG (Aug. 1953) 1954-2026 23. Croft, H. 0. : Thrnnr)fl~~~crt~li~.~,Fluid NON (UK/ Hotrr Trrr!wri.c\io/l, McGraw-Hill Book Company Inc.. New York City (1938) 129. The Permeability of Porou\ Media to Liquids 24 Klinkenberg. L-J.: and Gates. Drill. crrtd Prod. Pnrc.. API. Dallas ( lY4 1) 200- 13. Water Pemrcability of Rcxr25. Johnston. N. and Beeson. C.M.: voir Sands, Trw~s.. AIME (1945) 160. 43-S? Reduction in Permeability waith Ovcr26. Fatt. I. and Davis. D.H.: burden Pressure, Trtrm.. AIME (19.52) 195. 329. Application of Electrical 21. Wyllie. M.R.J. and Spangler. M.B.: Resistivity Measurements to Problem of Fluid Flow in Porow Media, Bull., AAPG (Feb. 19.52) 359-403. Saturation 28. Kennedy. H.T., VanMeter, O.E., and Jones. R.G.: Determination of Rotary Cores. P<,f. Efr~r. (Jan. 1954) B.52-B.64 oj Oil29. Gates, G.L., Morris, F.C., and Caraway. W.H.: f$cr Bose Drilhg Nuid Fihrurc on Aucr/~sis of Corr v,frm Sour/~ C&s Lewe, Crdijimtiu urld Rcmgc/~. C&r&> /G/d, technical report, Contract No. RI 4716. USBM (Aug. 1950). The Restored-state Method for 30. Welge. H.J. and Bruce, W.A : Determination of Oil m Place and Connate Water. Drill. crrrrl Prod Prm.. API, Dallas (1947) 166-74 USC of Ccn31. Slobod. R.L., Chambers. A.. and Prehn. W.L. Jr.: trifuge for Determining Connate Water. Residual Oil, and Capillary Pressure Curves of Small Core Samples. Tram.. AIME (1951) 192. 127-34. Capillary Pressure Investigations. Trms., 32. Brown. H.W.: AIME (1951) 192, 67-74. Well Logging Study-Quinduno Field. Roberls 33. Owen, J.D.: County, Texas, paper 593-G presented at the 195.5 AIME Formation Evaluation Symposium, Houston. Oct. 27-28. Capillary Behawor in Porous Solids, Trmv., 34. Leverett, M.C.: AIME (1941) 142, 152-68. Formation Evaluation 35. Wright. H.T. Jr. and Wooddy. L.D. Jr.: of the Borregas and Seeligson Fields, Brooks and Jim Wells Cowties, Texas. paper 591-G presented at the 1955 AIME FormatIon Evaluation Symposium. Houston. Oct. 27-28. Flow of Gases Through Con36. Cornell, D. and Katz. D.L.: solidated Porous Media. hi. cd Etzgr. Chew (Oct. 1953) 45. The Generalized Kozeny 37. Wyllie, M.R.J. and Gardner, G.H.F.: Carman Equation. World Oil (March and April 1958). Rewtiwty of Brme-Saturated Sands in 38. Winsauer. W.O. PI ol.: Relation to Pore Geometry. Bull.. AAPG (Feb. 1952) 253-77. Electrical Resistivity Measurements on Reservoir 39. Rust. C.F.: Rock Samples by the Two-Electrode and Four-Electrode Methods, Trms., AIME (1952) 195, 2 17-24. Porosity Index in Limestone from Electrical 40. Tlxier. M.P.: Logs-Part I , Odrmci Cm .I. (Nov. 15. 1951) 140. Formation Factors of Un41. Wyllie, M.R.J. and Gregory. A.R.: consolidated Porous Media: Influence of Particle Shape and Effect of Cementation, Truns.. AIME (1953) 198, 103-09. Saturation and Porosity from Electrical Logs in 42. dewitte. A.J.: Shaly Sands-Part I, Oil and GNS J. (March 4. 1957) 89. Effect of Clay and Water Salinity 43. Hill, H.J. and Mdhum. J.D.: on Electrochemical Behavior of Reservoir Rocks. Tmm., AIME (1956) 207. 65-72.
ii.fzj~~pj, In rclr,,
References
I. Fraser, H.J. and Graton, L.C : Systematic Packing of Sphere-With Particular Relation to Porosity and Permeability, J. G4. (Nov.-Dec. 1935) 785-909. 2. Tickell. F.G., Mechem. O.E. and McCurdy, R.C.: Some Studies on the Porosity and Permeability of Rocks, Trmv.. AIME (1933) 103. 250-60. 3 Core Laboratories Inc.. Dallas, TX. 4: Nutting, P.G.: Physical Analysis of Oil Sands. Buli., AAPG, (1930) 14. 1337-49. 5. Russell, W.L.: A Quick Method for Determining Porosity, B!i//ril.. AAPG (1926) 10, 931-38. 6. Stevens, A.B.: A Luhoruror~ Munuulfor Perrdrum Eqinrerirzg 308. Texas A&M U., College Station, TX (1954). 7. Washburn, E.W. and Bunting. E.N.: Determination of Porosity by the Method of Gas Expansion. J. Ant. Ccram. SW, 5. 48. 8. Beeson. C.M.: The Kobe Porosimeter and Oilwell Research Porosimeter. Trms., AlME (1950) 189. 313-18. 9. Dotson. B.J. et ni.: Pomsity Measurement Comparison by Five Laboratories. . Trcins., AIME (1951) 192. 341-46. Analysis of Fractured Limestone Cores, Truns.. IO. Kelton. F.C.: AIME (1950) 189. 225-34. Il. Krumbein. W.C. and Sloss. L.L.: .Srrcui~rciph~ and .Scdiww~turim, Appleton-CenturyCrofts Inc., New York City (1951) 218. I?. Geertsma. J.: Effect of Fluid Pressure Decline on Volumetric Changes of Porous Rocks, Truns.. AIME (19.57) 210. 33 1 and 339. 13 Hall. H.N.: Compressibility of Reservoir Rocks. Trmn.c., AIME (1953) 198. 309. 14. Fatt. I.: Pore Volume Compressibilities of Sandstone Reservoir Rocks, Trun;. , AIME (1958) 213. 362-64 15. Hammerlindl, D.J.: Predicting Gas Reservoirs in Abnormally Pressured Reservoirs. paper SPE 3479 presented at the 1971 SPE Annual Meeting, New Orleans. Oct. 3-6.
Chapter 27
Introduction
The early-day analysis of cores was largely an art, a qualitative matter of odors and tastes, sucking on the rock, and visual examination. The science of core analysis has evolved from such early beginnings, using developments in instrumental methods of chemical and physical analyses as they became available. Electron microscopy, mass spectrometry, gas chromatography, high-frequency phase analysis, acoustic wave train analysis, and nuclear magnetic relaxation analysis are among the tools being used in the more sophisticated core testing today. Many other techniques are available now to assist the geologist and petroleum engineer in the completion of wells and the evaluation and operation of oil and gas reservoirs, but core analysis still remains the basic tool for obtaining reliable information on the rock material penetrated. Study of representative core samples of an oilor gas-bearing formation provides the only means for direct measurement of many important properties of the formation. The minimum basic measurements made on cores generally comprise determination of porosity at no confining pressure, permeability at low confining pressure, and residual fluid saturations. Various supplementary routine tests such as chloride, oil gravity, directional permeability, grain density, and grain size frequently are made as an aid in interpretation and evaluation. These data are the subject of this chapter. tent, by saturation with a liquid, or by calculation from Boyle s law upon compression or expansion of gas in the pore spaces of the sample. The other widely used method involves the separate determination of the gas, oil, and water contents of the sample, and the summation of these three values to obtain PV. Most of the porosity data reported in the tables here were determined by the summation-of-fluids method. Comparison of porosity values obtained on samples from several thousand feet of core where measurements were made by both the summation-of-j&ids method and by a Boyle s law method showed agreements, in general, of 0.1 to 0.5% porosity. Extensive checks of porosity values by resaturation with brine have shown values slightly lower than by the other procedures, indicating approximately 98 to 99% resaturation.
Permeability
The permeability of a formation sample is a measure of its ability to transmit fluid. The permeability determination involves measurement of the rate of flow of a fluid of known viscosity through a shaped sample under a measured pressure differential. Air is the fluid normally used because of its convenience, availability, and relative inertness toward the core material. For many years, airpermeability measurements were corrected to an equivalent liquid permeability by use of the well-known Klinkenberg corrections. The permeability values reported in Tables 27.1 through 27.11 have been corrected to the equivalent liquid-permeability values, except as noted in the next paragraph. In the whole-core orfill-diameter core analysis procedures, permeability is frequently measured in two horizontal directions. One measurement is made in the direction of the major fracture planes and is reported as k. This value indicates the effectiveness of the fractures as flow channels. The core sample is then rotated 90 and the second measurement is made in a direction of flow perpendicular to the direction of the first measurement. This
(continued on page 9)
Porosity
Porosity is a measure of the void space or storage capacity of a reservoir material. Normally it is expressed as a percentage of bulk volume (%BV). Porosity may be determined by measurement of any two of the three quantities-grain volume, void volume, and bulk volume. Various generally acceptable methods and techniques for determining porosity are used by different laboratories. The void volume may be determined on a previously cleaned and dried sample by extraction or gas or air con-
27-2
PETROLEUM ENGINEERING
HANDBOOK
Formation Blossom Cotton Valley Glen Rose Graves Meakin Nacatoch Paluxy Pettit Rodessa+ Smackover* Tokio Travis Peak Tuscaloosa
Fluid Production C/O Cl0 0 Cl0 G/:0 Cl0 0 0 Gl:lO c/o c/o Cl0
l
Range of Production Deoth 2,190 5,530 2,470 2.400 3: 145 2,270 1,610 2,850 4.010 5:990 6.340 2,324 2,695 3,020 to to to to 2,655 8,020 3,835 2.725 31245 2,605 2,392 4,690 5.855 6;120 9,330 2,955 5,185 3,140
Average Production Deoth 2,422 6,774 3,052 2.564 31195 2,485 2,000 3,868 4.933 61050 8,260 2,640 3,275 3,080
Range of Permeability (md) 1.6 0.6 1.6 1.2 6.5 3.0 0.7 5 0.1 0.1 0.1 0.5 0.4 0.4 to to to to to to to 8,900 4,620 5,550 4,645 51730 6,525 6,930 13,700 698 980 12,600 11,500 6,040 3,760
How
to
to to to to to to to to to
to
to to to
to
to to
Indicates fluid Droduced: G = aas: C = condensate. 0 = oil Specific zone not identified k&ally Includes data from Mitchell and Glcyd zones. Includes data from Smackover Lime and Reynolds zones
TABLE 27.2-EAST
TEXAS AREA
Formation Bacon Cotton Vallev Fredericksburg Gloyd Henderson Hill Mitchell Mooringsport Nacatoch Paluxy Pecan Gap Pettit * Rodessa Sub-Clarksvillet Travis Peak* Wolfe Citv Woodbine Young
Fluid Production c/o C 0 Cl0 GICIO Cl0 0 0 0 : G/C/O Cl0 0 Cl0 C410 C
Range of Production Depth ffB 6,665 8,448 2,330 4,612 5,976 4,799 5,941 3,742 479 4,159 1,233 5.967 4,790 3,940 5,909 981 2,753 5,446 to to to to to to to to to to to to to to to to to to 7,961 8,647 2,374 6,971 6,082 7,666 6,095 3,859 1,091 7,867 1,636 8.379 81756 5,844 8,292 2,054 5,993 7,075
Average Production Depth (W 7,138 8,458 2,356 5,897 6.020 5;928 6,010 3,801 743 5.413 1:434 7,173 6,765 4.892 6,551 1,517 4,373 6,261
Range of Permeability fmd) 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.4 1.9 0.1 0.5 0.1 0.1 0.1 0.1 0.3 0.1 0.1 to to to to to to to to to to to to to to to to to to 2,040 352 4.6 560 490 467 487 55 4,270 9,600 55 3,670 1,180 9,460 180 470 13,840 610
Small amount of Navarro data combined with Nacatoch Data for Pinsburg, Potter, and upper Pettit combined wlfh Peltil Small amount of Eagleford data combined with subClarksvW *Data for Page cambmed with Travis Peak.
27-3
Average Permeability (md) 1,685 333 732 1.380 1;975 1,150 142 1,213 61 135 850 2,100 460 506
Range of Porosity w 15.3 to 40 11.3 to 34 17.3 to 38 9.8 to 40 14.4 to 41 17.1 to 40 9.9 to 41 15.1 to 32 6.2 to 28 5.1 to 28 1.1 to 34 13.6 to 42 9.4 to 36 15.6 to 39
Average Porosity w 32.4 20.3 23.4 34.9 30.9 31.8 30.5 26.9 15.4 16.5 14.2 32.1 24.3 27.3
Range of Oil Saturation w 1.2 to 36 0.9 to 37 4.0 to 52 0.3 to 29 2.6 to 56 0.6 to 43 0.2 to 52 7.5 to 49 9.1 to 29 0.7 to 26 0.7 to 41 0.9 to 57 0.5 to 36 0.3 to 53
Average Oil Saturation (W 20.1 13.1 21 .o 16.8 19.9 12.9 4.9 21.2 12.7 14.8 12.8 25.6 14.3 14.0
TABLE 27.2-EAST
TEXAS AREA (continued) Range of Calculated InterstitialWater Saturation toa 9 to 13 to 35 to 16 to 21 lo 23 to 15 to 29 to 24 to 22 to 30 to 10 to 6 to 12 to 17to 23 to 14 to 13 to 22 32 43 45 44 47 47 48 55 47 56 35 42 60 38 68 65 27 Average Calculated InterstitialWater Saturation to4 16 25 41 31 27 33 29 40 41 30 46 23 23 33 28 46 35 21
Average Permeability 0-W 113 39 1.2 21 19 70 33 5 467 732 6 65 51 599 42 32 1,185 112
Range of Porosity w 1.5 to 24.3 6.9 to 17.7 11.9 to 32.6 8.0 to 24.0 7.0 to 26.2 6.4 to 32.2 7.2 to 29.0 5.3 to 19.6 13.4 to 40.9 6.3 to 31 .l 16.3 to 38.1 4.5 to 25.8 2.3 to 29.0 6.2 to 38.0 5.6 to 25.8 17.1 t0 38.4 9.7 t0 38.2 4.4 to 29.8
Average Porosity Pi8 15.2 11.7 23.1 14.9 15.2 15.6 15.5 14.6 27.1 21.6 26.8 14.7 14.5 24.8 15.0 27.9 25.5 19.7
Range of Oil Saturation W) 2.7 to 20.6 1.1 to 11.6 3.3 to 39.0 trace to 24.3 0.8 to 23.3 0.9 to 26.7 1.8 to 25.9 2.8 to 26.6 0.6 to 37.4 2.2 to 48.7 3.5 to 49.8 0.9 to 31.6 trace to 25.3 1.4 to 34.6 0.1 to 42.8 1.5 to 37.4 0.7 to 35.7 trace to 4.5
Average Oil Saturation Wol 8.6 2.5 20.8 8.2 10.6 12.2 12.5 13.8 14.5 24.1 12.9 9.8 5.3 17.9 12.5 15.6 14.5 0.8
27-4
PETROLEUM ENGINEERING
HANDBOOK
TABLE 27.3-NORTH
LOUISIANA
AREA
Formation Annona Chalk Buckrange Cotton Valley a Eagleford Fredericksburg Haynesville Hosston Nacatoch Paluxy PettitC Pine Island d Rodessae Schuler Sligog Smackover Travis Peakh Tuscaloosa
Fluid Production 0 c/o GlClO C G/C C Cl0 0 Cl0 c/o 0 G/C/O GICIO Cl0 Cl0 c/o GICIO
Range of Production Depth m 1,362 to 1,594 1,908 to 2,877 3,850 to 9,450 8,376 to 8,417 6,610 to 9,880 10,380 to 10,530 5,420 to 7,565 1,223 to 2,176 2,195 to 3,240 3,995 to 7,070 4,960 to 5,060 3,625 to 5,650 5,500 to 9,190 2,685 to 5,400 9,960 to 10,790 5,890 to 7,900 2,645 to 9,680
Average Production Depth (fu 1,480 2,393 7,450 8,397 8,220 10,420 6,480 1,700 2,717 5,690 5,010 4,860 8,450 4,500 10,360 6,895 5,184
Range of Permeability W) 0.1 to 2 5 0.1 to 2,430 0.1 to 7,350 3.5 to 3,040 1.6 to 163 0.1 to 235 0.4 to 1,500 27 to 5,900 0.2 to 3,060 0.1 to 587 0.2 to 1,100 0.1 to 2,190 0.1 to 3,180 0.1 to 1,810 0.1 to 6,190 0.1 to 2,920 0.1 to 5,750
iDat. reported where member formatlon of Cotton Valley group not readlfy tdentlflable Data reported as Eutaw in come areas Includes data reported es Pettlt. Upper Petilt, and Mid-Pettit. eometlmes considered the same as Sllgo Sometimes referred to as Woodruff. ; Includes data reported localy for Jeter, HIII. Kllpatrlck, and Fowler zonee includes data reported focally for Bodcaw, Vaughn, Doris, McFerrin, and Justiss zones. ; Includes data reported as BIrdsong-Owens Frequently considered the same as Hosston
TABLE 27.4-CALIFORNIA
Area San Joaquin Valley a Los Angeles Basin and Coastal b San Joaquin ValleyC Los Angeles Basin and Coastal d San Joaquin VaIleye Los Angeles Basin and Coastal San Joaquin Valleyg Coastal h San Joaquin Valley Los Angeles Basin and Coastal
Range of Production Fluid Depth Production (fi) 6,820 to 8,263 0 0 2,870 to 9,530
60 to 450
165
Miocene, upper
4,210 4,100
10 to 1,200 5 to 1,040
245 130
4 to 7,500 86 to 5.000
1,000 1.110
Miocene, lower
0 0
5,300 4,430
30 to 154 20 to 380 5 to 80 -
76 134
700 842
Oligocene
0 0 0 0
33 100
Pliocene
aMainiy dala from Gatchell zone %zludes Uooe, and Lowe, Terminal. Umon Paclflc Ford. 237. and Sesnon zonee clnctudes Kernco, Repubhc, and 26R zo ee. d Includes Jones and Maw zO eS ; Includes JV. Obese. and Phamdes zones Manly data from Vaqueros zone z Manly data from Oceanic zone Mmly data fro Sespe zone lnciudes Sub Mullma and Sub Scaler No 1 and No 2 zones : Includes Ranger and Tar zone?, O&based data show high oil ?.at ,at,o [average 61%) and low water ( 3 to 54%. average 15%) O,l-based data show range 27 6 to 52 4 and average of 42.3% not Included I above 011Saturation Values
27-5
TABLE 27.3-NORTH
LOUISIANA
AREA (continued) Range of Calculated InterstitialWater Saturation Average Calculated InterstitialWater Saturation to4 37 35 24 36 :A 28 47 35 29 22 30 25 31 25 31 43
Average Permeability
Range of Porosity
Average Porosity
b-4
0.7 305 135 595 z; 140 447 490 26 285 265 104 156 220 357 706
14.3 13.4 3.5 12.8 12.8 5.5 8.8 25.8 9.6 4.5 8.5 5.1 3.6 7.3 3.4 7.0 10.7
(O/o)
to to to to to to to to to to to to to to to to to 36.4 41 34 28 23.1 23.1 29 40 39 27 27 34 27.4 35 23 27 36
(O/o) 26.8 31.4 13.1 22.9 19.9 13.4 18.6 31.4 27.2 14.3 20.6 19.1 15.0 21.1 12.9 19.4 27.6
W)
6.0 0.7 0.0 1.6 1.7 1.1 0.0 2.5 0.1 0.1 13.3 0.0 0.0 0.6 1.1 0.1 0.0 to to to lo to to to to to to to to to to to to to 40 51 14 28 4.3 14.5 35 33 48 59 37 31 24 27 22 35 37
(ON
22.0 22.6 3.1 4.3 2.7 5.1 8.6 19.5 11 .a 15.6 24.1 2.9 4.8 9.8 7.2 8.6 8.5
to4
24 to 40 29 to 47 11 to40 35 to 49 31 to 41 18 to 37 45 IO 54 23 lo 55 10 to 43 16to30 21 to 38 8 to 51 12 to 47 9 to 47 26 to 38 31 to 61
TABLE 27.4-CALIFORNIA
(continued) Range of Calculated InterstitialWater Saturation W) 15 to 49 15 to 72 Average Calculated InterstitialWater Saturation co4 35 36
17 to 40 19.5 to 39
28.2 30.8
9 to 72 10 to 55
32k 25
20 to 6Bk 22 to 72
5ok 44
12 to 62 12 to 61
30 30
13 to 34 11 to 33
23 21
20 to 38 21 to 29
26.4 24.3
4 to 40 13 to 20
19 15.8
25 to 80 32 to 67
51 53
14to67 27 to 60
36 37
15 to 40 34 to 36
34 35
19 to 34 15to22 30 to 36 24 to 41
12 to 40 6to 17 7 to 43 15 to 80
22 11.8 24.1 45
2 to 60 19 to 56 33 to 84 19 to 54
43 46 54 38
3 to 45 15 to 52 10 to 61 10 to 40
30 42 34 21
37 to 38 18 to 44 12 to 23
38 25 24 15
27-6
PETROLEUM ENGINEERING
HANDBOOK
TABLE 27.5-TEXAS
GULF COAST-CORPUS
CHRISTI AREA
(ft)
to to to to to to to to
Range of Permeability (md) 45 to 2,500 5 to 9,000 5 to 2,900 7 to 300 25 4 to 2,900 1,800 1 to 380 6 to 1,900
Dist. 4 Jim Wells. San Patricia, Webb, Brooks, Nueces, Jim Hogg. Hidalgo, W~llacy,
TABLE 27.6-TEXAS
GULF COAST-HOUSTON
AREA
Range of Production Depth 4,000 4,600 7,100 4,700 2,900 2,400 7,400 6,900 5,800 2,300 4,100 4,400 3,700 to to to to to to to to to to to to to 11,500 11,200 8,300 6,000 6,000 8,500 8,500 8,200 11,500 10,200 4,400 8,700 9,700
(fb
8,400 7,800 7,800 5,400 4,000 3,700 8,100 7,400 9,100 7,900 4,300 6,800 6,600
(W
uu
Range of Permeability 18 33 308 355 124 71 50 190 3.0 9.0 14 24 23 to to to to lo to to to to to to to lo 9,200 9,900 3,870 1,210 13,100 7,660 105 1,510 1,880 2,460 680 5,040 4,890
Average Permeability
(md)
810 1.100 2;340 490 2,970 2,140 86 626 96 195 366 750 903
Woodbine Yegua
2 to 2 to 4 to 4 to 3 to 2to 1 to 3 to 5 to 3 to 6 to 3 to 2 to
50 34 28 10 a 16 6 18 94 29 13 63 59
12.3 10.4 17.5 5.7 5.5 7.2 2.0 9.3 19.1 10.0 8.2 11.0 8.5
TABLE 27.7-LOUISIANA
GULF COAST
Range of Production Depth (fi) 5,208 to 14,900 2,700 to 12,700 7,300 to 14,600 6,700 to 12,000 17,533 to 18,906
27-7
TABLE 27.5-TEXAS
GULF COAST-CORPUS
Average Porosity
Range of Oil
w 30
27 27 24 28 24 19 29
W)
1 to 2 to 3 to 1 to 9 to 1 to 0 to 4 to 30 38 32 4 30 17 IO 40
fs5
58 32
TABLE 27.6-TEXAS
GULF COAST-HOUSTON
AREA (continued) Range of Calculated InterstitialWater Saturation w 20 to 63 12 to 61 14 to 31 25 to 47 23 to 53 21 to 55 53 to 61 26 to 36 20 to 54 20 to 50 24 to 59 17 to 59 17 to 53 Average Calculated InterstitialWater Saturation (%I 34 33 21 36 3 : 56 35 30 iz 33 34
Range of Porosity
Average Porosity
(Oh)
18.3 21.8 35.0 20.5 28.6 23.5 26.5 29.5 14.5 16.2 23.5 23.4 22.9 to to to to to to to to to to to to to 38.4 37.1 37.0 37.3 37.6 38.1 31.0 31.8 27.4 34.0 26.7 37.8 30.5
28.6 29.8 35.9 32.6 33.2 35.2 27.1 30.4 19.6 21.9 25.5 30.7 31.6
Range of Oil Saturation VW 0.1 to 6.0 4.6 to 41.2 0.2 to 0.8 8.1 to 21.8 0.2 to 1.5 11 .o to 29.0 0.0 to 1.5 14.4 to 20.3 0.2 to 10.0 4.6 to 20.5 10.7 to 27.4 0.1 to 15.5 3.5 to 21.8
Average Oil Saturation (04 1.o 13.5 0.5 15.3 0.5 16.6 0.2 15.3 1.5 9.7 20.1 1.2 11.4
TABLE 27.7-LOUISIANA
GULF COAST (continued) Range of Cal&ated InterstitialWater Saturation rw 20 to 74 18to50 19 to 57 23 to 60 55 Average Calculated InterstitialWater Saturation ( w 35 32 32 35 -
Range of Porosity Pw 15.7 to 37.6 18.3 to 39.0 16.7 to 37.6 22.1 to 36.2 5 to 29
Range of oil Saturation (Oh) 0.1 6.5 0.5 5.2 26 to to to to to 4.7 26.9 8.9 20.0 44
27-8
PETROLEUM ENGINEERING
HANDl3OOK
TABLE 27.8-COMPARATIVE
DATA-SIDEWALL (S.W.) VS. CONVENTIONAL LOUISIANA GULF COAST AREAS Average Depth (ft) 8,945 9,037 7.174 8.622 4:902 6,789 5,456 6,399 8.148 $826 8,276 8,415 7,240 7,693 7,369 7,099 3,861 4,194 2,824 3,625 10,664 11,500 8,996 10,171 4,286 4,040 4,504 4,383 Average Permeability W) 62 813 317 1,895 238 1,496 681 641 75 235 176 791 147 277 302 603 119 558 634 576 312 748 327 1,300 180 578 346 867
(CONV.) ANALYSIS,
TEXAS AND
Formation Frio
Area Houston
Fluid Production C 0
Type Analysis SW. Cow. SW. Conv. S.W. Conv. S.W. Conv. SW. Conv. S.W. Conv. S.W. Conv. SW. Conv. SW. Conv. SW. Conv. S.W. Conv. S.W. Conv. SW. Conv. S.W. Conv.
Average Porosity to4 27.5 26.7 30.8 27.7 27.2 26.5 29.5 28.5 27.3 26.8 27.1 28.7 27.9 29.7 29.9 31.6 26.8 31 .a 33.3 31 .a 28.2 27.4 28.2 26.6 28.5 29.0 30.4 29.8
Average Oil Saturation (% pore space) 0.7 0.7 14.6 14.6 0.8 1.1 19.5 16.3 4.2 1.9 10.0 7.9 0.2 0.7 10.5 11.7 3.2 1.7 20.9 19.9 2.5 2.1 10.1 14.8 0.5 0.7 17.7 20.0
Corpus Christi
C 0
Louisiana
C 0
Houston
C 0
Corpus Christi
C 0
Louisiana
C 0
Corpus Christi
C 0
second value is normally reported as kw , and it is usually representative of the matrix permeability. Values for kw are reported in the following tables for formations that are normally subjected to the whole-core or full-diameter core analysis procedures. These values are not corrected to equivalent liquid-permeability values.
Liquid Saturations
In the coring process, the core is exposed to the drilling fluid at a pressure greater than formation pressure. If the core contains oil or gas, some portion of this is flushed out and replaced by the drilling-fluid filtrate. As the core is brought to the surface and the external pressure is reduced, the expansion of free gas or dissolved gas expels both oil and water from the core. As a result, the pore spaces of the cores recovered at the surface contain free gas, water, and oil if oil is present in situ. The oil and water contents normally are called residual liquids. The residual oil and water contents of core samples normally are determined by retorting, vacuum distillation, or solvent extraction and distillation. The oil and water contents are converted to oil and water saturations as percentages of PV. The oil and water saturation values reported in these tables represent data obtained by the retorting or the vacuum distillation procedures. The water content of the core as recovered is generally called tofal wafer, and it may include some drillingfluid filtrate or invasion water. The water saturation actually existing at a given interval in a reservoir may be spoken of as the connate water or interstitial water. This interstitial-water saturation value, as reported in the ta-
bles, was determined in some cases by an empirical correlation factor applied to the total water value and in some cases by the use of capillary-pressure data for the specific reservoirs. The API oil gravity values reported normally were measured on the oil recovered in the retorting or vacuumdistillation procedures. Comparison of gravity values obtained in oil recovered from cores with values obtained on produced or drillstem test (DST) oil indicates general agreement to within f2 API. The liquid saturation data presented in the tables are from formations interpreted to be hydrocarbon-productive to some degree. In some cases, it was feasible to make a distinction between gas-, condensate-, and oil-productive zone characteristics. Table 27.9 shows core analysis data for zones identified as transition zones. These represent intervals or zones where an appreciable water cut is encountered during the life of a field. Such transition zones are present in many other areas and fields, but the available data did not permit a similar breakdown. It should be pointed out that the relative average depths reported for the gas-condensate, oil, and transition zones do not contradict the basic premises that gas overlies oil and that oil overlies water. The condensate-producing zones in the major formations in the U.S. gulf coast area, as presented in Tables 27.6 and 27.7, frequently are found at greater depths than are the oil-producing zones of the same formations. In a similar manner, the gas, oil, and transition zones shown in Table 27.9 for the extensive geologic groups and formations in the Oklahoma-Kansas area are found at different subsurface depths in different parts of the area.
27-9
names were selected in an effort to represent generally recognized nomenclature over large areas rather than local terminology. Some important producing formations are not included because of the lack of sufficient data at this time or because of their proprietary nature.
Reference
I. European Continenfd ShelfGuide, Oilfield Publications Ltd., Ledbury, Herefordshire, England (1982).
27.10
PETROLEUM
ENGINEERING
HANDBOOK
TABLE
27.9-OKLAHOMA-KANSAS
AREA*
FormatIon
Arbuckle
FluId ProductIon G To G 0 T : T ci 0 : T
AtokaC
Bartlesville
Range of ProductIon Depth IfU 2,700 to 5,900 500 to 6,900 600 to 11.600 3,700 to 3,800 500 to 4,500 300 to 3,700 700 to 7,400 200 to 5,700 500 to 2,600 4,800 to 5,100 3.700 to 7,800 2,600 lo 3,200 1,000 lo 3.800 2,700 lo 3,300 300 6,800 3,700 6,000 6,900 4,500 4,400 1,300 2,800 4,200 4,700 4,800 2,200 300 1,900 4,300 600 2,200 1,800 1,900 3,800 1,000 2,900 1,800 2,500 1,900 to to to to lo to to to to to to lo to to to to to to to to
IO
Burgess First Brom!ded Second Bromide Burbank Chester 2 G 0 T G 0 T 0 T G 0 L 0 a 0 E T Hoxbar : T 0 T 0 a 0 T 0 G 0 T E L 0 T 2,800 7,600 13,800 13,200 16,200 11,200 13,300 4,500 3,700 6,700 6,700 6,100 5,700 6,400 3,900 11,800 10,000 6,800 2.100 2,000 8,800 10,300 3,000 9,600 8,700 5,800
Cleveland
Deese 9 Hoover
Hut-don
Lansing Layton
Marmaton Misner
to to to to to 700 to 6,100 500 to 6,300 1,800 to 5,700 4,300 to 4,600 8,100 2,600 to 6,500 4,900 to 6,200 1,800 to 5,100 800 to 5,200 1,200 to 5,200 900 to 8,800 600 to 6.600 400 to 7,200 3,600 to 17,000 1,600 to 11,200
Average Production Depth (fU 4,500 3,500 3,600 3,700 2,600 2,100 2,600 1,500 1,200 5,000 6,500 2,900 2,600 3,000 1,600 1,800 7,200 8,600 11,500 12,800 9,000 9,700 2,800 3,000 5,700 5,700 5,700 3,500 3,200 3,100 6,500 5,200 4,000 2,000 2,000 6,300 4,200 3,000 4,600 4,900 3,800 3,300 3,900 2,900 3,200 4,400 8,100 4,300 6,000 4,000 3,100 3,900 4,600 4,100 4,000 10,100 8.100
Range oi ProductIon Thickness (ft) 5.0 to 37 1.O to 65.5 2.0 to 33 1.O to 9.0 3.0 to 16 2.0 to to 1.5 to 42 1 .O to 72 4 to 40 4 to 48 2.3 to 50 5 to 8 2 lo 26.5 4 to 5 2.5 lo 9 3.0 10 19.5 2.0 lo 82 15 to 161.3 20 to 53.6 3.0 to 69 5 to 44.5 3 to 48 31019 2 to 45 2 to 23 4 to 20.5 2to17 1 to 70 3 to 22 5 to 55 2 to 60.3 4 to 49 3 to 37 2to17 9to 11 2 to 63 3to13 2 to 77.3 2 to 73 3 to 16.2 410 18 1 to 57 3 to 15.5 1.5 to 7.5 3to 14 2 to 56.5 8 to 21 2 to 34.4 2 to 48.1 1 to 43 3 to 27.1 1.5 to 95.3 4 to 70.1 14to58 3 to 42
Average ProductIon Thickness m 18.3 11.8 14.3 4.0 7.8 6.5 11.4 14.0 14.5 19.0 12.5 6.5 8.8 4.5 20 5.8 11.3 18.7 65.1 37.9 16.2 18.4 17.3 9.1 10.9 8.6 10.0 9.0 13.4 7.7 19.3 11.7 16.6 11.9 8.4 10.0 14.4 9.3 14.0 14.7 6.5 22.0 9.3 10.3 7.4 4.7 8.5 10.6 15.8 16.1 12.2 10.9 13.3 12.0 17.4 35.3 12.2
Range of Permeablllty 3.2 0.2 0.1 1.3 0.3 9 0.2 0.2 0.1 0.1 0.3 1.4 0.3 3.1 0.2 0.6 0.1 0.9 3.4 2.0 0.8 0.1 0.1 0.1 0.1 0.1 2.5 0.1 0.1 7.8 0.4 1.9 1.3 55 6.4 0.1 0.5 0.1 0.1 0.3 0.2 0.3 1.1 24 37 0.1 0.1 0.4 0.1 0.2 0.1 0.1 0.1 12 0.7 (mdi to 544 to 1,530 to 354 to 609 to 920 to 166 to 36 to 537 to 83 to 43 to 664 10 6.6 to 160 to 13 10 104 lo 62 to 2,280 to 40 to 72 to 585 to 42 to 226 to 4.8 to 269 to 61 to 13 to 338 to 135 to 112 to 232 to 694 to 200 to 974 to 766 to 61 to 1,620 to 31 to 678 to 48 to 390 to 210 to 280 to 143 to 105 10 171 to 803 to 120 to 516 to 361 to 229 to 129 to 1.210 to 135 to 98 to 157
Average Permeablllty (md) 131 140 57 174 144 67.3 10.4 32.7 18.2 24.4 36.0 4.0 19.3 8.0 142 19 31.3 175 18.3 21.4 118 12.9 8.64 1.53 33.0 9.11 2.38 50.6 15.4 12.9 94.1 62.8 61.8 288 372 33.7 277 14.4 34.5 5.3 101 14 26.3 54.1 23.8 46.4 104 89.7 41.8 33.5 21.9 21.3 22.2 43.5 7.5 48.0 39.0
Range of Permeablllty k,o Wi 0.1 to 1,270 0.1 to 135 0.6 to 2.8 5.5 1.5 0.07 0.1 to 2.2 22 0.4 0.2 to 7.4 1.40 0.3 to 0.9 to to to to
OYIO 1.10 to to to to -
0 0.1 0.3
0.5
162
0.20 0 to 2.1 0.2 to 74 0 to 216 0 to 163 0.1 to 89 0.1 to 185 0.1 to 36 6.2 to 8.8
TYPICAL
CORE
ANALYSIS
OF DIFFERENT
FORMATIONS
27-l 1
TABLE 27.9-OKLAHOMA-KANSAS
AREA (continued)
Range of Calculated InterstitialWater Saturation PM 28 to 62 20 to 79 37 to 91 32 to 65 19 to 61 40 23 to 66 17to72 43 to 67 26 to 62 15to59 50 15to59 44 19to58 36to 72 12to87 28 to 45 8 to 44 40 31 to 73 45to 81 19to81 17to81 40 to 89 30 to 64 IO to 74 32 to 77 19 to 49 13to57 19to76 14to47 31 to 42 34to 39 t3to68 17to93 16to89 28 to 69 34 to 83 23 to 76 31 to 69 42 to 66 18 to 22 14 to 87 20to 51 60 to 93 27 to 95 43to 85 22 to 93 16 to 85 32 to 94 19 to 77 14 to 52 Average Calculated InterstitialWater Saturation w 40 47 52 45 37 40 48 40 54 40 32 50 37 44 35 40 54 34 32 25 43 51 43 33 61 42 44 49 37 33 Liz 35 ii 46 48 49 47 41 43 53 20 38 32 77 58 63 53 46 61 44 31
Average Permeability, $n?, 67.8 21.6 1.7 5.5 1.5 0.07 0.45 22 0.40 2.23 1.40 0.60 1.67 0.21 1.18 1.65 oio 1.10 5.24 2.04 52.3 6.7 23.3 0.20 0.62 13.9 13.7 14.2 13.2 9.44 4.23 -
Range of Porosity w 9.0 10 20.9 2.1 lo 24.3 3.7 to 23.1 8.5 to 17.3 5.9 to 28.6 11.9 to 18.6 8.4 to 21.1 8.5 to 25.8 8.5 to 20.1 3.8 to 19.8 1.2 to 19.3 11.9 to 14.8 8.3 to 21.4 16.9 to 18.1 8.1 to 22.8 1.5 to 6.5 1.4 to 15.7 1.5 to 10.9 3.5 to 14.5 5.6 to 11.7 5.6 to 11.4 6.4 to 21.6 7.1 to 17.0 2.6 to 20.7 2.3 to 16.0 3.2 to 17.8 9.8 to 23.5 7.4 to 24.6 11.0 to 20.4 9.8 to 22.6 4.7 to 26.4 11.7to23.4 12.7 to 24.1 16.7 to 22.5 13.9 to 18.2 3.1 to 29.7 14.3 1.6 1.1 8.4 lo 22.7 to 33.6 to 19.5 to 16.0 5.1 to 25.9 4.6 lo 27.2 14.2 lo 21.3 1 8 lo 21.4 11.0 to 12.1 2.1 lo 20.9 1.9 lo 11.3 6 5 10 37.8 5.7 to 39.3 1 5 lo 38.0 1.5 to 23.6 1.3 lo 34.1 1 1 to 26 1 2.8 to 9 6 5.5 to 16.5
Average Porosity VW 14.4 12.0 9.2 12.9 14.5 14.9 15.8 17.6 14.6 12.2 7.2 13.4 15.6 17.5 14.2 13.2 4.0 9.8 6.5 6.8 9.3 7.4 15.7 13.7 12.2 10.1 7.7 16.9 15.2 15.6 16.7 17.4 16.3 19.7 20.5 16.1 16.5 18.5 10.9 7.3 12.2 7.2 14.5 17.8 17.1 140 11.6 11.9 8.1 21 .o 22.3 18.7 10.3 13.4 9.3 6.7 11.0
Range of Oil Saturation PM 0.7 lo 9.4 5.2 to 42.3 0 to 23.6 0 to 8.1 5.1 to 35.1 5.8 to 21 .l 0 to 11.1 3.3 to 60.6 0.9 lo 35.7 0 lo 6.7 3.3 lo 25.8 4.6 IO 8.8 4.8 to 49.7 7.4 lo 7.8 16.2 lo 33 0 to 7.6 3.1 IO 24 0.4 to 6.8 0 lo 6.9 2.4 IO 24.2 0 Io 13.6 9.3 lo 26.6 2.0 to 15.7 0 lo 7.5 7.2 lo 35.9 0 IO 11.1 0 lo 7.1 5.8 lo 35.5 0 IO 21.1 2.2 lo 6.3 5.9 lo 46.4 0 to 7.0 126to231 6 6 to 17.1 0 7 to 4.4 3.2 to 48.7 3.3 lo 11.4 1.6 to 34.5 0 to 61.1 6.5 to 28.9 Oto78 1.6 to 37.3 0 to 14.3 6.4 to 18.1 2.1 to 2.3 4.1 to 41.6 0 to 8.2 0 to 6.8 1.4 to 30.0 1.1 to 18.3 0 to 9.3 2.1 to 56.5 0 to 41.2 4.0 to 14.7 5 1 to 27.7
Average Oil Saturation w 37 17.1 7.1 2.0 20.7 12.1 47 16.2 12.2 4.3 15.0 8.7 21.5 7.6 8.3 21.5 3.8 11 2.2 4.0 11.5 4.8 15.3 11.2 1.1 19.1 1.2 4.1 13.1 7.8 3.8 20.4 0.8 160 14.5 2.6 21.4 6.8 15.3 10.6 18.1 12.8 2.4 15.3 6.9 11.7 2.2 14.8 4.7 2.4 12.9 7.6 2.8 15.0 6.9 7.8 132
Range of Total Water Saturation w 34.5 to 62.7 20.6 to 79.3 37.2 to 91.9 36.4 to 65.2 18.4 to 61.5 42.7 to 55.4 23.4 to 70.0 17.4 to 85.2 43.9 to 88.0 32.9 to 82.4 14.6 to 58.5 50.0 to 51.3 15.3 to 60.0 47.3 to 55.2 19.3 35.7 12.8 29.5 28.2 8.9 21 .l 31.5 45.7 20.9 17.7 40.9 40.0 10.2 32.9 19.1 14.0 41 .l 14.6 34.8 40.1 13.6 50.5 16.7 160 37.4 to 65.4 to 71.8 to 67.2 to 78.6 to 45.7 to 44.9 to 57.6 to 73.4 to 80.7 to 80.7 to 80.8 to 89.2 to 64.4 to 74.0 to 77.2 to 54.9 to 58.6 to 77.1 to 48.5 to 50.7 10 40.6 to 68.5 lo 69.8 to 93.4 lo 687 lo 68.6
Average Total Water Saturation w 43.1 52.4 69.2 47.2 36.7 47.0 54.1 44.4 63.5 42.6 32.4 50.7 40.0 51.3 37.3 42.2 53.6 35.4 48.3 37.9 25.1 43.5 47.2 57.8 48.8 42.1 61.7 48.9 48.7 55.3 42.1 37.8 53.8 40.2 42.9 40.4 45.1 57.9 48.6 54.5 51.9 75.5 54.1 45.5 45.9 55.5 21.4 41.5 33.0 76.7 84.0 71.5 63.2 50.7 67.6 43.9 32.1
42 41
42
56
42 42
42 42
38.2 to 83.7 28.0 lo 76.3 33.2 IO 69.4 42.8 to 66.4 19.8 to 22.9 16.9 to 86.7 21.4 to 51 7 60.3 to 93 4 27.1 to 94.8 47.4 lo 84.9 22.6 to 93.5 18.9 to 85.3 32 9 to 94.0 19.3 to 76.5 148to522
42
45
35 to 48
27-12
PETROLEUM ENGINEERING
HANDBOOK
TABLE 27.9-OKLAHOMA-KANSAS
AREA (continued)
FormatIon Morrow
FluId ProductIon G 0 T G 0 a 0 T G 0 T G 0 T 0 a 0 a 0 i 0 a 0 0 2 T 0 a 0 T G 0 A G 0 T G 0 T 0
Range of Producllon Depth lfli 4,300 4,100 5,500 7,100 5,100 8,400 4,500 300 1,200 1,200 200 700 3,000 600 3,000 4,200 to 9,700 to 7,500 to 6,900 to 14,000 to 11,700 to 13,700 to 4,600 to 6,300 IO 5,800 to 5,300 to 3,200 to 2,500 to 6,600 to 6,700 to 5,400 to 7,400 3,500 to 3,600 2.100 to 3.700 3,600' 2,300 to 7,400 300 to 7,600 1,200 to 3,800 1,000 to 5,300 1,000 to 5,800 2,400 10 4,600 to 7,400 to 6,700
Average Production Depth m 6,100 5.700 6.100 10,900 8,300 12,300 4,600 3,800 3,300 3,100 1,200 1,500 4,000 3.100 3:700 4.500 4:200 3,800 3,600 3,600 4.300 3,100 3,100 3,700 3,200 3,400 1,100 3,500 4,600 5,600 4,800 2,700 2,200 2,800 13,400 8,000 8,600 5,400 4,900 4,600 800 4,300 4,900 3,900 6,700 6,500 6.000 4,600
Range of ProductIon Thickness (ff) 2 to 64 2to37 3to 30 14 to 149 3to 71 8 to 27 8 to 9 3.6 to 34.1 2 to 21 4to17 2 to 42 4 to 21 5 to 22 2 to 81 3 to 18 3 to 30 2 to 13 t to 32 5 to 7 4to 19 1 to 63 2 to 9 4 to 29 1 to 42.5 6 to 35.9 2 to 2 to 2 to 2 to 4 to 2to 8.9 to 21 to 2 to 3 to 3to 2 to 2 to 3.1 to 2 to 2 to t .9 to 5 to 1.3 to 1.5 to 2.6 to 40.5 84 27.5 28.5 9 14 16 268.4 136 86.5 73 111.7 117 34 35 28 29 28 32 5 30.4
Oil Creek
Oswego
Peru
Prue
Purdy Reagan
Redfork
Skinner
Viola
1,000 2,600 5.000 2,400 2,300 1,300 2,700 7,200 700 1,400 4,300 2,100 2,600 300 2,800 2.800 3,200 5,000 3,700 4,700 4,100
to 7,100
to to to to to to to 5,700 3,100 2,900 2,900 16,700 16,800 12,900 to 7,300 10,300 2,800 5,400 7,400 6,100 10,000 to 6,400 to 7,500 to 5,000
to 11,100
to to to to to to
Average Producllon Thickness (fU 11.0 9.8 9.5 46.3 12.6 15.0 8.5 12.3 10.6 9.8 12.4 10.3 13.6 14.6 11.7 14.8 4.8 7.4 11.0 6.0 7.9 10.5 5.3 11.8 9.2 11.5 12.0 12.4 26.4 9.8 8.7 7.0 7.8 12.5 78. I 15.3 20.0 39.1 17.2 19.6 10.8 11.3 10.0 7.7 13.4 11.3 4.4 16.2
Range of Permeabll~ty (md) 0.1 to 1,450 0.2 to 1,840 0.1 to 410 0.1 to 132 0.1 to 615 0.1 to 87 2.4 to 151 0.2 to 296 0.1 to 117 3.1 to 42 0.2 to 264 1.7 to 804 0.7 to 42 0.1 to 254 0.5 to 133 7.4 to 500 1.1 to 173 0.2 to 2,740 19.0 to 37 0.1 to 160 0.1 to 668 0 to 23 0.1 to 127 0.1 to 255 0.3 to 16 0.1 0.1 0.3 1.4 1.3 2.1 4.3 0.9 0.1 2.0 3.6 0.1 0.1 0.2 0.7 0.2 0.3 0.2 0.4 0.4 1.4 to to to to to to to to to to to to to to to to to to to to to 599 3.1 283 278 406 123 252 24 1,470 143 23 1,150 997 133 145 445 418 154 2,960 756 250
Average Permeablllty (md) 115 117 34.4 32.0 131 22.1 76.7 27.3 27.0 15.0 20.8 205 18.3 22.6 42.6 182 195 39.3 255 38.0 23.4 14.2 6.3 27.7 20.6 6.0 71.0 58.1 0.67 46.7 96.6 106 36 128 7.63 154.0 44.6 10.8 52.3 45.1 22.2 72.1 91.3 84.1 76.2 214.0 246.0 87.1
Range of Permeablllty. k,o (md) 0.3 to 55 0.1 to 48 0.2 to 230 0.1 to 86 0 to 41 51 to 266 -
2 to 6.6 2.40 0 to 1.3 8 to 22 53 0.5 lo 1.0 0.2 to 1.8 0.40 3.40 0.2 to 186 0.03 to 49 0.80 2.4 to 156
TYPICAL
CORE
ANALYSIS
OF DIFFERENT
FORMATIONS
27-13
TABLE 27.9-OKLAHOMA-KANSAS
AREA (continued)
Range of Calculated InterstitialWater Saturation W)
16
Average Permeability,
Range of Porosity w
4.2 to 24.4 5.7 to 23.2
Average Porosity w
148 14.6
7.5
23.1
28.0 75.6 9.24
to to to to to 2.6 to
5.5
5.0 1.6
13.0 2.6 5.8 15.0 5.0 4.1 14.7 12.0 55
1.3 to 29.5
0 to 5.6
21.7
39.8
5.1 0 0 0.1
to to to to
16.2 41.7
39.1
46.6 47.7 41.5 63.4 52.5 50.6 50.7 42.2 41.6 47.1 41.5 56.2 44.4 32.9 40.0 45.8 43.7 52.6 40.8 40.3 52.4 61.8 45.6 45.5 44.5 45.0 49.0 40.7 61.2 33.2 34.9 45.7 30.7 54.4 65.7 51.3 43.9 32.0 41.7 30.9 36.9 42.5
16 31 12 14 21
to to to to to to
77 54 90 40 76 74 73 89 56 73 56 49 72 60 50 68 72
45
33 to 43 29 to 42 35 to 46 25
34 to 55
to to 10 1 to 13.3 to
3.8 6.6 11.7 to
7.4 to 21.7
to to to to to
18 3
4.38 0.80 79.2
21.4 22.9 16.9 20.3 19.5 11.9 25.0 26.0 10.1 16.1 18.8 24.9 15.6 20.5
17.0 17.5 16.7 17.6 10.8 13.3 11.7 14.5 162 15.3 15.7 15.3 15.5 21.3 16.8 13.3 16.4 18.4 17.1
15.6 15.7 6.1 11.6 11 .o 9.3 8.4 7.1 16.6 10.8 12.0 10.9 11.2 12.4
16.9 19.0
20.0 13.6 4.2 14.2 6.2 4.7
15 to 42 to 44 to 28 to 36 to 25 to 20 to 32 to 16 lo 28 lo 12 to 29 to
to 43 to 46 41
34
39 to 44
16.9 9.9
4.2 20.1 0.5 9.9 15.1 21.1 2.0 12.5 11.4 16.0 9.0 4.1 12.2 2.6 5.0 15.5 8.6 18.6
to 57.7 to 69 7
16 lo 63 27 to 55
to 43 32 to 48
24
41 to 69 26 to 47 14 to 78
39 to 71 -
30 to 46 -
0 to 6.1
7.5 to 16.5 6.9 to 17.3 7.3 IO 29.6 7.1 to 10.9 0 to 6.6 3.0 to 44.5 0.7 to 7.7 1.7 to 9.4 3.2 to 41 0 0 to 33.7 8.1 to 33.8 0.7 to 6.3 3.6 to 40.5
to to to to to to to to to to to to
61.5 61.6 58.3 78.0 52.6 50.1 64.3 54.8 63.0 82.6 37.2 85.5
to 56 32 to 62 27 to 56 31 to 78 4-4 to 52 33 to 43 52 to 62 23 to 55 9 to 63 15to82 19 to 37 24 to 88
22 39 to 90 28 to 67 29 to 80 14to58 171043 18 to 58 40 to 60 43 to a7
31 to 44 33 to 36
40 to 45 29 to 40 49 5
to 90.8 to 68.0
3.6
11.7 7.9 1.5 10.2 6.1 11.8
29 7 to 60.5
15.0 to 58.2 24.6 to 63.6 17.7 to 45.0 19 0 to 56.3 41.4 43.0 to 60.5
to 50 28 to 48 29 to 42 32 33 to 50 34
12.9
4.4
8.3 to 16 7
to 87.9
60.1
to 42 41
a General geologic sections take at dtfferent points I Oklahoma-Kansas areas lndlcate some var!at!o s I the properties and a apprec~abfe variate I the occurrence and relative depths of many of the more m~portanl 011.and/or gas-producing zones. formations, geologic groups, and thelr members The general !de tlflcatlo of core samples from thee producing lntewals reflects local condmons or actlwt~es slgnlflcantly In the development Of average data values. an attempt has bee made to combine data orlgmally reported for locally named zones Into more generally recognued formatlow or geologic groups In some mstances (I e Deese. Cherokee) data are reported for a major geologic group as well as for $ome of 11svndlwdual members The values designated by the maw group name represent areas where the general character~stlcs permit Identlficatlo as to the gealognc group but not as to group member In other areas the group members or zones are readily ldentfffable The combmatlons of data and the use of local rather lha regmnal geologic names I some instances are emplaned 1 the footnotes b T represents transitlo zone or productlo of both water and &her gas or 011 fncludes data reported as Dornlck Hllfs and Dutcher Includes Bromide first and second as reported on McClaln County area g Data reported locally as Bromide third. Bromide upper third. and Bromide lower have bee ConsIdered as part of the Tuhp Creek Includes data reported as Cleveland sand, Cleveland lower. and Cleveland upper fncfudes the numerous zones (Deese first. second. third, fourth, fifth, Zone A, Zone 6. Zone C. and Zone 1) reported locally for the Anadarko, Ardmore, and Marietta Basm areas. I northwest Oklahoma. these different zones are normally referred to as Cherokee In other areas the zones are frequently Identlftable and properties are reported as for Redfork. Bartleswlle. etc
27-14
PETROLEUM ENGINEERING
HANDBOOK
TABLE 27.10~ROCKY
MOUNTAIN AREA
FormatIon Aneth Boundary Buite Cliffhouse D Sand Dakota Desert Entrada Frontier Sands Gallop Hermosa Hospa lsmay J Sand Leadville McCracken Madison Manefee Meeaverde Morrison Muddy Nugget Paradox Phosphoria (formerly Embar) Pictured Cliffs Point Lookout Shannon Sundance Sussex Tensleep Tocito
Fluid Production
Range of Production Depth 0) 5,100 to 5.500 to 5,400 to 3,600 to 4,350 to 500 to 653 to 5,400 to 3,600 to 265 to 1,5M1 to 500 to 4,900 to 5,300 to 4,800 to 4,600 to 5,544 to 4,470 to 6,970 to 9,950 to 8,264 to 3,400 to 5,200 to 1,500 to 1,600 930 9,900 9,500 5,100 5,300 to to to to to to 5,300 5.600 5,900 5,800 5.050 7,100 7,293 5,500 3.700 8,295 6,900 6,400 7,700 6,000 7,100 5,100 5,887 5,460 8.040 10,100 9,466 6,200 5,700 6,100 6.900 8,747 10,300 10,800 9,500 6,100
Average Production Depth (fi) 5,200 5,600 5,600 4,800 5,800 5,700 5,600 3,640 2,950 5,000 4,600 5,600 5,600 5,500 4,800 5,707 4,900 7.500 9,950 8,820 4,900 5,400 4,700 300 4,500 1,845 10,100 10,375 6,900 5,700 4,600 3,400 2,900 5,500 4,700 4,900 3,100 4,500 4,700 7,900 4.600
Range of ProductIon Thickness (f1) 3.8 to 23.1 8 to 27 2 to 68 2 to 56 7 to 33 2 to 75 13 to 75 11.6 to 18.3 4to10 8 to 100 5 to 25 2 to 43 5 to 30 3 to 36.2 3to17 6to 1El 10 to 90 151062 20 to 76 2 41 7 2 24 7 60 250 4 2 to to to to to to to to lo to 142 450 25 22 54 75 700 700 44.2 66
Average Production Thickness (W 14.0 17.5 16.2 13.7 15.0 32.0 32.0 14.9 6.0 46.0 11.6 12.4 .14.1 15.1 10.5 133 36.4 25.0 45 0 15.0 56 4 186.0 12.7 10.0 4.0 40.0 20.0 385.0 475.0 12.2 14.8 64 0 17.0 23.0 22.9 7.0 15.0 44.0 20.0 118.0 7.0 17.3
Range of Permeability (md) 0.7 01 0.1 0.1 0 0.1 0.1 1 .o 5 0 0.1 0.1 0.1 0.1 0.1 07 0 1 0 0.01 0 0.01 0 0 1 0.1 0 0 0.6 0.5 0.2 0.1 to 34 to20 to 114 to 3.7 to 900 to 915 to 915 to 11 to 300 to 534 to 324 to 2,470 to 91 to 37 to 70 to 25 lo 142 to 1,795 to 0.50 lo 21 to 272 to 1,460 to 20 to 17 to 1,250 to 2,150 to 65 to 85 to 42 to 119
Average Permeability OW 9.35 1.05 13.3 0.94 192 106 106 4.4 100 105 26.5 48.2 18.6 7.32 18.2 8 63 to.4 330 0.20 3.0 5.8 13 5.04 3.57 60 43 173 : 11.6 10.4 3.7 7.7 0.5 1.74 2.90 0.8 100 1.0 120 230 3.36
Range of Permeablhty. km VW 0.2 to 23 0.2 to 23 0.4 to 2.4 0.3 to 20 0.1 to 3.2 45.0 0 to 26 -
0.1 to 28 0 to 57 I
700 to 10.500 1,200 to 5,800 4,300 to 6.500 4,700 1,100 4,300 600 to to to to 5,500 6,860 5,100 11,800
1,400 to 5,100
OF DIFFERENT FORMATIONS
27-15
MOUNTAIN AREA (continued) Range of Calculated InterstltialWater Saturation W) 13 to 31 23 to 35 7 to 45 lOto 9 to 46 14to22 26 to 45 20 to 54 14 to 77 12to45 12 to 60 8 to 49 31 to 45 6 to 42 22 to 33 15 to 43 15 to 64 15 to 41 5 to 47 20 to 56 18 to 40 10 to 58 10 to 61
Average Perm_eability,
4.43 4.57 -
Range of Porosity W) 4.4 to 10.5 4.3 to 6.5 5.4 to 21 6 7.0 to 16 2 8.6 to 29 5 4.5 to 21 6 5.0 to 23.3 11.9 to 13.6 12.0 to 27 0 6.3 to 29 6 8.5 to 20.8 6.9 to 23 1 5.5 to 16 5 2.7 to 17 9 7.4 to 11 9 6.6 to 14.8 0.5 to 22 2 5.9 to 32.7 3.0 to 20.0 2.0 to 16.0 0.5 to 15.1 1.6 to 26.4 8.7 to 13.5 10.0 to 19.8 9.9 to 25.5 2.3 to 32 9 10 0 to 18.0 10 0 to 18.0 1.4 to 19.4 3.3 to 21.8
Average Porosity W) 6.1 47 11 .o 11.3 21.6 14.8 t 1.2 12.7 25.0 20 0 t 3.3 12.5 to.2 8.3 10.5 11.3 7.6 19.6 10.0 3.0 6.5 11.9 11.2 146 26.2 17.5 22.3 13.7 13.4 7.4 10.5
Range 01 Oil Saturation W) 14.5 to 35.9 4.7 4 8 lo 26.7 0 to 19.8 8 4 10 39.5 0.0 to 7.8 13.8 lo 54.5 13.4 to 16.6 trace to 6.0 7 6 lo 37.6 0 to 25.6 0.5 to 43.7 0 to 6.5 3.9 to 29.1 0.5 to 23.6 20.4 to 29.8 1 6 to 26.4 8.8 to 46.5 0.0 to 22.2 trace to 6.0 0.0 to 50.1 6.0 to 43.5 0.3 to 5.3 Oto68 5.0 7.6 0.0 50 0 36 to to to to to to 26.0 48.5 5.0 10.0 10.1 36.7
Average Oil Saturation wd 25.0 4.7 12 5 4.5 13.2 3.5 24.4 15.2 3.0 14.9 5.7 25.3 3.0 10.8 7.5 25.0 a.4 13.9 2.0 7.5 14.4 17.4 1.6 3.3 8.3 13.1 30.8 3.6 6.4 3.1 12 4
Range of Total Water Saturation w 12.5 to 30.5 23.8 to 35.0 9.3 to 48.8 10.2 to 60.3
Average
Total Water Saturation w 23.6 29.4 26.3 36.9 40.6 31 .o 19.2 40.0 35.7 32.7 35.6 36.1 36.0 3Co 27.5 42 0 61 .O 28.0 32.5 34.7 33.6 47.0 40.9 36.7 40.9 45.0 45.0 25.7 51.6
Range of Gravity
(OAPI) 41 40 to 41 36 to 42 38 to 43 31 to 50 39 36 to 42 41 to 42 40 36 to 42 45 to 46 21.6 to 30 29 to 56 26 to 42 40 to 43
Average
Gravity (OAPI) 41 41 1 36 40 41 2 39 40 ;: 38 45.5 26 42 38 48 41
14.8 to 55.3 11.6 to 44.3 14.8 to 24.7 20.7 to 59.2 17.2 to 76.9 14.2 to 45.3 11.6 to 60.0 8.7 to 49.7 32.3 to 44.6 10.0 to 65 0 14.5 to 45.1 145to664 20.0 to 60.0 20.0 to 50.0 9.9 to 57.9 10.8 to 60 5 16.0 to 11.9 to 40 30.0 to 35.0 to 7.0 to 40.6 to
2 0 to 25.0 3 1 to 31.0 5.6 to 6.0 to 15.0 to 8 0 to 5.0 to 12.6 to 21.6 15 0 25.0 20 0 27.0 17.8
2.40 -
8.9 17.5 11.4 10.9 13.3 12.0 19.0 130 13.6 20.2 14.7
3.0 to 40.0 0.0 to 21 .l 0 to 9.1 3.0 to 22.0 8.0 to 25.0 5 0 to 20 0 6.0 to 30.0 11 9 to 26.6
22.5 2.6 23.2 2.9 23.8 16.0 17.0 11.0 23.3 4.0 21.3
5 to 30 121055 20 to 49 5 to 50 40 to 55
21 36 41 35 19 43 46
88.0 55.6 60 0 60 0 59 55
15 to 42.3 f 22 to 63 40 to 43 17 to 56.5
46.3
36 to 40
25.4 55 39 39 44 39 42 26.2 36
27-16
PETROLEUM ENGINEERING
HANDBOOK
TEXAS-SOUTHEASTERN
FluId Form&on Bend Conglomerate Group 1 2 Blwbry Cambrran Canyon reef Canyon Clearfork sand 1 2 Dean 1 2 3 Ellenburger Fusselman Glorletta (Paddock)g Granite wash Grayburg 1 2 3 1 2 3 All All 3.4.6.7 a 4 5.6.7 1.2.3 2.3.4 All 8 5.6 1.2.3 4.7 1 2.5.8 47 2 All : 0 0 G 0 G 0 0 : 0 G 0 0 : 0 Ei G 0 G 0 G : 0 0 0 G 0 San Andres Area 1.2.4 5.6.8 3 8 3 2.3.4 5.6.7 z3.4.7 3.4.7 mart) 5.6 2(w) 2.4.7 1C 2.50 2.5e Production 0
Average Production Depth (fl) 6.000 10400 5.480 5.610 5.900 7.100 5,000 5 500 2,400 4.400 6,600 8.200 4.800 4.200 11.400 11.800 9.200 7.700
11.100
Range of ProductIon Thickness lfll 3 to 22 10 to 28 23to 50 410 95 20 to 95 40 to 222 3010 40t0180 3010 259 6 0 to 68 52 to 39 3 0 to 52 1410 117 8 to 299 Et0 19 to 65 to 11 to 30 to 18 to 8 to 3to 113 34 954 18 347 51 49 44 57
Average Productfan Thvzkness (f1) 13.2 20 0 36 43 30 3 36 a 80 16 9 95 41 33 26.2 18.6 14.5 54 99 17 34 27 69 14 3 55 34 32 16 3 22 3 51 15 6 42 274 20.8 45 50 22 3 99 10.2 18 6 40 1 30 2 56 1s 5 21 7 IS 5 34 4 36 7 16 8 15 1 14 335 10 6 41 7 22.5 28 0 59 10 8 16 6
Rangeof Permeablllty (W 4 to 311 16toll 1.6 to 3 8 0.1 10 5 3 0.8 to 1,130 0.6 to 746 01 to 477 -
Average Permeability (4 150 5.7 24 1.8 173 42 17 38 11 4.6 5.8 0.12 12.9 245 10.5 4.0 0.4 14.9 1.1 177 225 75 8.4 10.3 5.6 11.5 477 609 6.5 13.7 25 55 37.7 349 64 123 61 69 9.7 12.2 51.4 63 48 179 43 11 4 47 45 276 23 419 57 60 204 19.3 427
0
G 0 0 0 G 0 G : 0 E G 0 G 0 C 0 C 0
10.300 to 10.500 5 383 to 5.575 5 262 to 5.950 5 500 10 6.300 4.200 to 10.400 30001010000 1.500 5.400 to 6.800 to 8.300 9.100 5,000 5.100 11,600 12.300
Delaware
33 84 36 23
5.500 to 9.900 11 000 to 11:zoo 7 800 4 100 5.500 8.700 9.500 2200 2,300 3.000 2,300 3.600 2.400 to to to to 12.800 10.600 16.600 12,700
2.5 to 50 <OllOZ2 1.0 to 2,840 203to 246 0 1 to 2,250 12 to 26 0.5 to 25 46to 12 04to 223 11 to 2,890 5 to 3.290 t6to93 05to159 0.6 to 3 7 02to 118 0 3 to 1.430 0.3 to 462 loto 318
11.200 7400 10.100 10.300 12.000 2.400 4300 4.700 3.000 3.800 4.100 4.400 4400 2.700 9100 3100 3500 4.500 4.500 3.300 3900 2.600 7100 6 900 5600 5.900 7.800 5.200 3938 6500 9 800 8.800 3.600 3.500 9700 2800 2.300
4.400 3.000 to 4 800 1.300 to 3 900 4.100 to 11.400 3 4 4 5 200 900 700 300
sand
(Penrose)g
4 0 to 29 15 to 38 610 39 47to124 30 to 197 301080 40 to 136 2 0 to 59 20 to 120 11 to 57 2010101 3 0 to 39 20 to 76 6 to 21 1510 43 13 to 129 4oto119 20 to 114 45 to 204 30 to 53 30 to 66
0.2 to 4,190 0 3 to 461 0 3to 295 0 2to 0.6 to 04to 0.2 to 0 1 to 4 5to 19to 0 3to 02to 593 23 428 71 124 310 196 42 718
5 100 to 4 100
Seven Rivers
Sprayberry Strawn Straw llme sand
1 2 -
I300 to 4 000 4.800 to 8 500 5oooto 9 200 5.200 3.800 I.100 to 6 700 to 10.500 to 11.300
915to 7 366 6.100 to 7 300 8.400 2.500 2.400 9.000 1.400 1.4oa to 9 200 to 4 100 to 4 100 to 10.600 to 3 500 to 4 000
1 oto 400 02to135 23to9410 0 1 to 1.380 1 0 to 1,270 02to147 02to 145 1 oto 4 000
27-l
TABLE 27.11-WEST
TEXAS-SOUTHEASTERN
Average Permeabtlity,
Average Porostty (Oh) 150 10 9 12 7 78 120 89 15 1 14 3 13 5 92 58 10 3 179 21 0 33 43 67 15 2 50 60 42 36 33 3.3 15.0 13 6 14.4 17.7 124 11 3 6.4 79 119 77 16.6 172 85 71 15 5 160 16 5 158 11 7 129 72 6.9 126 16 2 49 43 99 153 15 5 81 179 18 8
Range of 011 Saturation 1%) 8.1 lo 8 6 9.5to161 2.1 to 9 1 9.6 to 19 3 6.9 to 21 2 3 6 to 39 2 4.8 lo 27 7 7.5 to 31 4 5.6 lo 27 1 ,22 to 2.0 to 3.910 2.1 to 44 103 156 6.6
Average 011 Saturation 1%1 8.3 11.5 4.9 12.8 11 8 11 6 58 13 7 57 156 16 5 33 7 60 11 2 37 92 57 11 0 40 129 42 a4 17 104 37 154 52 14 7 18.6 176 47 13.9 182 97 74 156 187 147 132 64 16.2 15 3 15.5 5.9 11 2 30 122 14 1 12 9 19 6 97 46 14.3 14.4 56 16 0
09to51 ozto42 0 3 to 249 <O.l 10.1 0.1 to 1.3 0.1 to 5.8 0.2 to 18 <O.l to 0 9 0.3 io 1,020 1.4 54 <Ol to396 03to13 02 to 17 02to126 0 2 to 48 0.3 to 2 1 0 1 IO 110 0.1 to 228 0.1 to 168 0 1 to 462 0.1 to 208 0.2 to 510 70 0.5 37 27 7 22.9 09 3.9 93 8.1 53 30 52 13 27 143 147 10 53 38 84 03 80 40 108 19 1 02 11 7 05 04 274 34 43 54 27 8 to 24 to 109 0.5 08 2.2 1.4 17 -
3.0 to 21.5
5 5 to 22 1 7.8 2.5 3.1 4 1 to 20.6 19to194 75to127 138to218 152to254 17to53 13to68 5 5 to 27 7 2 2 to 7 7 1 e to 25 2 37to46 13to 138 26to37 14to107 14 to 182 5 2 to 20 9 121 to204 3 5 to 26 1 11 1 to 143 7 0 to 20 0 631066 27to162 5 3 lo 24 3 2710139 10 7 to 22 2 5 7 to 27 0 32to140 31 to128 3 3 to 25 1 15 5 5 9 101 4 4 10 9 to 16 to 28 to233 to 20 IO 14 6 9 6 8
40 28 32 39 40 49 42 42 47 48 33 42 36 32 31 41
3.3 to 16 7 6.8 to 22 9 3.1 to 4 8 5 3 to 24 6 08to76 1.010 192 02to39 5.2 to 16 39to44 3 1 to 22 1 29toa7 4 8 to 22.5 7 1 to 42 6.2 to 37 9 24to71 4.8 to 22 1 8 3 to 34 47to 188 2.6 to 7 6 4.2 to 34 7 8 9 to 33 9 4 9 to 30 6 3 5 to 24 2 34to95 42to41 7 7 0 to 24 5 7.0 to 30 55tof33 4 9 lo 26 3 17to52 2 7 to 279 5 0 to 27 8 5 to 25 3 66to 168
19 to 53 41 45 22 47 40 32 25 39 to to to to to to to to 76 69 65 67 84 47 65 60
to
24 to 72 39 to 66 42 lo 71 22 26 55 32 31 28 to to to to to to 53 56 68 84 78 58
24 lo 71 39 to 66 35 to 66 22 25 55 32 31 28 to to to to to to 52 55 68 84 78 58
36 to 62 32 to 68 21 to 49 26 to 69 39 to 74 51 38 32 25 38 to to to to to 66 70 68 72 39
35 to 58 30 to 66 19 to 49 25 to 69 37 10 74 46 36 30 25 38 15 46 23 to 10 to to to to lo to 65 61 67 72 39 66 60 77
30 to 42 34 to 38 30 to 37 26 to 37
33 37 33 32
15 to 66 46 to 60 23 to 77 25 to 60 37 to 64 32 to 56 30 lo 64 32 10 65 28 to 56 id to 79 31 to 75
011011
1.5 to 6.210 0 2 to 36
23 10 59 37 to 64 31 to 56 26 to 64 29 to 64 28 to 56 36 to 76 31 to 75
27-18
PETROLEUM ENGINEERING
HANDBOOK
Fig. 27.1-Area
27-19
Fluid Production G : 0 E 0
Range of Production Depth fftl 4,500 6,100 6,200 8,300 2,850 6,950 4,400 to to to to to to to 8,100 10,800 6.700 8,800 7,500 7,800 14,800
Average Production Depth (fo 5,640 8,600 6,200 8,600 6,230 7,200 6,150
Range of Production Thickness (fi) 40 to 106 20 to 1,300 30 to 80 350 to 630 22 to 130 36 to 92 90 to 1,000
TABLE 27.12-ALASKA
10 to 18
15.0
35 to 44
40
TABLE 27.13-AUSTRALIA
(GIPPSLAND
Formation (Reservoir)
L-l L-l M-l M-l M-l M-l N-l N-4
Production
Fluid Production : G G : G
0
Range of Production Depth (m) 2,299 1,650 to 1.521 fo 11299to 2,352 1,018 to 2,396 1.950 1.556 1:377 2,396 1,151
Average Production Thickness b-0 80.4 6.0 7.5 59.1 40.0 37.2 99.0 2.7
Average Range of Permeability VW 800 to 3000 600 500 to5000 Average Permeability VW 2.occr 3;oO0 5,ocO
1,000
Interstitial-
SatZion W) 0 0 0 -
Water Saturation w f: 25 10 26 24
16
1,186 to 1,383
1,330 to 1,339
1,000
40
27-20
PETROLEUM
ENGINEERING
HANDBOOK
TABLE 27.14-ALBERTA.
CANADA Average Calculated Permeaiility 0-4 3.7 7.4 32.2 1.0 40.1 208.0 95.0 187.0 302.0 682.0
1,000.0 14.0
Formation Cardium A Cardium A Beaver Hill Lake A&B Father (conglomerate) Falher (sandstone) Gilwood Keg River Keg River Leduc 03 Leduc 03A Taber Viking Viking A
Pool Barrington Willesden Green Swan Hills Elmworth Elmworth Nipisi Rainbow Rainbow Red Water Bonnie Glen Taber Viking Kimsella Gilbey
Flutd Production 0 : E 0 G 0 0 : G 0
Average Production Depth WI 6,634 6,225 0,345 6,500 6,500 5,651 6,082 6,381 3,208 6,000 3,500 2,400 6,401
16.1
19.4
238.0
23.0 35.0 42.5 14.0 19.9 25.6 24.2 25.0 35.0 35.6
TABLE 27.15~MIDDLE
EAST
Range of Reservoir Water Saturation W) 9 to 100 8 to 16 151040 25 to 45 151035 Average Reservoir Water Saturation
Fluid ProductIon 0 0 0 0 0
Range of ProductIon Depth (fo 7,400 10 7,980 4,125 lo 4,422 10,000 to 12,000 10,000 to 12,000 10,000 to 12,000
Range of Permeability OW 0.3 to 6,000 2.0 to 10 0.5 to 1,000 0.2 to 500 0.5 to 1,500
Range of
Porosity W) 5 to 34 27 to 37 15to22 lot020 15to30
Average
Porosity W) 21 33 18 15 20
P/o)
25 10 25 35 20
TABLE 27.16-NORTH
SEA
Average Reservoir Water Saturation w/o) 23 20
Range of Porowty W) 25 to 30 7 to 37 IO to 26 -
Average Porosity W) 27 28 23 30
0
0 0
Chapter 28
Relative Permeability
Walter Rose.con\ultant *
Introduction
This chapter was written as an overview of relative permeability: the basic ideas are given and their evolution is traced. Also presented are some laboratory measurement details and comments on the use of relative permeability information in problem solving. Many unresolved issues still exist, regardless of the fact that the literature is quite rich with the descriptions of what previous workers have thought about this complex sub,ject Fluid flow is the major transport process that is involved in the recovery of oil, gas, and associated formation waters from subsurface petroleum reservoirs. As a consequence, process descriptions are needed to understand. to forecast. to manage, and to control production operations. Relative permeability is the concept that is often used as a framework for describing two- and threephase flow of immiscible fluids through porous sedimentary rocks. The term permubility historically has been adopted as a measure of the porous rocks ability to conduct fluid. If only one fluid is present in the interstices, this transport coefficient is called the speciJi(, permeability, but otherwise one must make reference to the c$kcti\v permeability of each of the immiscible fluids in the connected pore space. Relative permeability by convention is the ratio of the effective to the specific permeability. In transport theory one deals with U%.re.s. ,@(vs, and the coefficients by which these variables arc intcrconnetted. As soon as the force-flux relationships are established for particular cases, they can be written in such a way that the various permeability functions will appear explicitly. In other words, just as specific permeability has the sense of a transport coefficient that appears in Darcy s equation for single-phase flow, effective and corresponding relative permeability functions also can be thought of analogously as important transport coefficients by which multiphase flow processes are best described. Central to the relative permeability idea is that we arc dealing with material response types of parameters that cannot be derived from theory alone. On the contrary. laboratory measurements generally also will have to be made. All measurements are to be made in accordance with how the experimental variables properly are defined and used later. To simplify the discussion, assume that we are dealing with the steady flow of an incompressible fluid that is moving macroscopically in a horizontal direction. Then. kAAp 4=- I.IL ,
.... ....
. ..... .
. . (1)
Aulhor
of the chapter
where Ap is the pressure drop that can be measured across a specimen sample of length L and cross-sectional area A. and q is the volumetric rate of discharge of the flowing fluid whose viscosity is CL. Eq. I serves to prescribe an experimental measurement methodology by which values for k can be determined unambiguously. In particular. it is seen that a plot of q vs. AI, should be a straight line whose slope is proportional to the specific permeability. k, as long as certain limiting conditions prevail. These are that (1) the fluid is homogeneous: (2) the temperature is constant: (3) the transport process is free of electrokinetic effects (as occur when certain dilute electrolytes are being moved through electrically nonconducting media), and of film surface flow (as oc-
28-2
PETROLEUM ENGINEERING
HANDBOOK
curs when a gas is moving at such low mean pressure that the frequency of molecular collisions at the interstitial surface boundaries becomes important); and (4) the porous rock is sufficiently rigid and inert to preclude rapid changes in the pore geometry. Accordingly, in what follows, we adopt the idea that if Darcy s law adequately describes single-phase flow, analogs of it can be postulated just as well as useful descriptions of multiphase flow phenomena.
Historical Background
In the 1930 s and 1940 s, serious ideas about the meaning and measurement of relative permeability phenomena first appeared. Principal authors were Muskat, Botset. Hassler, Leverett, and perhaps a dozen more whose names punctuate the milestones described in the standard references. .* These people were connected predominantly in one way or another with the U.S. petroleum industry at a time when quantitative study of recovery of hydrocarbon fluids from subsurface sedimentary environments was in its infancy. In focusing on how reservoir fields of fluid flow would be influenced by the nature and number of the coexistent interstitial fluids, the early workers took advantage of the fact that the existing and well-developed understandings about single-phase flow in porous rocks, when generalized. seemed to provide credible descriptions of multiphase flow situations. When L.A. Richards3 published his classic 1931 paper on the flow of capillary-bound moisture in so-called unsaturated soils, Darcy s pioneering work already had stood the test of three-quarters of a century of scrutiny, use, and amplification by diverse groups of technologists that included groundwater specialists as well as petroleum, chemical, and civil engineers,
where Vfi is the volume of fluid j and V,,, is the total pore volume. In other words, the product C&S,can be thought of as an effective porosity of that portion of the partitioned pore space occupied by the jth fluid phase. It follows that locally (i.e., in any representative volume element of the reservoir rock system of interest) Es,, = 1, even when saturation is changing with time and position. In any case, it would appear that the relative permeability, k,, has a first-order dependency on the saturation level, Si. But this is only one of the dependencies, because usually many interstitial fluid phase distributions are possible for each level of saturation. Also, the reduced pore space occupied by each of the several pore fluid saturants is not necessarily everywhere bounded by interstitial (solid pore wall) surfaces; hence, it is possible that the influence of prevalent fluid/fluid interfacial boundaries also may have to be considered. Thus, the statement O<kr;<l forO<Sj<l, . . . . . . . . . . . . . . . . . . . (4)
while always true, ignores the fact that k,., may equal zero even when Sj is finite, and may be greater than unity even for values for S, less than unity.4. Moreover, Eq. 4 is too crude to indicate explicitly that krj can have more than one value for each value of si (e.g.. because of hysteresis).
Hysteresis From the thermodynamic point of view, the fluid flow processes under consideration are irreversible (i.e., nonequilibrium); therefore, they are inherently pathdependent. One consequence is that the equilibrium states approached in one direction can be different from those approached in another. This phenomenon is called hysteresis, and explanations for it are not hard to find. For example, while specific permeability depends mostly on the interstitial pore geometry, effective (hence, relative) permeability depends on the fluid saturation geometry as well. Usually there will be more than one way a given fraction of the pore space can be occupied by each fluid phase of interest. The result is that relative permeability data give values that are functions of the history and sequence of the prior saturation changes as well as being merely functions of the fluid saturation levels. As will be seen later in this chapter, little laboratory work has been published to prove that Eq. 2 truly describes multiphase flow phenomena. Nor can much reassurance be taken from the fact that the validity of Eq. I as a description of the characteristics of single-phase flow has been confirmed experimentally for a large number of cases. This is because the unsteady states are of the greatest interest when dealing with petroleum recovery problems, but this aspect is ignored by the Darcian modeling. In other words, Eq. 2 by itself does not make it possible to account for and to predict the saturation changes that occur as flood fronts move through reservoir space during production. And in the real processes to be modeled, the various phases may be flowing in separate directions rather than colinearly. The greatest limitation of Eq. 2, however, is that it gives no explicit prescription
Framework Ideas
By analogy to Eq. 1, a set of separate equations can be written to describe multiphase flow phenomena under the restricted conditions that no gravity forces are affecting the steady flow of each of the incompressible immiscible fluids, or 9, =(kjA)(AP,)l(~jLjL)=(Wc~jA)(APj)i(~jL), where the subscript j refers to the jth fluid phase (oil, k, is gas, water). Following the usual terminology. called the effective permeability, krj is called the relative permeability, and Pj denotes values of pressure locally measured in the various fluid phases that are separated by interfaces of contact. In particular, kj and kti are to be considered measures of the flow conductivity afforded by the porous rock when saturated with the immiscible fluid phases in some particular way. As will be seen, reference must be made to fluid saturation configurations and distributions as well as to fluid saturation levels. For example, and in analogy to the porosity concept where 4 is given by the local ratio of pore to bulk volume, the saturation of the jth fluid, Sj, is defined as
SQV8 /Jr
,,...............................(3)
RELATIVE PERMEABILITY
28-3
of how to avoid end effects (the buildup of wetting fluid saturation levels at surfaces of capillary discontinuity such as core end-faces) in laboratory work.
Pore Fluids
Whenever wetting and nonwetting immiscible fluids compete to occupy the same pore space, it is clear that at inflow and outflow surfaces of so-called capillary discontinuity, there will tend to be a buildup of wetting fluid saturation levels during the course of multiphase flow processes. This has been understood from the earliest days. 3.6 The central idea is that immiscible fluids that are co-existent in contiguous capillary pore space generally will be separated by curved interfaces of contact instead of by stress-free flat interfaces at contacts that occur exterior to the flow system. In fact, the inter-facial curved boundaries are a retlection of the balance between capillary and gravity forces in the static (stationary) cases, and of the viscous forces as well in the dynamic cases. This means that locally there usually will be a pressure difference between the immiscible fluids. This pressure difference, commonly called the capillary pressure, P,., by convention is defined as the local difference between p,! and pIr (where the subscripts n and MJrefer to the nonwetting and the wetting fluids, respectively). In systems at equilibrium. immiscible fluids tend to be distributed such that the free surface energy of the system is at a minimum, subject to the constraints imposed by the S, levels and by the hysteretic path of saturation change being followed as successive equilibrium states are being established. This usually means that the wetting fluid will be found in the smaller pore spaces, and that the interfaces of contact will be concave toward the nonwetting fluid (hence, P, will have positive values). In fact. capillary pressure, like kj and k, , frequently is assumed to be primarily saturationdependent, but it depends substantially on the fluid/fluid interstitial configurations at each saturation level. The Buckley-Leverett equation6 takes the following special form in describing two-phase displacement processes involving one-dimensional (horizontal) flow of incompressible fluids. With density p, =constant and aq,l&=O, we have
such as porosity, fluid viscosities, the specific permeability, and the total flow rate given by (q,,, +q,l)]. Indeed, the importance of having relative permeability information is to make it possible to solve reservoir engineering problems modeled by process descriptions such as given by Eq. 5. The more elaborate modeling technique involving equations with the transport and interaction coefficients described by Rose require considering more than simple relative permeability information, however. Even so, the experiments prescribed by the Buckley-Leverett and Rose equations will not be easy to perform. The laboratory difficulties obviously multiply if unsteady states are to be considered (where Sj changes with time and spatial position). This is why only state-of-the-art methodologies will be described here. Nonetheless, it is clear that future workers will continue to be challenged by the need to develop and to perfect measurement methods that are based on modeling that goes beyond the simplistic Darcian reasoning embedded in Eq. 2.
Measurement Methodologies
A number of measurement methodologies have been described in the literature; these are classified and explained to some extent as follows, along with some of the data that have been reported. Two controversial questions are (1) do the various methods yield equivalent data, and (2) if not, which are the most trustworthy ones? Laboratory measurement techniques for relative permeability determination are of two sorts. In the socalled steady-state methods, the effective permeability as a function of saturation is calculated from the flow data that are obtained on the assumption that Eq. 2 is correct in form. The trick is to make direct measurements prescribed by the theory, of parameters such as volumetric flow rates, pressure drops, and fluid saturation levels. In one variant of the method, known as the Hassler technique, provision also is made to control and to measure the local values of capillary pressure to avoid the troublesome end effects. The so-called unsteady-state methods are based on using integrals of Eq. 5 as the process model. The idea is to observe the consequences (i.e., the outcomes in terms of cumulative production) of controlled multiphase displacement experiments, and then to back-calculate the relative permeability values that are consistent with, and serve to explain, those outcomes. The cumulative production data also are processed to provide a basis for calculating average saturation levels to be associated with the relative permeability values. As will be seen, the steady-state methods are more time-consuming than the unsteady-state methods; still, the data obtained by them are at least as believable as the plausible model on which they are based (Eq. 2), especially if convincing measures are taken to minimize the capillary end effects mentioned previously. On the other hand, the unsteady-state methods, comparatively speaking, supply the wanted data quickly and cheaply. This latter advantage, it may be argued, is only a partial compensation for the uncertainties in data interpretation inherent in the indirect nature of such measurements. To expose the rationale for selecting one type of procedure over another, the ideas of recent authors are reviewed below.
=o.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...(S)
where 4 is the local porosity independent of time, Af is the cross-sectional area of flow path, P,. is the capillary pressure, L,, is the horizontal distance parameter, and t is the time (both independent variables); and h4, and M,,. are the mobilities of the nonwetting and wetting fluids, respectively (i.e.. M, =k,llLj). As can be seen from the derivations of Eq. 5 given in Refs. 1, 2, and 6. the saturation levels vary with position and time in a way that depends on the variation of P,. with S,,.. on the saturation gradient, on the relative permeabilities, and on the initial and boundary conditions of the problem [as well as on certain controllable and disposable variables
28-4
Fig. 2&l-Details
of steady-state
relative
permeability apparatus.
and Braun provide a comprehensive statement of how some people think the steady-state relative permeability method should be practiced when given the understandings and instrumentation opportunities available at the start of the 1980 s. Fig. 28.1 is a schematic ofthe laboratory system that can be used. Positive displacement or other types of constant-rate pumps, one for each fluid, discharge a fixed-ratio mixture into the core sample. Regardless of the initial saturation conditions within the core sample, the effluent fluid mixture eventually will be identical in composition to that being delivered by the pumps upstream. At this steady-state condition, effective permeabilities for the immiscible fluids can be calculated because the separate pump rates (4(, and q,,) will be known, and because an approximation of the pressure drop across the core sample can be used as the indicated driving force for each fluid. For example, suppose that the reservoir process under consideration involves edgewater encroachment into a uniform sand that has a certain level of interstitial water saturation, with the rest of the pore space filled with an unsaturated oil. A core sample representative of the formation could be selected and. after cleaning. an appropriate initial condition with respect to water and oil \aturat(on could be established by the so-called restoredstate capillaty pressure technique. Alternatively (and perhaps ideally), a fresh core sample could be brought into the laboratory as recovered by a pressure core barrel so that the wanted initial saturation conditions already would prevail. Or, as still another preparation
procedure, an as-received core could be processed so that the mud filtrate was replaced by simulated formation water that, thereafter, was displaced down to interstitial levels by flooding the sample with live oil. That is, in one way or another, the imbibition water/oil relative permeability data could be obtained where the experiment is started from a proper initial condition in fluid saturations and saturation distributions. And, if reservoir conditions of wettability were to be somehow preserved or restored as preparations for the experimental work are made, and if the ensuing displacement process were to be undertaken under conditions where reservoir-like overburden stress, pore fluid pressure, and/or temperature were simulated, then so much the better. However, the final saturation levels at steady state still would have to be measured. Several methods are available from which this information can be obtained.
Measurements. What the Braun and Blackwell method9 does is have a downstream water/oil separator where (as shown in Fig. 28.1) the bulk of each effluent phase is directed back to the inflow sides of the respective pumps, but where the differential amount between inflow and outflow of each phase is collected in a column and gravimetrically or volumetrically measured. In other schemes, X-ray absorbers or radioactive tracers can be added to one or more of the flowing fluids so that external instrument scans of the core will give the information that can be converted back to saturation levels. Even more simply, the core sample can be removed from the core holder after each steady-state condition is
Saturation
RELATIVE PERMEABILITY
28-5
DRAINAGE CURVE
Fig. 28.2-The Hassler sandwich. C denotes the core sample of length L and 6 denotes the inflow and outflow capillary end barriers. NW and W designate the ports through which the nonwetting and wetting fluids are directed collinearly as shown by the arrows.
reached, and then weighed so that saturation values can be calculated from independent knowledge of fluid densities and core sample PV. More details about the steady-state procedure under discussion are available in the references. Suffice it to say that entire curves of relative permeability vs. saturation are to be obtained. following well-defined imbibition or drainage paths, by proceeding stepwise from one steady-state condition to the next. For example, in the water influx experiment under discussion, the q,./q,, ratio delivered by the pumps could be set at succeedingly higher values while going from one steady state to the next until finally the high produced WOR s would confirm that a condition of residual oil saturation had been achieved. The previous discussion provides a brief description of the sense of the steady-state relative permeability measurement scheme as used when no special effort is made to control end effects (e.g., the buildup of wetting fluid saturation levels at surfaces of capillary discontinuity such as core end-faces). Early data obtained by equivalent steady-state procedures are reported in Muskat s classic work. lo Remarkably, there does not appear to be much learned from more recently obtained data than was at least qualitatively apparent in the beginning. For example, one conclusion that can be drawn consistently from laboratory observations is that the inequalities given previously as Eq. 4 can be written more exactly as O<k,.;<l forS;j<Si<l, . (6)
. \
0.2
0.4 SW
0.6
(0.8 E
1.0
Fig. 28.3-Schematic
where S;, is the minimum (irreducible) value of S, when the jth fluid of interest no longer has phase continuity over sensible distances within the pore space. On the other hand, the idea that relative permeability can be greater than unity for saturations less than unity has been reported. and explained as a reflection of the lubrication provided as one fluid slides by an adjacent immiscible one.
On the other hand, when attention needs to be given to end effects, some workers have proposed that long cores be used so that measurements could be confined to an inner portion. Even in such a so-called Penn State arrangement, however, the pressure-drop terms of Eq. 2 still are not measured separately for each of the immiscible fluids, and that is why it makes sense to refer back to Hassler s almost-forgotten work. The Hassler method for relative permeability determination has had a curious history. The patent was granted in 1944 and although cited occasionally, it has been ignored by most workers intent on developing simple state-of-the-art procedures. Scheidegge? referred to the method as superior, but difficult and time-consuming to apply, and at one point Rose I2 gave an analysis of why operational difficulties were to be expected in some applications. More recently, however, a patent has appeared that teaches how Hassler s ideas indeed might be reduced to a practical operating scheme. I3 Whether or not these claims eventually are substantiated by the facts, it is nonetheless of interest to examine in some detail the principle of the Hassler method for relative permeability determination. Thereby, a reference frame can be established to which other steady-state relative permeability procedures can be compared.
28-6
PETROLEUM ENGINEERING
HANDBOOK
0.6 kw k. 0.4
- 0.4
0.2
0.4 svl
0.6
0.6
1.0
Fig. 28.4-Schematic
l------
0.3
0.4
0.5 5,
0.6
0.7
0.8
Fig. 28.5-Hysteresis
in
waterflood
vs.
oilflood
relative
permeability curves.
Fig. 28.2 is a schematic that shows the sandwich arrangement where the wetting fluid (denoted by W) passes in series through an inflow capillary barrier as it enters the core sample, and then exits through a similar downstream barrier. The nonwetting fluid (denoted by NW) flows in parallel with the wetting fluid in the core sample but does not enter the pore space of the barriers. This exclusion is achieved by never letting the pressure in the nonwetting phase locally be so great that the barrier threshhold pressure is exceeded; hence, the endflow barriers will always remain 100% saturated with the wetting fluid during the course of the test. Fig. 28.3 is a schematic of a representative capillary pressure drainage and imbibition curve for the reservoir rock sample of interest where the relative permeability relationships are wanted. Suppose that, as an initial condition, the core (as well as the end barriers in series with it) is 100% saturated with water (say with a simulated oilfield brine). Let the threshold pressure for the core sample be PcA and that for the barrier material be greater than Pee. * Threshold pressure is being defined here as the lowest capillary pressure (P,., -PC.,,.), where the nonwetting fluid will enter a wetting-fluid-saturated porous medium. Since the capillary end barriers are made of material considerably less permeable than the core sample material, P,., $ P,, Hassler s way of performing the relative permeability experiment is to measure the effective permeability to each phase by using Eq. 2 at various successive capillary pressure conditions, such as PA, PSI, PSI PC, PDI, PD2. PE. The result is a succession of saturation states such as S,.,, , Ssl, SB2 SC,, SD,, SD2 SE in the sense that from A to C a drainage curve is being traced since the wetting-phase saturation is always decreasing, while from C to E an imbibition curve is being traced since the wetting-phase saturation is always increasing. Note that at A the wetting-phase saturation, S4, is 1.O, at C the wetting-phase saturation is close to the interstitial water level, S;,,, while at E the wettingphase saturation is (1 -S,,Z), where S,,, designates the residual (nonproducible) nonwetting fluid saturation level. Corresponding to the saturation points where capillary pressure values have been measured and shown in Fig. 28.3 during drainage and imbibition, the associated relative permeability data to be obtained by the Hassler method are shown in Fig. 28.4 in schematic format. It is instructive to draw some comparisons between these presentations. Capillary pressure as well as relative permeability functions are saturation-dependent. They both appear as parameters in the Buckley-Leverett description of multiphase flow and displacement processes (Eq. 5). Both functions are a direct reflection of the network structure of the pore space (size, shape, orientation, mode of branching, tortuosity, etc.). It is for these reasons that early workers were quick to search for the direct dependency between capillary pressure and relative permeability phenomena and that Hassler, with such great insight, was provoked to emphasize the interconnections that still earlier had been recognized by Leverett I5 and Richards.
*Pee IS the highest value of Ihe lowest value of S,
capillary pressure 10 be
10 ob,a,n
RELATIVE PERMEABILITY
28-7
The process under discussion is complicated and not easy to describe without many details. Since these are fully covered in the original references, J. it is enough to end the discussion here with the observation that there are at least two ways, in principle. to practice the Hassler method. One is to impose fixed boundary conditions in pressure upstream as well as downstream, and then to observe flow rates needed for the calculations of permeability data. * The other is to have constant flow rate as the upstream boundary condition while keeping constant pressure as the downstream boundary condition. What is gained thereby is circumvention of the practical difficulties of avoiding end effects encountered when the boundary conditions are set in terms of upstream and downstream pressure only.
OIL
ISOPERMS
GAS
/
YRTER /
\
' OIL
Unsteady-State
k, Method
methods to be described are suite straightforward The following analyses Fig. 28.6-Three-phase
rangingfrom saturation trajectory and oil isoperms 1x10- to2x10-3.
Compared to the unsteady-state now, the steads-state methods and involve few uncertainties. will reveal why this is so.
Procedure. In so-called unsteady-state procedures, effluent production from a core sample during the course of an imposed displacement process is recorded, and relative permeability functions are generated on the basis of a mathematical modeling of the process that is supposed to be consistent with what is being observed. In practice, the mathematical model usually selected is a simplified form of an integral of the Buckley-Leverett Eq. 5. By linearizing the equation by dropping the capillary pressure (end effect) terms, backcalculating the relative permeability functions appears to be possible. Since the aim is to get more resolution for calculation of intermediate values of relative permeability and for calculation of the associated saturation values, one needs to spread out the effluent production data. The whole enterprise is compromised, however. when a very unfavorable mobilit? ratio is chosen as the way to prolong the transient penod before total breakthrough of the displacing phase. The reader is referred to one analysis of this situation for further information. Methods of Calculation Institut Francais du Petrole Method. The reader also
is referred to the definitive paper by authors at the Institut FranGais du P&role I6 for full details about how to practice the unsteady-state method for relative permeability determination. Constant-flow-rate and constant-pressure schemes are described, precautions are enumerated, and calculation schemes are given. Fig. 28.5 is taken from another paper that shows how representative data and the associated relative permeability curves might look in a representative case.
Experimental
time, a back-calculation of relative permeability can be made (which, again, will be at most of qualitative value because of the limiting assumptions that have been used in the mathematical modeling). Claims made for the centrifuge technique are (1) it does not suffer from viscous fingering distortions, as in the case of the conventional unsteady-state procedures, and (2) it is faster than the competing steady-state procedures. However, the capillary end-effect problem still has to be faced. Another disadvantage is that in any given run, the information obtained applies only to the relative permeability of the invading phase. Since the conventional unsteady-state procedures give relative permeabilities for the displaced (as well as the displacing) fluids, and since larger samples can be processed than is possible in the centrifuge, the conventional approaches are the ones that are considered to be state-ofthe-art. In fact, the appeal that the centrifuge technique holds for experimenters is related to its suitability for automation, and to the fact that it is already generating three-phase data. Fig. 28.6, for example, is a display of some of the data that have been reported.
Critique of Methods
To summarize what has been published about relative permeability measurement methodologies, the critique originally given by Scheidegger, 2 and more recently affirmed, * is accepted enthusiastically here. The consensus seems to be that the steady-state methods. Hassler s in particular, give the most believable results since they are based on a plausible (however naive) definition of relative permeability (i.e., as embedded in Eq. 2). Unsteady-state methods, while quicker and easier to apply from the laboratory operational point of view, nonetheless are beset with enormous interpretation difficulties. This is mostly because they involve drawing inferences (by way of back-calculations) from a roughintegral form of the Buckley-Leverett Eq. 5. As noted previously, Eq. 5 is at most an imperfect model of the flow process under study.
discussion
Automated Centrifuge Technique. To finish with the of the variants of the unsteady-state methodologies, it will be useful to cite the work of
O Meara and Lease, I8 who gave three-phase data obtained with an automated centrifuge technique. By spinning core samples in a centrifugal field of known strength and observing effluent volumes as a function of
PETROLEUM
ENGINEERING
HANDBOOK
0 i
0 P *CO (Hellurn
,THEORETICAL
CURVE
2 SL,-
.4
WETTING
LIOUID
SATURATION (fracl~onal)
permeability
Fig. 28.8-Matching
Other k, Methods
Still other methods to arrive at relative permeability information remain to be cited for completeness. For example, there are the so-called stationary fluid methods I9 and the calculation methods based on capillary pressure and endpoint displacement data associated with Corey et al. *O and Stone.2 Figs. 28.7 and 28.8 display some of the results given by these investigators. In the stationary fluid methods, the effective permeability of one phase is measured by flowing that fluid at such a low pressure gradient that the contiguous immiscible fluid is left (it is hoped) unaffected. Clearly, the method can be trusted most when the stationary fluid has a saturation level close to its minimum (irreducible) value (i.e., Sj-+SV for the stationary fluid). Otherwise, it may be concluded that the data generated by applying the method will be distorted by the ever-present end effects. On the other hand, the method is easy to apply operationally, and the data generated giving k, at S;,, (irreducible water saturation) and k,. at S;, (irreducible oil saturation) are needed along with other parameters when calculated values of relative permeability are to be made by the popular Corey et al. and Stone methodologies. In any case, values for k, at Sit,. give an indication of the initial productivity of oil wells from horizons where there is only initial oil plus interstitial water saturation. Similarly, values for k, at Si, can be used to indicate the level of the effective water/brine permeabilities when residual oil saturation conditions are reached at the end of a waterflood recovery process. To understand better the rationale for using relative permeability calculation schemes, reference can be made to the early papers where they were first advocated. 14~ 2 All along, the idea underlying them has been that petrophysical properties (e.g., relative permeability and capillary pressure relationships, specific surface area,
electrical resistivity parameters, etc.) depend in one way or another on the nature of the pore structure, and on how it will be partitioned in representative cases between the intertwining filaments of the adjacent immiscible fluid phases. Of the pore textural properties enumerated previously, relative permeability turns out to be the most difficult one to measure in the laboratory. This has been the origin of the thought (and hope) that an economy could be expected if relative permeability information somehow could be extracted (by way of calculations) from the more easily obtainable data. Wyllie, a foremost advocate of these ways of thinking, is the authority to be consulted when more information is
needed.
Eqs. 7a and 7b are the ones given by Corey et al. 2o for calculating k, under three-phase saturation conditions, and Eq. 8 is the one proposed by Stone. I For SL > SLrr k,.~~=(1-S,-S,V)3(1-SS,+S,,,-2SL,)/(l-SLr)4. . . . . . . . . . . . . . ..I........... and for SL 2 SLr k,=(1-S,-S,.)/(l-SL,)4. Also, .. . .(7b) (74
km =(k,w +k,)(k,,
+k,)-(km,
+k,),
. (8)
where SL = total liquid saturation, fraction, SLr = residual liquid saturation left in pore space, fraction, k,,. = relative permeability to oil in a gas-free system, and k i-q = relative permeability to oil in a water-free system.
RELATIVE PERMEABILITY
28-9
-I
- LEGEND I .RUNI
SW 4 9.2
2 : .
h
5 6
7
2 E ; .Ol,
w 2 5 d E
.OOl
40 20 OIL SATURATION,
60 PERCENT
t 1
three-phase relative mobility Fig. 28.10-Unsteady-state. (relative permeability to oil, Berea sandstone).
The recommendation for using Eqs. 7 and 8 to obtain calculated values for the difficult-to-measure k,,, under three-phase saturation conditions is that one only needs to enter more easily obtained two-phase data. Inspection will show that while such equations give plausible results (e.g.. compatible with the indications of Eq. 6), the results are no more trustworthy than the modeling assumptions when measured relative permeability values are needed but are otherwise unavailable. to be drawn from this analysis is that if there is any merit to taking short-cuts, it is only when the benefits outweigh the risks. While many (perhaps most) of the readers of this chapter will never have to go to the laboratory to make relative permeability measurements, it is important for them to realize that someone, some day, has to make such measurements before any of the calculation schemes can be used with confidence. In other words, computer guesswork in general will be no substitute for laboratory work. This is because direct observations always will be needed before reliable predictions can be made for processes that inherently depend on material response parameters such as relative permeability.
7
80 60
Comments. A conclusion
0.2 t
cm
0 I
Recent Literature
More than 30 years have elapsed between an early and a later time when the author of this chapter was provoked to write papers on the then-current problems of relative permeability measurement. X It is as though some of the problems have continued to be unsolved and/or unsolvable, while other vexing ones have emerged to take the place of any of the earlier ones that somehow were resolved. In the meantime. work continues on many fronts such as those critiqued in this paper. Note that all the papers now to be referenced have appeared after
0.11
IO
20
GAS SATURATION.
30 %
40
28-10
PETROLEUM ENGINEERING
HANDBOOK
CONTACT ANGLE -
--
leo0 410
20
40
80
1Ul
Fig. 28.12--Relative permeability to oil and water showing the effect of displacement (Curve 1) vs. carbonated water displacement (Curve 2)
Wyllie wrote the original chapter on Relative in the first edition of this handbook (in Permeability 1962) referencing the literature through the 1950 s. z3
20
40
60
80
1
0 10
WATERSATURATION,PERCENT POREVOLUME Fig. 28.13-A comparison of restored-state and native-state water/gas relative permeability data.
RELATIVE PERMEABILITY
28-11
kro
Fig. 28.15-Gas/oil relative permeability calculated from production data from experiments.
permeability while Schneider and Owens3 and Owens and Archer34 were concluding that because of wettability effects, experimenters were advised to make use of socalled native-state cores. The work of Bardon and Longeron- also dealt with this subject. And somewhat later Sigmund and MacCaffery36 were trying to sort out the influence of reservoir rock heterogeneity on relative permeability characteristics. Some of the data presented by these authors are shown in Figs. 28.13, 28.14, and 28.15. In the 1980 s, numerous relative permeability papers continue to appear as though a renascent interest in the old subject is developing. A noteworthy one was by Hagoort, 37 who used a centrifuge technique to show the high efficiency of the gravity drainage recovery process inin water-wet cores. Bogdanov and Markhasin* troduced the less familiar subject that speculated that viscosity changes (because of molecular-surface interactions with the rock matrix) could distort relative permeability data. At the same time, Ashford,j9 in a very comprehensive paper, was reopening the dubious issue of how relative permeability and capillary pressure data can be linked directly. (As implied above, it is nice to have a conceptual theory available to explain relative permeability effects, but to expect that calculations avoid the need for careful experimental work is a kind of wishful thinking that can be justified only when there is a great urgency to have qualitative inputs for reservoir process simulations). Fig. 28.16 is representative of the fits reported by Ashford between calculated and measured values. Delshad et al. currently have addressed the interesting question of whether the transport of low-tension micellar fluids will significantly change classical relative permeability trends. They show that the residual (endpoint) saturations decrease and relative permeabilities increase as interfacial tension decreases. The predicted nearly 45 relative permeability curves are shown for one case in Fig. 28.17. Yokoyama4 has been dealing
Fig. 28.16~-Relative permeability relations for imbibition and drainage experiments, with values calculated by Naar et al. method.
Fig. 28.17~lmbibition
relative permeability curves for microemulsion and decane vs. oil saturation in Berea sandstone.
28-12
PETROLEUM ENGINEERING
HANDBOOK
no I-. 1
-*
/-
a2
Fig. 28.19-Comparison of drainage relative permeabilities from randomized network model with experimental data for 100 to 200 mesh sand.
with the equally complex problem of accounting for transverse and longitudinal capillary imbibition during displacements in stratified media. In another direction, Carlson4* extends Land s4 earlier prescriptions for calculating flow from independent measurements of rock properties. Chierici4 does the same thing, on the basis of the bundle of capillary tubes model described earlier by Brooks and Corey. 45 Some of these data are shown in Fig. 28.18. Finally, Lin and Slattety46 and Mohanty and Salterj carry pore structure (network) modeling as a basis for arriving at calculated relative permeabilities to still higher plateaus of sophistication. Extensive bibliographies are provided by these latter authors. Some of their results are given in Figs. 28.19 and 28.20. For example, Fig. 28.20 indicates that (1) relative permeabilities to the nonwetting phase k, during secondary imbibition (SI) and imbibition (IM) are essentially the same but lower than the values that refer to the primary drainage (PD) conditions, and (2) relative permeabilities to the wetting phase k, during primary drainage is lower than that in either seconda drainage or imbibition. In other words, those authors conclude that the ratio of conductivity or nooks and crannies to that of a full throat feature influences the wetting fluid permeabilities only at low saturation. Other recent papers are those of Salter and Moharty,48 who made observations to justify a modeling of multiphase flow that postulates flowing, dendritic, and isolated configurations for each phase, and of Maini and Batycky, 49 who claim (see Fig. 28.21) that temperature influences both the endpoint saturation and the shape of the relative permeability curves. A different view about the importance of temperature effects had been expressed earlier by Sufi et al. 5o
The paper of O Meara and Lease I8 has been cited previously in connection with the unsteady-state determination of three-phase relative permeabilities using the centrifuge technique. It, along with the Maini and Batycky paper, 49 are just two of nine other papers on relative permeability presented at the 1983 Society of Petroleum Engineers Annual Technical Conference and Exhibition in San Francisco, Oct. 5-8. Listing them by topic is one way to expose the depth and breadth of current interests in the subject. Thus, Kortekass describes displacement in cross-bedded reservoirs. Meads and Bassiouni5 speak of combining production history and petrophysical correlations to enhance the representativeness of relative permeability data. Miller and effects for Ramey 53 deal further with temperature oil/water systems. Mohanty and Salters4 extend their work on oil mobilization, transverse dispersion, and wettability effects. Fulcher et al. and Harberts6 deal further with low interfacial systems. Heiba er al. 57 address uestions. the wettability And in companion papers, Heaviside et al. 51 cover the experimental and theoretical aspects of relative permeability phenomena. Such are the extensive details currently being discussed.
RELATIVE PERMEABILITY
ble fluids of differing densities, and counterflow imbibition. (3) reservoir rock anisotropy and heterogeneity by which sample size and sampling frequency requirements are to be determined, (4) fines movements. (5) overburden stress simulation and related stress-relaxation and creep compaction phenomena, (6) viscous drag at interstitial interfaces between contiguous immiscible fluids, (7) chemical precipitation and dissolution phenomena, (8) chemical reactions, (9) high Reynoldsnumber conditions (nonlinear laminar and turbulent flow regimes) (10) Klinkenberg gas-slippage effects. (I 1) concentration and/or thermal gradients being superimposed on fields of flow primarily caused by mechanical energy gradients, (12) non-Newtonian rheology, (13) systems characterized by more than one local pore space type, (14) fluid/solid interactions, for example, as related to the mineralogy of interstitial clays, (15) viscous fingering, and (16) hysteresis related to wettability changes. While theoretical considerations may permit a qualitative prediction of the nature of some of these effects, in the final analysis, any truly quantitative assessments should be based on directly undertaken experimental work, if possible. This is not to say that in every case the laboratory data obtained on small hand specimens will reveal everything that needs to be known about large composite petroleum reservoirs, but rather that observations generally are more trustworthy indicators than are blind guesses.
50
100
NONWETTING
PHASE
SATURATION
Conclusions
Several questions come to mind whenever the subject of relative permeability is raised: how is it to be defined ? where can the information be obtained? why is it needed and by whom? What are the proper (fruitful) ways to use it? Only partial answers to some of these questions have been given in this chapter. This is because the subject is too vast to be dealt with fully in limited space and because the details not covered are probably too specialized for the average reader, whose concern is only with the ordinary applications. Here we have asserted that definitions of permeability (specific. effective, relative) are embedded in the differential equations that describe the transport equations governing fluid flow in petroleum reservoirs. Eqs. I and 5 are examples that apply to special situations (steady single-phase flow and unsteady multiphase flow of immiscible fluids, respectively). Obviously, the differential equations of transport will have other forms for more general cases of fluid flow, such as those observed when there is coupling with chemical diffusion and/or heat transfer processes. This is to say that there may be different kinds of relative permeability to deal with according to the nature of the process under consideration. In all cases. however, one must start by constructing the defining equations to be compatible with the underlying principles of nonequilibrium thermodynamics. Relative permeability information can be obtained in two major ways. The preferred method is to have an cxperiment performed on a representative sample of the reservoir rock according to the procedure prescribed by the appropriate integral form of the defining differential equation. Eqs. 1 and 2 are examples of integrals that involve measurable terms (such as volumetric flow rates
Fig. 28.20~-Relative
permeability curves.
RUN: I1
TEMPERARATURE: i02' C
WATFP SATIJPATIZN
28-14
PETROLEUM ENGINEERING
HANDBOOK
and pressures. pi, at bounding sample surfaces). Since the parent differential equations themselves have terms that cannot be directly measured in the laboratory (such as velocity and mechanical energy gradients), integral forms that apply to the particular initial and boundary conditions of the problem must be derived and used for each specific case. In these connections. it will be revealing (and disturbing) to mention that the degree of equivalence of relative permeability functions obtained by different methods (e.g., the steady- vs. the unsteady-state schemes mentioned previously) so far has not been established fully. Eq. 2 prescribes how the steady-state results are to be obtained, while integrals of Eq. 5 prescribe how the unsteady-state results are to be obtained. Until the definitive laboratory work is done and the comparisons are reported in the literature, users of relative permeability information today will continue to be left in the dark as to which procedure can be used with the most confidence when applications are being made. As for ordinary applications, relative permeability information is used by petroleum reservoir engineers when interpretations and assessments are being made about the outcomes of observed and probable petroleum recovery processes. These applications are referenced in other chapters in this handbook (e.g., on reservoir simulation. well testing analysis. etc.). and will not be discussed further here. Suffice it to say that the same governing transport equations and their integrals by which relative permeability is defined, necessarily are used again in the ensuing analyses of reservoir performance; hence, relative permeability will be needed as input data whenever analytical studies are undertaken. Among other things, this means that using drainage relative permeability data to describe an imbibition process (such as waterflooding) should be avoided. A second way to obtain relative permeability information is to develop suitable models of the processes under consideration for use as calculation algorithms. Eqs. 7 and 8 are examples that in use require laboratory data more easily obtained than the relative permeability functions themselves. For example. to calculate three-phase oil relative permeability as a function of oil saturation (where S,, = I -S,,.-S,) by Eq. 7, all that is needed is prior knowledge of the interstitial water saturation. Naturally, such calculation schemes cannot be used blindly except for cases where it has been independently validated that calculated and experimental values are equivalent. In other words, even the most carefully constructed calculation scheme does not circumvent the need to have experimental measurement methods also developed and available. calculation schemes usually involve However, analytical functions that can be entered directly into the computer software used for such things as qualitative economic forecasting. Similarly, the related way to obtain relative permeability information, namely by inferring it from the values that force history matches of observed field data, also has a utility when qualitative assessments are being made using reservoir simulators. The points being made here are both subtle and selfevident. The aim has been simply to convey a certain set of useful ideas to users of relative permeability information such as (I) that relative permeability methodologies
are still in the developmental state even after more than a half century of well-intentioned labor by hundreds of workers and (2) that users of relative permeability information must put the burden of proof on those that supply it. to demonstrate that credible schemes have been used.
Acknowledgment
Any author who has been writing narrowly on a specialized subject like relative permeability for more than a third of a century (specifically from 1948 to 1984l) has to be deeply indebted to all those inspiring workers who have managed to keep the issues alive and in focus over such a long period. A few of them, but not all, are named in the abbreviated references that appear here. Likewise, any teacher for a similarly long period must be deeply indebted to the generations of inquisitive students who were willing to ask the provocative questions before they graduated and left to disappear into oblivion. What a lucky thing it was to have had so many companions on what otherwise would have been a dreary Orwellian journey.
Nomenclature
A= Af = k= k, = k,.i = k ,-oh = km,,. = k,.,,. = K ,11< = L= L/l = M,, = M,,, = P, = Ap = P,. = 4= s, = s,l II Sin = Sj = SL = s1.r =
SLU srr, s,,. = = =
s wr = I= J p, = =
Ap = c=
cross-sectional area cross-sectional area of flow path specific permeability effective permeability, Fluid j (gas. oil. or water) relative permeability, Fluid j (gas, oil, or water) relative permeability to oil in water-free system relative permeability to oil in gas-free system relative permeability of wetting fluid microemulsion dispersion coefficient length horizontal distance mobility of nonwetting phase mobility of wetting phase pressure, Fluid j pressure drop capillary pressure volumetric flow rate residual gas saturation irreducible (minimum) value of S.; irreducible (interstitial) water saturation fractional saturation of jth fluid total liquid saturation residual liquid saturation wetting liquid saturation residual (nonproducible) nonwetting fluid saturation water saturation residual water saturation time volume of Fluid j total PV pressure drop fluid viscosity
RELATIVE
PERMEABILITY
28-15
(T = inter-facial 6 = porosity
Subscripts
g = gas
tension
28.
Sawn, bv
A.M.:
Relative
Permeability
Measurements 1966)
Unsteady-State
Snr.
199-205, 29. Saraf, Measurement Estimating 235-42; 30 31. Balint, ment Phase (1973) 32. Lefebvre Relative V.
T&s.,
D.N. and Using Fluid Trans., et nl.: V.G.
Sot
Permeability Koola~
j = n = 0 = M=
es Fold& of Carbon
Panteleev,
et al.:
by 11-13. Prey,
Water,
References
I. Bear. J .: Dw~tmi~s
Publishing Co. Inc.,
E.J.: 1973)
Permeabilities
of a Consolidated
Sot.
ofFluids
New Phwics Capillary Pitx\ics
and
Owens, Flow
2. Scheidegger,
edmon.
A.E.: L A.:
Following
U. of Toronto
Press.
Tmns..
Wettabili-
3 Richards.
Porous
Effect
of Rock
4. Odeh.
ty, 5. Yuster. Idealized The 6. 7.
Ratro 346-52.
Permeability
J. Per.
Inter-
Tiwis.. ST.
Tech. (July
35. Bardon. facial 1980) 36 C.
Trans., AIME,
D.G.: Influence Permeability, F.G.: Porous
Theoretical Systems,
Consideration
Pmt.,
M.C.: AIME Connected
Hague
Sigmund, State
McCaffery, for
in Sandh. Problems
107716.
Determining
Characteristics
Descripttons ml,\
Displacement
J. (Feb.
Hagoott, J. 3x (June
in Porous Me&r.
(1972) 2.345,935 A Permeability presented 229-40. No. R.J.:
by Gravity
Publishmg x. 9. Has\ler. Braun. Measuring Condittons. Technical IO. II. Muskat. Book Oaoba. Rose. Rose. Rose. Co.
Amsterdam Patent
11944). Technique for Curves at the at Reservoir 1981 Oct. Annual 4-7.
I.L.:
Determinatton for the Change with No. Rock. IO. 57-60. Phase Imbibition and
Steady-State
to Oil
to Molecular F.E.:
Surface
10155
U&b.
39 Ashford, Relative on 40
Conference M.: J.S. W.. Inc.. rf Some Method. U.S. Patent and Bmce. Rock. M.C.: (1940) New cl/.: Trctns.,
Multiphase Based
fh~sic~~/ Prin+lc.s
Laboratory AIME
Cycles
Capillary
(1981) Delshad,
1, 77-94.
Experiments,
(1951) m
Problems No.
Applymg (July
Per.
The
Eng.
J.
(Aug
Permeability W.: W.
J. Per. Twb.
4..506..542 Evaluation AIME W.A.:
1161-63. Character
Y.
and SPE
Effects in 1981
of SPE Oct.
Immiscible
Displacements
Stratified
Trcm.. Capillary
(1949)
at the
Behavior
m Porous by
Technical 42 Carlson,
Conference F.M.:
142. of
Simulation Phase,
Relative SPE
10157
R~WP de / /nrri/ui
and Rosrelle. W.O.:
Fmquis
Graphical From
du Pdrdr
Techniques Displacement Trcrhs..
( 1973)
for ExAIME, 44 43
Technical
Conference
and
Exhibition.
San Antonio.
5. 695-714.
17.
Land, Two-
Relative
Permeability Sx.
for
Determimng periment\. 26.5. IX. O Meara. Pemteabillty paper 19. 20. 21. Rose. Corey. Tw/r. Stone, Rslativc 22. Rose. W.: SPE W.: A.T. (Nov. H.L.: Conference ~rrrd Pr,x/
Relative
Permeability 1978)
Prr.
807-17:
Eng. J. (June
Chierici. Relative 45 Brooks. State 46 Lin, U. C. G.L.: R.H
Trans.. AIME.
for
Drainage
Permeabilities.
Per. Enx.
Hydraultc
J. (June
Paper
1984)
Measurements 12128 and Pi-w.. 01 trl.: 1956) presented and Dallas Exhibition. API,
(1964). and Slattery. for and Annual 26-29. Mohanty. at the New and Core ci of 1982 1982 K.K.: SPE Scpt J.P.: Multiphaw and Annual 26-29. Effect of Temperature 1985) on SPE on Flow Technical in Porous SPE Observations Modelmg. paper J.C.: Two-Phase No. Salter, Technical S.J.: Three-Drmenaional. Flow Through Randomized, Porous Flow I1018 Media. in Porous at Model K.K. II.
Francisco.
Permeabtlity
(1948)
127-3.5.
AIChE J , 28 (1982)
Pore-Level
Modelmg,
Probability
Esttmating (Oct.
and Exhibition.
New
Permeability, Theoretical Perme;ibility. Relattve SPE. Some Thtrd Formatmn Regronal and
J. Cdz. Pri.
Generalizations
Twh.
Leading (1949) 2,
and
of Relative 23. 24. Wyllre. Hrrnrllxwk. Rose. ment. 2s Rorc. Sample Rocky 26. Loomis. &r/i. 27. Naar. wltdated ~iwt.r. 599. J. W. W..
I. Macroscoprc presented
M.R.J.:
Prrrol~ut~~ Prdwrion
I- I4 Measure2. 446-59. and 1983 Fluid SPE 50. 22-25. 49. Permeability t 1951) Reservoir Lake Set Rock at the City. of
Conference
Orleans,
Prm
Heavy-Oil/Water
May
Inveatigattons. 51.
Conference
Bureau
of Mmes
Orleans. Reservoir
Sept.
Relative
Permeability
in 1962)
Uncon1317:
Water/O11 Zones.
Porou\ . AlME.
J. (Dec.
28-16
PETROLEUM
ENGINEERING
HANDBOOK
52.
~xI
Z.:
Production More
Hl\tq
57.
Heiba.
A.A.
Pconcahdmes presented hlhitmn. Hcawside. Relative tion\~ Techmcal at the San J.,
I983 Black.
Contcrence
Conlcrcnce
Francwo.
Oct.
S-8. and Ramey. Relative Sot,. Salter. H.J. / pi. S.J.: Jr.: 0,~. El l ect of J. (Dec. or Temperature Unconsolidated 190) Flow 945-54. I Porw\ and on and
J.F.:
Pem?rabilitics
Sands. and
Conference
San
Francwo.
Multiphase Tranhvew
3-011 paper
Dlspenmn.
Wettahility. Technical 55. Fulchcr, ~ucnts Tcih. 56. Harben. ty. Techmcal on (Feb. R.A.
presented
at the Number
Exhihltion. ot Cupillaly
San
Franasco. Cunes.
General References
Relative 249-60.
Pem~cab~l~ty
J. / PI
PcnneahdiWhitaker. S.: Flow in Porous Media II: The Governing Equations t or
SPE Oct.
Annual 5-n.
Immi~cihle, 1. 105-25.
Two-Phase
Flow.
( 1986)
Conterence
Chapter 29
Introduction
A reservoir trap is a combination of physical conditions that will cause hydrocarbon liquids and/or gases and water to accumulate in porous and permeable rock and prevent them from escaping either laterally or vertically because of differences in specific gravity, pressure, fluid/gas characteristics, and/or lithology. It has the capability of collecting, holding, and yielding hydrocarbon fluids and water. The portion of the trap that contains oil and/or gas accumulations is the petroleum reservoir. It generally occupies a limited portion of the trap capacity, the remainder being occupied by formation waters that underlie and are interspersed within the petroleum accumulation. Traps are formed by an infinite variety of structural and stratigraphic conditions of rock formations combined with pressure differentials among the various fluids within the reservoir rock. A trap consists of an impervious cover or roof rock overlying a porous and permeable rock. Reservoir pressure gradients and fluid flow within the reservoir rock can create traps that do not have structural closure. The boundary between oil and water or between gas and water need not be flat or level when these pressure gradients are present. Generally. however, traps do have structural closure, and as viewed from below, the impervious cover is concave, preventing the oil and gas, if present. from escaping vertically or laterally. The water underlying the oil and gas exerts a buoyant force on the oil/water boundary or contact, lifting and holding the oil and gas to the crest of the structure or area of minimum hydrostatic pressure. made by geologists attempting to include all factors and conditions that account for petroleum reservoirs. Many reservoirs have unique features that cause the oil to accumulate at a given location. The purpose of this chapter is to illustrate the more common geological conditions that cause traps and to point out a few of the infinite variety of minor variations that help create and hold petroleum accumulations in place. Structural Traps Structure implies some form of rock deformation, commonly expressed as a positive uplift, which may result in four-way dip closure. With the proper stratigraphy, structural traps may be present. Domes, anticlines, and folds are common structures. Fault-related features also may be classified as structural traps if closure is present. Structural traps are the easiest to locate by surface and subsurface geological and geophysical studies. They are the most numerous among traps and have received a greater amount of attention in the search for oil than all other types of traps. In new areas of exploration the prime search is for potential reservoir rock, source beds for hydrocarbons, and structural deformation. This structural deformation provides opportunities for several types of structural traps. Domes, Anticlines, and Folds. Domes, anticlines, and folds in general must have structural closure to become effective traps. The reservoir rock must dip away in all directions from the crest of the structure. If there is not dip in all directions away from the crest but hydrocarbons are present, there are other contnbuting physical factors that helped complete the trap. Domes, anticlines, and folds caused by structural deformation of sedimentary rocks generally create many potential traps because the deformation extends vertically through potential reservoirs. Thus a single well can reveal many possible pay zones when drilled on the crest of a domal structure.
Trap Classification
Classification of traps logically falls into three broad general groups: (I) structural, (2) stratigraphic, and (3) combination. More detailed classifications have been
29-2
PETROLEUM ENGINEERING
HANDBOOK
KREYENHAGEN HILLS
KETTLEMAN PLAINS
KETHiXl-;AN
SANvJL$+J N
TEMBLOR
KREYENHAGEN
CROSS SECTION
Fig. 29.1-Example
L*IN 8Y 5NdO
-. - . - ..-- -, INPLVXEHE
1FII rAtr
3.000
SANTA
FE SPRINGS
FIELD, CALIFORNIA
Fig. 29.3-Example of anticlinal folds creating many separate reservoirs: Santa Fe Springs field.
Figs. 29.1, 29.2, and 29.3 are cross sections of Kettleman Hills, Midway, and Santa Fe Springs fields, CA. t These are examples of single folds creating many separate accumulations. The separation between the various reservoirs is demonstrated in each case by different oil/water and gas/oil contacts in most reservoirs. The Midway field also illustrates stratigraphic traps formed on the flanks of the anticlinal fold. Folds, anticlines, and domes arc the easiest to interpret in subsurface studies. They vary in size from a few acres to several thousand acres. Folds and anticlines were created by compressional or tensional forces in the earth s crust or by differential compaction of the sediments. Asymmetrical anticlines, overturned anticlines, thrust faulting, and fracturing generally indicate areas of compression. Symmetrical folds and anticlines, low-angle normal faulting, monoclines, homoclines, and low-relief domal structures generally indicate areas of tensional forces or compaction. Mountainous areas usually result from compressional forces. Torsion and shearing help cause local complex structures but are generally forces resulting from the more regional compressional forces of the earth s crust. Stable areas or areas of subsidence ate the counterpart to mountain-building compressional areas. They are areas where structures arc caused by differential
29-3
m
FAULTOF REGIONALDIP
THRUSTWT;EVERSE
downwarping, as in the midcontinent area of the U.S. and areas where structures are created by the lengthening of the earth s crust, such as the gulf coast of Texas and Louisiana. This lengthening causes regional horizontal tensional forces that create simple and more predictable local structures. The many prolific structural and stratigraphic oil trends that parallel the U.S. gulf coast today are the result of regional downwarping and tensional forces. Fault Traps. Fault traps are classified as structural traps where closure is effected in one or more directions by faulting or where faulting has caused definite changes in the reservoir configuration (such as along strike-slip faults). Many structures are faulted without being limited by the faults or without changing the reservoir configuration. Fault traps can occur in both the up- and downthrown blocks. Closure against the fault can result from faults striking across regional dip or across anticlines or domes. Horsts and grabens and other closed fault blocks can result in traps with relatively no structural closure. Normal or gravity faults (Fig. 29.4) occur as a result of tensional or gravitational forces. The angle of the fault plane with the horizontal generally ranges between 25 and 60. Normal faults involve horizontal lengthening of the earth s crust and are recognized in the subsurface by loss of stratigraphic section in wells drilled through the fault plane. Geophysically, they are recognized by interruptions in the continuity of reflective interfaces. Two common types of normal fault-related traps are: (1) fault closures and (2) rollover fault closures. Any structural nosing cut at right angles by a fault results in a fault closure. The direction of throw on the fault is not important but the closure created by the fault is. For example, a south-plunging structural nose cut at right angles by a fault will result in a potential trap. The fault throw may be in either direction. A trap will result if the fault acts as a seal or if the potential reservoir is thrown against a shale or other impermeable member on the opposite side of the fault. Gravity-type faulting commonly occurs in areas of tension and over the crests of domes and anticlines because of the stresses involved. Fault traps are common in such an environment, and hydrocarbon accumulations may occur on either the up- or downthrown blocks, in horsts, and/or in grabens. Rollover fault closures are common in sedimentarv basins receiving great quantities of sediments. Closure $ created on the downthrown block by contemporaneous sedimentation and fault movement. Through this interaction, more deposition takes place next to the active fault plane, resulting in a downbending of the deposits into
the fault. This downbending creates a reversal in dip and this results in closure. This type of trap is extremely common in the Cenozoic formations of the U.S. gulf coast. Reverse or thrustfaults (Fig. 29.5) result from compressional forces and involve horizontal shortening of the earth s crust. The angle of the fault plane with a horizontal plane can vary from a few degrees to 90 and can be recognized in the subsurface by repetition of stratigraphic section in wells drilled through the fault plane. Structural traps of this nature are common on both the east and west coasts of North America. The occurrence of a trap against a fault depends on the fault plane sealing the porous reservoir rock and preventing migration across or along the fault plane. Fractured formations usually are caused by local deformation, faulting and folding, reduction in overburden permitting expansion of the underlying rock, and differential compaction. Brittle rocks are more commonly affected because of their inelasticity. In many cases minor joints, fractures, and fissures are modified by solution and combine with primary and secondary porosity to give a greater effective reservoir porosity and permeability. Fractures in reservoirs increase the wellbore radius and permit extremely tight and impermeable areas to bleed into the fractures over a wide area and thus be connected with channels leading to the wellbore. Production is sometimes obtained from igneous and metamorphosed rock as a result of fracturing. The fractures provide the reservoir space as well as the permeability to permit oil and gas migration, accumulation, and production from the reservoir. For a trap to occur in a fractured formation, it must be overlain by a more pliable or less brittle rock that has not been fractured by the deformation. Otherwise, migration would occur up through the fractures and there would be no trap. Where faulting caused the fracturing, production is limited to a narrow band along the fault. When folding or other deformation has caused the fracturing, the reservoir can become very complex in shape and unpredictable in production performance. Generally, the areas of greatest deformation have the greater number of fractures, which results in better well performance and recovery of more oil or gas. Stratigraphic Traps Traps created by changes in stratigtaphy have the same physical requirements as structural traps. There is an updip limitation or termination of the reservoir rock, creating an area of minimum hydrodynamic potential or .concave closure. In case of structural limitations, this is
PETROLEUM ENGINEERING
HANDBOOK
Stratigraphic traps can be classified as primary or secondary. Primary traps are those formed during sedimentary deposition: lenses, facies changes, shoestring sands, offshore sandbars, reefs, and detrital limestone or dolomite reservoirs can be classified as primary. Secondary traps arc those resulting from later causes such as solution, cementation, erosion, fracturing, and chemical alteration or replacement. Primary Stratigraphic Traps. These traps result from deposition of elastic or chemical materials. Shoestring sands, lenses, sand patches, bars, channel fillings, facies changes, strand-line (shoreline) deposits, coquinas, and weathered or reworked igneous materials are classified as elastic sedimentary deposits and can result in stratigraphic traps. Fig. 29.6 is a structural/stratigraphic interpretation of the northwest Atkinson field in Live Oak County, TX. 2 An ancient sand-filled stream channel meander has cut into older south-dipping shales and created a perfect stratigraphic trap. Fig. 29.7 is a cross section3 across the Yoakum Channel in Lavaca County, TX. This is an example of a channel filled with shale. The shale plug served as the seal for reservoirs within a west-plunging structural nose. Hydrocarbons are trapped in the truncated updip portions of the reservoirs. Organic reefs or biohenns and biostromes are the primary chemical stratigmphic traps; they are built by organisms and are foreign bodies to the surrounding deposits. A cross section of Scurry field in Scurry County, TX (Fig. 29.8) gives an example of a primary chemical stratigmphic trap.4 The Strawn and CiscoCanyon series are limestone reefs that have had younger sediments deposited on the flanks and eventually over the crest of the reef deposits. The shale serves as the seal. Differential compaction of the thicker shales on the
interpretation; Northwest
obtained by faulting or a turnover of the reservoir rock. In stratigraphic traps, this limitation is accomplished by changes in porosity and permeability, which result from erosion and overlap, facies, and nondeposition, lithological changes caused by depositional variations, truncation, and differential compaction.
115J164V
CHANN
FILL
LOG
SECTION
A-A
MIDDLE
WILCOX
CHAN NEL
Fig. 29.7-Cross
29-5
Fig. 29.8-Example
flanks of the reef as compared with the thinner shale at the crest has created structural closure in younger overlying formations. Hydrocarbon accumulations have occurred in the Cisco and Fuller formations as a result of this differential compaction. Additional traps in other reservoirs arc the result of updip permeability and porosity barriers and are either primary or secondary stratigraphic traps. Secondary Stratigraphic Traps. Traps of this type were formed after the deposition of the reservoir rock by erosion and/or alteration of a portion of the reservoir rock through solution or chemical replacement. Secondary stratigmphic traps actually should fall into the combination-trap classification because most are associated with or are the result of structural relief during some stage of development of porosity and permeability or limitation of the reservoir rock. However, many of the so-called typical stratigraphic traps fall into this category, and it is felt that it would be impossible to change the historical usage of this term. Therefore, secondary stratigraphic traps are defined for this discussion as those traps created after deposition and having limitations caused by lithology changes. Erosion creates a major part of these through truncation of the reservoir rock. On-lap deposition (when the water is encroaching landward), off-lap deposition (when the water is regressing), and the chemical alteration of limestone result in many secondary stratigraphic traps. The East Texas field (Figs. 29.9 and 29.10) is perhaps the most famous field in this classification. It is a truncation of the Woodbine formation as it approaches the regional Sabine uplift. 5 A certain amount of leaching of the cementing material by waters over the unconformity
has resulted in increased porosity and permeability in the field as compared with similar Woodbine sands in the deeper portions of the East Texas basin. Combination Traps Combination traps are structural closures or deformations in which the reservoir rock covers only part of the structure. Both structural and stratigraphic changes are essential to the creation of this type of trap. Traps of this nature are dependent on stratigraphic changes to limit permeability and structure to create closure and complete the trap. Updip shale-outs, strand-lines, and facies changes on anticlines, domes, or other structural features causing dip of the reservoir rock create many combination traps. Unconformities, overlap of porous rocks, and truncation are equally important in forming combination traps. Faulting is also a controlling factor in many of these traps. Asphalt seals and other secondary plugging agents may assist in creating traps. Salt Domes. These structures are of enough importance to justify a separate classification. However, sometimes they are difficult to identify, and many of the traps result from both stratigraphic variations and structural deformation. Intrusion of rocks into overlying sediments may result in many different types of traps. Salt intrusions are more commonly associated with petroleum traps, although some igneous intrusions have also resulted in the formation of petroleum traps. Salt domes are classified as piercement, intermediate, and deep-seated domes. Salt plugs or masses have moved up from greater depths through overlying sediments, forming traps in the sediments that have not been penetrated by the salt. Most salt intrusions took
29-6
PETROLEUM ENGINEERING
HANDBOOK
EAST
TEXAS
FIELD
IZONTAL SCALE
Fig. 29.9-Example of a secondary stratigraphic trap: structure and cross of East Texas field.
section
EAST
TEXAS
FIELD
apparently grew steadily and contemporaneously with the deposition. Many times the salt masses of some domes must have reached the surface or near surface, where groundwaters could act on the intrusive salt mass. Some of the domes are very near the surface of the ground. Some have reached the surface and are currently extruding salt. In areas of very low rainfall, such as southwest Iran, salt has reached a height of 5,000 ft above the surrounding terrain. Deep-seated salt domes are normally those at considerable depths where the salt may not have been penetrated by drilling. These can be identified by the overlying characteristic structure or by geophysical data, which help prove salt is present at depths of 12,000 ft or more and can be assumed to have caused the overlying complex structure. Intermediate domes can be defined arbitrarily as those where the salt is deeper than 2,500 ft but has been penetrated by drilling at depths less than 12,000 ft. Traps occur on the flanks of salt domes where sands have been faulted and deformed or terminate against the salt mass and where facies changes have resulted because of the associated uplift. These conditions are illustrated in Fig. 29. 11.6 Traps occur in the cabrock, which consists of calcite, anhydrite, and limestone. Caprock is the insoluble residue on top of the plug that results from the dissolution of the salt from the crest of the plug. Porosity and permeability in the capmck result from fracturing, solution, chemical alteration, or any combination of these and are generally restricted to the calcite or limestone portions of the caprock. Traps also overlie the salt mass and may result entirely from structural closure, faulting, differential compaction, or stratigraphic variations combined with the deformation, as indicated in Fig. 29.12.l
Fig. 29.10-Example of a secondary stratigraphic trap: structure and cross section of East Texas field.
considerable geologic time to reach their cut-tent position in the crust of the earth. Some are still growing, and apparently all have grown intermittently, allowing sands to be deposited over the crest of the structure at certain times and limiting deposition to the flanks of the sttucture at other times. The movement of the salt mass up through the surmunding rock creates many complex structures and sedimentary variations. Radial and peripheral faulting provide the avenue for the salt pushing up through overlying sediments. At times, the overlying formations were competent enough to stop or delay the growth of the salt plug. At other times, the salt
29-7
,PLIESTOCENE
ing on variations in source rock, transport distance, and depositional environment. Quartzose-Type. Quartzose-type sediments occur in periods of geologic quiescence, with relatively flat coastal plains bordered by shallow seas. Weathering and chemical decay are at a maximum and erosion is at a minimum. Only stable minerals remain, and these are well sorted and generally uniform in texture and composition. Blanket sands and shales over extensive areas are general factors and the sandstones demonstrate high vertical as well as horizontal permeability. Waterdrive reservoir mechanisms can be expected and high recoveries by primary methods of production are the general rule because of the homogeneity of the reservoir rock. The coastal plains, embayments, and continental shelf along the Gulf of Mexico from Texas to Florida are typical of the physical requirements for this type of deposition. Graywacke Sediments. These sediments occur in periods of moderate geologic disturbances. The coastal region is moderately uplifted and the depositional basins are somewhat deeper with a shorter continental shelf. More rapid erosion and shorter distances of transportation prevent the complete weathering and chemical decay of the sediments, and some of the more unstable minerals are able to remain. The land and adjacent basin areas are unstable, and minor isostatic adjustments occur from time to time, causing abrupt changes in the sediments being deposited. This causes poor sorting, lenticularity, irregular porosity and permeability variaand heterogeneous deposition. Vertical tions, permeability is poor, limiting water drive and gravity drainage. Production is normally gas-depletion drive, and the opportunities for secondary recovery operations are excellent. The New England coast is typical of the environment necessary for these types of deposits. Arkose Sediments. Arkose sediments result from deposition during periods of intense erogenic movements. Certain land areas are sharply elevated above other land areas and/or the shoreline. Faulting and major isostatic adjustments occur frequently. The con-
i A -OLIGOCENE
JACKSON
BARBERS
HI1 YEGUA -
gL1000 FT
Fig. 29.1 l-Example of piercement-type dome showing termination of sands on- the flanks of the salt plug and the resultant reservoir traps.
Clastic or detrital rocks are created from fragments transported by wind or water and allowed to settle out of suspension when the weight of the fragments is sufficient to exceed the carrying capacity of the transporting agent (wind or water). Chemical rocks are the result of precipitation of materials out of aqueous solution by organic growth and deposition or evaporation of seawater in closed basins, which precipitates salt and other evaporites. A list of reservoir rocks is given in Table 29.1. Detrital Reservoirs Clastic or Detrital Granular Reservoirs. These reservoirs can be classified according to rock types, depend-
Fig. 29.12-Example of traps caused by a combination of structural and stratigraphic variations and of the complex faulting occurring above an intrusive salt mass.
29-8
PETROLEUM ENGINEERING
HANDBOOK
ROCKS
29.2-TYPES OF RESERVOIRS
1. Sand, conglomeratic sand, and gravel (clean, argillaceous, silty, lignitic, etc.) 2. Porous calcareous sandstone and siliceous sandstone (incomplete cementation) 3. Arkosic (feldspathic) sand, arkose, arkosic conglomerate (granite wash) 4. Detrital limestone and dolomite, oolitic and pisolitic limestone, coquina, and shell breccia Fractured Porosity 1. Fractured sandstone and conglomerate 2. Fractured limestone, shale, and chert Crystalline Porosity 1. Crystalline limestone and dolomite 2. Sugary dolomite saccharoidal porosity Solution Porosity 1. 2. 3. 4. 5. Crystalline limestone and dolomite Cavernous limestone and dolomite Porous caprock Honeycombed anhydrite Ooliclastic limestone
Sometimes present in brittle, siliceous fractured shales Anhydrite Evaporites Develop porosity from leaching lqneous or Metamorphic Rock by circulating waters
1. Very uncommon 2. Sometimes contain oil when secondary porosity is developed by fracturing or weathering 3. Best-known igneous reservoirs are the serpentine plugs of Bastrop and Caldwell counties, Texas
moundlike or lenslike form. As an organic layer, such as a bed of shells, crinoids, or corals or a modem reef in the course of formation. Brecciat A coarse-grained elastic rock composed of angular broken fragments held together by a mineral cement or in a fine-grained matrix. Closure: In a subsurface fold, dome, or other structural trap, the vertical distance between the structure s highest point and its lowest closed structllre contour. Four-way dip is determined by in-line and cross-line right angle control demonstrating dip in four directions away from the crest of the closure. Coquina: A dettital limestone composed wholly or chiefly of mechanically sorted fossil debris that experienced abrasion and transport before reaching the deposition site, and weakly to moderately cemented but not completely endmated. Deftifalr Pertaining to or formed from detritus of rocks, minerals, and sediments. The term may indicate a source outside or inside the depositional basin. Facies: The aspect, appearance, and characteristics of a rock unit, usually reflecting the conditions of its origin; especially as differentiating the unit from adjacent or associatedunits. Graben: An elongate, relatively depressedcrustal unit or block that is bounded by faults on its long sides. Horst: An elongate, relatively uplifted crustal unit or block that is bounded by faults on its long sides. Minimumhydrodynamicpotential: As used here, a geologic position or condition due to impermeability in the reservoir rock where the dynamic action of fluid movement is abated. Minor isosfafic adjustment: The minor adjustment of the lithosphere of the earth to maintain equilibrium among units of varying mass and density; excess mass above is balanced by a deficit of density below and vice versa. Norn~lfuult: A fault in which the hanging wall appearsto have moved downward relative to the foot wall. The angle of the fault is usually 45 to 90. A low-angle normal fault is a normal fault with the angle of the fault less than 45. Ofilop deposirion: The progressive offshore regression of the updip terminations of the sedimentary units within a cornformable sequence. of mcks in which each successively younger unit leaves exposed a portion of the older unit on which it lies. The successive contraction in the lateral extent of strata (as seen in an upward sequence) resulting from their being deposited in a shrinking sea or on the margin of a rising land mass. On-fop deposition: The regular and progressive pinching out toward the margins or shores of a depositional basin of the sedimentary units within a conformable sequence of rocks in which the boundary of each unit is transgressed by the next overlying unit and each unit, in turn, terminates farther from the point of reference.
tinental shelf is extremely narrow or nonexistent. Maximum erosion and the short distance of transportation virtually eliminate chemical decay and weathering. Sediments are deposited and covered over by younger sediments before any appreciable sorting and weathering can take place. Unstable minerals are present in the thick heterogeneous deposits. Highly porous stratigraphic traps are developed by lensing, pinchouts, and unconformities. Depletion-drive reservoirs are the general rule and recoveries are usually low. Much of the California coast is typical of the depositional environment for the deposition of arkosic sediments. Chemical Reservoirs Limestones and Dolomites. Limestones and dolomites also are deposited in quiescent geologic environments. Deposition of limy deposits is occurring along the west coast of Florida and some of the Bahama Islands while elastic sediments are being deposited in other nearby local areas. Carbonate Reservoirs. Carbonate reservoirs include limestones, and reefs, elastic limestones. chemical dolomite. Porosity may be intercrystalline, intergranular, oolitic or ooliclastic, vuggy fractured, fossiliferous, cavernous, or saccharoidal. Production characteristics are highly variable in carbonate reservoirs. depending almost entirely on the type of porosity and fracturing developed and the resultant permeability. Other types of reservoirs are given in Table 29.2.
Glossary of Terms
Eioherm: A moundlike, domelike, lenslike. or reeflike mass of rock built by sedentary organisms (such as corals, algae, foraminifers, mollusks, and gastmpods), composed almost exclusively of their calcareous remains and enclosed or surrounded by rock of different lithology. Biostrom: A distinctly bedded and widely extensive or broadly lentitular, blanketlike mass of rock built by and composed mainly of the remains of sedentary organisms and not swelling into a
PETROLEUM
RESERVOIR
TRAPS
29-9
Oolicnsr: One of the small, subspherical openings found m an oohtlc mck, produced by the selective solution of ooliths without destmction of the matrix. Oolicusric porosity: The porosity produced in an oolitic rock by removal of the ooids and formation of oolicasts. Oolith: One of the small round or oval accretionaty bodies in a sedimentary rock resembling the me of fish, usually formed of calcium catbonate and having a diameter of 0.25 to 2 mm. Pisolirh: One of the small, round or ellipsoidal accretionary boches in a sedimentary rock, resembling a pea in size and shape, and constituting one of the grains that make up a pisolite. It is often formed of calcium carbonate and some are thought to have been produced by a biochemical algae-encrustation process. A pisolith is larger and less regular in form than an oolith. Pisoliric: Pertaining to pisolite or to the texture of a rock made up of pisoliths or pealike grains. Pisolitic limestone: A limestone with pisolitic texture. Saccharoidal: Said of a granular or crystalline sugar. texture resembling
9. Krynine, P.D.: The Megascopic Study and Field Classification of Sedimentary Rocks, J. Geol., 56, No. 2.
General References
Aguilera R. : Natural Fractured Reservoirs, Petroleum Publishing Co., Tulsa (1980). Bates, R.L. and Jackson, J.A.: Glossary ofGeology, second edition, American Geological Inst. (1980). Beebe, Warren B.: Natural Gases of North America, Vols. 1 and 2, Memoir 9, AAPG (1968). Bouma, H., Moore, G.T., and Coleman, J.M.: Framework, Facies and Oil Trapping Characteristics of the Upper Continental Margin, AAPG (1978) Studies No. 7. Braunstein, J.: North American Oil and Gas Fields, AAPG (1976) Memoir 24. Busch, D.A.: Stratigraphic Memoir 2 1. Traps in Sandstone, AAPG (1974)
Shell breccint A breccia composed of angular broken shell fragments. Shoestting sands: A shoestring of sand or sandstone usually buried in the nudst of mud or shale as in a buried sandbar or channel fill. Strand line: The ephemeral line or level at which a body of standing water meets the land; the shoreline, especially a former shoreline now elevated above the present water level. Strike-slipfault: A fault on which the movement is parallel to the fault s strike. Torsion: The state of stress produced by two force couples of opposite movement acting in different but parallel planes about a common axis. Torsion faults are wrench faults or lateral faults in which the fault surface is more or less vertical. Tnmcnrion: An act or instance of cutting or breaking off the top or end of a geologic stmcture or land form, as by erosion. Unconformity:A substantial break or gap in the geologic record where a mck unit is overlain by another that is not next in stratigraphic succession, such as interruption in the continuity of a depositional sequence of sedimentary rocks or a break between eroded igneous rocks and younger sedimentary strata. It results from a change that caused deposition to cease for a considerable span of time, and it normally implies uplift and erosion with loss of the previously formed record.
Geologic Formation and Economic Development of Oil and Gas Fields of California, California Department of Natural Resources, Sacramento (1943). Halbouty, M.T.: Cilant Oil and Gas 1968-1978, AAPG (1980) Memoir 30. Fields of the Decade
Halbouty, M.T.: Salt Domes, Gulf Region, United States and Mexico, second edition, Gulf Publishing Co., Houston (1979). Halbouty, M.T.: The Deliberate Search for the Subtle Trap, AAPG (1982) Memoir 32. Hanna, M.A.: Gulf Coast Salt Domes, Geology, AAPG (1934). Problems in Petroleum
Hubbert, M.K.: Entrapment of Petroleum under Hydrodynamic ditions, Bull., AAPG (Aug. 1953) 37, 1954-2026.
Con-
King, R.E.: Stratigraphic Oil and Gas Fields: Classification, Exploration Methods and Case Histories, AAPG (1972) Memoir 16, S.E.G. Special Publication No. 10. Levorsen, A.I.: Geology Francisco (1954).
to Hydrocarbon Ex-
References
1. Galloway, T.J.: Bull. 118, California Division of Mines, Sacramento (Aug. 1957). Field; Good Example of Subtle 2. Sams, H.: Atkinson Stratigraphic Trap, Oil and Gas J. (Aug. 12. 1974), 145-63. 3. Hoyt, W.V.: Erosional Channel III the Middle Wilcox Near Yoakum, Lavaca County, Texas, Trans., Gulf Coast Assn. of Geological Societies (Nov. 1959) 9, 41-50. 4. Occurrence of Oil and Gas in West Texas, F.A. Herald (ed.) Bureau of Economic Geology and West Texas Geological Sot. (Aug. 1957). 5. Occurrence of Oil and Gas in Northeast Texas, F.A. Herald (ed.) Bureau of Economic Geology and East Texas Geological Sot. (April 1951). 6. An Introductionro Gu[fCoa~r Oil Fields, Houston Geological SOC. (1941). 7. A Guide Book, Houston Geological Sot. (1953). 8. Pirson, S.J.: Oil Reservoir Engineering, second edition, McGrawHill Book Co. Inc., New York City (1958).
Russel, W, L. : Structural Geologyfor Petroleum Geologists, McGrawHill Book Co. Inc., New York City (1955). Scholle, P.A., Bebout, D.G., and Moore, C.H.: Carbonate tional Environments, AAPG (1983). Deposi-
Structure of Typical American Oil Fields, Bull., AAPG (1929). Weeks, L.G.: Habitat of Oil, AAPG (1958). Wilhelm, 0.: Classification of Petroleum Reservoirs, (Nov. 1945) 29, 1537-79. Bull., AAPG
Woodland, A.W.: Petroleum and the Continental Shelf of Northwest Europe, Vol. I, Geology, John Wiley Sons Inc., New York City (1975). Young, A. and Galley, I.E.: Fluids in Subsurface Environments, AAPG (1965) Memoir 4.
Chapter 30
Bottomhole Pressures
G.J. Plisga,
Sbhio Alaska Petroleum Co.*
Introduction
The practice of using bottomhole pressure (BHP) to improve oil production and to solve petroleum engineering problems started in about 1930. Pressures in oil wells were first calculated from fluid levels and later by injecting gas into the tubing until the pressure became constant. The earliest BHP measurements were made with pressure bombs and with maximum-indicating or maximum-recording pressure gauges that did not have the accuracy, reliability, or durability now demanded. These early pressure measurements were occasional. or spot tests rather than systematic diagnostic engineering measurements. are numerous mechanical self-contained pressure devices available (Table 30.1) only the most commonly used continuously recording BHP gauges are discussed fully. Regardless of the type of force summing device incorporated into the BHP gauge, whether physical displacement (piston elements) or deformation (bellows/bourdon tubes), the generated force is coupled to a recording device. The Amerada pressure gauge has a helical bourdon tube as a pressure element that is of sufficient length to rotate the stylus the full inside circumference of the cylindrical chart holder without multiplication of movement. A clock moves the chart longitudinally. The gauge is made in both 1% - and l-in. diameters with a length of approximately 74 in. A vapor-pressure-type recording thermometer can be run in combination with this to obtain continuously recorded temperatures and allow correction of pressure measurements. This will also ensure that thermal stabilization has occurred. The Humble gauge pressure element has a piston, which moves through a stuffing box against a helical spring in tension. Attached to the inner end of the piston is a stylus that records longitudinally on a chart in a cylindrical holder, which is rotated by a clock. The instrument is made in two sizes, with 1 l/4- and 15/16-in. OD s, and is approximately 60 in. long. Thermometer elements are available for both sizes. Other recording gauges have been described in the literature, two of which were continuously recording, but they are no longer available on the market. The Gulf BHP gauge has a pressure element consisting of a long metallic bellows restrained by a double helical spring in tension. The recording mechanism is a cylindrical chart holder rotated by a clock. The USBM BHP gauge pressure element is a multiple-bellows type with a movement of about 0.6 in., which is multiplied through a rack and gears to about 5% in. of stylus movement. The stylus records longitudinally on a cylindrical chart that is
BHP Instruments
The development of precision recording pressure gauges small enough in diameter to be run through tubing made it feasible to make BHP measurements in sufficient number to develop the science that now makes them indispensable to petroleum engineering. BHP now is determined with continuously recording pressure gauges, which are either self-contained or surface-recording. Self-Contained Gauges
Mechanical self-contained pressure gauges are used universally. The pressure element and recording section are encased and sealed against external pressure except for an opening to communicate the pressure to the element. The entire instrument is run to the depth at which the pressure is to be measured, allowed to stabilize thermally, and then returned to the surface and the pressure determined from the chart. Modem pressure measurement systems incorporate force summing devices that convert energy into physical displacement or deformation. These force summing devices can take many forms, three of which are shown in Fig. 30.1. Although there
The ongmat chapter on the topic bn the 1962 edition was written by C.V. Mfllikan.
30-2
PETROLEUM ENGINEERING
HANDBOOK
TABLE
30. 1--MECHANICAL
RECORDING
BHP GAUGES
RPG-3 OD, in. Length, in. Type pressure element* Maxtmum pressure, psi Accuracy, %FS t Resolution, %FSr Maximum service temperature, OF Maximum clock running time, hours 1.25 77 B 25,000 +0.2 +0.05 500 360
Amerada RPG-4 1 76 B 25,000 f 0.2 f 0.056 500 144 RPG-5 1.5 20 B 20,000 k 0.25 + 0.05 450 120 KPG 1.25 73
B
Kuster AK-1 2.25 36 B 30,000 * 0.25 f0.025 350 120 K-2 1 41 B 20,000 f 0.25 +0.05 700 120 K-3 1.25 43 B 20,000 f 0.25 -f 0.04 700 120 Leutert 1.42 139 RP 10,000 f 0.025 f 0.005 300
Johnston J-200
2.88
B = bardon tube, AP = Rotating piston, P = Piston. Normally, elements are avarIable in several ranges. tFS= Full scale. NS = Not stated
PRES.SU{~~-
;E
DIAPHRAGM
BELLOWS
pressures readings are made on a chart. When a number of readings are to be made from a chart, it is advantageous to use one of the several available chart scanners. Some engineers have used microscopic comparators to read pressure deflection to 0.0001 in., but the inherent errors of a pressure element even under most careful handling are usually greater than the added accuracy of such precise measurement of the chart. New electronic chart scanners have improved the readability and accuracy of mechanical BHP gauges. Calibration Self-contained pressure gauges, like all pressure gauges used for precision work, must be calibrated on a deadweight tester at regular intervals. To obtain maximum accuracy, the pressure gauge is calibrated before a survey at the anticipated bottomhole temperature, (BHT), the survey conducted, and then the chart read using the presurvey calibration lines. New pressure elements should be calibrated frequently until they have become seasoned in service and their ability to retain calibration has been established. Before calibration, pressure equal to the maximum range of the element should be applied and released several times. The number of calibration points should be more on a new element, and two or more curves should be run as a check. The element should also be calibrated at the reservoir temperature at which the pressures are to be determined, or a temperature-correction factor should be determined to correct pressures measured at other than calibration temperature (Fig. 30.2). During calibration, the gauge should be tapped lightly to relieve residual friction in the moving parts of the element. Under normal operating conditions, pressure determination in a well should have an accuracy within a range of 0.2% of the maximum range of the element. The pressure element range should be selected to operate in the upper two-thirds range when at bottomhole conditions. Greater accuracy can be obtained by greater attention to details of calibration and the use of instruments that are wellseasoned by service. Pressure increase inside the gauge caused by an increase in temperature is considered only when extra effort is made to obtain precision pressures. Temperature Effect Temperature effect is an inherent property of metals and is present in all gauges, although for some alloys it is very small. Except for such alloys, temperature change
Fig. 30.1-Force
summing devices
rotated by a helical spring but controlled by a watch movement. BHP gauges, although rugged and capable of service in severe conditions, must be considered precision instruments. Proper attention to adjustment, calibration, and operation is required to obtain consistently reliable and accurate pressure measurements. A comparison of the commonly used mechanical continuously recording BHP gauges is shown in Table 30.1. Charts Charts used in a BHP gauge are paper or metal. Paper charts have an abrasive coating and are marked by a brass or gold stylus. Metal charts, which are made of brass, copper, or aluminum, are generally preferred because they are not affected by humidity. Plain metal charts require a sharp pointed stylus. Coated metal charts are generally preferred because they produce less stylus friction. Black-coated charts are marked with a steel or jewel stylus, which burnishes the coated surface. Whitecoated charts are used with a brass or gold stylus. A finer line can be made on the black chart, but it is more difficult to read. A brass or gold stylus used with whitecoated charts, paper, or metal, must be sharpened very frequently. A small magnifying lens and steel scale with O.Ol-in. divisions are most frequently used for reading charts with static pressure-i.e., where only one or two
BOTTOMHOLE PRESSURES
- - - - _ _- - - - - - - - -
- - TKlle Marlmum
output
- - -
- -
- -
Time
--\
outpur \ \ \ \ \ \ \ \
4
Temperature Error / / / / , , /
.-T,mE?
Fig. 30.3-Hysteresis.
/ .II/
Fig. 30.2-Temperature
must be considered in pressure measurements. The preferred method is to calibrate the pressure element at the temperature of the reservoir in which pressures are to be measured. The calibration curve for most pressure elements is practically a straight line; therefore, a temperature-correction coefficient may be determined for a given pressure element and used to correct for temperatures other than the calibration temperature as follows. For a given pressure, preferably about threefourths of the maximum for the element, determine the pressure deflection and the temperature. For the same pressure, determine the deflection at a higher temperature, preferably 100F higher. Then
If it is more convenient, pressure readings may be substituted for deflections in Eq. 2. Gauges with a steel pressure element usually have a temperature coefficient of about 0.0002 in./F. When the pressure gauge is run to the depth of the pressure determination, it should remain long enough to stabilize thermally, usually 15 to 20 minutes. If the instrument cannot remain long enough to reach temperature equilibrium, a maximum-indicating thermometer run in a closed container as part of the gauge will give a satisfactory reading for temperature correction. Many BHP gauge elements are made of an alloy with a very low temperature coefficient such as Ni-Span C@, and a temperature correction may be neglected up to 200F except where extra precautions are taken to obtain very precise pressure measurements. For temperatuf?s above 2OOF, most elements require a varying correction that can be determined only by actual calibration. Hysteresis Hysteresis is a characteristic of metals under strain that must be recognized in pressure gauges. Because of hysteresis, the calibration of a gauge made with increasing pressures will differ slightly from a calibration made with decreasing pressures (Fig. 30.3). If only static pressures are to be determined, a calibration at increasing pressure is satisfactory. When a flowing pressure starting from a static condition is to be determined, hysteresis may be of sufficient magnitude to take into account. To determine the hysteresis effect, the pressure element should be pressurized somewhat higher than the highest anticipated well pressure and released several times before running the calibration, first with increasing pressures, then with decreasing pressures. Flexing the element several times will substantially reduce the hysteretic effect and should be done each time just before the gauge is run into a well.* Operating Equipment The BHP gauge, a self-contained instrument, is run on a wireline and depth is measured by the line running in contact with a calibrated measuring wheel, which operates a counter. The most frequently used calibration of the measuring wheel is 2 ftirev. When the contact of the line is tangent to the measuring wheel, the wheel
(11
where CT T, T2 dl
= = = = d2 =
temperature coefficient, lower temperature, higher temperature, deflection at T, for given pressure, and deflection at T2 for same pressure.
d,.=
1 +CT(To -T,)
.. ..
(2)
where
d,.= d, = T, = T, =
deflection at calibration temperature, observed deflection, calibration temperature, and observed temperature.
30-4
PETROLEUM ENGINEERING
HANDBOOK
TABLE 30.2-WIRELINE
Tensile Strength (Psi) Plow steel ~ai$3ss steel 232,000 150,000 170,000
Diameter Line (in.) 0.066 -945 794 513 582 11.4 11.8 13.1 611 692 14.0 14.1 15.6 0.072 0.082 1,225792 090 18.0 18.3 20.2 0.092 1,542 1,130 997 22.8 23.0 25.4
Nominal Weight per 1,000 ft of Line (Ibm) Plow steel Stainless steel Monel
diameter in inches is 241~. When the contact is an arc of the wheel, the wheel diameter is D=(24/n)-d, where d is the diameter of the wireline. This applies if the greatest distance of the chord of the contact arc from the periphery of the wheel is greater than the diameter of the wireline. The measuring wheel diameter should be checked at reasonable intervals to maintain accuracy of depth measurements. A decrease in the diameter of the wireline caused by wear or by permanent stretch resulting from a hard pull will also cause errors in depth measurements. The most common wirelines have diameters of 0.066, 0.072, 0.082, or 0.092 in. In areas having noncorrosive well fluids, plow steel lines are most satisfactory, but for corrosive conditions, stainless steel and Monel@ lines are used. Both plow steel and stainless steel are subject to hydrogen embrittlement, but are satisfactory for short runs such as static-pressure tests, except under severe conditions. For an operation that requires the line to be in the hole under corrosive conditions for several hours, a Monel line should be used. Nominal tensile strengths of wire lines and their weights are given in Table 30.2. Equipment for operating a wireline varies greatly. The most frequently used unit is a trailer-mounted reel driven by a 2- to 4-hp air-cooled gasoline engine. Power is transmitted to the reel by V-belt, through a disk clutch, or by hydraulic drive. On smaller equipment, an idler pulley permits the V-belt to serve as a clutch also. Braking may be by friction disk, brakeband, or hydraulic
pump. When the equipment is used for a continuous program it may be mounted in a pickup truck, with housing to protect against the weather. Pressure Bombs Pressure bombs were used to some extent before recording pressure gauges small enough in diameter for oilwell use became available. They were usually made from tubing approximately 1 L/2 in. in diameter with a small needle valve in the top and a ball and seat in the bottom to hold maximum pressure in the well. When the bomb was recovered, the pressure was determined by attaching a pressure gauge to the valve. The bomb had to be long enough to leave a volume of gas (or air) in the top to reduce the error of filling the bourdon tube of the pressure gauge. There was also substantial correction for temperature unless the bomb was raised to BHT before the pressure was read. An ordinary commercial maximum-indicating pressure gauge enclosed in a pressure-tight container was used occasionally, and in some cases a recording mechanism and clock were added. Such instruments were 3 in. or larger in diameter, which limited their use to wells without tubing. Surface Recording Gauges
Surface recording pressure gauges can be used either permanently installed or wircline retrievable. All surface recording pressure gauges must be run on a single-
BOTTOMHOLE PRESSURES
30-5
TABLE 30.4-SUMMARY
Pressure Accuracy SGTISOI Capacitive Excitation AC-DC special Output Level PM 0.02 Range (Psi)
OF TRANSDUCER CRITERIA
Frequency Response (Hz) 0 to>100 Oto Temperature VI +I85 Range Shock and Stability WY0 0.05
>
and Effects
0.01 to 5,000
with <0.05% calibration shift 0.5 30 to 10,000 >I00 Oto +165 0.5
Differential
transformer
AC special
poor
> to6
cycles life
Force balance
AC line power
0.01
1 to 5,000
oto
<5
poor
O.OS%/month
Piezoelectric
medium
level
0.1 to 40,080
1 to > 100.000
excellent
1 use effects
with amp
Potentiometer
AC-DC regulated
high level
5 to 10,000
oto
>50
-85
to +300
poor
05
(O.Ol%lF)
Strain gauge
AC-DC regulated
Unbended
IOV AC-DC
0.25
0.5 to 40,000
0 to >2,000
good
0.5
compensated
Bonded
foil
1 OV AC-DC
0.5
5 to 10,000
0 to >I,000
-65
to +250
very good
0.5
(O.O1%/F over limited range) Thm film IOV AC-DC 3 mVN 0.25 15 to 10,000 0 to > 1,000 - 320 to + 525 (O.OOS%/~F over limited compensated range) Diffused semiconductor 10 to 28 V DC medium mVN level 3 0.25 15 to 5.000 0 to > 1,000 - 65 to 250 (0.005%/F limited over 0.05 < 0.25% calibration shift after 10 B cycles very good 0.05
to 20 mVN
compensated range) 1OV AC-DC medium level 3 mVN to 20 mVN 0.25 5 to 10,090 0 to > 1,000 -65 to t250 (O.Ol%lF over limited range) Variable reluctance AC special 40 mVN 0.5 0.04 to 10,000 0 to > 1 ,OiJo -320 to +800 0.5 0.5 < 0.5% calibration shift after lo6 cycles
QQQd
> lo6
cycles life
(O.O2%/OF over limited compensated range) AC special high level and frequency 0.02 1t0100 oto >I00 -85 to +200 requires temperature Vlbratlng quartz AC special high level and frequency 0.01 1 to 10,000 0 to >I00 oto control 0.005 0.01
Vibrating
we
and tube
+I302
> lo6
cycles life
Stability
and calibration
shift should
be considered
together.
30-6
PETROLEUM ENGINEERING
HANDBOOK
L&g5
BOURDON TUBE
:::: ..-. q I
OUTPUT
/ BASE
CRYSTAL
transducer.
Variable Inductance Transducer conductor armored cable that carries a direct current from the surface to the transducer in the bottomhole instrument. Oscillating current returns through the same circuit from the transducer to surface instruments that determine and record its frequency. A transducer is any device that converts energy from one form to another. There is a large variety of transducers that allow nonelectrical variations to be converted into changes in resistance, current, voltage, capacitance, etc. Some examples are the strain gauge, thermistor, and the microphone. Readings are made at selected intervals of 1 second to 30 minutes or more. The frequency recorded in cycles per second is translated to pounds per square inch from a calibration curve. Table 30.3 shows a comparison of commonly used surface recording pressure gauges. A permanently installed surface recording pressure gauge requires a gauge carrier or receiver and either a single conductor line or small diameter tubing (0.092 in.) strapped to the production tubing. The pressure gauge can be run with the tubing or by wireline retrievable, which sits in a gas lift mandrel or some other device to complete the circuit. The surface instruments may be connected permanently, or one set can be used to monitor BHP in several wells. Modem precision pressure-measuring systems incorporate force summing devices that convert gas or liquid entry into physical displacement or deformation. The following sections discuss various concepts of pressure transducer technology shown in Table 30.4. In the variable inductance transducer, a flux linkage bar is mechanically linked to a spiral bourdon tube, diaphragm, or bellows (Fig. 30.1). The flux linkage bar is in the magnetic path of an E-core transformer (Fig. 30.4). Displacement of the flux linkage bar by pressure changes the E-core flux density resulting in a transformer output proportional to the pressure applied. The advantages are medium-level output and rugged construction. Disadvantages are a requirement for AC excitation, poor linearity, and susceptibility to stray magnetic fields. Piezoelectric Transducer The piezoelectric effect is the property exhibited by certain crystals of generating voltage when subjected to pressure (Fig. 30.5). When a strain is applied to an asymmetrical crystalline material such as barium, titanite, quartz, or rochell salt, an electrical charge is generated. When a piezoelectric crystal is connected to a diaphragm, bellows, or bourdon tube, the generated charge can be made proportional to the applied pressure. Advantages are very high frequency response (250 kHz), small size, rugged construction, and ability to accept large overpressures without damage. Disadvantages are temperature sensitivity, inability to make static measurements, and special electronics required. Potentiometric Transducer This transducer is constructed by coupling the wiper of a multitum potentiometer to an amplifying mechanical linkage to a diaphragm, bellows, or bourdon tube. Advantages are low-cost, high-level output and simple electronic circuits. Disadvantages are limited life, poor resolution, large hysteresis, and low frequency response. Vibrating Wire Transducer A thin wire is connected in tension to a diaphragm, bellows, or bourdon tube and is caused to vibrate under the influence of a magnetic field (Fig. 30.6). The wire s frequency of vibration is directly related to its tension.
BOTTOMHOLE PRESSURES
30-7
/ OUTPUT MAGNETS
Fig. 30.7-Thin
Fig. 30.6-Vibrating
wire transducer.
Advantages are very high accuracy, low hysteresis, and excellent long-term stability. Disadvantages are sensitivity to shock and vibration, temperature sensitivity, and additional electronics. Strain Gauge Transducers
A strain gauge transducer is a strain-sensitive resistor mounted to a diaphragm, bellows, or bourdon tube. When pressure is applied, the resistor changes its physical length thereby causing change in resistance. This effect is expressed by Arlr
F,=- ~,L,
..
. . . . . . . . . . . . . . . . . . . . . . . (3)
where = Ar = r = aL = L =
F,
gauge factor, change in resistance, unstrained resistance, change in length, and unstrained length.
term stability characteristics, and ease of vibrational excitement. Fig. 30.8 shows various modes of motion of quartz crystal. Advantages are excellent accuracy, resolution, and long-term stability. Disadvantages are sensitivity to temperature and extremely high costs. New technology is constantly evolving a new generation of surface recording pressure gauges that are more and more advantageous to the petroleum engineer and petroleum engineering problems.
There are four basic types of strain gauge transducers; unbonded wire, bonded foil, thin film, and semiconductor. A rule that applies to strain gauge transducers is the larger the gauge factor, the higher the output of the device. For unbonded wire, the gauge factor is four. Bonded foil and thin film (Fig. 30.7) have factors of two. For semiconductor transducers, the factor ranges from 80 to 150. Vibrating Crystal Transducer In vibrating crystal transducers, a crystal is forced by external electronic circuits to oscillate at its resonate fre quency when external stress is applied to the crystal by mechanical linkage to the diaphragm, bellows, or bourdon tube. The resonate frequency of the crystal shifts in proportion to the stress. In at least one transducer of this type, the pressure is applied directly to the crystal itself. The vibrating crystal is usually manufactured out of quartz because of its excellent elastic properties, long-
Calculated BHP
BHP calculated from surface pressure and fluid level, although less accurate than measured pressure, is sufficient for many practical uses. In an open hole or open tubing, the fluid level can be determined by a float run on a measuring line. In pumping wells, the fluid level can be determined by sound reflection. There are four types of commercial instruments available. These are the deptbograph, echometer, sonoloy, and acoustical well sounder. Each of these instruments records sound reflection initiated by firing a blank shotgun shell or pistol cartridge or venting pressure into a chamber attached to the casing head. A sound reflection is received and recorded from each tubing collar. By counting the collar reflections and knowing the tubing tally, the depth to the fluid can be calculated. In deep wells, attenuation of the collar reflection makes accurate counting difficult. In some oil wells,
30-8
PETROLEUM ENGINEERING
HANDBOOK
usually those having considerable gas, a foaming condition makes the fluid level difficult to identify or may indicate a fluid level much higher than actual. A foaming condition is usually indicated when the fluid level changes several feet on tests made at short intervals of time. Fluid level can also be calculated from the time interval for the reflection to be received from the fluid level, but the variation in the speed of sound through gases of different compositions and the effect of temperature make the procedures more laborious and usually less accurate than the simpler method of counting collar reflections. In calculating the pressure caused by the column of fluid, allowances should be made for gas in solution in the oil, which will reduce its specific gravity below that measured in the stock tank. This can reduce the gradient 5 psi/100 ft or more where much gas is in solution under high pressure. For wells producing water, it is customary to calculate the fluid head on a basis of a column of oil and water in the same ratio as normal production for the well, but this calculation is less reliable for low-capacity wells with high casinghead pressure and for pumping wells in which the pump is several hundred feet above the producing formation. If the surface pressure is low, the pressure caused by the weight of a column of gas may be too small to warrant consideration, but under high pressure it should be calculated and added to the hydrostatic head by the same equation used for calculating BHP in a gas well, where D is the depth of the fluid level. Producing BHP of a pumping well that is sufficiently accurate for practical use may be calculated by shutting in the casing head until the gas pressure depresses the fluid level to the inlet of the pump, at which time fluid delivery is stopped. The casinghead pressure at that time, plus the head of the column of gas from the casing head to the pump inlet, plus the head of the column of liquid from the pump inlet to the producing formation is the producing BHP. A check of the determination should be made by releasing and controlling the pressure a few pounds less than the maximum pressure read and determining the rate of production under such conditions. The BHP in a gas well can be calculated with an equation developed by Pierce and Rawlins4:
pw5=(pwh)e
0 m347y *D
where T is the average temperature in the borehole, F+460. Deviation of a gas from Boyle s law will affect the calculated BHP enough to be considered only in high pressure deep wells. USBM Monograph 7 presents the following equation.
Application of BHP
The importance of pressure analysis in projecting and enhancing the performance of producing oil and gas wells emphasizes the need for precision pressure measurement systems. Today s petroleum engineer must have sufficient information about the reservoir to adequately analyze current performance and predict and optimize future performance. More specifically, such pressures are a basic part of reservoir calculations, rate of equalization of pressures, interference tests for well spacing or rate of development, formation damage during completion, rework or workover operations, and indication of deposition of salts, sediments, or other restrictions at the wellbore. Other applications are design of downhole equipment for artificial lifting, efficiency of operation of such equipment, and evaluation of drillstem test (DST) information. Static Pressure Static pressure is the most frequent BHP measurement. Most such measurements are made as a pressure survey of a pool where the pressures in all wells are determined in a short period of time either by cooperation of the operators or by order of a conservation commission, usually as a result of a recommendation by the operators. Pressures are taken under reasonably uniform conditions after the wells have been shut in a specified length of time such as 24 or 48 hours, or longer, if the pressure buildup is at a slow rate. The pressures should be measured at or adjusted to a common data plane. In many pools, the pressures will not reach equilibrium in the specified shut-in time. However, if the pressures are determined for several surveys under the same conditions, the indicated rate of decline of the reservoir pressure should be reasonably accurate. Tests in representative wells which have been shut in long enough to reach pressure equilibrium will show the relation of the measured pressure to the actual reservoir pressure. Pressures in inactive wells may be used to confirm the actual pressure and the rate of decline. Average Reservoir Pressure
..
. . . . . I. I.. . . . . (4)
where p,,,$ = static BHP, psia, p,+h = wellhead pressure, psia, e = base, natural logarithms, YR = specific gravity of gas (air= l), and D = depth of well, ft. This equation is based on an average temperature of the column of gas of 60F. While the temperature gradient in a producing well is rarely a straight line, the average temperature at a depth below seasonal effect (20 to 30 ft) and at a depth of the pressure is sufficiently accurate for practical purposes. The equation can then be written5 :
The average reservoir pressure for a pool may be determined by arithmetically averaging the pressures of all
BOTTOMHOLE PRESSURES
30-9
wells. For some pools it is preferable to determine a weighted average by weighting each pressure by the productive thickness of the reservoir at that point. When for any reason the pressure cannot be determined in substantially all wells or where wells are irregularly spaced, a better average reservoir pressure is determined by recording the pressures, either actual or weighted, on a map of the area and drawing isobars from which the average pressure weighted for an area is determined by planimeter, grid system, or other means. Static Pressure from Partial Buildup Many low-permeability reservoirs require excessive shut-in time to reach static or equilibrium pressure. Several methods have been proposed for calculating the reservoir pressure from partial buildup of pressure. Muskat proposed plotting log (p,,,$ -pr) vs. time, where pbrs is an estimated static BHP and pr is the measured pressure at different times, p,, , pt, , p,, , etc. When plotted on semilog cross-sectional paper with pressure on the log scale, the selected pws is the static pressure when the plot is a straight line. Arps and Smith7 proposed plotting increments of pressure increase for uniform time periods against measured pressure on rectangular cross-sectional paper, and extrapolating the curve to intersect the zero line of the incremental-pressure scale, which gives the static pressure. Both the Arps and Smith and the Muskat methods are more commonly used in cases of rapid pressure buildu Miller et al. .r presented an equation to calculate the static pressure from a partial buildup curve:
P~,~=P*+ (Psd-*!j')@ o~oo70*
from which A@is determined in the Miller et Fig. 30.9-Curves al. equation for calculating static pressure from buildup pressure curve. Solid lines assume influx at fd and dotted line indicates direction of curves when no influx at rd.
kh
. . .
. . . .
where
PWS
p* = last pressure on buildup curve, psi psd = log,(rdlrw) for constant pressure at radius
4 p B k h @
of drainage, or = log,(rd/r,)-0.75 when no influx across external boundary, = production rate at shut-in, B/D, = reservoir fluid viscosity, cp, = formation volume factor for total fluid produced, RBISTB = permeability of reservoir, md, = effective thickness of reservoir, ft, = dimensionless pressure variable from curves (see Fig. 30.9),
ds = Pnf-Pws, p4 = BHP during buildup, psi, PWS = BHP at time of shut-in, psi, and
t = time after well was shut in. hours. The value of p* is determined by plotting the buildup pressure vs. the time on semilog paper, with time on the log scale. When afterflow is completed following shutin, the points should fall on a straight line and p* is the
highest measured pressure lying on the straight line. This is the same straight line by which the slope, m, is determined in the equation for permeability from buildup pressures by the same authors and is discussed further under that topic. The straight line by which the slope m is determined comes from the middle time region of a Homer plot. Afterflow causes lack of development of the middle time region (with long periods of aftefflow), early onset of boundary effects, and development of several false straight lines that could be mistaken for the middle time region. This makes the middle time region difficult for the buildup test analyst to recognize. Recognition of the middle time region is essential for successful buildup curve analysis based on Homer plot method. The line must be identified to estimate reservoir permeability, to calculate skin factor, and to estimate static drainage area pressure. A log-log graph of the pressure change, pws -pwf (Q), in a buildup test vs. shut-in time t presents a good estimation of where the straight line portion, or middle time region, begins (Fig. 30. 1O).9 A log-log graph of pressure change pws -pwf vs. At I6 is an even more diagnostic indicator of the end of afterflow distortion. Fig. 30.11 shows a semilog plot of theoretical buildup test data. The use of type curves has greatly improved identification of the straight line portion of the buildup curve after wellbore storage effects have ended. HomerlO plotted buildup pressure vs. (t+At)lAt on semilog paper with (t+At)lht on the log scale where t=total producing time since well completion, hours, and At=time since the well was closed in, hours. Extrapolation of the curve to a value of (t + At)/At = 1 is the approximate static pressure.
30-10
PETROLEUM ENGINEERING
HANDBOOK
SEYKOG
Fig. 30.11-Plot
of typical
ti
LOG t
t2
The value of I is determined by dividing the cumulative production of the well by the rate of production per hour at the time the well is shut in. The uncertainty of the value oft increases with the age of the well. Experience indicates that using the time of the flow period before the well is shut in for the buildup test is often more reliable than using the total time since completion provided fully stabilized conditions exist both when the well is opened and when it is shut in. Thomas, l1 with Homer s basic equation, preferred using the reciprocal of (t+At)lAt and therefore plotted pressure vs. At/(l+At) on semilog paper with Atl(t+At) on the log scale, and extrapolated the curve to a value of Atl(t+Ar)= 1, which gives the approximate reservoir pressure. Hurst * plotted the buildup pressure vs. the shut-in time in minutes on semilog paper with time on the log scale. A straight line is drawn through the points and extrapolated to an intercept time value of 1. The slope is the value of pressure change in one log cycle. Expressing his equation in English units, the static pressure is calculated:
p,,,=b+mlog625m, ... .. ....... . . . . (8)
curve is frequently masked by one or more factors, such as skin effect, afterflow, and the early onset of boundary effects. Chap. 35, Well Performance Equations, addresses these problems. The capacity of a well to produce can be estimated from the BHP drawdown on a flow test. For gas wells, the open-flow capacity is calculated by the procedure proposed in Ref. 5 except that both the static pressure and the flowing pressure are measured with a BHP gauge. Open-flow capacity of oil wells having large capacity and high pressure is rarely of value, and little work has been reported on such determinations. Theoretically, so long as single-phase flow maintains, the rate of flow should increase in proportion to the drawdown pressure. But, because of gas coming out of solution below the bubblepoint, turbulence in the flow, and borehole restrictions, the flow conditions change and the proportionality does not hold. Engineers who have investigated the problem usually consider that little increase will be obtained after the drawdown pressure is one-half the static pressure. For low-pressure wells, the rate of production will usually continue to increase until the flowing pressure is equal to or close to atmospheric pressure. Productivity Index PI is defined as the barrels of oil produced per day per pound decline in BHP. To determine the PI, a well is shut in until static or reservoir pressure is reached. The well then is opened and produced until the BHP and rate of production are stabilized. Since a stabilized pressure at surface does not necessarily indicate a stabilized BHP, the BHP should be recorded continuously from the time the well is opened. The PI is then
.... ... (9)
where pas = static BHP, psi, m = slope of buildup curve, and h = intercept of curve with time value of 1 minute. Correct identification of the straight-line portion of the buildup curve is necessary for the interpretation of pressure buildup data. The straight-line portion of the
where
J = PI, B/D-psi, qO = rate of oil production, PWS = static BHP, psi, and
B/D,
BOTTOMHOLE PRESSURES
30-l 1
stabiliza-
Specific PI is defined as barrels of oil produced per day per pound decline in BHP per foot of effective reservoir thickness, and is expressed as
J,= (p, " -Pwf)h, . . .., . . . . . . .. .. ..
and therefore the PI. Calculating a PI based on total fluid mass instead of liquid production has given more consistent results in some pools. Permeability Permeability of the reservoir rock can be calculated from the PI. Wycoff et al. I3 used
. (10)
where J,y is the specific PI and h is the effective reservoir thickness, ft. On a flow test the time required for a well to reach a stabilized BHP and rate of production, that is, the transient period, may require several hours or even days and occasionally several weeks. The duration of the transient period and the rate of the pressure decline and the PI decline during the transient period will indicate the quality of the reservoir. A short transient period indicates a high-quality reservoir, and a long transient indicates a low-quality reservoir which will have a comparatively low recovery of the amount of oil calculated to be in place. Reservoir quality is not related to the numerical value of the PI. The nature of the transient period is most conveniently expressed by plotting the productivity index by hours on log-log paper. Typical well test data are presented in Figs. 30.12 through 30.15. The negligible transient period of the test shown in Fig. 30.12 indicates a high-quality reservoir from which relatively high recovery may be expected. The short transient period in the flow test in Fig. 30.13 indicates a comparatively high recovery of the original oil in place. In Fig. 30.14, the transient period is quite long but the continuous flattening of the slope of the PI curve indicates eventual stabilization. Fig. 30.15 is the flow test of a well in which the continuous decline in pressure, production, and PI shows that the flow will not stabilize and therefore the ultimate recovery will be lower than that reasonably expected from a consideration of the producing formation and its apparent productivity. Flow tests and PI tests are conducted by other procedures. Some prefer to run the test at two or more different rates of flow. There is usually some difference in the PI at different rates, sometimes more than can be accounted for by inherent limitation of accuracy of pressure measurement and production gauging. Changing GOR or WOR will affect the relative permeability
=325 J,pB
log=, rw
..
. . . . . . . . . . . . . (11)
where
k = rd = rw = q = p = B = h = Pws = p,,,, =
permeability, md, radius of drainage area, ft, wellbore radius, ft, production rate, B/D, viscosity of produced fluid, cp, formation volume factor, RBISTB, effective reservoir thickness, ft, static BHP, flowing BHP, and J, = specific PI.
Permeability can be calculated from the buildup curve obtained when a producing well is shut in following a flow test. Muskat discussed this in 1937 and presented an equation in 1949. l4 The equation in commonly used units is 40.37 wui*pB log(rdlr,.) hy where m =slope of log (p, -p,,,,) vs. t, d=diameter of flow pipe, in., t=time, hours, since well was closed in, and other parameters are as previously defined. The equation has limitations, but it may be helpful in cases of severe reservoir damage such as interpretation of DST buildup curves where recovery is low.
k=
, ,,..........,..
(12)
30-12
PETROLEUM ENGINEERING
HANDBOOK
HOURS
Fig. 30.13-Flow-test
from a BHP
ko=
162.5qo~oBo hm ( .,,...._...........~..
(13)
where k, = effective oil permeability, md, k = permeability, md, 40 = oil production rate, B/D, = viscosity of reservoir oil, cp, i = formation volume factor of oil RB/STB, i = effective reservoir thickness, f;, and m = slope of buildup curve. The slope m can be determined most conveniently by plotting BHP against time in hours on semilog paper, with time on the log scale. The initial part of the curve is affected by the afterflow into the wellbore, and the last part of the buildup curve may be too flat and therefore unreliable because of interference of drainage areas or limited reservoir. The calculation is simplified by extrapolating the buildup curve to encompass one complete cycle of the log scale. The slope is then the difference in the pressure reading at the beginning and at the end of the cycle. The part of the curve representing the slope of the buildup curve is usually evident, but occasionally interferences and irregularities in the reservoir make the slope uncertain. We consider that if the value
0.0002637kt
falls to the range 10 ~ r to 10 p2, the slope m is proper and the calculated value of k is valid. This tacitly assumes that the values of the other factors are known and that the conditions exist for which the equation was derived. A valid calculation of permeability from buildup curves requires stabilized conditions of pressure and rate of production at the time the well is closed in. These equations for calculating permeability are based on liquid single-phase flow. When a small amount of free gas is produced with the liquid, adjustment by the relative permeability will give an acceptable answer. When gas only is produced from the reservoir, the equations for permeability are as follows. Radial flow equation:
k=
. (15)
Pressure buildup t5 :
k=
.(I61
where md, = rate of production, Mcf/D at 14.7 psi and q&T 60F, = viscosity of reservoir gas, cp, Fi = reservoir temperature, F+460, z = gas deviation factor, reservoir conditions, h = effective reservoir thickness, ft, PWS = static BHP, psia, pwf = flowing BHP, psia, rd = drainage radius, rw = wellbore radius, and, mg = slope of pressure buildup curve, pw 2 vs. log t, where t=time after shut-in, hours, pw =BHP at t.
k = permeability,
m=
,,.__..,.................
(14)
4ct,rd2p where md, t = time, hours, from closed in to end of straight-line portion, 4 = porosity, fractional, Cl. = liquid compressibility, psi - , p = viscosity, cp, r d = drainage radius, ft,
k = permeability,
BOTTOMHOLE
PRESSURES
30-13
HOURS
Fig. 30.14-Flow-test
15-w
DRAW-DOWN
PRESSURE
-e-t
I bl i 1 hi
/ II1
I 1 /PRODUCTION
I
I
tttl'
1w IO HOURS
Fig. 30.15-Flow-test
Permeability Damage The permeability of the reservoir adjacent to the borehole is frequently reduced by the invasion of drilling mud, water blocking by invasion of filtrate water from drilling fluid or other source, swelling of clay particles, or deposition of salts or wax. Blinding or partial clogging of screens or perforations will give a similar effect. A substantial restriction of flow can be observed on the BHP chart by the straight-line buildup until near the maximum pressure instead of the normal exponential shaped curve. On the other hand, permeability adjacent to the borehole may be increased as a result of acidizing, fracturing, or shooting. The amount of damage or improvement to the permeability adjacent to the borehole is determined by comparing the permeability calculated from flow test with the permeability calculated from the buildup and is expressed as the productivity ratio:
where
FP = productivity ratio, kj = permeability calculated from productivity
index, and kb = permeability calculated from buildup pressure. Substituting the equated expressions of kj and kb and simplifying:
F, = 2m log(rdlrw) P ICS -P wj . ........ .. . . . . . . . (18)
A fractional Fp indicates restricted flow at the borehole and an Fp greater than unity indicates better permeability at the borehole, usually the result of stimulation. Dolan et al. I6 presented an empirical equation to determine damage factor that does not involve the use of the amount of production. It is of particular value for interpretation of drillstem test charts that have an acceptable buildup but negligible fluid recovery. His equation expressed as productivity ratio is
4 Fp= o.183(p, -pw~), . ..... . . . . . (19)
30-l 4
PETROLEUM
ENGINEERING
HANDBOOK
where
FP = productivity ratio, = static BHP, P W S pti = flowing BHP, and
Ap = slope of plot ofp vs. log (t+AtlAt), where t = time well was open, minutes, At = time after well was shut in, minutes, and p = pressure at t+At. Hurst I2 restrictions borehole, pressed in for oil: has developed eq uations for calculating to flow through the reservoir adjacent to the which they have called skin effect. Exoilfield units, these equations are
It is usually difficult to delineate specific heterogeneities from well test results only because different heterogeneities may cause the same or similar well test response. A higher degree of confidence is achieved when the interpretation of test results is confirmed by geological and geophysical evidence of heterogeneities. Linear discontinuities, faults, and barriers affect a pressure buildup behavior and are manifested by a second straight line, the slope of which is double the initial straight line. For a well near a linear fault, drawdown testing can be used to estimate reservoir permeability and skin factor in the usual fashion, as long as wellbore storage effects do not mask the initial straight-line section. If the well is very close to the fault, the initial straight-line section may end so quickly that it will be masked by wellbore storage. As the drawdown proceeds and the pressure at the producing well falls below the initial semilog straight line, the following equation indicates that ~,,f vs. log t plot will have a second straight line portion with a slope double that of the initial straight line.
PS =m
(Pr -Pwf>
171
-log
93,
...
( 10.4mhqbc,r,,.
>I .. .
(201
pi+m
0.86859s+log
. . (22)
Mg
m=
-log
Plhr
(21)
where
Ps = pressure loss due to skin effect, psi, Pws = static BHP, psi, Pwf = flowing BHP, psi,
= pressure on straight-line portion of semilog plot one hour after beginning a transient test, psi, pi = initial reservoir pressure, psi, s = skin effect, and L = distance to a linear discontinuity, ft.
Pt = 40 = qg =
B, =
z= h= TR = 4= co = rw = m= mg =
well pressure 1 hour after shut-in, psi, production rate for oil, B/D, production rates for gas, Mcf/D at 14.7 psia and 60F, formation volume factor of oil, RBISTB, compressibility factor (gas deviation factor), effective reservoir thickness, ft, reservoir temperature, F+460, porosity reservoir rock, fraction, compressibility of reservoir oil, psi - , borehole radius, ft, slope, pt vs. log t, and slope pt2 vs. log t, where pt is the well pressure at time t and t is time after well shut-in, hours.
However, the simple occurrence of a doubling slope in a transient test does not guarantee the existence of a linear boundary near the well. To estimate the distance to a linear discontinuity, we use the intersection time, t,, of the two straight line segments of the drawdown curve. The following equation applies for drawdown testing. I7
(23) where c, is compressibility of the total system, psi- The effects of reservoir heterogeneities and method of determining them are discussed in Chap. 26. Lifting Equipment BHP measurements are valuable to the engineer for determining the size and type of artificial lift to install and to monitor the efficiency of such equipment. The efficiency of a gas lift depends, among other factors, on the pressure at the depth of gas injection. From
Positive values of ps indicate damage and negative values of pr indicate improvement of permeability adjacent to the wellbore in terms of pressure loss due to skin effect.
BOTTOMHOLE
PRESSURES
30-15
the flowing pressure or PI, the pressure can be calculated for a given rate of production and after a gas lift is in operation. Knowing the PI, the producing efficiency can be calculated from the rate of production. The change in pressure gradient in a well being gas lifted will determine the point at which gas is entering the flow string, which is of interest where multiple flow or unloading valves are installed. The PI is used to determine the amount of fluid available to be pumped from a well and therefore the size of the pump that should be run and the depth at which it should be set. After a well has been on the pump for some time and declined in production, the question frequently arises as to whether the productivity of the well has declined or the pumping equipment has decreased in efficiency. The correct condition can be determined by measuring the static and producing pressures, either with a bottomhole gauge or from fluid-level calculation by sound reflections as described previously. If fluid-level determinations are known to be unreliable for a given pool or are proved uncertain in a given well because of fluctuating fluid-level by several determinations, a BHP gauge should be run. Many operators have successfully run a pressure gauge in the annulus between the tubing and casing. During such runs the wireline will sometimes wrap around the tubing or the gauge will wedge between the tubing and casing. It can usually be released by starting the pumping equipment, but sometimes it is necessary to move in a pulling unit to lift the tubing to free the gauge. Many operators use an eccentric tubing head to position the tubing against one side of the casing instead of its normal central position and thus minimize chances of the gauge s becoming hung up in the annulus. The BHP gauge can be run on the rods in most wells by attaching it, preferably rigidly, to the standing valve. A pressure gauge under these conditions is subject to severe vibrations resulting from both the vibration of the pumping equipment and a waterhammer effect from the liquid at the pump level. Drillstem Tests The most common use of BHP gauges in a drilling well is in evaluating DST s. Pressure gauges were first used to determine that the valve functioned properly and was open during the test. Subsequently the pressure information has become important. The detailed pressure information obtained on a DST can be used to determine the productivity of the formation by calculating the PI from the amount of fluid recovered during the test and the static and drawdown pressures. A more recent use has been to confirm indicated productivity from the buildup after the valve is closed. It is not unusual to have a very poor recovery of fluid and obtain a good buildup curve such as can be obtained only when the producing capacity is much better than indicated by the quantity of fluid recovered. The permeability can be calculated from the buildup curve. While subject to a high probable error because of the short time of the test and usually high reservoir damage adjacent to the borehole, it is sufficient to indicate whether further testing may be warranted. Mud Weight The pressure at the depth of a DST, before seating the packer and also after the test is completed and the packer
unseated, permits calculating the average weight of mud in the hole and is used to verify mud weight as measured by routine tests.
Nomenclature
b = intercept of curve with time value of 1 minute E = formation volume factor for total fluid produced B, = formation volume factor of oil CL = liquid compressibility CO = compressibility of reservoir oil ct = compressibility of the total system C, = gas compressibility CT = temperature coefficient d = diameter of flow pipe d,. = deflection of calibration temperature d, = observed deflection d, = deflection at Ti for given pressure d2 = deflection at T2 for same pressure D= depth of well e= base, natural logarithms F, = gauge factor FP = productivity ratio h= effective reservoir thickness J= PI J, = specific PI k= permeability of reservoir kb = permeability calculated from buildup pressure kj = permeability calculated from productivity index k, = effective permeability of oil L= distance to a linear discontinuity and unstrained length AL= change in length m= slope of buildup curve, pt vs. log I or slope of log(p,, -pwf) vs. I ml* = slope of pressure buildup curve, pm, vs. log t P= pressure at r+Ar Ap = slope of plot of p vs. log t+ Al/At or
AP=P~~-P~, Lcp = dimensionless factor from curves p* = last pressure on buildup curve Pi = initial reservoir pressure PA = pressure loss due to skin effect P.sd = log e rd/r,,, for constant pressure at radius of drainage or log c (rd/r,.)-0.75 when
pKh = wellhead pressure pH.,, = static BHP = pressure on straight-line portion of semilog Plhr plot one hour after beginning a transient test
30-16
PETROLEUM
ENGINEERING
HANDBOOK
4= 4K qn = =
r=
Ar = t-d =
rw = .Y= t= t=
At = t, = T=
T,. = 7-o = T, = T, = T, = z=
production rate at shut-in rate of production rate of oil production unstrained resistance change in resistance drainage radius wellbore radius skin effect time time from closed in to end of straight-line portion of buildup curve time after well was shut in intersection time of the two straight line segments of the drawdown curve average temperature in the borehole calibration temperature observed temperature reservoir temperature lower temperature higher temperature compressibility factor (gas deviation factor) specific gravity of gas (air=l) viscosity of produced fluid viscosity of reservoir gas viscosity of reservoir oil porosity reservoir rock
m=
kt +ct,rd CL
...
(14)
254.359q,p,TRz k= h(p,&p,,f2)
log ...
(15)
k=
127.2q,~KTRZ hm,
.(16)
(20)
ps =mR
Pt-P4 m<s
log
142.817q,zTR
$hm fic 8 r II ,2j
il
.(21)
L=
kt, ,,78,~c,~,
(23)
Y,D
67.37T
.........
(5)
in in in in in in in in in in in
p ,, =p*+ I,I
(P v/ -MqbB
2ahk
(7)
References
I. Millikan, C.V.: Bottom-Hole Pressures, Petroleum Production Hundbook (1962) 2, 27-l-27-14. 2. Brownscombe. E.R. and Cordon, D.R.: Precision in BottomHole Pressure Measurement. Trans., AIME (1946) 165, 159-74. 3. Bergman, J.C.. Guimard. A., and Hagernan, P.S.: High Performance Pressure Measurement Systems, Johnston-Macco (1980) 10. 4. Pierce, H.R. and Rawlins. E.L.: The Study of Fundamental Basis for Controlling and Gauging Natural Gas Wells, Part I , RI 2929, USBM (1929). 5. Buck-Pressure Data on Natural Gus Wells and Their Application to Pmductiotz Pructices, USBM Monogrdph 7 (1935) 168. 6. Muskat, M.: Use of Data in the Build-Up of Bottomhole Pressures. Trans., AIME (1937) 123, 44-48. 7. Arps, J.J. and Smith, A.E.: Practical Use of Bottom-Hole Pressure Build-Up Curves, Drill. and Prod. Pruc., API (1949) 155.
kz
. 27l
(II)
2*4p,,,
-P,,t.)
44.330 dtLB
k= hY
log . .
(12)
k,, = 4OtcOB~,
5,45751,n7
(13)
BOTTOMHOLE
PRESSURES
30-17
8. Miller. C.C.. Dyes. A.B.. and Hutchinson. C A. Jr.: Estimation of Permeabihty and Reservoir Pressure from Bottomhole Preasurc Bmld-Up Charactenstics, Trum., AIME (1950) 189. 91-104. 9. Earlougher, R.C. Jr., Kench, K.M., and Ramey, H.J. Jr.: Well&e Effects in Injection Well Testmg,J. Pet..Tech. (Nov.
1973) 1244-50.
13. Wycoff, R.D. er 01.1 Measurement of Permeability of Porous Media. Bull.. AAPG (1934) 18. 161. 14. Muskat, M.: Physical P&r&s ofOil Production. McGraw-Hill
Book Co. Inc., New York City (1949).
10. Homer, D.R.: Pressure Build-Up in Wells. Proc.. Third World Pet. Gong. (1951). 11. Thomas, G.B.: Analysis of Pressure Build-Up Data, J. Pet. Tech. (April 1953) 125-28; Trans., AIME, 198. 12. Hurst, W.: Establishment of the Skin Effect and Its Impediment to Fluid Flow into a Wellbore, Pet. Eng. (1953) 25. No. 11, B-6.
15. Tracy, G.W.: Why Gas Wells Have Low Productivity, Oil and Gas .I. (Aug. 6, 1955) 54, No. 66, 84. 16. Dolan, J.P., Einarsen, C.A., and Hill, G.A.: Special Applications of Drill Stem Test Pressure Data, J. Pet. Tech. (Nov. 1957) 318-24; Trans., AJME, 210. 17. Gray, K.E.: Approximating Well-to-Fault Distance fmm Pressure Build-Up Tests, J. Pet. Tech. (July 1965) 761-67.
Chapter 31
Temperature in Wells
G.J. Plisga,
Sohio Alaska Petroleum Co.*
Introduction
Frequently, when working with a wildcat or deepening old production into zones that are relatively unknown, it becomes necessary to know the underground temperatures expected at a predetermined depth. In the 1920 s, API Research Project 25 investigated the relationship between geothermal gradients and the geologic structures of oil fields. Initially, interest in subsurface temperatures in oil fields focused on high temperature in deep wells, which caused cement to take initial set before it was placed behind the casing. More recently, cements with low hydration heat have been developed to protect permafrost intervals in northern frontier areas. Temperature surveys in wells were used to determine the top of the column of cement behind the casing.2 With continual development of temperature devices, the reliability, accuracy, and speed of response have opened new horizons to temperature logging. Temperature logs are used currently to identify fluid entry into the wellbore, fluid migration behind the casing, tubing/casing leaks, and the extent of hydraulic fracturing and to monitor injectivity profiles. a piston, which extends through a packing gland against a tense helical spring. A stylus arm attached to the end of the piston extends into the cylindrical chart holder of the recording mechanism. The temperature range may be changed by varying the diameter of the cylinder and piston. To prevent the well pressure from affecting the temperature element, the mercury container is enclosed within an outer tube, which is filled with mercury to reduce thermal lag. With reasonable care in calibration and operation, temperature readings are accurate to 2F and differential temperatures of 0.5F can be read. Amerada Gauge Temperature Element. The temperature element for an Amerada gauge is a pressure element with a bulb attached to the pressure end of the helical Bourdon tube, but thermally insulated from the gauge to reduce thermal lag. The bulb contains a liquid that has a substantial vapor pressure in the temperature range of interest. For various temperature ranges, different liquids and different ranges of helical Bourdon tubes are selected, preferably in such a combination that the maximum temperature range is near the critical temperature of the liquid, which gives maximum defection per degree change in temperature. Ranges of approximately 120 and 200F are used most frequently, with the minimum or maximum temperature as requested on the order. For maximum chart readability, the span between the minimum and maximum should be no greater than 200F. Thermometers with a range of 120 to 2OOF, as ordinarily calibrated, have a sensitivity of about 0SF and temperature changes of 0.1 F can be detected. The absolute accuracy of the Amerada temperature gauges is +2F. The time required for thermal equilibrium is 20 minutes, but some 70% of the change in temperature will be recorded in 30 to 45 seconds when the instrument is immersed in liquid. A faster-responding gauge design, which increases the
Thermometers
Self-Contained Recording Thermometers Self-contained recording thermometers, as used in the oil fields, use the same mechanism to record as the bottomhole pressure (BHP) gauge, with a thermometer element substituted for a pressure element. Temperature elements are made for some of the commercially available BHP gauges. Humble Gauge Temperature Element. The temperature element for the Humble gauge is a container filled with mercury. With an increase in temperature, the mercury expands into a small-diameter cylinder at the end of
Author of the original chapter on this topbc m the 1962 edltm was C.V. Mllhkan
31-2
PETROLEUM ENGINEERING
HANDBOOK
temperature sensing area and reduces the heated mass, reaches thermal equilibrium in 8 to 10 minutes. The response of a liquid-vapor element is not a straight line, and therefore the accuracy and sensitivity of the element depend on the temperature to be measured. Time Response. The time response of a thermometer to a change in temperature is directly related to the rate of movement through fluid, or flow of fluid past the thermometer. When a long section of hole is to be surveyed for a possible anomaly, the thermometer can be run at 50 to 100 ftimin followed by a second nm at 2 to 5 ftimin through intervals of interest indicated on the first run. Thermometers in Gas. The thermal conductivity of a gas is much lower than that of liquid. Therefore, a thermometer in gas has greater thermal lag for a given change in temperature. However, in most wells in which a thermometer is run through gas, any anomaly present is caused by expansion of gas, and the change in temperature is much greater than normally found when the anomaly is caused by migration of liquids. Because of the greater change in temperature, the presence of an anomaly is recorded as quickly in gas as in liquid. If a wellbore contains gas through the interval of interest and no gas expansion is present, such as a survey to determine migration of fluid behind the pipe, it is preferable to fill the well with liquid. If this is not feasible, then the thermometer must be run much slower (one-fifth to onetenth) than the normal rate in liquid. Electrical Surface-Recording Thermometers
vantages. Self-contained thermometers have the advantages of portability and low investment. Disadvantages are much greater thermal lag and the necessity of retuming the instrument to surface and reading the charts before the results are known. Electrical surfacerecording thermometers have the advantage of quick response to temperature change, which permits running faster, plotting temperature against depth as the survey progresses, and checking any temperature anomaly without having to recover the instrument from the well. Disadvantages include a much greater investment, larger and heavier equipment, and delicate instrumentation.
Thermometry
Introduction Two types of temperature surveys are used in the oil fields. One determines the true temperature at the depth of interest, and the second determines the depth or interval of a change in temperature. Usually, true temperature is measured with a maximum-recording thermometer. However, the use of electrical surface-recording thermometers is becoming more widespread for measuring true temperatures. Determination of a change in temperature requires a continuous record some distance above and below (as well as through) the interval of interest. In this case, the use of a differential thermometer will show the existence of a change in temperature rather than the true temperature or the actual magnitude of the change. Actual Temperature The geothermal gradient is very different in the various sedimentary basins, but within a given basin the change is gradual from one part to another.3b4 In most oilproducing areas, the gradient is usually within the range of 1 to 2F increase for each 100 ft of depth. Temperatures just below the seasonal effect, ordinarily 30 ft below the surface, are about 1.5F higher than the isotherms of the average annual temperature, as shown by climatological data of the U.S. Weather Bureau (Fig. 3 1. I). An exception to this is in northern latitudes where continuous permafrost exists. A geothermal gradient determined from such surface temperatures and the temperature of the producing formation is sufficiently accurate for most practical uses (Fig. 31.2). The rate of temperature increase in some areas is greater with depth, especially below 10,000 ft, and marked increases have been reported below 18,000 ft. When a precise geothermal gradient is to be determined, the hole selected must not have been disturbed for several months. In any event, the survey must be conducted while the well is in operation, since the passage of the thermometer will alter the static gradient. The thermal conductivity of geological strata varies. The average heat conductivity of common sediments is given approximately by the figures in Table 3 1.1. 4 When fluid movement continues in a borehole for periods of time, such as in drilling operations or a producing well, the temperature effect will be different on each formation. Any effect on a given formation will depend on its thermal conductivity, the difference in temperature between the moving fluid and the formation, and the length of time such movement continues. When fluid movement is stopped, temperature equalization
Electrical surface-recording thermometers have a therresistance wire, or thermistor as a mocouple, temperature element. As normally calibrated for oil well use, electrical surface-recording thermometers have a sensitivity of 0.5F and a thermal lag of only a few seconds. They are run on armored, insulated cables and the measuring wheel is geared to drive a chart recorder, camera, or computer to record temperature against depth. Differential thermometers have been developed that record very small changes in temperature, 0.1 F or less, and are useful for identifying an anomaly in a long section when surveying at logging speeds of 100 to 150 ftimin. Once an anomaly is recorded, the thermometer can be run at slower speeds to completely define the anomaly. The differential thermometer is usually run in conjunction with an electrical thermometer to allow the absolute temperature to be measured in conjunction with the differential temperature. By using electrical surface-recording thermometers, any temperature change noted can be checked by a rerun without returning the instrument to the surface. Very small anomalies under static conditions may be disturbed by the movement of the instrument, and therefore when a check of such condition is run, it should be delayed long enough to reestablish temperature equilibrium in the hole. Advantages and Disadvantages Self-contained thermometers and electrical surfacerecording thermometers each have advantages and disad-
TEMPERATURE
IN WELLS
31-3
Fig.
31.1-Average annual temperature, OF, for the period 1899 to 1938. lsolines are drawn
through points of approximate equal values.
begins, but considerable time, usually several months, is required to approach temperature equilibrium. In temperature surveys of wells, such temperature irregularities can be confused with an anomaly caused by some operating condition. Normally, however, the irregularities in the gradient resulting from normal operating conditions are small and the abnormal condition being investigated, such as a hole in the casing, fluid migration behind the pipe, or cement top, is of such size
character that there is no uncertainty as to cause. The actual temperature at depth is very important in many problems in drilling, production, and reservoir work. Drilling mud is often adversely affected by high temperature. The type of cement and additives are determined by the temperature at the casing seat or zone of interest. In oil reservoirs, the amount of gas in solution, the bubblepoint, and the viscosity are all related to the temperature, as is the amount of condensate formed and
or
SURFACE
TEMP.
Fig. 31.2-Contour
31-4
PETROLEUM ENGINEERING
HANDBOOK
TEMPERATURE
INCREASES
36 ;02 16 12
Shale
the amount that remains as a liquid wetting the reservoir rock. The volume of gas per unit of reservoir rock and the supercompressibility of the gas are also related to the actual temperature. Temperature Surveys
Fig. 31.3-(a) Example gradient gradient fluid flow. gas flow. (b) Example
A temperature survey of a well is made either by running the thermometer continuously at a slow speed or by stopping it for a short time at regular intervals. For a survey through a long interval of hole, a continuous run is often preferred, while for a short interval numerous stops of 1 to 2 minutes are made. Temperature readings should be taken through the interval of interest, while running the thermometer in and while pulling up. Both runs should be at the same speed or at the same stops to determine more accurately the depth of any anomaly. On runs made very slowly, 2 to 5 ftimin, the actual thermal lag may be too small to warrant surveying the opposite direction. When a survey is started at a given rate, that rate should not be changed during the survey. To do so will cause a change in gradient or anomaly on the chart that may mask an actual anomaly or change in gradient in the well. Since the temperature chart is not available until the survey is completed and the thermometer is removed from the well, thorough notes that record time and depth of the instrument are required to correlate temperature and depth. The location of a temperature change on a temperature survey is, in most cases, much more significant than the actual temperature or the amount of change in temperature. Normally, the temperature gradient in a well is reasonably uniform, and any deviation is indicative of an abnormal condition at that depth. The deviation may be an irregularity or anomaly in an otherwise uniform gradient, or it may be merely a change in the gradient. The primary causes of temperature change in a wellbore are expansion of gas, hydration of cement, and migration of fluid. Expansion of liquids in producing wells and heat of solution and chemical reaction are often present, especially in drilling wells, but the net effect on the temperature in the borehole is normally too small to be recognized. Gas expansion will cause an anomaly on the gradient, and migration of fluid will cause a change in gradient. Gas expanding as it enters the borehole from the reservoir formation is much cooler than the adjacent formations, and therefore the particular intervals from which the gas is flowing can be identified by a temperature survey (Fig. 3 1.3A). A typical temperature survey conducted in a well producing minimal or no gas can be identified by a temperature survey (Fig. 3 1.3B). Greater detail will be recorded if the thermometer is run across the open hole or perforated sections while the well is pro-
ducing. When the tubing tail is below the producing interval, any gas produced that travels downward to the bottom of the tubing, then up the tubing past the thermometer, mixed with any other fluid that may be flowing into the wellbore, will mask anomalies that exist at the producing interval. With this type of completion, the tubing must be shut in and the annulus allowed to produce with the thermometer being run in the tubing. If a packer is set or, for other reasons, the well cannot be produced through the annulus, the survey can be run after the well is shut in. A temperature survey on a shutin well relies on gas cooling by expansion to lower the temperature of that part of the formation producing gas below that of the formation not producing gas. Some time is required for the temperature to equalize, and a temperature survey run long enough after shut-in to complete afterflow will record lower temperatures in the principal intervals of gas production. A sequence of runs through the interval will permit a more reliable interpretation. The lowest interval producing gas can be identified from a temperature survey by the cooling effect through the intervals producing the free gas and the return to the normal gradient below. Water-producing intervals can be determined only if enough free gas is produced with the oil to give a change in gradient at the lowest interval producing oil. A hole in the casing through which fluid is moving can usually be found by running a temperature survey. In addition to the depth of the hole in the casing, the depth of the formation, which is the source of the migrating fluid, and the depth of the formation into which the fluid is moving must be determined when a hole is known or suspected to exist. A permeability or injectivity profile of a waterinjection well can be determined from a temperature survey. A typical temperature survey of a water-injection well with a homogeneous formation shows a cooling across the injection interval (Fig. 31.4). Two procedures are used to obtain injection-well temperature surveys. After injecting water for a period of time, injection is discontinued and, after a few hours, the survey is run. A more reliable answer would be expected from a series of
TEMPERATURE
IN WELLS
31-5
TEYPERATURE INCREASES
TENPERATURE
INCREASES
INJECTING
WATER
2 HOUR SHUT - IN
STATIC
\ - -,
YJECTION ZONE \ \ \ \ 1
12 HOUR SHUT-IN
PRODUCING
ZONE
MELED
ZONE !
Fig. 31.4-Water-injection
gradient.
Fig. 31.5-Fluid
channel gradient.
surveys at intervals of a few hours. Continued injection for many months can cool the entire formation and make it difficult to identify the relative injection capacity of the different parts. Under some conditions, another method that may give more detail is to discontinue injection for a day or more, then run a survey while injecting water at a very low rate, such as 1 to 5 bbl/hr. Because of the residual variations of temperature from normal waterinjection operations, a survey before injection is recommended for comparison. If water were channeling from below, the temperature survey would show a warmer anomaly at the base of the producing interval (Fig. 31 S). Fig. 31.6 shows idealtemperature curves for various conditions of migration of fluid through a hole in the casing. Certain assumptions were made in drawing these ideal curves. Where gas is migrating, some expansion and, therefore, cooling is presumed as the gas leaves the formation. Also, it is assumed there will be a drop in pressure and, therefore, expansion and cooling at the depth of the hole in the casing. If there is no expansion at either point, the curves for gas would have the same appearance as the curves for
liquid. In cases where both formations are either above or below the casing hole, the movement of fluid from the hole to the farthest formation will mask any gradient between the hole and the nearest formation. There are times when the volume of migrating fluids is not sufficient to affect the temperature. Gas leaking through a small hole in the casing will cause a very sharp temperature drop, but the volume may be too small to affect the gradient. Migration of fluids behind the casing will create a lower rate of temperature change than the normal gradient. If the flow is upward the gradient temperature will be higher than the normal gradient, and for downward flow the gradient temperature will be lower. Preparation of the well is essential for a successful and reliable interpretation of the survey results. Usually the maximum anomaly is of the order of 2F, and under less than optimal conditions the anomaly may be so small or so masked that a reliable interpretation is impossible. In these instances, the use of an electrical surface-recording differential thermometer may be the only method of obtaining a successful survey.
Gradient
from which fluld is migrating into which fluld is migrating upward upward downward downward downward downward to casmg hole, then up to cas.mg hole. then up
(C) Gas mtgrat,ng (0) Liqwd migrating (E) Gas rmgratmg (F) Llquld migrating (G) Gas m,grat,ng (H) Liquid migrating
Fig. 31.6-Ideal-temperature
31-6
PETROLEUM ENGINEERING
HANDBOOK
Casing-leak and channeling-of-fluid temperature surveys are run with the well shut in. The shut-in time must be long enough for the entire wellbore to approach an even temperature gradient. A minimum of 24 hours should be allowed. However, the shut-in time required should be determined from past experience, type of problem to be identified, and location. Gas production and water production intervals are located with the well flowing at the maximum practical rate. If the survey cannot be run while flowing the well, the time interval between shutting in the well and running the survey may be critical. All afterflow must have ceased, but not so long as to allow the temperature to approach the normal gmdient, which would mask any anomaly. A temperature anomaly can be created by injection of fluid, usually oil or water, when normal conditions will not give a temperature change. Examples of this are identifying a channel below the producing interval and locating a packer or tubing leak by pumping cool fluids. Drilling Wells In drilling deep wells or wells that encounter high temperatures, especially in excess of 250F, a xepresentative bottomhole temperature (BHT) is required for selection of a proper mud program, cement, and additives to ensure proper cementing of the casing. In development drilling, temperature data from neighboring wells can be used in estimating the BHT of a new well. In exploration drilling, an estimation of the BHT may be all that is available. Unless offset data are available a
geothermal gradient must be assumed that will permit estimation of the BHT. A technique exists for determining static BHT s by plotting on semilog paper T,, vs. (tk + At)/At where
T ws = bottomhole shut-in temperature measured at
At, OF, tk = circulation time, hours, and Ar = time after circulation ceases or shut-in time, hours.
The data required for this graph usually are obtained on successive openhole logging runs and allow an approximation of static BHT.5 Although this Homer-type analysis is not mathematically correct, when assuming short circulating times the technique provides reliable estimates of static temperature. This technique is most applicable in regions of high geothermal gradient, where log-recorded temperatures can be significantly lower than the static temperature. The hydration of cement is an exothermic reaction, and sufficient heat is generated that the presence of cement behind a string of casing can be determined by a temperature survey for up to several days after cementing. The character of the anomaly at the top of cement in a particular field is fairly uniform but varies greatly in different fluids. The anomaly may be a large, sharp increase (Fig. 31.7A) in some cases 35 to 45F, or it may be a very slight increase in gradient (Fig. 31.7B).
TEMPERATURE
INCREASES
>
TEMPERATURE
INCREASES
-t
TEMPERATURE
IN WELLS
31-7
The principal influence on the survey is the time elapsed between placement of the cement and running the survey. Other influential conditions include fineness of cement, chemical composition, rate of hydration, mass of cement in place, and the thermal conductivity of the adjacent formation. The maximum temperature usually occurs 4 to 9 hours after cementing, but reliable data can be determined in most areas after 48 hours. Any temperature change is affected more by the rate of hydration than by the total amount of heat liberated. Although hydration continues indefinitely, the rate decreases rapidly from the peak. A washed-out section of hole may be responsible for a large, sharp increase in temperature and can indicate a false cement top. A small temperature change or slight change in gradient could be caused by a small annular area or dilution of the cement with drilling mud. These factors, which influence the size of the temperature anomaly at the top of the cement in a given well, vary widely in their effect. However, even under an unfavorable combination enough heat is generated to permit a determination of the cement top. A new cased-hole logging method exists for detecting vertical flow outside the casing resulting from faulty cement. The radial differential temperature (RDT) log measures variations in temperature in the plane of the casing radius on the inside of the casing.6 Normally, two sensors are used, placed 180 apart; one sensor may be used at the wall of the casing and the other sensor in the body of the tool. An anchor spring at the top of the logging tools prevents the entire tool from turning as the sensors rotate. A motor rotates the tool at a speed of one revolution every 4 minutes. The RDT logging tools are designed to allow attachment of a perforating gun, which can be adjusted to perforate into the suspected channel or abnormality located by logging, If a channel is suspected in a perforated well, the well should be produced long enough to ensure that channel fluid is being produced before running the RDT log. The RDT sonde is placed at depths in the well where the channel is suspected. The arms are extended and the instrument revolves once or twice. Before moving to another depth the arms are retracted. As many measurements as required to delineate the channel can be made on one run. In some cases better results are obtained by injecting fluids at the surface to cool the channel. Temperature surveys can be used to locate depth of lost circulation in an area where formations above the depth of drilling are known to have taken fluid. The
temperature survey will show a sharp increase in temperature immediately below the point of loss of fluid. The temperature break will be even greater if slow losses are occurring while running the survey. At times when the hole is considered dangerous, the survey can be run through open-ended drillpipe over the suspected interval.
Summary
Wellbore temperature surveys are an inexpensive method to determine problem well conditions. The data obtained from a temperature survey are often the only data available and usually are accurate and reliable. When an anomaly occurs, one of these conditions must exist: (1) expansion of gas, (2) migration of fluid, or (3) some type of chemical reaction. With the exception of measuring the actual temperature at a point in a wellbore, temperature surveying is highly qualitative. In the majority of surveys, consistency of procedures, past experience, and the engineer s ingenuity allow reliable information to be collected and unique analyses to be performed. Since no quantitative relationship between temperature and depth exists that covers all areas and sedimentary basins, an assumed gradient of 1 to 2F/lOO ft depth is appropriate.
References
I. Heald, K.C.: Study of Earth Temperatures m Oil Fields on Anticlinal Structures. Bull. 205, API, Dallas (1930) I. The Economic Utility of Thermometric 2. Leonardon, E.G.: Measurements in Drill Holes in Connection with Drilling and Cementing Problems, Geophysics (1936) 1, 115. 3. Van Orstrand. C.E.: Normal Geothermal Gradient in the United States, Bull., AAPG (1935) 19, 79. 4. Nichols, E.A.: Geothermal Gradients in Midcontinent and Gulf Coast Oil Fields, Trans. AIME (1947) 170, 44-50. 5. Dowdle, W.L. and Cobb, W.M.: Static Formation Temperature From Well Logs-An Empirical Method, J. Per. Tech. (Nov. 1975) 1326-30. 6. Cooke, C.E. Jr.. Radial Differential Temperature (RDT) Loggmg - A New Tool for Detecting and Treating Flow Behind Casing, J. Pet. Tech. (June 1979) 676-82.
General References
Romero-Juarez. A. : A Simplified Temperature Changes in Deep Wells, 763-68: Trirrzs., AIME. 267. Method for Calculating J. Pet. Tech. (June 1979)
Smith. R.C. and Steffensen. R.J.: Interpretatwn files in Water-Injection Wells. J. Pet. Ted. Trmu.. AIME. 259. Wooley. G.R.: Computing Injection, and Production 1509-22.
Chapter 32
A potential test is a simple production test under stabilized flowing conditions to determine the ability of a well to produce. Potential or production tests are used to determine the rate of production through a given size of choke. This production rate and GOR then are used to determine if the well is capable of producing the assigned daily allowable production rates. Production-rate tests are conducted on a well so that its producing capabilities can be determined and a record of its producing abilities maintained. The results of these tests are used in diagnosing and evaluating a producing well. Their impottance cannot be overemphasized since they are used in every phase of reservoir and equipment analysis in which a knowledge of the productivity of the reservoir is essential. The test results aid in determining the parameters shown in Table 32.1. Production tests usually are considered part of routine field operations. Because they are performed on every type of producing well, a standard method of procedure that would cover every well cannot be set forth in detail. In many cases, the method or procedure for obtaining the results desired is left to the engineer s initiative and judgment and, in many cases, ingenuity. In every production test an initial understanding of the equipment employed and the method of completion, and a general knowledge of the producing reservoir and the results of previous tests on the well or on comparable wells will greatly simplify the testing procedure and aid in obtaining the desired results. Potential or production rate tests generally are required periodically by most state regulatory bodies in the U.S., such as the Texas Railroad Commission and the Louisiana Dept. of Conservation. The state regulatory bodies, authorized by the various states to control the production of oil and gas, set up a daily allowable or maximum rate at which each individual well may be produced. The well and reservoir conditions are considered
in setting up this maximum producing rate. Usually one other condition is involved in setting the producing allowable: the produced or re$ned oil and gas storage or market capacity. These regulatory bodies meet periodically and set a maximum number of producing days for a given period. The allowable is specified in two ways: the maximum amount of oil that can be produced each day and the maximum number of days each set period (usually 1 month) that this maximum daily allowable may be produced. It is a responsibility of the engineer or person in charge to set up and supervise the proper testing of all producing wells. When the production is controlled by state regulatory bodies, it is also the responsibility of the engineer or person in charge to find out the requirements set up by these regulatory bodies as to the proper method for testing and reporting the tests on producing wells. The importance of testing and information required to be supplied to the state regulatory agencies cannot be overlooked. Form W-2 of the Texas Railroad Commission s Oil and Gas Div. is an example of the information required (see Fig. 32.1). As a general rule in Texas and Louisiana, the daily allowable is determined by the producing horizon of the well, but there are exceptions to this rule. The reservoir characteristics, previous reservoir performances, and completion procedures are considered before a definite allowable is set for the well. The general case in Texas is covered by a depth allowable set up in 1947. These allowables are set up for wells completed in proven areas and known reservoirs.
32-2
PETROLEUM
ENGINEERING
HANDBOOK
TABLE
32.1-DETERMINATIONS
FROM
TEST
RESULTS
a. 9. 10. Il.
Optimal or maximum efficient productionrates. Correlation and identification of producing horizons. Results of recovery methods. Estimates of oil and gas reserves; for example, gauged rateproductiondecline vs. time. Decline trends and performance productionsin the ability of the reservoir to produce. Qualitative determinations of gas and/or liquid contacts. Determinationsrelative to artificially imposed harmful wellboreor reservoir conditions such as gas or water coning, sanding or bridgingaction, and paraffin depositions. Analyses and comparisons of well-completion practices and equipment. Performance of and comparison between subsurface well equipment and installation principles. Analyses and comparisons of artificial lifting practices and equipment. Determining the necessityof and evaluating the results of remedial measures.
The allowables set up by the state regulatory bodies are not necessarily the proper rate to produce the wells. The producer should work very closely with the reservoir engineers and geologists to see that the test data, along with all other information available, are used to determine the most efficient producing rate of the well or
reservoir.
The allowable given a well completed in a new field or new reservoir is called the Discovery Allowable: It is usually set up on a producing-depth basis. This allowable is for a certain period of time or until a certain number of wells are completed into the reservoir. For example, in Texas the onshore discovery allowable is for a 24-month period or until the 11th well is completed into the reservoir. The offshore discovery allowable is for an l&month period or until the sixth well is completed into the reservoir. The discovery allowable in Texas is set up as in Table 32.3.
When several companies produce from the same reservoir, it is common procedure to combine their knowledge and arrive at a maximum efficient producing rate (MER). It is permissible for a company to request a change of allowables to conform more closely to the MER as determined by the company s engineers. Before any change in the allowables is made by the regulatory bodies, a meeting or hearing of all the companies involved in producing from the reservoir is proposed. At this time all the information available is evaluated to determine if a new and different allowable is warranted. This new allowable (to conform more to the MER) usually involves a reduction in the normal allowable set up by the regulatory body.
Fig. 32.1-Sample
POTENTIALTESTS
OFOILWELLS
32-3
TABLE
32.2-ALLOWABLE
YARDSTICK 65 Yardstick'
SCHEDULE 66Offshore" 160 238200 238 238 249 265 288 310 331 353 380 402 435 471 515 562 621 679 735 789 843 905 1,000 40 200 200 245 245 275 305 340 380 420 420 465 465 515 515 565 565 620 620 675 675 735 735 .-80 160 590 590 590 640 705 785 865 950 1,050 1,150 1,150 1,260 1,260 1,380 1,380 1,500 1,500 1,625 1,625 1,750 1,750 1,880 1,880 330 330 330 360 400 445 490 545 605 665 665 730 730 800 800 a75 875 950 950 1,030 1,030 1,115 1,115
47 Yardstick Depth (ft) 0 to 1,000 1,000 to 1,500 1,500 to 2,000 2,000 to 3,000 3,000 to 4,000 4,000 to 5,000 5,OOOto 6,000 6,oooto 7,000 7,000 to 8,000 a,oooto 8,500 8,500to 9,000 9,000 to 9,500 9,500 to 10,000 10,000 to 10,500 10,500 to 11,000 11,000 to 11,500 11,500 to i2.000 12,000 to 12,500 12,500 to 13,000 13,000 to 13,500 13,500 to 14,000 14,000 to 14,500 14,500 to 15,000 10 18 27 36 45 54 63 72 81 91 103 112 127 152 190 20 28 37 46 55 64 73 82 91 101 113 122 137 162 210 225 225 290 330 375 425 480 540 40 57 66 75 84 93 102 111 121 133 142 157 182 230 245 275 310 350 395 445 500 560 10 -21 21 21 22 23 24 26 28 31 34 36 40 43 48 20 39 39 39 41 44 4% 52 57 62 68 74 81 88 96 106 119 131 144 156 169 181 200 -
40 74 74 74 78 84 93 102 111 121 133 142 157 172 192 212 237 262 287 312 337 362 400 -
80 129 129 129 135 144 158 171 184 198 215 229 250 272 300 329 365 401 436 471 506 543 600 -
1965 yardstick effective tt field d!scovered aiter Jan 1966 offshore yardstick efiective Jan 1. 1966
1, 1965
Wide-open flow may cause water or gas coning, influx of sand into the wellbore, collapse of tubing or casing, and/or many other harmful results. The ability of a well to produce usually is determined by use of the PI. The use of the PI was first mentioned by Moore in 1930. In a 1936 paper M.L. Harder states that the relative ability of a well to produce shows the PI to be superior. 2 API states in Recommended Practice for Determining Productivity Indices3 that the PI is calculated from the observed production rates and subsurface pressure measurements obtained. Special applications and modifications by the user to conform to individual requirements and conditions are normally used. The following discussion of PI is not meant to cover all applications but only to show how the PI may be used. By definition, the PI is equal to the barrels per day of stock-tank oil production per pound force of pressure differential between the wellbore opposite the producing horizon and the static reservoir pressure, which is the average pressure of the well drainage area. Therefore, the PI is, in barrels of oil produced per day per psi decrease in reservoir pressure, the difference between the average pressure in the drainage area of the well and the flowing bottomhole pressure (BHP) of the well. According to the accepted concepts of flow, the rate of flow in a system containing a single fluid under steady-state conditions should be directly proportional to the pressure drop. Using this concept, the PI would be the slope of the line resulting from plotting the rate of flow against pressure drop. On such a plot the wide-open flow quantity or well potential would be measured at the maximum pressure drop available. Such a case is referred to as the ideal PI. Observed values of production rates vs. pressure differentials do not give straight lines. PI data on nonflowing wells are usually more linear than the data
on flowing wells. Experience has shown that the line will be curved. This is because PI is defined to occur under steady or pseudosteady-state flow conditions. The curvature results because pB is not constant if the flow is single phase and/or gas evolution or water coning exists around the wellbore - i.e., relative permeability effects. PI is higher than theoretical when calculated erroneously before pseudosteady-state flow exists. PI is meaningless unless the radius of wellbote damage is fixed - i.e., pseudosteady-state flow is established. The flow of compressible fluids (oil and water) into a wellbore after the drainage area has been established is, strictly speaking, described by a pseudosteady-state flow equation.
TABLE
32.3-DISCOVERY
ALLOWABLES DailyWell Allowable WI) 20 40 60 80 100 120 140 160 180 200 210 225 255 290 330 375 425 480 540
Interval of
Depth (fu 0 to 1.000 1,000 to 21000 2,000 to 3,000 3,000 to 4,000 4,000 to 5,000 5,000 to 6,000 6,000 to 7,000 7,000 to 8,000 8,000 to 9,000 9,000 to 10,000 10,000 to 10,500 10,500 to 11,000 11.000to 11,500 11,500 to 12,000 12,OOOto 12,500 12.500to 13,000 13,000 to 13,500 13,500 to 14,000 14,000to 14,500
32-4
PETROLEUM
ENGINEERING
HANDBOOK
However, some people in the oil industry describe the flow by a Darcy-type equation, which is referred to as steady-state flow. The difference in the flow rates determined from the two equations is very small. We discuss PI by use of both the steady-state and pseudosteady-state flow equations. Steady-State Flow For a radial system under steady-state flow, the equation giving the flow rate is
where J is the PI, STBID-psi, and AJJ is the pressure difference between p e and p wf, psi. In analyzing the PI, it can be seen that it is a function of the formation characteristics, k and h, the fluid characteristics, pLo and B,, and the system characteristics, h, re, r,,,, and s.
Specific PI
The term specific PI is frequently used and usually means the PI per foot of pay. When the term specific is used, it is necessary to state why. The productivity could just as well have been made specific to any other variable in Eq. 2. Well systems do not operate at any time under steadystate conditions, but they do under pseudosteady-state conditions. Oil, water, and gas are compressible fluids; therefore, only pseudosteady-state occurs. Thus, we cannot expect Eq. 2 to yield exact correlation. The primary correlation sought has been with permeability so that the PI could be predicted from core analysis. An example of an early correlation is given by Fig. 32.2. 4 Oil flowing into a wellbore will practically always seem to enter the wellbore from a formation of lower permeability than the homogeneous fluid value determined in the laboratory. In many cases the relative permeability to oil at interstitial water saturation is one. What actually happens is that the permeability measured in the laboratory is too high because the confining pressure is lower than in the reservoir. Muskat states that the PI should not be used to predict production of high differential pressures by simply multiplying the PI by the pressure drop of interest. He states that it is doubtful that calculated potential tests would agree with actual tests. The relative comparison should reflect the comparative production capacity with fair approximation. The PI times reservoir pressure equals the open flow potential. Muskat also states that the productivity index is an excellent tool to determine well problems such as: 1. Comparison before and after well treatments IO evaluate these treatments. J should increase. 2. Stable GOR, R, with a declining J indicates plugging of wellbore. 3. R increasing markedly without decline in J indicates entry of extraneous gas. This would be the same if R changes with various production rates and J stays constant. 4. Rapid increase in water production should bring a decline in J if water is entering through typical strata within oil pay. If 1 is maintained, this should indicate the water is not coming through the oil-producing strata. 5, Decline of J should take place during normal reservoir depletion and parallel the normal growth of GOR or water/oil ratio (WOR). If not, plugging of the wellbore should be considered. These are guidelines for further investigation.
where 40 = oil production rate, STB/D, k= permeability of formation, md, h= thickness of formation, ft, Pe = pressure at the effective drainage radius re normally approximated by PR, average reservoir pressure in drainage area, psi, Pwf = flowing BHP, psi, oil viscosity, cp, ;I I oil formation volume factor, RB/STB, r, = effective drainage radius, ft, r, = wellbore radius, ft, and S= skin effect (zone of reduced or improved permeability), dimensionless The term equivalent to the PI is
JLL 7.08x10-3kh
Ap
/1,B,[ln(r,/r,)+s]
(2)
IO3
PERMEABILITY, md
Theoretical PI
Muskat and Evinger6 first showed that a theoretical PI could be worked out using the steady-state formula as developed for a radial system flowing oil and gas. By
POTENTIAL
TESTS
OF OIL WELLS
32-5
such a system it can be shown that the J for a given well system can be expressed in terms of three parameters: (1) the producing GOR, (2) the pressure gradient in the well system, and (3) the absolute reservoir pressure. It can be shown that for the steady-state flow of oil and gas in a radial system, the following equation expresses the rate of oil flow. 7.08~10-~
40= kh PR k,/k -dp, s In re/rw +s pi PoBo
.... ....
(3)
Pw
l.OC+JPe
Pe 3.m
py
PR k,/k -4
hobo
7.08~10-~
(p,
khA,
-p,fHW,~~,)+4
J=
10 -3 kh X-0.75+s)
fi,B,(ln
. . . . . . . . . . . . . . . . . . . . . . . . . (4) where A, is the area under the curve. By using Fig. 32.3 and the definition of the PI, it can be seen that I will not double if (p, -p,,,& is doubled because the area under the curve will not double. Also pe is determined by reservoir conditions and cannot be varied. Note that for a definite value of (p, -pwf) taken at a high absolute pressure, J will be greater than for the same (p, -pg) taken at a lower pressure because the area under the curve will be greater. It is not readily apparent, but can be shown, that J depends on the producing GOR, R. A simple explanation is that an increase in R means that the oil saturation is less, thus k, is smaller. In Fig. 32.3, the curves labeled R t , RI, and R3 are for different GOR s with R, >
R2 > R3.
Pseudosteady-State Flow
The steady-state equation is used frequently; however, this would only apply if the pressure at the outer radius stayed constant, which would only happen if a complete pressure maintenance program were maintained. If the well has a closed boundary or is operating with an established drainage radius, then pseudosteady-state flow occurs. The pseudosteady-state equation is 7.08~10-~
40=
where x is a factor for noncircular drainage area and well location. Most reservoir engineering flow equations assume radial geometry. This assumes the drainage area of the well is circular and the well is located in the center. Experience has shown that in many cases the drainage area of the well is rectangular, triangular, or other shapes. As stated previously, the PI, J, is a function of the system characteristics, which can be applied to the productivity functions can be determined by reservoir index. Sha limit tests. r Van Everdingen originally defined the skin effects as the additional resistance concentrated around the wellbore that result from the drilling and completion techniques employed. 9 This skin effect detracts considerably from a well s capacity to produce. More recently the skin effect is also used to indicate improved permeability around the wellbore that results from acidizing and/or fracturing. The skin effect can be defined as
......... ...
..
. (7)
where
r, = radius of area around the wellbore affected
kh (PR-p,,,,) . . .
.(5)
p$,[ln(r,/r,)-0.75+s]
affected by skin effect, md. for a circular drainage area and The skin effect, s, normally is determined from pressure transient analysis. The reader is referred to Chap. 35, Well Performance Equations, for a detailed treatment of skin effect.
. .@I
32-6
PETROLEUM
ENGINEERING
HANDBOOK
The aim of the production engineer is to make the PI, J, as high as possible; the equation for J indicates this may be done by several ways that include: 1, Remove the skin effect through acid treatment or the use of various completion or drilling fluids, depending on the formation. 2. Increase the effective permeability by fracturing or propping. 3. Reduce the viscosity by formation heating. 4. Reduce formation volume factor, B, , by production techniques and surface separation system. 5. Increase the well penetration, h, by completing across the entire producing formation. Care should be taken not to complete across a zone of excessive gas or water production. 6. Reduce the ratio re/rw. Since it appears as a logarithmic tey4 this has little influence. Underreaming is seldom considered as a means of well stimulation. The above equations should indicate that the most important step in determining and analyzing the performance of any well, especially flowing wells, is to determine the well production rate for any given flowing BHP. It is now readily apparent that to compare a PI of a well it is necessary to know what is being compared, which includes the permeability, sand thickness, well radius, drainage radius, fluid characteristics, and flow relations. A comparison should also be made on reservoir pressure and pressure drawdown for a similar GOR. The standard procedure for conducting PI tests mainly consists of following the directions that have been set forth in the procedures for conducting static and flowing reservoir pressures and gauged-rate production tests. The most popular wireline pressure gauge is the Amerada recording pressure gauge. In some cases, the use of artificial-lift equipment prevents the passage of subsurface gauges; therefore, other means must be found for determining these pressures. It is possible with the use of sounding devices-i.e., Echometer or Sonolog - to determine the liquid level. Knowing the liquid level, the static or flowing pressures can be approximated by gradient and depth calculations. Care should be exercised in determining the static and flowing pressures to be sure they are the equilibrium pressures. If there is any doubt regarding the equilibrium conditions, two or more pressure readings should be made several hours apart to be sure they are the same. Some formations stabilize in one hour but most take four to 24 hours. Tight formation could take several days. In determining the actual J, the flowing rates should have wide enough variation to compensate for any errors in measurement. When artificial-lift equipment is used, the gauged production tests for determining the J must be lower than the production limitations of the lift equipment. Methods of determining production rates are: (1) stock-tank measurement; (2) portable well testers, including batch-type meters, positive-displacement meters, turbine meters, and flow meters; and (3) stationaty test equipment.
ing the amount of liquid produced by an oil or gas well is the manual gauging of the production in stock tanks. A single well producing into a tank battery presents no testing problems, as it is a simple process to measure the liquid in the stock tanks at the start of the test and the liquid in the stock tanks at the end of the test. Most tank batteries are arranged so that one well may be tested at the battery while the other wells are being produced. This requires, of course, the addition of a test separator in addition to the production separator and also sufficient stock-tank volume so that no commingling of the production is required. A separate gas run and meter would also be required for measurement of the gas produced while the well is on test. If this meter run is not available, a portable orifice well tester could be used. If testing facilities are not available at the lease, it is necessary to shut in the entire lease while individual wells are being tested into the tank battery. In the latter case, it is usually better to use portable test equipment to determine the production rate. Before the production test is started, with stock tanks as a means of measurement, it is necessary to determine the amount of basic sediment and water (BSSrW) at the bottom of the tank, in addition to the liquid level at the start of the test. If at all possible, stock tanks should be clean, since errors may be introduced in determining BS&W. Dirty stock tanks can cause the tank tables to be in error. If possible, produced water should be produced and gauged in separate stock tanks. After gauging the test tank or gun barrel, the usual pmcedure would be to proceed to the separator, check and record the operating pressure, and on large-volume separators, if the well being tested has a low productivity, record the liquid level in the separator. The choke size at the wellhead should be carefully determined and if there is a question, the choke should be checked and calibrated. If a treater or heater is used, its operating characteristics should be noted so that any conditions not uniform may be considered on the gauged production test. All tests should be made after the production has stabilized and under conditions that are as uniform as possible. No change should be made at the wellhead or tank battery during the duration of the test. Standard tests should range from 24 to 168 hours in length, depending on the well and reservoir characteristics. All data should be observed and recorded at less frequent intervals. The time between intervals will vary depending on the length of test. In cases where short tests (6 to 8 hours duration) are necessary, consecutive data recordings should be made hourly. Tests should be for 24 to 168 hours duration so that fluctuations in the GOR, as a result of heading tendencies and temperature variations, may be considered and the test results averaged. It is not uncommon to observe a 40 to 50F temperature variation between night and day atmosphere and to observe a 10 to 20% variation in the gas/liquid ratio. Recognizing these facts, it becomes necessary to measure the temperature of either the liquid or vapor section of the separator. The preferred location to obtain the temperature would be in the flow line immediately before the separator. This latter temperature would more nearly reflect the ac-
Stock-Tank Measurement
The oldest and most widely accepted means of determin-
POTENTIAL
TESTS
OF OIL WELLS
32-7
(II Degassing element FLOATE 1 OIL LLC WAT 77+&i/ r 20 r WATER OUTLET ET (2) Mist extractor (vane-type) (3) Relief valve (4) Safely head (optional) (5) Elkpt~cal manway (6) 011.dump valve (7) Water-dump (6) 011 meter (9) Water meter (IO) Check valves (11) Strainers (optional) (12) Plug Yalws Z&WELL FLUID (13) Pressure gauge (14) 011 outlet-wew nipple (removable) (15) Interface LLC pIlot (16) 011 LLC pIlot WATER &$#$~OIL;UTLET (18) Gas outlet (from separator) (19) Orifice fittmg (optmnal) (20) Gas back-pressure FRONT ELEVATION SIDE ELEVATION (21) Liquid-level sight glasses (22) Trailer assembly valve OUTLET (17) Water outlet (from separator) valve
tual liquid-phase temperature prior to the flash liberation process. The production should be sampled each time the data are recorded during the production test. This interval sampling is very necessary if storage facilities are such that the water produced must be disposed of as it is produced. Refer to Chap. 17 for sampling information. Wells that produce with a high WOR should always be tested at length because of the usual uneven oilproducing rate. Depending on the separation pressure and the characteristics of the crude for shrinkage, some consideration should be given to determining the amount of gas produced that is not measured by the gas meter but is vented from the stock tanks. Points to check to ensure the accuracy of stock-tank gauges include: 1. Correct and accurate strapping table is used with the test tank. 2. No dents or damage has occurred to the stock tank since strapping table was made. 3. Tank is clean with no encrustations or deposits on the walls. 4. If foamy crude is being produced the liquid level will be almost impossible to determine correctly with a gauge tape. Chemical addition or settling time may be required to minimize the foam. 5. BS&W at bottom of tank is determined as accurately as possible before and after the test. 6. Gauge tape must not be kinked and plumb bob must be carefully touched at the bottom of tank. 7. Temperature of oil in the stock tank must be considered. 8. The oil level must be steady and undisturbed while a gauge reading is taken.
32-8
PETROLEUM
ENGINEERING
HANDBOOK
(1) B?gasswlg element (2) Mist extractor (vane-type) (3) Relief valve (4) Safety head (optional) (5) Elliptical manway (6) OIldump (7) Oil meter (6) Check valves (9) Strainer (optional) valve
-8
FLUID
\T
(10) Plug valves (11) Pressure gauge (12) Oil outlet (from separator) (13) Gas outlet (from separator) (14) Orifice fittmg (optional) (15) LLC pilot (16) Trailer assembly
U
FRONT ELEVATION
,
SIDE
ELEVATION
(17) Liquid-level
(I) Degassmg element (2) Mist extractor (vane-type) (3) Relief valve (4) Safety head (optIonal) (5) EllIptical manway (6) Water-dump (7) Water meter NOTE. INTERFACE MUST BE WEIG SINK IN OIL ANU FLOAT tMU -PLAN (8) 0114nlet motor (9) 0~1.outlet motor valve (lo) 011 meter (11) Pressure gauge (12) Oil-outlet wer nipple (removable) (t 3) Water outlet (from separator) (14) Gas outlet (horn separator) WELL FLUID INLET -5 (I 5) Orlflce iittmg (optional) (1.5) Interface LLC Plkx (17) Trailer assembly (18) Liqwd-level slght glasses llne valve
(2,) Gas back-presSure FRONT ELEVATION SIDE ELEVATION (22) Check valve (23) Check valves (24) Siram?r (+WXMt)
YalYe
POTENTIAL
TESTS
OF OIL WELLS
32-9
(1) oegassmg
element
(2) Misl extractor (vane-type) (3) Relief valve (4) Safety head (optmal) (5) Elliptical manway valve molar valve motor valve
< 3 ,Ll!?l
SECTION %B
(10) Oil-Inlet motor valve PLAN I ---- ] (11) Oil-outlet (12) Oil meter (13) Pressure gauge (14) Oil-outlet war mpple (removable) (15) Water outlet (irom separator) (16) Gas outlet (horn separator) (17) Orifice fitting (optlonal) (18) Interface LLC pilot DRAINSM (19) Trailer assembly (20) Llquld-level slght glasses llne motor valve
FRONT ELEVATION
SIDE
ELEVATION
valve
-ET
(1) DegassIng element (2) Mist extractor (vane-type) (3) Gas outlet [from separator) (4) Orifice fItring (optional) (5) 011 outlet (from separator)
.ET
Ii----l/L17
(6) 011 meter (7) Oil-Inlet mOtOr Valve (6) Oil-outlet motor valve (9) Relief valve (lo] Safety head (optional) (11) Pressure gauge (12) Gas-equalizing line valve
PLAN -
sight glass
manway
(161 Plug valve (17) Check valve (apt~onal) (16) Trailer assembly
FRONT ELEVATION
SIDE
ELEVPTION
32-10
PETROLEUM
ENGINEERING
HANDBOOK
TABLE
32.4-WELL-TESTER
SPECIFICATIONS
Shell OD (in.) 16 24 30 36 48 16 24 30 36 16 20 24 30 14 16 20 24 12 14 16 20 12 14 16 20
Oil Plus Water (B/D) 500 1,200 1,800 2,800 4,800 500 1,200 1,800 2,800 500 850 1,200 1,800 400 500 850 1,200 300 400 500 850 300 400 500 850
Gas
(MMscf/D) 1.6
3.6 5.5 8.0 14.0 3.5 7.7 12.5 17.0 5.5 8.3 12.5 19.0 6.1 7.7 11.4 16.5 5.5 6.8 8.0 12.2 5.3 6.5 8.0 12.4
Oil (B/D) 5001,200 1,800 2,800 4,800 500 1,200 1,800 2,800 500 850 1,200 1,800 400 500 850 1,200 300 400 500 850 300 400 500 850
Water (B/D) 250 600 900 1,400 2,400 250 600 900 1,400 250 425 600 900 200 250 425 600 150 200 250 425 150 200 250 425
Gas
(MMscflD) 1 .o 2.2 3.5 5.1 8.7 2.1 4.9 7.5 11 .o 3.3 5.0 7.5 12.0 3.8 5.0 6.5 9.1 3.5 4.3 5.5 7.0 3.4 4.2 5.3 7.0
W)
500 1,200 1,800 2,800 4,800 500 1,200 1,800 2,800 500 850 1,200 1,800 400 500 850 1,200 300 400 500 850 300 400 500 850
Approximate Weight (Ibr?$ 1,200 1,500 1,700 2,300 41600 1,900 2,200 2,400 2,800 2,600 2,800 3,000 3,200 2,900 3.100 31400 3,600 3,000 3,400 3,800 4,100 3,500 4,000 4,500 4,800
The type of fluid to be tested must be considered in determining the retention time. If the crude foams, the necessary retention time for the gas to break out of solution may be 5 minutes or longer. For three-phase separation, additional retention time may be required for the oil and water separation depending on the type of emulsion produced. Many times the addition of heat and/or chemicals is required to produce proper separation of the oil and water or gas and oil. Well testers are available in low working pressures that utilize either electric or gasfired heaters to heat the well fluid and improve the separation processes. The meter used must be of sufficient metering capacity not to limit the capacity of the well tester. Working pressures are available up to 4,000 psi. The listing in Table 32.4 of sizes and capacities of standard well testers is not complete but may be used as a guide to determine the approximate size and capacity of the unit required for your specific testing purpose. The types of meter available for use on well testers are (1) batch-type meters, (2) positive-displacement meters, and (3) flow meters, including standard and mass flow meters. Batch-Type Meters The batch-type meter works by means of cyclic accumulation, isolation, and discharge of predetermined volumes. Each dump volume is registered on a counter. The counter reading is then multiplied by the dump volume to determine the total measured volume. When metering vessels such as batch meters are used to measure liquid hydrocarbons, four factors must be ob-
tained and maintained: 1. An unchanging volume in the metering vessel must be maintained consistently. This means that there can be no foreign material deposited in the vessel and that the vessel itself must not change shape or size. 2. Exact upper and lower dumping levels must be obtained and maintained in the metering vessel. These dumping levels must be the same for each cycle. 3. Proper valve arrangement must be maintained so that no liquid may slip through the vessel without being metered. The valve or valves should be arranged so that there is a period of time at the beginning and end of each cycle during which both the inlet and outlet to the metering vessel are closed at the same time. This assures that no unmetered fluid will slip through the metering vessel. 4. An appropriate and accurate meter factor must be used to compensate for temperature change of the liquid, shrinkage (volume reduction) of the liquid resulting from pressure reduction, mechanical metering error of the metering vessel, and BS&W content of the liquid. These various factors are usually combined into one factor known as the meter factor or meter multiplier. These meter factors are usually less than 1.0. In other words, the meter normally reads higher than the net stock-tank volume. Advantages and Disadvantages of Batch-Type Meters 1. A metering vessel can be used as its own meter proving tank if the vessel is inspected for encrustation, the dumping levels are observed during operations, and valves are checked for leakage.
32-11
TABLE 32.5-NOMINAL
Barrelsper discharge Metering capacity, bbl/24-hr day
0.25 300
2.0 1,440
5.0 2,000
10.0 4,000
20.0 8,000
30.0 15,000
720
2. A metering vessel will handle more sand and other foreign material without causing trouble than the positive-displacement (PD) meter. 3. A metering vessel will meter from zero flow to maximum rated rate of flow with the same degree of
accuracy.
4. Weight-type (hydrostatic-head) metering controls may be used to meter foaming oil. 5. The unit may be adjusted while in operation. 6. Free gas will not register as liquid if the controls should fail to function. 7. Initial and installation cost is slightly higher than with the PD meter. 8. The meter delivers an intermittent discharge of liquid. 9. Gas is requited to displace the liquid from the vessel. 10. Paraffin buildup on the vessel wall will cause inaccuracies. 11. The meter requires more space and is heavier than a PD meter. 12. With heavy viscous crudes a larger inlet or differential pressure arrangement between the separator and metering vessel may be required to maintain the rated capacity. Nominal rated metering capacity of various sizes of volume-type dump meters under normal field conditions is given in Table 32.5. Pressure ratings are up to 3,ooO psi. Volume meters are not used for gas measurement. PD Meters PD meters are quantitative instruments. They are termed positive-displacement because some sensing element is forcibly or positively displaced through a measuring cycle by the hydraulic action of the fluid on the element. l1 For a completed measuring cycle a known quantity is displaced by the sensing element. It is necessary to count the number of cycles and multiply them by the displacement volume to get the total liquid quantity that has passed through the meter. This latter function is carried out by the meter s gear train and register. Fig. 32.9 shows the basic types of PD meters: nutating disk, oscillating piston, oval gear, rotary vane, reciprocating piston. and bi-rotor. Probably 80 to 90% of all PD meters in service today are of the nutating-disk type. They are most popular because of the relative simplicity of the construction, ruggedness, accuracy over a wide range, and low cost. The accuracy of the nutating-disk type meter is not as high as that of the PD meters of the other types. Advantages and Disadvantages of PD Meters 1. Discharge of the metered liquid is continuous. 2. PD meters can be used to meter exceptionally viscous liquids.
3. PD meters do not require gas to displace the fluid through the meters. 4. Initial and installation cost is lower. 5. Temperature compensation can be applied with certain types of PD meters less expensively than with batchtype meters. 6. Paraffin may not reduce metering accuracy. 7. Liquids metered must be free of gas, since slugs of gas may damage or wreck the meter. Gas will register as fluid when passing through a PD meter. 8. Sand, mud, salt, or other foreign particles will cause wear on the PD meter and cause inaccurate meter readings. 9. Some type of meter-proving process is required to prove PD meters periodically. Actual stock-tank gauges are used in many cases to determine the accuracy of the meter. 10. Meters must be operated between a minimum and maximum specified rate of flow. High or low rates of flow may affect the accuracy of the PD meter. PD meters are available in pressure ratings to 5,000 psi. Capacities of PD meters depend on the size and type of meter. If possible the manufacturer of the PD meter should be requested to furnish information regarding the
NUTATING
DISK
OSCILLATING
EL
PISTON
OVAL
GEAR
ROTARY
VANE
RECIPROCATING
PISTON
Bl-ROTOR
Fig. 32.9-Basic
types of PD meters.
32-12
PETROLEUM
ENGINEERING
HANDBOOK
Turbine Meter Today most liquid measurement is done by the use of turbine meters. A turbine meter is a flow rate measuring device which has a rotating element that senses the velocity of the flowing liquid. t2 This liquid causes the rotating device to rotate at a velocity proportional to the volumetric flow. The movement of the rotating device is sensed either mechanically or electrically and is registered. The actual volume is then compared to the registered readout to arrive at a meter or register factor (see Figs. 32.10 and 32.11). The turbine meter is used because of its simplicity and costs. Each meter application will require different meter or register factors. Considerations in the selection of turbine meters include (1) properties of the liquid to be metered-viscosivapor pressure, corrosiveness, and ty, density, lubricating ability; (2) operating conditions, including pressure, flow rates and whether continuous or intermittent, temperatures (some meters have temperature compensators), and quantity and size of abrasive particles in the fluid; and (3) space availability (see Fig. 32.12). Items or conditions that normally affect the meter factor are shown in Table 32.7. Both turbine and PD meters should be connected so that meter factors may be periodically determined.
Size
(in.) 5/B to J/i Yi 1 1% 2
Average Capacity for Nutating-Disk Type PD Meters. The average rates shown in Table 32.6 are based on oil as the fluid being metered. The manufacturer should always furnish information about recommended capacity. When PD meters are used downstream of separators or vessels that arc not continually dumping, it is necessary to size the meter on the maximum rate of discharge while the vessel is dumping. PD meters are used in some cases for gas measurement The types used are the bellows and birotor. High cost and size have held the use of PD meters for gas measurement to a minimum.
(I) Upsiream (2) Upstream (3) Bearings (4) Shaft (5) Rotor hub
(6) Oownstrearn *tator supports (9) Meter housing (10) Pickup (11) End connectlorw
Fig. 32.10-Names
(2) Prsssure gauge (optional) (3) Filter, au eliminator 9 *.1ew , j . 7i Ii I +q ( I? I= ,h sl 6 - < > + $ hgf , :. 4 1 + e+ ;<I t Mfp&.Tl; y (4) Straight pipe (5) Straightening (6) Turbtne meter (7) Straight pipe (with straghtening (8) Pressure gauge (9) Thermometer (IOJ Proving connections (should be downstream of meter run) vane as required) vane (as required) and/or strainer as required
(12) Control valve (as required) (131 Check valve (as required)
Fig. 32.11-Turbine
POTENTIAL
TESTS
OF OIL WELLS
32-13
fxF= Fig. 32.12--Schematic operationdiagram of oilwell production meter installation with stock tank or open prover.
Flowmeters The standard-type flowmeters are: orifice plate, Venturi tubes, flow nozzle, Pitot tube, drag body, and lift surface. These meters or devices are used to create a flowing differential pressure. This flowing differential pressure is used to solve the flow equation for the rate of flow. Orifice plate, Venturi tube, flow nozzle, and Pitot tube are commonly used. Drag body and lift surface arc not as familiar. The net force resulting from a pressure difference is measured. This pressure difference is used to solve the flow equation. If the force is parallel to the flow direction, the force is called drag body. If the force is perpendicular to the flow direction, the force is called a lifting surface. In addition to differential pressure drop, six other factors must be considered and included in the integration to determine a basic flow rate or quantity. They are (1) static pressure, (2) flowing temperature, (3) specific gravity of the flowing quantity, (4) size of orifice run, (5) size of the orifice plate if orifice plate is used to create the differential pressure, and (6) supercompressibility, if applicable. These factors may be considered by the solving of the flow equation or they may be applied as a multiplying factor applied to the meter reading. Mass Flowmeter. In about 1942, W.J.D. VanDijik of The Netherlands constructed and evaluated the first mass flowmeter, as such. The mass flowmeter measures the quantity of matter passed through the meter. This mass is independent of all ambient conditions, which is not true of volumetric meters discussed previously. If any type of flowmeter mentioned above is compensated in any way (electrically, mechanically, or a combination of both) for fluid density, it is a mass flowmeter. As early as 1930, pressure and temperature compensators could be attached to a standard orifice meter. They were mass flowmeters, in a true definition of mass flowmetering, although they were not called by that name. On well testers, and for measurement of produced
water, oil, and gas, it is most common to use turbine, PD, and batch meters for liquid measurement and the standard orifice meter for gas measurement. Mass flowmeters may be used for both gas and liquid measurement. Because of the cost and special requirements for technicians to operate and maintain these meters, they are seldom used in field operations. Automation and remote readout requirements can be accomplished with the use of well testers in much the same way as with automatic custody units. The liquid measurement may be relayed by pneumatic or electrical impulses to a transmitter. These impulses in turn may actuate some type of recording unit at a central location. The gas measurement would require some type of pressure transducer to convert the differential pressure to an electrical signal. More often, when remote recording of gas flow is required, an integrating flowmeter is used and the transmitted signal will read in volumetric units. Fig. 32.13 shows a well tester installed at a tank battery for permanent test or lease automation. Stationary Metering Installation Stationary metering installations are those that include metering separators, metering treaters, and any other
TABLE
THE
METER
1. Mechanics of meter as to tolerances. 2. Change in clearance due to wear or damage. 3. Flow rates and variations in flow. 4. Temperatures of liquids. 5. Viscosity of liquids. 6. Pressure of liquids. 7. Pressure drop across the meter as a resistance to flow. 8. Foreign materiallodged or deposited in the meter or connecting piping. 9. Inlet condition changes, such as changes in the entrance to the meter, which change the flowingfluid profile. 10. Lubricating properties of the liquid. 11. Accuracy and conditions of meter-provingsystem and meter-factor test.
32-14
PETROLEUM
ENGINEERING
HANDBOOK
oil and water. PD meters may be used for metering all produced fluids. Pressure ratings range to 3,000 psi. Capacities will be the same as shown for standard oil and gas separators in Chapter 12. The metering capacity will be as shown for the type of meter used.
GOR
The GOR may be defined as the rate of gas production divided by the rate of oil production. It is usually expressed as standard cubic feet of gas per barrel of stocktank oil produced under stabilized flowing conditions for a 24-hour period. The term cubic feet of gas or standard cubic feet of gas means the volume of gas contained in one cubic foot of space at a standard pressure base and standard temperature base. This standard base is normally 14.65 psia and 60F. Most tables published for orifice well testers, Pitot tubes, flow provers, and other gas measurement means are referred to a base pressure of 14.65 psia and 60F temperature. Whenever the conditions of pressure and temperature vary they may be converted to a standard or base condition by the use of the real gas law. The volume of gas used should be the total gas produced from the reservoir through either the casing or tubing. Any gas that is injected back into the reservoir for artificial lifting purposes such as gas lift should be subtracted from this total gas produced.
type of meter used in conjunction with test separators or emulsion treaters. These units are installed as an integral part of the tank battery. The metering separator combines two functions of separating and metering the produced fluid (see Figs. 32.14 and 32.15). The metering separation is divided into one compartment for separating the liquid and the gas and into one or two other compartments for metering the
RELIEF VALVE
MIST EXTRACTOR
OLJTLET
RELIEF VALVE
SEPARATING
CHAMBER
GAS EQUALIZING
GAS BOIL
INLET -
PRESSUREGAUGE BAFFLE PLATEINTERFACE LLC WITH WEIGHTED OIL-WATER INTERFACE BAFFLE PLATE COMPARTMENTS PIPING COPPER TUBING -OPTIONAL ITEMS l ERT lNTERFC = (r PRESSUREGAUGE FOR -s
I,., II
COUNTER,
Flg. 32.15-Three-phase metering separator with integral metering compartments. Gauge glasses and valves are furnished for separating chamber and both metering chambers. Automatic BS&W samples can be furnishedas an option.
32-15
The volume of oil produced should be determined by any of the means discussed in the first part of this chapter. Procedures for Well Testing Flowing Wells. The oil flow should be stabilized during the 24-hour period immediately preceding the test. This stabilized flow should be very close to the assigned allowable or the daily producing rate. If the well being tested is a discovery well, the producing rate should be as close to the assigned discovery allowable as possible. Any adjustments should be made during the first 12 hours of the stabilization period and no adjustment made during the last 12 hours or during the time the well is on test. All gas withdrawn from the reservoir must be included as produced gas. If the oil has a great deal of shrinkage after it is placed in the stock tanks, some means should be considered for measurement of the gas that breaks out of solution. Any gas used for operation of machinery or for any other purpose must be considered as produced gas. Tests should range from 24 to 168 hours duration to consider any uneven flow. Intermittent Flowing Wells (Stopcocked). The procedure for testing should be as outlined for flowing wells, except the shut-in casing and tubing pressures should be approximately equal to the pressures recorded at the beginning of the test. The Texas Railroad Commission states, The closed-in casing pressure at the end of the 24-hour test period shall not exceed the closed-in casing pressure at the beginning of the test by more than six-tenths (0.6) pounds per square inch per barrel of oil produced during the test. This rule also applies to flowing wells. This is true because of the loading characteristic of some wells while shut-in. Before an accurate test can be made the flow must be stabilized and stabilization cannot occur while producing from a loaded annulus or tubing. Gas-Lift or Jetting Wells. The volume of gas used should be the net produced gas or formation gas. Formation gas will be equal to the total gas produced minus the injection gas. Pumping Wells. In computing the GOR for pumping wells, the total volume of gas produced during the 24-hour period, ending with the closing in of the well at the conclusion of the tests, and the total barrels of oil that are produced in order to obtain the daily allowable must be used regardless of the actual pumping time in the 24-hour period. If the gas produced is not enough to measure accurately, this should be indicated on the test report as gas too low to measure. In some states the regulatory agencies will lower the assigned allowable of the well if the daily or producing allowable is not produced while making GOR tests. Average GOR To obtain an average GOR for several wells or for all the wells in a field, one cannot take the arithmetic average value of the ratios. For example, two wells with GOR s of 2,000 and 4,000 would not necessarily have an
average ratio of 3,000. For the average ratio to be 3,000 the wells would have to produce the same amounts of oil. The average GOR of several wells must be obtained by dividing the total gas production of all wells involved. For example, if the 2,000-ratio well produced 50 B/D oil and the 4,000-ratio well produced 200 B/D oil, the average GOR for the two wells would be
R=(2,000X50)+(4,0tXlx200) 50+200
=3,600
cu ftibbl.
(8)
where
R 1w = individual
i? = average GOR, cu ftlbbl, well GOR, cu ft/bbl, and 4iw = individual well daily production, STBID. GOR
Cumulative
The cumulative GOR is defined as the total amount of gas produced and kept from the reservoir up to a ceflain time divided by the cumulative oil produced up to the same time. Therefore, G,-G
NP
Rp= -,
. ..I....I....................
'(9)
where
R, = cumulative GOR, cu ft/bbl, G, = total gas produced, cu ft, G = gas reinjected, cu ft, and NP = total oil produced, bbl.
32-16
PETROLEUM
ENGINEERING
HANDBOOK
GOR.This shows that maintaining the GOR as low as possible will increase the cumulative production for the same amount of produced gas. Consider the internal-gas-drive reservoir. As oil is produced from the reservoir the space is taken over by gas volume. The presence of gas within the reservoir decreases the ability of oil to flow and increases the ability of gas to flow. After a certain minimum gas saturation (about 5 to 10%) is exceeded, the ease with which gas flows increases to such an extent that it flows concurrently with the oil. This process continues until finally the only flow is almost all gas. This allows the reservoir energy to escape and causes the reservoir to cease production by natural means. Fig. 32.16 shows how the stock-tank cumulative production almost ceases as the reservoir/GOR increases.
B, = 011 formation volume factor, res mJiSTm+ t, = effective drainage radius, m r, = wellbore radius, m s = skin effect (zone of reduced or improved pwmeability), dimensionless J = productivity index (PI), m3 I(d.kPa)
References
I. Moore, T.V.: Definitions of Potential Productions of Wells Without Open Flow Tests, Bull.. API, Dallas (1930) 20.5. 2. Harder, M:L.: Productivity Index, API, Dallas (May 1936). 3. API Recommended Practice fir Determining Productivity Indices, API RP 36, first edition, API, Dallas (June 1958). 4. Calhoun. J.C. Jr.: Fundamentals of Reservoir Engineering, Evised edition, U. of Oklahoma Press, Norman (19.53). 5. Muskat, M.: Physical Principles of Oil Production, Intl. Human Resources Development Corp., Boston (1981). 6. Muskat. M. and Evinger, H.H.: Calculation of Theoretical Productivity Factor, Trans., AIME (1942) 146, 126-39. 7. Odeh, A.%: Pseudosteady-State Flow Equation and Productivity Index for a Well With Noncircular Drainage Area, .I. Pet. Tech. (Nov. 1978) 1630-32. 8. Earlougher, R.C. Jr.: Estimating Drainage Shapes From Reservoir Limit Tests, J. Pet. Tech. (Oct. 1971) 1266-75; Trans.. AIME, 251. 9. van Everdingen, A.F.: The Skin Effect and Its Influence on the Productive Capacity of a Well, J. Pet. Tech. (June 1953) 171-76; Trans., AIME, 198. IO. Dake, L.P.: Fundamentals of Reservoir fkgineering, Elsevier Scientific Publishing- Co.. New York Cite (1978). I I. Measurement of Petroleum Liquid Hy~rocarbotu by Positive Displacement Meter, API Standard 1101, first edition, API, Dallas (Aug. 1960). 12. Manual of Petroleum Measurement Standards, API, Dallas (1961) Chap. 5.
@,[ln(r,lr,,,)+s]
5.427x 10-4kh
. .
J2!?=
Ap where
~oB,[ln(r,/r,,)+.~]
..t. .
(2)
m2 h = thickness of formation, m pe = pressure at the effective drainage radius I,, normally approximated by PR , kPa p,+,f = flowing bottomhole pressure, kPa PO = oil viscosity, Pa 1s
Chapter 33
Introduction
The gauging or testing of gas wells arose from the need to measure the productive capacity of a well. The earliest response to this need was to open the well to flow to the atmosphere and then to measure the flow rate. However, it soon became apparent that such practices were wasteful of gas, dangerous for personnel and well equipment, and frequently damaging to the reservoir. In addition, such tests provided very little information for estimating production rates into a pipeline. As a result, the practice of gauging gas wells by opening the well to flow to the atmosphere decreased and now is almost completely confined to stripper gas areas where pressures are very low and the rates of flow are small. every bailer run or at the end of each 5 ft of formation drilled. Upon completion, data were available to consttuct a chart showing a relationship between the rate of flow and depth. The chart is useful in determining the depth of the major gas-producing zones. Such data were valuable in planning remedial work that may be necessary during the life of the well. Pitot-tube gauges were useful in determining rate-of-flow increases resulting from each stage of acid treatment. In many cases the pitot-tube gauge after acid treatment provided data from which the desired flow rates for a backpressure test could be selected. Fig. 33.1 shows a pitot-tube and flow nipple arrangement that is suitable for gas measurement. The pitot tube should be made of %-in-ID pipe shaped to measure impact pressure at the center and in the plane of the opening of the flow nipple. The flow nipple should be at least eight pipe diameters long, free from burrs or other obstructions, and must be round. The impact pressures are measured with water or mercury manometers or a pressure gauge, depending on the pressure to be measured. The impact pressure is converted to rate of flow by suitable equations or tables such as those published by Reid. Subsequent experimental work by the USBM* is in reasonable agreement with the Reid data. The equations published by Reid were investigated by Binckley, 3 who concluded that they were based on sound theoretical principles. Reid s equations and tables have been adjusted to a pressure base of 14.65 psia for the purposes of this handbook. The adjusted equations for impact pressures less than 15 psig are ~/,~=34.81~&* K, qa =1284d; and
The author also wrote the original chapter on this topc fn the 1962 edItIon
.. , . .
.. . .
. (1) . (2)
Jh,
y,C=183.2dt
&,
33-2
PETROLEUM
ENGINEERING
HANDBOOK
60F, and for discharge into an atmospheric pressure of 14.65 psia. Corrections can be made when desirable by multiplying values from the equations or tables by the following factors.
F,=
J J
0.600 YR 520
and
FT=
(460+ Tf)
where = specific gravity correction factor, YY = specific gravity of gas being measured, air= 1.000, FT = flowing-temperature correction factor, and Tf = temperature of flowing gas, F. TO MANOMETER ORPRESSURE GAUGE CONNECTION The atmospheric-pressure correction factor for values from Table 33.1 and Eqs. 1, 2, and 3 is
Fbar = F,
J
-
Pa
14.65
L=8di
where Fbar is barometric correction factor and pa is atmospheric pressure, psia. The value of pressure used for p, in Eq. 4 is the absolute pressure and is computed by adding the barometric pressure to the gauge pressure. The correction factor for barometric pressure for Table 33.2 is
and pitot tube for gas
nipple
Fbar =
Pi +Pa
pi + 14.65
In ordinary usage, rates of flow are taken from pitot tables or formulas without correction. where qx dj h,. h,, pi
=
= = = =
of gas flow, Mcf/D (14.65 psia and fjoF), ID of flow nipple, in., height (manometer reading), in. water, height (manometer reading), in. mercury, and impact pressure, psig.
rate
Example Problem 1. Given an impact pressure of 27.2 in. of water on a flow nipple with ID~2.441 in., determine the rate of flow. Rate of flow from Table 33.1 for ID = 1.OOO= 182 Mcf/D. Rate of flow for ID = 2.441 in. is q,=182(2.441)2=182x5.958=1,080Mcf/D. Or, by Eq. 1, the rate of flow is
For impact pressures more than 15 psig, the adjusted Reid equation is q&,=23.89d?p, , . . . . . . . . . . . . . . . . . . . . . . . . . . (4)
where p, is impact pressure, psia. Values of rates of flow for various impact pressures are given in Table 33.1 for a flow nipple with an ID of 1.000 in. Rates of flow in Table 33.1 were computed by Eqs. 1 through 3. The range of impact pressures is from 0.1 in. of water to 15 psig. Rates of flow for impact pressures from 15 to 200 psig were computed by Eq. 4 and are given in Table 33.2 for a flow nipple with an ID of 1.000 in. Impact pressures measured on larger flow nipples can be converted to rates of flow by multiplying the rate of flow from the table corresponding to the impact pressure by the square of the ID (in.) of the larger nipple. Rates of flow taken from Tables 33.1 and 33.2 or computed by Eqs. 1, 2, 3, and 4 are for gases with a specific gravity of 0.600 (air = 1.OOO),flowing temperatures of
=(34.81)(5.958)(5.215)
= 1,080 McfiD.
Example Problem 2. Given an impact pressure of 65 psig on a flow nipple with ID=4.082 in. with discharge into a barometric pressure of 13.2 psia, determine the rate of flow. Rate of flow from Table 33.2 for ID of 1 .OOOand atmospheric pressure of 14.65 psia is 1,904 Mcf/D. For ID of 4.082 in. and barometric pressure of 13.2 psia, the rate of flow is qg = 1,904(4.082)*(65+13.2)/(65+14.7) =(1,904)(16.663)(0.9812)=31,100 Mcf/D.
OPEN
FLOWOFGASWELLS
33-3
Or
by
Eq. 4. 13.2)
Backpressure Testing
Before the development of the backpressure method for testing gas wells, the open-flow capacities of gas wells were determined by actual open-flow tests. The flowing of wells at their wide-open rate results in waste and possible damage to the well. In addition, the open-flow
test yields very little information regarding the capacity of a well to deliver gas into a pipeline system. The backpressure method of testin gas wells was developed by Rawlins and Schellhardt. 9 Results of tests on 582 wells as reported in their study and other work on many wells reported elsewhere show that when the rates of flow are plotted on logarithmic coordinates against corresponding values of (pi -p,,,,)-the difference of squares of the shut-in pressure FR and the flowing sandface (bottomhole) pressure (BHP) p,f-the relationship may be represented empirically by a straight line.
TABLE 33.1-RATES OF FLOW FOR IMPACT PRESSURES LESS THAN 15 PSIG MEASURED WITH A PITOT TUBE FOR FLOW NIPPLE WITH ID = 1 .OOO in. *
Impact Pressure
Water (in.) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.6 0.9 1.0 1.25 1.36 1.6 1.6 2.0 2.2 2.4 2.7 3.0 3.5 4.1 4.5 5.0 5.4 6.0 6.6 6.2 9.0 9.5 10.0 Mercury (in.) psig
qg, IO3 cu ft/D (14.65 psia and 60F) 11 .o 15.6 19.1 22.0 24.6 27.0 29.1 31.1 33.0 34.6 38.9 40.6 44.0 46.7 49.2 51.6 53.9 57.2 60.3 65.1 70.5 73.6 77.6 60.9 65.2 90.6 99.7 104.4 107.3 110.1 309 314 327 340 352 363 374 365 396 406 410 430 446 466 466
Impact Water (in.) -~10.9 12.0 12.2 13.9 15.0 16.3 17.7 19.0 20.4 21.6 24.5 27.2 29.9 32.6 -
Pressure
Mercury On.) 0.60 0.66 0.90 1.02 1.1 1.2 1.3 1.4 1.5 1.6 1.8 2.0 2.2 2.4 2.6 2.6 3.0 3.2 3.4 3.6 3.6 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 15.3 16.3 17.3 16.3 19.3 20.4 22.4 24.4 26.5 26.5 (Psig)
q,,103 cu ft/D (14.65 psia and60F) 115 121 122 130 135 140 146 152 157 162 172 182 190 199 207 215 222 230 237 244 250 257 263 269 275 261 267 293 296 304 502 516 522 549 564 560 606 634 661 677 710
0.10 0.12 0.13 0.15 0.16 0.18 0.20 0.22 0.26 0.30 0.33 0.37 0.40 0.44 0.50 0.60 0.66 0.70 0.74 5.6 6.0 6.5 7.0 7.5 6.0 6.5 9.0 9.5 10.0 10.2 11.2 12.2 13.2 14.3
2.0 -
of the diameter
33-4
PETROLEUM
ENGINEERING
HANDBOOK
TABLE 33.2-RATES OF FLOW FOR IMPACT PRESSURES. 15 TO 200 PSIG, MEASURED WITH A PITOT TUBE FOR FLOW NIPPLE WITH ID = 1 .OOO in. Impact Pressure Wg) 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 32 34 36 38 qg, lo3 cu ft/D (14.65 psia and 6OOF) 710 733 757 781 805 829 853 877 901 925 948 972 996 1,020 1,044 1,068 1,116 1,163 1,211 1,259 impact qg, 103 cu ft/D (14.65 psia and 60F) 1,307 1,426 1,546 1,665 1,785 1,904 2,023 2,143 2,262 2,501 2,740 2,979 3,218 3,457 3,697 3,935 4,174 4,412 4,651 4,890 5,129
Pressure
Wg) 40 z: 55 60 65 70 75 80 90 100 110 120 130 140 150 160 170 180 190 200
lo3
test showing
is especially suitable for determining the performance characteristics of wells producing from reservoirs with low permeability. High-Pressure Gas and Gas-Condensate Wells
Multiply rate of tlow from table by the square 01 the diameter for Row nipples with ID s mere than I.000 in
The backpressure method of testing wells requires that a series of flow rates and corresponding pressure measurements be obtained under stabilized conditions or at certain fixed time intervals. Testing under stabilized pressure and rate-of-flow conditions or according to a fixed time interval has become known as multipoint or flow-after-flow backpressure testing. As the original backpressure or multipoint method came into general use, it became evident that the method of testing was applicable to those wells that approached stabilized producing conditions within a relatively short time. However, performance characteristics could not be determined by this method for wells that approached stabilized producing conditions slowly over a considerable period. This characteristic of slow stabilization has been associated generally with wells producing from reservoirs with low permeability and resulted in the development of the isochronal method of backpressure testing by Cullender. 4 The procedure used to obtain the necessary performance data for the isochronal testing method is to open the well from a shut-in condition and allow the well to flow without disturbing the rate by changing the mechanical adjustment of chokes or valves for a specific period of time. The well is then shut in and allowed to return to a shut-in pressure comparable with that existing before the well was first opened, after which the well is again opened at a different rate of flow. In isochronal testing, each rate of flow starts from a comparable shutin condition, which provides a means of maintaining a simple pressure gradient throughout the drainage area of the well during testing. The isochronal method of testing
All the instructions for testing wells in this chapter apply to gas wells that produce a single-phase gas into the wellbore or to wells that are predominantly gas wells and the fluid flowing in the reservoir has a high in-place gas/liquid ratio (GLR). However, these methods for testing gas wells have been applied to high-ratio oil wells with some degree of success. The chief difference between testing methods for highpressure gas and gas-condensate wells and low-pressure wells is the care used in taking the data and methods used in computing the results. The effect of liquids is usually more pronounced in high-pressure than in low-pressure wells. Consequently, special care should be used to measure GLR s in high-pressure wells. Often it is necessary to determine the GLR at each rate of flow during a backpressure test. If the ratio was not constant during testing, the well probably was accumulating liquid in the wellbore during testing or unloading liquid. In either case the test is probably not acceptable and the well should be cleaned by flowing at a high rate and retested at rates of flow high enough to keep the well free of liquid. Temperature effects during testing of high-pressure wells may be troublesome in interpreting test results. For example, Well B (Fig. 33.2) has a shut-in wellhead pressure of 4,173 psia at a wellhead temperature of 117F. Maximum wellhead pressure was observed 3 minutes after the well was shut in. If wellhead pressure has been observed for an extended period of time, the wellhead pressure would have decreased to about 4,140 psia. The decrease in wellhead pressure is caused by the cooling of the gas in the well. In general, better tests can be obtained on such large-capacity wells if the testing is done after a preflow period. The preflow period should be run long enough to bring wellhead temperatures to a normal operating range of temperature. Wellhead
OPEN
FLOW
OF GAS WELLS
33-5
temperatures should be recorded during testing at periodic intervals so that actual measured temperatures can be used in computing subsurface pressures by methods outlined under Example 3 in the section on computing subsurface pressures. Official Testing Official testing of gas wells for state regulatory bodies is usually a multipoint test of short duration. In addition to the multipoint test, a single rate of flow for a period of 24 to 72 hours is required in some cases. The tester is referred to the test manuals of the various states, provinces of Canada, or appropriate countries for exact procedures, and no attempt is made here to outline official testing. Backpressure Equations
where C, the performance coefficient, and n, the exponent, are different from those in Eq. 5 for a given well. prs and ptf represent wellhead shut-in (static tubing) pressure and working (flowing tubing) pressure on the flowing-gas column at the wellhead, respectively. Eq. 6 is useful especially in estimating the capacity of a well to deliver gas into a pipeline under specified conditions.
In either multipoint or isoehronal backpressure testing, the rates of flow and the corresponding values of the difference of squares of the average formation (reservoir) pressure pR and the sandface pressure [bottomhole flowing pressure (BHFP)] p,,f are plotted on logarithmic coordinates and a straight line is drawn through the points. The equation for the relationship is qg =C(F,2 -p,J), . . I. , . . . . . . . . . . . . . . . (5)
where the performance coefficient is represented by C and the exponent of the backpressure curve by n. The industry by common usage has referred ton as the slope of the backpressure curve, even though n is the reciprocal of the mathematical slope of the line. Here n is referred to as the exponent of the backpressure curve. Eq. 5 is an empirical relationship for both the multipoint test and the isochronal test and has resulted from the study of results of many tests. Values of the exponent vary for individual wells in the range of 0.5 to 1 .O. Tests that result in exponents less than 0.5 or more than 1.O should be rerun. Exponents of less than 0.5 resulting from multipoint tests may be caused by the slowstabilization characteristics of the reservoir or by the accumulation of liquids in the wellbore. Exponents greater than 1.O may be caused by the removal of liquid from the well during testing or by a cleaning of the formation around the well, such as the removal of drilling mud or stimulation fluids. Also, a multipoint test run in decreasing rate sequence may have an exponent of more than 1.O for wells in slow-stabilizing reservoirs. Erratic exponents in isochronal testing are caused by either accumulation or cleaning of liquids from around the well. Erratic alignment of data points from multipoint or isochronal testing is usually caused by changes in actual well capacity during testing. Such changes may be caused by accumulation of liquids or the cleaning of the wells. The effects of the liquids in the well on multipoint testing have been given in detail by Rawlins and Schellhardt. Eq. 5 represents the capacity of a well to deliver gas into the wellbore, and it is useful especially in evaluating reservoir conditions. The capacity of a well to deliver gas at the wellhead may be represented by qy =c(p,\? -pl,2) i, . . (6)
-p, )
= Cl,
The maximum rate of flow [open-flow potential (OFP)] is given when Ct is a maximum-that is, whenp,f = 0. Eq. 7 incorporates the properties of the reservoir and the gas and can be extended to noncircular areas as given in Ref. 5.
33-6
PETROLEUM
ENGINEERING
HANDBOOK
Determination
The terms calculated absolute open flow (CAOF) and OFFare the rate of flow in thousands of cubic feet of gas per 24 hours that would be produced by a well if the pressure against the face of the producing formation in the wellbore were zero. The value of the OFP is usually determined graphically by plotting rates of flow qg against the corresponding values of a,$ -p wf2. The straight-line relationship between qg and jji -pwf2 is extended so the rate of flow qg corresponding to the value ofp,$ can be read by extrapolation. Then qs is the AOFP of the well in cubic feet per 24 hours. The AOF can be computed from Eq. 5 or read directly from plotted relationships. In wells producing from reservoirs with low permeability, the reported AOF must be identified further by the time involved in the test and the type of test. For example, the OFP of such a well as determined by a 3-hour multipoint test (each rate of flow lasting 3 hours) would be less than that determined by a 2-hour multipoint test. The open flow determined by an isochronal test of 3 hours would be different from that determined by a multipoint test. A good example of the relationship between AOF and type of test is given by Cullender.4 Reported OFP s on wells in low-permeability reservoirs are more or less meaningless without an indication of the type of test involved. Determination of the Exponent n
for 24 hours or longer to equalize the reservoir pressure in the vicinity of the well. Wells with slow pressurebuildup characteristics should be shut in 48 to 72 hours, if possible. While the well is shut in, the gas-measurement equipment should be prepared for use. If the gas is to be measured with an orifice meter, the meter should be calibrated, the diameters and condition of the run and plate verified, and the differential pen should be zeroed in accordance with good meter practice. If a critical-flow prover (see later section on gas measurement) is used, it should be placed in a vertical position at the wellhead or downstream from the separator so that the gas will flow up and away from the test area. If a separator is used, control the rate of flow with a production choke and maintain pressure on the separator with a critical-flow prover or backptessure regulator when an orifice meter is used. If a separator is not used, control rate of flow and pressure at the wellhead with the critical-flow prover. Always install thermometer wells at the wellhead and at gas-measuring equipment so that temperatures may be measured with a thermometer or calibrated recording device. The thermometer wells should be filled with water or oil to obtain accurate temperature measurement. Shut-In Pressure All shut-in or flowing pressures should be measured with a dead-weight or piston gauge, because spring gauges are usually not accurate enough for backpressure tests. Determine and record the pressure at the end of the shutin period, prepare the well for testing, and redetermine the shut-in pressure as a check on the first measurement and to obtain the rate of pressure buildup. Report each pressure and time the well was shut in prior to each pressure measurement. After the second pressure measurement, either the isochronal or multipoint test may be started. Subsurface pressures in gas wells may be measured directly with pressure gauges or computed from wellhead pressures. Subsurface-pressure gauges are very useful in wells where liquids accumulate in the wellbore during shut-in. However, the use of subsurface gauges limits the rates of flow during the backpressure test to velocities that will not lift the gauge in the flow string. The use of subsurface gauges is limited to rather low rates of flow in 2%-in. OD tubing, but there is practically no limitation on their use in 7-in. casing. The use of subsurface gauges in the annular spaces of dual completions is practically impossible. In cases where largecapacity wells are being tested, correction must be made for the effect of hysteresis on gauge readings, or the BHP must be measured at each rate of flow by a separate run of the gauge. The accumulation of liquid in the wellbore is probably the most serious cause of erroneous calculated BHP s. Other sources of error are uncertainties in temperature gradients and specific gravities of the fluids flowing in the well. Before a backpressure test is begun, special care should be taken to remove the liquids from the wellbore by flowing at rates large enough to lift the liquid. If possible, each rate of flow used in the backpressurc test should be large enough to lift continuously any liquid that may move into the wellbore during production. Temperature gradients can be
1%qg2
log@j -p&
-1%
481
-log(J,If
-p&,
. . .
Values of qg and corresponding values of pi -p,,f* , either actual experimental points or values read from the straight-line relationship, are substituted in Eq. 9. Usually the data points do not fall exactly on a straight line; so the best practice is to read values of qg and jji -pwf2 directly from the straight line. Determination of the Performance Coefficient C
After the value of the exponent n has been determined, the value of the performance coefficient C may be determined by substitution of a corresponding set of values for qg and a,$ -pwf2 and the value of n into Eq. 5. The value of C is found by solution of the resulting equation. Graphically the value of C may be determined by extension of the straight-line relationship to jj,$ -t,f2 = 1 and reading the corresponding q!. When jr~ -pwf2 is unity, C is equal to qg. In practtce, a determination of the value of C is seldom necessary for routine analysis of backpressure tests. Preparation of Well for Testing The wellbore should be cleaned of liquids by flowing at a high rate to a pipeline for a period of 24 hours. If the well does not have a pipeline connection, it may be blown to the air for a short period of time, provided blowing is considered safe. Extra precautions should be taken on new wells to remove drilling mud, solids, and stimulation fluids from the wellbore. The well should be shut in
OPEN
FLOW
OF GAS WELLS
33-7
established for a new area only by actual measurement. Usually a flowing-temperature gradient can be estimated by assuming a straight-line gradient between flowing wellhead temperature and bottomhole temperature. Uncertainty in the specific gravity of the fluid flowing in the well can be eliminated to a large degree by careful measurement of gas/hydrocarbon liquid ratio and determination of the specific gravity of the separator gas, separator liquid. and stock-tank liquid.
changed and when the well opened on each rate of flow were carefully reported. The remarks column gives the results of the specific-gravity measurement and the condition of the flow with regard to whether the well was producing water. All the observations recorded in Table 33.3 are necessary for accurate analysis of test results. Computation of the results of a backpressure test on a gas well involves the following steps. I. Compute rates of flow and pressures at the face of the producing formation from pressure and volume observations made at the wellhead. 2. Determine values of p,: -prf2 and p,$ -p,+f* and rates of flow corresponding to these pressure factors. Then, PR and pwf are calculated at the midpoint of the sandface in wells without tubing. If the well has tubing, they are determined at the entrance to the tubing, provided the entry to the tubing is no more than 100 ft from the midpoint of the sandface. 3. Plot values of q8 and corresponding values of PR-pwf2 and pt: -pti2 on logarithmic coordinates. 4. Determine values of the exponent n and the performance coefficient C of the flow equations qg =C(Pl? -Pwf2Y and 4x =c(P,.? -Pff2Y*. For most routine analyses of backpressure tests. determination of the value of C is not necessary. 5. Determine the CAOF. Computations for rate of flow and pressures at the producing formation are explained in separate sections. A convenient form for reporting the results of a multipoint test is illustrated in Table 33.4 for the test data taken on Well A and reported in Table 33.3. Table 33.4 shows general well information, a summary of test data, calculation of rates of flow, data for determining compressibility, and the difference of squares of pressures for wellhead and bottomhole conditions. The calculated OFP of 25,000~10~ cu ft/D was determined in Fig. 33.3 where the rate of flow is the abscissa and PJ -p w,2 (in thousands) is the ordinate on logarithmic coordinates. The data points were connected by a straight line and extrapolated to a value of PJ -p,,,f2, where p&f2 is zero. In this case, the value is j~i = 230.9 (thousands). The corresponding rate of flow is 25,000 x lo3 cu ft/D. The AOF of 25,000~ lo3 cu ft/D for Well A is for a 3-hour four-point test. If the test were for a lesser-time fourpoint test, the resulting AOF would have been more than 25,000 x lo3 cu ft/D. The exponent n was determined by taking values of qs andpd -pwf2 from the straight line in Fig. 33.3 and Eq. 7 as follows. qs, lo3 cu ft/D 20,000 4,ooo n= log 20,000-log log 168-log =0.699. 4,000 16.8 p,$ -p wf2 (thousands) 168 16.8 log 5.00 =-=log 10.0 0.699 I .ooo
33-a
PETROLEUM
ENGINEERING
HANDBOOK
TABLE
33.3-FIELD
TEST (WELL A)
rema
carnty.
okuha
Well P
No.Taps
WELLHEAD WORKING PRESSURE Csiq. An;&us Tey. Tbg. Psi0 W u*ll hut lo fo a dam I /
REMARKS
PAGEL
OF JDATA
BY J. it.
J.
The performance coefficient C was determined from the exponent n=0.699, Eq. 5, and one of the corresponding values of qg and j12 -pd* as follows. 20,000 C log C log c c = = = = = C(168)0.699 20,000/(168)~699 log 20,000-0.699 4.3010-1.5555 557.
33.4 where qg and corresponding pt: -ptf* values ate plotted on logarithmic coordinates. The straight line has been extended to show a wellhead OFP of 22,000 X lo3 cu ft/D. The exponent is 0.672 and C is 621. The backpressure relationship corresponding to Eq. 6 is qR =627(pr,2 -P~~)O.~ *. This wellhead performance equation for Well A, illustrated in Fig. 33.4, is a measure of the ability of Well A to deliver gas at the wellhead through 5 %-in. casing as indicated by the multipoint test given in Table 33.4. The relationship is influenced by the size of the flow string and hydrostatic head of the gas column as well as the productive capacity of the well. An example of the bottomhole performance as indicated by a multipoint test is given in Fig. 33.2 for an extremely large-capacity well. Well B (Fig. 33.4) had a shut-in pressure~R of 5,169 psia at a depth of 10,658 ft and a wellhead pressure of 4,173 psia. The calculated OFP was 280,000x103 cu ft/D. The corresponding wellhead performance for Well B producing through 2X-in.-OD, 6.5-lbm/ft tubing is illustrated in Fig. 33.5 where the data points for the test are plotted as circles.
log 168
The value of 557 may be checked b extrapolating the straight line on Fig. 33.3 top2 -p,,,, z- - 1 and reading the corresponding value of qs. Note that the value of C=557 is for qg in units of lo3 cu ft/D and for jr2 -pg* in units of thousands. The backpressure equation for the results of the multipoint test on Well A given in Table 33.4 and illustrated in Fig. 33.3 is
The wellhead Performance of Well A as determined by the test results given in Table 33.4 is illustrated in Fig.
OPEN
FLOW
OF GAS WELLS
33-9
TABLE
33.4--RESULTS
OF MULTIPOINT
BACKPRESSURE
TEST (WELL A)
COMPANY ADDRESS DISTRICT LOCATION CASING SIZE5 Tens CmmtJ, WT.x WT.--2Lb5 o.wl6105 TO 2620 DATE OF COMPLETION BAROMETER Okldwm I D I.0 ~ Ii A 5.012 FIELD
LEASE
DATE
WELL 6-17
N0.I 19k
~goton
RESERVOIR
Uwota
SET SET
Al
269
PERF PERF. 90
2665-2620
Ca
25fQ I -
CASING ACRES
F,
REMARKS:
13.2
P,.
uB.5
Pa
Le3.5
Pb230.92xlO
Potmtbl n
25.m 0.699
IO C fm
3001------
I01 1000
qg, lo3 Fig. 33.3-Multipoint Well A. test showing cu ft/D bottomhole performance for Fig. 33.4-Multipoint Well A.
II,,,, 10,000
30,cQo
PETROLEUM
ENGINEERING
HANDBOOK
3-
3-
3-
:0
0
q9,
10,000
lo3 cu ft/D wellhead
IOC
Fig. 33.5-Multipoint
test
showing
performance for
Well 0.
Data points represented as squares (Fig. 33.5) are flow tests of several days duration with Point 1 taken shortly after production started and Point 3 taken over a year later. The position of the data points in Fig. 33.5 indicated that the performance of Well B improved after the well was placed on production, which was probably caused by the removal of drilling fluids from the area around the wellbore. The wellhead OFP of Well B was 41,000~ IO3 cu ND, which was the approximate capacity of the tubing. A different wellhead performance curve would result if the tubing were changed in Well B. The wellhead performance for a different string of tubing can be calculated by starting with the bottomhole performance curve in Fig. 33.2 and calculating the pressure drop caused by friction for the different string of tubing.
establishment of a simple pressure gradient around a producing well during the test period, which prevents the variation of the performance coefficient with time from obscuring the true value of the exponent. The determination of the relationship between performance coefficient and time permits the estimation of the rate of flow of a given well into a pipeline over long periods of time. The term isochronal was adopted as being descriptive of the method, because only those conditions existing as a result of a single disturbance of constant duration are considered as being related to each other by Eqs. 5 and 6. The expression single disturbance of constant duration is defined as those conditions existing around a well as a result of a constant flow rate for a specific period of time from shut-in conditions. Under actual test conditions this requirement is rarely satisfied. However, this condition may be approximated by starting a well on production and allowing the well to produce without further outside or mechanical adjustments in rate of flow. Thus a simple pressure gradient is established around the wellbore as opposed to a complex pressure gradient resulting from a multipoint backpressure test. The presentation of isochronal test data as a series of parallel curves with a constant exponent n and a constant performance coefficient C for a specific time interval involves certain assumptions. The exponent of the performance curves for a gas well is assumed independent of the drainage area. It is established immediately after the well is opened. The variations of the performance coeffcient with respect to time are believed to be independent of the rate of flow and the pressure level under simple gradient conditions. The procedure employed to obtain the necessary performance data for an isochronal test is to open a well from shut-in conditions and obtain rate-of-flow and pressure data at specific time intervals during the flow period without disturbing the rate of flow. After sufficient data have been obtained, the well is shut in and allowed to return to a shut-in condition comparable with that existing at the time the well was first opened. The well is again opened at a different rate of flow with data being obtained at the same time intervals as before. The procedure may be repeated as many times as necessary to obtain the desired number of data points. With the exception of starting each rate of flow from shut-in conditions, the procedure for running isochmnal tests is the same as that for the multipoint test. The necessity for cleaning the well, calibrating the gasmeasuring equipment, and accurately measuring pressures and temperatures remains the same. At least four rates of flow should be taken; the lowest rate should reduce the pressure at the wellhead about 5% and the highest rate of flow should reduce the pressure about 25%. The results of an isochronal test are computed in the same manner as those for a multipoint test. The data points are plotted on logarithmic coordinates as illustrated in Figs. 33.6 and 33.7. The isochronal curves are drawn so that the points taken at a constant time for the various rates of flow are joined by a straight line. For example, all the points on the line labeled Time, 3 hr in Fig. 33.6 represent the performance of Well A after flowing at the various rates of flow for 3 hours from shut-in conditions.
OPEN
FLOW
OF GAS WELLS
33-11
3oor--,
0
; .P zi N^ ct I N :
loo-
IO 1000
I
l0,000 qg. IO3 cu WD
III
60,000
IO_ 1000
* 10,000
II 50,000
qg.
performance for
lo3
Fig. 33.6~lsochronal
Well A.
test showing
bottomhole
Fig. 33.7~lsochronal
Well A.
test showing
The results of isochronal tests can be analyzed in two ways. One way is to use Eqs. 7 and 8 and the properties of the gas to determine the properties of the reservoir and the skin factor. The second way is to use the results as a basis for comparison of well performance at the time of the test with performance as measured previously or to set a base against which future performance is to be compared. The isochronal type curves shown on Fig. 33.6 can be used to estimate the pressures that would have been observed if the test had been a constant-rate drawdown test. Test periods longer than the 3-hour periods on Fig. 33.6 are much more desirable for this purpose. With this information the k,h value for the reservoir and the total skin value (s,=s+F,~q~) are calculated as given in Chap. 35. This results in several values for the total skin, s,, as a function of the rate of flow, qg, from which s and F,D can be obtained for use in Eq. 7.6 The multipoint test can be analyzed to obtain k,h, s, and F,JJ as indicated by Ref. 7. A discussion of the performancecomparison method follows. A copy of actual field data for an isochronal test is given in Table 33.5 for Well A, which is the same well used in the example of a multipoint test.* Four rates of flow of 3 hours duration were used with each flow starting from shut-in conditions. Shut-in pressures reported varied from 359.6 psig after 48 hours for the first rate of flow to 357.6 psig, which was just previous to the fourth rate of flow. The results of the isochronal test are summarized in Table 33.6. Bottomhole and wellhead performance curves are illustrated on Figs. 33.6 and 33.7, respectively. The isochronal test on Well A (Fig. 33.6) shows that the calculated OFP for a BHP of 399.1 psia was 5 1,500, 41,500, 35,ooO, and 31,500~ lo3 cu ft/D at the end of 0.5, 1 .O, 2.0, and 3.0 hours, respectively. The calculated potential after 3 hours flow was only 61% of the potential after 0.5 hour of flow, A similar figure for the wellhead performance of Well A is 66% (Fig. 33 -7) If Well A were opened into a pipeline with a constant
This test, used I the 1962 edltlan, was run many classtc tsochronal test example today. years ago. It still stands as a
pressure, the rate of flow at the end of 3 hours would be 66% of the rate of flow at 0.5 hour. Experimental data not given here show that the production at the end of 72 hours has decreased to about 48% of that at 0.5 hour. The figures showing change-of-performance characteristics with time illustrate the need for isochronal test data for estimating the delivery from a particular well into a pipeline. Accurate estimation of pipeline deliveries from wells producing from reservoirs with low permeability is practically impossible without isochronal test data. Examination of the field notes under the Remarks column in Table 33.5 indicates that Well A started to produce water during the flow test taken on Dec. 20, 195 1, which was the largest rate of flow. The effect of water production on well performance is illustrated by the irregularities in the corresponding data in Figs. 33.6 and 33.7. Water production and accumulation of water or liquids in the wellbore cause the performance characteristics of a well to deteriorate. The data represented as squares in Fig. 33.5 are isochronal points taken after Well B has been flowing from 5 to 30 days. Their close agreement with the data from the multipoint test indicates that the performance of Well B does not vary appreciably with time. Well B produces from a reservoir with high permeability and the radius of drainage is established quickly after the well is opened to flow.
33-12
PETROLEUM
ENGINEERING
HANDBOOK
TABLE
33.5-FIELD
TEST (WELL
A)
company
Location 2 Prover
Tmaa
CounW.
LKUba
I I
I I
I I
I PAGEA
I OFAOATA
I BY J. If. J.
TABLE
33.6--RESULTS
OF ISOCHRONAL Flow Duration (hr) 0.5 1 .o 2.0 3.0 0.5 1 .o 2.0 3.0 0.5 1 .o 2.0 3.0 0.5 1 .o 2.0 3.0
TEST ON WELL A
Pts
Date Dec. 17, 1951 (psi4 372.6
(M%D) 4,159 4,120 4,078 4,047 5,552 5,485 5,461 5,423 7,019 6,982 6,847 6,777 8,599 8,153 8,048
-P,,$ 10.60 13.19 15.79 17.63 14.85 18.10 21.58 23.56 17.93 22.17 26.48 26.80 24.96 34.10 37.00
PrsZ - Pn2 10.60 12.82 15.08 16.55 15.18 17.97 20.87 22.58 20.59 24.14 27.76 29.81 27.93 35.22 37.63
372.0
372.0
370.8
The results of the 3-hour multipoint test and the 3-hour isochronal test on Well A are shown together in Fig. 33.8 as wellhead performance curves. The exponent (0.672) for the multipoint test is less than the exponent (0.848) of the isochronal test. In general, exponents of multipoint curves run in increasing rate sequence are less than those for isochronal curves for the same well. The first data point on the multipoint test ( qn =4,928 Mcf/D) is on the isochronal curve (Fig. 33.8) because the first rate of flow for the multipoint curve was started from shut-in conditions. Thereafter, the position of each succeeding point of the multipoint test is influenced not only by the rate of flow but also by each preceding point. The initial points of each multipoint test on wells producing from reservoirs with low permeability represent the formation characteristics, while other points represent complex conditions that are almost impossible to interpret. The characteristic exponent of the isochronal curve still applies to the complex points, with the only difference in performance being in the performance coefficient C. If the exponent of 0.848 is applied to the complex points of the multipoint test (Fig. 33.8). it can be
OPEN
FLOW
OF GAS WELLS
33-13
TABLE 33.7-BASIC ORIFICE FACTORS FOR CRITICAL-FLOW PROVER (USBM plate design) F, - McflD
Base temperature, OF 60 Base pressure, psia 14.65 Flowing temperature, OF 60 Specific gravity 1.000 2-in. Prover Orifice Diameter (in.) Factor (F,) Orifice 4-k Prover Factor (FP) 1.074 2.414 4.319 6.729 9.643 13.11 17.08 21.52 26.57 31.99 38.12 52.07 68.80 88.19 110.6 2% 3 136.9 168.3
Diameter (in.)
0.06569
0.1446 0.2716 0.6237 0.8608
1.115 1.714 2.439 3.495 4.388 6.638 9.694 13.33 17.53 22.45 28.34 34.82 43.19
seen that the coefficient obtained in each case can be considered the result of an effective time, which has no permanent significance because it is not equal to the elapsed time or to the elapsed time since the last change in flow rate. An examination of the multipoint and isochronal data presented in Fig. 33.5 for Well B shows that there are certain gas wells that stabilize so rapidly that there is no necessity for obtaining isochronal performance data. As the time required for stabilization increases, the differences between data obtained by the isochronal test and the multipoint test increase.
F,=
i, 7s
. . . . . . . . . . . . . . . . . . . . . . . . . . ...(n)
Gas Measurement
Orifice Meters The recommended specifications for orifice meters and methods for computing rates of flow are those published by the American Gas Assn. 9 It should be noted that the basic orifice factors are for a pressure base of 14.73 psia. Multiplying the basic orifice factors in Ref. 9 by 1.0055 changes volumes to a pressure base of 14.65 psia. Basic orifice factors,for a pressure base of 14.65 psia have been published in the test manual of the Corporation Commission of the State of Kansas lo and the Interstate Oil Compact Commission. Critical-Flow Provers The following method for measurement and computation of rates of flow for critical-flow provers is a modification of the method published by Rawlins and Schellhardt.* The equation computing rates of flow from measurements with a critical-flow prover is qg=psFpFgFTFpv, ............ .... . .(lO)
where yg is specific gravity of the flowing gas, air = l.COO. Factors to correct for an assumed flowing temperature of 60F to the actual flowing temperature of the gas at the point of measurement are given in Table 33.9 and may be computed by FT= 520 -, Tf __. _. . . . . . . . . . . . . . . . . (12)
where Tf is actual flowing temperature of the gas, (F +460). The supercompressibility factor to correct for the effect of gas compressibility is computed from the compressibility by F,,\,= r i, z . . . . . . . . . . . . . . . . . . . . . . (13)
where z is compressibility of the gas at ps and Tf or the pressure and temperature at point of measurement. Methods for estimating gas compressibilities are given in Chap. 20.
where p,, is static pressure on critical-flow prover, psia. Basic orifice factors, F,, , for 2- and 4-in. critical-flow provers are given in Table 33.7. These factors apply only to plates designed according to USBM specifications. The adjustment factor (Table 33.8) to correct for an assumed specific gravity of 1.000 to the actual specific gravity of the gas flowing through the prover may be computed by
33-14
PETROLEUM
ENGINEERINGHANDBOOK
TABLE 33.8--SPECIFIC-GRAVITY
ADJUSTMENT
FACTOR
Specific Gravity 0.550 0.560 0.570 0.580 0.590 0.600 0.610 0.620 0.630 0.640 0.650 0.660 0.670 0.660 0.690 0.700 0.710 0.720 0.730 0.740 0.750 0.760 0.770 0.780 0.790 0.800 0.810 0.820 0.830 0.840 0.850 0.860 0.870 0.880 0.890 0.900 0.910 0.920 0.930 0.940 0.950 0.960 0.970 0.980 0.990
0.000
1.348 1.336 1.325 1.313 1.302 1.291 1.280 1.270 1.260 1.250 1.240 1.231 1.222 1.213 1.204 1.195 1.187 1.179 1.170 1.162 1.155 1.147 1.140 1.132 1.125 ,118 ,111 ,104 .098 ,091 ,085 ,078 ,072 1.066 1.060 1.054 1.048 1.043 1.037 1.031 1.026 1.021 1.015 1.010 1.005
0.001
0.002
0.003 1.345 1.333 1.321 1.310 1.299 1.288 1.277 1.267 1.257 1.247 1.237 1.228 1.219 1.210 1.201 1.193 1.184 1.176 1.168 1.160 1.152 1.145 1.137 1.130 1.123 1.116 1.109 1.102 1.096 1.089 1.083 1.076 1.070 1.064 1.058 1.052 1.047 1.041 1.035 1.030 1.024 1.019 1.014 1.009 1004
0.004 1.344 1.332 1.320 1.309 1.298 1.287 1.276 1.266 1.256 1.246 1.237 1.227 1.218 1.209 1.200 1.192 1.183 1.175 1.167 1.159 1.152 1.144 1.137 1.129 1.122 ,115 ,108 ,102 ,095 ,089 ,082 ,076 ,070 1.064 1.058 1.052 1.046 1.040 1.035 1.029 1.024 1.019 1.013 1.008 1.003
0.005
1.342 1.330 1.319 1.307 1.296 1.286 1.275 1.265 1.255 1.245 1.236 1.226 1.217 1.208 1.200 1.191 1.183 1.174 1.166 1.159 1.151 1.143 1.136 1.129 1.122 1.115 1.108 1.101 1.094 0.088 1.081 1.075 1.069 1.063 1.057 1.051 1.045 1.040 1.034 1.029 1.023 1.018 1.013 1.008 1.003
0.006 1.341 1.329 1.318 1.306 1.295 1.285 1.274 1.264 1.254 1.244 1.235 1.225 1.216 1.207 1.199 1.190 1.182 1.174 1.166 1.158 1.150 1.143 1.135 1.128 1.121 1.114 1.107 1.100 1.094 1.087 1.081 1.075 1.068 1.062 1.056 1.051 1.045 1.039 1.034 1.028 1.023 1.017 1.012 1.007 1.002
0.007 1.340 1.328 1.316 1.305 1.294 1.284 1.273 1.263 1.253 1.243 1.234 1.224 1.215 1.206 1.198 1.189 1.181 1.173 1.165 1.157 1.149 1.142 1.134 1.127 1.120 1.113 1.106 1.100 1.093 1.087 1.080 1.074 1.068 1.062 1.056 1.050 1.044 1.039 1.033 1.028 1.022 1.017 1.012 1.007 1.002
0.006 1.339 1.327 1.315 1.304 1.293 1.282 1.272 1.262 1.252 1.242 1.233 1.224 1.214 1.206 1.197 1.188 1.180 1.172 1.164 1.156 1.149 1.141 1.134 1.127 1.119 1.112 1.106 1.099 1.092 1.086 1.080 1.073 1.067 1.061 1.055 1.049 1.044 1.038 1.033 1.027 1.022 1.016 1.011 1.006 1.001
0.009
1.338 1.326 1.314 1.303 1.292 1.281 1.271 1.261 1.251 1.241 1.232 1.223 1.214 1.205 1.196 1.188 1.179 1.171 1.163 1.155 1.148 1.140 1.133 1.126 1.119 1.112 1.105 1.098 1.092 1.085 1.079 1.073 1.067 1.061 1.055 1.049 1.043 1.038 1.032 1.027 1.021 1.016 1.011 1.006 1.001
1.3471.346 1.335 1.334 1.323 1.322 1.312 1.311 1.301 1.300 1.290 1.279 1.269 1.259 1.249 1.239 1.230 1.221 1.212 1.203 1.194 1.186 1.178 1.170 1.162 1.154 1.146 1.139 1.132 1.124 1.117 1.110 1.104 1.097 1.090 1.084 1.078 1.072 1.065 1.059 1.054 1.048 1.042 1.036 1.031 1.025 1.020 1.015 1.010 1.005 1.289 1.278 1.268 1.258 1.248 1.238 1.229 1.220 1.211 1.202 1.194 1.185 1.177 1.169 1.161 1.153 1.146 1.138 1.131 1.124 1.117 1.110 1.103 1.096 1.090 1.083 1.077 I.071 1.065 1.059 1.053 1.047 1.041 1.036 1.030 1.025 1.020 1.014 1.009 1.004
usually necessary to compute the specific gravity of the fluid flowing in the wellbore. The shrinkage of the liquid between the separator and the stock tank is usually unknown and apparently can be neglected. The equation for computing the specific gravity of the flowing fluid, ysJ, is: R &Yg +4m3
-rff= R,L~ L YL ) . .. . .
yL = specific gravity of hydrocarbon liquid referred to water, and VL = vapor volume equivalent of 1 bbl (60F) of hydrocarbon liquid, cu ftibbl. The specific gravity and the approximate vapor volume of the hydrocarbon liquid can be calculated from the API gravity by (14b)
OPEN
FLOWOF
GAS WELLS
33-15
TABLE 33.9-FLOWING-TEMPERATURE
ADJUSTMENT
FACTOR
Observed
Temperature
lOFl
0 1.063 1.052
1 1.062 1.051
2 1.061 1.050 1.039 1.028 1.018 1.008 0.9981 0.9887 0.9795 0.9706 0.9619 0.9535 0.9452 0.9372 0.9294 0.9217 0.9143 0.9069 0.9000 0.8931 0.8863 0.8797 0.8732 0.8668 0.8606
3 1.060 1.049 1.038 1.027 1.017 1.007 0.9971 0.9877 0.9786 0.9697 0.9610 0.9526 0.9444 0.9364 0.9286 0.9210 0.9135 0.9063 0.8992 0.8923 0.8856 0.8790 0.8725 0.8662 0.8600
4 1.059 1.047 1.037 1.026 1.016 1.006 0.9962 0.9868 0.9777 0.9888
5 1.057 1.046 1.035 1.025 1.015 1.005 0.9952 0.9859 0.9768 0.9680
6 1.056 1.045 1.034 1.024 1.014 1.004 0.9943 0.9850 0.9759 0.9671
7 1.055 1.044 1.033 1.023 1.013 1.003 0.9933 0.9840 0.9750 0.9662
8 1.054 1.043 1.032 1.022 1.012 1.002 0.9924 0.9831 0.9741 0.9653
9 1.053 1.042 1.031 1.021 1.011 1.001 0.9915 0.9822 0.9732 0.9645
50 60 70 80 so 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240
0.9905
0.9813
0.9896 0.9804
0.9715
0.9723 0.9636
0.9551 0.9469 0.9388 0.9309 0.9233 0.9158 0.9085 0.9014 0.8944 0.8876 0.8810 0.8745 0.8681 0.8619
0.9628
0.9543 0.9460 0.9380 0.9302 0.9225 0.9150 0.9077 0.9007 0.8937 0.8870 0.8803 0.8738 0.8675 0.8613
0.9602
0.9518 0.9436 0.9356 0.9279 0.9202 0.9128 0.9055 0.8985 0.8916 0.8849 0.8784 0.8719 0.8656 0.8594
0.9594
0.9510 0.9428 0.9349 0.9271 0.9195 0.9121 0.9048 0.8979 0.8910 0.8843 0.8777 0.8713 0.8650 0.8588
0.9585
0.9501 0.9420 0.9341 0.9263 0.9187 0.9112 0.9042 0.8972 0.8903 0.8836 0.8770 0.8706 0.8644 0.8582
0.9577
0.9493 0.9412 0.9333 0.9256 0.9180 0.9106 0.9035 0.8965 0.8896 0.8830 0.8764 0.8700 0.8637 0.8576
0.9568
0.9485 0.9404 0.9325 0.9248 0.9173 0.9099 0.9028 0.8958 0.8889 0.8823 0.8758 0.8694 0.8631 0.8570
0.9560
0.9477 0.9396 0.9317 0.9240 0.9165 0.9092 0.9020 0.8951 0.8882 0.8816 0.8751 0.8687 0.8625 0.8564
VL=369+5yAP,
+O.O4(y Apr)2,
...
.(14c)
where yAPr is stock-tank oil gravity, API. The derivations of Eqs. 14a and 14c were given by Smith. I2 Equations for Computing Subsurface Pressures
where L = length of flowstring in well corresponding to H, fit H = vertical depth in well, ft, and 98 = rate of gas flow at 14.65 psia and 60F, lo6 scf/D.
Pressures at the sandface or at the inlet to the tubing in shut-in or flowing gas wells may be measured with BHP gauges or computed from wellhead pressures. However, most subsurface pressures in gas wells are calculated by equations. The most usable and realistic equations available are those of Cullender and Smith, I3 which have been adopted by the Kansas Corp. Commission, the Interstate Oil Compact Commission, and the New Mexico Conservation Commission, and by the Railroad Commission of Texas for cettain fields. The equations were revised * recently for use with programmable calculators and small computers. The revised flow equation for gas wells is
=(F rq 1 )*
...
. . Wa)
J
where
-72 f
4 log 7.4ri
K , . . . .. . .. ..
(16b)
p Tz-2.082F*+--
yg qg
*dp > I 1,000
f = coefficient of friction (friction factor), ri = internal radius of pipe, in., K = absolute roughness characteristic = 0.0006 in., and r/K = relative roughness. Refer to Ref. 12 for the background of Eqs. 15, 16a, and 16b. The second term in the numerator on the right side of Eq. 15 is the kinetic energy term that heretofore has been set at zero because the computations were made manually. Although the kinetic energy term can be neglected without appreciable error in the majority of cases, there is no need to do so when programmable
d,% H (~/Tz)~ L
. . . . . . . . . . . . . . . . . . . . . . (15)
33-16
PETROLEUM
ENGINEERING
HANDBOOK
TABLE
F, VALUES
FOR VARIOUS
FLOW STRING
(K = 0.0006
in.)
d,
(in.) (Ibmlft)
d,
(in.)
Minimum N Rs 139,000 189,000 224,000 284,000 355,000 445,000 525,000 605,000 626,000 612,000 659,000 638,000 Nominal Size (in.) Casmg
f=, 0.09505 0.04643 0.03105 0.01776 0.01050 0.006180 0.004184 0.002985 0.002755 0.002905 0.002442 0.002633
1.315 1.660 1.990 2.375 2.875 3.500 4.000 4.500 4.750 4.750 5.000 5.000
1.80 2.40 2.75 4.70 6.50 9.30 11.00 12.70 16.25 18.00 18.00 21 .oo
1.049 1.380 1.610 1.995 2.441 2.992 3.476 3.958 4.082 4.000 4.276 4.154
4%
(2)
(Ibmlft) 13.00 15.00 14.00 15.00 15.50 17.00 20.00 23.00 25.00 15.00 17.00 20.00 23.00 26.00 20.00 22.00 24.00 26.00 28.00 31.80 34.00 20.00 22.00 23.00 24.00 26.00 28.00 30.00 40.00 26.40 29.70 33.70 38.70
d, (in.)
Minimum N Fe
d,
(in.) (Ibmlft)
d,
(in.)
Fr
Minimum N fle
F,
5%
5.000 5.000 5.500 5.500 5.500 5.500 5.500 5.500 5.500 6.000 6.000 6.000 6.000 6.000 6.625 6.625 6.625 6.625 6.625 6.625 6.625 7.000 7.000 7.000 7.000 7.000 7.000 7.000 7.000 7.626 7.625 7.625 7.625
4.494
4.408 5.012
696,000 681,000 784,000 778,000 773.000 764;OOO 744,000 726,000 710,000 872,000 860.000 843;OO0 823,000 806,000 964,000 953,000 941,000 930,000 919,000 899,000 885,000 1.035.000 1;025;000 1,019,OOO 1,014,OOO 1.003.000
0.002146 0.002257 0.001617 0.001647 0.001670 0.001722 0.001830 0.001942 0.002043 0.001256 0.001301 0 001363 0.001440 0.001514 0.0009922 0.001018 0.001049 0.001080 0.001111 0.001171 0.001215 0.0008380 0.0008579 0.0008691 0.0008798 0.0009018
7=/a
7.625 8.000 8.125 8.125 8.125 8.125 8.625 8.625 8.625 8.625 8.625 8.625 8.625 8.625 9.000 9.000 9.000 9.000
45.00
26.00 28.00 32.00 35.50 39.50 17.50 20.00 24.00 28.00 32.00 36.00 38.00 43.00 34.00 38.00 40.00 45.00 36.00 40.00 43.50 47.00 53.50 58.00 33.00 55.50 61.20 32.75 35.75 40.00 45.50 48.00 54.00
6.445 7.386 7.485 7.385 7.285 7.185 8.249 8.191 8.097 8.003 7.907 7.825 7.775 7.651 8.290 8.196 8.150 8.032 8.921 8.835 a.755 8.681 a.535 8.435 9.384 8.908 8.790 10.192 10.136 10.050 9.950 9.902 9.784
1,033,OOO 1,199,ooo 1,216,OOO 1,199,ooo 1,181,000 1 ,I 63,000 1,353,OOO 1,342,OOO 1,326,OOO 1,309,OOO 1,292,ooo 1,277.OOO 1,268,OOO 1,246,OOO 1,360,OOO 1,343,ooo 1,335,ooo 1,314,OOO 1,473,OOO 1,458,OOO 1,444,OOO 1,430,OOO 1,404,OOO 1,386,OOO 1,557,OOO 1,47l,OOO 1,450,OOO 1,704,OOO 1,694,OOO 1,678,OOO 1,660,OOO 1,651,OOO 1,830,OOO
0.0008417 0.0005911 0.0005710 0.0005913 0.0006126 0.0006349 0.0004438 0.0004520 0.0004658 0.0004801 0.0004953 0.0005089 0.0005174 0.0005394 0.0004382 0.0004513 0.0004579 0.0004756 0.0003623 0.0003715 0.0003804 0.0003888 0.0004063 0.0004189 0.0003178 0.0003637 0.0003764 0.0002566 0.0002602 0.0002660 0.0002730 0.0002765 0.0002852
8%
5%
6%
6.049 5.989
5.921 5.855
5.791
5.675 5.595 6.456 6.398 6.366 6.336 6.276 6.214 6.154 5.836 6.969 6.875 6.765 6.625
6%
9.625 9.625 9.625 9.625 9.625 9.625 10.000 10.000 10.000 10.750 10.750 10.750 10.750 10.750 10.750
9%
.992;000 982,000
926,000 1,125,OOO 1,108,OOO 1,089,OOO 1,064,OOO
0.0009253 0.0009489
0.001089 0.0006872 0.0007119 0.0007423 0.0007837 10
7 h
calculators or computers are used. Eq. 15 is based on the assumptions that the flow is completely turbulent, the coefficient of friction, f, is a constant, the compressibility of the gas at base pressure and temperature conditions (14.65 psia and 60 F) is 1 .OOO,and only a gas phase is flowing. Eq. 15 has a subtle but important concept in the value of the quantity H/L at the wellhead, where both Hand L are zero. For a vertical wellbore, H = L and
H/L=~l&(~/L)=l.ooO. + In a deviated wellbore, H is less than L, and for a horizontal pipeline, H = 0, and as a result the term for the head of gas drops out of Eq. 15. For a complete guide to the algebraic convention for H and L, refer to Ref. 12.
OPEN
FLOWOF
GAS WELLS
33-17
TABLE33.11-F,VALUESFORVARlOUSANNULl Casing ID (in.) 4.154 4.276 4.408 4.494 4.580 4.670 4.778 4.892 4.950 4.976 5.012 5.140 5.240 5.352 5.450 5.524 5.595 5.675 5.791 5.836 5.855 5.921 5.989 6.049 6.154 6.214 6.276 6.336 6.366 8.398 6.445 6.456 6.625 6.765 6.875 6.969 7.185 7.285 7.385 7.386
(K=
O.OOlin.)
Tubino 1.900 0.005082 0.004576 0.004107 0.003838 0.003593 0.003361 0.003109 0.002872 0.002761 0.002713 0.002649 0.002438 0.002289 0.002137 0.002016 0.001931 0.001854 0.001773 0.001883 0.001623 0.001607 0.001551 0.001497 0.001452 0.001376 0.001336 0.001296 0.001259 0.001241 0.001222 0.001195 0.001189 0.001099 0.001032 0.0009830 0.0009439 0.0008619 0.0008273 0.0007946 0.0007943 2.375 0.008901 0.006oa7 0.005356 2.875 0.01093 0.009268 0.007867 0.007119 0.006473 0.005886 0.005281 0.004738 0.004492 0.004389 0.004251 0.003809 0.003509 0.003213 0.002983 0.002825 0.002684 0.002538 0.002346 0.002277 0.002249 0.002155 0.002064 0.001988 0.001865 0.001799 0.001735 0.001676 0.001647 0.001618 0.001576 0.001566 0.001429 0.001327 0.001255 0.001198 0.001079 0.001030 0.0009839 0.0009834 3.500
0.004948
0.004583 0.004242 0.003880 0.003544
0.01250 0.01086 0.009289 0.007980 0.007419 0.007187 0.006883 0.005947 0.005343 0.004770 0.004342 0.004055 0.003806 0.003552 0.003226 0.003111 0.003065 0.002911 0.002764 0.002643 0.002450 0.002349 0.002251 0.002161 0.002119 0.002074 0.002012 0.001998 0.001796 0.001651 0.001549 0.001469 0.001306 0.001240 0.001178 0.001177 0.01245 0.01012 0.008738 0.007506 0.006634 0.006074 0.005601 0.005133 0.004552 0.004354 0.004274 0.004012 0.003768 0.003570 0.003260 0.003100 0.002947 0.002810 0.002745 0.002878 0.002584 0.002583 0.002266 0.002057 0.001912 0.001800 0.001577 0.001487 0.001405 0.001404 0.01098 0.009783 0.008658 0.007351 0.006924 0.006755 0.006215 0.005726 0.005341 0.004757 0.004466 0.004193 0.003952 0.003839 0.003724 0.003565 0.003529 0.003041 0.002710 0.002486 0.002316 0.001987 0.001857 0.001740 0.001739
0.003390
0.003324 0.003235 0.002946 0.002746 0.002545 0.002385 0.002274 0.002175 0.002070 0.001930 0.001880 O.OOla59 0.001790 0.001722 0.001665 0.001572 0.001522 0.001472 0.001427 0.001405 0.001382 0.001349 0.001342 0.001234 0.001153 0.001095 0.001049 0.0009524 0.0009120 0.0008739 0.0008736
0.01017 0.009455 0.009176 0.008301 0.007528 0.006935 0.006057 0.005630 0.005235 0.004892 0.004734 0.004573 0.004352 0.004302 0.003639 0.003201 0.002910 0.002692 0.002276 0.002116 0.001972 0.001970
0.009582 O.OOa132 0.007451 0.006837 0.006313 0.006074 0.005835 0.005508 0.005436 0.004486 0.003879 0.003486 0.003196 0.002655 0.002450 0.002268 0.002266
Eq. 15 does not lend itself to mathematical integration without making assumptions regarding T and z, but it may be integrated over definite limits by the trapezoidal mle. If we let
(~3+~2)+...+(pn-pn-l)(zn+zn-1)],
. . . (18)
PI
s[
Pn p/Tz-2.082FZ+--
= pnz(jp= s 53.356
p,
+...+(pn-pn-,)(I,+Z,-*)], then
. . . . . . . (17)
whereZt,Z2,Z3 . . . I, is the trapezoidal rule interval corresponding to the respective pressure. If we make the assumptions that the kinetic energy term is zero or that the temperature, T, and the gas compressibility factor, z, are constant, the equations given in Chap. 30 of this handbook can be derived. However, the numerical examples that follow will make use of Eqs. 15 through 18. The details of computations of a BHFP and a shut-in BHP by Eqs. 16a, 16b, 17, and 18 are illustrated by Example Problem 3 on Page 33-18. To utilize the equations, it is necessary to evaluate the factor F, for various flow strings. The value may be determined by several correlations; however, the values given in Tables 33.10 and
33-16
PETROLEUM
ENGINEERING
HANDBOOK
TABLE 33.12-WORK
Company 0.615 ^lg 49 11.29s %CO,
SHEET FOR CALCULATION OF SUBSURFACE FLOWING PRESSURE BY EQS. 15,16a, AND 16b
Lease 2.5 % N2 , 10,490 10,658
PPC
B 361 Temperature
10,658
H -02.4413,913.0117 0
1,000 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8.000
L
1,000 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8.000 9,000 10,000 10,490 10,490 10,658
d 2.441 2.441 2.441 2.441 2.441 2.441 2.441 2.441 2.441 2.441 2.441 2.441 1.995 1.995
P
4,023.l 4,023.3 4.134.0 4,245.0 4,3X3 4,468.0 4580.0 4,692.3 4.805.0 4,917,s 5,031.l 5,086.7 5,086.7 5,112.0
T
122 122 127 132 137 142 147 152 157 162 167 169.5 169.5 170.3
2
0.8776
I
104.719 104.346 104.343 103.992 103.659 103.337 103.036 102.743 102.464 102.196 101.943 101.693 101.583 76.546 76.446
AP
0 110.1 110.3 110.7 111.0 111.3 111.7 112.0 112.3 112.7 112.9 113.2 55.6 0 25.3
23,018 23,059 23,063 23,049 23,039 23,052 23.047 231045 23.065 23,047 23,052 11,302 3,871
23,018 23,059 46,121 69,170 92,209 115,261 138,308 161,352 184.417 207,464 230,516 241,818 241,616
Y,L
0
23,053 46,105 69,158 92,211 115,263 138,316 161,369 184,421 207,473 230,526 241,822 241.822
Line 1 2 3 4 5 6 7 8 9 IO II 12 13 14 15
0.8894 0.8894
0.9012 0.9129 0.9244 0.9359 0.9472 0.9584 0.9695 0.9805 0.9913 0.9966 0.9966 0.9989
9,000
10,000 10,490 10.490 10,658
245,689
245,695
33.11 were calculated by the methods published by Smith. * To compute subsurface pressures where the well is equipped with tubing set without a packer, the preferred practice is to calculate the flowing subsurface pressure from the wellhead pressure measured on the static gas column by means of the static column equations. If the well has a packer, it is necessary to calculate the flowing subsurface pressure by means of the equations for flowing gas columns. Depths for calculating or measuring subsurface pressures in wells are determined in practice by the equipment installed in the well. Where a well is equipped without tubing or with tubing set without a packer, the proper depth for pressure determinations is the distance to the midpoint of the productive sandface. If the well has tubing set with a packer, the pressures are determined at the entrance to the tubing provided the entry to the tubing is no more than 100 ft from the midpoint of the productive sandface. Otherwise, appropriate corrections would be made to determine the pressure at the midpoint of the sandface. An explanation of the computational procedures used in Tables 33.12 and 33.13 will be helpful before going into the details of the calculations. The recent advances in computing equipment or, more realistically, the dramatic decrease in the cost of computations have given the average engineer access at least to a handheld programmable calculator or mom likely a microcomputer. Therefore, the emphasis in the past has been to simplify equations by making assumptions regarding pressure, temperature, and gas compressibility, but that has not been done here. Now the factor F, and compressibility factor, z, become subroutines, the results of which are never seen by the user. In this case, Tables 33.10 and 33.11 may seem redundant. The compressibility factors given in Tables 33.12 and 33.13 were calculated by the equation of state published by Hall and Yarborough t4 and Yarborough and Hall. l* The results of the computations in Tables 33.10 through 33.13 have been rounded,
and the rules for rounding vary from one piece of computing equipment to another. The algorithm used for solving Eqs. 15, 17, and 18 seems to work for all cases, but users may wish to devise their own algorithm.
Example Problem 3-Flowing Well. Details of the method for calculating a flowing subsurface pressure for Well B are given in Table 33.12. The wellhead flowing pressure for Well B was 3,913 psia at a flow rate of 11.299 x lo6 cu ft/D. The annular space between the tubing and casing was packed off and filled with mud so that it is necessary to calculate the flowing subsurface pressure at a depth of 10,658 ft down the flowing column of gas. Gas properties are those given in Table 33.12. The flow string measures 10,490 ft of 27/s-in.-OD, 6.50-lbm/ft tubing with 168 ft of 2%-in.-OD, 4.70-lbm/ft tubing at bottom of flow string. Also, H = L, or the flow string is vertical. Computation of the required pressure is done in two major steps because of the change in size of the flow string at a depth of 10,490 ft. Computations are given in the following steps. Step 1. Obtain the ID s from Table 33.10 and enter at top of Table 33.12 2% in.
0%ID =
2.441 in.
2% in. OD-ID
= 1.995 in.
Step 2. Determine the temperature gradient applicable to the problem. In this example, the flowing temperature of the gas at the wellhead was 117F, and the subsurface temperature at 10,658 ft was 170F. The temperature was assumed to be a straight-line relationship between 117F at H = 0 and 170F at H = 10,658 ft for a temperature gradient of 5F per 1,000 ft.
OPEN
FLOW
OF GAS WELLS
33-19
33.13-WORK
SHEET
FOR CALCULATION
SHUT-IN 0
PRESSURE
BY EQS. 15,16a,
AND 16b
Date of Test Equations Gradient WP) x Used 15,16a, 5Fll,000 37.464 x Line 0 23,053 46,105 69,156 92,211 115,263 136,316 161,369 184,421 207,474 230,526 245,695 1 2 3 4 5 6 7 6 9 IO 11 12 13 16b ft
679 N.A.
H
0 1,000 1,000 2,000 3,000 4,000 5,000 6,000 7,000 6,000 9,000 10,000 10.658
L
---ii 1,000 1,000 2,000 3,000 4,000 5,000 6,000 7,000 6,000 9,000 10,000 10,656
d,
N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A. N.A.
PO
4,173.o 4,266.0 4,266.l 4,359.3 4,452.6 4,546.1 4,639.7 4,733.4 4,627.2 4,921.l 5,015.l 5,109.3 5,171.3
T
117 122 122 127 132 137 142 147 152 157 162 167 170.3
I
123.931 123.753 123.751 123.573 123.410 123.245 123.061 122.929 122.766 122.645 122.500 122.362 122.266
AP
0 93.0 93.1 93.2 93.3 93.5 93.6 93.7 93.6 93.9 94.0 94.2 62.0
(1 +/n-l)
23,036 23,059 23,051 23,043 23,062 23,056 23,051 23,046 23,046 23,044 23,066 15,166
u,
+/n-1)
Y,L
0.6963 0.9071 0.9071 0.9177 0.9262 0.9365 0.9486 0.9566 0.9665 0.9762 0.9677 0.9971 1.0033
23,036 23,059 46,i 10 69,153 92,215 115,271 136,322 161,370 164,416 207,460 230,526 245,694
are zero.
Step 3. Enter wellhead data on Line 1 where Hand L Calculate f t from definition of I in Eq. 17. From Eq. 17. I is: ;-2.082(g) F2 + H L (P& 1,000 .
At the wellhead, where H=O and L=O for a vertical wellbore. H=L, then H/L=lim (H/L)=l.OOO. H&L+0 For a deviated wellbore, H is less than L, and for a horizontal pipeline, H = 0, and the term for the head of gas drops out of the term for 1. FZ+~(plTz) /l,OOO
;=(3913)/(577)(0.8776)=7.72747. Note that z was calculated by methods given in Refs. 14 and 15 (see also Chap. 20). 2.082(0.615)( 11.299)* 2.082(ygq,2/dpp>= (2~l41)~(3913) =0.00118. [p/~z-2.082(y,q,2/dpp)]=7.72629. Using Eqs. 16a and 16b:
=0.014067+(1.000)(0.059714) =0.073781. Then Z=(7.72629)/(0.073781)=104.719. If the F2 value determined from F, (taken from Table 33.10) is substituted above, I becomes 104.708, which compares well with 104.719. Step 4. Determine trial Ap (Line 2) for a depth of 1.000 ft by 37.484~7, XL = 37.484(0.615)(1,000) 2( 104.719)
Ml=
Step 5. Complete calculation of first trial 12 (104.346) on Line 2 where the temperature is 122F, and the first trial pressure is 3913.0+ 110.1=4023.1 psia. At these conditions, the compressibility factor, Z, is 0.8g94. Estimate the second trial Ap by: 37.484(0.615)(1,000)
&2=
= 1 1o 3
(p/Tz)*/1,CKIO=(7.72747)*/1,000=0.059714.
104.719+104.346
33-20
PETROLEUM
ENGINEERING
HANDBOOK
Step 6. Complete calculation of second trial 12 (104.343) on Line 3 where the temperature remains at 122F, and the second trial pressure is 39 13 .O+ 110.3 = 4023.3 psia. Under these conditions, the compressibility factor, Z, remains at 0.8894. Estimate the third trial Ap by 37.484(0.615)( 1,000) *~3= 104.719-t 104.343 = 110.3.
Production Rate
Estimation of the steady production rate of a well into a pipeline operating at a relatively constant pressure requires both test data and a general knowledge of the producing characteristics of the well. For example, an estimate is required of the capacity of Well A (Fig. 33.7) to deliver gas into a pipeline operating at a pressure so that pts * -ptf * in thousands equals 20. Starting from shut-in conditions, the delivery rates would be 6,950, 6,000, 5,150, and 4,730x103 cu ft/D. The rate at 72 hours would be 3,340 x lo3 cu ft/D from data given in the text. The steady reduction rate would be about 2,000 to 2,400x10 Pcu ft/D. Although theoretical methods have been published for estimation of stabilized production rates, they would require more data than is available for the well. Well B (Fig. 33.5) would produce about 4,300~ lo3 cu ft/D into a pipeline when
2 PlS -prf* =500
Since the third trial Ap is the same to within 0.04 psi, the pressure at a depth of 1,000 ft was determined by trial and error to be 3,913.0+110.3 = 4,023.3 psia. (Note that the third trial was not entered in Table 33.12.) Step 7. Repeat Steps 4 through 6 to calculate the pressure at a depth of2,OOO ft. Only the final step was given in Table 33.12. Table 33.13 illustrates the calculation of subsurface shut-in pressures in a gas well by Eqs. 15, 16a, and 16b by the same procedure used in Example 1. The only difference is that for the shut-in well the rate of flow, qg , is zero and, as a result, the pressure loss caused by friction is zero. Therefore, the inside diameter of the pipe has no effect on the calculations. Size of Integration Interval The integration interval was 1,000 ft in Tables 33.12 and 33.13 for a moderately high-pressure well and, for the flowing example (Table 33.12), the rate of 11.299 x lo3 cu ft/D gave an effective or average velocity of 14.7 ft/sec near the wellhead. Also, the compressibility factor, z, of the gas at wellhead conditions is in that portion of the z vs. pressure curve where z is very nearly a linear function of pressure. At this low velocity and the nearly linear relationship of z with pressure, an integration interval of 1,000 ft is probably more than enough. Likewise, at low pressures where z is again almost a straight-line function of pressure and at low velocities, the integration interval could be extended to 3,000 ft without undue error. However, even moderate computation facilities eliminate the necessity for expanding the integration interval to more than 1,000 ft.
as long as the well remained in good condition. Actually, the performance of Well B increased during production and the rate of flow would have increased. The performance of Well B did not deteriorate with time. Estimation of the sustained production rate of a particular well against fixed pipeline conditions requires a general knowledge of well performance and a definite knowledge of the performance characteristics of the particular well. The accuracy of such estimations is dependent to a large extent on the amount of proper test data available for study.
OPEN
FLOW
OF GAS WELLS
33-21
qg,
IO3 cu ft/D
qg,
lo3
cu fVD
Fig. 33.9-Effect of tubing installation on performance of Well C; Points 1 through 4 are before tubing installation, Point 5 is after.
Fig. 33.10-Effect of obstruction in tubing on performance of Well D; Points 1 through 6 are before removal, Point 7 is after.
bottomhole chokes, injection of chemicals such as the alcohols or glycols into the flow string, or by the installation of downhole heating equipment. The accumulation of hydrates in the flow strings may be alleviated to some extent by elimination of obstructions in the flow string, use of proper valve sizes at the surface, elimination of sharp bends in surface lines, and proper placement of chokes in surface lines. Remedial action consists in lowering of hydrate-formation temperatures by chemicals or by maintaining the temperature of the flowing gas above the hydrate-formation temperature. Heating of the flow string in a well is usually accomplished by the circulation of hot oil in the casing around the tubing of the well. However, it must be emphasized that hydrate troubles are very easily confused with liquid troubles in low-temperature wells. A careful study should be made of flowing temperatures in a well before recommendations are made for hydrate prevention. Liquids Most performance difficulties in gas wells are caused by the accumulation of liquids in the wellbore. Liquid troubles may be caused by hydrocarbons (condensate and etude oil), salt water, or brines coming into the wellbore from the producing formation, brines from foreign sources through casing leaks, or fresh water. Occasionally, the production of formation water or crude oil may be eliminated by plugback operations. Liquids in wellbores may be removed by tubing strings of proper design, siphon strings (tubing with jet holes or gas-lift valves), and plunger lifts. Periodic flowing of the well at high rates to the pipeline may eliminate liquid troubles. Remedial action for water troubles requires an identification of the source of the water. This is done by water analyses. If it is decided from analyses that the water is native to the formation, there is a choice between plugback work and water removal by various means. Salt water that is foreign to the producing forma-
tion or fresh water in excessive amounts indicates a casing leak that should be repaired. However, moderate amounts of fresh water usually condense in the flow strings of gas wells. Fresh water that occurs naturally should not be confused with fresh water from a foreign source. Cavings Cavings that consist of shale and pieces of the formation are usually most troublesome in openhole completions. The presence of cavings in wells without tubing can be determined easily by comparison of measured depth with drilled depth. Remedial action consists of cleaning out, installation of liners in openhole, and acid washes where the formation is soluble in acid. Unconsolidated sand is troublesome in many Gulf Coast wells. Sand may damage the performance of a well in addition to causing severe damage to equipment. Remedial action consists of cleaning out, installation of treatment for the special liners, or consolidation formation. Deposition of Salts Salts (sodium chloride or other chemical compounds) may be. deposited in the flow strings or wellbores of gas wells. Sodium chloride and water-soluble salts often may be removed by water or light acid washes. Occasionally it is necessary to replace the flow string with clean pipe. Heavy crude oil (not a salt) may be removed From the flow string and to a limited extent from the face of the wellbore by washing with kerosene. Casing and Tubing Leaks Casing leaks usually permit the migration of gas into another formation, but occasionally in low-pressure areas water may come into the wellbore through leaks. The migration of gas into another formation is wasteful. Casing leaks, depending on their size in relation to well
33-22
PETROLEUM
ENGINEERING
HANDBOOK
capacity, cause deterioration in well performance. Positive identification usually can be made with subsurface temperature surveys or special techniques supplied by service companies. Tubing leaks, where there is no packer, tend to defeat the purpose of the tubing. Liquid removal becomes difficult if the hole in the tubing is large. Small leaks in tubing are usually the result of corrosion. In wells where the annular space is packed off, tubing leaks may allow casing pressures to build up to dangerous levels. Foreign Objects Foreign objects such as swab rubbers, stud bolts, or pieces of metal may remain in the flow string of a well after completion. Such objects should be removed from the flow string because they can seriously affect the delivery capacity of a well. The removal of foreign objects ftom the upstream side of chokes is common.
ratios were normal. Thus it was concluded that liquids were not the source of trouble. After Points 4, 5, and 6 were taken, it was decided to blow the well. Shortly after the well was opened, a swab rubber and several pieces of metal were blown from the well. Afterward, the performance of Well D returned to normal, as indicated by the positions of Point 7 and later performance points that are not numbered on Fig. 33.10. Space does not permit a complete description of reconditioning procedures. However, it is hoped that this brief outline does illustrate the importance of adequate performance tests in the maintenance of well productivity and planning reconditioning procedures.
Nomenclature
C = performance coefficient di = internal diameter, in. f= coefficient of friction (friction factor) F= term in Eq. 16a F, = specific-gravity adjustment factor F = non-Darcy flow factor ;; = basic orifice factor for critical-flow prover, lo3 cu ft/D at 14.65 psia, 60F, specific gravity = 1.000 adjustment factor Fpu = supercompressibility F, = factor defined by Eq. 16a FT = flowing-temperature adjustment factor h, = height (manometer reading), in. mercury h, = height (manometer reading), in. water H= vertical depth in a well, ft (in untubed wells H is the vertical depth to the midpoint of the productive formation; in tubed wells H is the vertical depth to the entrance to the tubing) I= terms in Eq. 17 K= absolute roughness characteristic, in. L= length of flow string in well corresponding to H, ft n= exponent of the backpressure equation or slope of the backpressure curve P pressure, psia Pi = impact pressure on a pitot tube, psig PI = impact pressure on a pitot tube, psia Ppc = pressure, pseudocritical, psia PR = average pressure in the reservoir at vertical depth H Ps = static pressure on critical flow prover, psia Ptf = flowing pressure at wellhead measured on a flowing column of gas, psia Pts = shut-in pressure at wellhead, psia Pwf = subsurface (bottomhole) flowing pressure in the wellbore at vertical depth H, psia qg = rate of flow, lo3 cu ft/D or lo6 cu ft/D (14.65 psia and 60F) ri = internal radius of pipe, in. R gL = gas to hydrocarbon liquid ratio, cu ft/bbl T = temperature, OF+460 Tf = temperature of flowing gas, OF+460
OPEN
FLOW
OF GAS WELLS
33-23
Tpc = temperature,
pseudocritical,
OF+460
TR = reservoir temperature
I , = vapor volume equivalent of 1 bbl (60F) of hydrocarbon liquid, cu ft/bbl z = compressibility factor for gas A = difference between two values 78 = specific gravity of separator gas or gas being measured, air = 1.000 yg = specific gravity of the flowing fluid, air = 1.000 -ye = specific gravity of hydrocarbon liquid referred to water
References
1. Reid, W. : Open Flow Measurement of Gas Well, Western Gas (Nov. 1929) 32. 2. Rawlins, E.L. and Schellhardt, M.A.: Back-pressure Data on Natural Gas Wells and Their Application to Production Practices, USBM Monograph (1935) Washington, DC. 3. Binckley, C.W.: Methods of Approximating Open Flow of Gas, Proc., Southwestern Gas Measurement Short Course (1954) 304 4. Cullender, M.H.: The Isochronal Performance Method of Determining the Flow Characteristics of Gas Wells, J. Pet. Tech. (Sept. 1955) 13742; Trans., AIME, 204. 5. Odeh, A.S.: Pseudosteady-State Flow Equation and Productivity Index for a Well With Noncircular Drainage Area, J. Per. Tech. (Nov. 1978) 1630-32. 6. Ramey, H.J. Jr.: Non-Darcy Flow and Wellbore Storage Effects in Pressure Build-Up and Drawdown of Gas Wells, J. Per. Tech. (Feb. 1965) 223-33; Trans., AIME, 234. 7. Odeh, A.S., Moreland, E.E., and Schueler, S.: Characterization of a Gas Well From One Flow-Test Sequence, J. Per. Tech. (Dec. 1975) 1500-04; Trans., AIME, 259. 8. Smith, R.V.: Unsteady-State Gas Flow into Gas Wells, J. Pet. Tech. (Nov. 1961) 1151-59; Trans., AIME, 222. 9. Orifice Metering of Natural Gas, Gas Measurement Commitfee Report 3, American Gas Ass., New York City (April 1955). 10. Manual of Back Pressure Testing of Gas Wells, State of Kansas, Topeka (1959). Il. Manual of Back Pressure Testing of Gas Wells, Interstate Oil Compact Commission, Oklahoma City (1962). 12. Smith, R.V.: Practical Natural Gas Engineering, Pennwell Publishing Co., Tulsa (1983). 13. Cullender, M.H. and Smith, R.V.: Practical Solution of GasFlow Equations for Wells and Pipelines with Large Temperature Gmdients, J. Pet. Tech. (Dec. 1956) 281-87; Trans., AIME, 207. 14. Hall, K.R. and Yatirough, L.: A New Equation of State for ZFactor Calculations, Oiland GasJ. (June 18, 1973) 71, No. 2.5, 82-85. 15. Yarborough, L. and Hall, K.R.: How to Solve Equation of State for Z-Factors, Oil and Gus J. (Feb. 18, 1974) 72, 86-88.
YL=
131.5+yAPl
141.5
. .
. .
. . . .
(14b)
vLy65.7+0.89
-fApI+0.~7(YApI)2.
(14C)
. (15)
F==
. . . . (16a)
Al-
7.4 Ti l - =4 log-. K f
...
..
(16b)