Professional Documents
Culture Documents
L I F T
tional area. Upstream from the nozzle, the gas has high pressure and low velocity; downstream from the nozzle, the gas has low pressure and high velocity. The throat is where the interchange of momentum occurs between the power (gas) and well fluids. The diffuser changes the energy from kinetic back to potential with the gradual increase in area to move the mixture (power and well fluids) to the surface. The gas-to-liquid momentum transfer is less efficient than the liquid-to-liquid or liquid-to-gas transfer because of the high velocity required of the gas to accelerate the well fluid. Therefore, the gas/liquid ratio of mass flow rate must be higher to provide a good transfer of momentum. The volumetric gas-flow rate (power fluid) should be two or more orders of magnitude of the volumetric flow rate of well fluid to obtain an efficient momentum transfer.
INSTALL ATION
Diffuser
Throat
Nozzle
Gas entrance
The injection-gas volume is regulated through the gas-lift manifold before the gas enters the annulus of the well, where it travels to the circulating sliding sleeve or operating mandrel. From the sleeve, the high-pressure gas passes through the nozzle and mixes with the well fluid along the throat. Then, the mixed fluid passes into an expansion area or diffuser directing the fluid to the surface. As shown in Fig. 1, the GLJP has three components. The nozzle changes the energy from potential to kinetic because of a sudden increase in cross-sec-
This article is a synopsis of paper SPE 48840, Gas-Lift-Jet Applications Offshore Lake Maracaibo, by Juan Faustinelli, Wilfredo Briceo, and Aaron Padron, PDVSA E&P originally , presented at the 1998 SPE International Oil and Gas Conference and Exhibition in China, Beijing, 26 November.
30
The GLJP can be installed either in the circulating sliding sleeve or in the bottom mandrel. To install a GLJP in a sliding sleeve, the operating valve is retrieved and replaced with a dummy valve. The sliding sleeve is opened to circulate gas to clean the annular space. The GLJP with a lock man, drel incorporated, then is installed. To install a GLJP in an operating mandrel, the operating valve is replaced with a large-port orifice or circulation valve. The gas-injection volume is regulated by the nozzle diameter of the GLJP A collar stop is . installed just below the operating mandrel. The lower packoff, plus a spacing tube and a receptacle, are installed. A landing shoe, the GLJP, and the upper packoff are installed. The last trip into the well is to install a tubing stop. Installing the GLJP in the circulating sliding sleeve instead of the operating mandrel is recommended because it requires fewer wireline trips and operational problems have been experienced because of gas filtration in the upper or lower packoff.
ADVANTAGES
It creates a suction force at the exit of the nozzle, in the throat, decreasing the wells flowing pressure. In wells where the reservoir pressure is low and does not require unloading valves, all the energy (potential), high-pressure gas, is converted to kinetic. It has no moving parts. It can be installed in a well having a high-gas/oil-ratio zone above the main oilproduction zone, providing the capability to produce without gas injection (autolift).
DISADVANTAGES
The disadvantages of the GLJP are associated with the need to run tools below the pump. Static or dynamic tests for pressure and temperature below the GLJP cannot be run. The holdup depth cannot be verified. Recompletion to a new zone is difficult.
MAY 1999
A R T I F I C I A L
L I F T
The GLJP must be removed to perform these tasks. The momentum interchange of the GLJP is the least efficient when compared with liquid-to-liquid or liquid-togas cases.
SELECTION CRITERIA
The general selection criteria change as new results and more experiments are analyzed. The following criteria were used. Water cut less than 30%. Low sand content. Production rate less than 800 BOPD because only two throat diameters (3/4 and 7/8 in.) are available. Productivity index greater than 0.5 B/D-psi. Gas/liquid ratio less than 1,500 scf/bbl.
FIELD INSTALL ATIONS
GLJPs were installed in wells operating under continuous gas-lift offshore Lake Maracaibo. The experience gained from those installations improved the selection criteria for future candidates. The results obtained indicate that the GLJP was successful in three wells and failed in four wells. Also, one installation of the GLJP was for a separate zone with a high gas/oil
ratio near the production zone that enabled production without gas injection. The tests showed a production increase in three cases (Wells 2, 4, and 7). However, the GLJP was not successful in four cases (Wells 1, 5, 6, and 8). Comparing the results between Wells 1 and 2 shows that the formations high gas/oil ratio makes Well 1 less efficient. Because the throat diameter is a function of the total volume (liquid and gas), in the high-gas/oil-ratio case, the larger throat diameter could have been used instead of the smaller diameter. For Well 5, the GLJP did not obtain an oil-production increase because of its low productivity index and a possible downhole restriction (small throat diameter). Well 7 was successful because of its high productivity index. However, the system would be more efficient if the larger throat were used. Well 6 suffered from the formations high gas/oil ratio plus a small throat diameter. Well 8 seems to have an inefficient momentum transfer. It needs additional gas mass or momentum to accelerate the fluid coming from the reservoir. It is important to mention that the differential between the injection and tubing pressures should be kept as high as possible
A significant production increase with the same gas-injection volume was obtained in two wells during the field tests. Studies are needed on such combined parameters as nozzle/throat ratio, differential between injection and tubing pressures, and the total-fluid-volume/throat-diameter ratios that influence momentum transfer and the applicability range. The productivity index and the formation-gas/liquid ratio are the most important parameters to select well candidates. The GLJP performed successfully when installed in a well containing a sand with a high gas/oil ratio above the main liquid-producing zone, providing the capability to produce without surface gas injection.
Please read the full-length paper for additional detail, illustrations, and references. The paper from which the synopsis has been taken has not been peer reviewed.
32
MAY 1999
A R T I F I C I A L
L I F T
Wellhead
Two phase
One solution is application of artificial-lift methods to supply additional energy to the fluids and generate adequate drawdown at the formation while maintaining a high wellhead pressure to transport the fluids to the host platform. A common method for subsea production is gas lift (GL), where gas is injected into the tubing string to reduce the hydraulic head without increasing the frictional losses so that the net result is an increase in the wellhead pressure for a fixed bottomhole pressure. This method benefits vertical flow, but not horizontal flow. For long-distance horizontal multiphase flow, the net result of the increase in the gas/liquid ratio (GLR) is detrimental because the frictional loss increases and the hydraulic head does not decrease.
Well
Reservoir
This article is a synopsis of paper SPE 48934, Modeling a Jet Pump With an Electrical Submersible Pump for Production of Gassy Petroleum Wells, by P .M. Carvalho,* SPE, A.L. Podio, SPE, and K. Sepehrnoori, SPE, U. of Texas at Austin, (*also with Petrobrs Intl.), prepared for the 1998 SPE Annual Technical Conference and Exhibition, New Orleans, 2730 September.
34
Pumping methods include the electrical submersible pump (ESP), progressing-cavity pump (PCP), and jet pump (JP). In the ESP method, a multistage centrifugal pump, driven by a coupled electric motor, is installed at the bottom of the tubing string. The JP with no moving parts, con, sists of a body with a nozzle, throat, and diffuser, all set in a nipple in the tubing string. In traditional systems, clean power fluid is pumped from the surface to the pump through the tubing. This power fluid passes through the nozzle, creating a low-pressure region connected to the pump intake so that the well fluid is suctioned into the throat region of the JP The mixed fluid (i.e., . power and produced fluids) exits the pump through the diffuser into the tubing/casing annulus with sufficient pressure to overcome the hydraulic head and frictional losses and move to the surface. Most systems must operate with a minimum of free gas present at the pump intake. Without the annulus flow path, the annular space cannot be used as a downhole separator and vent the gas at the casing head.
MOTIVATION
Application of an ESP to a well that has a large free-gas volume at the pump intake requires installation of a gas separator. Use of the gas separator requires installation of an extra flowline to vent the separated gas to the host platform. To avoid installation of the vent line and still use the ESP a method , was developed to reinject the separated gas into the fluid after the fluid has been pumped through the ESP .
METHOD
The basic idea is to use a JP at the ESP discharge to allow the gas (separated by the rotary gas separator and segregated to the annulus) to be compressed and injected back into to the liquid stream being pumped by the ESP to the surface. Fig. 1 shows a schematic of a potential deepwater system.
MODEL
The first subsystem is an ESP a multistage , centrifugal pump installed below the liquid level. It is driven by an electric motor and has a gas separator at the intake. To model
MAY 1999
A R T I F I C I A L
L I F T
Mixture
Jet Pump
Water P. Sensor
rally; therefore, energy must be supplied to the fluids to transport them to the surface with the desired flow rate. Because GL is not recommended for long-distance horizontal flowlines, the ESP was considered. Without installation of the ESJP the ESP , alone is not applicable because of the high free-gas content at the pump intake.
PROTOTYPE TEST
ESP
the performance of the ESP an approximate , correlation is used to overcome the difficulty in solving the complicated dynamic model and to correlate the pressure increase at each stage, the pump-inlet pressure and void fractions, and the liquid flow rate. Multiphase Flow in Pipes. To calculate the pressure gradient inside the tubing, a multiphase-flow simulator developed by Petrobrs was used. First a vertical multiphase-flow correlation is chosen; then, the required tubing intake pressure for each flow rate is calculated from the selected correlation. JP Model. Fig. 2 is a diagram of the liquid/gas jet pump (LGJP). The inlet region is inside the pipe upstream from the nozzle. The gas stream enters the device at the pump-suction region. Under normal operating conditions, the liquid jet enters the throat region surrounded by a gas annulus. The two streams mix in the throat region, and a homogeneous mixture of gas bubbles in liquid then is decelerated in the diffuser region. The mixing process (in which the disintegrating liquid jet entrains, accelerates, and compresses the gas) occurs at a point in the throat region that is controlled by the discharge pressure for a given nozzle rate and suction pressure. High pressure at the diffuser forces early mixing; lower pressure moves the mixing zone downstream. From observations made with a Plexiglas model of the LGJP the liquid jet is sur, rounded by a gas annulus with a distinct boundary between the two phases at the beginning of the throat region. At some point in the throat, the phases start to mix
MAY 1999
and, if the throat is long enough, the stream exiting the pump is a homogeneous bubbly mixture. In the throats mixing process, the transfer of momentum from the liquid primarily serves to compress the gas; this contrasts with the liquid/liquid jet pump, where significant momentum transfer is involved to increase the kinetic energy of the secondary liquid stream. The pressure recovery in the LGJP diffuser is significantly reduced because the liquid performs most of the work in compressing the entrained gas bubbles. Model results show that, as the volumetric-flow ratio at the throat entry increases, the efficiency and mixing-zone length also increase. However, a point of maximum efficiency does exist. After this point, the measured efficiency drops sharply and the comparison between experimental and theoretical results is very poor. It was observed that, whenever the mixing zone is in the throat region, the experimental data compare very well with the theoretical results. The increase in volumetric-flow ratio causes an increase in the mixing length so that, when it is greater than the throat length, the model no longer represents the process. Generally, as volumetric-flow ratio increases, pump efficiency increases and compression decreases.
ESJP APPLIC ATION
A prototype test was conducted in the production laboratory at the U. of Texas at Austin. An ESP from Reda Pump Co. was installed with a JP from Trico Industries. While the ESP equipment was conventional, the JP was a modified liquid/liquid JP . The modification allowed the power fluid to enter from the lower part of the device while the produced fluid was supplied from the tubing/casing annulus. The throat length, as well as nozzle, throat, and diffuser diameters, was unchanged. Fig. 3 shows the test-well configuration. Completion of the 550-ft-deep well was done so that the ESP was installed at a depth of 512 ft and the jet pump was installed 129 ft below the surface. The well was equipped with pressure sensors along the tubing string and in the annulus at different depths to monitor the pressure profile. Test results presented very good repeatability and reproducibility. The observed behavior was somewhat different than expected from the theoretical model. Although the overall behavior was as predicted (i.e., as the volumetric-flow ratio increases, the compression ratio decreases), the experimental results show a sharp decrease for the lower volumetric-flow ratios while the theoretical model predicts a flat behavior. One possible explanation for this anomaly is the use of a liquid/liquid jet pump to perform gas compression. The geometry of the device may not be appropriate, resulting in a poor efficiency.
SUMMARY
The system simulator was used in a case study for an offshore well in the Campos basin. To investigate the feasibility of this design, a simple application to a typical well was calculated. The results show that application of this system to a typical gassy oil well is feasible. This well is not capable of producing natu-
This study demonstrated, theoretically and in the prototype test, the feasibility of the ESJP artificial-lift method. Wells with larger GLRs should be considered for this method. The authors believe that the ESJP is a promising technology for deepwater development. The next step is completion of the ESJP simulator. The final step will be development of a 2D model for the ESJP . Please read the full-length paper for additional detail, illustrations, and references. The paper from which the synopsis has been taken has not been peer reviewed.
35
A R T I F I C I A L
L I F T
A 65/8-in. liner, with a 4-in. liner as a contingency. A standard 95/8-7-in. liner polishedbore receptacle (PBR), used to ensure continuity into the liner through the seal assembly. A 7-in. tieback-seal mandrel stabbed into the PBR. A 95/8-in. hydraulically set production packer. A packer-anchor latch. A full-bore PBR with a chevron-seal assembly and circulating ports. A 65/8-in. full-bore permanent downhole-gauge carrier (if required). A 7-in. tubing string (L80, 13% Cr or plastic coated). A 7-in. WRSCSSSV. A spool tree. This design provides maximum flexibility for the future plans of all wells in Brent. The completion can accommodate the following special situations. Zonal isolation from the top down, with either Weatherford or Liner Flex casing-/tubing-patch technology. Allowance for the insertion of a gas-lift insert string (GLIS) at a later date through the spool tree. Deployment of ESPs during the deepdepressurization phase of Brent. Running of dual concentric completions. Furthermore, the monobore-completion design facilitates the implementation of additional new technology, such as well tractors, as cost-effective alternatives to coiled tubing for well-intervention activities and for drilling. The Brent monoborecompletion design has reduced cost by 30% and installation time by 65% vs. the previously run MSA completion.
GLIS
This article is a synopsis of paper SPE 50585, Cost-Effectiveness and Flexibility Through the Brent Gas-Lift Insert-String Completions, by J.D. Brakel, I. Taggart, and Z. Bin Yahia, Shell U.K. E&P originally presented at , the 1998 SPE European Petroleum Conference, The Hague, The Netherlands, 2022 October.
36
To ensure that a true monobore design would be effective, the only given in the existing completion design was the wireline-retrievable, surface-controlled subsurface safety valve (WRSCSSSV) with a minimum inside diameter of 53/4 in. Other components used in the new completion design included the following.
The selection of a monobore Christmas tree to replace the conventional Y-block tree was not taken lightly. Well-bay engineering studies were carried out to ensure the monobore tree would fit alongside the existing trees. Tests included running a 23/8-in. jointed tubing string with 23/8-in. gas-lift mandrels,
MAY 1999
A R T I F I C I A L
L I F T
tubing-retrievable SSSV (TRSSSV), ASV , travel joint, and tubing hanger through the tree. A specially adapted Y-block tree was used to hang off the string. Full-functionality trials were carried out, including locking the string in place with the existing control lines, spacing out the completion to hang off in the Christmas tree, and retrieval upon completion of the tests. Gas-lift dummy valves were recovered and replaced with valves by use of wireline, and tests were carried out by pumping both water and nitrogen to establish flow rates that could be expected through the equipment from each vendor. All three GLIS systems tested worked to their individual specifications, with the final system selected on the basis of cost (Fig. 1). For offshore deployment, consideration was given to use of either a jointed-tubing string or coiled tubing. Although coiled tubing appeared to give the greatest functionality with regard to running and pulling times, research revealed that large-bore coiled tubing (2 in. and larger) is difficult to handle and connect with the numerous gas-lift mandrels. Integral-connection jointed tubing was selected to provide the strength required, the possibility of wireline access to the reservoir, and to allow the string to be deployed under pressure by use of a hydraulic workover unit (HWU). To allow the GLIS to be hung off in the spool tree, an adapter block was built to fit on top of the tree allowing the GLIS hanger to line up with a gas-injection-valve assembly.
INSTALL ATION
ventional tubing hanger. The conventional tree was removed and the modified (overshot) spool tree installed in a free-standing mode by use of the platform crane. Next, the additional valve block, the GLIS block, which accommodates the gas-lift-line connection and GLIS hanger, was installed. The snubbing blowout-preventer stack and HWU were rigged up subsequently on top of the GLIS block. Following the rig up of the HWU, the installed barriers were retrieved and the protection sleeve installed in the SSSV nipple and GLIS block. Full reservoir access then was reinstated by retrieving a deep-set tubing plug with wireline and pushing another deep-set tubing plug to bottom with the HWU work string. The latter plug could not be retrieved because of a casing patch that restricted the tubing bore.
PERFORMANCE
Owing to a delay in the planned installation of the GLIS, no actual performance data were available at the time of writing this paper. Only theoretical performance is available. Choice of GLIS Sizes. The performance of the GLIS is affected by the gas rate injected down the insert string and by the verticalflow performance up the tubing/casing annulus. Changing the insert-string size will affect both. Therefore, it is important to model the complete system properly. The majority of the wells in Brent are completed with a 7-in. completion tubing. A few older wells contain 51/2-in. tubing. Analyses of well performance for various GLIS sizes were made that concluded that the optimum GLIS size is 23/8 in. The optimum gas rate that can be injected down a 23/8-in. insert string is 5 to 6 MMscf/D. GLIS vs. Conventional Gas Lift. One of the biggest disadvantages of GLIS vs. conventional gas lift is that, as the reservoir pressure drops, the gas-lift operation with a GLIS becomes inefficient. This inefficiency occurs because more gas is needed at deeper depths of injection to lift the liquid out of the well. Because of friction, gas injection down an insert string becomes more restricted as the injection point gets deeper. Likewise, more friction loss will occur in the annulus side as well. In the long term, the GLIS is a better option than conventional gas lift. As the reservoir depressurization progresses, the gas caps in the reservoirs will expand and most of the wells on gas lift will become more gassy and more likely to flow natural-
As part of the Brent redevelopment project, a significant program of well-engineering activities needed to be executed, ranging from further development drilling to wellintervention operations for reservoir-management purposes, to preparing wells for ESP installation to facilitate reservoir depressurization. These activities must be performed within tight operating budgets, whereas timing of most of these activities is critical as well. Therefore, the drilling sequences for the various Brent platforms depend on the use of activity-matched intervention equipment, such as coiled tubing and the HWU. This equipment also allows activities to be executed concurrently with main-rig operations. The installation of the first GLIS in Brent, Well BD31, required the conventional Christmas tree to be changed to a spool tree. Because the completion tubing was still in good condition, it could be left in place. Hence, a modified spool tree was required that could be fitted over the con MAY 1999
ly. Therefore, the gas-lift systems are likely to become redundant. For the GLIS, it would be a simple matter to retrieve the insert string to provide a larger flow area for the gassy liquid.
Please read the full-length paper for additional detail, illustrations, and references. The paper from which the synopsis has been taken has not been peer reviewed.
37
A R T I F I C I A L
L I F T
In chamber lift, the injection gas initially contacts the top of the liquid slug in the dip-tube/chamber annulus and displaces this slug into the tubing above the chamber before injection gas enters the lower end of the dip tube. Chamber-lift operation prevents water accumulation in the production conduit because the water is U-tubed from the chamber first, followed by the oil then by the injection gas. The chamber-lift principle has been applied in some wells to lower the point of gas injection inside a given tubing string. A dip tube is installed and packed off immediately above the bottom GLV. The depth of gas injection is lowered to the bottom of the dip tube. The principle has been used in low-permeability wells that produce sand to prevent sand accumulation in the wellbore.
GA S-INJECTION GA S CYCLE
When the chamber-operating GLV opens, the standing valve (SV) closes. The liquid in the dip-tube/chamber annulus is Utubed into the dip tube and tubing above the chamber. Most of the slug is displaced to the surface by the injection gas. However, because of injection-gas breakthrough, some liquid fallback occurs during displacement.
This article is a synopsis of paper SPE 52120, Re-examine Insert-Chamber Lift for High-Rate, Low-BHP Gassy , Wells, by H.W. Winkler, SPE, Texas Tech U., originally presented at the 1999 SPE Mid-Continent Operations Symposium, Oklahoma City, Oklahoma, 2831 March.
38
While the SV is closed and the liquid slug moves to the surface, the reservoir pressure increases and formation production enters the chamber/casing annulus. Gas separation occurs in this annulus with the gas accumulating below the packer. This gas is compressed by the new production entering the wellbore and restricts the volume of produced-liquid accumulation. After a liquid slug reaches the surface, the injection gas in the tubing exhausts into the flowline and the pressure in the chamber decreases. The SV opens when the chamber pressure becomes less than the reservoir pressure beneath the SV. The liquid in the chamber/casing annulus flows into the chamber, followed by the trapped gas until the pressures in the chamber/casing annulus and chamber equalize. The chamber/casing annulus below the packer now is filled with gas. The gas-injection-cycle frequency depends on the well deliverability. When maximum cycle frequency is required, the gas-injection cycle begins when the wellhead tubing pressure approaches the production-header (separator) pressure. Timecycle control of the injection gas is required to ensure maximum cycle frequency. Soon after beginning the surface gas-injection cycle, the chamber-operating GLV opens and the SV closes. With high gas-injectioncycle frequency, most of the formation production enters the wellbore while the SV is closed and the liquid slug is moving to the surface. The free gas in the chamber/casing annulus beneath the packer must be vented into the tubing above the chamber to prevent the gas from being trapped and compressed. Venting the trapped gas allows the chamber/casing annulus to fill with liquid so that when the SV opens after a slug surfaces, liquid, rather than free gas, enters the chamber.
CHAMBER SIZE
The published guideline for selecting a chamber size is to select the largest-diameter pipe that can be run safely in the casing or openhole section of a well. The recommendation is valid for low-production-rate
MAY 1999
A R T I F I C I A L
L I F T
wells. However, the larger-diameter chamber may be a poor choice for a high-production-rate, low-FBHP gassy well. The , capacity of the chamber/casing annulus, between the packer and the SV should , equal or exceed the chamber capacity in high-production-rate wells, provided the free gas below the packer is vented into the tubing above the chamber.
PREDICTED OPERATIONS
Analyses of three insert-chamber installations in a low-FBHP high-production-rate , well with a formation-gas/-oil ratio of 400 scf/STB illustrates the problem of free gas with this method of lift. The chamberlength calculations are based on an injection-gas pressure equal to 70% of the initial opening pressure of the chamber-operating GLV and on the chamber-/tubing-capacity ratio of the tubing above the chamber. Without Venting Free Gas. The first scenario assumes that no check valve exists above the chamber for venting the free gas from the chamber/casing annulus below the packer. Fig. 2 illustrates a typical capacity relationship between the chamber/casing annulus and chamber when a largediameter insert chamber is used. The volume of formation gas trapped in the chamber/casing annulus below the packer at a minimum FBHP of 140 psi is 191 scf. The total volume in the chamber/casing annulus to accommodate formation production at a maximum FBHP of 252 psi is only 8.0 ft3 (<0.27 STB of oil and 98 scf of formation gas) because of the trapped free gas. When the SV opens, <0.27 STB of oil enters the chamber followed by trapped free gas from the chamber/casing annulus.
MAY 1999
Venting Free Gas. The second scenario uses the previous chamber design, except that free gas in the chamber/casing annulus below the packer is vented. The insertchamber installation shown in Fig. 3a has an annulus vent valve between the bypass packer and the top of the chamber. It is a large conventional GL check valve run upside-down for gravity closure. Trapped free gas beneath the packer can flow from the chamber/casing annulus into the tubing above the chamber through the vent valve. The chamber/casing annulus fills with formation liquid rather than compressing trapped free gas between gasinjection cycles. The capacity of the chamber/casing annulus between the packer and the SV in Fig. 3a is 3.5 STB. By venting the chamber/casing annulus, 3.5 STB of oil, rather than <0.27 STB, enters the chamber when the SV opens. The total capacity of the chamber is 7.2 STB. The chamber capacity is more than double that of the chamber/casing annulus above the SV depth. Smaller-Diameter Chamber with Vent Valve. The third design, shown in Fig. 3b, uses a smaller-diameter pipe for the chamber. The dip tube is the same diameter as in the previous scenario but much longer. The capacity of the chamber/casing annulus between the packer and the SV depth is 10.2 STB, and the capacity of the insert chamber is 7.1 STB. Because the capacity of the chamber/casing annulus exceeds the chamber capacity, the chamber fills with liquid production immediately after the SV opens, if the trapped injection gas from the previous gas-injection cycle enters the tubing above the chamber through the cham-
ber bleed valve. The success of chamber-lift operation with a high gas-injection-cycle frequency depends on the gas-throughput capacity of the chamber bleed valve.
CHAMBER BLEED VALVE
A liquid seal at the lower end of the dip tube occurs when a liquid slug surfaces, and the injection-gas velocity in the tubing begins to decrease. The injection gas from the previous U-tubing cycle in the chamber is trapped in the dip-tube/chamber annulus and must be vented before the chamber can fill with liquid. Without venting the trapped injection gas, most of the production entering the chamber will enter only the dip tube. As a result, the major benefit of an accumulation chamber is nullified. For maximum gas-injection-cycle frequency, rapid venting of the trapped injection gas from the dip-tube/chamber annulus into the dip tube is required. The chamber bleed valve allows the diptube/chamber annulus to fill with liquid by venting injection gas from the previous gas-injection cycle. A high-gas-throughput-capacity chamber bleed valve is necessary for efficient chamber-lift operation in a high-cycle-frequency chamber installation. The chamber-bleed-valve performance is less critical to efficient chamber lift in a low-cycle-frequency, low-capacitychamber installation.
Please read the full-length paper for additional detail, illustrations, and references. The paper from which the synopsis has been taken has not been peer reviewed.
39
A R T I F I C I A L
L I F T
Four producers, including a water-supply well, were drilled from this central pad site. Hole trajectories for the shallow reservoir depth required a directional program for maximum angles possible in the surface and main holes. For Well 8-24-17-15 W4M, after reaching an inclination angle of 72 at 602 m MD and setting and cementing surface casing, a tangent angle of 70 to 72 was maintained into the reservoir, where casing was set at 2873.2 m MD. Dogleg severities were maintained at less than 2.83/30 m through the entire section. Well Completions. All production wells were completed with tubing-conveyed perforating systems. Initially, sufficient reservoir pressure allowed the wells to flow, but shortly after implementation of water injection, water cuts increased, necessitating artificial lift. Three artificial-lift methods were investigated for suitability in these prolific slant wells: gas lift, ESPs, and PCPs. Gas lift for Pad-5 wells was not an option because supply gas was unavailable. Lift selection and design were based on the reservoir inflow-performance relationship (IPR), impact of the downhole mechanical configuration relative to landing depth, and maximum achievable drawdowns. ESP and PCP Installations. Following an extensive post-perforation debris-cleanout program, several wells were completed with ESP systems. The ESP assembly installed in Well 8-24 included a pressure-sensing instrument (PSI), motor, three-chamber seal, rotary gas intake, a 104-stage centrifugal pump, and a tubing drain. A retrievable check valve was not installed because of the potential for scale buildup and the inability to retrieve by wireline at such high hole angles. All ESP components were Monel coated and fully centralized to prevent dragging and damage during installation. In other wells, electrically powered topdrive PCP systems were deployed with conventional rod strings. Insufficient PCP pres-
sure and excessive rod/tubing contact experienced when surface speeds exceeded 300 rev/min constrained production. Consequently, the PCP assemblies were set at depths shallower than 1200 m to minimize well interventions. Economics of infrequent workovers and low operating costs had to be weighed against the reduced oil-production rate. Failure History. The main factors leading to system inefficiencies and pump impairment were scale deposition, gas locking, wellbore emulsions, and mechanical pump failure attributed to formation debris.
SELECTING A PCP
The PCP was chosen because of its ability to operate successfully in the harsh conditions. Normally, scale does not deposit inside the pump. The operation of the pump does not increase the emulsion in the fluid and can tolerate free-gas slugs for short periods of time while providing lift to the fluid. Also, PCPs can tolerate formation debris. The abrasives in the fluid will deteriorate PCP life slower than submersible centrifugal pumps. The primary problem of the PCP system is with the rotating sucker-rod string causing excessive wear on the production tubing. Because these PCPs must be set high in the production string, productivity suffered. To achieve maximum wellbore productivity at Well 8-24, a decision was made to use a bottom-motor-drive PCP system.
DRIVE TYPE
Use of a four-pole induction motor was compared with an electronic motor. Although requiring a speed-reducing gearbox not needed with the electronic motor, the four-pole induction motor and drive system was chosen because the operator has considerable experience with these systems. Also, the system was available for immediate installation.
DEPLOYMENT
The high deviation (72) in the pump landing section of the well is a problem when deploying a wireline-retrievable system.
MAY 1999
A R T I F I C I A L
L I F T
Wireline tools are designed to latch and unlatch the pump by use of gravity. Wellbores with designated landing areas greater than 49 from vertical will experience poor tool operation. Coiled-tubing deployment required additional developmental work at the time of the workover decision. A tubing-deployed system represented the least-risk option. This method uses the production tubing to suspend the downhole pump/drive system inside the casing. The fluid is produced up the tubing, and any free gas separated ahead of the pump intake is allowed to travel up the annulus. One disadvantage of this system is that the power cable must be banded to the outside of the tubing string. This configuration is susceptible to cable damage during installation in a deviated wellbore. Another disadvantage is that the entire tubing string must be pulled to surface when any component in the pump/drive system fails.
GA S-AVOIDER INTAKE
Following extensive PCP modeling to determine the desired pump size, loading, and horsepower requirements, the ESP assembly in Well 8-24 was replaced with a bottom-motor-drive electrical-submersible PCP (ESPCP) system. System Configuration. The completion configuration for the first well was kept simple. The traditional stainless-steel capillary tube for injecting de-emulsifier treatments was eliminated. To prevent the intake of large volumes of gas, a gas-avoider intake was fabricated for the high-anglewellbore installation. Special modifications were implemented to the ESPCP to prevent installation and operational complications. The base of the PCP stator was machined to accommodate 88.9-mm eight-round pin ends. This provision provides extra tolerance for rotor oscillation during excessive speed. Grooved slip-on cable-protector clamps were installed on the stator to house and shield the three-conductor flat cable. A top centralizer and bottom motor guide were essential for cable protection during installation operations. The downhole configuration consisted of a PSI, four-pole motor, modified modular protector, 4:1 gear-reduced flex drive with bottom intake, and an ESPCP The PSI was . installed to measure the suction-intake pressure required to match respective IPR drawdowns. The selected ESPCP contains a high-nitrile, high-speed-geometry rotor/ stator combination. The existing galvanized power cable, the surface transformer, switchboard, and variable-frequency-drive unit were retrofitted to operate the bottommotor-drive ESPCP system.
Intake of high volumes of solution gas leads to decreased pump efficiencies. As shown in Fig. 1, this device, when landed in a highly deviated orientation, will draw well fluid from the bottom of the casing, avoiding free gas being drawn into the suction of the pump. Fig. 2 is a cross-sectional view of the gas avoider. The slot at the bottom of the inner cam rotates freely while the unit is being deployed. The counterweights allow gravity to position the intake slots at the bottom of the casing at the final landing depth. The wellbore liquid is redirected to the low-side intake, below the gas/liquid interface. When the unit is brought to the surface, the intake must be refurbished with new bearings to ensure smooth operation for the next installation.
MAY 1999
ESPCP-System Analysis. PanCanadian had been able to achieve only minimum drawdown (7100 kPa) under ESP operations at Well 8-24 because of gas-locking and formation-debris issues. The ESPCP system was designed to draw the well down to 1000 kPa. A composite IPR (Vogel correlation) was established that was used to predict the oil, water, and total-liquid rates at the target inflow conditions. The pump was started on 24 April 1998 at 30-Hz operating frequency (221 rev/min). To monitor system performance effectively, speed increases in increments of 2.5 to 5.0 Hz were maintained every 2 to 3 weeks following pump-rate stabilization and evaluation of downhole torque parameters. Five speed increases were implemented from 30 to 47.5 Hz (221- to 345rev/min speed range), resulting in approximately 105 m3/d incremental-oil production. To prevent pumped-off conditions while maximizing reservoir-pressure response, a final speed of 47.5 Hz (345 rev/min) was maintained, capturing 92% of the wells potential. Drive-voltage-output and primary-sidecurrent data were captured. The measured data were converted to downhole-motor horsepower and torque-load values necessary to monitor the performance of the ESPCP motor/drive assembly. Efficiencies ranged from 48% at startup to stabilized values of 68 to 70% at higher speeds. Maximum operating torques were well below the operating limit. Please read the full-length paper for additional detail, illustrations, and references. The paper from which the synopsis has been taken has not been peer reviewed.
41
A R T I F I C I A L
L I F T
For any artificial-lift system, power demand can be calculated by applying hydraulichorsepower calculations that take into consideration differential pressure through the system, total fluid rate, and efficiency according to the type of lifting system. Basically, the artificial-lift system should provide the fluid with enough energy to reach the gathering facilities. GL is used to reduce the pressure profile in the tubing to allow fluids from the reservoir to reach the production facilities. Reducing the weight of the fluid column is the principle used to combine the artificial-
lift systems. By injecting gas into the fluid column, pressure-discharge requirements for an RP or ESP are reduced. Less differential pressure through the pump means less energy needs to be supplied by the system to the fluids to reach the production facilities. The benefits are smaller surface and subsurface equipment, less power consumption, and reduced capital expenditure. All the principles of GL design related to pressure injection, depth of injection, available gas volume, and gas/liquid ratio are applied to the proposed combined artificiallift system. Furthermore, as in the conventional GL system, optimal points for gas injection exist that are related to the phenomena of vertical and horizontal multiphase flow in pipes. To evaluate the net effect of the combined system, the concept of equivalent depth for the pump can be used. A new functional depth, instead of the operational (actual) depth, should be used to evaluate the performance of the system. This value is the result of comparing the pressure profile of a conventional artificial-lift system with that of the combined system. In Fig. 1, Point A is the actual discharge point of a conventional ESP or RP Point B .
This article is a synopsis of paper SPE 53966, Production Optimization by Combined Artificial-Lift Systems and Its Application in Two Colombian Fields, by Hubert Borja, SPE, and Ricardo Castano, Hocol S.A. (a Nimir Petroleum Co.), originally presented at the 1999 SPE Latin American and Caribbean Petroleum Engineering Conference, Caracas, 2123 April.
42
Fig. 1Pump equivalent depth by lightening of the column with gas injection.
MAY 1999
A R T I F I C I A L
L I F T
is the discharge point for a lightened fluid column resulting from gas injection. The Point-B pressure value extended to Point C on the gradient curve of a conventional system represents an equivalent point in pressure, but at a different depth, for the subsurface pump. This point is the new pressure required to design the pump to lift the corresponding fluid rate. The actual depth of the pump has not changed. Point D is the equivalent pump-setting depth of the combined system with respect to the conventional system. The pressure difference between Points A and B represents the power reduction for an artificial-lift system in terms of dynamic head. The equivalent-depth concept is particularly important to design and evaluate the benefits of the combined artificial-lift system. Most importantly, the system can be adjusted to the continuously changing conditions of a reservoir (i.e., during waterflood or pressure-maintenance programs) that will impose a continuous change in the pressure profile inside the tubing. To design a combined artificial-lift system for specific well conditions, the conventional artificial-lift system is designed first, because it will perform at Point A in Fig. 1. Then, with the same well inflow performance, the artificial-lift system is designed for a shallower depth. The shallower depth is inferred from the reduced pressure requirements obtained by injecting gas at specific points in the tubing. All the primary artificial-lift requirements can be calculated for each equivalent depth between Points A and D in Fig. 1. The equivalent depth is determined with conventional GL-design procedures, taking into account the injection pressure, volume of gas, and depth of the injection point. Fig. 2 shows, for a specific case, how increasing the gas-injection rate, while keeping the other parameters constant, reduces the equivalent depth. Fig. 3 shows, for the same case, how the equipment requirements (power consumption, size)
MAY 1999
Fig. 3Effect on the power and equipment requirements resulting from changes in the equivalent depth.
are reduced when designed for shallower equivalent depths. Large reductions are seen in the pumping-unit and electricmotor sizes and the sucker-rod grade for the RP system; and pump and motor sizes along with cable and variable-speed-driver requirements for an ESP system.
EVALUATION AND APPLIC ATION
One of the main restrictions was the need to continue separating the gas and liquid downhole, to avoid gas-lock problems, with simultaneous gas injection for GL operations. Two alternatives were evaluated. The first was to inject the gas into the tubing with a parallel gas conduit connected to the GL mandrels. The second considers use of a gas-handler device to pump the gassy liquid with the ESP The gas-conduit . diameter is restricted by the casing size and, because of the pressure losses inside the gas conduit, the gas-injection point is considerably shallower than when the gashandler device is used (between 2,000 and 5,000 ft). Currently, four Balcon field wells are producing with the ESPGL system. Installation of the combined system resulted in a combined production increase of 3,300 BOPD. In general terms, injecting larger gas rates or having deeper injection points reduce the equivalent depth. As in a conventional GL system, technical and economic optimum points exist to operate the system. The key issue for the changes and benefits, as a result of injecting gas to lighten the fluid column, is the reduction in the pumpdischarge pressure. Pressure reductions were as great as 1,885 psi, an average of 44% of a conventional ESP system. All the
other results and benefits are proportional to this pressure reduction. In two wells, the functional depth reduction (operating minus equivalent depth) was more than 5,000 ft (more than 50% of the actual depth). The total power-consumption saving is approximately 400 kW. The required pump and motor sizes were reduced between 33 and 67%. The average total dynamic head of the combined system is 57% lower than in a conventional ESP system. Another advantage is the ability to reduce production losses by maintaining a reduced level of production, by continuing the gas injection, when the ESP is shut in. At least 30% of the combined-system production level can be obtained with the GL system alone. The operating envelope is wider than with single systems, allowing adjustments to changing well conditions. Problems, such as tubular-size restriction and operating depth, were solved with the installation of the combined system. Internal tubing corrosion decreased because of the reduced partial pressure of corrosive gases as a result of decreasing the pump-discharge pressure For the Tello field, the implementation of the RP-GL combination is being evaluated. In addition to the peak-torque reduction, a decrease in power requirements and stress on the rod string has occurred.
Please read the full-length paper for additional detail, illustrations, and references. The paper from which the synopsis has been taken has not been peer reviewed.
43
A R T I F I C I A L
L I F T
The surface equipment of a tubingless rodpumping system is the same as a conventional system. The downhole section includes the pump, packer, pump centralizers, guide, hollow-rod string, rod centralizers, and a hollow polished rod. The tubingless rod-pumping system eliminates the tubing used to hang the pump and act as the oil-flow passage in the conventional system. This system sets the pump at a predetermined depth in the casing and uses a hollow-rod string to transmit motion and energy from the surface pumping unit to the downhole pump. The area between the rods and the casing, above the packer, is used as the oil-flow passage. When installing a tubingless rod-pumping system, the packer is lowered to a predetermined depth in the casing with a tubing string. After the packer is set, the tubing string is pulled to the surface. Then, the pump is lowered into the packer inner bore by use of the hollow-rod string. The pump is set, and a seal is formed between the pump and packer by means of the setting and sealing mechanisms on the lower outer This article is a synopsis of paper SPE 54304, Rod-Pumping Technique Without Tubing, by Xu Xiufen, Jiang Minzheng, and Zhu Jun, Daqing Petroleum Inst., Peoples Republic of China, originally presented at the 1999 SPE Asia Pacific Oil and Gas Conference and Exhibition, Jakarta, 2022 April.
44
surface of the pump barrel and on the packer. Two rigid centralizers are installed on the outside of the pump barrel to ensure its alignment with the inner bore of the packer when lowering the pump. The discharge port of the pump opens to the casing. No changes are required to the wellhead assembly. Also, the pump can be pulled with the hollow-rod string for inspections. When the plunger of the pump travels downward, oil passes the traveling valve into the upper chamber of the pump. When the plunger travels upward, reservoir oil enters the lower chamber of the pump. At the same time, oil in the casing above the packer is lifted to the surface. When the well needs cleaning, well-flushing fluid is pumped through the hollow-rod string. The check valve on the lower end of the rod string opens and the well-flushing fluid returns to the surface through the rodstring/casing annulus.
PREVENTING WEAR
through the tubing string. The back-flow resistance in the tubing string and valves can cause some of the flushing fluid to seep into and contaminate the oil reservoir. In a tubingless system, flushing fluid is pumped through the hollow-rod string and the check valve at the end of the rod string, then into the rod/casing annulus, and back to the surface, forming a closed loop. The flushing fluid does not contact or contaminate the oil reservoir. Each tubingless well can save the cost of tubing, less the costs of the packer and other special auxiliary tools. The pump and rods can be pulled and lowered simultaneously. Drilling a slimhole well requires less horsepower and lighter drilling equipment, reducing drilling costs by 30 to 50%. Because the tubingless rod-pumping system eliminates the need for a tubing string, it is suitable for slimhole wells.
FIELD TEST
Eccentricity of the hollow-rod string and wellbore, wellbore inclination, and bend deformation of the rod string during the down stroke can cause wear between the rods and the casing or tubing. When rod/tubing wear occurs in conventional systems, the worn tubing is easily replaced with new tubing. If the casing becomes worn in a tubingless system, repair is very difficult. Therefore, rod centralizers are installed along the rod string to prevent wear to the casing. System Features. Because this system changes the flow passage from a conventional system, it simplifies the structure. In a conventional system, one factor that affects pump efficiency is elongation of the rod and tubing strings that decreases the effective stroke length of the pump. In a tubingless system, tubing-string elongation is eliminated, preserving pump efficiency. When the conventionally completed well needs cleaning, the flushing fluid is pumped down the tubing/casing annulus, passes the standing and traveling valves in the pump, then flows back to surface
Conventional systems have been replaced with tubingless systems in five wells in the Daqing oil field. The tubingless system works smoothly, steadily, and effectively. At the time this paper was written, two of the systems had been operating for approximately 1.5 years. Test data from three wells indicate an average pump efficiency of 77.2%. The cumulative increased production of the test wells was 1,825 tons. Because the inner diameter of casing is larger than that of tubing, precipitated wax and filth on the casing has had little effect. Use of a tubingless system can lengthen the time between well-cleaning treatments. With conventional systems, Wells Nan 4J2-342 and Nan 3-1-231 were cleaned every 275 and 100 days, respectively. With the tubingless systems, the two wells had been working normally for 480 and 330 days, respectively. Well Fang 52-60 had been working normally for 4 months with a steady flow rate. Please read the full-length paper for additional detail, illustrations, and references. The paper from which the synopsis has been taken has not been peer reviewed.
MAY 1999
A R T I F I C I A L
L I F T
considered. For all the alternatives, a floating production, storage, and offloading (FPSO) system having sufficient space and capacity to accommodate the required facilities is stationed at the field. Also, it provides salvage value when the field is abandoned. Wells are to be completed with conventional wellheads on a production platform. Continuous Gas Lift. Simulation studies indicated that the optimum gas-injection pressure for the field is 1,280 psi. Calculated peak gas injection for the field is 45.9 MMcf/D, and fuel-gas requirement of the field is estimated at 1.1 MMcf/D. The expected total peak gas requirement for the field is 47 MMcf/D. The following alternatives were evaluated. Alternative 1. This system assumes that the processing facilities are at the field, the gas supply is from the processing platform, and oil is transported by pipeline. Initial oil/gas separation takes place on the FPSO, and no compression is needed. Gas to fuel the generators and for gas lift is transported through a 56-mile-long gas line from the processing platform. All recovered gas is flared. Oil pressure is boosted at the field from 100 to 700 psi for transportion to the processing platform by pipeline. The gas-lift design of all seven wells will change as reservoir and production parameters change throughout the life of the field. A loss of revenue of U.S. $192 million will occur because of the gas flaring. The cost impact of this alternative on the total project cost would be approximately U.S. $270.78 million. Alternative 2. This scenario is a variation of Alternative 1. However, oil is transported from the field to the processing platform by shuttle tanker. Shuttle-tanker costs will form a major part of the operating cost. The cost impact of this alternative would be the same as for Alternative 1. Also, bad weather conditions can hamper tanker transportation, which, in turn, can disrupt the production schedule of this field because both the FPSO and shuttle tanker have limited oil-storage capacities. Alternative 3. This system assumes that the processing and gas-compression facili-
Various lifting methods were assessed. Appropriate initial selection of a lift system is crucial to successfully exploit an isolated, marginal field. Generally, selection depends on several factors, such as geographical location of the field, capital and operating costs, reliability, and flexibility in handling varied production rates. For an isolated, marginal development program, the proper lift system maximizes ultimate recovery and profitability. Sucker-rod-, hydraulic-jet-, and electrical-submersible-pumping (ESP) systems were evaluated along with continuous lift with either natural gas or nitrogen. With respect to their applicability to an isolated, marginal offshore field, only continuous lift with natural gas and ESP qualified for further analysis and economic evaluation.
SCENARIO EVALUATION
ties are on the FPSO at the field and oil and gas transportation is through separate pipelines. Oil pressure must be boosted at the field for transport to the processing platform. Produced gas will be used for injection and fuel gas. Excess gas, will be transported to the platform through a gas pipeline. The cost impact of this alternative would be U.S. $120.8 million, after considering capital and operating costs and the revenue generated from gas sales. Alternative 4. This scenario is a variation of Alternative 3, with oil transported by shuttle tanker from the FPSO to the processing facility. The shuttle-tanker costs, compressor-maintenance expenses, and gas-lift-workover cost form the major part of the operating cost. The cost impact of this alternative would be U.S. $128.72 million, including capital and operating costs as well as revenue generated from gas sales. ESP Major surface equipment for this sys. tem includes generators, control room, step-up transformers, and a supervisorycontrol and data-acquisition (SCADA) system. A downhole pump is installed in each well, and surface facilities are common for all wells. The SCADA system enables monitoring of ESP operations from the processing facility without deploying manpower to the field. Few production-platform visits should be required. Less downtime and more consistent production should occur because the system permits quick reaction to changes in operations or well conditions. Alternative 5. This system assumes that the processing facilities are on the FPSO at the field and oil is transported by shuttle tanker. Also, surface equipment required for the ESP system is on the FPSO. Associated gas from the field is used for generating power required for ESP operations, processing facilities, and other related utilities. The ESP-package cost would be U.S. $12 million. ESP workovers are expected every 2 years for each of the seven wells. This option experiences a loss of revenue because of the gas flaring. Also, bad weather conditions can affect tanker transportation severely. In late field life, produced gas from the field
MAY 1999
Various scenarios of continuous-gas-lift (with natural gas) and ESP systems were This article is a synopsis of paper SPE 54375, Downhole ESP and Surface Multiphase PumpCost-Effective Lift Technology for Isolated and Marginal Offshore-Field Development, by Prasanna V. Mali, SPE, Rajiv Singh, SPE, S.K. De, SPE, and M.K. Bhatta, SPE, Oil & Natural Gas Corp. Ltd., originally presented at the 1999 SPE Asia Pacific Oil and Gas Conference and Exhibition, Jakarta, 2022 April.
46
A R T I F I C I A L
L I F T
might provide insufficient fuel gas and diesel fuel will have to be used. The combined impact of capital and operating costs and gas flaring on the total project cost would be U.S. $119.86 million. Although manpower is required for maintenance of the processing facilities, none is required for supervising ESP operations because of the SCADA system. Also, the required higher wellhead pressure of 300 psi will limit ultimate oil recovery from the field. Alternative 6. A variation of Alternative 5, this scenario transports oil by pipeline. Initial oil/gas separation takes place at the field. Booster pumps are used to transport oil to the processing platform. Annual booster-pump operating costs are estimated at 3% of capital cost. This alternative suffers a revenue loss because of gas flaring. The total cost impact of this alternative is estimated at U.S. $111.94 million. However, pipeline transportation of crude oil ensures uninterrupted oil transportation to the processing platform. Manpower requirements are the same as Alternative 5. Alternative 7. This scenario uses a multiphase pump on the FPSO at the field and transports the oil by pipeline. The multiphase-pumping system is well suited for exploitation of an isolated, marginal field. Unprocessed well fluid (water, oil, and gas) is transported from the wells to an existing processing facility by means of a single rotating machine and a single pipeline, allowing an unmanned operation. This system makes it possible to increase oil production by reducing wellhead pressure, thereby extending economical access to small and medium-sized fields. The multiphase-pumping system can be designed to handle changing reservoir pressures and production parameters over the life of the field. Downhole pumps ensure steady wellfluid production. The surface multiphase pump boosts the well-fluid pressure for transport to the processing facility. Gas from the processing platform serves as an uninterrupted fuel supply for the generators. Major operating costs include ESP workovers and multiphase-pump maintenance. Each well requires workover once every two years. Annual multiphase-pump maintenance cost is estimated at 3% of capital cost. No manpower is required at the field to supervise ESP or multiphase-pumping operations. A smaller FPSO can accommodate the required facilities because the processing and compression facilities at the field will be eliminated. The cost of this alternative would be U.S. $103.2 million, which includes capital and operating costs as well as revenue accrued from gas sales.
MAY 1999
RESULTS
A downhole ESP and a surface multiphase pump, when placed in series, would ensure maximum oil recovery from this isolated pool. The wellhead pressure can be kept low with the help of the multiphase pump. Economic analyses of these alternatives show that the use of a downhole ESP and a surface multiphase pump in series is the best choice. Alternative 7 is the most attrac-
Please read the full-length paper for additional detail, illustrations, and references. The paper from which the synopsis has been taken has not been peer reviewed.
47