You are on page 1of 6

60

IEEE Transactions on Power Systems, Vol. 13, No. 1, February 1998

POWER FLOW CONTROL AND POWER F L O W S T ~ ~ E S S Y S T E ~ SFACTS DEVICES FOR


Douglas J. Gotham gotham@ecn.purdue.edu
Student Member State Utility Forecasting Group (SUFG) Purdue University West Lafayette, IN

G. T Heydt . heydt@enuxsa.eas.asu.edu
Fellow Center for the Advanced Control of Energy and Power Systems (ACEPS) Arizona State University Tempe, AZ

Abstract - In this paper, the modeling of flexible AC transmission system (FACTS) devices for power flow studies and the role of that modeling i the study of FACTS devices for power flow control are n discussed. FACTS devices are solid state converters that have the capability of control of various electrical parameters in transmission circuits. A number of power flow study programs were developed in order to model various types of FACTS devices. Three main generic types of FACTS devices are suggested and the integration of those devices into power flow studies,studies relating to wheeling, and interchange power flow control are illustrated.
Keywords: Flexible AC transmission systems, FACTS, power flow control, wheeling power flow studies.

these categories, but these are selected to indicate how FACTS devices should be analwed in power flow studies.
I

"

UPFC [3-51

2. Types of FACTS devices

FACTS FERC GTO

svc

TCSC UPFC

Nomenclature Flexible AC transmission systems Federal Energy Regulatory Commission Gate turn-off thyristor Static var compensator Thyristor controlled series compensator Unified power flow controller

As listed in Table(l), three main types of FACTS devices are considered here. They are: Type A devices: Figure (1) shows a "Type A" FACTS device inserted in a line i-j. It is assumed that in a Type A FACTS device, the active and reactive line power flows are controlled, and the bus phase angle and voltage magnitude are not specified and are independent power flow study states.
The scheduled line flow metered at bus i is Fi + . The notation Pi + J ' jQi and Pj + jQj shown at buses i and j represent the load power specified at those buses. The Type A FACTS device is modeled as shown in Figure (2)

jG..

1. Introduction The concept of using solid state, power electronic converters for power flow control at the transmission level has been known as FACTS [l]. The idea has had some success in certain areas such as var dispatch and control [1,2]. However, the full use of FACTS for power flow control has had limited application in part due to reliability concerns, and in part due to availability of components. Perhaps the most salient consideration is the cost of these devices. A potential motivation for the accelerated use of FACTS is the deregulation / competitive environment in the contemporary utility business. The potential ability to control the path of the flow of electric power, and the ability to effectivelyjoin electric power networks that are not well interconnected, suggests that FACTS may find new applications as es merge and as the sale of bulk power between distant exchange partners become more common. The potential improvement of the transient response of a system with FACTS devices is a very important consideration (perhaps the most important consideration in many applications); however, the focus of this paper is on steady state power exchange in systems with FACTS devices. The types of FACTS devices c w e d y available can be categorized into devices that control certain electrical parameters. For example, the UPFC can be used to control active and reactive line flows ( , ). The phase angle regulator type can be used to control active power flow. The several types of existing and proposed FACTS devices can be categorized into three types. These types are termed A, B, and C here for convenience. Table (1) lists the types modeled with the parameters controlled and examples of actual devices that most closely match the models. The literature cited in column three of Table (1) gives a more complete description of the devices cited The type denignatiom in the firat column are desigtdons used in this paper for power flow studies. Not all FACTS devlces fall into
PE-813-PWRS-0-04-1997 A paper recommended and approved by the IEEE Power System Analysis, Computing and Economics Committee of the IEEE Power Engineering Society for publication in the IEEE Transactions on Power Systems. Manuscript submittad December 23, 1996; made available for printing April 11, 1997.

In this model, a fictitious bus labeled i' is introduced to force

T.. + j(z.. 1J 1 J

to flow in the given line. The main inaccuracies of this ideal model are: device losses are ignored; actual FACTS devices do not usually control and 0 in the manner indicated; and actual devices such as the UPFC may also control other parameters such as bus voltage. FACTS Bus i Bus j Device

. v
Pi+jQi

---+Fi;

-Cj

1 Scheduled J line power

-"

P.+jQ.
J

Figure (1) Type A FACTS device


Bus i

Bus i'

Bus j

Pi+Fij

+jQ,+jq,j -F,j jQ,j Figure (2) Model of a Type A FACTS device

P.+jQ
J

Type B devices: Figure (3) shows a Type B FACTS device. In this device,

only active line power

F1J

is controlled. This type of FACTS device has a

model similar to a phase angle regulator. The model used for this device is depicted in Figure () In this model, the active power flow depicted at the 4. top of Figure (4)is shown using a fictitious bus i' in a way similar to the Type A device. The reactive flow depicted at the bottom of Figure (4) is not controlled. This model does not consider device losses and control of other circuit voltages, currents, or impedances. The Type B model differs fiom a phase angle regulator model in that no additional unknown variables are introduced (ix., no phase angle setting nor transformer tap setting). The Type B model is similar to the TCSC in that active power flow is controlled,

0885-8950/98/$10.00 0 1997 IEEE

61
but the Type B model shown in Figure (4) does not model variation of line impedance. Type C devices: This FACTS device is a bus reactive power controller. The SVC is an example of a device in this category. It is assumed that the device adjusts injected bus reactive power to control the voltage magnitude at that bus. Figure 5 depicts the Type C device and its model. Note that the specified reactive power load at bus i, jQi, is combined with the Type C device reactive power output and this reactive power varies as I Vi is controlled. The main inaccuracy of this model is that device losses are ignored. The Type C model is essentially a PV bus with P=O. This FACTS device is not the focus of attention in this paper since it is desired to emphasize the steady state power control capability of FACTS devices.

Bus i

FACTS

Dcvicc

Bus j

7..
P.GQ. J J

Pi+jQi

Type B: This device results in an added A' mismatch equation for each Type B device. The resulting submatrix dimensions are (ng represents the number of Type B devices): J1 @pq+ npv+ nB) b @pq+ npv+ nB) 53 @pq)b @pq+ npv+ nB) Y Y 52 @pq+ npv+ "B) by (ne91 J4 (npq)by ("pq). The notation used is desmbod as follows: PQ bus I has a Type B device installed, bus i remains a PQ bus and bus i' becomes a Type B bus. Type C: The Jacobian submatrix dimensions for this type are similar to Type B, namely: JI @pq+ npv+ nc)by (npq-k npv+nc) J3 (npq)by @pq+ npv+ nc) J2 @pq+ "pv+ nc) by (npq) J4 (npq)by (npq). However, note that when PQ bus i has a Type C device installed, it no longer is a PQ bus: after imtzillation of the Type C device, it is counted as a Type C bus. For each additional row and column of the Jacobian matrix due to a FACTS device, there i!; a corresponding additional power mismatch equation that must be satisfied in order for the power flow study to converge to the solution. Since these equations are similar in form to the original power mismatch equations

Figure (3) Type B FACTS device


Bus i
Bus

Bus j

Pi+Fij

-rij

P.

Bus i

Bus j

j'lj

AQ(s,~v~) =o m6, =0 the overall convergence characteristics are largely unaffected (this was verified through actual testing of a large number of representative cases). However, the specific con&uction of the Type B FACTS device model brings with it special considerations. The calculation of the voltage magnitu& and phase angle of the FACTS bus is achieved by solving two simultaneous nonlinear equations. These equations are solved with a separate Newton-Raphson study within each iteration of the large power flow study. Therefore, another set of mismatch equations must be met for each iteration of the larger study. It is porsible that this smaller Newton-Raphson study might not converge, particulzuly when voltage magnitudes are significantly less than rating. Thus, the overall solution may not converge due to divergence of the inner search.
4. Discussion of losses attributable to FACTS devices

kb

rh
6

Figure (4) Model of a Type B FACTS device Bus i Bus i

Type C device Typc C dcvicc modcl Figure ( 5 ) Model of a Type C FACTS device 3. Jacobian matrix structure and load f o study convergence in the lw presence of FACTS devices The Newton-Raphson power flow method was used to perform the required studies. In this technique, the Jacobian matrix J usually (i.e., no FACTS devices, no phase shifters, no tap changing transformers under load) has a structure given by
BP/86
. [aQias I =

aP/alv
aQ,ald]
=

J1
[J3

J2
J4]

where the dimensions of J1 through 54 are: J1 (npq + "pv)by ("pq + "pv) J4 (npq)by (npq) npq = number of PQ buses J2 (npq + npv) by (npq) npv = number of PV buses. J3 ("pq) by (npq + npv) It is assumed that there is a single swing bus and phase sh&ers and other bus types are absent. The structure of the Jacobian matrix must be modified in the presence of FACTS devices. The required modifications are: Type A: Each one of these devices cause the addition of a PQ bus, with F correspondingA' and AQ mismatch equations. Therefore, ifnA denotes the number of Type A devices are present, the Jacobian submatrices becomes: J1 ("pq+"pv+nA) by @pq+ npv+nA) J3 @pq+ nA) by @pq+ npv+nA) J2 C"pq+ npy+ "A) by ("pq + "A) 54 @pq+ "A) by. ("pq + "A). For this notation note that when PQ bus i has a Type A device mstalled. PQ bus i remains a PQ bus, and bus i' is a Type A bus.

The addition of FACTS devices to the transmission system is likely to impact the losses associated with transmitting power in that system. The change in losses will OCLU both directly, from losses occurring in the actual device, and indirectly, from the change in system parameters that result from the operation of the device. The major source of losses that are directly attributable to FACTS device are primarily caused by the solid state switching components of the device. The most prevalent type of solid state switch used in FACTS devices is the gate turn-off thyristor (GTO), which is used because of its ability to handle large currents and high voltages [15]. The losses occurring in a GTO can be classified into three types: on-state, off-state, and switching losws. These losses are proportional to the current through the device and can he modeled by adjusting the active power flow into the i' bus. All three types of losses will be small under normal operating conditions [151. It should be noteld that the insertion of FACTS devices need not increase overall system losses. In fact, such devices may significantly reduce losses. This may seem counterintuitive and contradictoryto Tellegan's theorem, which states that the naturally occurring solution of a system is also the solution with the lowest total losses. However, the insertion of the device results in a physical change in the system. Therefore, the solution of the new system may have lower losses than that of the old system. An example of this is a Type C FACTS device, such as a SVC, that reduces the flow of reactive power throupJ~ system. the
5. Further considerations concerning Type B FACTS devices

The particular construction of the Type B FACTS device model requires additional modifications to the power flow study. For the purpose of active power calculations, hp=O, the transmission line containing the FACTS device is connected to the fictitious bus i' at one terminal and to the receiving bus j at the other. However, the line is connected to the actual sending bus i and the receiving bus j for the reactive power calculations, A Q=O. This requires two separate admittance bus matrices for the calculation of the first and fourth quadrants, J1 and J4, of the Jacobian matrix of the fast decoupled power flowstudy. The admittance matrix used in the calculation of J is n by n where n indicates the number of actual buses. The admit4 ,

62
tance matrix used to determine J1 is (n+nB) by (n+ng) where nB indicates the number of Type B FACTS devices. While the active power of the FACTS device is specified, the reactive power is not. The reactive power coming from the device must be the same as the reactive power coming from the sending bus along that line. Therefore, the reactive power going from the sending bus to the receiving bus, measured at the sending bus, is calculated during each iteration of the Newton-Raphson power flow study sending bus by the parallel path. Either Type A or Type B devices are suitable for this purpose. Electronic fence: The concept of an electronic fence is an attempt by a utility to protect its property rights by preventing another utility from using its transmission system. This is analogous to a physical fence keeping trespassers from passing through an individuals property. In the case of the electronic fence, there are limited number of access points, these being the tie-lines connecting the utility to its neighbors. In the analogous case, consider an area that can only be accessed along certain paths. The ability to control tr&c flow through the area is dependent on the number and placement of gates at the access points. Theoretically, a gate would be needed at all but one access point in order to perfectly control the number of visitors passing through, assuming that traffic can only flow in one direction through an access point at a time. This analogy holds for the electronic fence. In order to completely control the amount of power passing through the utility, it would need to control the flow on all but one of its tie-lines, which would be prohibitively expensive. However, some success can be achieved by fewer devices if they are strategically located, especially if the number of interconnections is low. The key locations for these devices are those tielines that are most heavily loaded. Similar to a physical path that is difficult to access, a weak tie-line connection is not likely to be heavily used. Therefore, a FACTS device in that line will not have a great effect. Iftwo or more parallel paths exist, a single device will have little effect, since the flow will travel over the uncontrolled tie-lines. Either Type A or Type B devices are suitable for this purpose, unless a considerable amount of undesired reactive flow occurs. In such an instance, Type A devices are favored. Enhanced economic operation: There are two distinct means of placing a FACTS device in a transmission system for the purpose of increasing the systems ability to transmit power, thereby allowing for the use of more economic generating units. The first is to place the device in an underutilized transmission line. This allows more power to be pushed through the line. The second way of locating a FACTS device for this purpose is to place the device in the most heavily loaded line, using the device to limit the flow in that line. This allows more power to be sent through the remaining portions of the system while protecting the line with the device from being overloaded. The second method, which sites the devices in the heavily loaded line, is the most effective. In general, either Type A or Type B devices may be used. If reactive power flows are a significant portion of the total flow on the limiting transmission line, either a Type A device in the line or a Type C device located at the end of the line that receives the reactive power may be used to reduce the reactive power flow, thereby increasing the active flow capability. Obtaining a specijk operating conditzon: The placement of a FACTS device to obtain a desired operating condition, such as correcting an undervoltage or forcing a certain amount of current flow through a line. The proper location and type of device will be determined by the specific condition. For instance, ifa chronic undervoltage occurs at a given bus, a Type C device at that bus is an obvious candidate. Ifthe utility wants to be able to control the flow in a specific line, either a Type A device or a Type B device may be used in that line. The intended objective of a FACTS device will have a large impact on the optimal location of the device. A location that is best for one objective may be less than optimal for another. Additionally, since the characteristics of each utilitys system are unique, optimal locations may vary between utilities. Therefore, a series of guidelines were developed for choosing the type and location of FACTS devices for various possible objectives. These guidelines are summarized in Table (2).
7. An example based on regional flows and operating positions In this section, an example is provided to illustrate the value of FACTS devices for several power flow control objed:ves. The example L based on an actual system in the midwestem USA although the description has been condensed here. The recent merger of two weakly connected midwestem utilities brought about the questions of how the two companies would transmit power between them, how that transfer would affect neighboring utilities, and whether FACTS devices might be useful in facilitating that transfer. Therefore, a generation and transmission system model that is loosely based on seven interconnected midwestem utilities was developed. This model is used to exemplify the potential uses and control strategies of FACTS devices discussed in this work. It is important to note that the sys-

W(5 -

*-* 5)~ i ?>. j

This value is used for the scheduled reactive power flow of the device at each iteration for power mismatch purposes. During each iteration of a Newton-Raphson load flow study, the changes to the bus voltage magnitude and phase angle are calculated; however, there are no entries in 54 corresponding to the fictitious bus, if . Therefore A I V if I and A6 it are calculated separately. Since the active and reactive power flows coming out of the device and the voltage at the receiving bus j are known, the voltage magnitude and phase angle of the FACTS bus can be determined &om
P+jQ

(5,-

Vi) yij

*-*

7,

This equation can be viewed as two separate equations by breaking it up into its real and imaginary portions. To solve these two simultaneous nonlinear equations, the Newton-Raphson method was used. Therefore, during each iteration of the larger Newton-Raphson power flow study, an internal iterative Newton-Raphson study is performed. 6. Optimal siting of FACTS devices Having made the decision to install a FACTS device in the system, there are three main issues that must be addressed: what type of device should be used, how much capacity should it have, and where i the system n should it be placed. While this section is primarily concerned with the last of these issues, the first two will be briefly addressed. Assuming that the cost of a particular device is a function of the power transfer capability, it would not be desirable to install a device that is overly large for its intended purpose.. Anything larger than the rating of the transmission line in which it is going to be installed would not be economic since the line limit would prohibit the device from being used to its full potential. Likewise, ifthe device is too small and it cannot handle as much power flow as the transmission line, the utility has effectively reduced the rating of the transmission line. Therefore, the size of the transmission line should be determined by the rating of the associated transmission line, keeping in mind the potential for later upgrading of the line. Like the decision of where to place a FACTS device, the choice of which type of device is greatly impacted by the desired effect. For instance, a Type C device should be considered when reactive power control or voltage support is necessary. Type B devices may not perform well in lines with high reactive power flow. Also, the relative costs of the devices will have a considerable effect on which device is chosen. It is likely that the cost of a device is inversely proportional to the maturity of the technology. This would indicate that the SVC, a Type C device, is among the cheapest and the IJPFC, a Type A device, would be one of the more expensive. The decision of where to place a FACTS device is largely dependent on the desired effect and the characteristics of the specific system. One possible method for determining the optimal location of a device is to perform a separate power flow study with the device in each line. This would be quite time consuming for a power system of even moderate size. However, it is possible to identify places in the system that are potential candidates for FACTS controllers. Some general guidelines for the placement of FACTS devices for various objectives are: Prevention of loop flows: The placement of FACTS controllers in the power system to prevent loop flows depends primarily on the location of the loop flows. The device should be placed in one of the transmission lines on which the loop flow occurs, with power flow either being forced to zero or sent in the opposite direction ofthe loop flow. Ifa direct parallel path exists, such as another line between the same buses, the effectiveness of the device for this purpose will be greatly reduced. This is because some of the power that is being directed through the FACTS device will be shunted back to the

63
tem model was developed for the purpose of illustration and should not be construed as an accurate model of the interconnected utilities The pictorial in Figure (6)depicts the interconnections between utility companies in the example system. When multiple interconnections exist between utilities, they have been aggregated for purposes of illustration. However, the individual detail for each interconnection was maintained in the examples. The actual data used in the example studies was based on the 1994 FERC data on file. A power flow study was perfomed using the midwestem regional model without any FACTS devices or scheduled power transfer between companies. The voltage magnitudes of the illustrated 765 kV buses were set to 1.025 pu.. The higher operating voltage on the 765 kV portion of the system is consistent with actual utility operating practices. The results of the power flow study are also depicted in Figure (6)and this will be termed the base case which is used for comparison purposes.
b

in Figure (8). This example shows a 550 MW sale from utility D to utility C. This value was chosen since it is the largest amount of power that can be transferred without overloading any of the transmission lines in the total system. The limiting line is m e of the two 345 kV lines connecting utilities A and C. While the l i e directly connecting utilities C and D appears to be the most liely candidate for power flow, only 28 MW flows through it. This is because the line in question is a relatively high impedance 138 kV line. Lower impedance panhs exist through other utilities, particularly through utility A. This power transfer example will also be used in later examples to illustrate the modeling of FACTS devices in Newton-Raphson power flow studies for other . . purposes.

f-\

Table (2) Guidelies for siting FACTS devices Obiective I Guidelines I TvDe I IPrevention of loop flows IPlace in flow path I AorB I Avoid direct parallel paths Best when few tie-lines exist Electronic fence A or B Place in heavy flow lines AorB More economic oueration Place in heaw flow lines (with low reactive flow) More economic operation Place in high reactive flow line or A or C (with high reactive flow) at reactive flow receiving bus Obtain a specific operating Dependent on the specific condi- A, B, or condition tion desired C

Figure (7) Power flows (in MW) with B-D-A-B loop flow prevented using a Type. A FACTS device

Figure (6 ) Base case power flows (in MW) between-utilities The following studies exemplify the use of FACTS devices for various objectives: Prevention of Loop Flows The base case shown in Figure (6) contains numerous examples of loop flows. An example of the use of a FACTS device to prevent a loop flow involves the introduction of a Type A FACTS device in the single transmission line between company B and company D to stop the B-D-A-B loop flow. In this instance, the device is placed on the D end of the transmission line and is set so that 20 MW and 1 MVAr are forced to flow from utility D to utility B along that line (measured at D). The base case had a flow of -28 MW, 53 MVAr. The resulting power transfers are shown in Figure ( ) Note that the B-D-A-B 7. loop flow has disappeared and that the insertion of the FACTS device caused little effect on the rest of the system. Example of an electronic fence This is an example of the use of FACTS devices for the protection of a utilitys transmission assets from the undesired wheeling of power by the utilitys neighbors. The concept of an electronic fence is the prevention of unwanted flows through the utility, thereby allowing the utility to use its own trmmission facilities more filly. To illustrate the use of FACTS devices as an electronic fence, an example has been developed that involves a power transfer between two companies. Since the two utilities in question are not strongly interconnected, a significant portion of the transferred power flows through other utilities, as shown

Figure (8) Power flows (in MW) with utility D sellig 550 MW to utility C with no FACTS device present Many candidates exist for the placement of a FACTS device to act as an electronic fence. Perhaps the most obvious are the heavily loaded 345 kV line connecting utilities A and D and one of the three lines between utilities A and C. If utility A were to place a Type A device in one of its 345 kV tie-lines with utility C in an attempt to prevent the undesired wheeling of power through its system. The power flow setting of the device was adjusted to minimize the total amount of power transferred from utility D to utility C through utility A. The resultant power flows are shown in Figure (9). While the device: performs its function of controlling the power flow on that line, it is somewhat less successful in stopping the unwanted flow through utility A. Some of the displaced power flow now travels through the 138 kV tie-line and the other 345 kV tie-line between A and C. The device is marginally successful in this case in that the total amount wheeled through the system lis reduced h m 474 MW to 67 MW. As was the case in the previous example, the ability of the device to produce the desired result is diminished when other direct parallel flow paths exist. Additionally, the type of device used for this purpose may affect the result. A Type B device model was initially used, but the ability to directly control reactive power flows was necessary to significantly reduce the wheeled power. This was caused by excessive reactive power flow, e.&, several hundred MVAr,limiting the active power capability of the line. The Type A device prevented this problem since it allows direct control of the reactive

64
power flow. In addition, the inclusion of the FACTS device reduces the total power transfer capability between utilities D and C by reversing the flow through the most heavily loaded line. The power sent &om D to C is now only 210 MW, as compared to 550 MW in Figure (8). From the preceding example, it can be seen that FACTS devices can be used as electronic fences to varying degrees of success. In situations where the utility is somewhat isolated From the remainder of the system, they can be used quite successhlly. However, in cases where the utility is highly interconnected with its neighbors or where directly connected parallel paths exist, the efficacy of the electronic fence will be lower. Additional devices may be needed to control flow along other paths. This could be considered to be similar to a physical area with numerous possible entrances and exits. In order to control the amount of traffic through the area, one would need to place gates at many of them in order to restrict the flow through the area. However, if there were only a single entrance along with a single exit, one gate would suffice. Example of increasing power transfer capability This is an illustration of the modeling of different types of FACTS devices to increase the power transfer capability between two regions. This can be very si@icant economically in situations where transmission constraints prevent the most economic generating units fiom being fidly utilized. Currently, relatively cheap generation units in portions of the southern and midwestern regions of the United States are operated at less than full capacity while more expensive generation is used in the northeastern region of the nation. The example involves the use of a Type 3 FACTS device in the 13 8 kV transmission line that directly connects utilities C and D. Recall F o Figure ( 8 ) that this tie rm is relatively weak, resulting in low power flow levels. The inclusion of the FACTS device can increase the power flow through that line and allow greater use of utility Ds generation capacity. The results of this example are shown in Figure (10). The inclusion of the FACTS device, which is set at 240 MW and inserted on the D end of the transmission line, increases the overall power transfer fiom utility D to utility C lo 730 MW. In some instances, FACTS devices can be used to improve power transfer capability, thereby allowing more economic generators to be dispatched. The degree of success that can be achieved depends on the choice of the transmission line in which the device is installed. The greatest effect occurs when the device can reduce reactive flow on a key line,thereby allowing more active flow, or when the device protects that line from overloading while allowing the remaining lines to be more heavily utilized. In [16], alternative FACTS devices were used to study their effectiveness in increasing power transfer capability. The results of those studies appears in Table (3). Note that the lesser effect ofthe Type I device in Figure (10) is 3 a product of the location of the device rather than the type of device. The wheeling utilities would only add the device if it were economically advantageous, i.e., if they believed that the income that they would receive through wheeling charges would allow them to recoup their investment in the device. Additionally, it should be noted that while this system is based on seven midwestern USA utilities, a similar situation could exist within a single utility or power pool: a utility whose transmission constraints restrict its generation dispatch may be able to use a FACT device to obviate the problem. Example of unloading a selected line Another potential use of FACTS technology is to unload a selected transmission line. This can be accomplished in either the line containing the FACTS device or in another circuit. The first case is trivial simply set the power of the FACTS device to zero. The second case involves altering the setting of the device to match the voltage magnitudes and phase angles of the buses at both ends ofthe transmission line that is to be unloaded. Figure (11) shows the two 345 kV l i e s es D and E with their power flow levels and bus voltages fiom the example depicted in Figure (8). Figure (1 1) shows the megawatt and megavar flows in the absence of a FACTS device (solid arrows) and with a Type A device in one line (dashed arrows). The device settings were adjusted to unload the other 345 kV circuit. Table (3) Comparison of power transfer capabilities in the given example for various FACT devices
/
\

SA

Figure (9) Power flows (in MW) with Type A device installed as an electronic fence i an intertie between utility A and utility C n

Figure (1 1) Focus on regions D and E with D selling 550 MW to C: with no facts device (solid arrows); with Type A FACTS device installed in a lme in order to unload a circuit (flows in MW, MVAr) Note that the control settings of the device i Figure (11) are n larger than the sum of the active and reactive power flows of the two lines without the FACT device. As the transmission line containing the FACTS device assumes more of the power flow. it is not only the direct parallel line that is affected. Power that previously flowed through other parts of the system is shunted through this path. Therefore, the power transferred through the FACTS device is greater than the sum of the power transferred by the two lines in the absence of the device. Also note that the buses at both ends of the transmission lines in question are PV buses. This is important in that the bus voltage magnitudes are controlled to 1.00 pu by the generator at that bus. This makes the adjustment of the device settings easier in that only the phase angles need to be matched. It would be more difficult to match the voltages ifthe ends were PQ buses.

Figure 10) Power flows (in MW) with Type B device installed in 138 kV C-D intertie to increase power transfer capability

65
Example of directingflow between regions To illustrate the modeling of FACTS devices in a power flow study for the purpose of directing power flow between regions, two cases are analyzed The first case involves the transfer of 200 MW ofpower fiom utility C to utility D without any FACTS devices. This is done by generation dispatch. The results of the power flow study for this case are shown in Figure (12)using solid arrows to denote the active power flows in this case. Losses are low in this system, and only the lw 1) active power f o s sent are depicted in Figure ( 2 . Note that of the 200 MW transferred, only 2 MW flow directly to utility D. The remaining power is wheeled through other utilities. Almost all of the power is transferred through utility A, which wheels approximately 191 MW. The second case, also shown in Figure (12)with dashed arrows denoting megawatt power flows, transfers the same amount of power from utility C to utility D. However, a Type B FACTS device is modeled in the line between C and D, thereby forcing the power to flow directly fiom utility C to utility D. Note that utility A no longer wheels a substantial amount of power. On the other hand, a substantial amount of power is lost as large flows are forced over the relatively high resistance intertie.

[5]R.Nelson, J. Bian, S. Willliams, Transmission Series Power Flow Contro1,IEEE Trans.PowerDelivery, v. 10,No. 1, pp. 504-510, 1995. Jan. [6]M. Noroozian, G. Anderson, Power Flow Control by Use of Controllable Series Components, IHX Trans. Power Delivery, v. 8 No. 3, pp. , 1420-1429, 1993. July [7]E. Larsen, C. Bowler, B Damsky, S. Nilsson, Benefits of ThyristorControlled Series Compensatiion, CIGRE SC 14,Paris, 1992. [8]D. McDonald, B. Damsky, Modeling and Testing of a Thyristor for Thyristor Controlled Series Compensation (TCSC), IEEE Trans. Power January 1994. Delivery, v. 9,No. 1, pp. 352-359, [9]M. Iravani, et al., Applications of Static Phase Sh&ers in Power Systems, IEEE Trans.Power Dldivery, v. 9,No. 3 pp. 1600-1608, , July 1994. [lo]H. Fujita, H. Kurebayashi, H. Nohara, M. Goto, Y. %to, Loop Power System Control by High-speed Phase Shifter, Electrical Engineering in Japan, v. 115, No. 3 pp. 60-63, , July 1995. [ 111 IEEE Special Stability Controls Working Group, Static VAR Compensator Models for Power IFlow and Dynamic Performance Simulation, IEEE Trans.Power Systems, v. 9,No. 1, pp. 229-240, February 1994. [12] M. Pamiani, M. Iravani, Voltage Control Stability and Dynamic Interaction Phenomena of Static VAR Compensators, IEEE Trans. Power 0 August 1995. Systems, v. 1 ,No. 3,pp. 15!)2-1597, [13]D. Ramey, R. Nelson, J. Bian, T. Lemak,Use of FACTS Power Flow Controllers to Enhance Transmission Transfer Limits, Proceedings. April 1994. American Power Conference,v. 56,pt. 1, pp. 712-718, [14] Schauder, M. Gemhardt, E. Stacey, T. Lemak, L. Gyugyi, T. Cease, C. A Edris, Development of a Plus or Minus 100 MVAR Static Condenser for Voltage Control of Transmission Systems, IEEE Trans. Power DelivJuly 1995. ery, v. 10,No. 3,pp. 1486-1493, [ 151 N. Mohan, T. Undeland, W.Robbm, Power Electronics: Converters, Applications andDesrgn, John Wiley & Sons,New York, 1995. [16]D. Gotham, Flexible AC Transmission Systems in Power Network Analysis and for Power Flow Control, PbD Thesis, Purdue University, West Lafayette, IN, December 1996.

Figure (12)Active power flow (sent) in megawatts with C s z i g 200 MW to D solid arrows show case with no FACTS device present; dashed arrows show flows with a Type B FACTS device installed in the C-D intertie 8. Summary and conclusions This paper provides a series of examples that illustrate the modeling of FACTS devices in power flow studies. These power flow studies were performed using a model based on the actual interconnections of seven midwestern utilities. The examples were developed under the following scenarios: prevention of loop flows, implementation of an electronic fence, increase in power transfer capability, unloading a selected l i e , and directing power flows between regions. Additionally, some scenarios exemplify more than one use of FACTS technology. While there are definite benefits to be gained with each of these potential uses for FACTS devices, there are also disadvantages. In many cases, the use is specific to a certain operating condition that. Therefore, the high cost of a FACTS device may not be justified if that is the only purpose for installing the device. However, a combination of the benefits detailed here, along with transient considerations, could justify the cost of a device. Table (4)is a summary of potential advantages and disadvantages involved with the use of a FACTS device for the examples presented.
References

[I] N. Hingorani, Flexible AC Transmission, IEEE Spectrum, v. 30,No. 4,pp. 40-45, April 1993. [2]IEEE Special Stability Controls Working Group, Static VAr Compensator Models for Power Flow and Dynamic Performance Simulation,IEEE Trans.Power Systems, v. 9,No. 1, pp. 229-240, February 1994. [3]L. Gyugi, Dynamic Compensation of AC Transmission Lines by SolidState Synchronous Voltage Sources, IEEE Trans. Power Delivery, v. 9, NO. pp. 904-911, A p d 1994. 2, [4]R. Mihalic, P. Zunko, D. Povh, Improvement of Transient Stability Using Unified Power Flow Controller, IEEE Trans.Power Delively, v. 11, NO. 1, pp. 485-491, J~IIIUW 1996.

Biographies Douglas J. Gotham is from Baghaw, Michigan. He holds the BSEE degree from Michigan Tech University in Houghton, MI and the MSEE and PhD degrees f o Purdue University, West Lafayette, IN. He is presently holds a rm position with the State Utility Forecasting Group at Purdue. Gerald T. Heydt is from Ls Vegas, Nevada. He holds the BEEE degree a from the Cooper Union in New York, and the MSEE and PhD degrees from Purdue University. He spent approximately 20 years as a faculty member at Purdue and recently he took the position of Director of the NSF Center for the Advanced Control of Energy and Power Systems at Arizona State University, Tempe AZ. He IS a fellow of IEEE.

You might also like