446

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 1, JANUARY 2011

Damping Performance Analysis of IPFC and UPFC Controllers Using Validated Small-Signal Models
Shan Jiang, Student Member, IEEE, Ani M. Gole, Fellow, IEEE, Udaya D. Annakkage, Senior Member, IEEE, and D. A. Jacobson, Senior Member, IEEE

AbstractThe paper discusses the dynamic behavior of two different exible ac transmission system devices; the interline power-ow controller (IPFC) and the unied power-ow controller (UPFC) in a benchmark system. The small-signal model of the interline power-ow controller is developed and validated using detailed electromagnetic transients simulation. Using this validated model, the damping capabilities of the IPFC and the UPFC are compared and rationalized. From a small-signal dynamics point of view, it is shown that the series branches of these devices essentially segment the network creating a new structure. This structure change may be used to effectively improve system damping without requiring the design of a tuned feedback controller. The IPFCs two series branches in contrast to the UPFCs single series branch permit more opportunities for network segmentation. Hence, the IPFC has greater potential for improving the systems dynamic performance. Index TermsDecoupled control, electromagnetic transients (EMT), exible ac transmission system (FACTS), interline power-ow controller (IPFC), small-signal stability, unied power-ow controller (UPFC).

Fig. 1. Converter-based FACTS.

I. INTRODUTION

LEXIBLE AC transmission system (FACTS) controllers based on voltage sourced converters (VSCs) provide a potentially attractive solution to controlling power ow in the modern electric network [1]. Such devices can be used for improving power transfer limits, for congestion management in the network as well as for damping oscillatory modes. Two examples of VSC-based FACTS devices that have recently received attention are the unied power-ow controller (UPFC) and the interline power-ow controller (IPFC) and are shown in Fig. 1. The UPFC consists of a shunt VSC and a series VSC connected via a common dc bus which includes a dc capacitor for ripple control. The shunt VSC provides voltage support to the connected bus. The series VSC has the ability to precisely
Manuscript received November 10, 2009; revised March 01, 2010; accepted July 13, 2010. Date of publication September 30, 2010; date of current version December 27, 2010. This work was supported in part by Manitoba Hydro and in part by the Industrial Research Chair Program of the Natural Sciences and Engineering Research Council of Canada (NSERC).. Paper no. TPWRD-008312009. S. Jiang, A. M. Gole, and U. Annakkage are with the Electrical and Computer Engineering Department, University of Manitoba, Winnipeg, MB R3T 5V6, Canada (e-mail: sjiang@ee.umanitoba.ca; Gole@ee.umanitoba.ca; annkkag@ee.umanitoba.ca). D. A. Jacobson is with Interconnections and Grid Supply Planning Engineer with Manitoba Hydro, Winnipeg, MB R3C 2P4, Canada (e-mail: dajacobson@hydro.mb.ca). Digital Object Identier 10.1109/TPWRD.2010.2060371

control power ow in the line. In the IPFC [2], the two VSC converters are both inserted in series with two different lines and share a common dc bus. Hence, they have the capability to precisely control power ow in two different transmission lines. There is a signicant body of work on the UPFC. Since the original introduction of the concept [3], [4] , most aspects of the UPFC, such as its operation, modelling [5][9], and eld experience [10] have been well investigated. However, in contrast, the literature on the IPFC is relatively scarce. Reported work on the IPFC considers aspects such as power ow and constraints [11], [12] , and comparison with other FACTS devices [13]. There have been very few studies on the IPFCs small-signal stability performance. Reported studies in the literature have either considered an idealized simple system or omitted detailed validation [14], [15] [16]. This paper rst investigates the dynamic performance of the IPFC by developing a small-signal model of the IPFC that considers a decoupled control system [7], [17] , in which the direct (d) and quadrature (q) current control loops are decoupled, enabling independent control of real and reactive power. Next, the IPFC model is combined with a small-signal representation of a 12-bus four generator network [18] to investigate applications, such as providing damping in multi-modal power networks. A second objective of the paper is to compare the damping performance of the IPFC with a related FACTS devicethe UPFC. A third objective of the paper is to validate the combined IPFC-ac network small-signal model with a detailed electromagnetic transients (EMT) model. The selected network is ideal for this purpose, as it is of sufcient size to demonstrate multiple oscillatory modes (local and inter-area); and yet is small enough to be amenable to detailed electromagnetic transients simulation [18]. Of all simulation models, EMT models the network elements in the fullest detail. This is in contrast with transient stability programs which typically consider steady-state or quasi

0885-8977/$26.00 2010 IEEE

JIANG et al.: DAMPING PERFORMANCE ANALYSIS OF IPFC AND UPFC CONTROLLERS

447

Fig. 3. IPFC small-signal model.

Fig. 2. Twelve-bus power system.

TABLE I POWER FLOW OF THE SYSTEM

it should have a stronger inuence on the power ow and on oscillatory modes. The test system is used to validate the IPFC small-signal model and demonstrate potential performance improvements. III. SMALL-SIGNAL MODEL This section develops a small-signal model for the IPFC and incorporates into the small-signal model of the overall test system. The description of the various submodels and their synthesis into a model for the entire system will be discussed.

TABLE II OSCILLATION MODES WITHOUT FACTS

A. IPFC Model The IPFC small-signal model is shown in Fig. 3. The primary function of the IPFC is power-ow control with injection of a voltage in series with its host line. The IPFC controls the magnitude and phase angle of the injected voltages in each line, resulting in four degrees of freedom. One degree is used for regulating the capacitor voltage of the dc link. The reand reactive maining three are used to control the real power in one line, and the real or reactive power in the other. In this paper, the line on which and are controllable is called the master line, and the other line on which only either or is controllable, is called the slave line. All variables and ratios related to the master line and slave line are denoted with the subscripts and , respectively. It is important to note that one of three coordinate systems is employed in this paper. Initially, the master or slave lines sending-end voltage or are selected as reference vectors ( axis) for vectors the master or slave line equations, respectively. Hence, the , current components ( , , , and ) contribute to the , , and ), respectively. real and reactive powers ( , The other coordinate system is the networks coordinate system ( - system) which takes the swing (innite) bus voltage as its reference ( axis). All vectors of the network are nally converted into these - components for system-wide analysis. The dynamic behavior of the IPFC is represented by two groups of state equations. One represents the dynamics of its dc capacitor; the other, the control system. In this research, the popular decoupled control method [7], [17] is assumed, as shown in Figs. 4 and 5 for the IPFCs master and slave branches, respectively. The controllers attempt to maintain the real and reactive current of these branches at their set-point values. With decoupled control, the and axis current loops do not interact

steady-state phasor representations for the bulk of the ac network and use differential equation representations only for generators and a few other dynamic devices in the network. Hence validating the small-signal model with an EMT model provides the best comparison short of comparison with an actual system. II. TEST NETWORK FOR IPFC EVALUATION This system shown in Fig. 2 has earlier been proposed as a benchmark platform for FACTS and HVDC studies [18]. It consists of twelve buses (six 230-kV buses, two 345-kV buses and four 22-kV buses) and covers three areas. Area 1 is a generation center and area 3 is a load center. Power-ow studies (see Table I) show that the system has congestion following partial generation loss or tripping of a critical transmission line. Also, Table II, obtained from small-signal analysis, shows three of the least damped modes. It is clear that two oscillation modes are signicantly underdamped (i.e., 5%). These issues can be resolved with the installation of a FACTS device. One solution is to place a UPFC on line 16 or 78. Placing the UPFC series branch in series with line 78 alleviates the congestion in line 16 and improves damping [18]. Another solution is investigated further in this paper- that of placing an IPFC on two of the three lines (line 12, 16 and 78). Since the IPFC has one more series branch than a UPFC,

448

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 1, JANUARY 2011

The decoupled controllers (master line and slave line) are modeled by mathematical representation obtained from Figs. 4 and 5 (not including the nonlinearities) yielding (3)(10):

(3)

(4)
Fig. 4. Decoupled controller of the IPFCs master branch.

(5) (6) where , , , and current errors as given in are the integrals of IPFCs

(7) (8)
Fig. 5. Decoupled controller of the IPFCs slave branch.

(9) (10)

with each other and the real and reactive current can be independently adjusted without perturbation to the other. The controller to the also limits the injected voltage magnitude rated maximum (Magmmax and Magsmax, respectively, for the master and slave converters). The implementation of this block is shown in the Appendix. The charging equation for the dc capacitor is obtained by equating the power entering capacitor with the power in the dc link and losses

Differential (1)(10) are linearized as state (11) and algebraic (12), (13) and expressed in the global reference frame ( ), with the swing bus voltage as the -axis

(11) (12) (13)

(1) where is the capacitance of the dc bus capacitor, is the voltage of the dc bus, and are the real power owing out of the dc link into the transmission line through the master and is a resistor in parallel slave branch of IPFC, respectively. is with the dc capacitor and represents the converter losses. the system base megavolt-ampere. at the reference setting The PI controller that maintains is modeled as

where

is a 2-D - vector of the master line current. Similarly , , , , and are vectors that contain the -axis and -axis components of their corresponding variables. B. Generator Model

(2) is the axis current order for the slave line. This degree of freedom is allocated to maintaining the dc capacitor at its , reference voltage. The other three current orders ( and ) of the IPFC can be arbitrarily specied.

The generators and exciters are modeled and linearized as (see the Appendix for details) (14) (15)

JIANG et al.: DAMPING PERFORMANCE ANALYSIS OF IPFC AND UPFC CONTROLLERS

449

Fig. 6. Integration of models into the system.

C. System Model The small-signal models of the IPFC and generators are integrated with the network as shown schematically in Fig. 6. The transmission lines are represented as Pi-sections, and transformers as leakage impedances. Hence, the V-I relation of the complete network can be modeled as (16) where

and

are the current and voltage vectors of all buses of the network. is the admittance matrix of the network. The small-signal model can be obtained in the standard form and from the dynamic (17) and , (18) by eliminating devicess differential-algebraic (11)(15) (17) (18) Validation of the model developed in the paper uses current reference as the input to the IPFC/UPFC controllers. As the sending voltage is relatively xed, controlling current is qualitively similar to controlling power. If a power injection model can is required, additional equations such as be added and linearized. These can be added to (16) to generate a power model. IV. VALIDATION OF SMALL-SIGNAL MODEL The small-signal model state equations were implemented in MATLAB. This state variable formulation can be used for small-signal stability analysis through eigenvalue calculation. It can also be numerically integrated to produce time-domain results. Results from the numerical simulation can be compared with a detailed EMT model for validation purposes. This validation was carried out by comparing the simulation results from the small-signal model with an EMT program (PSCAD/EMTDC). The location of the IPFC device is based on power-ow control considerations. In the system, the main purpose of the added device is to relieve the overloading in line 16. This can be realized by placing a series branch of the IPFC in line 16. Congestion control is also possible with the use of a UPFC-type FACTS

device, in which the UPFC series element is inserted in line 16. However, in this case, the power spills over into alternate paths 78 and 12 based on the relative impedance of each branch. With an IPFC, the second branch can be inserted in alternate line 78, thereby allowing precise control in two transmission lines. Thus, from the perspective of power-ow control, the IPFC is a better option since it allows precise power transfer levels on two lines. In this paper, the UPFC is selected as the basis for comparison of the performance with the IPFC. The following ve cases are considered: Case 1) the original system without any FACTS device; Case 2) an IPFC is placed in line 16 and 78: it relieves congestion of line 16 and improves the utilization of the transfer capacity of line 78; Case 3) a UPFC is placed in line 16: it relieves congestion of line 16, but the extra power sent to Area 3 is on lines 78 and 12, which may result in the possibility of loop ow of power; Case 4) a UPFC is placed in line 78: it transfers more power through line 78; it will relieve the congestion of line 16, but also may result in undesirable loop ow in line 12; Case 5) an IPFC is placed in line 12 and 78: it transfers more power through line 78, but line 16 (congested line) is not directly controlled by the IPFC. directly control the In the aforementioned cases, cases power ow in line 1-6; while in Case 5, line 1-6 is not directly controlled by the IPFC. Although from a congestion-management viewpoint it is not a direct control option, it is presented in this paper to demonstrate the damping effects of IPFC later in Section V. The validation effort consisted of comparing the detailed EMT simulation of a full nonlinear system and small-signal simulation results for disturbances to the systems of cases 1 through 5 before. An agreement indicates that linearization is truly representative of the full detailed system at the operating point. It also adds condence that no algebraic errors were made during the derivation of the small-signal model. Results are shown below for a representative case (Case 2), where a 2% step change was applied on the IPFCs reference . Figs. 79 show representative traces of machine speeds and real powers of the three generators and the IPFC obtained from electromagnetic time-domain simulation of the full (nonlinear) system and from the linearized small-signal model. (also see Fig. 7), for step changes Fig. 10 shows the speed of 2%, 8%, and 14%, respectively. The curves are normalized to the step disturbance size, so they can be overlaid for comparison. The IPFC controller ceiling (see Fig. 5) was found to be reached only for the 14% step. From a visual inspection, the curves for 2% and 8% disturbance have similar shapes and, hence, similar frequency and damping to that predicted by the small-signal model (note that the curve for 2% damping was earlier shown to be identical to that from the small-signal model: Fig. 7, graph). However, once a controller limit is reached for the 14% step, the response become oscillatory and completely different from that predicted by the small-signal model, indicating that the small-signal model is no longer valid for this size of step change.

450

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 1, JANUARY 2011

TABLE III PARTICIPATION FACTORS-CASE 1 (NO FACTS)

Fig. 7. Generator rotor speeds.

TABLE IV OSCILLATION MODES

Fig. 8. Generator real powers.

Fig. 9. Real powers of the master line and slave line of the IPFC.

Fig. 10. Responses of the generator speed for different step changes.

V. DAMPING PERFORMANCE STUDIES The developed small-signal model was used to investigate the damping performance of the IPFC. This section summarizes the results of this study. Consider the base case (Case 1) where no FACTS device is installed. Each of the three generators (third order model) and exciter (rst order model) introduces four states. Therefore, without any FACTS device, the system has 12 states. Eigenvalue analysis shows that there are three pairs of complex conjugate eigenvalues corresponding to lower frequency electromechanical modes (see Table II). The participation factors of

these modes indicate that the three oscillation modes (0.85, 1.12, 0.75 Hz) are dominated by generators G2, G3, and G4, respectively (see Table III). In the following discussion, these modes are referred to as Mode 2, Mode 3, and Mode 4, respectively, to associate the mode with the generator that dominates the particular mode. All three modes have low damping, with Modes 2 5%). and 3 being critical This section discusses the IPFC damping performance in the test system. To show the advantages of IPFC, the UPFC is chosen as the basis for comparison. The ve cases presented in Section IV are used in this section. It can be noticed that in all of these cases, the UPFC and IPFC introduce additional modes but not in the critical low frequency electromechanical range of concern. The impacts of FACTS on electromechanical modes are shown in Table IV. It is clear from Table IV that once the series branch of UPFC or IPFC is inserted at the sending end of a line that has a generator connected close to its receiving end, it changes the dominant oscillation mode associated with that generator, both in frequency and damping. Those modes whose dominant generators are electrically remote to the FACTS installed line are marginally affected. To explain the aforementioned observation, the participation are investigated. The participation facfactors for Cases tors indicate the association of a state with a certain mode. According to Table V, Modes 2 and 3 are still dominated by G2 and G3 for their states , and have the largest participation factors in these modes. This means, the UPFC on line 16 did not change the association of G2 and G3 to the Modes 2 and 3, respectively. However, for Mode 4, in addition to G4, the also shows a large participation factor. This UPFC state is expected because the UPFC is located in line 16 which is connected to G4. is the integral of the UPFCs series branch Note that state axis current component . Therefore, the large participation factor means that the series branch of the UPFC has a signicant participation in Mode 4.

JIANG et al.: DAMPING PERFORMANCE ANALYSIS OF IPFC AND UPFC CONTROLLERS

451

TABLE V PARTICIPATION FACTORS OF CASE 3-UPFC IN LINE 16

TABLE VII CONTROLLABILITY INDICES OF CASE 2

5

TABLE VI PARTICIPATION FACTORS OF CASE 2-IPFC IN LINE 16 AND 78

Similar participation of the series branch can be noticed in other cases too. In the case of IPFC (Cases 2 and 5), the difference is that due to the coupling introduced by the IPFC, both generators at the ends of the two lines participate in the corresponding modes. This is evident in Table VI where Mode 4 has both G3 and G4 participating in it in addition to the two series branches of the IPFC. The UPFC, on the other hand, does not introduce such coupling. Another point of interest is determining the effectiveness of different FACTS device types for controlling low frequency oscillations. The inspection of the controllability indices [19] or participation factors of the low frequency modes permits this analysis. In small-signal analysis, the controllability index indicates the inuence of a given input to a mode. Table VII shows the . In cases 2 and 5, which controllability indices of cases consider the IPFC, the column entries are those described by , , ). For cases 3 and 4 that the IPFC headings (i.e., have the UPFC, the entries are those described by the UPFC , , and ). It is found that: headings (i.e., 1) for the IPFC cases (Cases 2 and 5), the largest controllability index, shown in bold, for any of the modes is in columns corresponding to real-power-related variable (i.e., or ). Similarly for the UPFC cases (Cases columns 3 and 4), the largest values appear in the column, again associated with the real-power ow. This is as expected because real power plays a dominant role in sustaining or damping rotor oscillations and, hence, these variables are signicantly more effective in damping than those associated with reactive power; 2) in Cases 2 and 5, the IPFC can control at least two modes, while in Cases 3 and 4, the UPFC can only control one

mode. Again, this is as expected, because the IPFC is able to directly control the real power ow in two lines, as opposed to the UPFC which can only control the ow precisely in one line. This shows the IPFCs advantage over the UPFC for low frequency damping control; 3) as seen from Table IV, Mode 2 has very low damping (1%) . This mode and remains poorly damped in cases is associated with G2 oscillating against the innite bus (see Fig. 2), and controlling ows in 16 or 78 (as with cases 2 to 4) has a negligible effect on it. On the other hand, controlling the ow in line 12, as in Case 5, directly affects the power exchange between G2 and the innite bus and, as seen from Table IV, immediately changes the mode frequency to 0.23 Hz and introduces signicant damping (35%). VI. INTERPRETATION OF THE DAMPING CHARACTERISTICS In a conventional damping controller, such as a PSS on an electrical machine, changing the PSS gains usually retains the resonance frequencies near their original values but increases the damping. In contrast, when a UPFC or IPFC is included, the aforementioned results (Table IV) not only indicate improved damping, but also show a signicant change in the resonant frequencies (eigenvalues) of the network. This section attempts to interpret this seemingly anomalous behavior. The UPFC is used for the purpose of explanation, but the argument applies to the IPFC as well. A. Impact of Series Branch on Dynamic Performance The series branch is the main contributing part of the UPFC and IPFC in controlling power ow. In the UPFC, the series converter is operated to provide the ordered real and reactive power into the line. For power-ow control and congestion management, the series branch is given a constant real and reactive power setpoint, as in sample cases 25 before. In this operation, (and the reactive power ) in the incremental real power the line is zero. From a machine acceleration point of view, this is tantamount to having no incremental power transfer through the line. With this control, the system behaves (incrementally) as if the line were not present. The validity of this statement can be checked by modelling the series branch as shown in Fig. 11(b),

452

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 1, JANUARY 2011

Fig. 11. Equivalent disconnected circuit of the series branch of the UPFC.

TABLE VIII EIGENVALUE INFORMATION-WITH THE MODEL AS IN FIG. 11(b) (CASE 6)

which shows the disconnection. The current sources merely represent steady-state real and reactive power levels and, hence, do not contribute to incremental small-signal dynamics. To validate the aforementioned idea, a test case (Case 6) was created by modifying the UPFC controller case (Section IV, Case 3). Line 16 (UPFC series branch) was cut and replaced with an equivalent circuit with one constant current source at each side of the line [Fig. 11(b)]. The current sources do not contribute to the dynamics but only ensure that the steady-state power ow at either end is maintained. If the proposed theory discussed before is valid, then the oscillation frequency and damping ought to be nearly identical to that with the UPFC in line 16. The eigenvalue information for the critical modes of Case 6 is shown in Table VIII. In comparison with Case 3 (see Table IV), Case 6 (Table VIII) shows that the frequencies and damping of the modes are, in general, close (though not exactly identical). This validates the premise that using a series UPFC branch with , control is tantamount to disconnecting the systems previously connected by the transmission line 16. Although Case 3 and Case 6 modes are close, there is some mismatch, particularly for Mode 4, where Case 3 shows a mode frequency and damping of 0.58 Hz and 29%; whereas Case 6 shows 0.59 Hz and 14%. This is due to the fact that in Case 3, the UPFC keeps the real and reactive power constant by regulating the injected voltage, the corresponding line is not really cut. Since this regulation process is not perfect, the parameter setting of the UPFC have impacts on the network dynamics. Nevertheless, the mode frequencies are nearly identical and damping values are signicantly higher than that for the base Case 1. The disconnected line visualization is thus still generally valid. The aforementioned analysis explains the observed mode frequencies and dampings in Table IV. Now the altered mode frequencies and damping can be identied as being those of a new network structureone in which existing transmission lines have been cut. If the cut is introduced at a suitable location, the new structure can exhibit better damping. For example, segmenting the network by virtual removal of connections to

distant networks could alter low-frequency interarea modes, which are often difcult to control [19]. Note that the improved dynamic behavior is a result of the segmentation of the network rather than by the selection of suitable controller parameters. This approach does not require classical feedback controller tuning methods for designing the controller, but merely requires the selection of a suitable line to cut. Sometimes, exibility may not exist in selecting the cut location (e.g., when a primary objective, such as congestion management, restricts the location. Then, it is possible that the new structure may not be suitably damped. However, there is still the possibility of improving damping using a regulator to modulate the power order of the UPFC branch in the conventional way. A similar conclusion can be arrived at regarding the series branches of the IPFC. However, since the IPFC has two series branches, it has even more potential for cutting (i.e., segmenting) the network, in comparison to the UPFC. As seen from Table IV, Case 5, a suitably located IPFC is able to damp all critical modes in the test network. The effect of segmentation has the advantage of mitigating any propagation of disturbances in one area into the wider network. One possible disadvantage that comes with segmentation is that it may reduce the synchronization torque on the generator shafts, resulting in reduced transient rotor-angle stability margins. This could be addressed by having an auxiliary controller to improve the transient stability during disturbances [20]. HVDC links operated in constant power control mode also provide the same decoupling between the sending and receiving end systems. The aforementioned discussion indicates that the UPFC or the IPFC devices operating in constant power ow also have a similar effect. The MVA rating of the UPFC or IPFC converter branches is only a small fraction of the power transmitted between the subsystem [21]. On the other hand, the converters in an HVDC link must carry the full load power. B. Demonstration of the Cutting Effect of the UPFC Series Branch The aforementioned visualization of the UPFC or IPFC as an element which disconnects the system (with the series branch operated in power control mode), can be demonstrated in another way. The system in Case 3 (UPFC in line 16) is remodelled, but now includes an additional gain KG in the error paths as shown in Fig. 12. Note that if KG is zero, control outputs and and, hence, the injected voltage is a constant. This effectively means that for incremental small-signal purposes, the UPFC is not present, as the incremental injected voltage is zero. If KG is increased, the controller begins to regulate the real and reactive currents and maintain constant power in the corresponding branch. As discussed before, from a small-signal point of view, this is like disconnecting the line that contains the UPFC series branches. Fig. 13 shows the frequency of critical Mode 4 of the system for different values of KG. For very small KG, the mode frequency is 0.75 Hz, which is the same as that of the system without FACTS (Case 1). As KG is increased, the mode frequency gets closer to 0.59 Hz for the equivalent disconnected system of Case 6. This shows that the controller has an impact on the mode frequency; however, when the gain is large enough,

JIANG et al.: DAMPING PERFORMANCE ANALYSIS OF IPFC AND UPFC CONTROLLERS

453

Fig. 14. Magnitude limiter.

Fig. 12. Remodelled controller of the series branch of UPFC.

4) it is theoretically possible that poorly damped modes still exist in the changed network; if so, a suitable damping controller can be introduced that modulates the power references of the FACTS device; 5) since the IPFC has more series branches than the UPFC, it provides more opportunities for network segmentation and, hence, has the potential for greater damping improvement. APPENDIX Generator Model: Bus 9 is chosen as the innite bus and, thus, G1 is treated as an ideal voltage source. Three generators (G2, G3, and G4) with their exciters are represented by the typical fourth order dynamic model (third order generator plus rst order exciter). For the th generator and exciter

Fig. 13. Evolution of the mode frequency of Mode 4 during tuning the series branch of UPFC and IPFC.

(A1)

the line ow is essentially constant and the disconnected visualization becomes valid. As mentioned earlier, the disconnection concept is a useful approximate visualization, and so there will be some differences with the full FACTS device representation.

(A2) (A3) (A4) These equations can be linearized and expressed in state vector form combining all generators as (5) and (6). The derivation of the model is found in [19] (A5) (A6) where

VII. CONCLUSION This paper presents the small-signal model of the IPFC. The model is validated via EMT simulation using a 12-bus network capable of modelling multiple oscillatory modes. The damping performance of the IPFC is evaluated by using the validated IPFC small-signal model. The damping performance is also compared with that of a UPFC. The following results are obtained: 1) the effect of installing an IPFC or UPFC in constant power control mode for the series branch is similar to that of disconnecting the transmission line that contains the series branch; this resulting change in network structure introduces signicant changes in the corresponding mode frequencies as well as mode damping. 2) with proper selection of the location of the series branch, the resulting network can be made to exhibit improved damping behavior; 3) the improved dynamic performance is essentially caused by a virtual change of the network structure rather than by the tuning of controller parameters as is the case with most traditional approaches such as the PSS; hence, the feedback damping controller design can usually be avoided.

is the change in current injected by the th generator to is the change in voltage of the network the network and node to which the th generator is connected. IPFC Controller Parameters: The limiting of the magniand orders in Figs. 4 and 5 is achieved by rst tude of and to polar form (i.e., magnitude and phase), converting passing the magnitude through a limit and then reconverting to rectangular coordinates as shown in Fig. 14.

454

IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 1, JANUARY 2011

TABLE IX IPFC CONTROLLER PARAMETERS (SEE FIGS. 4 AND 5)

The parameters of the controllers shown in Figs. 4 and 5, including the limit, are shown in Table IX.

REFERENCES
[1] Y. H. Song and T. J. Allan, Flexible AC Transmission Systems (FACTS). London, U.K.: Inst. Elect. Eng. Press, 1999. [2] L. Gyugyi, Converter-based FACTS controllers, in Flexible AC Transmission Systems-The FACTS (Ref. No. 1998/500). Inst. Eng. Technol. Colloq., Nov. 1998, pp. 1/1111. [3] L. Gyugyi, A unied ow control concept for exible AC transmission systems, in Proc. Int. AC and DC Power Transmission Conf., Sep. 1720, 1991, pp. 1926. [4] N. G. Hingorani, Flexible AC transmission, IEEE Spectrum, vol. 30, no. 4, pp. 4045, Apr. 1993. [5] A. Nabavi-Niaki and M. R. Iravani, Steady state and dynamic models of unied power ow controller (UPFC) for power system studies, IEEE Trans. Power Syst., vol. 11, no. 4, pp. 19371943, Nov. 1996. [6] R. Mihalic, P. Zunko, and D. Povh, Improvement of transient stability using unied power ow controller, IEEE Trans. Power Del., vol. 11, no. 1, pp. 485492, Jan. 1996. [7] I. Papic, P. Zunko, D. Povh, and M. Weinhold, Basic control of unied power ow controller, IEEE Trans. Power Syst., vol. 12, no. 4, pp. 17341739, Nov. 1997. [8] H. F. Wang and F. J. Swift, A unied model for the analysis of FACTS devices in damping power system oscillations. I. Single-machine innite-bus power systems, IEEE Trans. Power Del., vol. 12, no. 2, pp. 94146, Apr. 1997. [9] Z. Huang, Y. Ni, C. M. Shen, F. F. Wu, S. Chen, and B. Zhang, Application of unied power ow controller in interconnected power systems-modeling, interface, control strategy, and case study, IEEE Trans. Power Syst., vol. 15, no. 2, pp. 817824, May 2000. [10] C. D. Schauder, L. Gyugyi, M. R. Lund, D. M. Hamai, T. R. Rietman, D. R. Torgerson, and A. Edris, Operation of the unied power ow controller (UPFC) under practical constraints, IEEE Trans. Power Del., vol. 13, no. 2, pp. 630639, Apr. 1998. [11] X. P. Zhang, Modelling of the interline power ow controller and the generalised unied power ow controller in newton power ow generation, transmission and distribution, Proc. Inst. Elect. Eng., vol. 150, no. 3, pp. 268274, May 2003. [12] Y. Zhang, Y. Zhang, and C. Chen, A novel power injection model of IPFC for power ow analysis inclusive of practical constraints, IEEE Trans, Power Syst., vol. 21, no. 4, pp. 15501556, Nov. 2006. [13] S. Arabi, H. Hamadanizadeh, and B. B. Fardanesh, Convertible static compensator performance studies on the NY state transmission system, IEEE Trans. Power Syst., vol. 17, no. 3, pp. 701706, Aug. 2002. [14] A. M. Parimi, I. Elamvazuthi, and N. Saad, Damping of inter area oscillations using interline power ow controller based damping controllers, in Proc.2nd IEEE Int. Power and Energy Conf., Dec. 13, 2008, pp. 6772.

[15] A. Kazemi and E. Karimi, The effect of interline power ow controller (IPFC) on damping inter-area oscillations in the interconnected power systems, in Proc. IEEE Int. Symp. Industrial Electronics,, Jul. 913, 2006, vol. 3, pp. 19111915. [16] X. Jiang, X. Fang, J. H. Chow, A.-A. Erids, and E. Uzunovic, Regulation and damping control design for interline power ow controllers, in Proc. IEEE Power Eng. Soc. General Meet., Jun. 2428, 2007, pp. 18. [17] Schauder and H. Mehta, Vector analysis and control of advanced static VAr compensators, Proc. Inst. Elect. Eng., Gen., Transm. Distrib. C, vol. 140, no. 4, pp. 299306, 1993. [18] S. Jiang, U. D. Annakkage, and A. M. Gole, A platform for validation of facts models, IEEE Trans. Power Del., vol. 21, no. 1, pp. 484491, Jan. 2006. [19] P. Kundur, Power System Stability and Control. New York: McGrawHill, 1994. [20] S. Limyingcharoen, U. D. Annakkage, and N. C. Pahalawaththa, Effects of unied power ow controllers on transient stability, Proc. Inst. Elect. Eng., Gen. Transm. Distrib., vol. 145, no. 2, pp. 182188, Mar. 1998. [21] A. S. Mehraban, A. Edris, C. D. Schauder, and J. H. Provanzana, Installation, commissioning, and operation of the worlds rst UPFC on the AEP system, in , Proc. Int. Conf. Power System Technology, 1998, vol. 1, pp. 323327. Shan Jiang (S03) received the B.Sc. and M.Sc. (Eng.) degrees in electrical engineering from Chongqing University, Chongqing, China, in 1989 and 1993, respectively, and is currently pursuing the Ph.D. degree at the University of Manitoba, Winnipeg, MB, Canada. Between 1992 and 2001, he was a Project Engineer with Dongfang Electrical Corp., Deyang, Sichuan, China, in charge of design and commissioning of generator control systems. His expertise lies in the area of exible ac transmission systems, power system control, and power systems simulation.

Ani M. Gole (S77M82SM04F10) received the B.Tech. degree in electrical engineering from the Indian Institute of Technology, Bombay, India, in 1978 and the Ph.D. degree from the University of Manitoba, Winnipeg, MB, Canada, in 1982. He is Distinguished Professor at the University of Manitoba, and a Registered Professional Engineer in the Province of Manitoba.

Udaya D. Annakkage (M95SM04) received the B.Sc. degree in electrical engineering from the University of Moratuwa, Sri Lanka, in 1982, and the M.Sc. and Ph.D. degrees from the University of Manchester, Manchester, U.K., in 1984 and 1987, respectively. He is a Professor at the University of Manitoba, Winnipeg, MB, Canada, and a Registered Professional Engineer in the Province of Manitoba. He is an editor of IEEE TRANSACTIONS ON POWER SYSTEMS.

D. A. Jacobson (S84M90SM05) received the B.Sc. (Hons.), M.Sc., and Ph.D. degrees in electrical engineering from the University of Manitoba, Winnipeg, MB, Canada, in 1988, 1990, and 2000, respectively. Currently, he is the Interconnections and Grid Supply Planning Engineer at Manitoba Hydro. He is a Registered Professional Engineer in the Province of Manitoba.