4.7. MODELING OF THE SVC FOR POWER-SYSTEM STUDIES ‘The extent of modeling of the SVC and the power system is dependent on the nature of the power-system studies to be pertormed. In this section, the basic SVC modeling concepts involved in different power-system studies are presented. 4.7.1 Modeling for Load-Flow Studies ‘The SVC models in these studies should represent the fundamental frequency, steady-state, and balanced performance of the SVC [3], [11], [14], [25], (26). {t may be necessary to model the SVC in terms of its three individual phases when an unbalanced operation of the SVC is considered, such as during load compensation or voltage balancing. The features of conventional load-flow pro- grams are described in the text that follows. 4.7.1.1 SVC Operation Within the Control Range if the slope of the SVC is neglected, then the SVC is modeled as a PV bus, with P =O and V = Veer However, if the slope is considered (as in the analysis of weak ac systems), the same is modeled by connecting the high-voltage side of the SVC bus (0 @ fictitious auxiliary bus by means of a reactance equal to the slope expressed in per units on the SVC base. Such a model is shown in Fig. 4.26(a). ‘High-Voltage-Side Gey Bs (PQ) Xe High-Voltage-Side fre] Bus (PO) = tapes % == Medium-Volage “ = paniiary Bus (PV with ft —F— Bus (PV) remote control) ‘Svc svc fa) i) Figure 4.26 ‘The SVC models with slope representation using conventional power- flow PY buses: (a) without a coupfing transformer and (b) with a coupling transformer,MODELING OF THE SVC FOR POWER-SYSTEM STUDIES 135. PQNode (P_ 2° eye Bus Brin £08 1> hnax = jBmax fOr V< Vann Figure 4.27. The SVC model with slope for operation outside the control range. It may become necessary to model the coupling transformer should the SVC be connected to the tertiary winding. When the transformer is represented explicitly, the susceptance range of the SVC must be appropriately adjusted to represent the correct reactive-power rating as seen at the high-voltage bus [11]. The corresponding load-flow model is illustrated in Fig. 4.26(b). 4.7.1.2 SVC Operation Outside the Control Range The SVC is repre- sented as an appropriate shunt admittance, depending on which limit is violated (31. If Ive > Imax (the inductive-limit violation), then Qmnax B=Bye=— Vion (4.61) If V < Vinin (the capacitive-limit violation), then B= Bmw = - 20 (4.62) Vi where Qmax = the maximum inductive-reactive-power rating at Vsvc = Vinax Qin = the maximum capacitive-reactive-power rating at Vsve = Venin It may be noted that Qmax is a positive quantity, whereas Qin is a negative quantity, The SVC model in this case is shown in Fig. 4.27. Modeling of the SVC as a PV node with @ limits is not appropriate for representing the SVC under overload conditions. This characterizes an incorrect behavior outside the control range, as depicted in Fig. 4.28. In advanced load-flow programs, a comprehensive SVC model is used that models not only the slope in the controlled range but also the Q limits in the uncontrolled domain [3], [11].136 SVC CONTROL COMPONENTS AND MODELS PQ Node (P= 0 Q= 0) SVC Bus Voltage Veve == Normal SVC characteristic — Inappropriate representation of SVC as a PV inode with Q-Limit O=Onas, oo SVC Current, Igy Capacitive «+> Inductive Figure 4.28 Inappropriate SVC modeling outside the control range for load-flow stud- ies. 4.7.2 Modeling for Small- and Large-Disturbance Studies In these studies, only the positive-sequence behavior of the SVC-compensated system is modeled. The electromagnetic transients in the SVC (TSC, TCR) and the network can be neglected if the objective is to investigate the stability related to electromechanical oscillations. ‘The models corresponding to the different components of the control sys- tem have already been presented. Because the bandwidth of stability programs is limited, the thyristor firing control is not explicitly represented. In such a case, a suitable interface model is needed to convert the output signal Bree into a controlled network component connected to the SVC bus. This can be imple- mented using any of the two models shown in Fig. 4.29 and listed as follows: 1. The variable-susceptance model, in which the SVC current in response to the susceptance output is given by Isye = BretVsve (4.63) Isve, Ivo. « y =a Lo @ © Figure 4.29 Basic static var generator models for small- and large-disturbance studies: (a) the susceptance model and (b) the current-source model.MODELING OF THE SVC FOR POWER-SYSTEM STUDIES 137 2. ‘The controlied-current-source model, in which the SVC current is again given as Isyc = Bret Vsve (4.64) Although both models in the preceding list are equivalent, a significance difference exists in their implementation. In case of the variable-susceptance model, the system-admittance matrix B must be updated in case any changes occur in By, Whereas a constant B matrix can contiace to be used in case of dhe controffed-current-source model. 4.7.3 Modeling for Subsynchronous Resonance (SSR) Studies In subsynchronous resonance (SSR) studies [43], a wide bandwidth of elec- tromechanical frequencies is considered, so a need exists to mode? the network transients, 4s well as the thyristor-controlled and thyristor-switched elements. ‘Additional filtering in the measurement systems may be required to eliminate network-resonant frequencies close to fundamental for the satisfactory, stable operation of the SVC control system. These aspects are discussed in detail in Chapter 5. 4.7.4 Modeling for Electromagnetic-Transient Studies The requirements for modeling the SVC in a general electromagnetic-transient study [3}, {33}, 114}, (21}-(24] are representation of three phases; accuracy aver a wide frequeney range: representation of all system nonlinearities, as well as different controls, and protection functions; time-domain simulation; detailed modeling of the static var system components (e.g., reactors, capacitors, and transformer models—inckuding their saturation character istics); 6. representation of the GPG system and synchronizing system for studying the instant-to-instant behavior of SVCs; and 7, modeling of additional contro! and protection functions. yee yas 4.7.6 Modeling tor Harmonic-Performance Studies ‘The basic SVC model required for conducting harmonic-performance studies {2}, [3] comprises an ideal harmonic-current source (single-phase or 3-phase symmetrical) with a specified current spectrum, The magnitudes of the nth- order harmonic current, J,, are dependent on the SVC configuration and its138 SVC CONTROL COMPONENTS AND MODELS operating condition. A harmonic-voltage source mode} in series with a variable source admittance is proposed in ref. [27]. Harmonic-performance studies are based on certain assumptions, as follows: 2 a ‘The effect of multiple harmonic sources can be investigated by applying the superposition principle. The SVC harmonic generation can be modeled by positive-, negative-, and zeto-sequence harmonic sources. ‘The system can be represented by finear models at each harmonic fre- quency. _ The precise evaluation of harmonic distortion must have accurate load modeling.