An Electromagnetic Transient Simulation Model For
An Electromagnetic Transient Simulation Model For
1. INTRODUCTION
Voltage Sourced Converters (VSCs) have several advantages over conventional thyristor based converters.. These include: i) independent control of active and reactive power, ii) operation into a very weak ac systems, even one purely passive loads with no danger of commutation failure, iii) the ability to provide reactive power and voltage support, iv)the use of pulsswidth modulation to limit lower order harmonics resulting in a very limited need for harmonic filters. For this reason, several VSC-HVDC systems have been commissioned in recent years, including the 330 MW Cross Sound Cable PI. In addition to point to point bulk power transmission, . VSGHVDC has also been CCECE 2004 CCGEI 2004, Niagara Falls, May/mai 2004 0-7803-8253-6/04/$17.00 0 2004 IEEE
~
proposed for other applications [2] such as the connection of off-shore wind farms or oil drilling platforms into the mainland electrical network and for underground transmission or distribution systems within congested cities. It must be pointed out however, that because of limited ratings of gate-turn of today's semiconductors, the amount of power which could be sent by VSC-HVDC is presently lower than a conventional one and the losses are also higher. Also in the event of a fault on an overhead dc line, hecause of the anti-parallel diodes in the semiconductor switch, the VSC converters at both ends of an HVDC link have to be disconnected by opening the ac circuit breakers to enable the arc to be extinguished [ret]. The advantages of VSGHVDC can only be fully realized with a Foperly designed control system, which allows decoupled control of the real and reactive powers and the dc voltage. The paper describes a simple model of a VSC-HVDC system, schematically shown in Fig.1, which can be useful for the study of different control strategies. The model allows one to use different strengths for the sending and receiving end ac networks. Typical operating contingency scenarios are simulated in order to evaluate transient performance considering two different control methodologies..
- '1063 -
3. CONTROL OF VSC
Operationally, the VSC is similar to a very fast synchronous machine, in that the active and reactive power may be individually controlled within the capacity of the converter by adjusting the magnitude and phase of ac voltage behind the converter transformer reactance. The PWM scheme generates this ac voltage from the dc voltage through switching operations when fed the necessary reference waveform With a solitaty VSC bridge supplied from a fixed dc voltage source, it is possible to control the real and reactive powers to any value within the converter's ratings. However, the HVDC transmission scheme consists of two VSCs connected on the dc side. This adds the constraint that in the steadystate, the real power entering the VSC rectifier ("sending end") must equal (neglecting losses) the real power leaving the VSC inverter ("receiving end"). The control scheme used here allows independent specification of the real and reactive currents on the receiving end and only the reactive current on the sending end. The real current order at the sending end is set through a proportionalintegral controller that regulates the dc side voltage to its rated value. A straight forward method, referred to as "direct control" of achieving the control of real and reactive power is shown in F i g 2 Here, proportional-integral controllen act on the real and reactive power errors to generate the magnitude and phase respectively of the voltage references to the inner PWM controller.
Pref
The controller works on the principal that the real power flow is more affected by the phase angle and that the reactive power is more sensitive to the voltage magnitude. A phase locked loop tracks the phase of the ac voltage and provides the phase reference for the generation of the three phase reference voltages. Although the control system is simple, it has the drawback that a request for a change in real power causes a transient in the reactive power and vice versa. An improved method is that ofdecoupled control [4,5], in which the control system includes additional signal paths which ensure that a request foia change in the real (reactive) power is effected without any transient in the reactive (real) power. The method is shown schematically for the inverter in Fig. 3. The reference for real current is generated to control the active power and the reactive current is used to control the terminal voltage. The rectifier side controls are similar except that the real current reference is controlled by the dc capacitor voltage error, and the reactive current reference is zero for unity power factor.
Pref
w,ref
4. CASE STUDIES
i)-Direct vs. decoupled control Fig. 4, compares the performance of the two control options discussed above with respect to the real and reactive power transients .for a step change in the real power order. As is evident the direct control method manifests a significant transient in the reactive power when the real power is changed. No such kansient is observed for the decoupled control method. ii) Dependency on the Ac System Strength Studies were conducted using receiving end networks with two different strengths, the first a relatively strong system of short circuit ratio SCR = 2L75" and a much weaker system with SCR = 1.1L75'. The results for the active power, (rms) terminal voltage are shown in Fig. 5 . Although the weaker system shows a somewhat higher
Va
. .....
- 1064 -
overshoot, the performance in either case is very good. It should be noted that a convefitional H M C system would have significant difficulty in operating at SCR values less than 1.5. iii) Effect of Capacitor Size : Fig. 5, illustrates the systems dynamic behavior for two different dc bus capacitor values. The response to a step change in power for two values large: 667 pF (0.1 pu-s) and small 66.7 pF (0.01 pu-s). Here the pu values indicate the ratio of steadystate stored energy
(1/ 2 ) and the rated dc power. As can he seen, the performance with either value is good. ......... . . . . . . . -_.. - ... .... __ Enen d capacnor size --.. .......... .._.
cvz
c=o.Ipus
-0
i-
1
Fig. 4: Direct VS. decoupled control of P and Q
Fig. 6 : Effect of capacitor size on system behavior iv) Fault studies: The model can he used to demonstrate the effect of ac side faults on the performance of the dc transmission system. Fig. 7 shows the recovery of dc power and D voltage following a three phase to ground inverter side fault. The short circuit ratio of the ac system was 2.0. The fault duration is taken as 83.3 ms (five fundamental frequency cycles). The recovery to 90% power takes 35
- 1065 -
ms, which is much improved in comparison to that possible with a conventional HVDC converter feeding a comparable ac network. Specjal controls that are active during the fault include a valve current limit (to protect fiom damage to the IGBTs) and a voltage dependent power order limit that reduces the power order.
ACKNOWLEDGEMNETS
The authors gratefully acknowledge the support from Manitoba Hydro (Research Grants T261iT262); and the Natural Sciences and Engineering Research Council of Canada (Discovery Grant Program).
REFERENCES
1) N . Horle, K. Eriksson, A. Maeland, T. Nestli, Electrical supply for offshore installations made possible by use of VSC technology, CIGRd 2002 Conference, Paris, France, Aug 2002
2) G . Reed, R. Pape, M. Takeda, Advantages of Voltage Sourced Converter (VSC) Based Design Concepts for FACTS and HVDC-Link Applications, Proceedings of IEEE Power Engineering Society Transmission and Distribution Conference, July 2003.
3) M. Szechtman, T. Wess, C.V. Thio, First benchmark model for HVDC control studies, CIGRE WG 14.02. ElectraNo. 135, pp 54- 73, April 1991
5 . CONCLUSIONS
A transient simulation model for VSC based HVDC transmission system was presented. Sample studies fo : evaluating performance considering different controller types, different ac system -strengths and faults was investigated. The performance was seen to be adequate over all tested operating ranges. The model is capable of being adapted to a wider range of studies than those presented here.
Transformers: 350 MVA, 17W138 kV, Xc = 15%, Thevenin equivalents: Series R-L w i t h R4.210, L=8 1mH for SCR=2. T-line data (for each box shown in Fig. 1) are as follows: R=7 0, L=18.8 mH, C=51.12 pF for 50 h cable line (can he scaled for shorter/longer lines).
- 1066 -