Source of information: North american Windpower
http://www.amsc.com/newsroom/documents/NAWindpowerJune2007.pdf
With rapid growth in the wind power industry, independent electricity system operators (IESO) must make sure that large-scale wind farms can be installed safely on the IESO-controlled grid and maintain an appropriate level of reliability. Detailed system studies are performed to evaluate the impacts of wind farms on the grid and to determine whether or not all of the IESO requirements have been met. In particular, the reactive power requirements deal with the ability of the wind farm to dynamically absorb and inject reactive power into the grid. This article identifies the IESO’s reactive power capability requirements and examines the possible compensation solutions that can be used to meet these requirements.
In Ontario, wind farms employing induction wind turbine generators or asynchronous generators must have the same reactive power capabilities as a conventional synchronous generator of the same active power size, as measured at the point of common coupling (PCC) to the IESO-controlled grid. The PCC is defined as the point where the grid connects to the high side of the main output transformer. Some of the detailed requirements include:
- at all levels of active power output, the minimum continuous reactive power required to be injected by the facility at the PCC is 0,35 per unit (pu) of rated active power;
- at all levels of active power output, the minimum continuous reactive power required to be absorbed by the facility at the PCC is 0,33 pu of rated active power;
- the reactive power capability of the facility must be fast, dynamic and continuous and must be able to hit both ends of the required voltamp reactive (VAR) output range in a time comparable to that of a conventional generator. There are two basic types of wind turbine generators used in wind farms – those that operate at a constant leading power factor (type I turbines) and those capable of injecting or absorbing reactive power (type II turbines). At the PCC, a 100 MW synchronous generator is required to have the capability to dynamically inject 35 megavars (MVAR) and absorb 33 MVAR at a constant system voltage for all active power outputs. A 100 MW wind farm must achieve these same reactive output levels. To meet this IESO requirement, wind farms employing type I turbines need additional dynamic reactive compensation equipment, such as STATCOMs, SVCs and the D-VAR reactive compensation solution.
Wind farms employing type II turbines may not require any dynamic reactive compensation equipment if the dynamic reactive power range of the turbine is sufficient and fast enough. In either case, wind farms may need to install static reactive compensation equipment to meet the reactive power requirements at the PCC. The next two cases will show how IESO’s reactive requirements can be met utilizing different turbine types and dynamic and/or static reactive power compensation equipment. Type I wind turbines In modeling a wind farm for system studies, the collector system was simplified by aggregating all of the wind turbines onto two feeders, as illustrated in the one-line diagram, on the next page, for a 100 MW wind farm that uses type I wind turbines. The effective impedance of each feeder was chosen so that the simplified model has the same active and reactive power losses as the actual wind farm. The value of these impedances can be determined by using analytical methods or by carrying out load flow simulations. For this wind farm, a D-VAR system, consisting of two STATCOMs and impedance-based reactive devices (capacitors and reactors), was used to meet the IESO’s dynamic reactive compensation requirements.
To check that this wind farm meets the IESO’s reactive power re quirements, three loadflow simulations were completed. (See “Load Flow Simulations For Type I Wind Turbines.”) The first simulation checked that the wind farm could inject 35 MVAR at the PCC, at a system voltage of 242 kV because this is a typical voltage level on Ontario’s 230 kV transmission system. This simulation is usually carried out at fullrated MW output to ensure that it can meet this 35 MVAR injection requirement at all power output levels. The second simulation checked that the wind farm could absorb 33 MVAR at 242 kV. This simulation is usually carried out with the wind farm near zero megawatts because this level ensures that it can meet this 33 MVAR absorption requirement at all power output levels. Since the devices used in D-VAR systems are voltage-dependent, it is important to check that the wind farm can still provide the required reactive power at low system voltages. The third simulation, therefore, ensured that the wind farm was capable of injecting 35 MVAR at a system voltage of 220 kV, which is the lowest continuous operating system voltage. The flows given in the Low Flow Simulations table are in MVA, and the directions are as shown in the type I wind turbine diagram. The flows from the STATCOMs, shunt capacitors/reactors and feeders are measured at the collector bus. The flow out of the system is measured at the PCC, and the flow out of the turbines is measured at the terminals of the turbines. The simulation data show that this wind farm was capable of injecting and absorbing the required reactive power at 242 kV. However, the wind farm could not inject the required reactive power at low system voltages. To solve this problem, slightly larger shunt capacitors would be required.
Type II wind turbines The model for a different wind farm was simplified by aggregating all of the turbines on one feeder as illustrated in the oneline diagram of the sample wind farm for type II wind turbines. The same three loadflow simulations that were completed for the first wind farm were also completed for this wind farm. (See “Load Flow Simulations For Type II Wind Turbines,” next page.) These results show that this wind farm meets the reactive power requirements of the IESO. The flows given in this table are in MVA, and the directions are as shown in the type II wind turbine diagram. The flow through the capacitors and feeder is measured at the collector bus. The flow out of the system is measured at the PCC, which is at the high side of the main output transformer, and the flow out of the turbine is measured at the terminal of the machine. To ensure that the wind farm has sufficient dynamic reactive margin to respond quickly to a contingency on the power system, both capacitors normally would be in service. This wind farm can inject reactive power rapidly when required. If a contingency occurs that requires the wind farm to trip the normally inservice capacitors, the response time will be approximately one second, which is comparable to the MVAR ramping of a synchronous generator.
As the type I and type II wind turbine cases show, the reactive compensation requirements and solutions can vary immensely depending on the type of turbine being used. Type II turbines can meet the require-ments with static capacitors and reactors, while type I turbines need a dynamic solution like the D-VAR system. The IESO’s approach for reactive power compensation is that the wind farm should have the same reactive power capability as a conventional synchronous generator of the same MW size. This translates into the reactive power requirements stated previously.