7.1 Overview
Wind generation technology has matured over the past several decades into an economically viable and environmentally favorable source of energy. Today wind generation has become a significant portion of the generation mix in many countries around the world. This document has focused on describing the dynamic performance, behavior and modeling of this generation resource. In general, wind turbine generators tend to by quite different in both mechanical and electrical construction from traditional large thermal, nuclear and hydro power plants. A wind farm of comparable peak megawatt capacity to a large thermal power plant will consist of many tens to perhaps hundreds of wind turbine generators and span over many square kilometers of land or sea. Each wind turbine generator consists of the mechanical turbine, which typically has three rotor blades that can have a diameter in excess of 80 m, that is connected to a small generator through a slender shaft, often with a gear box in between. There are presently four major concepts for the actual generator:
• a conventional, constant speed, induction generator,
• a variable speed induction generator unit with a variable, external, rotor resistance, • a variable speed unit with a doubly-fed asynchronous generator, and
• a variable speed unit with a fully rated frequency converter connecting the generator to the electrical grid.
Each of these concepts, together with other emerging concepts such as the hydrodynamic gear drive train turbine, have been discussed and explained in detail in this document.
7.2
Performance, Control and Dynamics of Wind Farms
In the early years of wind turbine generator design, the units were mainly designed for application in distribution systems and as distributed resources. Thus, a typical requirement was for the wind turbine generators to disconnect from the system following a major system disturbance. Presently, most wind farms are of the tens to hundred megawatt range and are connected to major transmission systems. Thus, the expectation is for these generating units to help support the system during major disturbances. With the application of modern wind turbine generator technologies (and occasionally other supplemental devices such as static var compensators etc.) it is possible to build wind farms capable of riding through voltage transients caused by typical transmission system faults and disturbances and having adequate reactive reserves and automatic controls to provide voltage regulation at the point of interconnection. The fault ride through capability of some generator technologies has been described in detail in chapter 3 and some of the appendices.
Of course, the intermittent nature of the energy source (wind) is not controllable, thus this presently still constitutes the major challenge facing operating systems with large amounts of wind generation. Active power control systems have been proposed for wind generators that allow their contribution to frequency and/or tie-line regulation, but this is always at the expense of wasted wind power if no means of energy storage is available.
The exact amount of wind generation that may be incorporated into a system before the burden of operation becomes excessive (usually called maximum penetration of wind power) is highly system dependent, since it is affected by the weather patterns of the region, the type of installed generation capacity in the system, the available power transmission capacity of the system with its neighbors and the contractual obligations governing these
interconnections. The unique and unambiguous determination of such penetration limits is still an open question.
Much progress has been made, particularly with research and development in the science of wind generation forecasting but significant additional work remains in this area as well as considerations related to the potential of marrying wind generation with energy storage technologies that could help with active power regulation as mentioned above.
7.3 Modeling Recommendations
In steady-state power flow analysis, a wind farm should be modeled with at the very least a simplified, but full, representation of the collector system. That is, even if lumped into single components, each component of the wind farm should be modeled, namely:
1. the wind turbine generator,
2. any shunt compensation devices at the generator voltage level, or on the collector system and connection substation,
3. the generator step up transformer,
4. the collector system impedance (taking into account resistance, as well as reactance, and susceptance in the case of underground cables), and
5. the substation transformer stepping up the collector voltage to the transmission system voltage.
Proper modeling of all these components is important. Models that represent the entire wind farm as a single unity power factor generator at the transmission bus are inadequate and can grossly misrepresent the actual reactive consumption of the wind farm. Depending on the nature of the study and the electrical distance of the wind farm being modeled to the region of the electrical system under study, a hierarchy of possible models for the wind farm might be considered. From as simple as a single lumped model of all the wind turbine generators in the wind farm together with an equivalent representation of the collector system (modeling each component explicitly but as a lumped total equivalent, see Figure 6-4 in Chapter 6), to a representation consisting of multiple feeders within the collector system with equivalent groups of machines at the end of each feeder, to a detailed representation of the entire wind farm and collector system. Wind farms that utilize different wind turbine technologies will require at the very least a separate equivalent machine to represent each group of different technology wind turbine generators. In these cases one may lump turbine groups (of the same type) fed by individual feeders. Thus, a wind farm consisting of say fifty wind turbines might be represented by five feeders of groups of ten turbines.
The starting point for dynamic simulation is the steady-state power flow model developed for a wind farm. Again the amount of detail will depend on the study. For example, if one is studying localized issues within an electrical power system (e.g. local voltage stability), depending on the electrical network topology it may not be necessary to model in great detail wind farms that are quite remote to this region – such determination relies heavily on the nature of the problem and topology of the system under study. Conversely, for studies in the electrical vicinity of a wind farm (or group of wind farms) it may be desirable to lump fewer machines together and thus represent the wind farm in more detail because the wind farm may have machines scattered over several tens of kilometers and thus the voltage can vary enough throughout the wind farm to affect the number of units that trip on low voltage. This and other factors should be considered, and the level of detail adjusted to support the study objectives.
One final comment is pertinent with respect to power system studies. Most power system studies are presently performed in positive sequence stability programs. In these programs if one is studying system-wide phenomena such as power system oscillations (inter area modes of rotor oscillation), or frequency instability or slow voltage decay etc., then one is limited to
a bandwidth of up to roughly 2 Hz, because the network model is static and not adequate for studies that go outside this range. For example, one cannot study phenomenon such as subsynchronous resonance (SSR) with torsional modes, ferroresonance with electrical equipment etc. in such positive sequence stability programs. On the other hand, if one is looking at control interactions between controllers that are in the same substation or very close by and thus not affected by the network impedance (e.g. potential interactions between