By Dr. Bartosz Wojszczyk, Quanta Technology; David Herbst, Realtime Utility Engineers Inc.; and Mitchell Bradt, University of Wisconsin-Madison
During the last 30 years, the size of the largest commercial wind turbines has increased from around 30 kW to as big as 5 MW. The working principle of a wind turbine is based on two well-known processes. The first involves the conversion of kinetic energy from the power of air moving to mechanical energy. This is accomplished by using different size turbine rotor blades with diameters up to 360 feet. The mechanical output power of the wind turbine is directly proportional to the swept area covered by the blades. The second process includes mechanical energy conversion through a generator into electricity that is transmitted to the electric grid. Wind turbines can be classified by their mechanical power control and speed control.
The 150 MW Shiloh wind farm in the Montezuma Hills of Solano County, Calif. Photo courtesy PPM Energy.
Wind turbines can also be classified as either stall regulated or pitch regulated. Stall regulation is achieved by proper geometries of the turbine blades such that the airfoil generates less aerodynamic force at high speed, thus reducing the turbine’s torque. This is a simple, inexpensive and robust mechanical system. Pitch regulation uses pitching motors in the turbine hub that twist the blades around their own axes. As the wind speed changes, the blade quickly pitches to the optimum angle to control torque in order to capture the maximum energy or self-protect as needed. Some turbines are able to pitch each blade independently to achieve balanced torques.
So far as controlling the speed that blades travel, turbines can be fixed speed (Type A); limited variable speed (Type B); or variable speed with either partial (Type C) or full (Type D) power electronic conversion (Table 1).
Impact on Power Grid Operation
Even though wind generation technology has significantly improved through the years, it continues to present important system design, control and operation challenges. If not properly addressed, these may lead to or contribute to significant system disturbances, like the one that occurred in Europe on Nov. 4, 2006.
On that November night, a portion of Germany’s power interconnection was affected by a serious incident that originated in the North German transmission grid and disrupted power supply for more than 15 million households. The event split a synchronously connected network into three islands: two of them under-frequency and one of them over-frequency. As a result of cascading overloads and system tripping, two of three large separated systems, Western Island and North-Eastern Island, ended up with a significant amount of wind generation resources.
Western Island was in an under-frequency state. During the incident, about 40 percent of the wind power units tripped. Moreover, 60 percent of the wind power units connected to the grid tripped just after the frequency drop. Wind units were automatically reconnected to the grid when voltage and frequency conditions were restored to the acceptable range. However, this occurred without control from control centers.
North-Eastern Island was in an over-frequency state. The imbalance in this subsystem caused rapid frequency increase and triggered the necessary primary, standard and emergency control actions of tripping wind generation units sensitive to high frequency values. Tripping these units, which amounted to about 6,200 MW, helped decrease the frequency value during the first few seconds of the disturbance. At about the same time, wind turbines were automatically reconnected without control from the control center. This was contrary to the required decrease of generation on the whole island and contributed to the deterioration of system conditions in this area (over-frequency that lasted a long time). Based on the sequence of events from both regions, it can be concluded that a significant amount of wind generation without controlled reconnection influenced system operation/control during the restoration phase of the disturbance.
Wind resources are often located in remote areas, far from existing utility facilities, and display widely varying intensity. These factors make it a highly fluctuating energy resource that can experience major power swings of up to 25 percent. About 10 percent of the time, wind may produce an hourly output from 5 to 20 percent of capacity. Such variability may affect power systems on local and global scales.
Local and Global Impacts
Local impacts can occur inside a wind generation facility or affect the grid connected to it. This impact is attributed to the “abnormal” work conditions within the footprint of the generation facility and connected adjacent grid and does not depend on the overall wind power penetration of the grid system. Interconnecting wind generation to the utility infrastructure must be designed in such a way that ensures system protection and automatic reclosing schemes will continue to operate without jeopardizing safety and grid reliability. These include thermal limits, frequency control, low voltage ride-through requirements and other critical considerations. It is therefore necessary to select appropriate protection schemes for interconnection between independent power producers (IPPs) and utilities, based on the circumstances of each wind generation deployment. Examples of the most common protection schemes used between IPP facilities and utility networks include:
- Instantaneous and time delay over-current relays (50/51, 50N/51N) (relatively small MW wind generation connected to radial network)
- Distance protection (21) with Direct Transfer Trip (DTT) and Directional Comparison
- Distance protection (21) with directional unit (67)
- Ground time delay over-current (51N) and under-voltage (27) relays (back-up protection, connected on the high-side of the step-up transformer)
- Differential relay (87) (used on the step-up transformer as entrance protection for faults fed from the utility)
Global impact affects the behavior of a larger system beyond the boundaries of a wind generation facility. It is strongly related to wind energy penetration of the wide-area power system, such as significant contribution to the load dispatch. It can be said that the impact of wind power becomes more pronounced the higher the wind penetration level is. Massive wind deployment can significantly affect grid dynamics, its operation and voltage stability. Wind generation also offers challenges to the global aspect of frequency control and load flow due to the fact that wind is uncontrollable. Wind generation has a limited contribution to primary frequency regulation. The higher the penetration of wind generation, the more requirement constraints needed for remaining conventional power plants to keep frequency fluctuation within predetermined limits.
Wind Generation System Protection
A typical GSU transformer connected to a traditional generator has a grounded wye/delta configuration with the delta side connected to the generator. The transformer is a ground current source for the transmission bus and isolates the flow of the zero-sequence current across the transformer. In typical applications (for example, fossil generation plants), generator bus exposure to faults is relatively small. It is also known that most operating procedures limit the ability of the delta bus of the transformer to be energized with an ungrounded neutral point (configuration).
On the other hand, the exposure of the delta winding of the transformer to long distances in large wind generation applications can be significant. In some cases, it can exceed 20 miles. A single-phase-to-ground fault on the ungrounded system will result in increased voltage on the unfaulted phases. The large distance exposure requires the use of zig-zag or wye/delta grounding transformers to prevent damage to cables, arrestors, transformers and other equipment. Figure 1 shows a typical grounding configuration on a wind generation bus. The applied grounding transformers provide a ground current source to the delta system. This grounding keeps the unfaulted phase voltages at an acceptable level during single-phase-to-ground faults and allows the application of ground over-current relaying.
Wye/delta grounding transformers can be connected directly to the bus/line or may be protected by fuses (Figure 1). When protected with fuses, it is important to coordinate the fuse minimum melt-time and maximum clearing time curves with the line or bus protective relay curves. It is necessary for system faults to be cleared by relay and breaker action before the grounding transformer fuse operates. A maintenance check of a wye/delta grounding transformer secondary voltage should indicate the “health” of the grounding source. Operating sequences also need to be considered in the proper placing of grounding transformers (Figure 2). A grounding transformer placed on the bus side (Fig. 2 location 1) of the collector station introduces a problem with the line losing ground reference during fault conditions on a line. This can be avoided by placing a grounding transformer on the line side (Fig. 2 location 2) of the circuit breaker. In some cases, grounding transformers may be required on the bus and line side of a collector substation.
Another issue is the use of non-standard transformer configurations and their effect on applied protection schemes. The rapid growth of the wind energy market (like other renewable energy sources), presents problems not only in meeting the high demand for wind turbines, but also wth the long lead times required for the delivery of large power transformers. In many cases, transformers used by developers are based on what is available at the time of construction in order to meet turn-key project deadlines.
Case 1: Wye/Wye Autotransformer. Figure 3 shows an example of a 69/34/12 kV autotransformer used in a small wind generation application. This transformer has very low impedance (Z=4.25 percent), which complicates ground over-current protection coordination between the utility and the wind generation substation. Utility side protection includes phase distance (21) and a ground directional over-current (67N) relay. The high side of the transformer is protected by phase and ground instantaneous and time delay over-current relays (50/51, 50N/51N) and the low side of the transformer is protected by phase and ground time delay over-current units (51/51N).
A grounding transformer was also required on the line side of the substation even though the 34.5 kV bus was grounded. This grounding transformer was needed due to possible line fault conditions. During line fault conditions, the 34 kV line could become ungrounded if the line breaker trips before the wind turbines become disconnected from the system (the wind turbine transformers in this case were connected delta on the 34.5 kV side). A detailed model of the circuit that includes the wind turbine’s short circuit capability curves, relays settings and its coordination evaluation, and unfaulted phase voltage profiles may indicate that a grounding transformer is not needed. For this project however, wind generation interconnection studies were not completed before substation engineering began.
Case 2: Delta/Wye Standard Distribution Transformer. A delta/grounded wye transformer is a commonly used configuration by utilities to serve distribution load. Figure 4 shows this configuration connected to a large array of wind generation. Grounding transformers are not required in the arrangement since the wind turbine transformers are connected “wye” on the 34.5 kV bus side and the 34.5 kV network will not lose its ground reference point.
The high side of the network (161 kV) in this application is not a ground current source. The ground relay (at breaker B) looking away from the substation into the network will not see any zero-sequence current for a transmission line fault (F). In this case, the primary protection was Directional Comparison Unblocking over fiber scheme (21P/67NP) and back-up protection was distance relay (21, zones 1, 2) with a directional ground unit (67NI&T). This line protection scheme required a transfer trip signal to be sent any time the remote breaker (A) opened. Once the breaker (A) at the remote end of the line trips, the ground reference at breaker (B) is lost and the voltages on the unfaulted phases will rise. Transformer arrestors with a higher MCOV rating are also required. A ground fault on the 161 kV line can also be sensed at breaker (B) by wiring a PT secondary winding in an open delta configuration. A traditional over-voltage relay (59N) with a short time delay is not used since the transfer trip scheme over fiber provides faster tripping for a breaker (B).
Case 3: Different Transformer Phasing and Load-Tap Changing (LTC) Application. The transformer winding configuration and phase displacement between the low and high side of the transformer also need to be taken into consideration in order to properly set transformer differential relays (87T). Two wind generation projects in the same state were constructed about 18 months part. Foreign transformers (both a grounded wye/delta configuration) from the same factory were delivered. Transformer 1 showed the ANSI standard of the low side lagging the high side by 30 degrees. Transformer 2 showed the low side leading the high side by 30 degrees. This was initially missed when settings for the transformer 2 differential protection were calculated. Based on this experience, it can be concluded that a thorough review of the transformer nameplate is always prudent.
In addition, large amounts of wind generation are being required to meet increasingly tight voltage schedules. Transformer 1 was ordered with a 32-step load-tap changing (LTC) on the high side. The LTC controller senses voltage from the PT located on the low side (34.5 kV) of the transformer, while the LTC itself is physically located on the high side of the transformer. The LTC becomes impractical on the low side due to its high current values (5000A). Initially, the LTC controller was programmed with a voltage set point, certain bandwidth and delay time. A voltage sag on the low side of the transformer resulted in raising the LTC operation on the high side. This operation drove the low side voltage lower. Since LTC controls have an inverted mode for this situation, the inverted mode was enabled and the LTC operated as expected.
Interconnecting wind generation with the utility grid must be designed in such a way that ensures system protection and automatic reclosing schemes will operate correctly under “abnormal” grid conditions. During the protection scheme design/selection process, engineers should consider the characteristic performance of the actual wind turbine used on the project and the specifics of the wind generation station configuration. This includes the selection of power and grounding transformers. This is necessary because fault levels within wind generation collector stations and possible power system contributions to fault levels will differ between different wind turbine solutions and wind collector station configurations.
To ensure correct wind generation collector station designs, comprehensive interconnection studies must be conducted. This includes detailed circuit models with wind turbine short circuit capability curves, relays settings and its coordination evaluation and unfaulted phase voltage profile analysis. In addition, reclosure schemes must be enhanced to allow wind generation to re-connect to the system in a controlled manner based on the specific system conditions.
Authors: Dr. Bartosz Wojszczyk is senior director for Quanta Technology and a technology and business expert on renewable technologies, power system protection, automation, distributed generation, energy storage, “smart grid” and advanced technology applications. He has authored/co-authored 20 technical papers and has one U.S. patent.
David Herbst, P.E., has been a utility system protection and field commissioning engineer for 25 years and has substantial experience in fault/system event analysis and field troubleshooting. He has been involved with wind generation design and commissioning since 1999, is a licensed professional engineer and a graduate of the University of Wisconsin-Madison.
Mitch Bradt, P.E., develops and delivers continuing education for engineers on alternative energy, electric distribution, power electronics and rotating machinery at the University of Wisconsin-Madison. He has been a consulting engineer designing substations and wind farms and a commissioning engineer for STATCOM and SMES equipment.
Example of a Relay Communication Scheme
Figure 5 (below) shows a small wind generation connected to a utility substation through a radial network. A fiber optic cable was buried with the phase conductors between the substations for relay communication purposes. Initially, a direction comparison unblocking (DCUB) scheme using a distance (21P) and directional ground (67NP) relay was proposed. The main operational principle of the DCUB scheme is based on sending an unblocking signal (tripping signal) to the relay at the remote end of the protected line during fault conditions inside the protected zone. At this point, note that wind generation will not always supply sufficient fault current values due to varying wind conditions impacting wind turbine operation (for example, during low wind speed). This situation may affect sensitivity and speed of installed relays. The lack of an unblocking command received at the utility end of the line (remote end of the line) could result in a delayed trip of the breaker for line fault conditions. In this case, this was not acceptable to the utility. A combination of a DCUB and a directional comparison blocking (DCB) scheme was used to ensure proper high speed tripping from the utility source. A transfer trip scheme was also used that allows a trip signal to be sent any time the utility breaker opens.