Wind farms by the nature of their design and operating characteristics are susceptible to a variety of overvoltages.  Hence it is always important to conduct studies and tests of the various levels of overvoltages and how the equipment at the wind farm are able to withstand with or without mitigation measures.  In this Blog, we will provide an overview of the issues, the analytical approach and potential mitigation.  Then, we demonstrate how these are applied to a sample wind farm.

Types of Overvoltages

But before we go any further, let us begin with some definitions of different types of overvoltages:

• Transient overvoltage – short-duration highly damped, oscillatory or non-oscillatory overvoltage, having a duration of a few milliseconds or less.  Can be due to lightning, switching, and very fast front surges.  Equipment withstand capability are specified as Basic Lightning Impulse Insulation Level (BIL) and Basic Switching Impulse Insulation Level (BSL).
• Temporary overvoltage – Oscillatory phase-to-ground or phase-to-phase overvoltage of relatively long duration (seconds to minutes) and that is damped or weakly damped.  Due to switching or fault clearing operations, and includes events such as load rejection, and faults on high-resistance or ungrounded systems.  May also occur due to non-linearities such as harmonics and ferroresonance.
• Ground Fault overvoltage – a specific form of temporary overvoltage that occurs in three-phase ungrounded or high-impedance grounded systems.

Typical Study Items

A typical study of overvoltage for wind farms may include the following items:

• Energization – when the wind farm main transformer and collector system are energized from the external system.  Generally, this is done with the wind turbines offline, but may also be extended to include test conditions when individual turbines are energized from the grid.
• Load Rejection – when the wind farm is isolated from the grid due to a switching event, excluding those that occur from faults and automatic control response.  Used to confirm if the generator step up transformers have sufficient BSL capability and the cable/collector system and wind turbines have sufficient BIL/BSL.
• Faults (single line and three phase) – simulates short circuits at various locations in the substation and collector system, detected and cleared by protective relaying acting through circuit breakers.  The voltage impacts will cover both the transient and temporary overvoltage type results.  Used to confirm the same set of withstand capabilities as the load rejection tests with the addition of the main transformer.  In addition, the need and sizing of surge arresters are also determined in these tests.
• Collector ground faults – a specific form of single-line-to-ground fault applied to ungrounded or high-resistance grounded collector and turbine systems.  The key sequence tested is the opening of the collector feeders to clear the fault with the wind turbines remaining online until they trip or “spin-down”.  (This is discussed in detail in the Blog “Banking on grounding of solar and wind farms“, from March 2012.)

If any of the above tests indicate a potential impact from overvoltage, mitigation measures may include: adding or changing out lightning arresters, adding grounding banks, reducing spin-down times for the turbines, and adding direct transfer trips of the individual turbines before the collector breaker opens, among other options.

The above tests assume that there is no capacitor bank at the wind farm.  If there is a capacitor bank, then additional tests may be needed for switching of the bank.  Also, the tests above assume that insulation coordination has been verified and that the BIL and arrester ratings are compliant with the coordination design.

Sample Study

We applied the above concepts to a sample wind farm with the following characteristics: 24 MW wind farm interconnecting at 138 kV to a local utility using 12 turbine units rated 2 MW each, Type 3 (Doubly Fed Induction Generator or DFIG).  The collector system, including the location of surge arresters, is shown in Figure 1 below.

The system is modeled in the transient software PSCAD®/EMTDC (or simply “PSCAD”) developed by the Manitoba HVDC Research Centre, and the various tests are simulated.  The results for the energization and load rejection tests are summarized in Figure 2 below.

Although energization voltages are within the nominal ranges, load rejection shows some high transient overvoltages as shown in Figure 3 below.

However, these overvoltages are well within the withstand capabilities of the main transformer, collector cables and wind turbines even without surge arresters.  Likewise, overvoltages following faults are within withstand capabilities.

For ground fault overvoltages, the duration that overvoltages are present depends on the “spin-down” times of the turbines; i.e., the time that the turbines remain energized after the collector breakers have opened.  The capability of the surge arresters are critical to whether they will be able to withstand the resulting overvoltage.

The ground fault overvoltage tests and results are summarized in Figure 4 below.  The simulations were conducted with and without a grounding bank.  In the cases without a grounding bank the energy absorbed by the arrester exceed their capability for spin-down times greater than 12 cycles (0.2 sec).

Conclusion

A wind farm is subject of a variety of overvoltage as a consequence of its design and equipment.  The types of overvoltage may vary depending on how large the wind farm is and whether or not it is effectively grounded or uses capacitor banks.  A study system demonstrated the method for analysis using the PSCAD software and interpretation of results.  No transient or temporary overvoltage issues were found in the study system, however the need for either a grounding bank or a short spin-down time for the wind turbines were indicated.