Wind Farm Transformers are different!
Irfan Akhtar PE BE ME
63K Followers | Power & Energy | Renewables: Solar, BESS, Wind, Hydal | Thermal: IC Engines, Steam, CCPP, Coal | Green Hydrogen from Power Perspective
Wind power is being added to grids around the globe at a very fast pace these days. Since its first introduction for utility scale power production, sufficient operational and failure data is now available to suggest that Transformers used in Wind Farms require special Design, Procurement and O&M considerations. And they are different when to comes to issues faced by their distribution counterparts in utilities and industries.
Each wind turbine in a wind farm has a generator step up (GSU) transformer which raises turbine generator voltage of few hundreds volt to medium voltage level of collector system. Failure rate of these transformers has been very high compared to standard application distribution transformers resulting in sleepless nights for operators of wind farms and their staff. Initially Oil filled transformers were a norm and are still being used in the world but Dry type transformers are fast becoming popular.
IEC has already issued a new guide for Wind Transformers admitting these are DIFFERENT:
"IEC 60076-16 ed1.0 (2011-08) - Power transformers - Part 16: Transformers for wind turbine applications
This standard applies to dry-type and liquid-immersed transformers for rated power 100 kVA up to 10 000 kVA for wind turbine applications having a winding with highest voltage for equipment up to and including 36 kV and at least one winding operating at a voltage greater than 1,1 kV. "
Differences
Let us see how our Wind GSU Transformer is different and what can be done to increase its life and prevent costly failures which can affect financial modeling of these projects:
- Inconsistent Wind -It is a known fact that wind speeds vary around the year as well as in a day/night cycle. At times, wind speed goes below the cut off speed of the generator and it has to be taken offline. Many hours of the day, wind is not able to turn the generator. Even during evenings/night, it varies a lot. Transformer thus experiences lot of thermal cycling - no-load to load and then back to no-load. In case of oil filled transformers, gassing occurs. Each gas has different solubility in Oil but as transformer cools, most of the gases are released causing bubble formation. These bubbles can cause hot-spots and give rise to partial discharges, eventually causing deterioration of the insulation. For dry type transformers, thermal cycling causes fatigue in the winding assembly and insulation weakens.
-
Harmonics and Spikes - The step-up transformer used for wind power cannot be simply a standard distribution transformer. Such transformers have to be more robust. GSU transformer is fed from the output of an inverter. The voltage delivered by the inverter contains two important non-sinusoidal components, both of which are harmful to the transformer. Firstly, some spikes are present in the input voltage. The additional electrical stresses caused by these spikes (which contain a high-frequency component) would cause a rapid deterioration of the insulation system. Therefore the GSU must be suitably designed for these spikes. Secondly, when inverters are used, they produce a lot of harmonics of frequencies like 150, 300, 450 and up to 2000 Hz. These harmonics will cause increased eddy and stray losses within the windings as well as other metallic structures producing temperatures and stresses that are 2 to 3 times higher than normal 50 Hz power wave form. These extra losses will cause additional temperature rises and unless specially designed to handle these rises, it will give rise to loss of life, dissolved gases (DGA), partial discharges (PD) and premature failure. Thus, the GSU transformers have to be specially designed to mitigate these harmful effects of the harmonics. An electrostatic shield provided between the primary and the secondary windings creates capacitances between itself and the windings and acts like a flter to prevent the unwanted harmonics to pass through from the low voltage to the high voltage side (the collector bus).
-
Switching Operations - Another big concern is the transient over-voltages caused by switching. Due to the wide variations in wind speed and the consequent wide fluctuation in the power generated by the turbine, the unit can be switched off. In a given day, this can happen multiple times. The switching operation, particularly by vacuum circuit breakers, cause transient over-voltages due to the current chopping producing high di/dt in conjunction with the inductance of the transformer windings and the capacitances of the cables. Though the magnitudes of the switching over-voltages due to TRV are lower than the BIL of the transformer, such phenomena can trigger a large oscillatory voltage within the windings and can cause failures. This subject has been dealt with adequately in the IEEE standard C570.142.
-
Transient Over-voltages - In most wind farms the GSU transformers are daisy chained and accordingly their high voltage terminals are configured in loop-feed bushing arrangement. Since cable fault can happen in this type of arrangement, the transformer should be capable of “riding through” such faults, otherwise clearing such a fault would require disconnecting one group of turbines. If a single-line to ground or a double line to ground fault occurs, the voltage of that line is forced to ground potential and the other two phases face an overvoltage situation. This in turn results in overstressing of the insulation system within the transformer. Grounding transformers should be used in the high voltage side of the wind farm. Such grounding transformers provides a zero sequence impedance in the event of a single-line to ground fault, thus stabilizing the system voltage and thereby the over voltage on the GSU transformer is avoided. The grounding transformers can be either zig-zag connected, or a suitable delta-wye connected.
- Vibrations - For the transformers installed inside the nacelle, vibrations are another difference these transformers experience over their lifecycle. It goes without saying that vibrations can introduce mechanical stresses, causing looseness inside the transformers.
-
Transformer Sizing and Voltage Variation - Relying on standard distribution transformer design practices can lead to cases where over-voltages or over-currents could shorten the life of a wind turbine step-up transformer. Due to the economic reality of competitive pricing for renewable energy projects and because transformers are normally not operated at rated loading, there is little inclination to oversize the transformer. There may even be a tendency to underrate slightly and allow the transformers overload capability to carry it through. Since wind generators are not rated at standard distribution transformer ratings, trying to match it to a "standard" kVA rating becomes difficult. Typical wind generator ratings usually fall between standard distribution transformer ratings, so there seems to be a tendency to select the closest one, even if slightly undersized. The same is true for over-voltage capability. An over-voltage capability of 5 percent is normally required for the wind turbine step-up transformer to absorb any over-speeding of the generator. Transformer standards require a 10 percent continuous over-voltage capability. To prevent putting the core into saturation, the 5 percent over-voltage requirement should actually be added to the standard 10 percent capability.
- Gassing - Gases generated from oil decomposition include hydrogen, methane, ethane, ethylene and acetylene in increasing order of the energy required to produce the gas. Gases used to monitor paper decomposition include carbon dioxide and carbon monoxide. Since our concern is for the health of wind turbine step-up transformers, a quick look at the unique problems facing the step-up transformer, as related to the potential gases generated, is called for.
-
? Cyclical loading causes the gas to be absorbed into the fluid and then released from the fluid as the fluid heats and cools. Depending upon the rapidity of the temperature changes, the creation of bubbles may be such as to “short circuit” insulation clearances. Most likely these conditions would not be a complete insulation failure; but rather, a low level partial discharge occurrence. Most likely this would result in elevated levels of hydrogen gas in the oil.
-
? Harmonic content within the transformer, as noted earlier, can cause additional losses and heating within the transformer. If not properly addressed, this heating can lead to both paper and oil decomposition products. Since the heating is generated from the winding conductor outward, the predominate indication would be the appearance of carbon monoxide, thereby indicating thermal decomposition of the paper – also a possible contribution from oxidation of the oil with high CO2/CO ratios.
-
? Fault ride-through events produce fault level currents that will generate heating of the conductor. Without proper coordination of fault current levels and duration, excessive thermal conditions can exist within the transformer that will decompose the paper thereby producing higher than normal levels of carbon monoxide.
-
It has also been observed that Wind farm GSUs due to the reasons described above have higher than normal production of gases. As of now, there is no special guide to analyze DGA results for wind farm transformers. However, Doble, SDM, and certain utilities have come up with their revisions for gas limits to be taken as normal.
TRANSFORMER DESIGN CONSIDERATIONS
One can see the consistent theme here. Wind farm step-up transformers are often subjected to stresses and loading that departs from normal distribution transformer conditions. In the rush to cash-in on wind energy, developers are often trading low first costs for higher total costs of ownership to be shouldered later by the wind farm owners and operators. Now that the challenges of the application are more widely understood, more emphasis should be placed on specifying and designing transformers to meet the challenge. The correct application and protection of transformers for wind turbine step-up applications will take all these potential sources of gas generation into consideration.
WINDINGS Designing the coils with coolant ducts all around promotes smooth, laminar, convective flow. This shortens the path that the heat energy must travel to the coolant medium and eliminates the formation of hot-spots. Higher flow of the coolant, through the coils reduces the effects of thermo-cycling and better transports bubble formation away from the paper insulation.
CORE AND COIL GEOMETRY Transient voltages and system surges can expose the transformer to physical stresses, both chronic and catastrophic. Employing a round core leg shape and a round coil window will spread the radial forces in 360 degrees, maintaining shape and protecting the coil insulation.
CORE-FRAME AND END-BLOCKING The transformer can also be exposed to axial forces as a result of the same external events. Designing the core frame to contain these forces is important to the overall life of the transformer. Adding end-blocking without impacting coolant flow will eliminate telescoping of the coils and protect the integrity of the insulation.
TAP CHANGERS No Load Tap Changers can be affected, over time, by the rapid thermo-cycling present in a wind farm transformers duty cycle. Contact areas can be affected, creating a arcing point and gas generation. Consideration should be given to whether or not taps would ever need to be changed in this application.
All in all, it is pretty obvious that Wind Farm GSUs are not like their utility standard distribution transformers when it comes to their O&M, Life Cycle stresses and hence CBM requirements.
Condition Based Maintenance Needs
Some recommendations for Wind Farm transformer monitoring apart from going for Intrinsic Reliability (Reliability from Design):
- Oil sampling. Sample in one month when transformer is put in service. Initially sampling can be done 1-2 yearly basis to form a trend. Based on results for certain transformers, frequency can be adjusted. Watch out for increased levels of H2 and CO in particular.
- PD Survey - Due to formation of bubbles and moisture movement, partial discharges occur a lot. Hence it is a good idea to carry out PD surveys using RFI/TEV/HFCT/UHF or Acoustic sensors. Once every two years to yearly is fine.
- Temperatures should be monitored closely along with Max and Min readings. These would help track the insulation degradation.
- Power Quality audit to see if filters are working fine. A stitch in time saves nine.
- Furan analysis every 4 to 5 years can help judging the remaining life of paper.
- For dry type Transformers, Tan Delta Tipup, Excitation Current, TTR, WR, IR every 3 to 5 years would help assessing the condition. Same is recommended for Oil filled units.
- Thermography - Twice a year.
- Keep the cables in good condition as their failure causes lots of electrical stresses on the transformers. Cables can be tested for sheath continuity, PDs on MV side and Tan Delta.
- Last but not the least: Keep your transformer dry, clean and cool.
Bibliography:
1. IEEE R3 paper, Wind Farm Electrical Systems
2. Doble ICC 2013, Gassing in wind farm transformers
3. VA Transformer, Transformers for Wind Turbine Generators.
WTG Geek :)
2 年Grzegorz Swiezak that might interest you
Master of Science - MS at National Technical University "Kharkiv Polytechnic Institute"
3 年Thank you very much Mr. Irfan for your article, that make it easy for us to shortlist the issues with the many failed transformers at our wind farms. Your article is still useful even after 5 years of writing it. Thank you!
Gerente general en SEPROAE,S. de R.L., de C.V
7 年Thanks, to a very interesting read, i guess it′s a whole new ball game in O&M for substation engineers.
Electrical Engineer (EPIC & FM Maintenance). Looking for oppertunity in engg; BUSINESS or SERVICE in Middle east.
7 年Very informative, effort appreciated
Senior Asset & Maintenance Management Engineer II at New York Power Authority
7 年thanks Bruce for sharing. it is good and valuable information. but keep in mind that all transformer manufacturer know the application and behaviour of the Wind farm GSU and they are designing it accordingly. i am preparing specification for all type of transformers but never suggest/proposed winding type etc. for sure i am requesting for Grain oriented silicone sheet with vacuum type of tap changer and special type of insulation paper.