BESS Transformer RFQs: Evaluate Losses Against the Actual Duty Cycle

A BESS transformer bid should be evaluated using energized hours and the time integral of squared RMS current, including reactive-power operation. Annual exported MWh alone cannot distinguish a low-core-loss design from a low-load-loss design. Attach the same duty profile, reference temperature, auxiliary boundary and energy valuation to every request for quotation; then assess thermal feasibility separately from financial ranking.

1. Define the evaluation boundary

Count transformer losses on both charging and discharging intervals. Their direction does not reverse heat production. Standby is not necessarily zero current: voltage support can require substantial reactive current while battery active power is close to zero. Likewise, an open LV breaker does not remove core loss if the MV winding remains energized.

Eaton's loss guide supports the current-squared approximation. Use it for screening, with manufacturer loss components and temperatures for the final evaluation. This guide concerns comparing bids, while the four-quadrant BESS guide explains the operating envelope.

2. Supply a reproducible duty file

Specify timestamp, duration, terminal voltage, active power, reactive power, ambient temperature and cooling state. Calculate apparent power from simultaneous P and Q rather than independently adding their maxima. Identify which transformer winding carries each interval's current and include agreed overload events, scheduled outages and voltage variation.

For a sinusoidal balanced screening model, k is RMS current divided by rated current. At rated voltage it is also apparent power divided by rated apparent power. Harmonic current, DC components and differing tap voltages require further treatment; assigning every BESS transformer one fixed K-factor is not a substitute for its current spectrum.

3. Worked illustrative daily comparison

Assume a 5 MVA transformer remains energized for 24 h. Its daily loading is 4 h charging at k = 0.8, 4 h discharging at k = 0.8, 2 h reactive service at k = 0.4, and 14 h standby at k = 0. These are hypothetical currents, not product measurements.

Over 365 identical days, B saves 2336 kWh. At an illustrative constant loss-energy value of 0.12 EUR/kWh, that is 280.32 EUR/year before discounting. This small difference may not justify a large price premium. Actual dispatch and hourly energy values can change the result substantially.

4. Find the ranking crossover

A uses less core-loss energy, whereas B uses less load-loss energy. Their difference is 48 - 10Heq kWh/day, so the crossover is Heq = 4.8 h/day. Below that equivalent full-current duration A wins; above it B wins. Preserve this sensitivity in the commercial evaluation instead of publishing one efficiency percentage at rated load.

An annual model should price each interval's losses using the agreed commercial boundary. Battery revenue, purchased energy and metered export are different quantities. Do not double-count losses already included in a dispatch simulation or infer a universal tariff from this example.

5. Separate energy from thermal acceptance

The simplified comparison holds load-loss values at one agreed reference temperature and excludes cooling auxiliaries, harmonics and voltage dependence. It does not predict hot spots. Short peaks can matter thermally even when annual equivalent hours are low. Request the OEM's thermal calculation for the actual load chronology, ambient extremes and cooling failures.

IEC 60076-7:2018 addresses mineral-oil loading. For other insulation systems, obtain the applicable validated model rather than silently transferring that scope. Use the thermal modelling guide to frame the input discussion.

6. Procurement evidence and acceptance

The RFQ should state guaranteed P0 and Pk, their test/reference conditions, tolerance treatment, auxiliary consumption, evaluation file revision and the consequence of a measured deviation. Ask for winding loss and stray-loss treatment under the supplied spectrum. Tie the factory report to the purchased rating, tap and cooling design.

IEC 60076-1:2011 provides the general transformer framework; its public abstract does not establish this project's commercial penalties. Agree those contractually. Archive the guaranteed schedule, certified loss results, thermal assumptions and final financial workbook together so a later dispatch change can be reevaluated.

7. Frequently asked questions

Does charging count as negative loss? No. Current direction changes power flow, not resistive heating.

Can the transformer be switched off during standby? Possibly, but switching duty, restart time, auxiliary supplies and reactive-service obligations must first be resolved.

Is the cheapest annual-loss bid automatically acceptable? No. Thermal duty, insulation, fault withstand and interface requirements remain acceptance gates.

8. References

• Eaton: loss evaluation
• IEC: general requirements
• IEC: mineral-oil loading scope