The lowest purchase price is not always the lowest transformer cost. A transformer may remain energized for many years, accumulating core-related loss every hour and current-dependent loss while supplying load. A useful comparison combines verified loss values with operating profile, energy value, maintenance and the project's economic horizon.
1. Understand the loss components
No-load loss
No-load loss, P0, is mainly associated with energizing the magnetic core. It exists whenever the transformer is energized, even at low downstream demand.
P0 (kW) × energized hours per yearLoad loss
Load loss, Pk, is stated at rated current and a defined reference temperature. It includes winding I²R and additional stray components. A useful preliminary estimate at load fraction x is:
x² × PkActual loss also depends on winding temperature, harmonics and the guarantee/test basis.
Auxiliary and harmonic losses
Add fan, pump, heater and control energy where their duty is material. Non-sinusoidal current from UPS systems, drives and rectifiers can increase winding and stray loss; provide the current harmonic spectrum for thermal review.
2. Use the squared load profile
Average kVA alone can be misleading because load loss varies approximately with current squared. Operating at 100% load for half the time and zero for half gives an average load of 50%, but an average squared load fraction of 0.50. Constant 50% loading gives 0.25.
Annual loss energy ≈ Σ [P0 + xi²Pk + Paux,i] × hiHere xi is per-unit load and hi is hours in the interval. Use interval meter data or an equivalent RMS load factor where available.
3. Worked illustrative comparison
| Item | Offer A | Offer B |
|---|---|---|
| No-load loss | 1.00 kW | 0.75 kW |
| Load loss at rated current | 8.00 kW | 7.20 kW |
| Price difference | Baseline | USD 4,000 premium |
Assume both units are energized 8,760 hours/year, equivalent RMS load is 0.55, auxiliary energy is equal, and energy is valued at USD 0.12/kWh.
8,760 × [1.00 + (0.55² × 8.00)] ≈ 29,960 kWh/yearOffer B8,760 × [0.75 + (0.55² × 7.20)] ≈ 25,650 kWh/yearIllustrative difference≈ 4,310 kWh/year ≈ USD 517/year
Simple premium payback ≈ USD 4,000 / USD 517 = 7.7 yearsThis is not an ALBORGRID performance claim. It excludes temperature effects, harmonics, discount rate, escalation, tax, maintenance, outages and residual value.
4. Build the lifecycle model
Equipment, freight, duties, civil work, cabling, commissioning and spares.
No-load, interval-based load loss and auxiliary energy, including planned load growth.
Inspection, oil or cooling-system tasks, planned outages, critical spares and interruption exposure.
Decommissioning, oil handling where applicable, recycling and residual value.
PV = future cost / (1 + discount rate)^nState how energy-price escalation and the discount rate are handled; otherwise two models may appear comparable while using different economic assumptions.
5. Match loss emphasis to duty
Continuously energized, lightly loaded duty
No-load loss may dominate annual loss energy. A lower-P0 design, including an appropriate amorphous-core option, may deserve evaluation—but only with actual guaranteed values, price, acoustic requirements and the operating profile.
Sustained high loading
Load loss becomes more influential as current rises. Industrial, infrastructure and generation profiles should be evaluated across daily and seasonal intervals, not at one assumed average.
Oversizing
A larger transformer operates at lower per-unit current but may have greater absolute no-load loss and higher capital cost. Evaluate candidate ratings, growth timing, staging, redundancy and fault-current implications. Lifecycle cost must not be used to justify undersizing.
6. Align every comparison input
| Basis | Check |
|---|---|
| Rating | Same kVA/MVA, HV/LV ratio, frequency and tap condition. |
| Connection | Same system-compatible vector group and neutral arrangement. |
| Construction | Comparable oil-immersed or dry-type design and cooling state. |
| Impedance | Same or system-compatible percentage impedance. |
| Temperature | Same reference-temperature basis for load loss. |
| Standard | Same test, guarantee and tolerance framework. |
| Scope | Comparable enclosure, accessories and auxiliary cooling. |
Determine whether each value is a catalogue typical, a maximum guarantee, a measured FAT value or a corrected result. Use the approved quotation guarantee for commercial evaluation and the serial-numbered report for manufacturing verification.
7. Do not calculate from a series label
The current catalogue groups certain 10 kV S13, S20 and S22 designs against loss-performance levels under the cited GB 20052-2024 framework. That mapping should not be generalized to every market, voltage or manufacturer.
- Specify the applicable destination-market efficiency requirement.
- Request guaranteed P0 and Pk for the exact rating.
- Confirm the standard, edition and tolerance basis.
- Do not treat a higher series number as proof of compliance with another market's regulation.
8. Test the assumptions
Run low-, expected- and high-load cases. Vary electricity value, energized hours, load growth, discount rate, evaluation period and cooling duty. If a purchasing decision changes with a small assumption adjustment, describe it as sensitive instead of presenting one exact payback as certain.
Data needed for comparison
First confirm technical compatibility. Then apply the same operating profile and financial assumptions to each supplier's guaranteed loss schedule.