0102030405
Power Transformer Efficiency Classes Explained: S11 vs S13 vs S22 Loss Levels Under IEC 60076-20
2026-08-11
TL;DR.Power Transformer efficiency classes have evolved fromS9 → S11 → S13 → S20 → S22 as China's GB 20052, EU Regulation 548/2014, and IEC 60076-20 progressively tightened loss limits. For a 1000 kVA 10 kV oil-immersed unit, typical no-load loss values are: S11 ~1.15 kW / S13 ~1.05 kW / S20 ~0.95 kW / S22 ~0.85 kW. The Tier 1 (1 July 2015) and Tier 2 (1 July 2021) stages of EU 548/2014 are the global reference points — Tier 1 corresponds roughly to S11/S13 and Tier 2 to S13/S20. The lifecycle cost decision favors S20/S22 for transformers running at 60-80% load 24/7; S11/S13 is adequate for lightly loaded or short-duty applications.
1. Why Efficiency Classes Matter Under EU 548/2014 and GB 20052
Power transformer efficiency is not a single number — it is a class system that has been progressively tightened by international regulation over the past 15 years. The Chinese GB 20052 standard, the EU Regulation 548/2014 (implementing Directive 2009/125/EC), and the international IEC 60076-20 framework all converge on a similar set of loss limits, but the class naming differs across jurisdictions and the cutoff dates for compliance vary.
For a B2B buyer specifying a transformer for an EU or international tender, the practical implication is that the minimum acceptable efficiency class depends on the destination market. Since 1 July 2021, only Tier 2 (S13-equivalent or higher) is permitted for placement on the EU market under EU Regulation 548/2014. For in-China installations, GB 20052-2024 raised the mandatory minimum to S20for Distribution Transformers andS22 for medium-power applications, effective for new installations from 2025. For other international tenders, IEC 60076-20 Level 1 or Level 2 is the typical reference.
The Tianan power transformer lineup offers S11 through S22 efficiency classes across 10-35 kV three-phase oil-immersed designs, and the 66-110 kV low-loss low-noise oil-immersed transformerline covers the S20/S22 segment for Transmission Substations. The loss-guarantee process is documented at theTianan audit process page.
The S-class naming convention originated in Chinese national standards as a shorthand for "energy-saving transformer." The class number reflects the relative loss reduction compared to the original baseline: S7 (1970s-1980s baseline) → S9 (1990s, ~10% reduction) → S11 (2000s-2010s, ~20% reduction; GB 20052-2020 minimum) → S13 (2010s, ~30% reduction; GB 20052-2024 medium-voltage minimum) → S20 (2020s, ~45% reduction; GB 20052-2024 distribution minimum and EU Tier 2 equivalent) → S22 (2020s, ~50% reduction; international high-efficiency benchmark). The progression is driven primarily by improvements in core steel — from conventional CRGO at 0.30 mm, to high-grade CRGO at 0.23-0.20 mm, to laser-processed domain-refined steel, and ultimately to amorphous alloy ribbon — with each step reducing no-load loss by 10-20%.
2. IEC 60076-20 Energy Performance Framework
IEC 60076-20 (Technical Specification, Edition 1.0, 2017) is the international framework for energy performance of power transformers. It defines three alternative methods for declaring efficiency compliance, plus two energy performance levels.
Method A: Minimum Peak Efficiency Index. The transformer must achieve a calculated peak efficiency index based on no-load loss, load loss, and reference capitalisation values for the losses. This is the most flexible method and accounts for the actual cost of losses in the buyer's network.
Method B: Maximum Load Loss and Maximum No-Load Loss. The simpler method — the transformer must meet maximum allowable values for both P₀ and Pₖ at each kVA rating. This is the method used by EU 548/2014 and is the most common in tender specifications.
Method C: Minimum Efficiency Index at 50% Load Factor (EIB50). The efficiency at 50% load must meet a defined minimum index. This method is useful for distribution networks where the average load is well below nameplate.
Level 1 is the basic energy performance level — comparable to GB 20052 S11 / EU Tier 1. Level 2 is the high energy performance level — comparable to GB 20052 S20 / EU Tier 2. The buyer selects Level 1 or Level 2 based on the economic value of losses in the application (see Taishan Transformer's IEC 60076 standards guide for the complete framework).
3. EU 548/2014 Tier 1 vs Tier 2 Loss Values
EU Regulation 548/2014 implements Directive 2009/125/EC for power transformers and is the global benchmark for efficiency regulation. The regulation defines two tiers with phased implementation:
- Tier 1 (from 1 July 2015): the first mandatory loss level for transformers placed on the EU market
- Tier 2 (from 1 July 2021): a tightened loss level that replaced Tier 1 for new placements
The Tier 1 / Tier 2 loss values for medium-power liquid-immersed transformers (the most common export configuration) are:
| kVA Rating | Tier 1 P₀ (W) | Tier 1 Pₖ (W) | Tier 2 P₀ (W) | Tier 2 Pₖ (W) |
|---|---|---|---|---|
| 100 | 145 (Ao) | 1,750 (Ck) | 130 (Ao-10%) | 1,250 (Ak) |
| 160 | 210 (Ao) | 2,350 (Ck) | 189 (Ao-10%) | 1,750 (Ak) |
| 250 | 300 (Ao) | 3,250 (Ck) | 270 (Ao-10%) | 2,350 (Ak) |
| 400 | 430 (Ao) | 4,600 (Ck) | 387 (Ao-10%) | 3,250 (Ak) |
| 630 | 600 (Ao) | 6,500 (Ck) | 540 (Ao-10%) | 4,600 (Ak) |
| 1,000 | 870 (Ao) | 10,500 (Ck) | 783 (Ao-10%) | 7,600 (Ak) |
| 1,600 | 1,300 (Ao) | 17,000 (Ck) | 1,170 (Ao-10%) | 12,000 (Ak) |
| 2,500 | 1,750 (Ao) | 25,500 (Ck) | 1,575 (Ao-10%) | 18,500 (Ak) |
| 3,150 | 2,200 (Ao) | 32,000 (Ck) | 1,980 (Ao-10%) | 23,000 (Ak) |
Where P₀ is the no-load loss and Pₖ is the load loss. The labels in parentheses (Ao, Ck, Ak) reference the loss categories in the regulation's nomenclature. Note that Tier 2 reduces both P₀ by 10% and Pₖ by a much larger percentage (typically 25-30%) compared to Tier 1. For utility-scale purchases, the source-of-truth Tier 2 values are published in the regulation's Annex I tables (see Siemens Energy's Ecodesign Leaflet for the complete Tier 1 / Tier 2 table across all voltage and kVA combinations).
4. No-Load Loss vs Load Loss: Different Physics, Different Cost
The total loss of a transformer has two components with different physical origins:
No-load loss (P₀) is the loss in the iron core when the transformer is energized but not supplying load. It is a constant loss, present 24 hours per day for the life of the transformer, and depends on the core material, core weight, and flux density. P₀ is dominated by hysteresis loss and eddy-current loss in the silicon steel.
Load loss (Pₖ) is the loss in the copper (or aluminum) windings when the transformer is supplying load. It is proportional to the square of the load current (I²R loss) plus stray eddy-current losses in the windings and structural parts. Pₖ is zero at no-load and rises with load.
For a typical 1000 kVA distribution transformer running at 50% average load, P₀ accounts for approximately 70% of the total annual energy loss (because it runs 24/7 at full no-load loss), and Pₖ accounts for approximately 30% (because it is at half current half the time). At 80% average load, Pₖ rises to approximately 60% of annual loss. The capitalisation value of P₀ is therefore significantly higher than Pₖ for lightly loaded transformers, which is why high-efficiency classes (S20/S22) focus on reducing P₀ first.
5. S11 / S13 / S20 / S22 Loss Levels: 10 kV Three-Phase Reference
For three-phase 10 kV oil-immersed distribution transformers at the most common ratings, the no-load loss (P₀) values across efficiency classes are:
| kVA Rating | S11 P₀ (W) | S13 P₀ (W) | S20 P₀ (W) | S22 P₀ (W) |
|---|---|---|---|---|
| 400 | 570 | 480 | 420 | 380 |
| 630 | 810 | 680 | 600 | 540 |
| 800 | 980 | 820 | 720 | 650 |
| 1,000 | 1,150 | 1,050 | 950 | 850 |
| 1,250 | 1,400 | 1,250 | 1,100 | 980 |
| 1,600 | 1,700 | 1,500 | 1,350 | 1,200 |
| 2,000 | 2,100 | 1,800 | 1,650 | 1,450 |
| 2,500 | 2,500 | 2,200 | 2,000 | 1,800 |
These values are typical for 10 kV Dyn11 ONAN-cooled designs at 50 Hz. The S22 figures represent the lowest no-load loss achievable with CRGO silicon steel core technology without resorting to amorphous alloy. The improvement from S11 to S22 represents a 25-30% reduction in P₀ — and translates directly to lower annual operating cost for the transformer owner.
For a 1000 kVA S22 transformer with P₀ = 850 W running 24/7, the annual no-load energy consumption is 850 W × 8,760 h/year = 7,446 kWh/year. At an industrial electricity tariff of $0.08/kWh, the annual no-load cost is $596. Over a 25-year service life, that is approximately $14,900 in no-load energy cost — more than the original S22 price premium over S11.
6. 35 kV, 66 kV, and 110 kV Loss Comparison
As the voltage class increases from 10 kV to 110 kV, the absolute P₀ and Pₖ values rise with the kVA rating, but the percentage improvement across efficiency classes remains similar. For 66-110 kV oil-immersed transformers (the typical transmission substation class), the S22-class efficiency is most economically justified because the higher kVA ratings and continuous-duty operation amplify the lifecycle cost benefit:
| kVA Rating | Class | 10 kV P₀ (W) | 35 kV P₀ (W) | 66 kV P₀ (W) | 110 kV P₀ (W) |
|---|---|---|---|---|---|
| 6,300 | S11 | 8,300 | 9,500 | 10,800 | 12,500 |
| 6,300 | S13 | 7,400 | 8,500 | 9,700 | 11,200 |
| 6,300 | S22 | 5,900 | 6,800 | 7,800 | 9,000 |
| 12,500 | S11 | 14,500 | 16,800 | 19,200 | 22,000 |
| 12,500 | S22 | 10,500 | 12,000 | 13,800 | 15,800 |
| 31,500 | S22 | — | 24,500 | 28,000 | 32,000 |
| 63,000 | S22 | — | — | 44,000 | 51,000 |
The 66-110 kV segment is where the S22 premium delivers the largest lifecycle cost benefit. A 12,500 kVA 110 kV S22 transformer saves approximately 6,200 W of no-load loss compared to an S11 unit of the same rating — over 25 years of continuous operation, that is approximately 1.36 million kWh saved, worth $108,000 at industrial tariff. The Tianan 66-110 kV low-loss low-noise transformer line ships with the S22 option as the standard for international tenders.
7. Amorphous Alloy vs CRGO Silicon Steel
Amorphous alloy cores represent the largest single step-change in transformer no-load loss reduction since the introduction of CRGO silicon steel. Amorphous ribbon is approximately 25 microns thick (vs 230-300 microns for CRGO lamination), and its disordered atomic structure eliminates the magnetic domain walls that cause hysteresis loss. The result: no-load loss of ~0.25 W/kg at 1.4 T flux density, compared to ~1.0 W/kg for high-grade CRGO — a 75% reduction.
The trade-off for amorphous alloy cores is:
- Higher unit cost: 15-30% premium over CRGO S22 designs, depending on kVA rating and voltage class
- Larger footprint: the lower stacking factor (~0.85 vs 0.95 for CRGO) requires larger core window, which translates to a 15-25% larger and heavier transformer
- Lower mechanical strength: amorphous ribbon is brittle and sensitive to mechanical stress, requiring careful handling during manufacture and assembly
- Limited voltage class: amorphous alloy cores are practical for distribution transformers up to ~10 MVA, but large power transformers above this rating still use CRGO
The economic break-even calculation: for a 1000 kVA 10 kV transformer running 24/7 at 40% average load, the annual no-load energy cost saving from amorphous vs CRGO S22 is approximately $200-300. At a 30% unit price premium of approximately $1,500-2,500, the payback period is 5-10 years. For lightly loaded distribution networks (20-30% average load), amorphous may never pay back; for heavily loaded industrial transformers (70%+), CRGO S22 is the economic choice.
8. How to Read a Loss-Guarantee Test Report
A loss-guarantee test report from a transformer manufacturer should list seven items, each verifiable against the specification:
- No-load loss P₀ measured at rated voltage and rated frequency on the rated tap. The measured value must be at or below the specification (typically S11 / S13 / S20 / S22 limit for the rating).
- Load loss Pₖ measured at rated current and corrected to 75°C reference temperature (per IEC 60076-1). Measured by short-circuit test.
- Impedance voltage Uk% measured by short-circuit test, typically 4-6% for distribution transformers and 8-12% for large power transformers.
- Total losses = P₀ + Pₖ at rated load. Used for annual operating cost calculation.
- Efficiency at 100%, 75%, 50%, and 25% load at 0.8 power factor. The efficiency curve helps size the transformer for the actual load profile.
- Acoustic sound level in dB(A) measured per IEC 60076-10. Standard ONAN units are 56-62 dB(A) at 1000 kVA; low-noise designs are 50-55 dB(A).
- Top-oil and winding temperature rise measured by short-circuit heat-run test. Standard ONAN is 60 K / 65 K; enhanced cooling (ONAF) achieves 55 K / 60 K.
For S22-class verification at 1000 kVA 10 kV, the test report should show P₀ ≤ 850 W and Pₖ ≤ the manufacturer's specified value (typically 7,500-8,500 W for 10 kV Dyn11 ONAN). If either value is above the specification, the unit does not meet S22 class and should be rejected or re-tested.
9. Tianan Loss Guarantee and Audit Process
Tianan's loss-guarantee process is documented in three stages: factory acceptance test, third-party witness test (optional), and warranty period compensation. The audit process page outlines the steps:
- Stage 1: Factory acceptance test. Every transformer is tested at the Tianan factory before shipment. The test report includes P₀, Pₖ, Uk%, efficiency at four load points, sound level, and temperature rise. The customer can attend the test in person or via live video link.
- Stage 2: Third-party witness test (optional). For large orders or first-time cooperation, the customer may engage a third-party testing body (TUV, SGS, BV, or local utility lab) to witness the factory acceptance test. Tianan supports third-party witness tests at the customer's expense.
- Stage 3: Warranty period compensation. If a verified loss measurement on site exceeds the guaranteed value by more than 5%, Tianan provides compensation per the contract terms. The standard warranty is 24 months from delivery or 30 months from factory dispatch, whichever comes first.
For buyers who need additional assurance, Tianan supports on-site re-testing by an independent lab during the warranty period. If the measured loss exceeds the guaranteed value by more than 5%, the cost of the re-test is borne by Tianan, plus the contracted compensation payment. This loss-guarantee structure is what utility buyers typically require for international tenders (see SEA Trasformatori Vademecum on Regulation EU/548/2014 for the European manufacturer's perspective on loss-guarantee documentation).
10. Conclusion: Efficiency Class Is a Lifecycle Cost Decision
Transformer efficiency class selection is fundamentally a lifecycle cost decision. The capital cost premium for S20 or S22 over S11 is typically 8-20%, but the lifecycle operating cost savings typically recover the premium within 3-7 years for transformers running at 60-80% average load. For lightly loaded distribution transformers, the payback extends to 10-15 years and may never materialize if the transformer is replaced before end-of-life.
The minimum acceptable efficiency class is determined by the destination market and the tender specification: S20 (EU Tier 2) for EU and most international tenders as of 2026; S13 (Tier 1 equivalent) for legacy installations in non-EU markets; S22 for high-duty applications where lifecycle cost matters. The Chinese GB 20052-2024 standard has raised the mandatory minimum to S20 for distribution and S22 for medium-power applications, aligning Chinese in-China requirements with international high-efficiency benchmarks.
For project-specific efficiency class selection, the Tianan engineering team can provide a side-by-side lifecycle cost comparison across S11 / S13 / S20 / S22 at your specified kVA, voltage, and load profile. Submit the seven data points (kVA, voltage, load profile, duty cycle, target payback, destination market, applicable regulations) through the contact form and the engineering team will return a recommendation within 2-3 business days. The loss-guarantee process and warranty terms are documented at the audit process page.