How to Compare Solar Module Efficiency: STC, Area, Bifacial Gain and Lab Records

Short answer. A module efficiency figure is a ratio: electrical power out divided by light power in, both measured under stated conditions on a stated area. Change the conditions (irradiance, spectrum, temperature), the area (cell, aperture, total module) or the basis (lab-confirmed record versus datasheet nameplate) and the percentage changes. Headlines that compare a research-cell record with a commercial module datasheet, or a bifacial "with rear gain" rating with a front-only rating, are comparing different quantities. This guide explains the definitions and shows the arithmetic on two published datasheets.

Where the conditions come from

IEC 60904-3:2019 (edition 4.0, published 15 February 2019) describes the basic measurement principles for determining the electrical output of PV devices and relates performance ratings to a common reference terrestrial solar spectral irradiance distribution. That reference spectrum is what "AM 1.5" on a datasheet points to. The standard test conditions printed on datasheets combine it with an irradiance and a temperature; the JinkoSolar datasheet used below states them as irradiance 1000 W/m², cell temperature 25 °C, AM 1.5.

A single operating point is a convention, not a forecast of yield. IEC 61853-1:2011 (published 26 January 2011) defines a testing and rating system that gives module power at maximum power operation over a range of irradiances and temperatures, not just at one condition. For long-term qualification, IEC 61215-1:2021 covers design qualification of terrestrial PV modules for open-air operation; its test results are not a quantitative lifetime prediction. Efficiency, energy rating and durability are three different test families.

Research-cell records are not module efficiencies

The National Laboratory of the Rockies (NLR, formerly NREL) keeps the chart most often quoted in headlines. Its Best Research-Cell Efficiency Chart includes only efficiencies confirmed by independent recognised test labs (for example NLR, AIST, JRC-ESTI and Fraunhofer-ISE), measured under Standard Test or Reporting Conditions per IEC 60904-3 edition 2 or ASTM G173, at a reference temperature of 25 °C, with the area defined as the total cell area or an aperture area. The cell chart runs from 1976; NLR keeps a separate Champion Photovoltaic Module Efficiency Chart from 1988.

The module chart uses the same confirmed-measurement basis with module total area or aperture area, and it clusters modules by size: above 14,000 cm² large, 6,500–14,000 cm² standard, 800–6,500 cm² small, and 200–800 cm² mini-module. A record set on a small aperture is therefore flagged as such. When a press release quotes a record, three questions settle what it means: cell or module; which area definition; and which lab confirmed it. A manufacturer's own measurement is not a chart entry until a listed lab confirms it.

Reading a datasheet: the arithmetic

Datasheet efficiency is the module's STC power divided by its total area and by 1000 W/m². Two JinkoSolar datasheets (© 2024) show the calculation; they are evidence samples, not a recommendation.

Bifacial module, JKM625-650N-78HL4-BDV-F9-EN. The STC table lists maximum power of 625–650 Wp, module efficiency (STC) of 22.36–23.25% and a power tolerance of 0 to +3%. The module measures 2465 × 1134 × 30 mm. Dividing 650 W by 2.465 m × 1.134 m × 1000 W/m² gives 23.25%, and 625 W gives 22.36% — the printed figures (CNWattHub arithmetic). The efficiency, then, is defined on the full outer area including frame, which is the common datasheet convention and larger than the cell-area basis used in cell records.

Mono-facial module, JKM610-635N-66HL4M-(V)-F2-EN. This sheet lists STC power 610–635 Wp, module efficiency 22.58–23.51%, the same 0 to +3% tolerance and dimensions 2382 × 1134 × 35 mm. 635 W over 2.382 m × 1.134 m × 1000 W/m² reproduces 23.51%. Two modules whose efficiencies differ by a fraction of a point can differ simply because one has a slightly different frame or cell count; efficiency is not the same thing as power per module.

Tolerance sign. Both sheets print 0 to +3%, meaning the delivered module may exceed but not fall below nameplate at STC. A tolerance that also allows a negative deviation lowers that floor; always read the sign.

Temperature. The bifacial sheet gives a Pmax temperature coefficient of −0.29 %/°C. STC is 25 °C cell temperature; a module operating hotter produces less than nameplate, and the coefficient tells you how much per degree. Efficiency comparisons across technologies at STC alone ignore this.

Warranty is a separate curve. The mono-facial sheet's front page shows a 12-year product warranty and a 30-year linear power warranty with 1% first-year degradation, 0.40% annual degradation and 87.4% at year 30. (100% − 1% − 29 × 0.40% = 87.4%, CNWattHub arithmetic.) The warranted power is a fraction of nameplate at STC; it does not change the efficiency definition, but it is the number a lender will model.

Bifacial: two power tables on one datasheet

The bifacial sheet also prints a second table under "BNPI" — front 1000 W/m², rear 135 W/m², cell temperature 25 °C, AM 1.5 — with maximum power of 688–716 Wp, together with bifaciality coefficients φVoc 98±5%, φIsc 80±5% and φPmax 80±5%. The BNPI power is higher than the STC power because it adds a defined rear irradiance; it is not comparable with the STC rating of a mono-facial module, and the real rear contribution on site depends on albedo, mounting height and row spacing, none of which the datasheet can know.

The measurement procedure for bifacial devices is set out in IEC TS 60904-1-2:2024 (edition 2.0, published 7 November 2024), which describes the additional requirements for bifacial I-V measurement, while mono-facial devices are covered by IEC 60904-1; the 2024 edition replaces the 2019 first edition. A bifacial efficiency quoted "with gain" should state which edition and which rear irradiance it used; otherwise compare only the front-side STC figure.

What China's 2024 manufacturing thresholds mean

MIIT revised and issued the PV Manufacturing Industry Standard Conditions (2024 edition) on 15 November 2024. The 2024 edition raises average-efficiency thresholds for new projects: P-type cells and modules not below 23.7% and 21.8%, N-type cells and modules not below 26% and 23.1% (previously 23% for mono cells and 20% for modules). MIIT's own explanation states that the current mainstream N-type TOPCon average cell efficiency is 25%, so a 26% threshold for new N-type cell projects substantially raises the entry bar. The same revision raises the required workmanship-and-materials warranty period from 10 to 12 years and adds reliability-test requirements.

Two cautions. These are policy thresholds for new manufacturing projects, not product certifications, and a module listed at 23.1% is not "MIIT-approved" by virtue of the number. And the thresholds are averages for a project's output, which is a different statistic from a datasheet's top bin.

Checklist before comparing two efficiency numbers

1. Cell or module? Lab-confirmed record or manufacturer datasheet?
2. Which area: total module area (datasheet), aperture, or cell area (records)?
3. Which conditions: STC front-only, or BNPI / "with bifacial gain" — and which rear irradiance?
4. Which tolerance sign and which bin within the model range?
5. What is the temperature coefficient, and at what cell temperature will the module actually run?

Scope limits

Two datasheets from one manufacturer are used to show arithmetic; no ranking or cross-supplier comparison is intended. Standard texts were read at abstract level only. NMOT values are not printed on the two datasheets used, so no NMOT example is given. MIIT thresholds are quoted from the ministry's interpretation page, not from the 公告 text.