How to Assess Solar Module Bifacial Gain: Bifaciality Factor vs Bifacial Gain, Albedo Ranges, Measured Tracker Gains of About 4–12% at Natural Albedo (2–10% Typical), the IEC TS 60904-1-2 (BNPI 1,000/135 W/m²) Method and PVsyst Defaults

Short answer. Bifacial gain is the extra energy a bifacial system produces compared with a monofacial system of the same orientation and size; it rises with albedo, diffuse fraction, array height, row spacing and module spacing, the rear-side light is far less uniform than the front, which creates mismatch; and, separately, bifacial arrays run at higher DC current. It is not simply a module property: at system level IEA PVPS Task 13 defines it as rear-side energy divided by front-side energy, and it depends on mounting geometry, ground albedo and its homogeneity, the structure and the BOS design. On single-axis trackers over natural ground (albedo 0.2–0.3) gains are generally below 10%, and measured plant values cluster between about 4% and 12% at natural albedo, rising to about 17.5% at 0.6 albedo. What a datasheet gives you is the bifaciality factor (for example φPmax 80±5%) and a power rating at bifacial nameplate irradiance (BNPI: 1,000 W/m² front plus 135 W/m² rear), which is a measurement convention, not a yield forecast. This guide explains how to separate the two, what the Chinese adoption of IEC TS 60904-1-2:2024 measures, and which simulation inputs you must supply.

1. Bifaciality factor and bifacial gain are different things

IEA PVPS Task 13 (2021) separates three levels. Optical gain is rear irradiance over front irradiance (BGopt = Grear/Gfront). Module gain multiplies the rear irradiance by the bifaciality factor φ, the ratio of rear-side to front-side efficiency (BGmod = Grear·φ/Gfront), because the rear side is generally less efficient. System gain is rear-side energy over front-side energy (BGsys = Erear/Efront), and the report notes that φ itself may vary with irradiance level, rear-to-front ratio and operating temperature.

Typical bifaciality by cell type, as the 2021 report puts it: bifacial n-PERT modules mass-produced by four named Chinese companies with over 21% front efficiency and bifaciality of 80 to 85%, and HJT cells typically above 92%. A JinkoSolar dual-glass n-type datasheet states φVoc 98±5%, φIsc 80±5% and φPmax 80±5%.

CNWattHub reading: two modules with the same φ can deliver very different bifacial gain on different sites, and the same module on the same site will deliver different gain on a 1P tracker and on a low fixed-tilt rack. Compare bifaciality factors between modules; compare bifacial gain only between complete system designs.

2. What drives the gain

• Albedo. Albedo is the fraction of incident sunlight a surface reflects; it is not constant, varying with spectral and angular distribution, wetness and roughness. Table 7 of the 2021 report gives grass 0.15–0.26, snow 0.55–0.98, black soil 0.08–0.13, clay soil 0.16–0.23, sand 0.21–0.60, new asphalt 0.09 and weathered asphalt 0.18. Albedo and module bifaciality are generally the most important factors, and gains increase significantly under snow.
• Geometry. Module height, tilt and row-to-row distance, plus the homogeneity of the ground and the mounting structure, set how much light reaches the rear and how evenly.
• Structure. The 2024 Task 13 tracking report says portrait orientations (1P, 2P) are preferred for bifacial trackers because landscape mounting needs two purlins across the back of each module, and purlins crossing the rear can reduce rear irradiance and gain.
• Mismatch and current. Non-uniform rear light causes electrical mismatch within modules and possibly across strings, and bifacial arrays run at higher DC current, so wire, fuse and inverter sizing may need adjusting.

3. What measured plants show

The 2021 report compiles single-axis-tracker results: 10.4–12.4% at the 1.7 MW La Silla plant in Chile under desert conditions; Canadian Solar systems at 5.2–8.9% for 1P trackers at 0.2–0.3 albedo and 17.5% for a 1P 12 kW system at 0.6 albedo; NREL's demonstration plant at 4–8% for PERC and 6.5–11% for silicon heterojunction modules at 0.2 albedo; and a California system at 4.1% for 0.16 albedo and 9% for 0.56 albedo.

The 2024 report frames the market: over 90% of modules sold use bifacial cells, over 60% of installed systems use single-axis trackers, typical tracker gains of 15–20% and bifacial gains of 2–10% are additive, and a cited 2020 study found bifacial tracked systems gave 16% lower LCOE and up to 35% higher yield than fixed-tilt monofacial systems; in the report's simulated US comparison the largest relative increases appear to occur in snowy regions. On albedo enhancement it is cautious: enhancing materials typically have 60–75% albedo, in theory able to triple the gain (an extra 4–20% absolute on a 2–10% base), but, considering material cost and inverter clipping or curtailment, only half of this may realistically be achieved.

4. What the standard measures: IEC TS 60904-1-2:2024 and China's GB/Z draft

IEC TS 60904-1-2:2024 (edition 2.0, published 7 November 2024) adds bifacial requirements to IEC 60904-1 I-V measurement, replaces the 2019 edition and, among its changes, calculates equivalent irradiance from the ISC bifaciality rather than the minimum of the ISC and Pmax bifaciality. China is adopting it identically as a GB/Z guiding technical document: the consultation draft, proposed by MIIT and handled by SAC/TC90, defines bifaciality (双面率) as the ratio of front and rear characteristics (normally at STC; ISC, VOC and Pmax versions), defines equivalent irradiance, defines bifacial nameplate irradiance with the example of 1,000 W/m² front and 135 W/m² rear, and states that, unlike albedo, bifaciality is an intrinsic device property.

The method: bifaciality comes from separate front and rear STC I-V curves; crystalline-silicon bifaciality typically falls as irradiance falls, so non-STC values carry an irradiance subscript; several samples should be measured and their spread reported; and the rear-irradiance power gain BiFi is the slope of a linear fit of Pmax against rear irradiance passing through Pmax,STC, with Pmax,BNPI = Pmax,STC + BiFi·Gr reported at Gr = 135 W/m². Outdoors, rear non-uniformity must stay below 10%, so the draft suggests raising the module's bottom edge 0.5–1.0 m or laying a matte reflective cloth; indoors, single-side equivalent irradiance is GE = 1,000 + φISC × Gr, and the worked example (φISC 80%, Gr 200 W/m²) gives 1,160 W/m².

Reading a datasheet with this in mind: the JinkoSolar sheet lists STC power of 625–650 Wp and BNPI power of 688–716 Wp for the same bins. CNWattHub arithmetic: 688/625 ≈ 1.101 and 716/650 ≈ 1.102, i.e. the BNPI rating is about 10% above STC — the consequence of assuming 135 W/m² on the rear with φ ≈ 80%, not a measured field gain. Whether a site delivers 135 W/m² on the rear on average depends on everything in section 2.

5. Simulation inputs you must supply

PVsyst's documentation defines the module bifaciality factor as rear STC yield over front STC performance (a module specification), separates the ground albedo under the array from the project's far-terrain albedo (monthly values allowed for snow), requires a height-above-ground input, and recommends default values of 5% for structure shading of the rear (for structures adapted to bifacial modules) and 10% for rear-irradiance mismatch (for most cases), warning that both can increase drastically when structure sits directly behind the modules. CNWattHub reading: ask how the ground albedo, the height above ground and the two loss factors were set; a yield report that omits the height input or sets mismatch to 0% without justification deserves questions.

For China specifically, note that the 2024 PV Manufacturing Industry Norms count bifacial cells and modules by front-side efficiency when applying efficiency thresholds (for example 25% for existing n-type cells and 23.1% for n-type modules in new projects), so the Norms' efficiency figures say nothing about rear performance.

Reading table

Checklist

CNWattHub editorial method, not a sourced rule:

1. Ask for the module's bifaciality factors (φISC, φVOC, φPmax) with the irradiance they were measured at, and for the BiFi value and Pmax,BNPI from a test report to IEC TS 60904-1-2:2024 or the GB/Z adoption.
2. Treat the BNPI power rating as a convention (≈ +10% at φ ≈ 80%), not as expected yield.
3. Obtain a site albedo estimate by surface type and season, and ask how snow months were handled.
4. Check the structure: module height, pitch, portrait vs landscape, and whether purlins or torque tubes cross the rear.
5. Check the simulation inputs: ground albedo (monthly), height above ground, structure shading and mismatch factors, and the resulting BGsys.
6. Confirm DC design (cable, fuse, inverter) uses bifacial current, not STC front-only current.
7. If an albedo enhancer is proposed, ask for the clipping and curtailment analysis before accepting the theoretical gain.

Scope limits

Two IEA PVPS Task 13 reports, one consultation draft of a Chinese GB/Z adopting IEC TS 60904-1-2:2024, the IEC abstract, one software vendor's documentation, one manufacturer datasheet and one MIIT norm only. Measured gains are those reported for the named sites and are not transferable; no Chinese field data, NREL or Sandia validation paper, IEC 61215 rear-side test clause or price is cited. Not engineering or investment advice.