How Do Agricultural Environments Influence Bifacial PV Performance

Farms can change bifacial PV output because the ground and crops around the modules are never really fixed. Bare soil may be dry in the morning and wet after irrigation. A crop that is only 20 cm high in spring may cover most of the ground a few months later. After harvest, the soil becomes visible again. All of this changes the light reaching the back of a bifacial module.

For most agricultural projects, three things deserve the closest attention: what the rear side can actually see, how that view changes during the season, and whether the inverter and other electrical equipment can use the extra power. IEA PVPS also treats agrivoltaics as a system where crop production and PV operation need to be assessed together.^[1]

Tongwei lists both PV + agriculture and PV + animal husbandry among the applications for its PV modules. In these projects, the module is only one part of the picture. Ground cover, crop growth and mounting layout all affect the rear side.

Key Terms

Albedo tells us how much sunlight a surface reflects. Rear irradiance is different: it is the light that actually reaches the back of the module.

Bifaciality belongs to the module itself. It describes how strongly the rear side responds compared with the front under defined test conditions. IEC TS 60904-1-2 covers measurement methods for bifacial PV devices.^[2]

Bifacial gain belongs to the whole installed system. It measures the extra output of a bifacial system compared with a clearly defined reference system.

So these numbers should not be mixed up. 20% albedo is not 20% rear irradiance, and 20% rear irradiance is not 20% extra AC energy.

For example, the datasheet for Tongwei's TWMNH-66QD module lists mass-production bifaciality of 90±5%. That number describes rear-side response under test conditions. It does not mean a solar project will produce 90% more electricity.

Ground Reflectance

IEA PVPS reports that many natural soils and vegetation surfaces fall roughly in the 0.10 to 0.30 albedo range. Moisture, soil type, vegetation and sunlight conditions can move the real value higher or lower.^[3]

NREL measurements show why one number cannot represent every farm. In one dataset, native grass sites measured about 0.238–0.247, alfalfa about 0.221, pasture grass about 0.262, and a mixed grass-and-crop surface about 0.252.^[4]

A bright surface is only useful if sunlight reaches it and the module can see it. Dry soil directly under a deep module shadow may have good albedo, but it will not send much reflected direct light toward the rear cells.

Rear Irradiance

A simplified relationship is:

Equivalent irradiance ≈ front irradiance + rear irradiance × bifaciality

Take a module with 80% bifaciality:

If the front receives 1,000 W/m² and the rear receives 150 W/m², the simplified equivalent irradiance is:

1,000 + (150 × 0.80) = 1,120 W/m²

That still does not mean AC energy rises by exactly 12%. Cell temperature, mismatch, wiring losses, inverter efficiency and clipping all sit between sunlight and the final AC output.

Soil Moisture

Rain and irrigation can change how much light exposed soil reflects within a short time.

Wet soil is often darker than dry soil. If most of the ground is still visible, that change can be noticeable on the rear side.

The 0.24 and 0.17 values above are only calculation examples, not standard dry- and wet-soil values.

The same change in soil reflectance matters much less when a mature canopy hides most of the ground. Drip irrigation may wet narrow strips, while sprinkler or flood irrigation can change a much larger visible area.

Crop Growth

As crops grow, the rear side gradually sees less soil and more leaves.

The most useful field data are crop height, canopy density, row direction, visible soil, planting date and harvest date.

Height alone can be misleading. A 1 m crop with open leaves may expose plenty of soil. Another 1 m crop with dense foliage may hide almost all of it.

The crop heights and visible-soil percentages are examples, not standard values for all farms.

Early in the season, exposed soil controls much of the rear reflection. At full canopy, leaves and crop height become more important. Harvest changes the scene again, often within a few days.

Crop Type

The height ranges above are only examples used to explain array interaction.

A low but thick crop can hide more ground than a taller crop with open spacing. What matters is the shape of the crop around the module, not just the plant name.

Crop Rows

Tall crop rows running parallel to the PV rows can create long bands of vegetation and exposed soil behind the modules. Rows running across the PV array create a different pattern along the module row.

Row direction matters most when crops are tall or widely spaced. With short, even ground cover, the difference is usually much smaller.

Module Height

More clearance usually lets the rear side see a larger area of ground. It also leaves more room for crops and machinery.

These heights are comparison examples, not recommended minimums.

NREL includes module height, ground reflectance and shading geometry in its bifacial modeling tools because all three change rear irradiance.^[5]

Raising a very low array can make a clear difference. Once the rear already has a wide view of the ground, another meter may add much less light while still increasing steel use, foundation size and wind load.

Row Spacing

Wider rows expose more ground to sunlight and leave more room for tractors and other farm equipment. Fewer modules, however, fit on each hectare.

Ground Coverage Ratio, or GCR, describes how tightly the array is packed.

With about 2.0 m of module length per row, a simplified example looks like this:

This is only a simple GCR example. Real projects use the actual module and row geometry.

A wider layout may improve rear irradiance for each module while lowering the number of modules installed per hectare. More bifacial gain per module does not guarantee more electricity per hectare.

Rear Shading

Torque tubes, rails, posts, junction boxes, cables, irrigation pipes and tall plants can all block rear-side light.

Sandia modeling shows that module position, height, nearby rows, structures and ground shadows can change rear irradiance.^[6]

Torque tubes, rails and posts should be included in the rear-side shading model. Tongwei's module installation manual also gives mounting and clamping requirements that should be checked before the support structure is finalized.

Rear Uniformity

Two modules can have the same average rear irradiance and still behave differently.

Module B has much larger differences between cell areas, so electrical mismatch can be higher.

Sandia measured differences of more than 50 W/m² between rear cell locations in one sunny test configuration with the module around 0.6 m above the ground at latitude tilt.^[7]

The >50 W/m² figure belongs to that test setup and should not be used as a default mismatch value. On farms, uneven crops, posts, wet patches and irrigation equipment can create similar kinds of uneven rear lighting.

Diffuse Light

In most tilted bifacial systems, the rear receives both ground-reflected light and diffuse light from the sky.

Clouds and atmospheric particles scatter sunlight, so some light still reaches the rear even when direct ground reflection falls. Under cloudy conditions, diffuse light can make up a larger share of rear irradiance.

Temperature

Silicon PV modules usually lose power as cell temperature rises. The exact change is given by the temperature coefficient of the selected module.

Crops, wet soil, shade and airflow can change local temperature and humidity. DOE agrivoltaic research studies these crop-and-PV microclimate effects.^[8]

Do not assign a fixed cooling gain to crops. Module temperature still depends on ambient temperature, sunlight, wind, module construction and airflow around the array.

Use the selected module's actual temperature coefficient. Tongwei publishes this information in product data such as the TNC2.0-G12R-66 dual-glass module specification.

Irrigation

Irrigation changes more than the soil.

Sprinklers can wet module glass. If the water contains a lot of dissolved minerals, repeated drying may leave deposits. The real risk depends on water quality, spray direction and how often the glass is wetted.

Before fixing the PV layout, check:

• sprinkler height and reach
• prevailing wind
• where soil stays wet
• whether water reaches the module glass
• whether irrigation equipment remains accessible

Sprinkler reach, wet zones and service paths should be checked before post and module positions are fixed.

Dust and Soiling

Plowing, mowing, harvesting and vehicles on dry farm roads can create short periods of heavy dust.

If the rear provides 10% of effective irradiance and rear soiling reduces that rear contribution by 10%, total effective irradiance falls by about 1%, not 10%. The numbers are only an example, but the relationship is useful: rear dirt matters in proportion to how much energy the rear was supplying in the first place.

Base cleaning on measured loss, rainfall, farm activity and cleaning cost instead of using the same calendar schedule all year.

Vegetation and Livestock

Weeds growing close to the modules can shade lower rear cells, block inspection access and interfere with tracker equipment.

Grazing changes the ground in a different way. Animals keep some vegetation shorter, while trampling can expose patches of soil. The result can be a less even ground surface even when average vegetation height falls.

DOE includes livestock grazing among the main forms of agrivoltaic co-location.^[9]

Where animals enter the array, cables and electrical equipment also need suitable protection.

Snow

Fresh snow can reach an albedo close to 0.90 under suitable conditions, far above the roughly 0.10–0.30 range often seen for many natural soils and vegetation surfaces.^[10]

That gives the rear side access to much more reflected light, but snow can also cover the front glass, pile up near the lower edge and melt unevenly.

Winter output depends on both effects: more rear reflection and possible front-side snow loss.

Tracking and Vertical Systems

Single-axis trackers change module angle during the day, so both front irradiance and the ground visible to the rear keep changing.

On a farm, tracker position also changes crop shade and machinery clearance. IEA PVPS identifies agrivoltaics as an important use case for bifacial tracking systems.^[11]

Vertical bifacial arrays work differently again. In a typical east-west layout, one side receives stronger direct sunlight in the morning and the other side receives more later in the day.

Fixed-tilt results cannot be transferred directly to vertical or tracking arrays because sun angle, ground shadow and rear exposure are different.

Seasonal Change

The percentages are examples used to show seasonal change, not standard farm values.

A high-albedo period during a low-sun month may add less annual energy than a smaller reflectance increase during the high-sun growing season. Monthly or seasonal ground assumptions are therefore more useful than one fixed annual value when the farm changes sharply through the year.

Modeling

NREL's bifacial_radiance toolkit uses ray tracing to model reflective surfaces, module geometry, shading and rear irradiance.^[12]

If the crop reaches 1.8 m but the model assumes 0.8 m, even a detailed simulation is modeling the wrong field. The same problem applies to wrong albedo, clearance, GCR or row-pitch inputs.

Electrical Limits

Extra rear irradiance raises DC production, but inverter and current limits can stop part of that extra power from reaching the AC side.

Check:

• module current
• string current
• fuse ratings
• cable ratings
• combiner limits
• inverter MPPT current limits
• inverter clipping

This is a calculation example, not a recommended DC/AC design ratio.

Optical losses happen before light reaches the cells. Crop shade and rear racking are examples.

Electrical limits happen after the module has already produced DC power. Inverter clipping is one example.

Monitoring

Record farm conditions alongside PV data so changes in rear irradiance can be tied to crop growth, irrigation, harvest or cleaning.

• front irradiance
• rear irradiance at representative positions
• albedo where it matters to the model
• module temperature
• crop height and growth stage
• irrigation dates
• harvest dates
• cleaning events
• inverter and string data

NREL calculates measured albedo from ground-reflected irradiance divided by incoming global horizontal irradiance and publishes time-series, monthly and annual values.^[13]

If 80% of a field is covered by crops in summer but the albedo sensor sits over the remaining bare soil, the reading can be correct at the sensor and still give the wrong picture of the whole farm.

The same issue appears when the only rear irradiance sensor sits at the outside edge of a large array.

Check the Model

• Compare modeled rear irradiance with measured rear irradiance.
• Compare assumed crop height with actual crop height.
• Check whether the real ground surface matches the albedo assumption.
• Compare predicted energy with weather-adjusted measured energy.
• Compare expected clipping with inverter records.

If the model assumed a 1.0 m crop but the field reaches 1.6 m, lower rear irradiance may come from crop geometry rather than module degradation.

Troubleshooting

Bifacial gain also needs a clear comparison system. An 8% gain at one farm and a 12% gain at another do not prove that the second project is better. Weather, albedo, GCR, module type and the reference system may all be different.

Design Checklist

Check Tongwei's module installation manual for the selected module's mounting, electrical-connection and maintenance limits. Its PV quality white paper also provides information on product testing and quality control.

FAQ

Can crop rotation change PV yield?
Yes. Replacing a low crop with a taller, denser crop changes visible ground and rear shading even when the PV equipment stays the same.

Why can edge modules behave differently?
Modules near the edge can see illuminated ground outside the array that interior modules cannot. Module position is one of the factors included in Sandia bifacial modeling.^[14]

When is on-site albedo measurement worth doing?
It is most useful when rear-side gain has a meaningful effect on project economics and published values do not match the farm's real soil, crops, irrigation or snow conditions.

Conclusion

Agricultural bifacial PV changes with the farm. Natural soil and vegetation often sit around 0.10–0.30 albedo, while fresh snow can approach 0.90. Crop growth can hide most of the soil that was visible after planting, and irrigation can temporarily lower the reflectance of exposed ground. Wider rows and more clearance can improve rear-light access, but they also reduce module density or increase structural cost. Reliable yield estimates need seasonal ground conditions, real crop height, actual array geometry and electrical limits—not one fixed albedo number for the whole year.