Answer: Solar temperature derating estimates how module power changes as cell temperature moves away from the reference temperature. Use the selected module’s maximum-power coefficient and a defensible cell-temperature estimate. Include irradiance when estimating instantaneous DC power. A hot-hour calculation does not establish annual energy loss, AC output or a guaranteed financial return.
Calculate temperature loss without losing the irradiance term
Use a signed maximum-power coefficient, converted to a fraction per degree. A coefficient of −0.35%/°C becomes −0.0035/°C. The temperature-coefficient glossary explains why the power, open-circuit-voltage and short-circuit-current coefficients are separate quantities.
A simplified relationship is:
P_DC ≈ P_reference × (G_effective / G_reference) × [1 + gamma × (T_cell − T_reference)]
Here P is module DC maximum power, G is effective irradiance and gamma is the fractional power coefficient. The usual STC reference uses 25°C cell temperature and 1,000 W/m² irradiance. The historical PVWatts V5 manual documents this relationship and explains that its former low-irradiance correction was removed. The current V8 API documentation describes later module and thermal-model updates. This explanatory approximation does not specify every part of V8 or replace the selected software’s full electrical model.
Worked example with declared assumptions
Assume a hypothetical 450 W module, a −0.35%/°C power coefficient, 65°C cell temperature and 1,000 W/m² effective irradiance. The temperature difference is 40°C; the factor is 1 − 0.0035 × 40 = 0.86. Estimated DC power is 387 W. The previous FAQ’s 386 W was an arithmetic error.
| Hypothetical condition | Thermal factor | Estimated DC power |
|---|---|---|
| 65°C cell, 1,000 W/m² | 0.86 | 387 W |
| 65°C cell, 800 W/m² | 0.86 | 309.6 W |
| 15°C cell, 1,000 W/m² | 1.035 | 465.75 W |
The second row deliberately holds cell temperature constant to isolate irradiance’s effect. In an actual thermal model, changing irradiance can also change temperature. None of these rows represents measured weather or output in Phoenix, Mumbai or another city.
For an existing array, document the actual module model, coefficient source, irradiance definition and temperature method before replacing these illustrative inputs.
Estimate cell temperature separately from air temperature
Ambient temperature belongs to the weather record; cell temperature belongs to the module’s operating condition. The Sandia module-temperature model uses irradiance, ambient temperature and wind, with parameters tied to module construction and mounting. A roof-integrated assumption and an open-rack assumption should not silently share one parameter set.
A back-surface reading is also not automatically the cell temperature. Sandia’s cell-temperature model includes a separate irradiance-dependent difference between module and cell temperature. Record sensor location and the conversion method when comparing a forecast to measurements.
NOCT or NMOT is a reference-condition thermal descriptor, not a universal operating temperature. Read the selected datasheet’s test definition. Avoid treating a simplified datasheet-based estimate as validated for every wind speed, roof clearance or mounting arrangement.
What to check in the simulation
PVsyst’s steady-state thermal documentation describes an energy-balance model with heat-transfer coefficients and a wind term. Parameters require justification for the actual array. Merely choosing a sophisticated tool does not validate its inputs.
Use this review record:
| Input or result | Question to resolve |
|---|---|
| Module definition | Does the file match the purchased model and datasheet revision? |
| Weather record | Are temperature, irradiance and wind matched to the same location and timestamps? |
| Mounting model | Does it represent the actual rear ventilation and construction? |
| Thermal parameters | Are they measured, manufacturer-supported or declared defaults? |
| Reported power boundary | Is this module DC, inverter AC or energy at the project delivery meter? |
| Sensitivity run | What changes when only the thermal assumptions change? |
Keep all other assumptions fixed when comparing two thermal scenarios. Then inspect annual energy, seasonal differences and high-temperature hours. A single afternoon’s percentage is not the annual loss percentage.
Keep module, inverter and electrical design checks distinct
The Pmax coefficient does not replace cold open-circuit-voltage and hot operating-voltage checks. A hot-climate location can still have a cold design condition. Use the current equipment instructions and applicable adopted electrical rules for string sizing.
Inverter temperature derating is a separate equipment behavior. Obtain the exact model’s operating and derating curves; a universal 45°C threshold or 150 mm clearance is not defensible. Mounting clearances must satisfy the selected equipment instructions and project requirements.
Likewise, a better power coefficient alone does not prove that one cell technology is the best purchase. Compare actual models with the same weather, geometry, shading and system assumptions. Price premiums, warranties and integration requirements need their own evidence.
Avoid duplicate losses and unsupported savings
Before applying any extra percentage, identify what the output already includes. A temperature deduction added to thermally modeled energy can double-count the loss. Conditional sequential factors are multiplied; report-specific loss-diagram percentages may use different denominators and need their documented interpretation.
For example, four hypothetical sequential retained fractions of 0.90, 0.95, 0.97 and 0.98 leave 0.812763 of the starting quantity, a loss of 18.7237%. This arithmetic illustrates a specified sequence; it does not establish a project’s loss budget.
Carry the difference between complete simulation scenarios into the financial model using the actual applicable energy price and time profile. Do not convert an instantaneous power difference directly into a 25-year revenue claim. The solar-system-losses guide and PVsyst report review checklist provide related review paths.
When evaluating solar design software, ask to reproduce the same weather and module assumptions and inspect the reported loss boundary. That test is more useful than an unsupported promise of automatic accuracy.
Questions about temperature derating
What is solar temperature derating?
It is the temperature-related change in module power relative to its reference condition. For a module with a negative maximum-power coefficient, higher cell temperature reduces predicted power at otherwise comparable irradiance. Ambient air temperature alone does not establish that loss.
Does the temperature calculation predict AC output?
No. The simple calculation estimates module DC maximum-power output under stated assumptions. Inverter conversion, clipping, wiring, shading and other effects still need treatment in the system model.
Can colder modules produce above nameplate power?
A negative power-temperature coefficient gives a temperature gain below the reference cell temperature. Actual output also depends on irradiance and other operating conditions, so cold weather alone does not guarantee output above the nameplate.
Should a thermal loss be applied again to a simulation result?
First check the result boundary and the software loss sequence. If the simulation already includes the module thermal model, applying another generic temperature deduction would count the same effect twice.
Source review: September 30, 2026. Calculations are hypothetical illustrations; actual yield and electrical design need project-specific inputs and qualified review.
Where this fits
This article is part of SurgePV's Solar Technology hub, which works through the topic from first principles to the decisions a project team actually has to make.
