Energy yield assessment

An energy yield assessment predicts how much electrical energy a PV plant will produce. It usually gives the average annual output over the plant's lifetime, which is typically taken as 25 to 30 years. It forms the basis for calculating project revenue, and lenders generally want a bankable assessment carried out or reviewed by an independent expert. The method uses hourly or sub-hourly time-step simulation. It takes typical weather data for the site, transposes the irradiance onto the module plane, models how the plant responds to irradiance and temperature, and then applies losses. The losses include near and far shading, soiling or snow, wiring, inverter, mismatch, degradation and downtime. Results are given with confidence levels such as P50 and P90. Because yield follows plane-of-array irradiance closely, uncertainty in the solar resource data dominates the overall uncertainty. Overall uncertainty is often quoted at around 5 percent, but in some cases it can be more than 10 percent. A quick estimate multiplies the specific annual yield for the location by the system's nominal power. After commissioning, the model's expected yield can be compared with the measured specific yield using the performance ratio or the Energy Performance Index.

<!-- review notes

• sentence 4: support check: partial - The passages mention shading, soiling or snow, wiring, inverter, module degradation and downtime, and mismatch in u18, but they do not present these as one list of losses for yield assessment.
• sentence 7: support check: partial - The passages support the quick estimate and the EPI comparison, but nothing says the PR can be used after commissioning to compare the model's expected yield with measured yield.
• sentence 7: review: partial - The SMA example supports the quick estimate and the EPI passage supports comparing measured specific yield with model-expected yield, but 'after commissioning' and using the performance ratio as the comparison metric are not stated. -->