P50 and P90 energy yield

Every energy yield assessment for a wind or solar plant ends with two figures. P50 is the annual output the plant exceeds with 50% probability; P90 is the output it exceeds with 90% probability. The gap between them measures uncertainty: year-to-year resource variability, quality of the wind or irradiance measurement, wake losses, machine availability. A one-year P90 sits lower than a ten-year P90, because annual variability averages out over the decade. Lenders size debt on the one-year P90, sometimes on P99, with a debt service coverage ratio set against it; sponsors value the plant on P50. A short measurement campaign or a complex site widens the gap, and the debt amount drops accordingly, without anything changing at the plant. The link with insurance runs through business interruption. After physical damage, the indemnity is computed on the output lost during the outage. Some policies measure it against the wind or irradiance actually observed during the outage; others against the production budget, often the monthly P50. The difference shows up on every winter nacelle loss. A year without wind, with no damage at all, is not a loss under the property policy. Volume risk is covered elsewhere: weather derivatives, index-based production covers, or a proxy revenue swap, which guarantees a fixed revenue against actual output. Lenders see these as a way to bring P90 closer to P50 in their models, and the price of the hedge tells what the market thinks of the assessment.

• NREL (Dobos, Gilman, Kasberg), P50/P90 Analysis for Solar Energy Systems Using the System Advisor Model, 2012