Valuable Lessons from Outliers in a Wind Energy Resource Assessment Benchmark Study
10 Sep 19
5:00 PM
-
6:00 PM
Wind energy resource assessment (WERA) techniques, wind farm operations and investor confidence can be improved by evaluating the root causes of outliers in benchmark assessments. Systematic improvements to wind energy resource assessment that reduce the spread in wind farm performance ratio have long-term ramifications for pre-construction prediction accuracy. Thirty-six (36) WERAs from 2002 to 2015, updated to current techniques, are compared to actual production data. 26 of 36 projects are based on prior financing assessments in the ArcVera database and ten additional projects were evaluated as part of the NREL Wind Plant Performance Prediction (WP3) Benchmark Project. The benchmark projects have over 4300 MW of combined nameplate capacity and 258 calendar wind farm years of production data. Two benchmarked projects are in South America, one in Canada and the remainder are in the US. Benchmark comparisons are made with and without adjustments for annual wind variation, curtailment, and post-COD changes. The average benchmark performance ratio is close to 100% with standard deviation under 5%. Having de-biased the pre-construction WERAs (for now), the focus here is on the cause of outliers, with the goal of eliminating them. Lessons from the outlier investigations reveal owner-operator caveats, evaluation of predictable external factors, and improvements to observations and MCP practice, would reduce the frequency of outlier wind farm performance.