Posters

Abstract

Post-construction AEP estimates are used in validating pre-construction AEP methodologies, but how do we validate the post-construction AEP estimates themselves?! Two open operational wind farm datasets were used in a WeDoWind Open Data Exploration Challenge to compare results of various post-construction AEP methodologies by different challenge participants. This highlighted that deviations of 4% in the net P50 are not unusual. This can lead to significant variations in project valuations, increasing risk, and potentially undermines validation of pre-construction AEP estimates as well. Work was done to determine the causes of these variations, with the calculation of system availability, future degradation and availability assumptions, exclusions of one-off events, and source of long-term reference datasets key determinants, but, identifying these causes does not help us determine what the "correct" result should be. A data driven approach is proposed and tested that validates post-construction AEP estimates to future measured realities. For this work OpenOA has been used to estimate the P50 during an initial operational period, and then compared to actuals from a future period of operation. This validation methodology is tested on the two open operational wind farm datasets, and then applied to a much larger dataset, allowing the net P50 from the post-construction analyses to be compared against reality. A spread and positive bias is identified in the estimated P50 compared to the measured reality, implying that the P50 has been overestimated by the post-construction AEP estimate. The OpenOA analysis does not account for a reduction in performance due to system degradation or reduction in system availability in later life, and it is hypothesised that this could be the cause, however, further work will be required to determine the exact reasons. This validation method for post-construction yield analysis will be published as online open-source resource. It can therefore be used to validate other post-construction methodologies, act as an industry benchmarking tool, and in so doing can also support the validation of pre-construction AEP estimates too.