Data analytics for Canadian commercial wind energy

点击翻译按钮获取中文摘要

The Wind Energy Institute of Canada (WEICan) is both a research institute and a wind farm owner/operator. Like other Canadian commercial wind farms, WEICan is interested in leveraging intelligence from their operations data to help them improve farm profitability. A flexible tool capable of identifying the most significant wind farm operations variables to monitor and analyze would provide significant value for WEICan as a wind farm owner. This is also of critical interest to WEICan in its capacity as a national research institute where it is leading an initiative to standardize essential operations data collection for Canadian commercial wind operations. Subsequently, the immediate objective of the proposal is to create a proof-of-concept software tool capable of better characterizing, and improving wind farm performance. Such a tool will further enable WEICan to execute their leadership role in data collection standardization for the Canadian/International wind energy sectors. It will also empower WEICan as a wind farm owner to reveal strategic opportunities to improve their economic advantage over their competition; through the ultimate development of near-real-time operational analytics. A relationally structured, object-capable, database will be developed to integrate incongruent data elements (production data, meteorology, maintenance logs and reports, inspection photos, structural monitoring instrumentation, etc.) into a central location. A GIS referenced graphical interface will be created to make queries more intuitive to construct. Outcomes from more comprehensive queries and analysis could begin to reveal causation for underperformance and component wear. Structural ageing may be able to be mapped when combined with probabilistic loading models. Finally, the actionable intelligence derived from the tool may be leveraged in potential control optimization scenarios, i.e., the detection and subsequent sector management of wake flows, etc. The tool will be developed with sample data from two different commercial wind farm operations. Test queries and analyses will be evaluated against previously validated "hand assembled" historical analyses.