Approach to Identifying Behind-The-Meter Wind Power: A Case Study in Northern Ireland

As more and more electricity users install private renewable energy generators on the demand side, electricity demand changes into the difference between load and generation. The specific variation of generation and load cannot be directly obtained via prosumer meters, which challenges economic dispatch, emergency recovery, forecasting accuracy, etc. To address the data deficiency, this paper proposed an approach for separating out load and wind power generation from the user demand data. Specifically, separation processes firstly utilize discrete wavelet transform to decompose signals into several representative scales. Then, the operating capacity of wind turbines can be solved by matching the demand variation on a wavelet scale where disturbances like random behaviors and weather interaction are reduced. Consequently, generation data can be constructed by wind turbine models with capacity parameter and weather information. In Northern Ireland cases, behind-the-meter wind power is successfully separated out from user demand. The separated data of load and wind power better present the network status. Utilizing the data in demand forecasting, the forecasting errors of three general methods was effectively decreased by an average of 18.78%.