Quantifying the Economic and Grid Reliability Impacts of Improved Wind Power Forecasting

Wind power forecasting is an important tool in electricity market and power system operations. Accurate wind power forecasts are important in reducing the occurrence and length of curtailments, enhancing the market efficiency, and improving the operational reliability at the bulk power system level. This research quantifies the value of wind power forecasting improvements in the IEEE 118-Bus test system for the MISO, CAISO, and ISO-NE balancing areas. To measure the economic value, the commercially available production cost modeling tool PLEXOS was used to simulate the multi-timescale unit commitment (UC) and economic dispatch (ED) process for calculating the cost savings and curtailment reductions. To measure the reliability improvements, a software tool developed by the National Renewable Energy Laboratory (NREL) named FESTIV was used to calculate the system's Area Control Error (ACE) performance. The tools are cross-validated for robust inference on the value of improved wind power forecasting. With its value strongly tied to the network size, market structure, and the penetration level of wind energy, a Sobol Sequences Experiment Design was evaluated to fully capture these impacts. A total of 288 scenarios were evaluated to accommodate the variation of the following four factors: generation mixture, wind penetration level, wind forecast improvements, and ramp forecast improvements. Wind penetration levels were varied from 5% to 50% to capture a full range of future penetration levels. All of the designed scenarios were tested on three different modified IEEE 118-bus test systems with one-year of data at multiple timescales, including the day-ahead UC, four hour-ahead UC, and 5-min real-time dispatch.