Employing ARIMA models to improve wind power forecasts: A case study in ERCOT

Abstract

Wind power is growing significantly due to its favorable characteristics such as cost-effectiveness and environment-friendliness. This growth of integrating wind energy into the power grid imposes multiple challenges of scheduling and operations. The highly stochastic nature of wind energy due to the intermittent nature of wind speed makes it difficult to dispatch. Hence, it becomes more difficult to keep the balance of supply and demand. In this paper, we focus on the uncertainty in dispatch. Wind Power Forecasts (WPFs) are important for efficient dispatch and unit-commitment (UC). WPF improvement techniques include aerodynamic atmospheric models and time-series based model. This paper presents improvements to WPF using an autoregressive integrated moving average (ARIMA) model on available historical data of hourly wind power data - forecast and actual - from the Electric Reliability Council of Texas (ERCOT).