Assessment of the Cost Associated With Wind Generation Prediction Errors in a Liberalized Electricity Market

TLDR

Generators must buy or sell energy production deviations due to prediction errors when they bid in day‑ahead or hour‑ahead energy markets. The study proposes a probabilistic methodology to estimate the market energy costs for wind generators associated with wind prediction errors. The authors model wind prediction errors with a probability density function, compute hourly energy deviations and associated trading costs, and evaluate the approach on three real Spanish wind farm cases. The analysis reveals that prediction‑error costs can reach up to 10 % of a wind generator’s total energy income.

Abstract

In this paper, a probabilistic methodology for estimating the energy costs in the market for wind generators associated with wind prediction errors is proposed. Generators must buy or sell energy production deviations due to prediction errors when they bid in day-ahead or hour-ahead energy markets. The prediction error is modeled through a probability density function that represents the accuracy of the prediction model. Production hourly energy deviations and their associated trading costs are then calculated. Three study cases based on real wind power installations in Spain are analyzed. The three study cases show that the error prediction costs can reach as much as 10% of the total generator energy incomes.