Grid-scale Fluctuations and Forecast Error in Wind Power

TL;DR

This paper analyzes wind power fluctuations in the Irish grid, revealing self-similar, correlated turbulence structures, and introduces a simple method to reduce forecast errors related to temporal correlations.

Contribution

It identifies the self-similar structure of wind power fluctuations and proposes a memory kernel to improve forecast accuracy without prior model knowledge.

Findings

Wind power fluctuations exhibit self-similar, correlated turbulence.

A simple memory kernel reduces forecast errors related to timescale and scaling.

The approach improves forecast reliability in wind power generation.

Abstract

The fluctuations in wind power entering an electrical grid (Irish grid) were analyzed and found to exhibit correlated fluctuations with a self-similar structure, a signature of large-scale correlations in atmospheric turbulence. The statistical structure of temporal correlations for fluctuations in generated and forecast time series was used to quantify two types of forecast error: a timescale error ($e_{\tau}$) that quantifies the deviations between the high frequency components of the forecast and the generated time series, and a scaling error ($e_{\zeta}$) that quantifies the degree to which the models fail to predict temporal correlations in the fluctuations of the generated power. With no $a$ $priori$ knowledge of the forecast models, we suggest a simple memory kernel that reduces both the timescale error ($e_{\tau}$) and the scaling error ($e_{\zeta}$).