Short Term Electricity Load Forecasting on Varying Levels of Aggregation

TL;DR

This paper introduces an empirical scaling law that explains how electricity load forecasting accuracy varies with aggregation levels, highlighting the benefits and limits of aggregation for different methods and horizons.

Contribution

It presents a novel empirical scaling law for load forecasting accuracy across aggregation levels, supported by a decomposition of consumption patterns.

Findings

Aggregation improves forecasting performance up to a certain point.

Beyond the optimal aggregation level, additional aggregation does not enhance accuracy.

The model applies across various forecasting methods and horizons.

Abstract

We propose a simple empirical scaling law that describes load forecasting accuracy at different levels of aggregation. The model is justified based on a simple decomposition of individual consumption patterns. We show that for different forecasting methods and horizons, aggregating more customers improves the relative forecasting performance up to specific point. Beyond this point, no more improvement in relative performance can be obtained.