Long-Term Load Forecasting Considering Volatility Using Multiplicative Error Model

TL;DR

This paper introduces a novel long-term load forecasting method using the multiplicative error model (MEM), which accounts for volatility and improves forecast accuracy during economic fluctuations.

Contribution

It is the first to apply MEM to long-term load forecasting, incorporating volatility dynamics for more accurate predictions.

Findings

MEM outperforms traditional models in forecast accuracy

Volatility consideration improves predictions during economic downturns

Model demonstrates high directional accuracy during 2008 recession

Abstract

Long-term load forecasting plays a vital role for utilities and planners in terms of grid development and expansion planning. An overestimate of long-term electricity load will result in substantial wasted investment in the construction of excess power facilities, while an underestimate of future load will result in insufficient generation and unmet demand. This paper presents first-of-its-kind approach to use multiplicative error model (MEM) in forecasting load for long-term horizon. MEM originates from the structure of autoregressive conditional heteroscedasticity (ARCH) model where conditional variance is dynamically parameterized and it multiplicatively interacts with an innovation term of time-series. Historical load data, accessed from a U.S. regional transmission operator, and recession data for years 1993-2016 is used in this study. The superiority of considering volatility is proven by out-of-sample forecast results as well as directional accuracy during the great economic recession of 2008. To incorporate future volatility, backtesting of MEM model is performed. Two performance indicators used to assess the proposed model are mean absolute percentage error (for both in-sample model fit and out-of-sample forecasts) and directional accuracy.