Extreme prices in electricity balancing markets from an approach of statistical physics

TL;DR

This paper models the Italian electricity balancing market using statistical physics and agent-based simulations to analyze how increased renewable energy sources influence price fluctuations and the likelihood of extreme price spikes.

Contribution

It introduces a novel agent-based modeling approach to forecast price and volume fluctuations in balancing markets with high renewable integration, highlighting the rise in extreme price events.

Findings

Average prices are only slightly affected by renewables.

Probability of extreme price spikes increases with higher renewable share.

Renewables contribute to increased cost variability and peak risks.

Abstract

An increase in energy production from renewable energy sources is viewed as a crucial achievement in most industrialized countries. The higher variability of power production via renewables leads to a rise in ancillary service costs over the power system, in particular costs within the electricity balancing markets, mainly due to an increased number of extreme price spikes. This study focuses on forecasting the behavior of price and volumes of the Italian balancing market in the presence of an increased share of renewable energy sources. Starting from configurations of load and power production, which guarantee a stable performance, we implement fluctuations in the load and in renewables; in particular we artificially increase the contribution of renewables as compared to conventional power sources to cover the total load. We then forecast the amount of provided energy in the balancing market and its fluctuations, which are induced by production and consumption. Within an approach of agent based modeling we estimate the resulting energy prices and costs. While their average values turn out to be only slightly affected by an increased contribution from renewables, the probability for extreme price events is shown to increase along with undesired peaks in the costs.