Optimal BESS Scheduling for Multi-Market Participation in the Nordics

TL;DR

This paper presents a unified framework using GAMs and stochastic optimization for BESS scheduling across multiple Nordic electricity markets, improving revenue forecasting and market participation strategies.

Contribution

It introduces a novel integrated approach combining advanced forecasting with stochastic optimization for multi-market BESS scheduling in Nordic countries.

Findings

Spot markets are more predictable than frequency markets.

Forecast errors have limited impact on bid acceptance.

The framework helps optimize revenue and participation strategies.

Abstract

Battery energy storage systems BESSs can provide fast frequency reserves and energy arbitrage in Nordic electricity markets but their limited energy capacity requires accurate revenue forecasts and coordinated bidding across multiple submarkets. This paper introduces a unified framework that employs generalized additive models GAMs to generate one week ahead forecasts of spot and FCRN revenues in Denmark, Finland, and Norway using three years of hourly price and volume data. Forecast outputs feed a stochastic mixed integer optimizer that co-optimizes BESS participation in FCRN, FCRD, spot markets, subject to state of charge constraints, inverter losses, and differing pay as bid and pay as clear rules. Comparative analyses evaluate forecast accuracy and quantify the impact of forecast errors on BESS bid acceptance, market selection, and profitability under realistic seasonal price patterns. Results demonstrate that spot markets exhibit consistently higher predictability than frequency markets, and that forecast errors modestly affect bid acceptance but do not alter overall market participation. The proposed approach provides BESS operators and investors with a tool to assess revenue uncertainty and optimize multi market strategies in Nordic power systems.