Understanding Risk and Revenue in the Nordic 15-minute mFRR market: An EV Aggregation Study

TL;DR

The paper addresses how EV fleets can participate in the Nordic 15-minute mFRR Energy Activation Market by framing fleet flexibility as a virtual battery and applying a risk-aware stochastic optimisation that co-optimises day-ahead scheduling with quarter-hour mFRR bidding. The approach decomposes profits into mFRR activations, day-ahead energy costs, and imbalance settlements, and uses a CVaR-based risk objective $(1-eta) ext{E}_oldsymbol{ extomega}[oldsymbol{ ext{Pi}}] + eta ext{CVaR}_{oldsymbol{ extalpha}}(oldsymbol{ ext{Pi}})$ to balance expected returns against downside risk. Across two representative day-ahead price paths, co-optimisation improves both expected profit and the lower tail, primarily by buying less energy day-ahead and shifting procurement toward mFRR down, while flattening the charging profile to preserve eligibility for mFRR up. The study provides an operational guide for EV aggregators bidding in the Nordic mFRR market and highlights two extensions: rolling 45-minute re-optimisation and a V2G framework to further enhance revenue opportunities and manage degradation. Overall, the results suggest that coordinated day-ahead and mFRR participation can meaningfully increase profitability for EV aggregators in short-MTU balancing markets and offer practical guidance for implementation.

Abstract

Decarbonisation, decentralisation, and intermittency are driving the development of flexibility markets towards shorter market time units (MTU). Shorter MTUs and shorter gate closures lower the entrance barriers of demand side aggregators that face significant uncertainty on longer time scales. We study the business case for aggregated EV fleets participating in the Nordic 15-minute mFRR Energy Activation Market (EAM). Motivated by increasing system granularity and rapid EV uptake, we represent fleet flexibility as a virtual battery with time-varying power and energy envelopes and formulate a risk-aware stochastic optimisation that co-ordinates day-ahead scheduling with quarter-hour mFRR bidding. Using synthetic residential charging cohorts and observed day-ahead prices on two stylised days, we compare an independent day-ahead baseline to a co-optimised strategy under conservative availability and a CVaR-augmented objective. Across both price cases, co-optimisation increases expected profit and lowers downside risk: the model buys less energy day-ahead and shifts procurement toward mFRR down while flattening the charging plan to retain eligibility for mFRR up. Profit decomposition shows that the uplift is driven by higher mFRR down revenues and reduced reliance on unwinding day-ahead positions. We discuss operational implications for bidding and outline two extensions: rolling 45-minute re-optimisation and a V2G framework.

Figures (4)

• Figure 1: The quarter-hour timeline for mFRR: gate closure (current QH--45, transitioning to QH--25), notification, ramp up, full activation, and ramp down. Start and end of the operating quarter-hour in red, dashed lines.
• Figure 2: Individual EV charging parameters aggregated into a virtual battery, adapted from brinkel2023.
• Figure 3: Double-peak day. Most-likely scenario shown in both panels for like-for-like comparison between (a) independent and (b) co-optimised bidding; panels plot charging vs. envelopes, mFRR/imbalance volumes, and prices.
• Figure 4: Duck-curve day. Most-likely scenario shown in both panels for like-for-like comparison between (a) independent and (b) co-optimised bidding; panels plot charging vs. envelopes, mFRR/imbalance volumes, and prices.