Vehicle-to-Grid Aggregation in Swedish Electricity Markets Under Real-World Availability and Infrastructure Constraints - A Market-Based Optimization Analysis of EV Fleet Flexibility in Electricity Markets

Abstract

The increasing electrification of the transport sector presents both challenges and opportunities for modern power systems. While uncontrolled electric vehicle (EV) charging may increase peak demand and place additional stress on electricity networks, coordinated EV charging and vehicle-to-grid (V2G) operation offer potential flexibility services that can support grid stability and create economic value. This thesis evaluates the economic potential of aggregated EV-based flexibility in Swedish electricity markets from an aggregator perspective. A deterministic optimization framework was developed in Python using Pyomo and solved with Gurobi to model individual EV charging and discharging behavior under real-world mobility constraints. The analysis is based on event-level charging and state-of-charge (SOC) data from 152 EVs, of which 149 were successfully optimized, covering approximately 2.5 years of observed operation. Vehicle mobility behavior was converted from event-based data into hourly availability profiles through a preprocessing pipeline including event merging, SOC anchor reconstruction, and validation against the original dataset. Three market participation strategies were evaluated relative to a smart charging (V1G) reference: spot-market arbitrage, frequency containment reserve (FCR) participation, and co-optimized participation in both markets. Economic performance was assessed under Swedish SE3 day-ahead and FCR market conditions, while accounting for operational constraints including charging power limits, charging efficiency, home availability, grid fees, and simplified battery degradation costs. The results show that pure spot-market arbitrage generated a limited fleet-level benefit of 125 kSEK relative to the V1G reference case. In contrast, reserve-market participation created substantially higher value, with the FCR-only strategy generating 12.44 MSEK and the co-optimized strategy 12.56 MSEK in fleet value. Sensitivity analyzes demonstrate that spot-market profitability is highly dependent on price volatility and other operational varieties, whereas reserve-driven strategies remain comparatively more robust than pure spot-market arbitrage to changes in grid fees, availability perturbations, and conservative market assumptions for the analyzed time period. The findings suggest that under the modeled assumptions, the economic value of aggregated EV flexibility in Sweden is primarily driven by ancillary service participation rather than energy arbitrage.