Quantification and forecasting of reserve capacity from electric trains

This paper explores the quantification and forecasting of reserve capacity from electric trains for participation in power system ancillary service markets. We first map train electricity consumption – traction and non-traction – to suitable reserve products, considering operational and regulatory constraints. Using historical data from the Danish railway operator DSB, we estimate the available flexibility for frequency containment reserves, focusing on controllable non-traction loads such as heating and air conditioning. To support market participation, we develop a low-resolution stochastic forecasting model based on conformal prediction, capable of estimating reserve availability for both day-ahead and hour-ahead horizons. Results show that a fleet of approximately 60 active trains can provide up to 10 MW of downward regulation and 1.5 MW of upward regulation from non-traction loads. Additionally, traction power from 25 trains can provide up to 5 MW of upward reserve in certain time periods. The findings demonstrate a viable pathway for integrating electric trains into flexibility markets, offering new revenue opportunities for operators and enhancing grid stability. • Demonstrates the feasibility of using electric trains to provide Frequency Containment Reserve. (FCR-D). • Introduces a forecasting framework using conformal prediction to meet regulatory P90 compliance. • Quantifies reserve capacity from real-world Danish railway data under operational constraints. (data specificity). • Shows minimal impact on schedules and comfort when participating in ancillary service markets. (market readiness). • Advances sector coupling by integrating railway operations into future energy flexibility systems.