Dynamic electricity prices create opportunities for households to reduce electric vehicle (EV) charging costs, but uncertain driving demand and the option to postpone charging complicate decisions about when and how much to charge.
- Speaker
- Date
- Friday 6 Nov 2026, 12:00 - 13:00
- Type
- Seminar
- Room
- ET-14
- Building
- E Building
We study this problem using a finite-horizon Markov decision process that accounts for charging-rate limits, fixed costs per charging session, and costly public charging when battery energy is insufficient. We characterise the optimal charging policy and identify conditions under which the decision to charge follows a simple state-of-charge (SOC) threshold.
We also show that such threshold structures can fail, highlighting the limitations of commonly used charging rules. In numerical experiments with Dutch day-ahead prices, the overnight charging-cost function is nearly linear over typical charging amounts, and threshold and full-or-nothing policies perform close to the optimal policy.
However, their losses increase when the availability of low-price overnight charging is limited relative to charging needs. Our results suggest that simple charging rules can be effective when sufficient low-price charging opportunities are available, while adapting charging amounts to the vehicle's SOC and overnight electricity prices becomes more valuable as marginal overnight charging costs increase.
About the speaker
Tobias Kers is a 3rd-year PhD-candidate at the Econometric Institute supervised by dr. Ece Karakoyun, prof. dr. Rommert Dekker and prof. dr. Wilco van den Heuvel. His research focuses on problems related to the energy transition..
Coordinators
See also
- More information
Lunch will be provided (vegetarian option included).
For more information please contact the Secretariat Econometrics at eb-secr@ese.eur.nl
