Electric Vehicle Aggregator as an Automatic Reserves Provider in the European Market Setting

TL;DR

This paper introduces new models for electric vehicle aggregator scheduling that incorporate reserve activation uncertainty using real European data, significantly improving upon traditional deterministic approaches.

Contribution

It develops and compares deterministic, stochastic, and robust scheduling models for EV aggregators considering reserve activation uncertainty in European markets.

Findings

Deterministic models poorly represent reserve activation uncertainty.

Stochastic and robust models significantly improve scheduling accuracy.

Incorporating uncertainty enhances aggregator profitability and user comfort.

Abstract

Shift of the power system generation from the fossil to the variable renewable sources prompted the system operators to search for new sources of flexibility, that is, new reserve providers. With the introduction of electric vehicles, smart charging emerged as one of the relevant solutions. However, electric vehicle aggregators face the uncertainty of reserve activation on one side and electric vehicle availability on the other. These uncertainty can have a negative effect on both the aggregators' profitability and their users' comfort.State-of-the art literature mostly neglects the reserve activation or the related uncertainty. Also, they rarely model European markets or use real balancing data. This paper introduces a new method for modeling the reserve activation uncertainty based on actual data for the European power system. Three electric vehicle scheduling models were designed and tested: the deterministic, the stochastic and the robust one. The results demonstrate that the current deterministic approaches inaccurately represent the activation uncertainty and that the proposed models that consider uncertainty, both the stochastic and the robust one, substantially improve the results.