Chance-constrained scheduling considering frequency support from electric vehicles under multiple uncertainties

With the share of renewable generation increasing, considerable synchronous generators have been displaced. This trend results in the decline of system inertia and thus raises the frequency secure issue. Concurrently, electric vehicles (EVs) proliferate. Numerous EVs hold a huge potential of frequency regulation in case of contingency as EVs are able to achieve primary frequency responses (PFRs) via vehicle to grid chargers. However, the value of frequency responses from aggregated EVs has not fully exploited. This paper, proposes a chance constrained scheduling model incorporating the ability of PFRs from aggregated EVs. The proposed model characterizes uncertainties associated PFR capacities from aggregated EVs and wind power forecast errors by the Wasserstein-metric ambiguity sets, which do not rely on distributional assumption, and then formulates uncertainties-related constraints as distributionally robust (DR) chance constraints. These DR chance constraints are reformulated as tractable linear programs. Consequently, the proposed model leads to a mixed-integer linear program. Numerical results on a modified IEEE 39-bus system show that incorporating PFR from EVs can reduce total costs from $4,540,396 to $4,431,233, which is reduced by 2.4%. The proposed model also achieves better reliability compared to traditional methods, demonstrating the distributionally robust optimization in enhancing electricity scheduling and rising renewable integration.