A Stochastic Multi-period AC Optimal Power Flow for Provision of Flexibility Services in Smart Grids

The accelerated penetration of distributed energy resources (DER), most notably renewable distributed generation (DG) like wind and solar, in distribution networks (DNs) lead mainly to voltage and congestion issues. The unavoidable transition of passive DNs toward smart grids calls for fast-deployment of cost-effective flexible options, both grid controllable assets (e.g. on-load tap changer transformers) and DER (e.g. DG, storage batteries and flexible loads), to enable distribution system operators (DSOs) in maintaining reliable operation. To this end, this paper proposes a novel centralized stochastic multi-period AC optimal power flow (S-MPOPF) model allowing DSOs to procure at minimum cost from DER ancillary services for voltage control and congestion relief in medium voltage DNs. The main novelty lies on the joint optimization of a comprehensive set of flexible options. The full-flexibility options model is formulated as a mixed-integer nonlinear programming (MINLP) problem while some combinations of flexible options lead to simpler NLP problems. Furthermore, a meaningful approximation of MINLP problem as an NLP is explored. The results obtained for the optimal operation of a modified 34-bus benchmark DN on 24 hours demonstrate the relative effectiveness of various flexible options as well as that off-the-shelf MINLP and NLP solvers require generally low computation effort.