Kalantar Neyestanaki, MohsenCherkaoui, Rachid2021-07-232021-07-232021-07-232021-03-2610.1109/TSTE.2021.3068630https://infoscience.epfl.ch/handle/20.500.14299/180175Relying on the power flexibility of distributed energy resources (DERs) located in an active distribution network (ADN), this ADN will be able to provide power flexibility to the upper-layer grid at their point of common coupling (PCC). The power flexibility is defined as additional bi-directional active/reactive powers a resource can provide to the grid by adjusting its operating point. In this context, this paper presents a two-stage ADN management method to deliver, at the PCC, the power flexibility that the upper-layer grid operator would request minutes-ahead real-time operation. The first stage updates the power set-points of DERs considering their offer curves as well as the uncertainties stem from the short-term forecast errors of demand and renewable generation profiles. The inter-temporal constraints and losses of the grid are accounted for by exploiting a linearized dynamic optimal power flow model, whereby the first stage is implemented as a linear scenario-based optimization problem. Then, in real-time operation, relying on a linear optimization problem, the second stage adjusts the power flexibility injection of a utility-scale battery energy storage system (ESS) to mitigate the imbalance at the PCC inherent in the above-mentioned uncertainties. The performance of the method is tested on a real ADN.Active distribution network (ADN), ADN managementancillary servicesdistributed energy resources (DERs)energy storage system (ESS)linear optimization-based control strategypower flexibilitystochastic optimizationuncertaintytext::journal::journal article::research article