A Flexible Storage Model for Power Network Optimization

This paper develops a simple but flexible storage model for use in a variety of multi-period optimal power flow problems. The proposed model is designed for research use in a broad assortment of contexts enabled by the following key features: (i) the model can represent the dynamics of an energy buffer at a wide range of scales, from residential battery storage to grid-scale pumped hydro; (ii) it is compatible with both balanced and unbalanced formulations of the power flow equations; (iii) convex relaxations and linear approximations allow seamless integration of the proposed model into applications where convexity or linearity is required are developed; (iv) a minimalist and standardized data model is presented, to facilitate ease of use by the research community. The proposed model is validated using a proof-of-concept 24h storage scheduling task that demonstrates the value of the model's key features. An open-source implementation of the model is provided as part of the PowerModels and PowerModelsDistribution optimization toolboxes.