Techno-economic assessment of centralized and decentralized energy management strategies for energy sharing in collective self-consumption schemes

Abstract

The increasing deployment of individual and collective self-consumption systems is reshaping Energy Management Systems (EMSs) under evolving regulatory frameworks. This paper presents a techno-economic comparison between a centralized EMS and a decentralized EMS for flexible resources dispatching and sharing under collective self-consumption schemes. The centralized EMS is formulated as a Mixed-Integer Non-Linear Programming (MINLP) optimization problem, whereas the decentralized EMS employs a rule-based algorithm that requires no information exchange among members. Both strategies have been evaluated under the Spanish regulatory framework, a) using fixed allocation coefficients and b) introducing improvements borrowed from the Portuguese regulation, selected as a benchmark due to its advanced regulatory maturity. For the case of ex-ante allocation coefficients computation, an optimization-based methodology is proposed combining Mixed-Integer Linear Programming (MILP) with data clustering techniques. Results indicate that both EMS architectures achieve comparable energetic performance. The centralized EMS achieves the highest levels of self-consumption, self-sufficiency and energy sharing, particularly when proportional allocation coefficients are used, while the decentralized EMS performs closely. From an economic perspective, the centralized EMS provides the highest cost reductions, while the decentralized EMS yields lower economic savings but with significantly less computational effort, with runtimes up to eighteen times shorter. These findings highlight a clear trade-off between economic optimality and computational efficiency, positioning decentralized EMS solutions as a scalable and privacy-preserving alternative for individual self-consumers transitioning to collective self-consumption schemes in evolving regulatory frameworks.