Decentralised grid architectures for electricity distribution networks: A survey

The rapid growth of distributed energy resources (DER) such as rooftop photovoltaics (PV), battery storage, electric vehicles (EV), and flexible loads, is shifting power system coordination from centralised control centres to millions of prosumers and local controllers at the distribution level. This transition has led to many new coordination approaches across control, market, and learning-based models, which are often described as decentralised. However, this term is applied inconsistently: it may refer to decomposed optimisation, edge computing, peer-to-peer (P2P) trading, or distributed ledger technology, obscuring what is actually being decentralised, authority, computation, information, or topology. Existing surveys typically address one such concept in isolation, for example, microgrid control structures, energy management system (EMS) topologies, or market designs, without providing a unified, multi-dimensional view across the full coordination landscape. In this survey, we propose a six-tier graduated decentralisation scale for distribution level coordination architectures, accompanied by a set of classification criteria that we apply to systematically map and compare recent architectures. We discuss how topology, decision-making, autonomy, intelligence, information flow, and coordination mechanisms evolve as architectures move from centralised to more decentralised operation. We further identify concrete research gaps, and outline future directions for deployment grade, multi-actor grid coordination.