Answers are grounded in this paper and cite their sources.
Long-term energy-storage planning in distribution grids involves long time series, non-linear network constraints, and mixed-integer siting and sizing decisions. Especially detailed AC formulations are computationally challenging for long optimisation horizons. To address this challenge, we propose an operating-point-dependent linearisation of the AC power-flow equations using sensitivity coefficients and integrating them into a mixed-integer linear program for the joint siting, sizing and dispatch of different flexibility options over long time horizons. An iterative AC power-flow validation and relinearisation procedure is used to ensure approximation accuracy. Different approaches to select the linearisation points are benchmarked, either by using each operating point or by clustering approaches, using k -medoids while retaining the original injection and network data. The method is evaluated on SimBench distribution networks and benchmarked against non-linear AC-OPF formulations. Operating-point-dependent linearisation enables computationally tractable storage planning over long time horizons while preserving detailed network information. Clustering can reduce computational effort by about 40 % , provided that the original operating-point data are retained. The linearised model deviates from the lossless AC-OPF benchmark by approximately 0.3 % in storage power and 0.6 % in energy capacity. To demonstrate the use of the method, we develop a case study where household flexibility and grid scale storage are co-optimised in a full-year planning problem. The results show that aggregate household flexibility primarily affects storage sizing, whereas its spatial distribution influences local grid conditions and optimal siting.
No takes yet. Share an insight, caveat, or question.
Adelmann et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: