Distributionally Robust Joint Planning of Coastal Distribution Network and PV-Storage-EV Stations

TL;DR

This paper develops a tri-layer distributionally robust optimization framework for coastal distribution networks with PV, storage, and EV stations, effectively managing uncertainties to reduce costs and emissions.

Contribution

It introduces a novel tri-layer optimization model with a relaxation approach and an inexact CCG algorithm for robust planning of coastal energy infrastructure.

Findings

Reduces total investment and operational costs under uncertainty.

Ensures robust low-carbon planning in coastal distribution networks.

Demonstrates effectiveness on a real 47-node Chinese coastal network.

Abstract

The rapid integration of renewable energy resources, such as tidal and photovoltaic (PV) power, coupled with the growing deployment of electric vehicle (EV) charging infrastructure, necessitates coordinated planning for coastal urban distribution networks (DN). This paper presents a tri-layer distributionally robust optimization framework to jointly optimize the sitting of PV-storage-EV stations (PSES) and the configuration of coastal DNs, addressing uncertainties related to power load, PV generation, and EV charging demands. At the upper layer, optimal PSES siting and network topology decisions are made to minimize total investment and operational costs. The middle-layer formulation tackles worst case uncertainty scenarios via the optimal power flow model, utilizing ambiguity sets to capture correlated uncertainties. To handle non-convexities introduced by binary variables for energy storage systems, we propose and rigorously prove the exactness of a novel relaxation approach. At the lower layer, considering the dynamic pricing driven by tidal energy fluctuations, operational decisions-including electricity procurement and carbon emissions are optimized. An inexact column-and-constraint generation (i-CCG) algorithm is developed for efficient problem-solving. Numerical results from a realistic 47-node coastal DN in China illustrate that the proposed method effectively reduces costs and ensures robust, low-carbon planning under substantial uncertainties.