Code and data for the paper: "Event-informed Identification and Allocation of Distribution Network Planning Candidates with Influence Scores and Binary Linear Programming"

This repository contains datasets and code with a novel numerical approach aimed at finding a distribution network expansion plan (DNEP) that prevents future congestion and voltage issues. This approach is tested using the modified IEEE 33-bus network. Electricity demand and PV production data for a leap year with a 1-minute resolution was generated using the CREST model from the Loughborough University and is provided as a dataset of future high-load and high-production scenario. Using OpenDSS from the Electric Power Research Institute (EPRI), a power flow simulation was performed to identify constraint violation events (i.e., if there is over/undervoltage, or line overloading at each time step), the resulting dataset with line loading and voltage as a percentage of the limits is also available in this repository. The thermal and voltage constraint violation events (i.e., the duration and intensity of events) are used to determine a pool of potential candidates for infrastructure (i.e., line/cable) upgrade, voltage regulator, and energy storage system installations. A script is made available to dimension expansion planning candidates, and to obtain the minimum-cost list of these candidates that solves all forecasted constraint violation events using influence scores and binary linear programming. Three pools of candidates and three cost projections were considered. Further information is available in the description of datasets and scripts.