Deposit_STEg_power_500 MW

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:Provides a validated, reproducible framework for optimal PV siting with stochastic OPF, useful for power systems research.

🏢実務担当者:Offers a methodology for grid integration studies that can inform utility planning and renewable project siting decisions.

🏛政策担当者:Demonstrates how to evaluate large-scale renewable integration impacts on grid stability, informing policy for renewable targets.

=====================================================================  ZENODO DEPOSIT — METADATA FIELDS  Copy-paste into the Zenodo upload form (https://zenodo.org/deposit/new) ====================================================================== --- TITLE --- Replication code and data for: "Optimal PV Siting for 500 MW Integration in the Southern Tunisian Grid: Cross-Validated Stochastic Quasi-Dynamic OPF" --- UPLOAD TYPE --- Software --- DESCRIPTION (HTML — paste into the Zenodo description box) --- <p>This repository provides the complete MATLAB simulation framework for replicating all results reported in:</p> <blockquote> Y. Ben Salem and M. Aoun, "Optimal PV Siting for 500 MW Integration in the Southern Tunisian Grid: Cross-Validated Stochastic Quasi-Dynamic OPF," 2026. </blockquote> <p>The study addresses the optimal placement of 500 MW of utility-scale photovoltaic (PV) generation across six candidate sites in the southern Tunisian 150 kV transmission network operated by STEG. This capacity is directly comparable to the 598 MW concession pipeline contracted by STEG in 2024–2025. The framework combines particle swarm optimisation (PSO) with AC optimal power flow (AC-OPF) on a validated 31-bus MATPOWER model, extending a static 300 MW companion study to a temporal, stochastic, and parametric assessment at 500 MW.</p> <h3>Contents</h3> <ul>   <li><strong>MATPOWER case file</strong> (<code>case_steg_sud_v2.m</code>): 31-bus, 150 kV network model validated against 2021 operational measurements (MAE = 1.56%, max deviation = 2.76%).</li>   <li><strong>Quasi-dynamic 24-hour sequential OPF</strong>: 216 AC-OPF simulations across 3 seasonal days × 24 hours × 3 siting strategies.</li>   <li><strong>Probabilistic PV forecast generator</strong>: Gaussian-copula-based stochastic model combining Beta-distributed marginals, AR(1) temporal correlation (ρ = 0.70), and Cholesky-factorised spatial correlation (d₀ = 100 km).</li>   <li><strong>Monte Carlo stochastic evaluation</strong>: 24,000 AC-OPF solutions over 500 irradiance scenarios, with CVaR₉₅ risk assessment.</li>   <li><strong>PSO siting optimisation</strong> with 30-restart multi-start validation (CV = 0.21%).</li>   <li><strong>Cross-validation</strong> against Genetic Algorithm (GA) and Grey Wolf Optimiser (GWO).</li>   <li><strong>Impedance sensitivity analysis</strong>: per-site robustness index I_R under ±10% network parameter uncertainty.</li>   <li><strong>Temporal re-optimisation</strong> across 5 seasonal load scenarios (f_c ∈ {0.45, 0.60, 0.75, 0.80, 1.00}).</li>   <li><strong>N-1 contingency screening</strong>: 42-branch security assessment at summer peak PV and winter peak load.</li>   <li><strong>Statistical analysis</strong>: bootstrap confidence intervals, paired Wilcoxon signed-rank tests, Cohen's d effect sizes.</li> </ul> <h3>Key findings replicated by this code</h3> <ul>   <li>A reverse power flow regime emerges at 500 MW, with a load-factor crossover at f*_c ≈ 0.77.</li>   <li>PSO-S3 reduces expected losses by 27.6 MWh/day (−4.5%) with a CVaR₉₅ cost premium of only +0.75%.</li>   <li>Four of six candidate sites exhibit I_R < 5% (very robust), carrying 94% of the optimal capacity.</li>   <li>The siting recommendation is algorithm-independent (PSO–GWO gap: 0.18%, r = 0.99).</li> </ul> <h3>Requirements</h3> <ul>   <li>MATLAB R2024b or later</li>   <li>MATPOWER 8.1 (<a href="https://matpower.org">https://matpower.org</a>)</li>   <li>Statistics and Machine Learning Toolbox</li>   <li>Optimisation Toolbox (for GA comparator)</li> </ul> <h3>Reproducibility</h3> <p>All random seeds are fixed (<code>rng(42)</code>). A master script (<code>results/run_all.m</code>) reproduces all 16 tables and 10 figures from the manuscript. Expected runtime: ~45 minutes on an Intel Xeon workstation (single-threaded).</p> --- AUTHORS --- 1. Ben Salem, Yassine — University of Gabès, Tunisia (ORCID: xxxx-xxxx-xxxx-xxxx) 2. Aoun, Mohamed — University of Gabès, Tunisia (ORCID: xxxx-xxxx-xxxx-xxxx) --- AFFILIATIONS --- MACS Laboratory (Modelling, Analysis and Control of Systems), National Engineering School of Gabès, University of Gabès, Tunisia --- LICENSE --- MIT License --- KEYWORDS --- PV siting; Particle swarm optimisation; AC optimal power flow; Stochastic simulation; MATPOWER; Tunisian transmission network; STEG; Renewable energy integration; Monte Carlo; CVaR; Gaussian copula; Grey Wolf Optimiser --- RELATED IDENTIFIERS --- Type: "Is supplement to" Identifier: [DOI of the published article, once available] Type: "References" Identifier: [DOI of the companion paper, once available] --- GRANTS --- [Leave empty — the study did not receive dedicated external funding] --- COMMUNITIES --- Consider adding to:   - "Energy Research"   - "Power Systems"   - "Open Energy Modelling" --- VERSION --- 1.0.0 --- LANGUAGE --- English --- NOTES --- Bus names in the MATPOWER case file use internal numbering to preserve the confidentiality of unpublished STEG operational data. The electrical parameters (impedances, ratings, generation costs) are unchanged from the validated model.

• Zenodo https://zenodo.org/records/21968397first seen 2026-08-17 04:33:08 · last seen 2026-08-19 04:14:20