MEO37th/ML-Nigeria-330kv-grid-optimization-wind-integration

Wind integration analysis across Nigerian buses shows stable voltage, reduced losses, and strong economic gains. utilizing machine learning and python logic # Strategic Wind Integration for the Nigerian 330kV Grid **Author:** Michael Obuma **Repository:** ML-Nigeria-330kv-grid-optimization-wind-integration A high-fidelity power systems simulation project to determine optimal locations for wind energy injection to improve voltage stability, reduce system losses, and enable cost-saving thermal generation displacement on Nigeria's 330kV transmission network. ## Project Overview This research simulates the incremental integration of Doubly-Fed Induction Generator (DFIG) wind plants at key healthy buses within the Nigerian 330kV grid. The primary objective is to analyze the grid's response and identify the most effective locations for wind farms that not only add renewable capacity but also actively solve existing operational challenges. The simulation adheres to strict statutory voltage limits (0.95-1.05 pu) and models the proportional reduction of thermal power plants to quantify cost savings and ensure system balance. ## 🎯 Key Findings ### Best Performing Locations - **Gwagwalada (Bus 35):** Minimum losses of 8.11 MW at 30 MW wind integration - **Jebba TS (Bus 18):** Minimum losses of 8.16 MW at 30 MW wind integration - **Katampe (Bus 37):** Minimum losses of 8.16 MW at 30 MW wind integration - **Kaduna (Bus 32):** Minimum losses of 8.16 MW at 10 MW wind integration ### Economic Impact - **Maximum Cost Savings:** ₦78.97 million per hour - **Integration Capacity:** Up to 90 MW per location - **Voltage Stability:** Maintained across 0.95-1.05 pu range - **Fuel Displacement:** 1:1 wind-to-thermal ratio achieved ## The Challenge: Real-World Data Conditioning A significant part of this project involved overcoming the challenges presented by the raw grid data. A direct simulation was impossible due to two fundamental issues: 1. **Power Imbalance:** The initial data showed a ~458 MW surplus in scheduled generation compared to total load, creating an unsolvable base case. 2. **Numerical Instability:** The line data cont …