- Diagnosis of the Senegalese electricity transmission network- SENELEC

Abstract

The Senegalese electricity transmission network is undergoing a massive structural and technological transition, driven by an average annual demand growth of 11.9%, the rapid integration of intermittent renewable energies (targeting 40% by 2030), and the national Gas-to-Power strategy. These dynamics generate highly volatile and bidirectional power flows, shifting the system’s vulnerabilities towards static security and rendering traditional snapshot-based grid planning obsolete. This thesis addresses this challenge by providing a comprehensive prospective diagnosis of the interconnected grid for the 2025 reference year (culminating in a peak demand of 1,261 MW), achieved through the development of a fully automated digital twin in the open-source Python/Pandapower environment. To ensure industrial-grade accuracy and replicate physical grid operations, automated dynamic controllers (such as On-Load Tap Changers and shunt reactors) were explicitly modeled. The digital twin was rigorously calibrated against PSS/E using single-objective and multi-objective metaheuristic algorithms (Differential Evolution and Optuna). Crucially, this optimization framework acted as a physical anomaly detector, identifying and rectifying an underlying structural discrepancy in the transformer MVA base conversion. Once analytically corrected, the Pandapower model achieved near-perfect alignment with PSS/E reference data, reducing the nodal voltage Mean Absolute Error (MAE) to a negligible 0.0005 p.u. Leveraging this validated model, chronological steady-state (N-0) and contingency (N-1) simulations were executed over the 8,760 hours of the reference year. The systemic analysis revealed a deeply polarized grid. During off-peak periods, the lightly loaded 225 kV infras- tructure behaves as a massive capacitor, generating a net reactive power surplus (Ferranti effect) that forces urban 90 kV substations into chronic overvoltage states (reaching up to 1.15 p.u.). Conversely, evening load peaks and N-1 contingencies trigger critical bottlenecks on the Dakar peninsula, with evacuation corridors facing severe cascading overloads up to 154% of their ampacity. Furthermore, the analysis highlighted severe localized vulnerabilities, such as radial antenna weaknesses in the southern Casamance grid, where a single line trip systematically leads to localized blackouts. To secure the network, a multi-horizon operational and infrastructural roadmap was formulated. In the short term, zero-CAPEX preventive line switching and optimized static retuning of primary transformer off-load taps are proposed to provide immediate relief. For the medium term, the algorithmic sizing and deployment of 75 MVAr of Variable Shunt Reactors (VSR) at strategic nodes (Patte d’Oie and Tobene) is validated to definitively neutralize capacitive overvoltages. Finally, as a long-term structural solution, High-Temperature Low-Sag (HTLS) reconductoring is recommended for the congested 90 kV industrial corridors. This innovative approach doubles thermal ampacity while retaining the existing pylons, thereby avoiding the significant delays associated with building a new line.