Logo Lanfrica
  • Accueil
  • Atlas
  • Analyses
  • Documentation
  • Sign in

© 2026 Lanfrica. Tous droits réservés. Tous les droits d'auteur des ressources affichées sur le site Web Lanfrica appartiennent aux détenteurs de droits d'auteur d'origine, sauf indication contraire explicite.

medlouaynjima/Tunisia-Electricity-Analysis

Domaine:

environment and energyclimate

Type de record:

project
Créateur:
med
Hôte:
Why did Tunisia's grid collapse under a 47°C heatwave? A full data science pipeline analyzing 22,000 hours of synthetic grid data to forecast demand, predict blackouts, and model intervention ROI. # ⚡ Tunisia Electricity Crisis: Data-Driven Analysis & Solutions > A complete data science pipeline analyzing Tunisia's July 2026 electricity crisis — from demand forecasting and outage prediction to smart load shedding optimization and investment portfolio modeling. --- ## The Problem In July 2026, Tunisia faced its worst electricity crisis in decades. A severe heatwave pushed temperatures to **47°C in Tunis** and **49°C inland**, driving electricity demand to a record **~6,000 MW** against an installed capacity of only **~4,630 MW**. The resulting **1,370 MW supply gap** forced STEG (Tunisia's national utility) to implement rolling blackouts across neighborhoods, causing an estimated **$4.2 billion in economic losses**. Cross-border imports from Algeria and Libya were unavailable — both countries faced their own demand surges. With **95% of generation gas-dependent**, declining domestic gas production, and STEG burdened by **7.3 billion TND in debt**, the crisis exposed deep structural vulnerabilities. --- ## What This Project Does This repository contains a fully reproducible data science pipeline built on **synthetic but realistic hourly data** spanning January 2024 to July 2026. It demonstrates how data-driven approaches can inform crisis response and long-term planning. ### Models Built | Model | Purpose | Key Output | |-------|---------|------------| | **Demand Forecaster** | Predict next-hour demand from weather + calendar features | R² = 0.95, MAE ≈ 45 MW | | **Outage Predictor** | Classify whether an outage will occur | Identifies temperature > 35°C and supply margin as top predictors | | **Smart Load Shedding Optimizer** | Allocate power cuts to minimize economic loss | **40–60% cost reduction** vs uniform blackouts | | **Solution Impact Simulator** | Model 7 interventions (solar, wind, storage, demand response, etc.) | ROI ranking, payback periods, outage prevention | | **Portfolio Optimizer** | Find best solution combination within any bud …

Visit

github.com

Licenses

MIT