stilhere4huniid/Mall-of-Zimbabwe-Energy-Digital-Twin

A physics-based Digital Twin simulating the operational energy future of the 90,000m² Mall of Zimbabwe. Uses Python & XGBoost to forecast demand, test solar microgrids, and identify $1.66M in annual savings (Pre-Construction Feasibility Study). # 🇿🇼 Mall of Zimbabwe: Predictive Energy Digital Twin > **Strategic Feasibility Study:** Predicting the operational energy future of the 90,000m² Mall of Zimbabwe before construction begins in 2026. ## 🏙️ Executive Summary WestProp Holdings is developing the Mall of Zimbabwe, a $100M+ "Smart City" asset in Harare. To secure anchor tenants and optimize Capital Expenditure (CapEx), we built a **Digital Twin** to simulate the mall's energy consumption, costs, and grid resilience. **The Result:** We identified **$1.66 Million** in annual savings by integrating a 4MW Solar Microgrid and AI-driven HVAC controls into the pre-construction blueprints. ## 📊 Key Results | Metric | Baseline (Standard Build) | Optimized (Smart Build) | Impact | | :--- | :--- | :--- | :--- | | **Annual Demand** | 26.6 GWh | 14.5 GWh | **-45%** | | **Annual Bill** | $3.7 Million | $2.0 Million | **$1.66M Saved** | | **Grid Reliance** | 94% (High Risk) | 60% (Resilient) | **Energy Security** | ## 🧠 The AI Model Showdown We benchmarked three forecasting approaches to validate our financial projections. * **🏆 XGBoost (Winner):** 97.9% Accuracy (MAPE: 2.07%). Handling ZESA load shedding patterns best. * **❌ LSTM (Deep Learning):** 94.6% Accuracy. Slower to adapt to sudden grid outages. * **❌ SARIMA (Statistical):** 88.2% Accuracy. Good for seasonality but failed to capture holiday-specific spikes. * **❌ Prophet (Baseline):** 69.0% Accuracy. Failed to capture complex operational hours. ## 📂 Project Structure ├── Data/ # Raw & Calibrated Simulation Data ├── Visualizations/ # EDA & Model Performance Charts ├── Deliverables/ # Client-Facing Reports │ ├── Mall_of_Zimbabwe_Strategic_Brief.pdf (Executive Memo) │ └── strategic_energy_dashboard.html (Interactive Dashboard) └── Project_Notebook.ipynb # Full Source Code ## ⚡ Quick Start To reproduce the results or run the simulation on your local machine, follow these steps: # 1. Clone the Repository ` …