CarolN90/Sustainable-Power-Forecasting

Data-driven forecasting for Kenya’s renewable energy demand using SARIMAX and machine learning. Towards a Resilient Grid: Strategies for Variable Renewable Energy in Kenya Data-Driven Solutions for a Reliable and Green Energy Future --- Live Webapp: energyconsumptionapp.stream… >*This architecture supports scalability and modularity, making it easy to incorporate more data sources or integrate into real-time systems like Streamlit or PowerBI dashboards.* --- ## Project Overview Kenya is a global leader in renewable electricity generation, with over 80% of its power coming from sustainable sources like geothermal, wind, and solar. However, as the country accelerates toward its ambitious 100% renewable energy goal by 2050, critical challenges emerge—especially around variability, grid reliability, and equitable access. This project explores **data-driven strategies** to enhance grid stability and optimize renewable energy integration, leveraging machine learning and exploratory data analysis (EDA) on datasets sourced from Kenya's national energy agencies. --- ## Objectives This analysis seeks to answer the following key questions: 1. How have electricity consumption trends and grid connectivity evolved over time in Kenya? 2. How does energy generation by source evolve over time, and what trends emerge in the mix of renewables vs non-renewables? 3. How can we visualize Kenya’s energy landscape using interactive and static plots to reveal key insights? 4. What patterns and insights can be uncovered to inform future energy planning and improve grid resilience? --- ## Project Architecture & Data Flow The project follows a structured pipeline from data acquisition to deployment and feedback, enabling robust analysis and insights for Kenya’s energy resilience: ``` Data Extraction → Data Processing → Modeling & Analysis → Visualization & Reporting → Feedback Loop ↓ ↓ ↓ ↓ ↓ EPRA Excel → Cleaning → EDA & Stats → Jupyter Charts …