Solar energy potential assessment and investment profitability analysis: The case of North Macedonia
Year: 2026
Type: article
Abstract: This study presents a novel integrated framework that combines meteorological data analysis, machine learning forecasting models, clustering techniques, and economic modeling to offer the first data-driven investment roadmap for solar energy in North Macedonia. High-accuracy solar radiation forecasts for 2025 were generated using the Gradient Boosting Regression model (R² = 0.953) using NASA POWER data for the period 2020–2024. Cities were grouped into clusters based on their climatological characteristics; cities such as Štip, Veles, and Kavadarci demonstrated the highest solar potential and economic efficiency. Economic analyses identified payback periods of 3–5 years and average ROI (Return on Investment) values exceeding 24%. It was determined that increasing electricity prices linearly affected the ROI, reaching 33% with an increase in price to $0.12/kWh. The findings highlight the importance of a new integrated decision-support framework merging ML-based solar forecasting, spatial clustering, and dynamic economic modeling tailored for data-limited emerging markets such as North Macedonia
Author: Hakan Kaya
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7138113185