Artificial Intelligence for Early-Stage Planning of Carbon Capture, Utilization, and Storage (CCUS) Networks in Colombia: A Geospatial and Generative Modeling Framework

This working paper presents an integrative methodological framework for early-stage planning of Carbon Capture, Utilization, and Storage (CCUS) networks in Colombia. The framework combines geospatial clustering, machine learning forecasting, and generative artificial intelligence (AI) to support the identification of emission hotspots, the prediction of CO₂ flows, and the conceptual simulation of logistics configurations. Historical emissions data (2000–2022), spatial infrastructure layers, and sectoral indicators are used to define high-priority clusters through geospatial intelligence methods such as kernel density estimation and DBSCAN. Long Short-Term Memory (LSTM) and XGBoost models are employed to forecast CO₂ capture volumes under different macroeconomic and regulatory scenarios, while SHAP values improve model interpretability. Generative AI (GPT-4o) is integrated to simulate context-aware supply chain designs, yielding conceptual configurations that can guide stakeholder engagement and territorial planning. The proposed methodology provides a decision-support architecture that bridges spatial data analytics and artificial reasoning to inform CCUS deployment in fragmented and data-scarce environments. The study contributes to the international discourse on low-carbon transitions by offering a scalable and adaptable planning toolkit, grounded in Colombia’s institutional and territorial realities.