A bottom-up urban building energy model for evaluating thermal load electrification measures

Abstract

Urban Building Energy Modelling offers critical insights for urban decarbonization planning, but its adoption remains limited due to data challenges and model complexity. This paper introduces a fully automated, bottom-up UBEM framework that integrates LiDAR-derived 3D building geometry, tax and permit records, DOE and RECS-informed building templates, and utility data. Building models are generated and auto-calibrated using a computationally efficient 5R1C resistor-capacitor thermal model, augmented with a derivative-free optimization algorithm (BOBYQA) to minimize calibration error against seasonal energy billing data. Applied to the City of Ithaca, NY, the framework simulates over 5,000 buildings within minutes on a standard laptop, achieving sub-20% MAPE in calibrated zones. The model supports scenario-based analysis of electrification, envelope upgrades, rooftop PV adoption, and incentive schemes, incorporating detailed retrofit cost and payback calculations. This workflow demonstrates a scalable pathway for high-resolution, financially aware UBEMs suitable for deployment in small–to mid-sized jurisdictions with limited modelling capacity.