DELFT-UBEM
An automated database-driven pipeline for district-scale building energy modeling, demonstrated on Dutch district retrofit
Amin Jalilzadeh (TU Delft - Architecture and the Built Environment)
Azarakhsh Rafiee (TU Delft - Architecture and the Built Environment)
Thaleia Konstantinou (TU Delft - Architecture and the Built Environment)
Stefan de Graaf (Geodam, Amsterdam)
Henk Scholten (Vrije Universiteit Amsterdam)
Peter J.M. van Oosterom (TU Delft - Architecture and the Built Environment)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
District-scale energy transition planning requires building-level energy profiles, yet current Urban Building Energy Modeling (UBEM) workflows require substantial manual data preparation with no integrated path from building registries to scenario comparison. This paper presents DELFT-UBEM, an automated tool that generates dynamic simulation models from building registries and national performance standards through a lookup-table architecture that maps building age, function, and energy label to physical parameters, translating regulatory documents into simulation-ready parameters within a unified, database-driven pipeline. The pipeline shifts repeated project-level preparation into reusable database, lookup-table, and configuration components: once these components and a geographic boundary are available, data ingestion, geometric processing, archetype assignment, Input Data File (IDF) generation, parallel simulation, and post-processing execute as a connected automated workflow, producing analysis-ready results for downstream policy evaluation. Against national benchmarks, the model reproduces structural variation in Dutch residential heating demand across 35 archetypes without building-specific calibration. Comparison with measured postcode-level gas consumption across 63 postcodes (~3,000 buildings) corroborates this at aggregate scale, with residual errors systematically traceable to known data barriers and modeling limitations. To demonstrate downstream analytical capability, two contrasting Dutch districts were used to define custom retrofit and electrification applications, producing path-dependent marginal abatement cost curves with building-level sequencing across 3,404 upgrade steps. The paper discusses open challenges in UBEM validation standards and cross-tool comparability, highlighting how the pipeline’s modular architecture supports continuous refinement. The complete pipeline, archetype library, and parameter mappings are published as open-source resources, and DELFT-UBEM is accessible as a web-based digital twin platform.
help_outline