A Data-Driven Decision Support Tool for a priori Multi-Objective Optimization of Building Portfolio Assets
Preference-Based Decision Support for Sustainable Rooftop Strategies
F.D.J. Pallandt (TU Delft - Civil Engineering & Geosciences)
Ruud Binnekamp – Mentor (TU Delft - Real Estate Management)
GA van Nederveen – Graduation committee member (TU Delft - Integral Design & Management)
T. Chatzivasileiadis – Graduation committee member (TU Delft - Policy Analysis)
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
Cities increasingly face challenges regarding climate change, urban densification, biodiversity
loss and community needs. These challenges demand for adaptation strategies. Each
neighborhood faces unique challenges, requiring locally adapted strategies. Public buildings,
often owned by municipalities within a portfolio, can play a key role in these strategies to
improve resilience and livability. Yet, due to the diversity and number of assets, deciding where
and how to intervene is highly complex. Data and decision support tools such as Multi-
Objective Optimization (MOO) and Multi-Criteria Decision Analysis (MCDA) can reduce this
complexity, but traditional approaches often exclude stakeholders early, restrict evaluations to
predefined solutions, commit aggregation errors, and typically end with Pareto fronts that lack
a clear final design.
This thesis develops and tests a MOO decision support tool for the management of building
portfolio assets. Based on the Preferendus framework, it integrates stakeholders a priori,
avoids fixed solution spaces, applies mathematically sound preference modeling, and
converges on a single optimal configuration. A demonstrator case with 25 buildings with data
derived from The Hague was used to optimize rooftop interventions (sedum, biodiversity,
solar, water retention, and social-commercial roofs), balancing conflicting objectives such as
neighborhood needs, CO₂ impact, investment costs, and financial returns.
The tool was validated in a live workshop at Bress, where objectives, weights and preference
curves were adjusted in real time to search for a final optimal design configuration. Results
showed its ability to support portfolio owners in exploring trade-offs and aligning interventions
with objectives. A review by Arcadis confirmed its potential, while also pointing to
improvements for further development such as spatial impact modeling, data quality, a more
user-friendly interface, and prioritization of assets over time.
In conclusion, this preference-based MOO tool demonstrates that rooftop allocation is an
effective demonstrator case for sustainable portfolio asset management. Preferendus proves
suitable for structuring complex decision-making and translating stakeholder preferences into
optimized, data-driven design strategies, while overcoming key shortcomings in similar studies.