Umeå University's logo

umu.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Improved energy retrofit decision making through enhanced bottom-up building stock modelling
Umeå University, Faculty of Science and Technology, Department of Applied Physics and Electronics.ORCID iD: 0000-0002-7790-4855
Umeå University, Faculty of Science and Technology, Department of Applied Physics and Electronics.ORCID iD: 0000-0002-9310-9093
Umeå University, Faculty of Science and Technology, Department of Applied Physics and Electronics.ORCID iD: 0000-0002-8704-8538
Umeå University, Faculty of Science and Technology, Department of Applied Physics and Electronics.ORCID iD: 0000-0001-5776-5221
Show others and affiliations
2024 (English)In: Energy and Buildings, ISSN 0378-7788, E-ISSN 1872-6178, Vol. 318, article id 114492Article in journal (Refereed) Published
Abstract [en]

Modelling the performance of building stocks is crucial in facilitating the renovation at the building stock level. Bottom-up building stock modelling begins by detailing individual buildings and then aggregates them into stock level. Its primary advantage lies in capturing the inherent heterogeneity among distinct buildings, which enables tailored retrofitting. Naturally, this approach requires a comprehensive dataset with detailed building information such as geometry and envelope thermal properties. However, a common challenge is the incompleteness of available data in individual datasets. To address this, previous bottom-up studies have filled the missing data with representative or statistical data. Such practice could lead to homogeneous modelling of distinct buildings within the same statistical group. This limits the utilization of key ability of bottom-up building stock modelling in capturing heterogeneity, such as tailored retrofitting to explore potential retrofitting areas and strategies. To address this challenge of homogeneous modelling, we utilize data fusion framework for bottom-up building stock modelling, employing probabilistic record linkage and inverse modelling techniques to integrate multiple incomplete building performance datasets. This framework fills the missing data in one dataset with information from another, thus capturing inherent heterogeneity in the building stock. An empirical study was conducted in Umeå, Sweden, to investigate the framework's effectiveness by modelling building stock with various retrofitting strategies. This study contribution lies in enhancing bottom-up building stock modelling by capturing inherent heterogeneity, to provide tailored retrofitting solutions.

Place, publisher, year, edition, pages
Elsevier, 2024. Vol. 318, article id 114492
Keywords [en]
Bottom-up, Building stock modelling, Data fusion, Energy efficiency, Heterogeneity, Incomplete data, Inverse modelling, Record linkage
National Category
Other Civil Engineering
Identifiers
URN: urn:nbn:se:umu:diva-227827DOI: 10.1016/j.enbuild.2024.114492ISI: 001262604500001Scopus ID: 2-s2.0-85197361082OAI: oai:DiVA.org:umu-227827DiVA, id: diva2:1883950
Projects
Intelligent Human-Buildings Interactions Lab: Identify, Quantify and Guide Energy-saving Behavior at University CampusRESILIENTa Energisystem Kompetenscentrum
Funder
Swedish Research Council Formas, 2022-01475Swedish Energy Agency, P2022-00141Swedish Energy Agency, 52686-1Swedish Research Council Formas, 2020-02085Available from: 2024-07-12 Created: 2024-07-12 Last updated: 2026-03-18Bibliographically approved
In thesis
1. Heterogeneity-aware building stock modelling for urban energy transitions
Open this publication in new window or tab >>Heterogeneity-aware building stock modelling for urban energy transitions
2026 (English)Doctoral thesis, comprehensive summary (Other academic)
Alternative title[sv]
Heterogenitetsmedveten modellering av byggnadsbestånd för urbana energiomställningar
Abstract [en]

Bottom-up building stock modelling (BBSM) approach is widely used to assess the energy performance of urban building stocks by modelling individual buildings in detail and aggregating them to larger spatial scales. It plays an important role in supporting urban energy transition planning and policymaking. However, existing BBSM studies are constrained by several limitations, including incomplete building performance datasets, reliance on archetype-based averages in retrofit prioritization, simplified representations of occupant behaviour and building properties, and limited integration of modelling into public engagement tools. These limitations obscure inherent heterogeneity that determines which buildings benefit most from specific measures, leading to one-size-fits-all retrofit strategies and biased estimation of energy-saving potentials and policy effectiveness of energy transition initiatives.

This thesis advances heterogeneity-aware BBSM through an integrated and cumulative methodological pipeline that fuses incomplete building-performance datasets, enables localized retrofit prioritisation, and supports evaluation and communication of demand-side behavioural impacts. First, a data-fusion framework combines multiple incomplete datasets using probabilistic record linkage and inverse modelling, filling data gaps by transferring information across sources rather than relying on archetype-level averages. This improves stock representation by capturing building-to-building variation within the same urban context. Second, the thesis integrates data fusion with ensemble machine learning and explainable AI (SHAP) to identify impactful envelope retrofit measures in a local, data-driven manner. Across 81 building-stock clusters in three Swedish municipalities, the results demonstrate substantial variation in the most influential thermal components across municipalities and climate zones, underscoring the need for local-specific retrofit prioritisation. Third, the thesis incorporates occupant-behaviour diversity alongside building heterogeneity via an enhanced DOB-HUBS framework based on representative clustering, automated physics-based simulations, and surrogate machine learning. Empirical results of the Umeå building stock demonstrate that oversimplified behavioural and homogeneous assumptions can bias energy outcomes by up to 15% (standard deviation 3.06%). The framework is further used to assess Sweden’s forthcoming 2027 capacity-based electricity tariff, indicating that behavioural adaptations could reduce peak electricity demand by 6–17%, with heterogeneous impacts across clusters.

Building on these modelling advances, the thesis extends toward public engagement by developing data-driven benchmarking and interactive visual analytics platform that translate bottom-up modelling outputs into user-facing insights, including peer comparison and ‘what-if’ exploration of retrofit and behavioural scenarios. Collectively, the thesis contributes methods and empirical evidence for more credible, locally tailored, and publicly actionable building-stock analytics. This supports in designing targeted retrofit strategies, effective behavioural measures, and informed public participation in urban energy transition planning.

Abstract [sv]

Bottom-up-baserad modellering av byggnadsbestånd (BBSM) används i stor utsträckning för att bedöma energiprestandan hos urbana byggnadsbestånd genom att modellera enskilda byggnader i detalj och därefter aggregera resultaten till större spatial skala. Metoden spelar en viktig roll för planering och upprättandet av policys för urbana energiomställningar. Befintliga BBSM-studier begränsas av flera faktorer, däribland ofullständiga dataset gällande byggnaders energiprestanda, beroende av arketypbaserade genomsnitt vid prioritering av energieffektiviserande renoveringar, förenklade representationer av boendebeteende och byggnadsegenskaper samt begränsad integrering av modellering i verktyg avsedd för offentligt nyttjande. Dessa begränsningar döljer den inneboende heterogenitet som avgör vilka byggnader som har störst nytta av specifika åtgärder, vilket leder till generella renoveringsstrategier och snedvridna uppskattningar av potentialen för energibesparingars och policyers effektivitet i energiomställningsinitiativ.

Denna avhandling utvecklar heterogenitetsmedveten BBSM genom ett integrerat och kumulativt metodologiskt flöde som kombinerar ofullständiga dataset över byggnaders energiprestanda, möjliggör lokalprioritering av renoveringsåtgärder och stödjer utvärdering samt kommunikation av beteendebaserade effekter på efterfrågesidan. För det första kombinerar ett datafusionsramverk flera ofullständiga dataset genom probabilistisk postkoppling (probabilistic record linkage) och inverterad modellering, vilket fyller dataluckor genom att överföra information mellan datakällor i stället för att förlita sig på arketypbaserade genomsnitt. Detta förbättrar representationen av byggnadsbeståndet genom att fånga variation mellan enskilda byggnader inom samma urbana kontext. För det andra integrerar avhandlingen datafusion med ensemblebaserad maskininlärning och förklarbar AI (SHAP) för att identifiera effektiva klimatskalåtgärder på ett lokalt och datadrivet sätt. I analyser av 81 kluster av byggnadsbestånd i tre svenska kommuner visar resultaten betydande variation i vilka termiska komponenter som är mest inflytelserika mellan kommuner och klimatzoner, vilket understryker behovet av lokalt anpassad prioritering av renoveringsåtgärder. För det tredje integrerar avhandlingen variation i boendebeteende tillsammans med byggnadsheterogenitet genom ett utökat DOB-HUBS-ramverk baserat på representativ klustring, automatiserade fysikbaserade simuleringar och surrogate-maskininlärning. Empiriska resultat från byggnadsbeståndet i Umeå visar att förenklade beteendeantaganden och homogena antaganden kan snedvrida energiberäkningar med upp till 15 % (standardavvikelse 3,06 %). Ramverket används vidare för att analysera Sveriges kommande kapacitetsbaserade eltariff från 2027, där resultaten indikerar att beteendeanpassningar kan minska toppefterfrågan på el med 6–17 %, med heterogena effekter mellan olika kluster.

Med utgångspunkt i dessa modelleringsframsteg utvidgar avhandlingen även perspektivet mot offentligt nyttjande genom utvecklingen av en datadriven, interaktiv benchmarking- och visualiseringsplattform som översätter resultat från bottom-up-modellering till användarorienterad information, inklusive jämförelser med liknande byggnader och ”what-if”-utforskning av renoverings- och beteendescenarier. Sammantaget bidrar avhandlingen med metoder och empiriska resultat som möjliggör mer tillförlitlig, lokalt anpassad och praktiskt användbar analys av byggnadsbestånd. Detta stödjer utformningen av riktade renoveringsstrategier, effektiva beteendebaserade åtgärder och välgrundad offentlig medverkan i planeringen av urbana energiomställningar.

Place, publisher, year, edition, pages
Umeå: Umeå University, 2026. p. 69
Keywords
Urban energy transition, building stock, bottom-up, heterogeneity, tailored retrofitting, public engagement, machine learning
National Category
Energy Systems
Research subject
architecture, urban planning; climate change; sustainability; data science; sustainable development
Identifiers
urn:nbn:se:umu:diva-251241 (URN)978-91-8070-961-3 (ISBN)978-91-8070-962-0 (ISBN)
Public defence
2026-04-29, NAT.D.300 (Hörsal), Floor 3, Natural Sciences Building, Universitetsvägen, 901 87, Umeå, 09:00 (English)
Opponent
Supervisors
Funder
EU, Horizon 2020, 101016854Swedish Research Council Formas, 2020-02085Swedish Research Council Formas, 2022-01475Swedish Energy Agency, P2022-00141Swedish Energy Agency, 52686-1
Available from: 2026-04-08 Created: 2026-03-18 Last updated: 2026-03-18Bibliographically approved

Open Access in DiVA

fulltext(7040 kB)315 downloads
File information
File name FULLTEXT01.pdfFile size 7040 kBChecksum SHA-512
2f3c3df0fc6774b38c90ba11099dc2ac3299ee63b0b61fac8cbc0191bfc7f480864b65d3c29bb55e73d7f35478946d3db4858731b0c6500095a341a110ce6bf0
Type fulltextMimetype application/pdf

Other links

Publisher's full textScopus

Authority records

Penaka, Santhan ReddyFeng, KailunOlofsson, ThomasRebbling, AndersLu, Weizhuo

Search in DiVA

By author/editor
Penaka, Santhan ReddyFeng, KailunOlofsson, ThomasRebbling, AndersLu, Weizhuo
By organisation
Department of Applied Physics and Electronics
In the same journal
Energy and Buildings
Other Civil Engineering

Search outside of DiVA

GoogleGoogle Scholar
Total: 317 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 643 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf