Perceived Barriers to Adoption of Building-Integrated Photovoltaics in the Sustainable Renovation of Buildings: An Exploratory Configurational Analysis of Stakeholders and LLM-Generated Personas in Northern Portugal
Building-integrated photovoltaics (BIPV) can combine the renovation of the building envelope with on-site electricity generation, contributing to the sustainable renovation of buildings (reduction in carbon emissions and increased energy efficiency). Yet their uptake in existing buildings remains limited. This exploratory study examines how 16 property owners, architects and engineers in Northern Portugal, recruited through a purposive convenience sample, perceive fourteen barriers to BIPV adoption, grouped into economic, knowledge, technological and supply-chain categories, and how those perceptions relate to two single-item outcomes: the view that the benefits of BIPV outweigh its additional costs (perception) and the view that BIPV should play an important role in the energy renovation of the existing stock (perspective). Linguistic Likert answers were represented as triangular fuzzy numbers (TFNs), converted into crisp scores with the CFCS technique, and analysed descriptively and with fuzzy-set qualitative comparative analysis (fsQCA), accompanied by sensitivity analyses of calibration and thresholds. Sixteen persona-conditioned synthetic respondents, built from the participants’ profiles and generated five times under two prompts, were processed in the same way. Barriers were perceived as at least moderately important, and the lack of information and technical guidance ranked first. Respondents were more favourable towards the future role of BIPV (CFCS = 0.719) than towards its current benefit–cost balance (0.544). In the fsQCA, technological barriers were present in all configurations, leading to the absence of a positive perspective, but detailed configurations were sensitive to calibration choices and are interpreted cautiously. The synthetic samples reproduced the average outcome scores but compressed variance, rated technological barriers lower, depended strongly on the context given in the prompt for incentive-related items and showed no individual-level correspondence with the humans on whose profiles they were based. LLM personas may therefore complement, but not replace, human respondents.
Authors
- Carla Malta Pires (ORCID: https://orcid.org/0000-0002-5503-1192)
- Noelia Romero Castro (ORCID: https://orcid.org/0000-0002-8956-0367)
- Juan Piñeiro-Chousa
Institutions
- Universidade de Santiago de Compostela (ES)
Publication Details
- Journal
- Sustainability
- Published
- 2026-11-16
- DOI
- https://doi.org/10.3390/su182211660
- Primary Topic
- Sustainable Building Design and Assessment
- Type
- article
- Field-Weighted Citation Impact
- 0.00