2026-10-13T18:17:37Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34004586 2026-09-30T00:35:54Z category_26365 portal_12 item_type_5 month_year_09_2026

Ghanashyam Danam: Can we predict whether an infrastructure project quantifies its energy sustainability benefits? Ghanashyam Danam (22002926) Infrastructure engineering and asset management infrastructure sustainability benefit quantification internal project characteristics machine learning random forests energy early-stage decision making <p dir="ltr">Infrastructure projects promise benefits that their early-stage decision documents describe but never value. A benefit that is not valued in the same terms as cost cannot be weighed against it, so it drops out of the decision. This study asks whether a project's own characteristics determine which of its benefits get quantified. <br><br>One hundred public private partnership feasibility study reports, spanning nine asset types between 2008 and 2021, were coded for 83 energy related design, construction and operational features, grouped into 16 themes. Each report was separately recorded as having quantified an energy benefit or not: 48 had, 52 had not. A random forest was trained to predict the second from the first. <br><br>The model tells the two apart 84 times out of 100, well above chance under a permutation test. Which sustainability benefits get counted is therefore patterned, and predictable from the project itself.</p> 2026-09-30T00:35:54Z Image Poster 10.17608/k6.auckland.34004586.v2 https://figshare.com/articles/poster/Can_we_predict_whether_an_infrastructure_project_quantifies_its_energy_sustainability_benefits_/34004586 CC BY 4.0