Localising global construction delay risk factors: an optimisation-enhanced expert consensus framework for the Australian construction industry

Abstract

This study aims to investigate how globally identified construction delay factors can be translated into a framework that accurately reflects Australian project conditions. It aims to develop a validated set of delay risks by integrating systematic evidence synthesis with expert elicitation and optimisation-based consensus analysis. A systematic literature review consolidated 103 global delay factors, refined to 55 using a relevance-based filtering approach. Semi-structured interviews with senior Australian practitioners were conducted to assess contextual severity. Inter-rater agreement was measured using Kendall's Coefficient of Concordance (W), and a Genetic Algorithm (GA) was applied to optimise expert weighting and maximise consensus reliability, producing a statistically robust prioritisation. The optimisation process achieved strong expert agreement and generated a refined set of delay risks tailored to Australian project conditions. The most influential delays stem from governance and design-related issues, including late owner decisions, documentation delays, unclear drawings, and coordination challenges, alongside labour constraints and subcontractor performance. Cross-validation and triangulation with recent Australian studies confirmed the stability and contextual relevance of the final framework. The 22-factor framework provides practitioners with a clear basis for identifying and prioritising delay risks within Australian projects. By guiding planning and design decisions, it helps reduce schedule uncertainty and improve the predictability of project delivery. This research advances construction risk analysis by introducing an optimisation-enhanced localisation method that integrates global evidence, expert judgement, and evolutionary algorithms to produce a validated delay risk framework for Australia. The combined use of GA-based expert weighting and Kendall's W strengthens consensus reliability and offers a transparent methodological innovation to the field.