BUILDING SIMULATION VS. PERFORMANCE EVALUATION: A CRITICAL REVIEW OF CONVERGING AND DIVERGING OUTCOMES IN HOUSING STUDIES

Abstract

This article presents a critical review of fifteen peer-reviewed studies published between 2021 and 2025, focusing on the convergence and divergence between building performance simulation (BPS) outputs and post-occupancy evaluation (POE) data in housing studies. Using an integrative, semi-systematic approach, the review categorises the literature under five thematic areas: occupant-centric simulation models, surrogate modelling and machine learning, building envelope performance, urban building energy modelling (UBEM), and multi-objective optimisation. Findings reveal that convergence is strongest in studies that incorporate empirical calibration and behavioural variability, while divergence often arises from oversimplified input assumptions, lack of real-time feedback, and limited field validation. The review contributes to knowledge by identifying methodological patterns that enhance predictive reliability and by proposing recommendations for improving simulation accuracy in housing design. It concludes that greater integration of POE data, adaptive modelling strategies, and context-aware simulation tools is necessary to close the performance gap between predictive models and real-world outcomes.