Natural ventilation in large spaces: CFD simplified model validated with full-scale experimental data of Roman Baths

This study addresses the escalating need for energy-efficient and well-ventilated buildings by examining natural ventilation in large spaces. Validation of a CFD model was pursued through in-site experiments at the Roman Baths Museum in Chaves, Portugal. A sensitivity analysis aimed to determine the optimal number of monitoring points for model validation, crucial for establishing procedures in large-volume settings. Findings emphasized the feasibility of using a minimal number of monitored points, notably with a 3 × 3 test point arrangement, showcasing consistent temperature variations with low relative errors (0.50 %–1.75 %). Furthermore, the validated model assessed ventilation performance under diverse operational conditions, revealing slight enhancements in experimental settings, including an increase in air change rate (2.4 vs. 2.2 ACH) and a decrease in buoyancy dominance (Richardson number 197.3 vs. 241.3) compared to design conditions. Quantitative analysis highlighted similar temperature and velocity trends, with greater stratification in experimental conditions (temperature ratios 0.12 to 0.36 vs. 0.10 to 0.32). Qualitative assessments align with the quantitative analysis and enable the identification of stagnation zones and airflow distribution patterns. These findings affirm the methods' reliability in analysing ventilation in large spaces naturally ventilated, validating the model across diverse contexts, despite fewer data points. • Simplified model validated using is-situ data of large space. • Sensitivity analysis: monitoring with 26, 11, and 6 points. • Low relative errors: 0.5%–0.8 % (11 points), 1%–1.75 % (6 points). • Ventilation performance: ACH, Richardson numbers, temperature/velocity profiles. • Velocity maps reveal airflow patterns and recirculation zones.