Multiscale heat and mass transfer modeling and multi-objective optimization of CO removal in enclosed refuge chambers

• A framework integrates multiscale modeling with NSGA-II for CO removal. • Greater air flow, heating, and smaller particles improve removal but raise costs. • Optimal setup (3 mm particles, GHSV 6327 h⁻¹, ACH 9.7 h⁻¹) met CO reduction standards. • A 33 °C limit was identified as the energy-efficiency threshold for Hopcalite in CO removal. Refuge chambers are critical life-saving installations designed to protect personnel who cannot evacuate during emergencies, where carbon monoxide (CO) can pose a lethal threat. While existing regulations mandate specific CO removal requirements, a theoretical framework for optimizing system design remains underdeveloped. This study addresses this gap by developing a framework that integrates a multiscale heat and mass transfer model spanning catalyst, reactor, and chamber scales, with a multi-objective optimization algorithm. The framework accounts for key parameters such as particle size, gas hourly space velocity (GHSV), air changes per hour (ACH), and heating temperature. Bench-scale and chamber-scale experiments validated the applicability of the multiscale model for real-world scenarios. Using this model, we applied genetic algorithms to optimize system and operational parameters, balancing CO removal capacity with energy consumption to yield Pareto-optimal solutions. For the regulatory requirement of reducing CO from 400 ppm to 24 ppm within 20 min, the optimal energy-efficient setup involved 3 mm catalyst particles, a GHSV of 6327 h⁻¹, and an ACH of 9.7 h⁻¹. In extreme conditions involving severe CO leakage or catalyst deactivation, heating is an effective supplementary measure. However, for the investigated Hopcalite catalysts, temperatures above 33 °C result in disproportionately high energy costs. These findings provide a comprehensive framework for designing efficient CO abatement systems in enclosed refuge chambers, ensuring compliance with safety standards while optimizing operational performance.