Apportioning radon contamination sources in underground spaces using a grey-box model.

The management of radon in indoor environments is essential owing to its carcinogenic risk, with the main sources being gaseous radon from advection and aqueous radon from degassing. However, methods for apportioning indoor radon sources are limited, due to gaps between preliminary physical information and site-specific measurement. This study proposes a method for apportioning sources utilizing a grey-box model that integrates physical theories with data information. Monitoring of indoor air conditioning was conducted and a grey-box model for indoor radon contamination was established using a signal processing system and a physical-based model. Based on the optimized grey-box model, sources of indoor radon were apportioned. The average radon concentration during the monitoring period was 1104 Bq/m³, which was 7.5 times higher than 148 Bq/m³ , the action level recommended by EPA. The grey-box model showed good performance for predicting indoor radon concentrations. Degassing of aqueous radon was identified as the main source of indoor radon. The estimations provided by the grey-box model showed its potential to assess the contribution of each transport mechanism. Apportioning sources through grey-box modeling is a promising method for developing optimized mitigation strategies and bridging data gaps between the preliminary conceptual site model and site investigation.