Physics-based digital twin for real-time fault detection and diagnosis of a desiccant dehumidification system using metal–organic frameworks

초록

Metal–organic framework (MOF) desiccant wheels provide high adsorption efficiency but require careful monitoring because excessive regeneration temperatures can reduce adsorption performance and increase substrate deformation risk. This study addresses this challenge by developing a physics-based digital twin model for the real-time fault detection and diagnosis (FDD) of a solid desiccant dehumidification system utilizing a MOF-coated polystyrene desiccant wheel. By analyzing the transient thermodynamic behaviors under five representative component-level fault scenarios, distinct temperature-deviation patterns are identified as the primary indicators for fault classification. The developed digital twin model provides a validated baseline for FDD, and model validation and threshold-sensitivity analysis support the 10.0% temperature-deviation criterion. Under overload conditions, it successfully detects heater overheating at the 20 s sampling point and initiates heater-off corrective control. Accordingly, the proposed digital-twin-based FDD approach stabilizes operation below the upper regeneration-temperature limit of 70.0 °C, thereby preventing thermally critical operation. These findings provide a basis for real-time FDD and corrective control of thermally sensitive MOF-coated desiccant wheel systems in HVAC applications. © 2026 Elsevier Ltd.

키워드