Development of a digital twin for a liquid desiccant system with heat pump integration
Efficient humidity management has become a critical demand in modern HVAC systems due to rising outdoor humidity associated with global warming and the rapid growth of moisture-sensitive industries such as semiconductor and battery manufacturing. Liquid desiccant (LD) air-conditioning systems offer a promising solution by directly absorbing water vapor from ventilation air using a hygroscopic solution, thereby handling latent loads without the energy-intensive over-cooling and reheating processes required in conventional systems. As moisture absorption dilutes the solution, stable operation requires simultaneous cooling of the absorber and heating of the regenerator, making heat-pump integration highly effective. However, optimizing LD–heat-pump assemblies at the system level remains challenging because most prior studies address isolated components rather than the coupled thermal interactions that govern real performance. Digital twin technology provides a pathway to overcome this gap by enabling virtual replication of the physical system for supervisory control design, fault detection, and long-term performance evaluation. In this work, we develop a detailed digital twin of an LD air-handling unit equipped with two heat pumps and a cross-flow liquid desiccant contactor. The model is constructed using the Modelica language. Experimental data are obtained from an installed classroom ventilation unit featuring a 7.1 kW chilled-water heat pump and a 4.5 kW hot-water heat pump. The resulting digital twin platform enables fast scenario testing, energy-management optimization, and exploration of advanced control strategies that can further reduce carbon dioxide emissions in buildings and industrial facilities.