The energy consumption in the industry field is considerable. Saving energy has great potential in many industrial processes. For example, heat recovery from the spent coolingwater can be achieved by applying heat pump technology, and the recovered heat can be used for district heating. Previous study indicates that the compression-resorption cycle has higher efficiency than traditional vapor compression heat pumps due to the fact that it can take advantage of the temperature glide of the multi-component working fluid. The absorption process within the resorber and its performance should be further investigated. Ammonia–water mixture is commonly used as the working fluid in the compression-resorption heat pump. Recently the CO2 –NH3 –H2O mixture has been identified as a promising working fluid theoretically, relevant experiment should be conducted to verify it.

For the theoretical background, previous studies about kinetics of CO2 absorption in ammonia–water have been reviewed. Currently most of the theoretical and experimental studies are performed for post combustion capture usage. The working conditions deviate much from that for the compression-resorption heat pump cycle; therefore existent kinetic models are not suitable for predicting the absorption rate of the CO2 in ammonia–water within a compression-resorption heat pump cycle. Previous studies from TU Delft indicate that a more accurate model should be developed to predict the absorption process of the ammonia–water mixture in a multi-tube mini-channel heat exchanger.

In this study a steady-state theoretical model has been developed to predict the absorption process of ammonia–water mixture in a multi-tube mini-channel heat exchanger by assuming Nusselt falling film theory and annular flow pattern. The model is extended to cover all three different phases (superheated vapor, 2-phase, and sub-cooled liquid) of the ammonia–watermixture. Experiments with two different working fluids (ammonia–water mixture and CO2 –NH3 –H2 O mixture) on the tube side and cooling water on the shell side have been performed respectively. Results indicate that adding a small amount of CO2 (2.1 wt%) directly will lead to slightly better heat exchange on the water side when operated in optimum condition. But the operating status becomes much less stable than the experiment with ammonia–water as a working fluid. The pump becomes difficult to operate with constant mass flow. While when the flow directions on both sides are changed, more stable operating conditions can be achieved. This indicates the configuration of the heat exchanger and the flow direction influence the operating stability. Also it is possible to have CO2 desorbed at the pump even in low temperature. Therefore to put into practical use, more tests should be done to comprehensively investigate the feasibility of applying CO2–NH3–H2O mixture as working fluid in a compression-resorption heat pump.

The theoretical model has been validated by the experimental data. For the 2-phase flow condition on the tube side, the simulation results indicate that the assumption of annular flow pattern is reasonable for most of the heat exchange area. But when the vapor quality is low, the film thickness prediction based on the Nusselt falling film theory is not reasonable any longer, this indicates a transition of a flow pattern. A new flow pattern occurs in order to enhance the transport phenomena. The validation results show that the extended model can accurately predict the heat load and the temperature profile along the absorption process. However the pressure drop cannot be reasonably predicted. This can be caused by inaccurate friction factor estimation or ignorance of other effects which can cause extra pressure drop. The heat transfer performance at the superheated region is studied in detail. Results show an obscure relation between the weakened heat transfer phenomenon and the hydrodynamic instabilities on the tube side. Experimental data are also applied to validate the thermodynamic equilibrium models for the CO2 –NH3 –H2 O mixture. The model developed in this project can be used to predict the heat transfer performance for certain type of heat exchangers conducting ammonia–water absorption process, or be modified to apply for  other conditions. The experimental data are useful for ammonia–water and CO2 –NH3 –H2O mixture absorption related studies. As a follow-up for this research it is of interest to further investigate and identify the flow pattern transition when the vapor quality is low, and develop a more comprehensive model to predict the absorption process. Visual observations will help to understand the flow pattern transitions. A complete compression-resorption heat pump setup will contribute to more useful experimental result to estimate the performance using CO2 –NH3 –H2 O mixture as a working fluid.