The numerical simulation model of the premix cooling unit was developed in the framework of this dissertation. The simulation model is intended for predicting changes in all important parameters of the cooling system with ice bank in transient operating conditions, which occur during water cooling and ice bank formation with the variable heat load of the evaporator. The model consists of two subsystems that work alternately and exchange data at each time step. The first subsystem monitors changes of the refrigerant state during the circulation within the cooling system. The second subsystem monitors the flow and heat exchange inside the water bath, where the water initially cools down and later freezes around the pipe coil of the evaporator of the cooling system. The method of distributed parameters was used to simulate the heat exchangers of the cooling system (evaporator and condenser). The simulation results represent a number of easily measurable parameters of the cooling system such as evaporation and condensation pressures, mass flows through the capillary and compressor, individual temperatures within the system, but also some more difficult to measure parameters such as specific enthalpy in particular control volumes of the heat exchanger, velocity or vapour content. The most important parameters whose results are predicted by the simulation are the energy consumption and the mass of the formed ice bank. The evaluation of the numerical model was made by comparing the results of the simulation with the results obtained experimentally for 4 different configurations of the cooling unit. The evaluation showed that the deviations of the values of all important parameters obtained by simulation, are within acceptable limits and that such a simulation model can be used as a reliable tool for optimizing energy consumption of premix cooling units.