Energy efficiency of freezing tunnels: towards an optimal operation of compressors and air fans

Fish is one of Norway’s main exports, and can be shipped fresh, frozen or dried. This thesis examines the freezing of fish in batch tunnels and ways to increase the energy efficiency of this process. A fish freezing plant on the west coast of Norway was used as a baseline case and measurements were made of the freezing system. Different aspects of this system were simulated, mainly using MATLAB. The focus was on the compressors and the freezing tunnels of an industrial refrigeration system. The compressors and the freezing tunnel fans are the largest consumers of electricity, but they are often not operated at the highest efficiency. An analysis of the compressor operation showed that it was far from optimal, with several compressors often operating at part-load simultaneously. These were screw compressors regulated by slide valves, which provide easy capacity control, but also have low energy efficiency. The refrigeration system had several different sized compressors, and the results showed that it was possible to run the system with only one compressor at part-load operation. The total coefficient of performance was improved by as much as 29% for a low production period. A further analysis showed that installing a variable speed drive on one compressor would also improve energy efficiency and make capacity regulation straightforward. The freezing system included five batch freezing tunnels, each of which had a freezing capacity of more than 100 tonnes of pelagic fish. A typical freezing period lasted typically 20 h and decreased the fish temperature to -18◦C or below. The main task was to develop a computer program that could simulate the freezing process and the refrigeration system and locate opportunities for improvement. The air velocities inside the freezing tunnel varied with location, which were pinpointed using the computational fluid dynamics software program Airpak. These velocities were used in freezing time calculations. It was shown that a guide blade installed in the air flow at a critical location improved the air velocity distribution compared with no guide blade. Without the guide blade, the freezing times of the products were between 16 h and 32 h, but with a guide blade they were between 17h and 21 h, a span of only 4 h. These freezing times were calculated with a modified Plank’s equation. A numerical model was programmed in MATLAB and it was used to simulate the temperatures of the products. The model was a two-dimensional finite difference approximation of the heat conduction equation. The simulation results were compared to measured temperatures, to validate the model. The measured temperatures were also used for validation of another simulation program, programmed in Modelica. The final stage of this research involved testing different alternatives for reduced fan operation. The program for the product model was extended with models for calculating the energy consumption of the air fans and the compressors and using Airpak-simulated velocities. The air fan speed was reduced to 83%, 67%, 50% and 33% of full air fan speed. This was tested at 5 different points during the freezing period, to see how the freezing times were affected. Full air speed during the freezing period resulted in a total freezing time of 20 h. A reduction in air fan speed to 33% after 8 h resulted in an increase in total freezing time of 10 hours (47% longer) but reduced energy consumption to 73.8% of the baseline case. An alternative with only 4 h longer freezing time resulted in an energy consumption of 80.5% of the baseline case. It was assumed that the fans had variable speed drives. The effect of reduced air inlet temperature was also tested and the results show that this can reduce freezing times. The effect on the total energy consumption was not large and also depends on the rest of the refrigeration system. Issues raised by this thesis are relevant for future research. It is suggested that the main simulation program is expanded by incorporating more detailed models of the refrigeration system. Dynamic operation of the air fans is also a possibility, for example to gradually reduce fan speed with decreasing product heat load