An investigation of a jet-pump thermal (ice) storage system powered by low-grade heat
This thesis investigates a novel combination of a jet-pump refrigeration cycle and a thermal (ice) storage (TIS) system that could substantially reduce the electrical energy requirements attributable to comfort cooling.Two methods of TIS were identified; spray ice TIS would use evaporative freezing to store ice on a vertical surface,and encapsulated ice TIS would freeze a bed of encapsulated elements by sublimation freezing.Thestudy also investigates jet-pump refrigeration at partload and a convergent-divergent design manufactured from a thermoset plastic to make recommendations for performance enhancement for a system that has a low COP. An experimental rig was built to investigate the novel concepts in the laboratory. Encapsulated ice TIS was superior to spray ice TIS because, for the same nominal secondary flow, sublimation freezing causes an increase in coolth storage rate of about 10 % compared to evaporative freezing. Encapsulated ice stores experience difficulties in fully discharging their coolth (approximately 6% in this case), but spray ice TIS can be used to produce an ice/brine slurry enabling all of the ice to be used, and so may be more suitable if the unmelted ice represents a large proportion of the cooling capacity. Approximately 85 % to 90 % of the ice formed on the vertical surface during spray ice TIS testing was formed by evaporative freezing from a falling film. At high saturation conditions, heat is transferred mainly by conduction across the falling film. Both the growth of an ice layer on a vertical surface and freezing of encapsulated elements were found to be successful, but a large data spread was observed during spray ice TIS testing. It was thought that a variation in the steady-state saturation conditions in the evaporator/ice store was caused by variability of droplet size distribution from the spray nozzle flow, which may make a full-scale system unreliable. The COP of the spray ice TIS system was approximately 0.15 compared to a COP of approximately 0.25 found during encapsulated ice TIS testing. The difference was because of the use of an over-expanded primary nozzle, which restricted secondary flow and increased momentum losses. A primary nozzle that expands close to the design evaporator saturation conditions should be used to maximise entrainment ratio. The COP of a jet-pump TIS is low, but a system designed to operate at off-peak periods could increase the COP to about 0.8 by taking advantage of the lower ambient conditions. The measurement of entrainment ratio was used successfully to determine ice storage rate and COP. This was valid because of the assumption that the saturation conditions in the evaporator/ice store approached steady-state. However, over longer periods that would be found in large-scale systems, the ice storage rate and entrainment ratio may fall substantially. The steady-state assumption could still be used to observe the change in evaporator conditions by sampling over short time intervals (30 minutes). At part-load, increases in evaporator saturation temperature could increase entrainment ratio substantially (50 % increase) for only a small reduction in critical pressure lift ratio Ns *(15 % reduction). A variation in chilled water temperature could be used to boost entrainment ratio at the peak demand. The variation in Ns* is too small to use this strategy to control the jet-pump with respect to condenser operating conditions. The entrainment ratio is approximately proportional to the diff-user to primary nozzle area ratio. A doubling of entrainment ratio was attained for only a 15% reduction in Ns*. The change in geometry from a constant area throat to a convergent-divergent design caused the flow through the jet-pump to vary with outlet conditions indicating that secondary flow was not choked. Higher entrainment ratios and pressure lift ratios were observed, but the entrainment ratio varied with outlet conditions in the form of peaks and troughs, making its operation unpredictable. This was thought to be caused by the restriction in secondary flow area due to the interaction of the primary jet and the curved wall. The convergent-divergent design manufactured from a thermoset plastic was successfully tested, showing that a plastic material can be used as a material of construction. In principle, a large number of jet-pump units could be manufactured from a single mould, reducing the first cost. The investigation proved the concept of jet-pump TIS. Waste-heat could be utilised over 24 hours and year round, increasing the efficiency of the process. The use of a convergent-divergent throat design, multiple geometry jet-pumps and operation at off-peak periods can maximise the performance over a cooling season, and be competitive with other TIS and chiller systems. The mass production of jet-pumps using injection moulding techniques could reduce substantially the capital cost of a system. All of these factors should encourage the development of such systems, so that the harmful emissions caused by the use of air conditioning systems can be minimised.