Experimental analysis of waste heat recovery performance of plate exchangers manufactured from different materials

After completing their process, industrial liquids and gases comes idle state although that contain high amount of heat energy. Waste heat recovery can be carried out by using of equipments such as heat pipes, heat exchangers and waste heat boilers. In this way, much more efficiency will be obtained using with the same amount of energy input and also environmental pollution will be reduced and ecological balance will be maintained. In this study; a heat recovery device was designed and manufactured to provide air to air heat transfer using cross-flow plate heat exchanger. The low temperature clean and fresh air was taken from the external environment and the high temperature exhaust air was obtained from the internal environment. The aim is; heating the fresh air when these air passing through the plates. Therefore, heat exchangers made from three different materials as aluminum, polymer and cellulosic are supplied. Air speed, fresh air and exhaust air inlet temperatures were defined as operating parameters. As a result of these experiments, fresh air and exhaust air outlet temperatures were obtained and also thermal calculations were conducted with using these values. Parameters having an effect on thermal efficiency, the degrees of influence of these parameters and their interactions were examined with using Taguchi experimental design method. For these analysis Minitab (Release 16) statistical software was used. Thereby the interpretation of results was supported by statistical analysis. Analysis results have shown that; the most effective factor on the efficiency value is the air speed and there is no specific interaction between factors. The best thermal performances were achieved for all heat exchangers when fresh air inlet temperature 0 °C, air velocity 1,2 m/s and exhaust air inlet temperature 40 °C was selected. It was noticed that cellulosic heat exchanger compared to the other heat exchangers give much better results for all parameter values.