Evaporation in Flooded Corrugated Plate Heat Exchanger with NH3 and NH3/Miscible Oil

Abstract

Plate heat exchangers (PHEs) provide better heat transfer performance and operational flexibility and are being used in many process, chemical and refrigeration industries. However, their effective use in industries requires accurate heat transfer and pressure drop performance data. An extensive literature survey is presented to document existing literature available and highlight the importance of present study. This survey demonstrated that although considerable single phase work has been carried out on the performance of plate heat exchangers, however, existing single phase correlations have large disagreement and limited applicability. Limited two phase experimental results, using CFCs and HFCs as refrigerants are available on PHEs but very few data are reported for natural refrigerants in PHEs. The main objective of this project is to experimentally investigate thermo- hydraulic characteristics of a commercial chevron plate heat exchanger in single and two phase applications with and without lubricant effects and develop generalized correlations for Nusselt number and friction factor. The design and setup details of the experimental facility and data acquisition system are described. Steady state single phase (water to water) and two phase (ammonia evaporation) experiments were performed to investigate heat transfer and pressure drop characteristics of a commercial plate heat exchanger. Effects of plate geometry, heat flux and Reynolds number are considered in the single phase experiments. Experiments were performed for 30°/30° (soft), 60°/60° (hard) and 30°/60° (mixed) chevron plate configurations. For single phase experiments, Reynolds number (Re) is varied from 500 to 2,500 while the Prandtl number ranged between 3.5 and 6.0. Experimental data show strong influence of plate geometry and Reynolds number on the heat transfer and pressure drop in the plate heat exchanger. Considering the possible measurement errors, the experimental Nusselt number and friction factor data have an uncertainty of ±6.5% and 2.4% respectively. Based on experimental data generalized correlation to estimate single phase Nusselt number (Nusp) incorporating effects of plate geometry, Re and varying fluid properties has been proposed. The Nusp correlation represents experimental data within a ±2% error band for symmetric plate configurations with maximum deviation of ±4% for mixed plate configuration. The plate specific fsp correlations represent experimental data within a ±5% error band. The two phase steady state experiments, using ammonia as refrigerant, have also been performed on symmetric plate configurations, 60°/60°, 30°/30° and a mixed 30°/60° plate configuration. Experiments are conducted for -2°C, - 9°C, -14.5°C, -19.5°C and -25°C saturation temperatures. Majority of the experiments were conducted for 0.5 < exit vapor quality < 0.9 which is the range of interest. The range of other important parameters are; 1,225 < equivalent Reynolds number < 3,000 and 21 kW/m2 < heat flux < 44 kW/m2. For all three geometric configurations, the effects of mass flux, heat flux, saturation temperature, exit vapor quality (x) and equivalent Reynolds number (Reeq) are considered on the heat transfer and pressure drop characteristics of the plate heat exchanger. The two phase heat transfer coefficient (htp) is found to increase with an increase in the chevron angle for the entire range of heat flux and exit vapor quality. It also increased with an increase in saturation temperature. The uncertainty in experimental Nusselt number data is found to be maximum for low saturation temperature (-25°C) and is less than ±10%. Nucleate flow boiling dominated the low vapor quality regime (0.1 < x < 0.4) while convective boiling appears to play major role in the high vapor quality regime (0.5 < x < 0.8). Dry- out phenomenon is also observed for exit vapor quality beyond 0.8. Based on the experimental data a generalized correlation is proposed to estimate the two phase Nusselt number. The two phase Fanning friction factor decreased with an increase in Reeq. However, it is found to increase with an increase in saturation temperature. Considering all data measurement errors, an uncertainty of ±2.6% is found in the experimental friction factor data. Correlations to estimate two phase friction factor are also developed. Effect of miscible oil on the thermal-hydraulic performance of the plate heat exchanger has also been investigated. Experiments were conducted on the mixed plate configuration for five temperatures ranging between -2°C and -25°C with miscible oil concentrations of 3%, 6% and 9% by volume in liquid ammonia. The experimental results of miscible oil/ammonia mixture have been compared with pure ammonia data. The heat transfer coefficient is found to increase with an increase in oil concentration up to 3% but reduced with further increase in oil concentration. However, the effect of oil concentration is not significant on pressure drop, especially, at high saturation temperatures. At low saturation temperature and high oil concentration values, the pressure drop increased with a decrease in the saturation temperature. An uncertainty of about ± 10% is found in experimental two phase heat transfer coefficient data, while an uncertainty of less than ±3% is found in the experimental friction factor data. Generalized Nusselt number and friction factor correlations incorporating effects of oil concentration, Reeq, equivalent Boiling number, plate geometry and reduced pressure have been proposed. These correlations represent experimental data quite well. All experimental data are compared with previously published works.