Experimental study of ammonia flow boiling in a vertical tube bundle: Part 1 – Enhanced dimple tubeÉtude expérimentale de l’ébullition en écoulement de l’ammoniac dans un faisceau de tubes verticaux :Partie 1 : Tube nervuré amélioré

https://doi.org/10.1016/j.ijrefrig.2021.07.012Get rights and content

Highlights

  • Boiling of ammonia in a vertical shell and tube evaporator with enhanced dimpled tube bundle.

  • Study of wide range of experimental parameters.

  • Enhanced heat transfer coefficient compared to previous studies.

Abstract

This paper presents the experimental results of saturated two-phase flow ammonia in a vertical shell and tube heat exchanger comprising of seven dimpled enhanced tubes. The tests were conducted for a wide range of parameters. The heat load was provided by hot ethylene glycol/water solution flowing on the shell side. Saturation temperature of ammonia was -20°C ≤ Tsat ≤ 2°C, heat flux range 5 kW m−2q˙ ≤ 45 kW m−2, exit vapor quality range 0.1 ≤ x ≤ 0.9 and a mass flux range of 34.99 kg m−2 s−1 ≤ G ≤ 66.32 kg m−2 s−1. For each saturation temperature, the effects of heat flux and mass flux were studied for the saturated temperature range while the exit vapor quality at each mass flux was varied between 0.2 and 0.9. Two phase heat transfer coefficient increased with saturation temperature and heat flux. For all experimental conditions, performance showed initial increase and then a classic drop at the onset of dry out. The dimple tube showed better heat transfer coefficient compared to the previous studies performed on plain tubes with ammonia.

Introduction

Ammonia, Methyl Chloride, and Sulfur dioxide were the refrigerants utilized at the start of mechanical refrigeration era. The toxicity and flammability were the major issues with these refrigerants which prompted to introduce synthetic refrigerants such as chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs) and hydroflourocarbons (HFCs). Later on it was discovered that these refrigerants have a negative impact on the environment. In 2016 at Kigali, an amendment to Montreal protocol was signed by 195 countries to phase down the use of HFCs to 85% of 2013 baseline by 2045.

Natural refrigerants are attracting attention as a suitable replacement for synthetic refrigerants. Ammonia, among them is receiving prominent interest because its environment friendly, costs low and has excellent thermodynamic and transport properties. It has proven itself to be an excellent refrigerant for decades. Currently, it is the only refrigerant with zero Global Warming Potential (GWP) and Ozone Depletion Potential (ODP). The only major problem with ammonia is its toxicity and mild flammability. However, these concerns can be duly addressed with proper system design. In order to deal with the problems related to toxicity of ammonia, researchers are focusing on systems with low refrigerant charge and better heat transfer performance.

In order to augment heat transfer rate in heat exchangers Bergles et al. (1983, 1991, 1995) identified different enhancement techniques and classified them as passive and active. Passive techniques comprise of geometric modification such as surface structuring, fluid additives and flow changing devices. Active heat transfer enhancement techniques are methods which require external power source.

Heat transfer enhancement using passive techniques have been of interest to the industry. Special structured surface tubes are commonly used in evaporators, boilers, condensers and single phase application heat exchangers. In two phase flow boiling the heat transfer coefficient highly depends on dynamic behavior of bubbles. Different phenomena affecting the dynamic behavior of bubbles in two phase flow boiling have been investigated and presented in detail by Abdous et al. (2019) and Holagh et al. (2020).

There is limited experimental test data available on in-tube boiling and evaporation of ammonia in the open literature. One of the earliest ammonia in-tube boiling studies was conducted by Peavy (1950) on a flooded single horizontal tube evaporator and probably first ever to report the effect of oil on heat transfer coefficient (HTC). Data was reported for saturation temperature 1.06°C and 3.44°C. Shah (1974) used an industrial evaporator and experimentally studied the in-tube heat transfer and pressure drop of ammonia. Evaporator was made from commercial grade steel pipe with inside and outside diameter of 26.2 mm and 33 mm, respectively. The inlet and outlet temperatures were measured with thermocouples placed directly into the flow path. Three heating cables were mounted at 120° along the length of the evaporator. The leaving two phase ammonia from the test section entered a separation unit where vapor ammonia was drawn out and condensed via chiller. Liquid ammonia returned to the separation unit after condensation and pumped again into the test section. Experiments were carried out for a saturation temperature range of 0°C to -40°C, mass flux range 0.016 kg s−1 to 0.83 kg s−1 and heat flux ranged from 0.58 kW m−2 to 2.3 kW m−2. Shah (1976) also presented a graphical method for the prediction of two-phase heat transfer coefficient (HTC). This technique provided a general method for predicting the HTC for any refrigerant and condition. Ammonia was not present in the chart database but according to Shah (1976) the chart could be used to estimate the HTC for pure ammonia assuming it to be a Newtonian fluid and the properties lie between water and R-12.

Sandru and Chiriac (1978) investigated ammonia flow boiling in horizontal tube under test conditions of saturation temperature of -5°C to 5°C, heat flux from 7 kW m−2 to 25 kW m−2 and mass flux from 200 kg m−2 s−1 to 600 kg m−2 s−1. Thermocouples were mounted on the bottom and top portion of 0.5 m long test section. They presented two correlation for the saturation temperature (-5°C to 5°C) range.

Gungor and Winterton (1986) developed a general correlation based on a data bank containing more than 4300 data points for forced convection boiling. They developed a correlation for data extracted from research work of 28 authors. The data points were taken for refrigerants, ethylene glycol and water in vertical and horizontal tubes under saturated boiling condition. The correlation is also applicable to annuli and sub-cooled flow boiling.

Malek and Colin (1988) aimed to propose a design correlation for determining the size of the industrial evaporators in nuclear power plants or energy recovery systems. Experiments were performed on electrically heated tube 10 m long and 21.6 mm in diameter. The temperature of ammonia was measured with thermocouple placed directly into the flow of ammonia while the wall temperature was measured at 15 locations equally spaced 0.625 m apart on the tube. Experiments were conducted at saturation temperature range of 30°C to 75°C, range of mass flux 27 kg m−2 s−1 to 54 kg m−2 s−1 and heat flux 2 kW m−2 to 20 kW m−2. They presented two correlation for two phase HTC.

Kandlikar (1990) presented a correlation for the estimation of HTC in vertical and horizontal tubes. The correlation was based on ten fluids with 5246 data points extracted from 24 studies. The proposed correlation can be used for fluids other than those included in the data bank through the inclusion of fluid dependent parameter. Like Gungor and Winterton (1986), for single phase HTC Dittus-Boelter equation was used.

Liu and Winterton (1991) proposed a correlation for two phase HTC as a combination of forced convection flow boiling and nucleate flow boiling. They introduced the Prandtl number dependence on convective boiling while the nucleate boiling is independent of boiling number.

Kattan et al. (1998) presented a general correlation for flow boiling inside horizontal plain tube. They incorporated the effect of flow pattern in the presented model. The method predicts better accuracy especially for condition at higher vapor quality (x > 85).

Zürcher et al. (1999) performed experiments to study the heat transfer of pure ammonia using a plain horizontal stainless steel tube having 14 mm internal diameter and 3.26 m length. The heat transfer length of the test tube was 3.01 m centered in another tube. The system was configured as counter flow arrangement with hot water in the annular section. Thermocouples were installed in the cut shallow grooves at two different axial locations and integrated to a computer-controlled acquisition system. Experiments were carried out for a saturation temperature of 4°C and range of mass flux was kept at 20 kg m−2 s−1 to 120 kg m−2 s−1. The range of heat flux was 5.2 kW m−2 to 19.5 kW m−2 for lower heat flux and 17 kW m−2 to 71.6 kW m−2 for high heat flux over a vapor quality of 1% to 100%. The local HTC was predicted with the help of Kattan et al. (1998) model in different flow regimes i.e., stratified wavy, stratified, annular, intermittent, and annular with partial dry out.

Kabelac and De Buhr (2001) presented data for local HTC of ammonia boiling in horizontal smooth and finned tube. The experimental setup was operational in three different loops which are condensation and evaporation loop (main loop), heating loop and refrigeration loop. The vapor quality was controlled from 0 to 0.9 with the help of electric powered pre-evaporators. Two test tubes were used for experiments one of which was an aluminum tube with 10 mm inner and 11 mm outer diameter. The second tube was a low finned tube with 21 spirally square shaped fins on the perimeter. The fin spiral angle and height of the fin was 25° and 0.63 mm, respectively. Twelve thermocouples mounted circumferentially at three different locations were used to collect the test tube wall temperature. The range of saturation temperature was set from -40°C to 4°C. Measurements were conducted for the evaluation of local HTC in the range of mass flux 50 kg m−2 s−1 to 150 kg m−2 s−1 and heat flux range 17 kW m−2 to 75 kW m−2. At nearly 0°C the finned tube showed better enhancement than the smooth tube because of surface enlargement. With decrease in temperature the enhancement increased due to larger heat transfer area. Noticeable enhancement was also observed at lower mass flux.

Kelly et al. (2002) performed experimental investigation on double pipe heat exchanger with ammonia in the tube and R-134 on the annular side. The data was collected for both micro fin and smooth tube. Both tubes were aluminum with an outer diameter of 12.5 mm and 3.04 m length. The inner diameter of the smooth and micro fin tubes were 10.9 mm and 11.8 mm, respectively. The enhanced tube consisted of 60 fins with a fin height of 0.254 mm and helix angle of 17°. The experimental facility was run on saturation temperature of 5°C, -10°C, and -20°C under heat flux of 820 W m−2, 2710 W m−2 and 5430 W m−2 with mass fluxes of 9 kg m−2 s−1, 27 kg m−2 s−1, 47 kg m−2 s−1, and 61 kg m−2 s−1. The average HTC was evaluated at nominal vapor quality, 15% to 95%. They found that the micro fin tube enhanced the HTC relative to the smooth tube at low flow rates for laminar and stratified flow.

Vollrath et al. (2003) conducted tests with R-22 for enhanced and plain tubes made of aluminum. Majority of data was collected at saturation temperature of 35°C, with a few data points taken at 40°C and 53°C. The mass flux was kept in the range of 20 kg m−2 s−1 to 270 kg m−2 s−1 over a vapor quality of 0% to 95%. The test heat flux was 2 kW m−2 to 10 kW m−2. Experimental data showed an average increase of pressure drop by a factor of 1.4 for the enhanced test section regardless of mass flux or quality. Heat transfer data showed a decreased heat transfer coefficient in the enhanced test section. When coupled with the area enhancement, the enhanced test section performed much worse when compared to basic or smooth test section.

Boyman et al. (2004) carried out experiments on plain tube with ammonia boiling in the presence of immiscible oil. They studied the ammonia flow boiling inside plain horizontal tube and the effect of immiscible oil on the HTC. The test facility consisted of horizontal tube section having 2 meter length and an internal diameter of 14 mm. Ammonia boiling was accomplished by water glycol solution flowing counter to the refrigerant in concentric double pipe arrangement. Immiscible oil Gargoyle Arctic SHC 326 was employed with ammonia during experiments. The experimental study was carried out for a saturation temperature varying between -10°C to 10°C. The range of mass flux and heat flux was set from 40 kg m−2 s−1 to 170 kg m−2 s−1 and 10 kW m−2 to 50 kW m−2. The vapor quality was varied from 0.15 to 1. There was a significant decrease in the HTC with a minute amount of oil. When the oil concentration surpassed 0.1%, no further decrease was observed in the reduction of HTC.

Khir et al. (2005) performed flow boiling experiments with ammonia-water mixture in a vertical stainless steel co-axial tube (d = 28 mm). Three ammonia mass fractions (0.42, 0.55 and 0.61), three mass fluxes (707, 1590 and 2688 kg m−2 s−1) and three heat fluxes (8.2, 14.8 and 18.5 kW m−2) were varied simultaneously in their experiments, while the pressure ranged from 1.5 to 20 bar and the vapor qualities ranged from 0 to 0.6. They compared their experimental data with the correlations by Mishra et al. (1981), Bennett and Chen (1980) and the pure fluid correlation by Jung and Radermacher (1991), and found that only the Mishra et al. (1981) correlation gave reasonable predictions.

Komandiwirya et al. (2005) performed the experiments on enhanced and smooth tubes for pure ammonia. Ammonia mixed with oil was used only in plain smooth tubes. The tube internal diameter was 8.1 mm. The experimental facility consisted of three loops alike to Vollrath et al. (2003). The refrigerant saturation temperature at the inlet was fixed to 35°C. The mass flux and heat flux were varied between 20 kg m−2 s−1 to 270 kg m−2 s−1 and 2 kW m−2 to 10 kW m−2, respectively. It was concluded that for smooth tube, HTC increased with inlet quality except at mass flux of 80 kg m−2 s−1. For ammonia oil mixture the HTC decreased at high heat flux and vapor quality. In general, increase in the oil concentration caused a decrease in HTC. For enhanced tube the fin height could influence the HTC compared to the smooth tube. The heat transfer for altered enhanced tube was higher than that of a smooth tube at any value of mass flux and quality because of the enhancement factor greater than unity. The heat transfer results were two to three times lower for unaltered enhanced tube than the smooth tube.

Park and Hrnjak (2008) analyzed the experimental data of Komandiwirya et al. (2005).They compared the results of different correlation for ammonia with other refrigerants and showed higher HTC up to 300% for ammonia. When the mass flux exceeded 120 kg m−2 s−1, the heat transfer displayed a strong sensitivity to vapor quality. The impact of vapor quality on heat transfer was negligible when the mass flux was lower than 80 kg m−2 s−1.

Shah (2017) presented a modified version of his previous correlation Shah (2007) for calculation of local HTC during saturated boiling in bundles of plain and enhanced tubes. The correlation was verified with range of data of twelve different fluids including ammonia and for different variety of enhanced tubes of different materials having different arrangements. A wide range of experimental conditions were covered. It was reported that correlation predicts the HTC with mean absolute deviation of 15.2%.

Gao et al. (2018) performed experiments to investigate the flow boiling heat transfer and pressure drop characteristics of ammonia in 4 mm diameter horizontal plain tube. Heat flux was varied between 9 kW m−2 to 21 kW m−2, mass flux from 50 kg m−2 s−1 to 100 kg m−2 s−1, and saturation temperature from -15.8°C to 5°C. The heat transfer coefficient was reported to increase with heat flux. The heat transfer coefficient also increased with mass flux in low vapor quality regions while at higher quality regions the effect of mass flux was observed to reverse. Similarly, in low vapor quality regions, the effect of saturation temperature was little while at high quality regions and high mass flux the heat transfer coefficient decreases with decrease in saturation temperature. The two-phase flow pressure drop was reported to increase with increase in vapor quality and mass flux, while decreased with increase in saturation temperature.

Gao et al. (2019) Also studied the effect of miscible oil on heat transfer performance of ammonia in 8 mm diameter tube. The concentration of miscible oil was varied from 0% to 5.78%, the range of mass flux was between 51 kg m−2 s−1 and 99.5 kg m−2 s−1, with heat flux ranging from 9 kW m−2 to 21 kW m−2, and saturation temperature between -5.5°C and 5°C. It was reported that miscible oil has a negative effect of heat transfer performance of ammonia. The effect of miscible oil in low vapor quality regions was reported to be insignificant, while at high vapor quality regions the increase in oil concentration causes significant deterioration in heat transfer performance. It was also reported that miscible oil concentration has little or no effect on two-phase flow pressure drop.

Present study aims to investigate the two-phase boiling heat transfer characteristics of ammonia in a vertical evaporator with seven dimpled enhanced tubes under a wide range of experimental conditions. The saturation temperature was varied in four levels between -2°C and -20°C. At each saturation temperature, mass flux was varied between 34.99 kg m−2 s−1 and 66.32 kg m−2 s−1 with exit vapor quality varying between 0.1 and 0.9 at each level of mass flux.

Section snippets

Experimental setup

The experimental setup as shown in Fig. 1 consists of a refrigerant (ammonia) loop, a test loop and ethylene glycol/water solution loop. Ammonia flowed in a closed loop vapor compression refrigeration cycle of 65 kW capacity. The loop comprised of an oil separator, compressor, water cooled condenser, ammonia receiver vessel, vertical accumulator, and a control valve assembly. Ammonia at saturated condition circulated between the test evaporator and the vertical accumulator. Boiled off vapor

Vertical test evaporator

The experimental test evaporator used in the experimentation was a shell and tube type, single pass heat exchanger with enhanced dimpled tubes (Fig. 2), instrumentation probes and metering devices. The shell material was stainless steel SA312-TP304, 762 mm long with nominal diameter of 77.93 mm. The tube bundle consisted of 7 enhanced dimpled tube arranged in a 60° triangular pattern at 23.8 mm. The geometric detail and dimensions of the test evaporator and dimpled tube are given in Table 1.

Instrumentation

Temperatures of hot and cold fluid at inlet and exit were measured with PT-100 RTDs having an accuracy of ±0.1°C. All the RTDs were calibrated with a reference RTD before installation and were verified from time to time during experiments. A PC based data acquisition system with data logger was used to process and record data. Electronic turbine based digital flow meter was used for hot fluid flow measurement. The flow rate of hot fluid was measured with an electronic turbine flow meter and was

Data reduction

To determine the two-phase HTC in the tube side, the overall HTC, thermal resistance across tube wall and shell side single phase HTC should be known.

Considering the principle measurements of temperatures and flow rates and assuming no fouling on either side, the two-phase HTC (htp) on the refrigerant side can be determined by the following equation:1Uo=1htp+rikwln(rori)+riro1hsp

The two-phase HTC can be calculated from Eq. (2) as:1htp=1Uorikwln(rori)riro1hsp

The Overall heat transfer

Uncertainty analysis

According to the method presented by Moffat (1988) an experimental error analysis was performed for the evaluation of uncertainties in measurements and the results extracted from experiments. Eq. (10) is used to calculate the overall uncertainty “δR” for a calculated result. Different experimental parameters are listed in Table 2 with their range and experimental uncertainties.δR=[i=1n(RXiδXi)2]12

Where R is expressed as:R=R(X1,X2,X3,..Xn1,Xn)

Here each independent variable Xi has its own

Results and discussion

The effect of heat flux, saturation temperature, exit vapor quality and mass flux on two phase HTC are discussed here.

Comparison with previous studies

The results from this study are compared with previous studies as shown in Fig. 12 to Fig. 15. The comparison is performed at saturation temperature of -2°C and -20°C at mass flux of 34.99 kg m−2 s−1and 66.32 kg m−2 s−1. The comparison of the present data is performed with the studies performed and correlations developed by Malek and Colin (1988), Sandru and Chiriac (1978), Shah (2017), Cooper (1984), and Kattan (1998). Among these studies the correlations of Malek and Colin (1988) and

Conclusion

A vertical shell and tube heat exchanger comprising of seven dimpled enhanced tubes was tested under two-phase flow with saturated ammonia and a range of experimental parameters as follows.

  • Saturation temperature of ammonia -2°C to -20°C

  • Heat flux 5 kW m−2 to 45 kW m−2

  • Exit vapor quality between 0.1 and 1.0

  • Mass flow rate 0.0075 kg s−1 to 0.015 kg s−1

  • Mass flux 34.99 kg m−2 s−1 to 66.32 kg m−2 s−1

Ammonia under saturated conditions was flowing inside vertical dimpled tubes entering at the bottom and

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

The authors acknowledge Isotherm, Inc. USA for supplying the evaporator and other necessary measurement tools. Thanks are also due to Higher Education Commission (HEC) of Pakistan for support under SRGP-1757.

References (32)

  • C.Y. Park et al.

    NH3 in-tube condensation heat transfer and pressure drop in a smooth tube

    Int. J. Refrig.

    (2008)
  • M.M. Shah

    A correlation for heat transfer during boiling on bundles of horizontal plain and enhanced tubes

    Int. J. Refrig.

    (2017)
  • A. Abbas et al.

    Shell Side Single-Phase Experimental Heat Transfer Analysis of a Vertically Oriented Single Segmental Baffle Bundle With Dimpled Tubes

    Journal of Thermal Science and Engineering Applications

    (2020)
  • D.L. Bennett et al.

    Forced convective boiling in vertical tubes for saturated pure components and binary mixtures

    AIChE J.

    (1980)
  • A. Bergles et al.

    Bibliography on augmentation of convective heat and mass transfer-II. Iowa State Univ. of Science and Technology, Ames (USA)

    (1983)
  • Bergles, A., Jensen, M., Somerscales, E., Manglik, R., 1991. Literature review of heat transfer enhancement technology...
  • Cited by (3)

    • Experimental study of ammonia flow boiling in a vertical tube bundle: Part 4 – Comparative analysis

      2022, International Journal of Refrigeration
      Citation Excerpt :

      The variation of HTC with respect to exit vapor quality was similar to its variation with heat flux, because increasing heat flux also increased the vapor quality. This behavior has been discussed in detail by Abbas et al. (2021), and Ayub et al., (2021a, b)). Similarly, the HTC was higher at higher saturation and decreased with decrease in saturation temperature for given heat flux, mass flux and vapor quality as shown in Fig. 13 and Fig. 14.

    • Experimental study of ammonia flow boiling in a vertical tube bundle: Part 3 – Enhanced dimple tube with 2/3rd height solid round PVC nonconductive rod

      2022, International Journal of Refrigeration
      Citation Excerpt :

      These comparative results with plate evaporator indicate that the flow path created within the space between the tube inside surface and the solid PVC rod up to 67% tube height seem to be directing the flow in a similar way as within the chevron plates pair. The upper 33% of each tube with no rod, experienced similar flow characteristics as in Abbas et al. (2021). Further variation of tube geometry and optimizing the height of the rod to achieve similar performance as hard chevron plates would be preferable due to a few prominent advantages: (1) shell and tube exchanger is a rugged time-tested product, (2) carries only two gaskets versus hundreds in a plate and frame, (3) easy to maintain, (4) and surprisingly has a smaller footprint.

    View full text