- Email: [email protected]
Charging and discharging characteristics of cool thermal energy storage system with horizontal pipes using water as phase change material
Energy Conversion and Management 77 (2014) 755–762
Contents lists available at ScienceDirect
Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman
Charging and discharging characteristics of cool thermal energy storage system with horizontal pipes using water as phase change material H.H. Sait a,⇑, A.M. Selim b a b
Mechanical Engineering Department, King Abdulaziz University, Jeddah, Saudi Arabia Mechanical Engineering Department, College of Technology, Jeddah, Saudi Arabia
a r t i c l e
i n f o
Article history: Received 1 July 2013 Accepted 9 October 2013
Keywords: Falling film Ice formation Releasing Melting
a b s t r a c t An experimental investigation of ice formation on cold vertical banks of horizontal tubes subjected to falling-film– jet mode– is conducted. In the charging process, a set of internally cooled vertical banks of horizontal tubes of brine is subjected to a falling film of water. The formed ice is periodically observed, photographed and measured in falling-film jet mode at specific internal coolant (ethylene–glycol solution) flow rates and temperatures. In the discharge process, the same solution is heated and used internally to release ice. Different thicknesses of the released ice are observed and measured. The maximum quantity of released ice is obtained and the optimum ice formation is determined. The results indicate that the ice formation and the solid ice released are controlled by the thermal resistance of the ice, time and pitch between tubes. The maximum gained ice has a thickness that is approximately equal to half of the tube spacing between the tubes utilized, which is formed in approximately 45 min and released in 12.5 min. The variation in heating solution temperature has a slight effect on the gained ice and discharging time. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction There are two methods to store and reuse formed ice. The first method involves storing ice on the tubes of the freezing system. The second method involves releasing ice from the tubes and storing it in an isolated reservoir. In the first method, the freezing system is the storage system. Thus, the ice obtained by the system must be as thick as possible. In the second method, the ice has to be released with the minimum amount of melted ice and stored in an isolated receiver. In this case, the freezing system can be reused several times, which increases its capacity. This latter method is considered in this study. The topic of ice formation has been explored by some researchers [1,2,4–8,10]; the topic of ice melting has also been examined [3,9,11,12]. Sait et al. [1,2] performed an experimental investigation on the freezing of water falling film on a vertical bank of horizontal cold tubes. The brine flows into the concentric tubes in parallel. The authors focused their work on ice formation characteristics and heat transfer for the three main modes of falling film: droplets, jets and sheets. These researchers determined that the formation of ice depends on falling film and coolant flow rates. In addition, the overall heat transfer coefficients are controlled by the thermal ⇑ Corresponding author. Tel.: +966 26914730, mobile: +966 560007382; fax: +966 22564933. E-mail address: [email protected] (H.H. Sait). 0196-8904/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.enconman.2013.10.034
resistance of ice. Habeebullah [3] conducted an experimental study on ice formation around a horizontal long copper tube (L = 12.3 m with tree returned bends of 180° and d = 19.5 mm) immersed in water. This researcher found that there was a slope of the ice thickness, in which the axial distance depended on time but varied with coolant flow rate and Stanton and Biot numbers. The axial growth rate of ice was distinct for low values of the coolant Reynolds number and short freezing times. He also discovered an unexpected enlargement of ice thickness on the surface of the tube bends. Cabeza et al. [4] added stainless steel pieces, copper pieces and graphite matrix impregnated with water as phase change materials (PCM) to improve heat transfer. They founded that addition of stainless steel pieces in the PCM does not increase the heat flux significantly. However, addition of copper pieces and the use of graphite composite enhance heat transfer significantly. Ismail and Jesus [5] performed a parametric study of the solidification of PCM around a cylinder for an ice-bank application. They concluded that a lower initial temperature of the liquid phase seemed to accelerate the solidification. The thermal conductivity of the tube wall material can have a considerable influence on the velocity of the process. Kayansayan and Acar [6] analyzed ice formation around a finned-tube heat exchanger for cold thermal energy storage. They concluded that under identical flow and inlet conditions, the heat exchanger with finned tube stores a maximum of 45% more energy than the bare tube of a turbulent flow regime (Re > 3000).
756
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
Nomenclature A Cp d h k L M _ m Q_ T T U
aria (m2) specific heat (J/kg K) tube diameter (m) heat transfer coefficient (W/m2 K) thermal conductivity (W/m K) latent heat (J/kg) mass (kg) mass flow rate (kg/s) heat transfer rate (W) temperature (°C) average temperature (°C) overall heat transfer coefficient (W/m2 K)
Greek symbols ‘ tube length (m) s time (s)
exim exo h hi ho i io Lmice Mice o oo s smice sw tmi w 0
experimental ice melting overall heat transfer coefficient hot hot in hot out internal inner of outer tube latent heat of melted ice melted ice out outer of outer tube surface sensible heat of melted ice sensible heat of water theoretical melting ice water melting point
Subscripts cu copper
Eames and Adref [7] studied freezing and melting of water in the type of spherical enclosures that are used in thermal (ice) storage systems. They concluded that 90% of the cold could be extracted from the ice storage element within 70% of the time required for complete discharge. Hirata and Matsui [8] studied heat transfer with freezing and thawing for water flow around isothermally cooled cylinders in staggered and aligned arrangements. They demonstrated that the ice filling rate is strongly affected by the Reynolds number, cooling temperature, and cylinder pitch perpendicular to water flow. Melting occurred twice as fast as freezing occurred. Yingxin and Zhang [9] studied thermal processes for internal melting in an ice-on-coil tank, including the variation in ice-water density. They analyzed both charge and discharge processes for the ice-on-coil tank. They concluded that the discharge model is only applicable for processes that begin when the ice cylinder slightly overlaps and unfrozen water is present, which causes the ice to float. Vargas et al. [10] experimentally and analytically studied the fundamentals of melting ice-shell rides on heated horizontal cylinders. They assumed and proved that natural convection is negligible compared with the natural convection of the heated cylinder. The melted water is drained axially. Most of the melting is due to direct contact with the upper half of the hot cylinder. Their results indicate that the melting process consists of two distinct regimes: the first regime occurs when the cylinder is surrounded by ice, which consumes the majority of the melting time, and the second regime occurs when the cylinder cuts through the upper portion of the ice sleeve. The time until the ice falls off the cylinder can be obtained from a specific graph. Wu et al. [11] studied discharging characteristics by modeling cool thermal energy storage systems with coil pipes using n-Tetradecane as a phase change material. The results demonstrate that the higher the flow rate of the heat transfer fluid or the higher the inlet temperature of the heat transfer fluid, the higher the cool release rates, and less time will be required during the discharge process. The diameter of the coil pipes has little influence on the discharge process compared to the other variables previously mentioned. Masahiko et al. [12] studied the performance analysis of the liquid–ice thermal storage system for optimum operation. They concluded that intensive melting results when slush ice is pulled by buoyancy to the top of the hot capsule and when close contact occurs. The local heat transfer is most valuable at the top of the vissle
and decreases along the vissle wall due to the stratified layer and free convection. The average heat transfer coefficient increases as heat flux increases. The melting rate increases monotonically as a function of time, irrespective of heat flux and solution concentration. Tsuyoshi et al. [13] conducted an experiment on the melting of slush ice in a horizontal cylindrical capsule. Their results demonstrated that the COP was almost the same for all three operation modes, whereas the performance of heat release mode for POR decreases with an increase in running time or storage time. The system simulation suggests the potential for obtaining optimum operational conditions, such as the daily running time of thermal storage, among the employed operational modes. The previously described study examined freezing and thawing (melting) of stagnant or moving water flow. The freezing of falling water film on horizontal tubes and its characteristics were also examined. For cold thermal storage where, the accumulated ice is used during the peak electrical load periods, it is necessarily to investigate more about the behavior of the discharge cycle (melting process). The present study investigates both charging and discharging of ice on and from tube surfaces, but focusing more on the melting process since the freezing process was explained well by Sait et al. [1,2]. The quantity of optimally formed ice and ice releasing behavior are also investigated for a falling-film jet mode. 2. Experimental apparatus The experimental apparatus shown in Fig. 1 is designed and fabricated to allow falling film to freeze outside the tubes and to obtain the maximum quantity of released solid ice. A detailed description of the apparatus is provided in Sait et al. [1,2]. The main differences between the two apparatuses are as follows: (a) The coolant in the present study flows through the test concentric tubes in series only to achieve accurate temperature measurements, whereas the tubes were designed in series or in parallel in the previous apparatus. (b) A catching tube is used to get accurate temperature of the outlet falling film or the released iced. (c) A strainer is used under the tubes in this study to separate the released solid ice from the melted ice. (d) The hot solution flow rate and temperatures are recorded during the discharge cycle (melting process), which was not the case in the previous design of [1,2], to determine the heat of release.
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
757
Fig. 1. Schematic of experimental apparatus.
The test section, which consists of a supporting box and test tube arrangement, is comprised of seven concentric vertical horizontal tubes of copper, each containing inner tubes with diameters of 15.8 mm and outer tubes with diameters of 28.5 mm. All tubes are 235 mm long and 1 mm thick. They are fixed vertically underneath the feeding tubes at a pitch of 57 mm center-to-center. The outside of the concentric tube underneath is connected to the inside of the concentric tube on top for coolant series flow. Four fluid loops are used effectively to form ice on the tubes using falling film, as well as to release the ice from the tubes. The following loops are employed: the falling film loop, the coolant loop to freeze falling film, the discharging loop to release the gained ice and the pre-cooling loop to pre-cool water in the upper constant head tank to minimize sensible heat removal. The falling film loop activates when the pre-cooled water flows through the adjustable valve and flow meter to the upper feed tube in the test section, as shown in Fig. 2. The water originates from the upper feed tube and flows around the lower tube, where it is distributed into a uniform layer. It then falls and flows onto the test tube surfaces. Some of the water film freezes outside the cold tubes, and some of the water flows down to the catching tube and through the strainer to the water pool. The remaining water is pumped to the pre-cooled constant head tank through an adjustable valve and filter. The coolant loop, which is used to freeze falling film, activates a charging cycle when the cold solution (40% ethylene glycol by weight in water), at a temperature of approximately 10 °C, flows from the refrigeration machine tank (TD-30) through the feeding valve and flow meter to the test tube section. The coolant feeds the concentric test tubes from the inner tube underneath to the outer tube, to the inner tube of the outer tube located on top, and subsequently to the upper test tubes. The coolant returns to the refrigeration machine tank through a control valve. The discharging loop, which is used to release ice, is activated when the temperature of the worm solution ranges from 15 to 30 °C (40% ethylene glycol by weight in water), which is similar to the process in the coolant loop. Water is pumped from the worm reservoir through the control valve and then flows internally through the flow meter to the test tubes to release ice. It then returns to the worm reservoir through the return valve. Some of ice is melted, whereas other ice is released; both are dropped over
Fig. 2. Schematic of the path of falling film around tubes in the test section (copper tube; L = 235, S = d = 28.5 mm.).
the strainer. The melted ice water falls to the pool, whereas the released ice remains on the strainer. The height of the water in the pool is measured before and after the discharge process to obtain the quantity of melted ice. The released solid pieces of ice, which are collected on the strainer, are allowed to fall into the water pool. The height of the water in the pool is measured again to obtain the volume of water (considering 11% of the ice is floating). Activation of the pre-cooling loop, which is used to pre-cool the water in the constant head pre-cooling tank, allows the coolant to flow from the cooling machine through the pre-cooling feeding valve to the immersed cooling coil in the constant head pre-cooling tank. The water is cooled to a fractional degree (approximately 0.3 °C). It is then allowed to flow through the pre-cooling return valve to the refrigeration machine. 3. Measuring techniques The measuring points are shown in Fig. 3. The falling-film flow rate is measured by a calibrated flow meter with an accuracy of 1%. The splashed water from the tubes is collected and measured in a scaled container. The temperatures of the falling film at the entrance of the feed tubes, at the outlet of the test section (in the catching tube), and at the constant head tank are measured with calibrated T-type copper–constantan thermocouples, which are immersed in the tubes by a specially fabricated well with an accuracy of 0.05 °C. The coolant flow rate is measured by a calibrated flow meter with an accuracy of 1%.The coolant temperatures at the inlet and
758
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
The recirculation water pump is activated and its suction valve is adjusted to re-circulate the remaining water in the pool and to maintain suitable head in the tank. The coolant loop is then activated and the coolant solution is allowed to enter the test section. Ice begins to form around the test tubes. The flow rates of the coolant and falling film are observed and recorded periodically for every run. The pre-cooler, inlet and outlet temperatures for both falling film and coolant, as well as for the test tube surfaces, are recorded periodically. The ice formation is observed and photographed. At the desired time, the charging cycle is stopped and the level of stagnant water in the pool is measured. The discharging cycle is activated to release and obtain the quantity of gained solid ice. Some of ice is melted and drained to the strainer and subsequently drained to the pool. The remaining solid pieces of ice are released (depending on the ice thickness) and dropped on the strainer. The level of stagnant water in the pool is measured again (the difference in water level determines the melted ice volume). The solid pieces of ice are then allowed to fall from the strainer to the water pool and the water level in the pool is again measured. By considering that approximately 11% of the ice volume is floating, the difference in water level determines the solid ice volume. The ratio of the solid ice pieces to the total ice volume is obtained.
Fig. 3. Experimental measuring positions.
outlet of the test tubes are measured with a T-type copper–constantan thermocouple, which is immersed in a well-calibrated thermocouple with an accuracy of 1%. The surface temperatures of the specimen test tube are measured using calibrated 36-gage, copper–constantan thermocouples [1]. The differences in pool height of the test section before and after the discharge cycle are measured by a fixed scale with an accuracy of 1%. This step was performed to obtain the volumes of the melted and solid released ice. 4. Experimental ranges The flow of the falling film is controlled to maintain a steady-jet mode under uniform conditions for the majority of the seven test tubes. Its temperature is reduced to a certain value to minimize the sensible heat removed by the test tubes. The coolant flow rate and temperature are controlled to achieve the required absorbed heat to obtain reasonable ice formation. The hot solution flow rate and temperature are controlled to achieve the required added heat to release solid ice with minimum melted ice. Table 1 shows the variable ranges of flow rates for these fluids.
6. Observation remarks The falling-film jet mode, ice formation and released solid ice behavior are observed, photographed and studied. 6.1. Characteristics of falling-film jet mode Jet-mode flow is steady, uniform, and encompasses all test tubes. All surfaces are wetted and approximately 8% of water splashes out of the down test tubes. By adding a spot of color to the flow, it is observed that the jet streams diffuse as they move down the test tubes; however, they maintain their steadiness and their let mode along their path, as shown in Fig. 4. 6.2. Ice formation characteristics The formation of ice begins at the bottom of the test tubes because the lowest temperature of the coolant exists there and the falling film loses heat as it falls onto the cold tubes. The slope of
5. Experimental procedure The refrigeration machine is operated to reduce the coolant solution temperature to approximately 10 °C. The coolant is then admitted to the constant head pre-cooling tank to reduce its water temperature to approximately 0.3 °C. A warm solution heater is activated to obtain the required temperature. The flow rate of the falling film is adjusted to achieve a uniform and steady jet.
Table 1 Variable ranges of fluid flow rates. Fluid
Flow rate (kg/s)
Temperature (°C)
Falling film (water) Coolant (ethylene–glycol solution) Hot solution (ethylene–glycol solution)
0.0425 ± 0.005 0.162 ± 0.002 0.082 ± 0.002
0.3 ± 0.1 10 ± 1 15–30 ± 0.5
Fig. 4. Sample photograph of jet mode.
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
759
the ice thickness pulls the streams toward the bottom of the tubes. A greater accumulation of falling film exists at the bottom of the tube than at the top of the tube, which allows greater accumulation of ice at the bottom of each tube. The ice is nearly regular on most of the tubes, with the exception of the upper end of the tube due to water falling there. Fig. 5 shows a sample photograph of ice formation on the 7 tubes. Ice accumulates circumferentially on the tube surfaces until the tube spacing is filled with ice. By adding a spot of color to the flow, it is revealed that the slope of contaminated ice on the tube moves the flow streams toward the tube end, which causes ice accumulation at this end. 6.3. Ice releasing characteristics During the peak load, it is necessary to use the formed ice in the cold thermal storage for cooling. The formed ice can be stored either on the tubes or it can be collected in thermal storage. To obtain the desired quantity of ice to manage the peak load, many charging and discharging cycles need to be activated. As the ice
Fig. 7. Sample photograph of ice releasing.
is formed during the charging cycle and reaches its optimum quantity, a discharging cycle must be activated to release the ice from the test tubes and collect it in thermal storage. For the discharging cycle in this experiment, a solution similar to the solution used in the cooling loop—it exhibits a temperature range of 15–30 °C— is used to release the ice from the tubes and to measure its volume for evaluating the remaining ice to be stored. The ice begins to melt from the inner surface of the formed ice toward the outer surface. The melted ice, which is now in the form of water, seeks a way to escape the surrounding ice. A portion of the ice flows down the test tubes from both ends of the test tubes. The remaining formed water flows onto the ice layer; as a result, the water volume expands circumferentially. Due to the effects of the warm water, the weight of the ice and gravity, the top ice layer is milted faster than the bottom ice layer. The melting phenomenon continues and the ice thickness decreases until solid pieces of ice suddenly fall from the test tubes. Fig. 6 displays a schematic diagram for ice formation and releasing. If the ice thickness fills the space between the tubes, the melted ice increases and some ice is suspended between the tubes until more ice is melted, which leads to a greater amount of time required to release all of the ice. Fig. 7 shows a sample photograph of ice releasing. Fig. 5. Sample photograph of ice formation.
7. Heat transfer analysis during the freezing and melting process The heat transfer analysis during the solidification process was well explained by Sait et al. [1]. They demonstrated that the rate of heat transfer decreases as the amount of ice that accumulates on the test tubes increases, which causes the overall heat transfer coefficient to consequently decrease. The formed ice increases as both the coolant flow rate and the falling-film flow rate increase. The absorbed heat required to release ice consists of the following heats: sensible heat of sub-cooled ice, latent heat of melted ice and sensible heat of melted water. Thus, the experimental ice melting Q_ exim can be expressed as
Q_ exim ¼ Q_ smice þ Q_ Lmice þ Q_ sw
ð1Þ
where
Q_ smice ¼ M mice Cpmice ðT s T 0 Þ=smice
ð2Þ
and Fig. 6. Schematic for (a) ice formation and (b) ice releasing.
Q_ Lmice ¼ M mice Lmice =smice
ð3Þ
760
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
Table 2 Fluids properties, [14].
l (N s/m2)
Fluid Water (0.25 °C) Ice (5 °C) Cold solution (10 °C) Hot solution (25 °C)
1750 10 – 8 103
6
3.37 103
q (kg/m3)
k (W/m K)
1000 920 1060
569.6 10 1.88 0.467
1050
0.43
Cp (kJ/kg K) 3
4.2 2.040 3.3488 3.30
and
8.2. Characteristics of ice releasing
Q_ sw ¼ M mice Cpw ðT w T 0 Þ=smice
ð4Þ
This heat is added by the heating solution Q_ hav , which is expressed as
Q_ hav ¼
8.1.2. Effect of time on accumulated mass of ice Fig. 8 portrays the effect of time on ice mass formation. Ice formation distinctly increases with time. It begins increasing at a relatively high rate until it reaches 1.5 kg within 30 min. The rate then decreases until it reaches 2 kg within 45 min. The rate decreases further until it reaches 2.8 kg within 75 min. As the ice thickness increases, its thermal resistance increases and its heat transfer decreases. Because ice has a low thermal conductivity, it acts as an insulator for the heat transfer.
R t¼n _ Q dt t¼0 h time
ð5Þ
where t is the time interval, n is the indicated time, and
_ h Cph ðT hi T ho Þ Q_ h ¼ m
ð6Þ
The overall heat transfer coefficient Uexim of the experimental ice releasing is expressed as follows:
U exim ¼ Q_ exim =Aice ðT h T 0 Þ
ð7Þ
In these experiments commercial ethylene glycol solution (40% by weight) is used as a coolant and for heating also. The concentration is obtained by measuring its specific gravity experimentally, then using its value to get the concentration from the specific gravity graph of the ethylene glycol from Green and Perry [14]. All fluid properties are listed in Table 2.
The discharge cycle is used to determine the quantity of formed ice and the quantity of released ice. The quantity of released solid ice is measured and the percentage of released ice is calculated for various discharge solution temperatures. 8.2.1. Quantity of released ice In the discharge cycle some of the ice near the tube surface is melted. The melting continues until the remaining ice fall down from the tube and catches by the strainer. The quantity of the formed ice and consequently the ice thickness, affects the possible released of the solid ice. When ice thickness is small, less than 3.75 mm, no solid ice can be gained with discharge cycle. As the ice thickness increases to a maximum value of 17 mm, the collected solid ice increase and about 1.67 kg can be collected as shown in Fig. 9. As ice thickness increases more, the collected solid ice stays the same or decrease due to more melting is required to force the accumulated ice to fall down from the tubes. This effect is due to the relationship between the discharge heat flux and melting heat. When the quantity of ice is small, all of the ice is melted. This condition occurs because the discharge heat is greater than or equal to the melting heat of the ice surface that contacts the heated tube (tube surface area). When the ice quantity
8. Results and discussion For thermal storage applications, optimal ice quantities and optimum ice releasing at suitable intervals helps to store the required quantities of ice and accommodate required thermal load schedules. 8.1. Ice freezing characteristics Ice freezing quantity is affected by the falling-film mode (flow rate) and its temperature, as well as by coolant flow rate and coolant temperature. For the freezing process, the convection heat transfer coefficient (on the outer surface of ice) is affected by falling film flow rate as was stated by Sait et al. [1]. However, the variation of the ice thickness results in changing of the outside surface area of the ice which will affect outside convection heat transfer.
Fig. 8. Effect of time on ice thickness.
8.1.1. Effect of falling film on accumulated mass of ice As demonstrated by Sait et al. [1], the falling-film mode affects ice freezing. The outside heat transfer convection coefficient increases as the falling film flow rate increases and changes its modes from droplet, jet and sheet, which causes more ice to accumulate and consequently more ice thickness. However, this increase has a limit as was explained in Sait et al. [1], due to backsplash which causes some falling film liquid to step out of the test tubes. In addition as ice thickness increases, the ice resistance increases which limits the rate of ice formation. The moderate jet mode is steady for most of the tubes and contains a small amount of splashed water; therefore, it is chosen as the model for ice formation and ice releasing in this study.
Fig. 9. Effect of ice thickness on released ice quantity.
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
761
increases, some of the ice forms sleeves that suspend from the tubes (refer to Fig. 6). The solid ice sleeves expand as the quantity of accumulated ice increases. As the ice thickness increases, the sleeves become stuck in their tubes and in the tubes underneath. Thus, the quantity of released ice decreases due to melting from the bottom and top of the tubes. The maximum released quantity is approximately 1.67 kg with ice thickness of 17 mm. This thickness is slightly more than half of the tube spacing between the tubes (14.25 mm). 8.2.2. Released ice percentage Fig. 10 shows the percentage of the released solid ice for various average ice thicknesses. As described previously, the maximum released ice thickness is 17 mm and the maximum percentage of ice is 68%. A discharge solution temperature of 23 °C (room temperature, which is practically applied) is initially used for releasing ice. When the average ice thickness is small—approximately 3.5 mm—all of the ice is melted and no solid pieces of ice remain. When the ice thickness increases to 7.9 mm, approximately 54% of the ice remains. When the ice thickness increases by a maximum of 13.6 mm, approximately 63% of the ice remains. When the thickness increases to 17 mm, the percentage of released ice increases to 68%. When the thickness reaches 22 mm, the percentage of released ice decreases to approximately 50%. As the ice thickness reaches 27.9 mm, the percentage of released ice decreases to 41%. The maximum percentage occurs at approximately 17 mm of ice thickness, as shown in Fig. 10. 8.2.3. Effect of hot solution temperature on ice releasing During the discharging process, part of ice layer is melted and the rest falls down and collected by the strainer. With increasing the temperature of the heating solution, the melting process occurs fast, which makes the solid ice falls down more rapidly. However, as the temperature of the hot fluids increases, the melting process consumes more from the formed ice. That is why no real effect is
Fig. 12. Effect of ice thickness on the overall heat transfer coefficient for solid ice releasing for heating fluid flow rate = 0.162 kg/s.
noticed on the percentage of ice releasing within experimental temperature range of the heated solution which varies from 15, 23 and 30 °C as shown in Fig. 11. 8.3. Heat transfer coefficient for solid ice releasing As was explained earlier by Eqs. (1)–(7), the experimental overall heat transfer coefficient depends on heat transfer from the heated solutions to the solid ice which comprises of three quantities of heat transfer namely sensible heat of sub-cooled ice, latent heat of melted ice and sensible heat of melted water. As shown in Eq. (7), ice surface area and the temperature difference between the heating fluids and melting point affects the experimental overall heat transfer coefficient Uexim. Fig. 12 shows the variation of the experimental overall heat transfer coefficient Uexim for solid ice releasing with ice thickness. As the ice thickness decreases as melting is taken place, Uexim is increasing to reach its maximum value of 350 W/m2 K when the ice thickness decreases to 4 mm. 9. Comparison of time of ice formation with time of ice releasing
Fig. 10. Effect of ice thickness on percentage of released ice.
Fig. 11. Effect of heating solution temperature on released ice percentage.
Fig. 13 includes a comparison of the time of ice formation with the time of ice releasing. The figure indicates that when the average ice thickness is small—a maximum of 3.5 mm, which is formed in 5 min—all of the ice is melted in 2 min. When the thickness increases—a maximum of 7.9 mm, which is formed in 15 min— approximately 54% of the ice is released in 6 min. When the thickness increases further—a maximum of 13.6 mm, which is formed in 30 min—approximately 63% of the ice is released in 10 min. When the thickness increases to 17 mm, which is formed in 45 min, the percentage of released ice increases to
Fig. 13. Comparison between time of ice formation and ice releasing at certain ice thickness.
762
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
will make it very difficult to release the ice in discharging cycle. So the optimum value of building the ice should not exceed half of the tube spacing. 12. Conclusions
Fig. 14. Ice formation over a long period.
68% and is released in 12.5 min. When the thickness reaches 22 mm in 60 min, the released ice decreases to approximately 50% in 14 min. This result is due to the suspension of released ice underneath the tubes. If the ice thickness reaches 27.9 mm in 75 min, the released ice decreases to 45% in 21 min. These values indicate that the time for melting ice is shorter than the time for ice formation, which constitutes a decrease from 1/2 to 1/4 depending on the ice thickness. This finding is due to part of the ice melting as the upper part of the sleeves suspend from the tube while the lower part of the sleeves is free from the tubes. 10. The advantages of an ice releasing storage system In a solid ice releasing storage system, many charging and discharging cycles are activated. In the test tube storage system, only one charging and discharging cycle is activated. To compare these two systems, additional experiments are performed to obtain the maximum ice accumulation in the case of storing ice on the tubes of the freezing system. Fig. 14 shows the ice formation over a long period. Note that the accumulated ice reached 3.1 kg and diminishes after 120 min; thus, the power that was consumed has no gain. On the other side, the solid ice released in 120 min, i.e., two runs generate 2.27 kg of ice. If the released ice is used for 3 h (3 runs), 4.08 kg of ice is obtained. For the period far from the peak, such as for air conditioning or dairy farmers, 8 h (6 runs) can be conducted to obtain 8.16 kg of ice compared with 3.1 kg of ice in the other system. 11. Comparison with previous work The slope of ice formation on the tubes is less than the slope discovered by Habeebullah [3] due to the smaller surface area and shorter tube length in this study. Differences exist between the characteristics of the falling film and stagnant water that were employed. The unexpected enlargement of ice thickness on the surface of tube bends is not the case in this study due to the use of concentric tubes. Most of the observations of ice releasing are nearly equivalent to the assumptions and findings of Vargas et al. [10]. Ice freezing and melting is affected by the cylinder pitch, as demonstrated by Hirata and Matsui [8]. In the application of cold thermal storage, it is totally not advisable to fill up the tube spacing with ice to the point that ice surface contact with each other, which
An experimental investigation of optimum ice formation was performed to determine maximum solid ice releasing in fallingfilm jet mode on cold vertical banks of horizontal tubes. The main conclusions are as follows: (a) The ice formation and the solid ice released is controlled by the ice thermal resistance, time and the tube spacing. The maximum gained solid ice, within the scope of this study, is an ice thickness that is approximately equal to half of the tube spacing between the tubes, which is the same thickness of the tube diameter employed. This released solid ice formed in approximately 45 min and released in approximately 12.5 min. (b) The heating solution temperatures have a small effect on the gained ice and releasing time. (c) The heat transfer coefficient is affected by the direct contact area between the ice sleeves and the heated tubes. Acknowledgements This work was funded by the Deanship of Scientific Research (DSR), King Abdulaziz University, Jeddah, under Grant No. (829018-D1433). The authors, therefore, acknowledge with thanks DSR technical and financial support. References [1] Sait HH, Hussain A, Selim AM. Experimental investigation on freezing of water falling film on vertical bank of horizontal cold tubes. J Thermal Sci Eng Appl 2012;4:041006-1-7. [2] Sait HH. Heat transfer analysis and effects of feeding tubes arrangement, falling film behavior and backsplash on ice formation around horizontal tubes bundles. Energy Convers Manage 2013;73:317–28. [3] Habeebullah BA. An experimental study on ice formation around horizontal long tubes. Int J Refrig 2007;30:789–97. [4] Cabeza LF, Mehling H, Hibler S, Ziegler F. Heat transfer enhancement in water when used as PMC in thermal energy storage. Appl Thermal Eng 2002;22: 1141–51. [5] Ismail KAR, de Jesus AB. Parametric study of solidification of PCM around a cylinder for ice-bank application. Int J Refrig 2001;24:809–22. [6] Kayansayan N, Acar MA. Ice formation around a finned-tube heat exchanger for cold thermal energy storage. Int J Thermal Sci 2005;45:405–18. [7] Eames IW, Adref KT. Freezing and melting of water in spherical enclosures of the type used in thermal (ice) storage system. Appl Thermal Eng 2002;22: 733–45. [8] Hirata T, Matsui H. Freezing and thawing heat transfer with water flow around isothermally cooled cylinders in staggered and aligned arrangements. J Heat Trans 1992;114:681–7. [9] Yingxin Z, Zhang Y. Modeling of thermal processes for internal melt ice-on-coil tank including ice–water density difference. Energy Build 2001;33:363–70. [10] Vargas JVC, Bojan A, Dobrovicescu A. The melting of an ice shell on heated horizontal cylinder. Trans ASME 1994;116:702–8. [11] Wu S, Fang G, Chen Z. Discharging characteristics modeling of cool thermal energy storage system with coil pipes using n-tetradecane as a phase change material. Appl Thermal Eng 2012;37:336–43. [12] Masahiko Y, Shoichiro F, Tsuyoshi K. Performance analysis on the liquid–ice thermal storage system for optimum operation. Int J Refrig 2002;25:267–77. [13] Tsuyoshi K, Fukusako S, Yamada M, Itoh K. Experimental on melting of slush ice in a horizontal cylindrical capsule. Int J Heat Mass Trans 1999;42:2981–90. [14] Green DW, Perry Robert H. Perry’s chemical engineers’ handbook. 8th ed. McGraw-Hill; 2008.
Contents lists available at ScienceDirect
Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman
Charging and discharging characteristics of cool thermal energy storage system with horizontal pipes using water as phase change material H.H. Sait a,⇑, A.M. Selim b a b
Mechanical Engineering Department, King Abdulaziz University, Jeddah, Saudi Arabia Mechanical Engineering Department, College of Technology, Jeddah, Saudi Arabia
a r t i c l e
i n f o
Article history: Received 1 July 2013 Accepted 9 October 2013
Keywords: Falling film Ice formation Releasing Melting
a b s t r a c t An experimental investigation of ice formation on cold vertical banks of horizontal tubes subjected to falling-film– jet mode– is conducted. In the charging process, a set of internally cooled vertical banks of horizontal tubes of brine is subjected to a falling film of water. The formed ice is periodically observed, photographed and measured in falling-film jet mode at specific internal coolant (ethylene–glycol solution) flow rates and temperatures. In the discharge process, the same solution is heated and used internally to release ice. Different thicknesses of the released ice are observed and measured. The maximum quantity of released ice is obtained and the optimum ice formation is determined. The results indicate that the ice formation and the solid ice released are controlled by the thermal resistance of the ice, time and pitch between tubes. The maximum gained ice has a thickness that is approximately equal to half of the tube spacing between the tubes utilized, which is formed in approximately 45 min and released in 12.5 min. The variation in heating solution temperature has a slight effect on the gained ice and discharging time. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction There are two methods to store and reuse formed ice. The first method involves storing ice on the tubes of the freezing system. The second method involves releasing ice from the tubes and storing it in an isolated reservoir. In the first method, the freezing system is the storage system. Thus, the ice obtained by the system must be as thick as possible. In the second method, the ice has to be released with the minimum amount of melted ice and stored in an isolated receiver. In this case, the freezing system can be reused several times, which increases its capacity. This latter method is considered in this study. The topic of ice formation has been explored by some researchers [1,2,4–8,10]; the topic of ice melting has also been examined [3,9,11,12]. Sait et al. [1,2] performed an experimental investigation on the freezing of water falling film on a vertical bank of horizontal cold tubes. The brine flows into the concentric tubes in parallel. The authors focused their work on ice formation characteristics and heat transfer for the three main modes of falling film: droplets, jets and sheets. These researchers determined that the formation of ice depends on falling film and coolant flow rates. In addition, the overall heat transfer coefficients are controlled by the thermal ⇑ Corresponding author. Tel.: +966 26914730, mobile: +966 560007382; fax: +966 22564933. E-mail address: [email protected] (H.H. Sait). 0196-8904/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.enconman.2013.10.034
resistance of ice. Habeebullah [3] conducted an experimental study on ice formation around a horizontal long copper tube (L = 12.3 m with tree returned bends of 180° and d = 19.5 mm) immersed in water. This researcher found that there was a slope of the ice thickness, in which the axial distance depended on time but varied with coolant flow rate and Stanton and Biot numbers. The axial growth rate of ice was distinct for low values of the coolant Reynolds number and short freezing times. He also discovered an unexpected enlargement of ice thickness on the surface of the tube bends. Cabeza et al. [4] added stainless steel pieces, copper pieces and graphite matrix impregnated with water as phase change materials (PCM) to improve heat transfer. They founded that addition of stainless steel pieces in the PCM does not increase the heat flux significantly. However, addition of copper pieces and the use of graphite composite enhance heat transfer significantly. Ismail and Jesus [5] performed a parametric study of the solidification of PCM around a cylinder for an ice-bank application. They concluded that a lower initial temperature of the liquid phase seemed to accelerate the solidification. The thermal conductivity of the tube wall material can have a considerable influence on the velocity of the process. Kayansayan and Acar [6] analyzed ice formation around a finned-tube heat exchanger for cold thermal energy storage. They concluded that under identical flow and inlet conditions, the heat exchanger with finned tube stores a maximum of 45% more energy than the bare tube of a turbulent flow regime (Re > 3000).
756
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
Nomenclature A Cp d h k L M _ m Q_ T T U
aria (m2) specific heat (J/kg K) tube diameter (m) heat transfer coefficient (W/m2 K) thermal conductivity (W/m K) latent heat (J/kg) mass (kg) mass flow rate (kg/s) heat transfer rate (W) temperature (°C) average temperature (°C) overall heat transfer coefficient (W/m2 K)
Greek symbols ‘ tube length (m) s time (s)
exim exo h hi ho i io Lmice Mice o oo s smice sw tmi w 0
experimental ice melting overall heat transfer coefficient hot hot in hot out internal inner of outer tube latent heat of melted ice melted ice out outer of outer tube surface sensible heat of melted ice sensible heat of water theoretical melting ice water melting point
Subscripts cu copper
Eames and Adref [7] studied freezing and melting of water in the type of spherical enclosures that are used in thermal (ice) storage systems. They concluded that 90% of the cold could be extracted from the ice storage element within 70% of the time required for complete discharge. Hirata and Matsui [8] studied heat transfer with freezing and thawing for water flow around isothermally cooled cylinders in staggered and aligned arrangements. They demonstrated that the ice filling rate is strongly affected by the Reynolds number, cooling temperature, and cylinder pitch perpendicular to water flow. Melting occurred twice as fast as freezing occurred. Yingxin and Zhang [9] studied thermal processes for internal melting in an ice-on-coil tank, including the variation in ice-water density. They analyzed both charge and discharge processes for the ice-on-coil tank. They concluded that the discharge model is only applicable for processes that begin when the ice cylinder slightly overlaps and unfrozen water is present, which causes the ice to float. Vargas et al. [10] experimentally and analytically studied the fundamentals of melting ice-shell rides on heated horizontal cylinders. They assumed and proved that natural convection is negligible compared with the natural convection of the heated cylinder. The melted water is drained axially. Most of the melting is due to direct contact with the upper half of the hot cylinder. Their results indicate that the melting process consists of two distinct regimes: the first regime occurs when the cylinder is surrounded by ice, which consumes the majority of the melting time, and the second regime occurs when the cylinder cuts through the upper portion of the ice sleeve. The time until the ice falls off the cylinder can be obtained from a specific graph. Wu et al. [11] studied discharging characteristics by modeling cool thermal energy storage systems with coil pipes using n-Tetradecane as a phase change material. The results demonstrate that the higher the flow rate of the heat transfer fluid or the higher the inlet temperature of the heat transfer fluid, the higher the cool release rates, and less time will be required during the discharge process. The diameter of the coil pipes has little influence on the discharge process compared to the other variables previously mentioned. Masahiko et al. [12] studied the performance analysis of the liquid–ice thermal storage system for optimum operation. They concluded that intensive melting results when slush ice is pulled by buoyancy to the top of the hot capsule and when close contact occurs. The local heat transfer is most valuable at the top of the vissle
and decreases along the vissle wall due to the stratified layer and free convection. The average heat transfer coefficient increases as heat flux increases. The melting rate increases monotonically as a function of time, irrespective of heat flux and solution concentration. Tsuyoshi et al. [13] conducted an experiment on the melting of slush ice in a horizontal cylindrical capsule. Their results demonstrated that the COP was almost the same for all three operation modes, whereas the performance of heat release mode for POR decreases with an increase in running time or storage time. The system simulation suggests the potential for obtaining optimum operational conditions, such as the daily running time of thermal storage, among the employed operational modes. The previously described study examined freezing and thawing (melting) of stagnant or moving water flow. The freezing of falling water film on horizontal tubes and its characteristics were also examined. For cold thermal storage where, the accumulated ice is used during the peak electrical load periods, it is necessarily to investigate more about the behavior of the discharge cycle (melting process). The present study investigates both charging and discharging of ice on and from tube surfaces, but focusing more on the melting process since the freezing process was explained well by Sait et al. [1,2]. The quantity of optimally formed ice and ice releasing behavior are also investigated for a falling-film jet mode. 2. Experimental apparatus The experimental apparatus shown in Fig. 1 is designed and fabricated to allow falling film to freeze outside the tubes and to obtain the maximum quantity of released solid ice. A detailed description of the apparatus is provided in Sait et al. [1,2]. The main differences between the two apparatuses are as follows: (a) The coolant in the present study flows through the test concentric tubes in series only to achieve accurate temperature measurements, whereas the tubes were designed in series or in parallel in the previous apparatus. (b) A catching tube is used to get accurate temperature of the outlet falling film or the released iced. (c) A strainer is used under the tubes in this study to separate the released solid ice from the melted ice. (d) The hot solution flow rate and temperatures are recorded during the discharge cycle (melting process), which was not the case in the previous design of [1,2], to determine the heat of release.
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
757
Fig. 1. Schematic of experimental apparatus.
The test section, which consists of a supporting box and test tube arrangement, is comprised of seven concentric vertical horizontal tubes of copper, each containing inner tubes with diameters of 15.8 mm and outer tubes with diameters of 28.5 mm. All tubes are 235 mm long and 1 mm thick. They are fixed vertically underneath the feeding tubes at a pitch of 57 mm center-to-center. The outside of the concentric tube underneath is connected to the inside of the concentric tube on top for coolant series flow. Four fluid loops are used effectively to form ice on the tubes using falling film, as well as to release the ice from the tubes. The following loops are employed: the falling film loop, the coolant loop to freeze falling film, the discharging loop to release the gained ice and the pre-cooling loop to pre-cool water in the upper constant head tank to minimize sensible heat removal. The falling film loop activates when the pre-cooled water flows through the adjustable valve and flow meter to the upper feed tube in the test section, as shown in Fig. 2. The water originates from the upper feed tube and flows around the lower tube, where it is distributed into a uniform layer. It then falls and flows onto the test tube surfaces. Some of the water film freezes outside the cold tubes, and some of the water flows down to the catching tube and through the strainer to the water pool. The remaining water is pumped to the pre-cooled constant head tank through an adjustable valve and filter. The coolant loop, which is used to freeze falling film, activates a charging cycle when the cold solution (40% ethylene glycol by weight in water), at a temperature of approximately 10 °C, flows from the refrigeration machine tank (TD-30) through the feeding valve and flow meter to the test tube section. The coolant feeds the concentric test tubes from the inner tube underneath to the outer tube, to the inner tube of the outer tube located on top, and subsequently to the upper test tubes. The coolant returns to the refrigeration machine tank through a control valve. The discharging loop, which is used to release ice, is activated when the temperature of the worm solution ranges from 15 to 30 °C (40% ethylene glycol by weight in water), which is similar to the process in the coolant loop. Water is pumped from the worm reservoir through the control valve and then flows internally through the flow meter to the test tubes to release ice. It then returns to the worm reservoir through the return valve. Some of ice is melted, whereas other ice is released; both are dropped over
Fig. 2. Schematic of the path of falling film around tubes in the test section (copper tube; L = 235, S = d = 28.5 mm.).
the strainer. The melted ice water falls to the pool, whereas the released ice remains on the strainer. The height of the water in the pool is measured before and after the discharge process to obtain the quantity of melted ice. The released solid pieces of ice, which are collected on the strainer, are allowed to fall into the water pool. The height of the water in the pool is measured again to obtain the volume of water (considering 11% of the ice is floating). Activation of the pre-cooling loop, which is used to pre-cool the water in the constant head pre-cooling tank, allows the coolant to flow from the cooling machine through the pre-cooling feeding valve to the immersed cooling coil in the constant head pre-cooling tank. The water is cooled to a fractional degree (approximately 0.3 °C). It is then allowed to flow through the pre-cooling return valve to the refrigeration machine. 3. Measuring techniques The measuring points are shown in Fig. 3. The falling-film flow rate is measured by a calibrated flow meter with an accuracy of 1%. The splashed water from the tubes is collected and measured in a scaled container. The temperatures of the falling film at the entrance of the feed tubes, at the outlet of the test section (in the catching tube), and at the constant head tank are measured with calibrated T-type copper–constantan thermocouples, which are immersed in the tubes by a specially fabricated well with an accuracy of 0.05 °C. The coolant flow rate is measured by a calibrated flow meter with an accuracy of 1%.The coolant temperatures at the inlet and
758
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
The recirculation water pump is activated and its suction valve is adjusted to re-circulate the remaining water in the pool and to maintain suitable head in the tank. The coolant loop is then activated and the coolant solution is allowed to enter the test section. Ice begins to form around the test tubes. The flow rates of the coolant and falling film are observed and recorded periodically for every run. The pre-cooler, inlet and outlet temperatures for both falling film and coolant, as well as for the test tube surfaces, are recorded periodically. The ice formation is observed and photographed. At the desired time, the charging cycle is stopped and the level of stagnant water in the pool is measured. The discharging cycle is activated to release and obtain the quantity of gained solid ice. Some of ice is melted and drained to the strainer and subsequently drained to the pool. The remaining solid pieces of ice are released (depending on the ice thickness) and dropped on the strainer. The level of stagnant water in the pool is measured again (the difference in water level determines the melted ice volume). The solid pieces of ice are then allowed to fall from the strainer to the water pool and the water level in the pool is again measured. By considering that approximately 11% of the ice volume is floating, the difference in water level determines the solid ice volume. The ratio of the solid ice pieces to the total ice volume is obtained.
Fig. 3. Experimental measuring positions.
outlet of the test tubes are measured with a T-type copper–constantan thermocouple, which is immersed in a well-calibrated thermocouple with an accuracy of 1%. The surface temperatures of the specimen test tube are measured using calibrated 36-gage, copper–constantan thermocouples [1]. The differences in pool height of the test section before and after the discharge cycle are measured by a fixed scale with an accuracy of 1%. This step was performed to obtain the volumes of the melted and solid released ice. 4. Experimental ranges The flow of the falling film is controlled to maintain a steady-jet mode under uniform conditions for the majority of the seven test tubes. Its temperature is reduced to a certain value to minimize the sensible heat removed by the test tubes. The coolant flow rate and temperature are controlled to achieve the required absorbed heat to obtain reasonable ice formation. The hot solution flow rate and temperature are controlled to achieve the required added heat to release solid ice with minimum melted ice. Table 1 shows the variable ranges of flow rates for these fluids.
6. Observation remarks The falling-film jet mode, ice formation and released solid ice behavior are observed, photographed and studied. 6.1. Characteristics of falling-film jet mode Jet-mode flow is steady, uniform, and encompasses all test tubes. All surfaces are wetted and approximately 8% of water splashes out of the down test tubes. By adding a spot of color to the flow, it is observed that the jet streams diffuse as they move down the test tubes; however, they maintain their steadiness and their let mode along their path, as shown in Fig. 4. 6.2. Ice formation characteristics The formation of ice begins at the bottom of the test tubes because the lowest temperature of the coolant exists there and the falling film loses heat as it falls onto the cold tubes. The slope of
5. Experimental procedure The refrigeration machine is operated to reduce the coolant solution temperature to approximately 10 °C. The coolant is then admitted to the constant head pre-cooling tank to reduce its water temperature to approximately 0.3 °C. A warm solution heater is activated to obtain the required temperature. The flow rate of the falling film is adjusted to achieve a uniform and steady jet.
Table 1 Variable ranges of fluid flow rates. Fluid
Flow rate (kg/s)
Temperature (°C)
Falling film (water) Coolant (ethylene–glycol solution) Hot solution (ethylene–glycol solution)
0.0425 ± 0.005 0.162 ± 0.002 0.082 ± 0.002
0.3 ± 0.1 10 ± 1 15–30 ± 0.5
Fig. 4. Sample photograph of jet mode.
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
759
the ice thickness pulls the streams toward the bottom of the tubes. A greater accumulation of falling film exists at the bottom of the tube than at the top of the tube, which allows greater accumulation of ice at the bottom of each tube. The ice is nearly regular on most of the tubes, with the exception of the upper end of the tube due to water falling there. Fig. 5 shows a sample photograph of ice formation on the 7 tubes. Ice accumulates circumferentially on the tube surfaces until the tube spacing is filled with ice. By adding a spot of color to the flow, it is revealed that the slope of contaminated ice on the tube moves the flow streams toward the tube end, which causes ice accumulation at this end. 6.3. Ice releasing characteristics During the peak load, it is necessary to use the formed ice in the cold thermal storage for cooling. The formed ice can be stored either on the tubes or it can be collected in thermal storage. To obtain the desired quantity of ice to manage the peak load, many charging and discharging cycles need to be activated. As the ice
Fig. 7. Sample photograph of ice releasing.
is formed during the charging cycle and reaches its optimum quantity, a discharging cycle must be activated to release the ice from the test tubes and collect it in thermal storage. For the discharging cycle in this experiment, a solution similar to the solution used in the cooling loop—it exhibits a temperature range of 15–30 °C— is used to release the ice from the tubes and to measure its volume for evaluating the remaining ice to be stored. The ice begins to melt from the inner surface of the formed ice toward the outer surface. The melted ice, which is now in the form of water, seeks a way to escape the surrounding ice. A portion of the ice flows down the test tubes from both ends of the test tubes. The remaining formed water flows onto the ice layer; as a result, the water volume expands circumferentially. Due to the effects of the warm water, the weight of the ice and gravity, the top ice layer is milted faster than the bottom ice layer. The melting phenomenon continues and the ice thickness decreases until solid pieces of ice suddenly fall from the test tubes. Fig. 6 displays a schematic diagram for ice formation and releasing. If the ice thickness fills the space between the tubes, the melted ice increases and some ice is suspended between the tubes until more ice is melted, which leads to a greater amount of time required to release all of the ice. Fig. 7 shows a sample photograph of ice releasing. Fig. 5. Sample photograph of ice formation.
7. Heat transfer analysis during the freezing and melting process The heat transfer analysis during the solidification process was well explained by Sait et al. [1]. They demonstrated that the rate of heat transfer decreases as the amount of ice that accumulates on the test tubes increases, which causes the overall heat transfer coefficient to consequently decrease. The formed ice increases as both the coolant flow rate and the falling-film flow rate increase. The absorbed heat required to release ice consists of the following heats: sensible heat of sub-cooled ice, latent heat of melted ice and sensible heat of melted water. Thus, the experimental ice melting Q_ exim can be expressed as
Q_ exim ¼ Q_ smice þ Q_ Lmice þ Q_ sw
ð1Þ
where
Q_ smice ¼ M mice Cpmice ðT s T 0 Þ=smice
ð2Þ
and Fig. 6. Schematic for (a) ice formation and (b) ice releasing.
Q_ Lmice ¼ M mice Lmice =smice
ð3Þ
760
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
Table 2 Fluids properties, [14].
l (N s/m2)
Fluid Water (0.25 °C) Ice (5 °C) Cold solution (10 °C) Hot solution (25 °C)
1750 10 – 8 103
6
3.37 103
q (kg/m3)
k (W/m K)
1000 920 1060
569.6 10 1.88 0.467
1050
0.43
Cp (kJ/kg K) 3
4.2 2.040 3.3488 3.30
and
8.2. Characteristics of ice releasing
Q_ sw ¼ M mice Cpw ðT w T 0 Þ=smice
ð4Þ
This heat is added by the heating solution Q_ hav , which is expressed as
Q_ hav ¼
8.1.2. Effect of time on accumulated mass of ice Fig. 8 portrays the effect of time on ice mass formation. Ice formation distinctly increases with time. It begins increasing at a relatively high rate until it reaches 1.5 kg within 30 min. The rate then decreases until it reaches 2 kg within 45 min. The rate decreases further until it reaches 2.8 kg within 75 min. As the ice thickness increases, its thermal resistance increases and its heat transfer decreases. Because ice has a low thermal conductivity, it acts as an insulator for the heat transfer.
R t¼n _ Q dt t¼0 h time
ð5Þ
where t is the time interval, n is the indicated time, and
_ h Cph ðT hi T ho Þ Q_ h ¼ m
ð6Þ
The overall heat transfer coefficient Uexim of the experimental ice releasing is expressed as follows:
U exim ¼ Q_ exim =Aice ðT h T 0 Þ
ð7Þ
In these experiments commercial ethylene glycol solution (40% by weight) is used as a coolant and for heating also. The concentration is obtained by measuring its specific gravity experimentally, then using its value to get the concentration from the specific gravity graph of the ethylene glycol from Green and Perry [14]. All fluid properties are listed in Table 2.
The discharge cycle is used to determine the quantity of formed ice and the quantity of released ice. The quantity of released solid ice is measured and the percentage of released ice is calculated for various discharge solution temperatures. 8.2.1. Quantity of released ice In the discharge cycle some of the ice near the tube surface is melted. The melting continues until the remaining ice fall down from the tube and catches by the strainer. The quantity of the formed ice and consequently the ice thickness, affects the possible released of the solid ice. When ice thickness is small, less than 3.75 mm, no solid ice can be gained with discharge cycle. As the ice thickness increases to a maximum value of 17 mm, the collected solid ice increase and about 1.67 kg can be collected as shown in Fig. 9. As ice thickness increases more, the collected solid ice stays the same or decrease due to more melting is required to force the accumulated ice to fall down from the tubes. This effect is due to the relationship between the discharge heat flux and melting heat. When the quantity of ice is small, all of the ice is melted. This condition occurs because the discharge heat is greater than or equal to the melting heat of the ice surface that contacts the heated tube (tube surface area). When the ice quantity
8. Results and discussion For thermal storage applications, optimal ice quantities and optimum ice releasing at suitable intervals helps to store the required quantities of ice and accommodate required thermal load schedules. 8.1. Ice freezing characteristics Ice freezing quantity is affected by the falling-film mode (flow rate) and its temperature, as well as by coolant flow rate and coolant temperature. For the freezing process, the convection heat transfer coefficient (on the outer surface of ice) is affected by falling film flow rate as was stated by Sait et al. [1]. However, the variation of the ice thickness results in changing of the outside surface area of the ice which will affect outside convection heat transfer.
Fig. 8. Effect of time on ice thickness.
8.1.1. Effect of falling film on accumulated mass of ice As demonstrated by Sait et al. [1], the falling-film mode affects ice freezing. The outside heat transfer convection coefficient increases as the falling film flow rate increases and changes its modes from droplet, jet and sheet, which causes more ice to accumulate and consequently more ice thickness. However, this increase has a limit as was explained in Sait et al. [1], due to backsplash which causes some falling film liquid to step out of the test tubes. In addition as ice thickness increases, the ice resistance increases which limits the rate of ice formation. The moderate jet mode is steady for most of the tubes and contains a small amount of splashed water; therefore, it is chosen as the model for ice formation and ice releasing in this study.
Fig. 9. Effect of ice thickness on released ice quantity.
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
761
increases, some of the ice forms sleeves that suspend from the tubes (refer to Fig. 6). The solid ice sleeves expand as the quantity of accumulated ice increases. As the ice thickness increases, the sleeves become stuck in their tubes and in the tubes underneath. Thus, the quantity of released ice decreases due to melting from the bottom and top of the tubes. The maximum released quantity is approximately 1.67 kg with ice thickness of 17 mm. This thickness is slightly more than half of the tube spacing between the tubes (14.25 mm). 8.2.2. Released ice percentage Fig. 10 shows the percentage of the released solid ice for various average ice thicknesses. As described previously, the maximum released ice thickness is 17 mm and the maximum percentage of ice is 68%. A discharge solution temperature of 23 °C (room temperature, which is practically applied) is initially used for releasing ice. When the average ice thickness is small—approximately 3.5 mm—all of the ice is melted and no solid pieces of ice remain. When the ice thickness increases to 7.9 mm, approximately 54% of the ice remains. When the ice thickness increases by a maximum of 13.6 mm, approximately 63% of the ice remains. When the thickness increases to 17 mm, the percentage of released ice increases to 68%. When the thickness reaches 22 mm, the percentage of released ice decreases to approximately 50%. As the ice thickness reaches 27.9 mm, the percentage of released ice decreases to 41%. The maximum percentage occurs at approximately 17 mm of ice thickness, as shown in Fig. 10. 8.2.3. Effect of hot solution temperature on ice releasing During the discharging process, part of ice layer is melted and the rest falls down and collected by the strainer. With increasing the temperature of the heating solution, the melting process occurs fast, which makes the solid ice falls down more rapidly. However, as the temperature of the hot fluids increases, the melting process consumes more from the formed ice. That is why no real effect is
Fig. 12. Effect of ice thickness on the overall heat transfer coefficient for solid ice releasing for heating fluid flow rate = 0.162 kg/s.
noticed on the percentage of ice releasing within experimental temperature range of the heated solution which varies from 15, 23 and 30 °C as shown in Fig. 11. 8.3. Heat transfer coefficient for solid ice releasing As was explained earlier by Eqs. (1)–(7), the experimental overall heat transfer coefficient depends on heat transfer from the heated solutions to the solid ice which comprises of three quantities of heat transfer namely sensible heat of sub-cooled ice, latent heat of melted ice and sensible heat of melted water. As shown in Eq. (7), ice surface area and the temperature difference between the heating fluids and melting point affects the experimental overall heat transfer coefficient Uexim. Fig. 12 shows the variation of the experimental overall heat transfer coefficient Uexim for solid ice releasing with ice thickness. As the ice thickness decreases as melting is taken place, Uexim is increasing to reach its maximum value of 350 W/m2 K when the ice thickness decreases to 4 mm. 9. Comparison of time of ice formation with time of ice releasing
Fig. 10. Effect of ice thickness on percentage of released ice.
Fig. 11. Effect of heating solution temperature on released ice percentage.
Fig. 13 includes a comparison of the time of ice formation with the time of ice releasing. The figure indicates that when the average ice thickness is small—a maximum of 3.5 mm, which is formed in 5 min—all of the ice is melted in 2 min. When the thickness increases—a maximum of 7.9 mm, which is formed in 15 min— approximately 54% of the ice is released in 6 min. When the thickness increases further—a maximum of 13.6 mm, which is formed in 30 min—approximately 63% of the ice is released in 10 min. When the thickness increases to 17 mm, which is formed in 45 min, the percentage of released ice increases to
Fig. 13. Comparison between time of ice formation and ice releasing at certain ice thickness.
762
H.H. Sait, A.M. Selim / Energy Conversion and Management 77 (2014) 755–762
will make it very difficult to release the ice in discharging cycle. So the optimum value of building the ice should not exceed half of the tube spacing. 12. Conclusions
Fig. 14. Ice formation over a long period.
68% and is released in 12.5 min. When the thickness reaches 22 mm in 60 min, the released ice decreases to approximately 50% in 14 min. This result is due to the suspension of released ice underneath the tubes. If the ice thickness reaches 27.9 mm in 75 min, the released ice decreases to 45% in 21 min. These values indicate that the time for melting ice is shorter than the time for ice formation, which constitutes a decrease from 1/2 to 1/4 depending on the ice thickness. This finding is due to part of the ice melting as the upper part of the sleeves suspend from the tube while the lower part of the sleeves is free from the tubes. 10. The advantages of an ice releasing storage system In a solid ice releasing storage system, many charging and discharging cycles are activated. In the test tube storage system, only one charging and discharging cycle is activated. To compare these two systems, additional experiments are performed to obtain the maximum ice accumulation in the case of storing ice on the tubes of the freezing system. Fig. 14 shows the ice formation over a long period. Note that the accumulated ice reached 3.1 kg and diminishes after 120 min; thus, the power that was consumed has no gain. On the other side, the solid ice released in 120 min, i.e., two runs generate 2.27 kg of ice. If the released ice is used for 3 h (3 runs), 4.08 kg of ice is obtained. For the period far from the peak, such as for air conditioning or dairy farmers, 8 h (6 runs) can be conducted to obtain 8.16 kg of ice compared with 3.1 kg of ice in the other system. 11. Comparison with previous work The slope of ice formation on the tubes is less than the slope discovered by Habeebullah [3] due to the smaller surface area and shorter tube length in this study. Differences exist between the characteristics of the falling film and stagnant water that were employed. The unexpected enlargement of ice thickness on the surface of tube bends is not the case in this study due to the use of concentric tubes. Most of the observations of ice releasing are nearly equivalent to the assumptions and findings of Vargas et al. [10]. Ice freezing and melting is affected by the cylinder pitch, as demonstrated by Hirata and Matsui [8]. In the application of cold thermal storage, it is totally not advisable to fill up the tube spacing with ice to the point that ice surface contact with each other, which
An experimental investigation of optimum ice formation was performed to determine maximum solid ice releasing in fallingfilm jet mode on cold vertical banks of horizontal tubes. The main conclusions are as follows: (a) The ice formation and the solid ice released is controlled by the ice thermal resistance, time and the tube spacing. The maximum gained solid ice, within the scope of this study, is an ice thickness that is approximately equal to half of the tube spacing between the tubes, which is the same thickness of the tube diameter employed. This released solid ice formed in approximately 45 min and released in approximately 12.5 min. (b) The heating solution temperatures have a small effect on the gained ice and releasing time. (c) The heat transfer coefficient is affected by the direct contact area between the ice sleeves and the heated tubes. Acknowledgements This work was funded by the Deanship of Scientific Research (DSR), King Abdulaziz University, Jeddah, under Grant No. (829018-D1433). The authors, therefore, acknowledge with thanks DSR technical and financial support. References [1] Sait HH, Hussain A, Selim AM. Experimental investigation on freezing of water falling film on vertical bank of horizontal cold tubes. J Thermal Sci Eng Appl 2012;4:041006-1-7. [2] Sait HH. Heat transfer analysis and effects of feeding tubes arrangement, falling film behavior and backsplash on ice formation around horizontal tubes bundles. Energy Convers Manage 2013;73:317–28. [3] Habeebullah BA. An experimental study on ice formation around horizontal long tubes. Int J Refrig 2007;30:789–97. [4] Cabeza LF, Mehling H, Hibler S, Ziegler F. Heat transfer enhancement in water when used as PMC in thermal energy storage. Appl Thermal Eng 2002;22: 1141–51. [5] Ismail KAR, de Jesus AB. Parametric study of solidification of PCM around a cylinder for ice-bank application. Int J Refrig 2001;24:809–22. [6] Kayansayan N, Acar MA. Ice formation around a finned-tube heat exchanger for cold thermal energy storage. Int J Thermal Sci 2005;45:405–18. [7] Eames IW, Adref KT. Freezing and melting of water in spherical enclosures of the type used in thermal (ice) storage system. Appl Thermal Eng 2002;22: 733–45. [8] Hirata T, Matsui H. Freezing and thawing heat transfer with water flow around isothermally cooled cylinders in staggered and aligned arrangements. J Heat Trans 1992;114:681–7. [9] Yingxin Z, Zhang Y. Modeling of thermal processes for internal melt ice-on-coil tank including ice–water density difference. Energy Build 2001;33:363–70. [10] Vargas JVC, Bojan A, Dobrovicescu A. The melting of an ice shell on heated horizontal cylinder. Trans ASME 1994;116:702–8. [11] Wu S, Fang G, Chen Z. Discharging characteristics modeling of cool thermal energy storage system with coil pipes using n-tetradecane as a phase change material. Appl Thermal Eng 2012;37:336–43. [12] Masahiko Y, Shoichiro F, Tsuyoshi K. Performance analysis on the liquid–ice thermal storage system for optimum operation. Int J Refrig 2002;25:267–77. [13] Tsuyoshi K, Fukusako S, Yamada M, Itoh K. Experimental on melting of slush ice in a horizontal cylindrical capsule. Int J Heat Mass Trans 1999;42:2981–90. [14] Green DW, Perry Robert H. Perry’s chemical engineers’ handbook. 8th ed. McGraw-Hill; 2008.