- Email: [email protected]
Thermosiphon loop thermal collector for low-temperature waste heat recovery
Accepted Manuscript Thermosiphon Loop Thermal Collector for Low-Temperature Waste Heat Recovery Koji Matsubara, Professor, Yusaku Matsudaira, Itaru Kourakata PII:
S1359-4311(15)00897-2
DOI:
10.1016/j.applthermaleng.2015.09.004
Reference:
ATE 6986
To appear in:
Applied Thermal Engineering
Received Date: 11 June 2015 Revised Date:
1 September 2015
Accepted Date: 2 September 2015
Please cite this article as: K. Matsubara, Y. Matsudaira, I. Kourakata, Thermosiphon Loop Thermal Collector for Low-Temperature Waste Heat Recovery, Applied Thermal Engineering (2015), doi: 10.1016/j.applthermaleng.2015.09.004. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
Thermosiphon Loop Thermal Collector for Low-Temperature Waste Heat Recovery
Koji Matsubara, Yusaku Matsudaira and Itaru Kourakata
Department of Mechanical and Production Engineering, Niigata University
Corresponding author: Koji Matsubara, Professor Department of Mechanical and Production Engineering, Niigata University Ikarashi 2-nocho 8050, Niigata 950-2181, Japan E-mail [email protected] TEL & FAX (+)81-(0)25-262-7260
2
ACCEPTED MANUSCRIPT Abstract
This paper describes the thermal collector type loop thermosiphon for low-temperature A1 waste heat recovery. Water is used as the working fluid for heat transport at temperatures 100
and higher. The loop thermosiphon comprises of a thermal receiver, a condenser, and
riser and downcomer tubes. The thermal receiver is made of a cupper plate brazed by meandering heat transfer tube. This receiver collects the thermal radiation from the electric
RI PT
heater at the heat transfer area of 1,000 cm2 (40cm x 25 cm), and transports the heat by vaporization of water to the condenser having heat transfer area of 62 cm2. In the no inclination mode, the thermosiphon is upright so that the thermal receiver and the condenser are placed vertically. In this mode, the effective thermal conductivity exceeds 60 kW/(m K) when the thermal receiver temperature was higher than 125
for the water filling ratio
ratio
SC
=30% to 70%. Although the effective thermal conductivity is deteriorated for the higher filling =80% and 90%, the effects from the filling ratio is tiny for
=30% to 70%. The
experimental tests were also made for the negatively inclined mode where the inlet and exit
M AN U
ports of the receiver were directed downward and for the positively inclined modes where they were directed upward. The tests revealed that the negative inclination almost halted the heat transport. However, the tests also indicated that the positive inclination showed the performance comparable to the no inclination mode for the cases up to the inclination angle 90°.
TE D
Keywords
Loop thermosiphon, Low-temperature waste heat, Heat transfer experiment, Effective A2 thermal conductivity, Inclination angle 1. Introduction
EP
There is an increasing concern about reducing fossil fuel consumption. This concern arises from the diminishing energy resources and increasing carbon dioxide levels in the atmosphere.
AC C
Therefore, it is becoming more important to utilize low-temperature waste heat, which is now mostly unused. Figure 1 shows the statistics from Japan in 2011 [1]. The figure reveals that a large amount of waste heat was exhausted at temperatures below 200°C. The annual waste heat between 100°C and 200°C is 282 peta (1015) calories, which is 9.1% of the crude oil consumption there. The recovery of low-temperature waste heat is important for slashing fossil fuel consumption. Heat transfer devices are expected to be developed for waste heat recovery. A device with no external means of power is desirable because of the low density of the waste heat. Loop-style thermosiphons have been emerging as transportation devices. These have been applied for the removal of high-density thermal loads. Their apparent thermal conductivity exceeds 200 times that of copper [2]. Academic papers were published on a thermosiphon loop for electronics cooling [4-10]. Previous papers used isobutene, pentane, and nitrogen as
3
ACCEPTED MANUSCRIPT
thermal media, setting temperatures at about 50°C [4-10]. More recently, Khodabandeh experimentally examined a radio base station cooling loop that was filled with isobutene [11]. Dube et al. performed experiments on the effects of a noncondensable gas on loop thermosiphon heat exchangers with water as the working fluid [12]. Khodabandeh and Furberg [13] visualized the bubbly flow of R134a by using an evaporator with a polycarbonate window [13]. Sarno et al. [14] tested the passive cooling system of an R141 loop for in-flight n-pentane
thermosiphon
loop
for
cooling
outdoor
RI PT
entertainment avionics. Samba et al. [15] performed transient and steady-state analyses of an telecommunication
equipment.
Zimmermann and Melo experimented on a carbon dioxide thermosiphon loop for a Stirling A4 cooler [16]. Li et al. [17] performed a visualization experiment of an insert-type closed loop water thermosiphon for solar water heaters. Zhang et al. [18] proposed the integration of a
SC
mechanical refrigeration system and thermosiphon for the free cooling of data centers. Xie et al. [19] investigated a dual compensation chamber loop heat pipe subjected to an acceleration field for high-power and high-heat-flux electronic devices in aircraft and spacecraft.
M AN U
Increasing numbers of academic papers have recently been written on the thermosiphon loop. Most of these papers intended to contribute to electronics cooling and they treated refrigerants as the working fluid [4-11, 13-16, 18, and 19]. These papers set the working temperatures at levels from 0°C to 50°C, which are not high enough for waste heat transport. Few papers [12, 17] used water as the working fluid, setting a temperature as high as 100°C. A4 However, there are only limited cases of experiments for working temperatures higher than
TE D
100°C. There are still many unknown points concerning how a thermosiphon loop works and what phenomena occur under high-temperature conditions. The group at Niigata University [2] developed a water thermosiphon loop for waste heat recovery. The evaporator, consisting of three layers of copper blocks, was designed to resist the high saturation pressure of water. This thermosiphon loop of water works for
EP
high-temperature conditions in which the condenser wall temperature was increased beyond 150°C. This is because critical temperature of water is 374
and phase change occurs for the A3
AC C
temperatures under the critical temperature. This paper extends the previous study [2] to consider the thermal-collector-type thermosiphon loop with water. The heat transfer area of the evaporator was enlarged by using a meandering pipe attached to a copper plate. Series of experiments were conducted for this thermal collector loop to reveal circulation flow stability and heat transfer characteristics. 2. Experimental method Figure 2 shows a photograph and line drawing of the thermal collector thermosiphon loop. This apparatus comprises a thermal receiver plate, condenser, riser, and downcomer. The thermal receiver plate is made of copper. The riser and downcomer are bent stainless steel tubes. The thermal receiver is heated from the bottom surface by an infrared lamp heater. The condenser is cooled by a stream of coolant water. The working fluid is designed to
4
ACCEPTED MANUSCRIPT
circulate in the thermosiphon loop by natural convection. This can be confirmed by the temperature distribution, as will be discussed later. The heat transfer area of the receiver plate is 1000 cm2 (40 cm × 25 cm). That of the condenser wall is 62 cm2 (π × 0.8 cm × 25 cm). The low-density heat flux of the waste heat is thus intended to be increased at the condenser by shrinking the heat transfer area. The ratio of receiver plate to condenser wall area is A6
1000:62 = 16.2:1.
RI PT
As shown in Fig. 3, the thermal receiver plate is soldered to a meandering copper tube. This tube is made by pressing a round tube (φ = 8.0 mm, thickness = 5 mm) until it has an oval cross section with a height of 5 mm and width of 10 mm. The tube was pressed so that it A5 has a larger surface to contact the plate in order to reduce thermal resistance. The condenser is an annulus with a copper center pipe and aluminum alloy shell, as indicated in Fig. 3. The
SC
preliminary experiment revealed that the coolant water was boiled in the copper pipe and exited the condenser as a bubbly stream. This was not suitable for the determination of recovered heat because the difficulty of obtaining quality measurements. The copper
M AN U
condenser pipe was thus covered by a stainless steel jacket with a thickness of 6.0 mm to increase the thermal resistance. This could prevent the coolant water from boiling and allow the recovered heat to be calculated on the basis of the temperature difference between the inlet and exit of the condenser cooling side. The thermosiphon loop is insulated by foam A9 insulation on the riser and downcomer, and by sheet insulation on the thermal receiver and condenser. The temperature distribution is measured using thermocouples set on the thermal
TE D
receiver and tube joints, as illustrated in Fig. 2. The recovered heat was obtained from the temperature difference and flow rate of the coolant water. Table 1 summarizes the experimental conditions. The water filling ratio is defined as the ratio of the mass of filled water to the mass that can fill the entire volume of the loop Vmax = 105 mL, α = m / mmax . Measurements were performed for the case where the thermosiphon
EP
loop was mounted at an inclination angle of 0 degrees and the case where the loop was inclined from −45 to 90 degrees. The definition of the mounting angle is as shown in Fig. 4.
AC C
The thermosiphon loop with no inclination was tested with a water filling ratio that changed from 0.3 to 0.9, whereas the inclined loop was tested with the water filling ratio fixed at 0.3. The thermal load was regulated by changing the input voltage to the infrared lamp heater from 50 V to 100 V. Heat radiation occurred from the infrared lamp to the thermal receiver. Since the absorption rate of radiation at the thermal receiver plate is not very clear in the experiment, its temperature was used to represent the experimental condition. This temperature was set at 80°C to 200°C for the non-inclined thermosiphon loop and at 70°C to 170°C for the inclined loop. Temperature is measured by thermocouples at seven positions A12 indicated in Fig. 2. Pressure is measured at positions 4 and 5 indicated in Fig. 2. The pressure A10 sensor is a strain gauge type pressure transmitter, KM31, produced by Nagano Keiki. 3. Results and discussion
5
ACCEPTED MANUSCRIPT 3.1 Temperature distribution and pressure in the thermosiphon loop
Figure 5 shows the temperature distribution for the thermosiphon loop. This figure divides the experimental data into three groups according to thermal receiver plate temperature: (a)
T1 = 80°C–110°C, (b) T1 = 130°C–160°C, and (c) T1 = 180°C–220°C. The horizontal axis A12 indicates the positions of thermocouples as shown in Fig. 2. Thermocouple no. 1 is on the thermal receiver plate, nos. 2 and 3 are on the evaporator tube, no. 4 is in the bottom of the
RI PT
riser, and nos. 5 and 6 are in the inlet and exit of the condenser, respectively. No. 7 is in the bottom of the downcomer. If water circulates by climbing up the riser and descending the downcomer,
temperature
may decrease
according
to
thermocouple
number order.
Temperatures at 2 and 3, on the evaporator tube, may reverse even when the flow circulates stably since a temperature rise can occur between 2 and 3 because of heat absorption.
SC
In Fig. 5, the temperature generally decreases along with the thermocouple number. Thus, it is indicated that the fluid circulates in the thermosiphon loop by going up the riser and coming down the downcomer. The temperatures are reversed between 2 and 3 for conditions
M AN U
where the water filling ratio is high. This occurs because of irradiation and does not contradict the stable circulation of the working fluid as mentioned above. However, A11 temperature increases in the condenser (5 to 6) and downcomer (6 to 7) for some conditions of each group where the water filling ratio is relatively low at 30 % to 60 %. These cases suggested that the fluid flow did not circulate as designed.
The pressure was measured at the lower and upper ends of the riser, corresponding to
TE D
positions 4 and 5 respectively, as shown in Fig. 2. According to the measurements, pressure at the lower end is higher than at the upper end with an average difference of 2.3 kPa. It is thus suggested that the ascent of the working fluid is aided by pressure difference in the riser. A11 Since a pressure difference in the of order 2 kPa is quite small compared with the pressure change due to the heating conditions, the pressure distribution is considered negligible in the
EP
examination of how the heating conditions result in the phase changes. Figure 6 shows the A11 relationship between pressure and temperature at Position 5 together with the saturation
AC C
pressure and temperature. This figure also includes the temperatures at the upper and lower ends of the downcomer, indicated as Positions 6 and 7, respectively in Fig. 2. Since there is no experimental data for the pressures at 6 and 7, the temperatures at these positions were plotted against the pressure at 5 in Fig. 6. Therefore, the pressure indicated by T6 and T7 is not exactly the same as the pressure at Positions 6 and 7. However, for the reason mentioned above, the pressure distribution is be considered negligible compared with changes due to the heating conditions. The relationships between (T6, P5) and (T7, P5) approximate the relationships between (T6, P6) and (T7, P7) in discussing phase changes. Figure 6 (a) shows the results for a filling ratio of 30%. In this figure, the temperature at Position 5 is almost coincident to the saturation temperature at the same pressure. Therefore, the working fluid is indicated to be saturated at Position 5, along with the whole temperature region for the test. It is different from the coincidence between T5 and Tsat that the A11
6
ACCEPTED MANUSCRIPT
temperatures at Positions 6 and 7 are clearly lower than the saturation temperature at the same pressure. The working fluid is thus in a state of compressed water at 6 and 7. In Fig. 6 (b), the temperatures at 5, 6, and 7 are shown with the pressure at 7 for a filling ratio of 60%. The characteristics of the temperature and pressure relationship are similar to those for the lower filling ratio of 30% when the temperature is above 130°C. However, for the lower temperature case, the temperatures at 6 and 7 correspond to the saturation
RI PT
temperature rather than temperature at 5. In this case, the working fluid is assumed to be saturated in the condenser and in a state of compressed water in the upper riser. This A11 corresponds to the inversed temperature at condenser as discussed above. Figure 8 (c) shows the results for a filling ratio of 90%. In the case where the temperature is lower than 100°C for this condition, the temperatures at 5, 6, and 7 are lower than the saturation temperature. In
SC
these cases, the working fluid is assumed to be compressed water, and the thermal load is not high enough to saturate the water.
The presentation of temperatures using the measured pressure is summarized as follows.
M AN U
Normal circulation is assumed at all temperatures for a 30% filling ratio and at relatively high temperatures for 60 and 90% filling ratios. In these cases, the working fluid is assumed to circulate normally such that the fluid is vaporized in the riser and condensed water in the downcomer. However, circulation is assumed to be abnormal at low temperatures for the 60% filling ratio. The shortage of thermal load for such case is likely to cause the unstable A11 circulation flow. It is not clear how the working fluid moves in the loop for these abnormal Sentences
TE D
cases. The temperature distribution suggests that the downcomer introduces vaporized water restructured into the condenser in these unstable conditions. However, it is still ambiguous how the condensed water returns to the thermal receiver plate under these conditions. There is a possibility that the downcomer works as a heat pipe to raise and lower the water through one tube. There is also a possibility that the riser works to return the condensed water to the
EP
receiver plate. These are not in line with the water circulation as intended. The unstable circulation occurs for the low temperature conditions where thermal load is insufficient. It is
AC C
recommended that the thermosiphon loop is utilized for stable conditions with a sufficient thermal load.
The experimental tests discussed in this paper had no flow oscillation. However, other experiments showed temperature oscillations that were remarkable at the inlet and exit of the condenser (Positions 5 and 6). Figure 7 illustrates the temperature oscillation at these positions for a 60% water filling ratio and thermal receiver temperature T1 = 94.3°C. These temperatures were observed to oscillate in the order of tens of seconds. In the figure, the temperature at the condenser inlet (Position 5) tends to be higher than that at the exit
A11 (Position 6) but is reversed between 120 and 150 s. Such oscillations probabilistically occur for A11 iterative tests in the same conditions. Figure 8 summarizes the flow pattern based on three sets of experimental tests. The normal circulation and oscillating flow are distinguished by a temperature oscillation at the condenser inlet and exit (Positions 5 and 6). This temperature A11 A11
7
ACCEPTED MANUSCRIPT
oscillation occurs for cases where the thermal receiver temperature is lower than 130°C. The temperature oscillation tend to occur for low-temperatures cases with insufficient thermal load. It is in contrast to the lack of experimental reproducibility at low temperatures that A11 normal steady circulation occurs at high temperatures. The figure indicates that the thermosiphon loop works normally for the case with sufficiently high thermal load. The experimental data for the flow oscillation are included only in Figs. 7 and 8. Other figures
3.2 Thermal performance of the thermosiphon loop
RI PT
exclude the flow oscillation data.
Figure 9 shows the recovered heat at the condenser versus the thermal receiver temperature. This figure includes the results for water filling ratios from 30% to 90%. The
SC
recovered heat ranges from 100 to 600 W, which increases with the increase in thermal receiver temperature. The recovered heat for the 90% filling ratio lies below that for the lower filling ratio. The recovered heat for the 80% filling ratio decreases at low temperatures from A11
M AN U
the cases of the lower filling ratio. This indicates that the high value of the filling ratio tends to diminish the heat transported by the thermosiphon loop. It is suggested that the water flow A7, A8 rate decreases as the water filling rate increases since the heat transfer at the condenser essentially corresponds to the flow rate of latent heat. There are relatively high values of recovered heat for filling ratios from 30% to 70%. The recovered heat in these conditions essentially does not depend on the filling ratio except for the case where the thermal load is
TE D
attenuated at thermal receiver temperatures below 100°C. The circulation of water is A11 suggested to deteriorate for low temperatures but to increase along with the increase of the thermal load based upon the correspondence between the water circulation and the recovered heat.
This paper defines thermal resistance using the thermal receiver temperature TR and TR − TC . Q
(1)
AC C
Rt =
EP
the temperature on the outer wall of the condenser copper pipe TC as
This is plotted against thermal receiver temperatures shown in Fig. 10. The higher filling ratios are inferior to lower filling ratios in terms of thermal resistance. The thermal resistance for α = 90% and α = 80% is clearly higher than it is for lower filling ratios, for the A11 cases of thermal receiver temperatures below 150°C. The thermal resistance is as low as 0.14 K/W for filling ratios of 30% ≦ α ≦ 70%, which shows the superiority of lower filling ratios. The thermal resistance tends to surge for temperatures as low as 80°C when α = 40% and α = 60%. For α = 30%, a low thermal resistance is maintained for the entire range of thermal receiver temperatures. The superiority of low filling ratios is thus suggested by the stably suppressed thermal resistance for decreasing thermal receiver temperatures to 75°C. The effective thermal conductivity is defined using the cross-sectional area A of the riser and downcomer and the tube length L as
8
ACCEPTED MANUSCRIPT λ=
QL
(TR − TC )A
.
(2)
The cross-sectional area is determined by the outer diameter of the riser and downcomer tubes as 2A ( = 2πd 2/4) = 1.00 cm2, where d = 0.8 cm. The riser length of 97 cm is used as the tube length in equation (2). The effective thermal conductivity is as high as 60 kW/(m K) for cases of low water filling ratio. This tends to decrease when the thermal receiver temperature
RI PT
is lower than 125°C, or the filling ratio is higher than 70%. However, the effective thermal conductivity remains stable at high values for α = 30%, 40%, and 50%. This surpasses 80 kW/(m
K) maximally, which is 200 times the thermal conductivity of copper. The
thermosiphon loop for high temperatures is found to be effective for transporting heat at temperatures beyond 100°C.
SC
The experimental data of the recovered heat, the thermal resistance and the effective A11 thermal conductivity depended on the water filling rate and the thermal load. All of the three quantities are deteriorated for the low thermal receiver temperatures below 125
and the
M AN U
higher water filling rate of α = 80% and 90%. This deterioration occurs due to the stagnation of water circulation which was suggested by decreasing recovered heat. In contrast, all the three quantities showed preferable performance for the higher thermal receiver temperatures than 125
and for the lower water filling rate of α = 30%, 40%, and 50%. The stable water
circulation, suggested by the recovered heat, is attributed to the preferably maintained
TE D
characteristics of the thermosiphon loop. 3.3 Effects of mounting angle
The thermosiphon uses gravity to return condensed water to the evaporator. The mounting angle severely affects the thermal performance of the apparatus. This study tested the thermosiphon loop by changing the mounting angle from θ = −45 to 90 degrees. For the
EP
negative angle mode, the apparatus was inclined so that the evaporator inlet and exit direct downward. For the positive angle mode, the apparatus was inversely inclined. In the
AC C
experiments for the negative angle mode, there was almost no temperature increase in the coolant water through the condenser even when the thermal receiver plate was irradiated at maximum heat load. There was almost no recovered heat at the condenser when the apparatus was negatively inclined. An inclination of only −5 degrees almost eliminated any heat recovery, which suggests that negative inclination severely deteriorates water circulation. Figure 12 shows the recovered heat for inclination angles from α = 0 to 90 degrees. The water filling rate was set at 30% in this section. Unlike the significant deterioration in negative angle mode, a positive inclination moderately affects the recovered heat. In fact, recovered heat existed for all of the selected positive angles. However, this decreased as the mounting angle increased. The decrease in recovered heat was as much as 100 W in the case of the maximum mounting angle α = 90°. The positive inclination maintains the recovered A11
9
ACCEPTED MANUSCRIPT
heat up to α = 90°, to suggest that the water circulation is stable for the cases of positive inclination. It is advantageous that the weak dependency on positive inclination assures freedom in the setting of the apparatus. The thermal resistance of the thermosiphon loop is shown in Fig. 13 for α = 0 to 90 degrees. There are no remarkable changes in thermal resistance for temperatures higher than 100°C. This proves the usefulness of the thermosiphon in the positive angle mode together with the
RI PT
moderate decrease in recovered heat mentioned above. The thermal resistance surges for temperatures lower than 100°C. In the low-temperature region, thermal resistance was kept low for positively inclined angles α = 30 and 45 degrees. These cases have improved thermal resistance versus no inclination (α = 0 degrees). However, the thermal resistance is highest at α = 90 degrees. Therefore, a slight positive inclination may enhance the thermal performance
SC
at low temperatures, but a large inclination angle tends to deteriorate it. At the higher A11 temperatures than 100 , the thermal resistance for α = 90 degrees is slightly higher than other cases. This slight deterioration comes to the stagnation of water circulation suggested
M AN U
by the recovered heat.
The effective thermal conductivity is presented in Fig. 14 for the same cases as in Figs. 12 and 13. This quantity is kept high when the mounting angle lies between 0 and 45 degrees. Increasing the mounting angle to more than 45 degrees leads to deterioration of the effective thermal conductivity. In fact, from α = 0 to 90 degrees, the effective thermal conductivity is decreased by 30% when the thermal receiver temperature is higher than 120°C. However, the
TE D
performance of the thermosiphon is preferably maintained at a positive inclination. It is A11 attributed to sustained water circulation mentioned above. Positive inclination is shown to be acceptable for usage of the current style of thermosiphon loop. There was a clear contrast between negative inclination and positive inclination in thermosiphon performance. In the positive inclination mode, the thermal receiver plate inlet and exit were inclined upward. This
EP
configuration is considered to assist water circulation due to gravitational force. In the positive inclination mode, vaporized water ascends to exit the thermal receiver plate by
AC C
buoyancy. In this mode, condensed water descends back to the plate because of its weight. Conversely, it is considered that a negative incline leads to the collapse of water circulation by preventing the exit of evaporated water and the return of condensed water. The experimental results indicate that the mounting angle affects water circulation severely, but the optimum mounting angle to maintain heat transfer performance can be selected on the basis of water A11 circulation. 3.4 Error analysis Reproducibility of the experimental data was confirmed by iterative measurements. Figure 15 shows the combined recovered heat. This figure includes error bars showing twice the standard deviation from the five sets of measurements. These experimental data were acquired using the thermosiphon with no incline with a water filling ratio of α = 60%. The thermal resistance and the effective thermal conductivity are shown in the same way in Figs A11
10
ACCEPTED MANUSCRIPT
16 and 17, respectively. Whereas the unstable flow pattern may have harmed the experimental reproducibility for the low temperature cases, the measured quantities were reproducible at least for the high temperature cases. The recovered heat has a standard deviation of 12.8 W maximally, whereas the average value is 312 W. This standard deviation corresponds to 4% of the average value. The thermal resistance has a relatively high standard deviation for temperatures lower than 100°C. However, this quantity keeps the standard
RI PT
deviation low at higher temperatures. For thermal receiver temperatures above 100°C, the standard deviation is 0.008 K/W maximally, which corresponds to 6% of the average (0.140 K/W). The effective thermal conductivity resembles the thermal resistance in measurement A11 uncertainty. This quantity shows a large standard deviation for thermal receiver temperatures lower than 100°C and has lower standard deviation for higher temperatures.
SC
The effective thermal conductivity maximum standard deviation increases to 3.9 kW/(m K) for temperatures higher than 100°C. This is equivalent to 5.6 % of the average value (69.3 kW/(m K)). In all the three quantities the experimental uncertainty is modest for the higher
M AN U
temperatures than 100 , indicating the stable circulation of water. This low uncertainty vindicates the presented data and the analysis in the earlier sections. 4. Conclusions
This paper describes the heat transfer characteristics of a thermosiphon loop thermal collector. The working fluid is water for waste heat recovery at temperatures 100°C and
TE D
higher. The temperature distribution and inside pressure were measured to reveal the stability and performance of the thermosiphon loop. The conclusions thus acquired is summarized as follows:
(1) The temperature generally decreased along the thermal receiver plate, the lower end of the riser, inlet and exit of the condenser, and lower end of the downcomer. This indicated a
EP
normal circulation of working fluid: water vaporized in the thermal receiver plate, ascended in the riser, liquefied in the condenser, and descended in the downcomer.
AC C
However, the temperature was reversed between the condenser inlet and exit or between the upper and lower ends of the downcomer under some of the low-temperature conditions. This reversed distribution of temperatures occurred conspicuously when the thermal receiver temperature was lower than 140°C. This suggested that the circulation flow tends to be unstable for the cases where the thermal load was inadequately provided. (2) Temperature oscillation may occur at the inlet and exit of the condenser. The experimental tests were repeated three times to determine what conditions lead to flow oscillation. The experimental tests revealed that temperature oscillation is likely to occur for thermal receiver plate temperatures that are lower than 140°C. This oscillation probably occurs at low temperatures. However, the temperature is deterministically stable at the higher temperatures. The temperature oscillation is interpreted to occur for cases with inadequate thermal load.
11
ACCEPTED MANUSCRIPT
(3) The heat transport performance of the thermosiphon was evaluated by the recovered heat, with the thermal resistance and effective thermal conductivity changing the water filling ratio α = 30% to 90 %. These quantities were affected by the water filling ratio to indicate a deterioration of performance for α = 80% and 90%. However, they were not largely affected by the water filling ratios between α = 30% and 70%. The effective thermal conductivity exceeded 60 kW/(m K) when the thermal receiver temperature was higher
RI PT
than 125°C for α = 30% to 70%. This effective thermal conductivity is 100 times the molecular conductivity of copper.
(4) The effects of the mounting angle were examined by inclining the thermosiphon loop. For positive inclinations, the thermosiphon was tilted so that the evaporator tube inlet and outlet were directed upward. For negative inclinations, the thermosiphon was tilted
SC
inversely. The experiments in the negative inclination mode revealed that the coolant water showed almost no temperature increase, and there was almost no heat transport from the thermal receiver plate to the condenser. On the other hand, tests in the positive
M AN U
inclination mode indicated that the thermal performance of the thermosiphon loop is maintained for inclination angles of 75 degrees or less. Thus, it is possible that the thermosiphon is mounted in the positive inclination mode for practical usage. (5) The experimental uncertainty was examined by five iterative tests for a water filling ratio of 60%. There is a relatively large standard deviation for the low temperature cases where the thermal receiver temperature is set as high as 100°C. However, the standard
TE D
deviation is not remarkable for the higher-temperature cases. The ratio of the standard deviation to the average value is in the order of five percent for the recovered heat, thermal resistance, and effective thermal conductivity when the receiver temperature is set beyond 100°C. The stable circulation confirmed by the temperature distribution may reduce the scattering of the experimental data. The stability of the thermosiphon loop for
AC C
distribution.
EP
higher temperatures is observed by experimental reproducibility as well as temperature
References
[1] K. Maezawa, New power generation system by concentrating waste heat below 200 °C from factory, Clean Energy 20 (2011) 5-10. [2] S.W. Chi, Heat Pipe Theory and Practice, Hemisphere Publishing Corp., 1976, pp. 1-35. [3] P.D. Dunn, D.A. Reay, Heat Pipes, Pergamon Press Ltd., London, 1977, pp. 1-16. [4] C.C. Hsieh, B. Wang, C. Pan, Dynamics visualization of two-phase flow patterns in a natural circulation loop, Int. J. Multiph. Flow 23 (1997) 1147-1170. [5] R. Khodabandeh, Heat transfer in the evaporator of an advanced two-phase thermosiphon loop, Int. J. Refrig. 28 (2005) 190-202. [6] R. Khodabandeh, Pressure drop in riser and evaporator in an advanced two-phase thermosiphon loop, Int. J. Refrig. 28 (2005) 725-734.
12
ACCEPTED MANUSCRIPT
[7] A.M. Alklaibi, Evaluating the possible configurations of incorporating the loop heat pipe into the air-conditioning systems, Int. J. Refrig. 31 (2008) 807-815. [8] Y. Chen, L. Cheng, G. Xin, T. Luan, Steady state modeling of LHP and analysis, Proc. of 14th International Heat Transfer Conference, Paper No. IHTC14-22351, 2010. [9] H. Louahlia-Gualous, B. Mecheri, D. Nortershauser, S. Le Masson, Transient characteristics of a two-phase thermosiphon loop for cooling telecommunication outdoor cabinets, Paper No.
RI PT
IHTC14-22218, 2010. [10] Y. Dobriansky, Concepts of self-acting circulation loops for downward heat transfer (reverse thermosiphons), Energy Convers. Manag. 52 (2011) 14-425.
[11] R. Khodabandeh, Thermal performance of a closed advanced two-phase thermosiphon loop for cooling of radio base stations at different operating conditions, Appl. Therm. Eng. 24 (2004)
SC
2643-2655.
[12] V. Dube, A. Akbarzadeh, J. Andrews, The effects of non-condensable gases on the performance of loop thermosiphon heat exchangers, Appl. Therm. Eng. 24 (2004) 2439-2451.
M AN U
[13] R. Khodabandeh, R. Furberg, Instability, heat transfer, and flow regime in a two-phase flow thermosiphon loop at different diameter evaporator channel, Appl. Therm. Eng. 30 (2010) 1107-1114.
[14] C. Sarno, C. Tantolin, R. Hodot, Y. Maydanik, S. Vershinin, Loop thermosiphon thermal management of the avionics of an in-flight entertainment system, Appl. Therm. Eng. 51 (2013) 764-769.
TE D
[15] A. Samba, H. Louahlia-Gualous, S. Le Masson, D. Nörtenhäuser, Two-phase thermosiphon loop for cooling outdoor telecommunication equipments, Appl. Therm. Eng. 50 (2013) 1351-1360.
[16] A.J.P. Zimmermann, C. Melo, Two-phase loop thermosiphon using carbon dioxide applied to the cold end of a Stirling cooler, Appl. Therm. Eng. 73 (2014) 549-558.
EP
[17] J. Li, F. Lin, G. Niu, An insert-type two-phase closed loop thermosiphon for split-type solar water heaters, Appl. Therm. Eng. 70 (2014) 441-450.
AC C
[18] H. Zhang, S. Shao, H. Xu, H. Zou, C. Tian, Integrated system of mechanical refrigeration and thermosiphon for free cooling of data centers, Appl. Therm. Eng. 75 (2015) 185-192. [19] Y. Xie, J. Zhang, L. Xie, Y. Yu, H. Wu, H. Zhang, H. Gao, Experimental investigation on the operating characteristics of a dual compensation chamber loop heat pipe subjected to acceleration field, Appl. Therm. Eng. 81 (2015) 297-312. Acknowledgement A part of this study was supported by Niigata Prefecture financial assistance 2011 to 2013. The authors are grateful to express our appreciation for their support.
13
ACCEPTED MANUSCRIPT
RI PT
Table 1 Experimental conditions Thermosiphon mounting
No inclination mode
Inclination mode
Mounting angle
0°
-45°, -10°, -5°, 5°, 10°, 20°, 30°, 45°, 60°
Thermal
receiver
80°C–200°C
SC
temperature
70°C–170°C
0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9
0.3
AC C
EP
TE D
M AN U
Water filling ratio
Fig. 1 Exhaust heat in Japan, 2011.
14
SC
RI PT
ACCEPTED MANUSCRIPT
TE D
M AN U
Fig.2 Thermosiphon loop thermal collector.
℃
(a) Thermal receiver plate
(b) Inside of the condenser
AC C
EP
Fig. 3 Components of thermosiphon loop.
(a)
<0
(b)
=0
Fig. 4 Definition of mounting angle
(c)
>0
15
ACCEPTED MANUSCRIPT 120
80
30% 40% 50%
40 20
1
2
60% 70%
3 4 5 6 Measurement point
(a) T1 = 80°C–110°C
180
] Temp. [
140 120
30% 40% 50% 1 2
60% 80% 70% 90% 3 4 5 6 Measurement point
TE D
100 80
7
M AN U
160
80% 90%
RI PT
60
SC
Temp. [
]
100
7
(b) T1 = 130°C–160°C
EP
220
]
200
Temp. [
AC C
180 160 140 120
30% 40% 50% 1 2
60% 80% 70% 90% 3 4 5 6 Measurement point
(c) T1 = 180°C–220°C Fig. 5 Temperature distribution.
7
16
ACCEPTED MANUSCRIPT 300 T5 ,P 5 T6 ,P 5 T7 ,P 5 Psat
=30 P [kPa]
200
100
-100 20
40
60 80 Temp. [ ]
300
120
=60%
M AN U
P [kPa]
100
SC
(a) α = 30%
200
RI PT
0
100
T5 ,P5 T6 ,P5 T7 ,P5 P sat
0
TE D
-100 50
100 Temp. [
150 ]
(b) α = 60%
EP
300
=90
P [kPa]
AC C
200
100 T 5 ,P5 T 6 ,P5 T 7 ,P5 Psat
0
-100 50
100 Temp. [
150 ]
(c) α = 90% Fig. 6 Measurement of temperature and pressure.
17
ACCEPTED MANUSCRIPT
100 T5 T6
RI PT
80 70 60 50
0
50
100 time [sec]
SC
Temp. [ ]
90
150
200
M AN U
Fig. 7 Temperature transition of oscillating mode.
100
60 40
EP
20
TE D
[%]
80
AC C
0 50
Normal Temperature is oscillating
100 150 200 Thermal Receiver Temp. [ ] Fig. 8 Flow map.
250
18
ACCEPTED MANUSCRIPT 700 600
400 300
RI PT
Q [W]
500
30% 40% 50% 60%
200 100 75
70% 80% 90%
225
SC
100 125 150 175 200 Thermal Receiver Temp. [ ]
M AN U
Fig. 9 Recovered heat for θ = 0°.
0.6
30% 40% 50% 60%
0.4 0.3 0.2
70% 80% 90%
TE D
R t [K/W]
0.5
100 125 150 175 200 Thermal Receiver Temp. [ ]
225
EP
0.1 75
Fig. 10 Thermal resistance for θ = 0°.
AC C
90
[kW/(m K)]
80 70 60 50 40 30 20
10 75
30% 40% 50% 60%
70% 80% 90%
100 125 150 175 200 Thermal Receiver Temp. [ ]
225
Fig. 11 Effective thermal conductivity for θ = 0°.
19
ACCEPTED MANUSCRIPT 700 [°] 0 5 10 20 30
600
400 300
RI PT
Q [W]
500
[°] 45 60 75 90
200 100
80 100 120 140 160 Thermal Receiver Temp. [ ]
180
SC
0 60
Fig. 12 Recovered heat for positive inclination mode.
1.2
M AN U
[°] 0 5 10 20 30
Rt [K/W]
1 0.8 0.6
0.2
TE D
0.4
[°] 45 60 75 90
0 60
80 100 120 140 160 Thermal Receiver Temp. [ ]
180
EP
Fig. 13 Thermal resistance for positive inclination mode.
80
[kW/(m K)]
AC C
70 60 50 40 30 20 10
0 60
[°] 0 5 10 20 30 80 100 120 140 160 Thermal Receiver Temp. [ ]
[°] 45 60 75 90 180
Fig. 14 Effective thermal conductivity for positive inclination mode.
20
ACCEPTED MANUSCRIPT
700 600
400
RI PT
Q [W]
500
300 200 100 120 140 160 Thermal Receiver Temp. [ ]
180
SC
100 80
Fig. 15 Reproducibility of recovered heat.
M AN U
0.5
0.3 0.2 0.1
TE D
Rt [K/W]
0.4
0 80
100 120 140 160 Thermal Receiver Temp. [ ]
180
EP
Fig. 16 Reproducibility of thermal resistance.
80
[kW/(m K)]
AC C
70 60 50 40 30
20 80
100 120 140 160 Thermal Receiver Temp. [ ]
180
Fig. 17 Reproducibility of effective thermal conductivity.
ACCEPTED MANUSCRIPT Highlights The new thrmosiphon loop thermal collector is proposed for the waste heat recovery. Using water as working fluid enabled to transport heat as hot as 150℃.
RI PT
The effective thermal conductivity of this apparatus is 200 times as high as copper.
AC C
EP
TE D
M AN U
SC
This apparatus maintains preferable performance when inclined to horizontal setting.
S1359-4311(15)00897-2
DOI:
10.1016/j.applthermaleng.2015.09.004
Reference:
ATE 6986
To appear in:
Applied Thermal Engineering
Received Date: 11 June 2015 Revised Date:
1 September 2015
Accepted Date: 2 September 2015
Please cite this article as: K. Matsubara, Y. Matsudaira, I. Kourakata, Thermosiphon Loop Thermal Collector for Low-Temperature Waste Heat Recovery, Applied Thermal Engineering (2015), doi: 10.1016/j.applthermaleng.2015.09.004. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
Thermosiphon Loop Thermal Collector for Low-Temperature Waste Heat Recovery
Koji Matsubara, Yusaku Matsudaira and Itaru Kourakata
Department of Mechanical and Production Engineering, Niigata University
Corresponding author: Koji Matsubara, Professor Department of Mechanical and Production Engineering, Niigata University Ikarashi 2-nocho 8050, Niigata 950-2181, Japan E-mail [email protected] TEL & FAX (+)81-(0)25-262-7260
2
ACCEPTED MANUSCRIPT Abstract
This paper describes the thermal collector type loop thermosiphon for low-temperature A1 waste heat recovery. Water is used as the working fluid for heat transport at temperatures 100
and higher. The loop thermosiphon comprises of a thermal receiver, a condenser, and
riser and downcomer tubes. The thermal receiver is made of a cupper plate brazed by meandering heat transfer tube. This receiver collects the thermal radiation from the electric
RI PT
heater at the heat transfer area of 1,000 cm2 (40cm x 25 cm), and transports the heat by vaporization of water to the condenser having heat transfer area of 62 cm2. In the no inclination mode, the thermosiphon is upright so that the thermal receiver and the condenser are placed vertically. In this mode, the effective thermal conductivity exceeds 60 kW/(m K) when the thermal receiver temperature was higher than 125
for the water filling ratio
ratio
SC
=30% to 70%. Although the effective thermal conductivity is deteriorated for the higher filling =80% and 90%, the effects from the filling ratio is tiny for
=30% to 70%. The
experimental tests were also made for the negatively inclined mode where the inlet and exit
M AN U
ports of the receiver were directed downward and for the positively inclined modes where they were directed upward. The tests revealed that the negative inclination almost halted the heat transport. However, the tests also indicated that the positive inclination showed the performance comparable to the no inclination mode for the cases up to the inclination angle 90°.
TE D
Keywords
Loop thermosiphon, Low-temperature waste heat, Heat transfer experiment, Effective A2 thermal conductivity, Inclination angle 1. Introduction
EP
There is an increasing concern about reducing fossil fuel consumption. This concern arises from the diminishing energy resources and increasing carbon dioxide levels in the atmosphere.
AC C
Therefore, it is becoming more important to utilize low-temperature waste heat, which is now mostly unused. Figure 1 shows the statistics from Japan in 2011 [1]. The figure reveals that a large amount of waste heat was exhausted at temperatures below 200°C. The annual waste heat between 100°C and 200°C is 282 peta (1015) calories, which is 9.1% of the crude oil consumption there. The recovery of low-temperature waste heat is important for slashing fossil fuel consumption. Heat transfer devices are expected to be developed for waste heat recovery. A device with no external means of power is desirable because of the low density of the waste heat. Loop-style thermosiphons have been emerging as transportation devices. These have been applied for the removal of high-density thermal loads. Their apparent thermal conductivity exceeds 200 times that of copper [2]. Academic papers were published on a thermosiphon loop for electronics cooling [4-10]. Previous papers used isobutene, pentane, and nitrogen as
3
ACCEPTED MANUSCRIPT
thermal media, setting temperatures at about 50°C [4-10]. More recently, Khodabandeh experimentally examined a radio base station cooling loop that was filled with isobutene [11]. Dube et al. performed experiments on the effects of a noncondensable gas on loop thermosiphon heat exchangers with water as the working fluid [12]. Khodabandeh and Furberg [13] visualized the bubbly flow of R134a by using an evaporator with a polycarbonate window [13]. Sarno et al. [14] tested the passive cooling system of an R141 loop for in-flight n-pentane
thermosiphon
loop
for
cooling
outdoor
RI PT
entertainment avionics. Samba et al. [15] performed transient and steady-state analyses of an telecommunication
equipment.
Zimmermann and Melo experimented on a carbon dioxide thermosiphon loop for a Stirling A4 cooler [16]. Li et al. [17] performed a visualization experiment of an insert-type closed loop water thermosiphon for solar water heaters. Zhang et al. [18] proposed the integration of a
SC
mechanical refrigeration system and thermosiphon for the free cooling of data centers. Xie et al. [19] investigated a dual compensation chamber loop heat pipe subjected to an acceleration field for high-power and high-heat-flux electronic devices in aircraft and spacecraft.
M AN U
Increasing numbers of academic papers have recently been written on the thermosiphon loop. Most of these papers intended to contribute to electronics cooling and they treated refrigerants as the working fluid [4-11, 13-16, 18, and 19]. These papers set the working temperatures at levels from 0°C to 50°C, which are not high enough for waste heat transport. Few papers [12, 17] used water as the working fluid, setting a temperature as high as 100°C. A4 However, there are only limited cases of experiments for working temperatures higher than
TE D
100°C. There are still many unknown points concerning how a thermosiphon loop works and what phenomena occur under high-temperature conditions. The group at Niigata University [2] developed a water thermosiphon loop for waste heat recovery. The evaporator, consisting of three layers of copper blocks, was designed to resist the high saturation pressure of water. This thermosiphon loop of water works for
EP
high-temperature conditions in which the condenser wall temperature was increased beyond 150°C. This is because critical temperature of water is 374
and phase change occurs for the A3
AC C
temperatures under the critical temperature. This paper extends the previous study [2] to consider the thermal-collector-type thermosiphon loop with water. The heat transfer area of the evaporator was enlarged by using a meandering pipe attached to a copper plate. Series of experiments were conducted for this thermal collector loop to reveal circulation flow stability and heat transfer characteristics. 2. Experimental method Figure 2 shows a photograph and line drawing of the thermal collector thermosiphon loop. This apparatus comprises a thermal receiver plate, condenser, riser, and downcomer. The thermal receiver plate is made of copper. The riser and downcomer are bent stainless steel tubes. The thermal receiver is heated from the bottom surface by an infrared lamp heater. The condenser is cooled by a stream of coolant water. The working fluid is designed to
4
ACCEPTED MANUSCRIPT
circulate in the thermosiphon loop by natural convection. This can be confirmed by the temperature distribution, as will be discussed later. The heat transfer area of the receiver plate is 1000 cm2 (40 cm × 25 cm). That of the condenser wall is 62 cm2 (π × 0.8 cm × 25 cm). The low-density heat flux of the waste heat is thus intended to be increased at the condenser by shrinking the heat transfer area. The ratio of receiver plate to condenser wall area is A6
1000:62 = 16.2:1.
RI PT
As shown in Fig. 3, the thermal receiver plate is soldered to a meandering copper tube. This tube is made by pressing a round tube (φ = 8.0 mm, thickness = 5 mm) until it has an oval cross section with a height of 5 mm and width of 10 mm. The tube was pressed so that it A5 has a larger surface to contact the plate in order to reduce thermal resistance. The condenser is an annulus with a copper center pipe and aluminum alloy shell, as indicated in Fig. 3. The
SC
preliminary experiment revealed that the coolant water was boiled in the copper pipe and exited the condenser as a bubbly stream. This was not suitable for the determination of recovered heat because the difficulty of obtaining quality measurements. The copper
M AN U
condenser pipe was thus covered by a stainless steel jacket with a thickness of 6.0 mm to increase the thermal resistance. This could prevent the coolant water from boiling and allow the recovered heat to be calculated on the basis of the temperature difference between the inlet and exit of the condenser cooling side. The thermosiphon loop is insulated by foam A9 insulation on the riser and downcomer, and by sheet insulation on the thermal receiver and condenser. The temperature distribution is measured using thermocouples set on the thermal
TE D
receiver and tube joints, as illustrated in Fig. 2. The recovered heat was obtained from the temperature difference and flow rate of the coolant water. Table 1 summarizes the experimental conditions. The water filling ratio is defined as the ratio of the mass of filled water to the mass that can fill the entire volume of the loop Vmax = 105 mL, α = m / mmax . Measurements were performed for the case where the thermosiphon
EP
loop was mounted at an inclination angle of 0 degrees and the case where the loop was inclined from −45 to 90 degrees. The definition of the mounting angle is as shown in Fig. 4.
AC C
The thermosiphon loop with no inclination was tested with a water filling ratio that changed from 0.3 to 0.9, whereas the inclined loop was tested with the water filling ratio fixed at 0.3. The thermal load was regulated by changing the input voltage to the infrared lamp heater from 50 V to 100 V. Heat radiation occurred from the infrared lamp to the thermal receiver. Since the absorption rate of radiation at the thermal receiver plate is not very clear in the experiment, its temperature was used to represent the experimental condition. This temperature was set at 80°C to 200°C for the non-inclined thermosiphon loop and at 70°C to 170°C for the inclined loop. Temperature is measured by thermocouples at seven positions A12 indicated in Fig. 2. Pressure is measured at positions 4 and 5 indicated in Fig. 2. The pressure A10 sensor is a strain gauge type pressure transmitter, KM31, produced by Nagano Keiki. 3. Results and discussion
5
ACCEPTED MANUSCRIPT 3.1 Temperature distribution and pressure in the thermosiphon loop
Figure 5 shows the temperature distribution for the thermosiphon loop. This figure divides the experimental data into three groups according to thermal receiver plate temperature: (a)
T1 = 80°C–110°C, (b) T1 = 130°C–160°C, and (c) T1 = 180°C–220°C. The horizontal axis A12 indicates the positions of thermocouples as shown in Fig. 2. Thermocouple no. 1 is on the thermal receiver plate, nos. 2 and 3 are on the evaporator tube, no. 4 is in the bottom of the
RI PT
riser, and nos. 5 and 6 are in the inlet and exit of the condenser, respectively. No. 7 is in the bottom of the downcomer. If water circulates by climbing up the riser and descending the downcomer,
temperature
may decrease
according
to
thermocouple
number order.
Temperatures at 2 and 3, on the evaporator tube, may reverse even when the flow circulates stably since a temperature rise can occur between 2 and 3 because of heat absorption.
SC
In Fig. 5, the temperature generally decreases along with the thermocouple number. Thus, it is indicated that the fluid circulates in the thermosiphon loop by going up the riser and coming down the downcomer. The temperatures are reversed between 2 and 3 for conditions
M AN U
where the water filling ratio is high. This occurs because of irradiation and does not contradict the stable circulation of the working fluid as mentioned above. However, A11 temperature increases in the condenser (5 to 6) and downcomer (6 to 7) for some conditions of each group where the water filling ratio is relatively low at 30 % to 60 %. These cases suggested that the fluid flow did not circulate as designed.
The pressure was measured at the lower and upper ends of the riser, corresponding to
TE D
positions 4 and 5 respectively, as shown in Fig. 2. According to the measurements, pressure at the lower end is higher than at the upper end with an average difference of 2.3 kPa. It is thus suggested that the ascent of the working fluid is aided by pressure difference in the riser. A11 Since a pressure difference in the of order 2 kPa is quite small compared with the pressure change due to the heating conditions, the pressure distribution is considered negligible in the
EP
examination of how the heating conditions result in the phase changes. Figure 6 shows the A11 relationship between pressure and temperature at Position 5 together with the saturation
AC C
pressure and temperature. This figure also includes the temperatures at the upper and lower ends of the downcomer, indicated as Positions 6 and 7, respectively in Fig. 2. Since there is no experimental data for the pressures at 6 and 7, the temperatures at these positions were plotted against the pressure at 5 in Fig. 6. Therefore, the pressure indicated by T6 and T7 is not exactly the same as the pressure at Positions 6 and 7. However, for the reason mentioned above, the pressure distribution is be considered negligible compared with changes due to the heating conditions. The relationships between (T6, P5) and (T7, P5) approximate the relationships between (T6, P6) and (T7, P7) in discussing phase changes. Figure 6 (a) shows the results for a filling ratio of 30%. In this figure, the temperature at Position 5 is almost coincident to the saturation temperature at the same pressure. Therefore, the working fluid is indicated to be saturated at Position 5, along with the whole temperature region for the test. It is different from the coincidence between T5 and Tsat that the A11
6
ACCEPTED MANUSCRIPT
temperatures at Positions 6 and 7 are clearly lower than the saturation temperature at the same pressure. The working fluid is thus in a state of compressed water at 6 and 7. In Fig. 6 (b), the temperatures at 5, 6, and 7 are shown with the pressure at 7 for a filling ratio of 60%. The characteristics of the temperature and pressure relationship are similar to those for the lower filling ratio of 30% when the temperature is above 130°C. However, for the lower temperature case, the temperatures at 6 and 7 correspond to the saturation
RI PT
temperature rather than temperature at 5. In this case, the working fluid is assumed to be saturated in the condenser and in a state of compressed water in the upper riser. This A11 corresponds to the inversed temperature at condenser as discussed above. Figure 8 (c) shows the results for a filling ratio of 90%. In the case where the temperature is lower than 100°C for this condition, the temperatures at 5, 6, and 7 are lower than the saturation temperature. In
SC
these cases, the working fluid is assumed to be compressed water, and the thermal load is not high enough to saturate the water.
The presentation of temperatures using the measured pressure is summarized as follows.
M AN U
Normal circulation is assumed at all temperatures for a 30% filling ratio and at relatively high temperatures for 60 and 90% filling ratios. In these cases, the working fluid is assumed to circulate normally such that the fluid is vaporized in the riser and condensed water in the downcomer. However, circulation is assumed to be abnormal at low temperatures for the 60% filling ratio. The shortage of thermal load for such case is likely to cause the unstable A11 circulation flow. It is not clear how the working fluid moves in the loop for these abnormal Sentences
TE D
cases. The temperature distribution suggests that the downcomer introduces vaporized water restructured into the condenser in these unstable conditions. However, it is still ambiguous how the condensed water returns to the thermal receiver plate under these conditions. There is a possibility that the downcomer works as a heat pipe to raise and lower the water through one tube. There is also a possibility that the riser works to return the condensed water to the
EP
receiver plate. These are not in line with the water circulation as intended. The unstable circulation occurs for the low temperature conditions where thermal load is insufficient. It is
AC C
recommended that the thermosiphon loop is utilized for stable conditions with a sufficient thermal load.
The experimental tests discussed in this paper had no flow oscillation. However, other experiments showed temperature oscillations that were remarkable at the inlet and exit of the condenser (Positions 5 and 6). Figure 7 illustrates the temperature oscillation at these positions for a 60% water filling ratio and thermal receiver temperature T1 = 94.3°C. These temperatures were observed to oscillate in the order of tens of seconds. In the figure, the temperature at the condenser inlet (Position 5) tends to be higher than that at the exit
A11 (Position 6) but is reversed between 120 and 150 s. Such oscillations probabilistically occur for A11 iterative tests in the same conditions. Figure 8 summarizes the flow pattern based on three sets of experimental tests. The normal circulation and oscillating flow are distinguished by a temperature oscillation at the condenser inlet and exit (Positions 5 and 6). This temperature A11 A11
7
ACCEPTED MANUSCRIPT
oscillation occurs for cases where the thermal receiver temperature is lower than 130°C. The temperature oscillation tend to occur for low-temperatures cases with insufficient thermal load. It is in contrast to the lack of experimental reproducibility at low temperatures that A11 normal steady circulation occurs at high temperatures. The figure indicates that the thermosiphon loop works normally for the case with sufficiently high thermal load. The experimental data for the flow oscillation are included only in Figs. 7 and 8. Other figures
3.2 Thermal performance of the thermosiphon loop
RI PT
exclude the flow oscillation data.
Figure 9 shows the recovered heat at the condenser versus the thermal receiver temperature. This figure includes the results for water filling ratios from 30% to 90%. The
SC
recovered heat ranges from 100 to 600 W, which increases with the increase in thermal receiver temperature. The recovered heat for the 90% filling ratio lies below that for the lower filling ratio. The recovered heat for the 80% filling ratio decreases at low temperatures from A11
M AN U
the cases of the lower filling ratio. This indicates that the high value of the filling ratio tends to diminish the heat transported by the thermosiphon loop. It is suggested that the water flow A7, A8 rate decreases as the water filling rate increases since the heat transfer at the condenser essentially corresponds to the flow rate of latent heat. There are relatively high values of recovered heat for filling ratios from 30% to 70%. The recovered heat in these conditions essentially does not depend on the filling ratio except for the case where the thermal load is
TE D
attenuated at thermal receiver temperatures below 100°C. The circulation of water is A11 suggested to deteriorate for low temperatures but to increase along with the increase of the thermal load based upon the correspondence between the water circulation and the recovered heat.
This paper defines thermal resistance using the thermal receiver temperature TR and TR − TC . Q
(1)
AC C
Rt =
EP
the temperature on the outer wall of the condenser copper pipe TC as
This is plotted against thermal receiver temperatures shown in Fig. 10. The higher filling ratios are inferior to lower filling ratios in terms of thermal resistance. The thermal resistance for α = 90% and α = 80% is clearly higher than it is for lower filling ratios, for the A11 cases of thermal receiver temperatures below 150°C. The thermal resistance is as low as 0.14 K/W for filling ratios of 30% ≦ α ≦ 70%, which shows the superiority of lower filling ratios. The thermal resistance tends to surge for temperatures as low as 80°C when α = 40% and α = 60%. For α = 30%, a low thermal resistance is maintained for the entire range of thermal receiver temperatures. The superiority of low filling ratios is thus suggested by the stably suppressed thermal resistance for decreasing thermal receiver temperatures to 75°C. The effective thermal conductivity is defined using the cross-sectional area A of the riser and downcomer and the tube length L as
8
ACCEPTED MANUSCRIPT λ=
QL
(TR − TC )A
.
(2)
The cross-sectional area is determined by the outer diameter of the riser and downcomer tubes as 2A ( = 2πd 2/4) = 1.00 cm2, where d = 0.8 cm. The riser length of 97 cm is used as the tube length in equation (2). The effective thermal conductivity is as high as 60 kW/(m K) for cases of low water filling ratio. This tends to decrease when the thermal receiver temperature
RI PT
is lower than 125°C, or the filling ratio is higher than 70%. However, the effective thermal conductivity remains stable at high values for α = 30%, 40%, and 50%. This surpasses 80 kW/(m
K) maximally, which is 200 times the thermal conductivity of copper. The
thermosiphon loop for high temperatures is found to be effective for transporting heat at temperatures beyond 100°C.
SC
The experimental data of the recovered heat, the thermal resistance and the effective A11 thermal conductivity depended on the water filling rate and the thermal load. All of the three quantities are deteriorated for the low thermal receiver temperatures below 125
and the
M AN U
higher water filling rate of α = 80% and 90%. This deterioration occurs due to the stagnation of water circulation which was suggested by decreasing recovered heat. In contrast, all the three quantities showed preferable performance for the higher thermal receiver temperatures than 125
and for the lower water filling rate of α = 30%, 40%, and 50%. The stable water
circulation, suggested by the recovered heat, is attributed to the preferably maintained
TE D
characteristics of the thermosiphon loop. 3.3 Effects of mounting angle
The thermosiphon uses gravity to return condensed water to the evaporator. The mounting angle severely affects the thermal performance of the apparatus. This study tested the thermosiphon loop by changing the mounting angle from θ = −45 to 90 degrees. For the
EP
negative angle mode, the apparatus was inclined so that the evaporator inlet and exit direct downward. For the positive angle mode, the apparatus was inversely inclined. In the
AC C
experiments for the negative angle mode, there was almost no temperature increase in the coolant water through the condenser even when the thermal receiver plate was irradiated at maximum heat load. There was almost no recovered heat at the condenser when the apparatus was negatively inclined. An inclination of only −5 degrees almost eliminated any heat recovery, which suggests that negative inclination severely deteriorates water circulation. Figure 12 shows the recovered heat for inclination angles from α = 0 to 90 degrees. The water filling rate was set at 30% in this section. Unlike the significant deterioration in negative angle mode, a positive inclination moderately affects the recovered heat. In fact, recovered heat existed for all of the selected positive angles. However, this decreased as the mounting angle increased. The decrease in recovered heat was as much as 100 W in the case of the maximum mounting angle α = 90°. The positive inclination maintains the recovered A11
9
ACCEPTED MANUSCRIPT
heat up to α = 90°, to suggest that the water circulation is stable for the cases of positive inclination. It is advantageous that the weak dependency on positive inclination assures freedom in the setting of the apparatus. The thermal resistance of the thermosiphon loop is shown in Fig. 13 for α = 0 to 90 degrees. There are no remarkable changes in thermal resistance for temperatures higher than 100°C. This proves the usefulness of the thermosiphon in the positive angle mode together with the
RI PT
moderate decrease in recovered heat mentioned above. The thermal resistance surges for temperatures lower than 100°C. In the low-temperature region, thermal resistance was kept low for positively inclined angles α = 30 and 45 degrees. These cases have improved thermal resistance versus no inclination (α = 0 degrees). However, the thermal resistance is highest at α = 90 degrees. Therefore, a slight positive inclination may enhance the thermal performance
SC
at low temperatures, but a large inclination angle tends to deteriorate it. At the higher A11 temperatures than 100 , the thermal resistance for α = 90 degrees is slightly higher than other cases. This slight deterioration comes to the stagnation of water circulation suggested
M AN U
by the recovered heat.
The effective thermal conductivity is presented in Fig. 14 for the same cases as in Figs. 12 and 13. This quantity is kept high when the mounting angle lies between 0 and 45 degrees. Increasing the mounting angle to more than 45 degrees leads to deterioration of the effective thermal conductivity. In fact, from α = 0 to 90 degrees, the effective thermal conductivity is decreased by 30% when the thermal receiver temperature is higher than 120°C. However, the
TE D
performance of the thermosiphon is preferably maintained at a positive inclination. It is A11 attributed to sustained water circulation mentioned above. Positive inclination is shown to be acceptable for usage of the current style of thermosiphon loop. There was a clear contrast between negative inclination and positive inclination in thermosiphon performance. In the positive inclination mode, the thermal receiver plate inlet and exit were inclined upward. This
EP
configuration is considered to assist water circulation due to gravitational force. In the positive inclination mode, vaporized water ascends to exit the thermal receiver plate by
AC C
buoyancy. In this mode, condensed water descends back to the plate because of its weight. Conversely, it is considered that a negative incline leads to the collapse of water circulation by preventing the exit of evaporated water and the return of condensed water. The experimental results indicate that the mounting angle affects water circulation severely, but the optimum mounting angle to maintain heat transfer performance can be selected on the basis of water A11 circulation. 3.4 Error analysis Reproducibility of the experimental data was confirmed by iterative measurements. Figure 15 shows the combined recovered heat. This figure includes error bars showing twice the standard deviation from the five sets of measurements. These experimental data were acquired using the thermosiphon with no incline with a water filling ratio of α = 60%. The thermal resistance and the effective thermal conductivity are shown in the same way in Figs A11
10
ACCEPTED MANUSCRIPT
16 and 17, respectively. Whereas the unstable flow pattern may have harmed the experimental reproducibility for the low temperature cases, the measured quantities were reproducible at least for the high temperature cases. The recovered heat has a standard deviation of 12.8 W maximally, whereas the average value is 312 W. This standard deviation corresponds to 4% of the average value. The thermal resistance has a relatively high standard deviation for temperatures lower than 100°C. However, this quantity keeps the standard
RI PT
deviation low at higher temperatures. For thermal receiver temperatures above 100°C, the standard deviation is 0.008 K/W maximally, which corresponds to 6% of the average (0.140 K/W). The effective thermal conductivity resembles the thermal resistance in measurement A11 uncertainty. This quantity shows a large standard deviation for thermal receiver temperatures lower than 100°C and has lower standard deviation for higher temperatures.
SC
The effective thermal conductivity maximum standard deviation increases to 3.9 kW/(m K) for temperatures higher than 100°C. This is equivalent to 5.6 % of the average value (69.3 kW/(m K)). In all the three quantities the experimental uncertainty is modest for the higher
M AN U
temperatures than 100 , indicating the stable circulation of water. This low uncertainty vindicates the presented data and the analysis in the earlier sections. 4. Conclusions
This paper describes the heat transfer characteristics of a thermosiphon loop thermal collector. The working fluid is water for waste heat recovery at temperatures 100°C and
TE D
higher. The temperature distribution and inside pressure were measured to reveal the stability and performance of the thermosiphon loop. The conclusions thus acquired is summarized as follows:
(1) The temperature generally decreased along the thermal receiver plate, the lower end of the riser, inlet and exit of the condenser, and lower end of the downcomer. This indicated a
EP
normal circulation of working fluid: water vaporized in the thermal receiver plate, ascended in the riser, liquefied in the condenser, and descended in the downcomer.
AC C
However, the temperature was reversed between the condenser inlet and exit or between the upper and lower ends of the downcomer under some of the low-temperature conditions. This reversed distribution of temperatures occurred conspicuously when the thermal receiver temperature was lower than 140°C. This suggested that the circulation flow tends to be unstable for the cases where the thermal load was inadequately provided. (2) Temperature oscillation may occur at the inlet and exit of the condenser. The experimental tests were repeated three times to determine what conditions lead to flow oscillation. The experimental tests revealed that temperature oscillation is likely to occur for thermal receiver plate temperatures that are lower than 140°C. This oscillation probably occurs at low temperatures. However, the temperature is deterministically stable at the higher temperatures. The temperature oscillation is interpreted to occur for cases with inadequate thermal load.
11
ACCEPTED MANUSCRIPT
(3) The heat transport performance of the thermosiphon was evaluated by the recovered heat, with the thermal resistance and effective thermal conductivity changing the water filling ratio α = 30% to 90 %. These quantities were affected by the water filling ratio to indicate a deterioration of performance for α = 80% and 90%. However, they were not largely affected by the water filling ratios between α = 30% and 70%. The effective thermal conductivity exceeded 60 kW/(m K) when the thermal receiver temperature was higher
RI PT
than 125°C for α = 30% to 70%. This effective thermal conductivity is 100 times the molecular conductivity of copper.
(4) The effects of the mounting angle were examined by inclining the thermosiphon loop. For positive inclinations, the thermosiphon was tilted so that the evaporator tube inlet and outlet were directed upward. For negative inclinations, the thermosiphon was tilted
SC
inversely. The experiments in the negative inclination mode revealed that the coolant water showed almost no temperature increase, and there was almost no heat transport from the thermal receiver plate to the condenser. On the other hand, tests in the positive
M AN U
inclination mode indicated that the thermal performance of the thermosiphon loop is maintained for inclination angles of 75 degrees or less. Thus, it is possible that the thermosiphon is mounted in the positive inclination mode for practical usage. (5) The experimental uncertainty was examined by five iterative tests for a water filling ratio of 60%. There is a relatively large standard deviation for the low temperature cases where the thermal receiver temperature is set as high as 100°C. However, the standard
TE D
deviation is not remarkable for the higher-temperature cases. The ratio of the standard deviation to the average value is in the order of five percent for the recovered heat, thermal resistance, and effective thermal conductivity when the receiver temperature is set beyond 100°C. The stable circulation confirmed by the temperature distribution may reduce the scattering of the experimental data. The stability of the thermosiphon loop for
AC C
distribution.
EP
higher temperatures is observed by experimental reproducibility as well as temperature
References
[1] K. Maezawa, New power generation system by concentrating waste heat below 200 °C from factory, Clean Energy 20 (2011) 5-10. [2] S.W. Chi, Heat Pipe Theory and Practice, Hemisphere Publishing Corp., 1976, pp. 1-35. [3] P.D. Dunn, D.A. Reay, Heat Pipes, Pergamon Press Ltd., London, 1977, pp. 1-16. [4] C.C. Hsieh, B. Wang, C. Pan, Dynamics visualization of two-phase flow patterns in a natural circulation loop, Int. J. Multiph. Flow 23 (1997) 1147-1170. [5] R. Khodabandeh, Heat transfer in the evaporator of an advanced two-phase thermosiphon loop, Int. J. Refrig. 28 (2005) 190-202. [6] R. Khodabandeh, Pressure drop in riser and evaporator in an advanced two-phase thermosiphon loop, Int. J. Refrig. 28 (2005) 725-734.
12
ACCEPTED MANUSCRIPT
[7] A.M. Alklaibi, Evaluating the possible configurations of incorporating the loop heat pipe into the air-conditioning systems, Int. J. Refrig. 31 (2008) 807-815. [8] Y. Chen, L. Cheng, G. Xin, T. Luan, Steady state modeling of LHP and analysis, Proc. of 14th International Heat Transfer Conference, Paper No. IHTC14-22351, 2010. [9] H. Louahlia-Gualous, B. Mecheri, D. Nortershauser, S. Le Masson, Transient characteristics of a two-phase thermosiphon loop for cooling telecommunication outdoor cabinets, Paper No.
RI PT
IHTC14-22218, 2010. [10] Y. Dobriansky, Concepts of self-acting circulation loops for downward heat transfer (reverse thermosiphons), Energy Convers. Manag. 52 (2011) 14-425.
[11] R. Khodabandeh, Thermal performance of a closed advanced two-phase thermosiphon loop for cooling of radio base stations at different operating conditions, Appl. Therm. Eng. 24 (2004)
SC
2643-2655.
[12] V. Dube, A. Akbarzadeh, J. Andrews, The effects of non-condensable gases on the performance of loop thermosiphon heat exchangers, Appl. Therm. Eng. 24 (2004) 2439-2451.
M AN U
[13] R. Khodabandeh, R. Furberg, Instability, heat transfer, and flow regime in a two-phase flow thermosiphon loop at different diameter evaporator channel, Appl. Therm. Eng. 30 (2010) 1107-1114.
[14] C. Sarno, C. Tantolin, R. Hodot, Y. Maydanik, S. Vershinin, Loop thermosiphon thermal management of the avionics of an in-flight entertainment system, Appl. Therm. Eng. 51 (2013) 764-769.
TE D
[15] A. Samba, H. Louahlia-Gualous, S. Le Masson, D. Nörtenhäuser, Two-phase thermosiphon loop for cooling outdoor telecommunication equipments, Appl. Therm. Eng. 50 (2013) 1351-1360.
[16] A.J.P. Zimmermann, C. Melo, Two-phase loop thermosiphon using carbon dioxide applied to the cold end of a Stirling cooler, Appl. Therm. Eng. 73 (2014) 549-558.
EP
[17] J. Li, F. Lin, G. Niu, An insert-type two-phase closed loop thermosiphon for split-type solar water heaters, Appl. Therm. Eng. 70 (2014) 441-450.
AC C
[18] H. Zhang, S. Shao, H. Xu, H. Zou, C. Tian, Integrated system of mechanical refrigeration and thermosiphon for free cooling of data centers, Appl. Therm. Eng. 75 (2015) 185-192. [19] Y. Xie, J. Zhang, L. Xie, Y. Yu, H. Wu, H. Zhang, H. Gao, Experimental investigation on the operating characteristics of a dual compensation chamber loop heat pipe subjected to acceleration field, Appl. Therm. Eng. 81 (2015) 297-312. Acknowledgement A part of this study was supported by Niigata Prefecture financial assistance 2011 to 2013. The authors are grateful to express our appreciation for their support.
13
ACCEPTED MANUSCRIPT
RI PT
Table 1 Experimental conditions Thermosiphon mounting
No inclination mode
Inclination mode
Mounting angle
0°
-45°, -10°, -5°, 5°, 10°, 20°, 30°, 45°, 60°
Thermal
receiver
80°C–200°C
SC
temperature
70°C–170°C
0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9
0.3
AC C
EP
TE D
M AN U
Water filling ratio
Fig. 1 Exhaust heat in Japan, 2011.
14
SC
RI PT
ACCEPTED MANUSCRIPT
TE D
M AN U
Fig.2 Thermosiphon loop thermal collector.
℃
(a) Thermal receiver plate
(b) Inside of the condenser
AC C
EP
Fig. 3 Components of thermosiphon loop.
(a)
<0
(b)
=0
Fig. 4 Definition of mounting angle
(c)
>0
15
ACCEPTED MANUSCRIPT 120
80
30% 40% 50%
40 20
1
2
60% 70%
3 4 5 6 Measurement point
(a) T1 = 80°C–110°C
180
] Temp. [
140 120
30% 40% 50% 1 2
60% 80% 70% 90% 3 4 5 6 Measurement point
TE D
100 80
7
M AN U
160
80% 90%
RI PT
60
SC
Temp. [
]
100
7
(b) T1 = 130°C–160°C
EP
220
]
200
Temp. [
AC C
180 160 140 120
30% 40% 50% 1 2
60% 80% 70% 90% 3 4 5 6 Measurement point
(c) T1 = 180°C–220°C Fig. 5 Temperature distribution.
7
16
ACCEPTED MANUSCRIPT 300 T5 ,P 5 T6 ,P 5 T7 ,P 5 Psat
=30 P [kPa]
200
100
-100 20
40
60 80 Temp. [ ]
300
120
=60%
M AN U
P [kPa]
100
SC
(a) α = 30%
200
RI PT
0
100
T5 ,P5 T6 ,P5 T7 ,P5 P sat
0
TE D
-100 50
100 Temp. [
150 ]
(b) α = 60%
EP
300
=90
P [kPa]
AC C
200
100 T 5 ,P5 T 6 ,P5 T 7 ,P5 Psat
0
-100 50
100 Temp. [
150 ]
(c) α = 90% Fig. 6 Measurement of temperature and pressure.
17
ACCEPTED MANUSCRIPT
100 T5 T6
RI PT
80 70 60 50
0
50
100 time [sec]
SC
Temp. [ ]
90
150
200
M AN U
Fig. 7 Temperature transition of oscillating mode.
100
60 40
EP
20
TE D
[%]
80
AC C
0 50
Normal Temperature is oscillating
100 150 200 Thermal Receiver Temp. [ ] Fig. 8 Flow map.
250
18
ACCEPTED MANUSCRIPT 700 600
400 300
RI PT
Q [W]
500
30% 40% 50% 60%
200 100 75
70% 80% 90%
225
SC
100 125 150 175 200 Thermal Receiver Temp. [ ]
M AN U
Fig. 9 Recovered heat for θ = 0°.
0.6
30% 40% 50% 60%
0.4 0.3 0.2
70% 80% 90%
TE D
R t [K/W]
0.5
100 125 150 175 200 Thermal Receiver Temp. [ ]
225
EP
0.1 75
Fig. 10 Thermal resistance for θ = 0°.
AC C
90
[kW/(m K)]
80 70 60 50 40 30 20
10 75
30% 40% 50% 60%
70% 80% 90%
100 125 150 175 200 Thermal Receiver Temp. [ ]
225
Fig. 11 Effective thermal conductivity for θ = 0°.
19
ACCEPTED MANUSCRIPT 700 [°] 0 5 10 20 30
600
400 300
RI PT
Q [W]
500
[°] 45 60 75 90
200 100
80 100 120 140 160 Thermal Receiver Temp. [ ]
180
SC
0 60
Fig. 12 Recovered heat for positive inclination mode.
1.2
M AN U
[°] 0 5 10 20 30
Rt [K/W]
1 0.8 0.6
0.2
TE D
0.4
[°] 45 60 75 90
0 60
80 100 120 140 160 Thermal Receiver Temp. [ ]
180
EP
Fig. 13 Thermal resistance for positive inclination mode.
80
[kW/(m K)]
AC C
70 60 50 40 30 20 10
0 60
[°] 0 5 10 20 30 80 100 120 140 160 Thermal Receiver Temp. [ ]
[°] 45 60 75 90 180
Fig. 14 Effective thermal conductivity for positive inclination mode.
20
ACCEPTED MANUSCRIPT
700 600
400
RI PT
Q [W]
500
300 200 100 120 140 160 Thermal Receiver Temp. [ ]
180
SC
100 80
Fig. 15 Reproducibility of recovered heat.
M AN U
0.5
0.3 0.2 0.1
TE D
Rt [K/W]
0.4
0 80
100 120 140 160 Thermal Receiver Temp. [ ]
180
EP
Fig. 16 Reproducibility of thermal resistance.
80
[kW/(m K)]
AC C
70 60 50 40 30
20 80
100 120 140 160 Thermal Receiver Temp. [ ]
180
Fig. 17 Reproducibility of effective thermal conductivity.
ACCEPTED MANUSCRIPT Highlights The new thrmosiphon loop thermal collector is proposed for the waste heat recovery. Using water as working fluid enabled to transport heat as hot as 150℃.
RI PT
The effective thermal conductivity of this apparatus is 200 times as high as copper.
AC C
EP
TE D
M AN U
SC
This apparatus maintains preferable performance when inclined to horizontal setting.