- Email: [email protected]
Experimental study on solar thermal conversion based on supercritical natural convection
Renewable Energy 62 (2014) 610e618
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
Experimental study on solar thermal conversion based on supercritical natural convection Xin-Rong Zhang a, b, *, Yalong Zhang a, Lin Chen a a b
Department of Energy and Resources Engineering, College of Engineering, Peking University, Beijing 100871, China Beijing Key Laboratory for Solid Waste Utilization and Management, Peking University, Beijing 100871, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 September 2012 Accepted 19 August 2013 Available online 14 September 2013
In this paper, experimental investigation into the basic characteristics of solar thermal conversion using supercritical CO2 natural convection are presented. Natural circulation of supercritical fluids can be easily induced and even a small change in temperature can result in large change in density close to the critical point. The supercritical experimental system carefully designed and operated in this study. It is found that an obvious and continuous long-time drop of solar radiation would not affect the CO2 flow rate, temperature and pressure very much, if the solar radiation is in a relatively high-value level. This continuous drop can induce obvious drops in the CO2 flow rate, temperature and pressure only when the solar radiation is in a low-value level. Furthermore, it is observed that a long-time drop and low-value in the solar radiation may make the flow rate temporarily become zero, which should be paid more attention in future system design and operation. The collecting efficiency increases with the comprehensive coefficient and this pattern is contrary to that of water based system. In addition, it is found that there exist optimal flow rate and CO2 charge amount for system overall performance. This kind of solar thermal conversion has a higher collecting efficiency in spring and winter than summer and autumn; a better performance in cold and low-radiation region than hot and high-radiation region. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Supercritical fluid Natural convection Solar thermal energy Carbon dioxide
1. Introduction Among various solar utilization methods, solar thermal conversion has become one of the most promising kinds. Solar thermal conversion systems can be categorized into low-temperature solar thermal kinds, which usually do not use sunlight concentration, and medium-temperature and high-temperature solar thermal systems. Among those methods, low-temperature solar thermal systems have the potential to supply a significant number of households and commercial buildings with heating and/or cooling. Solar water heater is one of the most promising applications of lowtemperature solar thermal systems [1e7]. In recent years, a lot of studies have been carried out in the field of solar water heater and low temperature solar thermal convection [8e16], which are mainly for the solar water heaters of water-in-glass evacuated tube type [17,18].
* Corresponding author. Department of Energy and Resources Engineering, College of Engineering, Peking University, Beijing 100871, China. Tel.: þ86 10 82529066; fax: þ86 10 82529010. E-mail addresses: [email protected], [email protected] (X.-R. Zhang). 0960-1481/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.renene.2013.08.025
In traditional solar water heaters, both forced convection and natural convection are used. In natural-convective solar heaters, water is usually used as working fluid. Many previous studies were carried out to explore the flow dynamics, heat transfer and performances for natural circulation kind of solar water heaters [5]. From those studies, the natural-convective solar water heater owns a deficient characteristic, because water is an incompressible fluid and its natural convection is relatively weak under the solar heating condition. As an effort of improving the efficiency from solar to thermal energy, supercritical fluid was proposed as working fluid in solar collectors [19]. Supercritical fluid is a substance at a temperature and pressure above its thermodynamic critical point. Fig. 1 shows the thermo-physical properties of supercritical CO2. As shown in Fig. 1, the most important is that, close to the critical point, small changes in pressure or temperature can result in large changes in thermo-physical properties. The data shown in Fig. 1 is obtained from a Program Package for Thermo-physical Properties of Fluids (NIST Fluid Thermodynamic and Transport Properties database, version 8.0). Especially the density of supercritical fluid varies significantly as a function of temperature near the critical temperature. Such characteristics in thermo-physical properties allow a strong natural convection flow to be developed in the solar
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
611
Fig. 1. Variation of thermoproperties with temperature and pressure in the critical region, (a) specific heat; (b) density; (c) thermal conductivity; (d) viscosity.
thermal conversion system [19]. Furthermore, for supercritical CO2, its critical temperature is 31.1 C and that is low enough to be easily reached in the low temperature solar thermal conversion system. Solar water heater is the most widely used conversion unit around the world. Based on previous basic supercritical fluid tests [19e22], the solar water heater using supercritical fluid can be comprised of two loops: one is a supercritical fluid circulation loop and the other is a water loop. The two loops are coupled with each other by a heat exchanger. In the closed loop with supercritical fluid, the natural convective flow develops as a result of the heating process by solar radiation and the cooling process by the heat exchanging process with water. The supercritical fluid flow in the collector can absorb and transport heat, and then it transport the collected thermal energy to water. Therefore conversion from solar to thermal energy can be achieved. Inside the heat exchanger, supercritical fluid is cooled by the cold water flow (in real application, general tap water can be used). Water absorbs heat from the supercritical fluid and can be pipelined for related domestic or other costumer sides. A feasibility study was carried out in order to analyze whether the supercritical state can be achieved at the collector outlet and the Reynolds number was measured higher than 1900 in the previous experimental and numerical investigations [19e22]. Furthermore, some characteristics about this kind of solar water heater by supercritical CO2 were investigated by several experimental tests [20]. The obtained results revealed the supercritical CO2 flow rate is quite stable and will be less affected by the transient variations of the solar radiation. The solar thermal conversion process can be divided into three periods: starting-up, transition and stable period. High solar thermal conversion efficiency was found at a high mass flow rate and under operation pressure near to the critical point. However, in the field of supercritical fluid based solar thermal conversion systems, some basic characteristics of such natural
circulation based solar thermal conversion system design and effective operation are still not clear, such as the influence of solar radiation, supercritical fluid temperature and ambient temperature etc. on solar thermal conversion process and efficiency. Also the parameter description and seasonal performances of supercritical systems are still under development. As an extension of the previous study, in this paper, an experimental set-up is tested and systematic investigations are made to further clarify the basic characteristics of this solar thermal collection by supercritical natural convection flow. 2. Experimental set-ups Based on the concept described above, an experimental prototype of the solar thermal conversion system is tested. This experimental system is specifically designed for using supercritical CO2 natural circulation flow. A schematic diagram of the experimental set-up is presented in Fig. 3. The built experimental machine is mainly comprised of solar thermal collector, heat exchanger, valve 1, valve 2, hot water tank, valve 3, and measurement and data acquisition system. To make sure the solar water heater system is feasible, a solar collector with good heat collecting characteristics is required. Therefore, to effectively heat supercritical CO2 to relatively high temperature state in the experimental set-up, all-glass evacuated solar collector with a U-tube heat removal system. The collector is consisted of a glass envelope over glass tube that is coated with a selective solar absorber coating. The coating has a high solar absorbance of 0.918 and a low emissivity of 0.189. It is applied on the vacuum side of the inner glass tube. The transparence of the glass envelope is 0.92. Such design of the collector allows a maximum operating pressure as high as 15.0 MPa. In the present experimental set-up, evacuated solar collector of 1.69 m2 (gross area) is used which is manufactured basically based on the previous
612
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
ones using the supercritical CO2 forced convection [21e23]. The every U-shaped heat removal fluid tube utilized is 3.5 m long and 0.006 m internal diameter. Related system design and tests experiences in supercritical CO2 fluid systems [24e31] should be referred to as they may provide both experimental and theoretical information for the convenience of novel system development and manufacturing. In the supercritical fluid circulation loop, Valve 1 is operated by hand. Valve 1 can be used to adjust flow rate of the CO2 fluid. A concentric tube heat exchanger is utilized to couple the supercritical fluid circulation loop with the water loop. Supercritical fluid flows inside the inner tube of the heat exchanger and then solar heat is transferred through the wall of a stainless steel to water side. The concentric heat exchanger is a type of counter-flow circular pipe geometry. In this heat exchanger, the outer tube has an internal diameter of 25.0 mm and wall thickness of 1.2 mm and the inner tube is made of stainless steel, having an internal diameter and wall thickness of 6.0 mm and 1.0 mm, respectively. Heat exchanging area of the heat exchanger is 1.0 m2. A mass flow meter was used in this fluid loop to measure the mass flow of liquid CO2. This flow meter also has a maximum permissible operating pressure of 15.0 MPa. The flow meter was installed in the downstream of the heat exchanger in the supercritical fluid loop, as shown in Fig. 2. It provides a measurement range of 0.02e5.0 kg/min with an accuracy of 0.1%. It should be mentioned here that the supercritical fluid loop in the solar thermal conversion system is not symmetric, which can avoid an oscillation flow, based on our previous numerical studies [21e23,29]. As shown in Fig. 2, a hand-operated valve, valve 2, is installed in the water loop side. This valve is adjusted based on the water temperature which is measured at the outlet of the heat exchanger. In the water loop side, cool water flows into the heat exchanger and is heated by the supercritical CO2 fluid flow. Then hot water comes out of the heat exchanger. Hot water from the outlet of the heat exchanger flows into a hot water tank, as shown in Fig. 2. In the present experimental set-up, the hot water tank is made of stainless steel and its volume is 80 L. At the outlet of the hot water tank, there is a valve e valve 3, which can be used to collect hot water from the hot water tank. The valve 3 is also a hand-operated valve. In the measurement system, two T-type thermocouples and two pressure transmitters are respectively mounted at the inlet and outlet of the solar collector to measure CO2 temperatures and
pressures, with an accuracy of 0.1 C for temperature and 0.2% for pressure measurements. In addition, two platinum resistor temperature sensors were mounted to measure the inlet and outlet water temperatures of the heat exchanger. Its accuracy can be expressed as 0.15 þ 0.0002jtj C. Thermal insulation coating applied on all the CO2 and water loops to reduce heat losses from the piping. A meteorological measurement system was also installed. It is mainly comprised of sun radiation sensor, anemometer, and air temperature gauge and so on. In the present study, only measured solar radiation data is presented, and accuracy of the sun radiation sensor is 0.3%. This pyranometer is based on a thermopile sensor. Typically, the pyranometer output signal does not exceed 20 mV. The typical sensitivity is 10 mV/(W/m2). So the accuracy of the sun radiation sensor is less than 0.3%. In addition, a data acquisition system is used in the experimental set-up, which can be used to achieve real-time data measurement, acquisition, processing, and share. For each experiment, the solar collector is adjusted toward the direction of the Sun. Cool water flow is first pumped into the heat exchanger. Then according to the measured CO2 temperature, valve 1 is adjusted in order to achieve the expected extent. During the experiment, the two major sources of experimental error are: (1) temperature, pressure and flow rate measurement accuracy; and (2) errors resulting from data logging and reading by the computer. The error of data logging and reading are analyzed and listed as follows: 0.1 C for CO2 temperature; 0.2% for CO2 pressure; 0.2% for CO2 mass flow rate; 0.2 C for water temperature; 0.2% for water flow rate; 0.2% for solar radiation. Based on the above accuracies, accuracies of all the parameters defined in this paper are calculated to be less than 2.0% using the following equation:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pn 2 x21 þ x22 þ . þ x2n i ¼ 1 xi ¼ x ¼ n n
(1)
Where x is the average of the parameter x; xi is accuracy of every parameter involved in the calculation equation of x; and n is the number of the parameters in the calculation equation of x.
3. Results and discussion For first stage test, the solar collector is set declined, with an angle of 45 C with respect to the horizontal level in the present
Fig. 2. Flow diagram of the experimental set-up.
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
613
study, which indicates that the U-tubes in the collector has an angle of 45 from the horizontal level. The water inlet temperature is 20 C during all the experimental tests. The CO2 charge amount is 4.4 kg at most tests, except special mention. The following items are mainly measured: the solar radiation, the CO2 fluid temperatures at the inlet and outlet of the heat exchanger, the CO2 fluid pressures at the inlet and outlet of the solar collector, and the mass flow rates of CO2 fluid. Based on these measured quantities, the following parameters are defined to further learn the collecting efficiency, which evaluates the solar thermal energy conversion performance:
C ¼
Tf Ta
I
QC ¼ mCO2 ðhCO hCI Þ
hC ¼
QC IA
(2) (3) (4)
In this paper, CO2 enthalpy values at the different monitoring points were calculated based on the measured temperature and pressure values using a program package for thermo-physical properties of fluids: NIST Fluid Properties database (REFPROP 8.0). The experimental set-up was tested throughout the year of 2011 in the city of Shaoxing, Zhejiang Province of China. As the initial step of investigation into the fundamental characteristics of solar thermal conversion by supercritical CO2 natural convection, it should be noted that optimization structure and operation are not claimed in the present study. In the experimental test in the present study, the opening of valve 1 was adjusted according to the CO2 temperature at the outlet of the collector.
Fig. 3. Variations of the measured parameters with the test time on July 15, 2011. (a) Solar radiation and CO2 pressures; (b) CO2 temperatures and flow rate.
3.1. Transient characteristics of the solar thermal conversion system Based on the operation and measurement principles described above, the first focus point in this study is the transient characteristics of this kind of solar thermal conversion by supercritical natural circulation flow. Understanding transient characteristics is important to build up an optimal and high efficient energy conversion. The best way to obtain transient characteristics is to directly measure the variations of the operating parameters with time. Fig. 3(a) shows the variations of the measured solar radiation and CO2 pressures with the test time on July 15, 2011. Generally speaking, solar radiation in July in Shaoxing area indeed is representative for southeast part of China throughout a year. It is seen from this figure that this day is not sunny all the time and also not cloudy all the day, which represents a typical weather in Shaoxing area of Zhejiang Province. Sometimes in the day is sunny and sometimes is cloudy and its time-averaged solar radiation value is found at 610.0 W/m2, which also represents basic weather conditions seen in this area. From this figure, it is also observed that there is a sudden and obvious drop in the solar radiation around 12:30 and this solar radiation drop lasts shortly, which is marked by one circle in Fig. 3. In addition, a longer time period of obvious solar radiation drop is observed around 10:40, which lasts about 30 min. It is also seen that there are two obvious and continuous drops in the solar radiation, which respectively occur from 13:15 to 14:30 and from 15:00 to 16:30. These two continuous drops last about 75 min and 90 min, respectively. The measured variations of the mass flow rate and CO2 temperatures are presented in Fig. 3(b). From the curves shown in Fig. 3, it is seen that although the solar radiation drops slightly around 12:30, no obvious variation in CO2 temperature, pressures or CO2 mass flow rate happen. That means that the CO2
temperature, pressure and mass flow rate are less affected by a sudden and short-time solar radiation drop. The phenomenon was also partially observed in the previous study [20]. However, if such radiation drop lasts a relative longer time, as observed around 10:40, small drops in CO2 pressure, CO2 temperature at the collector outlet and CO2 mass flow rate are also seen near this time point. But the observed drops in the CO2 pressure, temperature and flow rate are later than the solar radiation drop. The phenomenon can be attributed to this solar thermal conversion being a thermal inertia system. This thermal inertia characteristic in supercritical fluid leads to the variations of the CO2 temperature, pressure and flow rate etc. are after the solar radiation heating. It can be concluded that the CO2 temperature, pressure and flow rate can be influenced by a longer time drop in solar radiation, although the observed drops in the CO2 temperature, pressure and flow rate are small and limited. But if we carefully observe that obvious and continuous drop in solar radiation occurring from 13:15 to 14:30, it is found that the CO2 temperature and flow rate only drop with a small amplitude although the solar radiation drop keeps a long time period almost 75 min. It is also seen no obvious drop in the CO2 pressure is observed due to this long time period drop in the solar radiation, and only two very small drop curves are observed. From Fig. 3, it is observed there are obvious variations in CO2 temperature, pressure and flow rate when a longer time drop in the solar radiation occurs around 15:00e16:30. That is different from the above observation. Although a continuous drop in the solar radiation also occurs for a long time period of 13:15e14:30, no obvious drops are found either in the CO2 temperature, flow rate or the CO2 pressure. The physical mechanism of that difference can be explained as follows: the continuous drop during 13:15e14:30 is in
614
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
a time period with relatively high level of the solar radiation. However, another continuous drop of 15:00e16:30 is close to the end period of the solar radiation heating during that day, and its solar radiation is in a low value level. The time-averaged solar radiation value during 13:15e14:30 is calculated at 550.0 W/m2, and the averaged value during 15:00e16:30 is found at 200.0 W/m2. The long-time and low solar radiation heating during 15:00e16:30 makes this thermal inertia system difficult to maintain its original flow rate, and the temperature and pressure would decrease. Therefore, obvious drops in the CO2 flow rate, temperatures and pressures occur due to this long and low solar radiation heating. Although a long time drop in the solar radiation also occurs from 13:15 to 14:30, the collected heat intensity is high enough to well support the CO2 flow, temperature and pressure. The low viscosity and high circulation rate of supercritical natural loops also contribute to this finding [29e31]. Therefore no obvious drops are observed during this experiment operation. In addition, it is also seen from Fig. 3(b) that during the continuous drop in the solar radiation from 15:00 to 16:30, the CO2 mass flow rate drops to very low values for three times. This observed phenomenon is much different from those obtained from the previous studies [19,20], where the supercritical CO2 flow rate is stable and is less affected by the sudden variation of the solar radiation. When the system reaches stable operation, the CO2 flow rate will keep at a high value even the solar radiation stays at a low level [20]. The stable period is defined as the time period starting from the first peak of the CO2 flow rate. Indeed no continuous solar radiation drop occurs in this period, as observed in Fig. 3, and also the solar radiation is in a low-value level. This long-time drop and relative low-value in the solar radiation lead to transient very low circulation rate in the system. The low-standard solar radiation heating is insufficient to sustain this supercritical natural convection flow. This kind of temporary very low flow rate phenomena is interesting and should be paid more attention in future engineering design and operation. At the same time, it is observed during the continuous drop in the solar radiation from 15:00 to 16:30, after the flow stops and then it can restart. After the transient flow rate drops, flow restarting process happens and such kind of oscillating flow situation are seen when the solar radiation values drops gradually in the late afternoon. When the flow restarts, the mass flow rate is seen to be at relative high standard or even high peaks. Three peaks in the CO2 mass flow rate are seen during this period. The phenomenon of reaching a peak value when the flow is started was also observed and verified in the theoretical analyses [21e23]. Here, it can be simply concluded that a very short time drop in the solar radiation will less affect the CO2 fluid flow rate, temperature and pressure. An obvious and longer time drop in the solar radiation may induce a small scale drop in the CO2 flow rate, temperature and pressure. A continuous long-time drop in the solar radiation would not affect the CO2 flow rate, temperature and pressure much, when the solar radiation is in relatively high-level. The continuous long-time drop in the solar radiation can induce obvious drops in the CO2 flow rate, temperature and pressure only if the solar radiation is in a low-value level, even the flow rate to near-zero values. Furthermore, it is seen that among the CO2 flow rate, temperature and pressure, the most affected parameter by obvious solar radiation variation is the CO2 flow rate, then the CO2 temperature and the minimum is the CO2 pressure. The physical mechanism is due to the sensitivity of supercritical fluids in response to the variations of solar radiation. After that, the basic viscosity drops (as show in Fig. 1) and the fluid flow will be susceptible to instabilities when high flow rate can generally be found in the system. Also, an obvious variation in the solar radiation heating directly affects the driven force and then the flow rate changes. The flow rate change
affects the heat transfer between the CO2 fluid and the collector and then the CO2 temperature vary. The evolution of heat collected from the water side and the transient collector efficiency as defined in Eq. (4) are plotted in Fig. 4 against test time. Also plotted is the evolution of solar radiation on the collector area (1.5 m2) in the same summer day (July 15, 2011). With the increase of solar radiation, the collected heat curve shows similar trend after supercritical CO2 flow started. It can be seen in Fig. 4 that the collector efficiency shows relative stable value around 60%, which indicates the relative stable behavior of the current test system from around 10:00 to 16:00. The detailed seasonal behaviors and efficiency will be discussed later in this paper. In addition, with the fluctuations of solar radiation as shown in Fig. 4, respective fluctuations of collected heat and efficiency are seen, which may due to a sensitive response of the system to the changes of flow rate (see Fig. 3(b)) and temperature curve (Fig. 3(a) and (b)). The collected heat is calculated according to the definition in Eq. (3) which can reflect the transient characteristics directly. First, the measured solar radiation and collected heat with the test time is presented in Fig. 3. The measurement is conducted on July 8, 2011. It is seen that the measurement starts from about 9:30 in the morning to 16:30 in the afternoon. 3.2. Time-averaged daily performance of solar thermal conversion system The collecting efficiency is the most important indicator of this solar thermal conversion performance. The main influencing factors to this solar thermal conversion system are solar radiation, ambient temperature, CO2 temperature, CO2 pressure and CO2 flow rate. The comprehensive coefficient C defined in the Eq. (2) is usually used to consider these influencing factors, which include the solar radiation, CO2 temperature in the collector and ambient temperature. In this session, this coefficient C will be utilized to show the influence of the factors on the solar thermal conversion. In addition, because the CO2 flow rate is an important design and operation parameter, its effect will also be focused on in this session. In order to figure out how the CO2 flow temperature affects the collecting efficiency, Fig. 5 shows the variations of the collecting efficiency with the measured CO2 temperature difference between the collector outlet and inlet. A wide range of the CO2 temperature differences from 2 C to 50 C are shown in Fig. 5, in that figure each point represents a time-averaged value of a day. It is seen from
Fig. 4. Variations of the measured collected heat, solar radiation on the collector area (1.5 m2) and collecting efficiency against test time on July 15, 2011.
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
Fig. 5. Variations of the solar thermal conversion efficiency with the collector temperature difference between outlet and inlet. Every point in this figure represents a time-averaged value per day.
Fig. 5 that the collecting efficiency increases with this CO2 temperature difference. When this CO2 temperature difference between the collector outlet and inlet is small, the collecting efficiency is found small. The collecting efficiency becomes large when the CO2 temperature difference is large. Based on Fig. 5, it is also found the collecting efficiency is at 0.18 when this CO2 temperature difference is 2 C; the collecting efficiency is 0.60 when the temperature difference increases to 27 C; the collecting efficiency increases to 0.83 when the temperature difference 53 C. The averaged collecting efficiency is calculated at 0.62 within the entire tests shown in Fig. 5. A physical explanation is also presented here for the phenomenon observed in Fig. 5. When the CO2 temperature difference increases, the more heat can be collected in the collector, according to the Eq. (3). The more collected heat quantity means increasing the collecting efficiency. Even the solar radiation level is the same, this more collected heat can be achieved by the variations in other parameters, such as ambient temperature. Therefore, the CO2 fluid temperature difference between the collector outlet and inlet should be increased as possible as we can in order to enhance the solar thermal conversion. Based on the previous test [20], the solar thermal efficiency is found to be increased when CO2 mass flow rate increases. Therefore efforts should be made in order to increase the CO2 mass flow rate so as to improve the system efficiency. Fig. 6 represents such an effort, where the measured variation in the CO2 mass flow rate with the solar radiation is presented. In this figure, every data point also represents a time-averaged value per day. Based on these measured daily-averaged values, it is found the CO2 mass flow rate increases
Fig. 6. Variations of the measured mass flow rate with the solar radiation. Every point in this figure represents a time-averaged value per day.
615
with the solar radiation during the entire tests. A good solar radiation heating provides a strong driven force for this supercritical natural convection flow. Then more high flow rate can be made corresponding to this good solar radiation. So, generally speaking, a good weather and high solar radiation are helpful to increase the collecting efficiency and to make an efficient solar thermal conversion. The collecting efficiency evolution trend for major representative experimental tests is shown in Fig. 7. It is found the collecting efficiency will decrease with the increase in the CO2 mass flow rate after this flow rate reaches to some value. This observed phenomenon reveals the previous test and observations are limited in the range. Indeed there exists an optimal CO2 mass flow rate, which is corresponding to the highest collecting efficiency. When the mass flow rate is less this optimal value, the collecting efficiency increases with the flow rate, because a quick flow can enhance the heat transfer and then the collecting efficiency. When the mass flow rate becomes larger than this optimal value, the collecting efficiency decreases with the flow rate increase, because when the flow becomes too quick, the flow cannot carry and transport the thermal energy out of the collector timely. Therefore, the collecting efficiency decreases when the mass flow rate is too high. Fig. 8 shows the evolutions of the collecting efficiency with comprehensive coefficient on August 5, August 19 and August 25, 2011, respectively. It is seen that the collecting efficiency is proportional to the comprehensive coefficient on August 25, 2011. For the measured data on August 5 and 19, though the collecting efficiency is not linearly proportional to the comprehensive coefficient, the general trend shows that the collecting efficiency will increase with the comprehensive coefficient. It is seen that the collecting efficiency points are scattered on both sides of the proportional line. This is much different from the case of water based solar collector. For water, the collecting efficiency is inversely proportional to the comprehensive coefficient, because an increase in the solar radiation or a decrease in the ambient temperature usually brings more heat loss from the solar collector. What physical mechanism leads to this difference found above? The physical explanation may be presented as followings: supercritical CO2 is one very much compressible fluid, and in the solar collector, the mass flow rate varies all the time, which is also different from the water case. For the water collector, the flow rate changes a little and that only depends on the pump power input. However, the flow rate of the supercritical fluid is affected by many factors, such as the solar radiation and ambient temperature etc. This compressibility of the supercritical fluid makes the relationship between the collecting efficiency and the comprehensive
Fig. 7. Variations of the solar thermal conversion efficiency with the measured CO2 mass flow rate. Every point in this figure represents a time-averaged value per day.
616
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
Fig. 9. Variations of the solar thermal conversion efficiency with the comprehensive coefficient under the CO2 charge amounts. Every point in this figure represents a timeaveraged value per day.
flow rate pattern observed in Fig. 6. There also exists an optimal CO2 charge amount, which owns the highest collecting efficiency and the best solar thermal conversion performance. The physical mechanism can be understood that a large CO2 charge amount would produce too quick flow to transport heat from the collector to the ambient. In addition, it is shown in Fig. 9 that the slope decreases when the CO2 charge amount increases, when this amount is less than 4.5 kg. As discussed, there is an optimal CO2 charge amount for this system, whose solar thermal conversion efficiency is the highest. This optimal amount is also considered as a ‘critical point’ or ‘saturation point’ to judge how much CO2 fluid should be charged into the system. When the charge amount is far from this ‘saturation point’, the line slope is larger, which means the collecting efficiency varies faster with C. Closer to the ‘saturation point’ the charge amount is, smaller slope the collecting efficiency has and slower variation will be seen between the collecting efficiency and C. 3.3. Seasonal performance of solar thermal conversion system
Fig. 8. Variations of the collecting efficiency with the comprehensive coefficient. (a) on August 25, 2011; (b) on August 5, 2011; (c) on August 19, 2011.
The measured seasonal variations of the collecting efficiency with C in four seasons are plotted in Fig. 10. It is found that generally the collecting efficiency will increase with C in four seasons. Furthermore, it is seen that the comprehensive coefficient C is
coefficient is contrary to the water case. This relationship observed in Fig. 8 should be pay attentions in the engineering design and operation. Furthermore, the measured variations of the collecting efficiency with C under different CO2 charge amounts are presented in Fig. 9. In this figure, four CO2 charge amounts are included and they are respectively 3.5 kg, 4.1 kg, 4.5 kg and 4.8 kg. For each CO2 charge amount, ten times tests are made. Every point in Fig. 9 represents a time-averaged value per day. Same with the pattern observed in Fig. 8, it is seen from Fig. 9 that the time-averaged collecting efficiency is also proportional to the comprehensive coefficient, under each CO2 charge amount. However, when the CO2 charge amount increases from 4.5 kg to 4.8 kg, a decrease in the collecting efficiency is observed. When the CO2 charge amount is less than 4.5 kg for the present system, the collecting efficiency will increase with the CO2 charge amount. Nevertheless, when the charge amount is larger than 4.5 kg, the collecting efficiency will decrease with the increase of CO2 charge amount. This phenomenon is similar to the
Fig. 10. Variations of the collecting efficiency with the comprehensive coefficient under four different seasons. Every point in this figure represents a time-averaged value per day.
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
generally larger in spring and winter than those in summer and autumn. Also, it should be noted that parameter C is relatively small for summer and autumn, which is due to high solar radiation value and ambient temperature condition in those seasons. Therefore, the collecting efficiencies in summer and autumn are generally smaller than those in spring and winter. The averaged collecting efficiency is calculated at 58% during the whole-year tests. Also it should be noted that the present experimental set-up is not optimized one and it is just based on the basic principle and designed for feasibility considerations. So the solar thermal conversion performance may be enhanced further in future studies, from both optimal system design and better operation strategies. In addition, from the comprehensive observation of Figs. 8e10, it can be seen that the solar thermal conversion performance will increase with C, however the measured data is transient, dailyaveraged or season-averaged. This revealed phenomenon indicates us this solar thermal conversion made by supercritical natural convection is more suitable to be used in the relatively cold or less radiation area. Indeed, Fig. 8 shows the transient behavior data of the collector and due to the sensitive properties of supercritical fluid across the collector, some points may go beyond the general region. However, the major data falls in the normal region. That figure shows the different behavior of efficiency to C parameter evolution trend compared with traditional systems. In Fig. 10, seasonal performances are shown and compared. The comparison in Fig. 10 indicates that relative better performances can also be found in spring and winter season for the current supercritical fluid based system. That is in accordance with the basic conclusion of maintained high performances of such systems in cold and low radiation areas/seasons as discussed in former sections. In these areas, this type of conversion and system are better than the solar collection by water, which generally has a higher efficiency in hot or more radiation area. More investigation and optimization are needed to well design this kind of solar thermal conversion using supercritical natural convection. 4. Concluding remarks In this paper, an experimental work is introduced and conducted to study the basic characteristics of solar thermal conversion made by supercritical CO2 natural convection. Based on the measured and arranged data in the transient, daily and seasonal patterns, the following concluding remarks are made: (1) A very short time drop in the solar radiation could not affect the supercritical CO2 flow rate, temperature and pressure. An obvious and longer time drop in the solar radiation may induce a small scale drop in the CO2 flow rate, temperature and pressure. A continuous long-time drop in the solar radiation will less affect the CO2 flow rate, temperature and pressure very much, if the solar radiation is in a relatively high-value level. This continuous long-time drop can induce obvious drops in the CO2 flow rate, temperature and pressure only if the solar radiation is in a low-value level. (2) Long-time drop and relative low-value in the solar radiation will lead to very low transient drop down and restart process in the system. This temporary flow rate drop phenomena should be paid more attention in the design and operation. (3) It is found that there exist optimal flow rate and CO2 charge amount, which own the highest collecting efficiency and the best solar thermal conversion performance. (4) The collecting efficiency is found to increase with the comprehensive coefficient. This phenomenon is contrary to that using water and compressibility of supercritical fluid may be one main mechanism to this difference. It is also found that
617
the solar thermal conversion supercritical natural convection has a higher efficiency in spring and winter than summer and autumn; better performance in cold and low-radiation area than in hot and high-radiation area is identified. Acknowledgment The support from the National Science Foundation of China (51276001) and Common Development Fund of Beijing are gratefully acknowledged. Nomenclature A C h hCO hCI I mCO2 QC Tf
Ta xi
hC
projected area of the collector (m2) comprehensive coefficient (K/W) enthalpy (J/kg) enthalpy values of the CO2 fluid at the outlet of the solar collector (J/kg) enthalpy values of the CO2 fluid at the inlet of the solar collector (J/kg) solar radiation (W/m2) mass flow rates of CO2 fluid (kg/s) collected heat quantity by the solar collector (W) fluid average temperature in the solar collector, which is calculated based on the inlet fluid temperature and outlet fluid temperature (K) ambient temperature (K) accuracy of every parameter (%) solar collector efficiency (%)
References [1] Davidson JH, Mantell SC, Jorgensen G. Status of the development of polymeric solar water heating system. Adv Sol Energy 2002;15:149e86. [2] Chen G, Dresselhaus MS, Fleurial JP, Caillat T. Recent developments in thermoelectric materials. Int Mater Rev 2003;48:45e66. [3] Greffet JJ, Carminati R, Joulain K, Mulet JP, Mainguy S, Chen Y. Coherent emission of light by thermal sources. Nature 2003;416:61e4. [4] Mancini T, Heller P, Butler B, Osborn B, Wolfgang S, Vernon G, et al. DishStirling systems: an overview of development and status. ASME J Sol Energy Eng 2003;125:135e51. [5] Kalogirou SA. Solar thermal collectors and applications. Prog Energy Combust Sci 2004;30:231e95. [6] Kumar A, Goyal AK, Sodha MS. Technoeconomic optimization of solar natural convection hybrid hot water systems. Int J Energy Res 1988;12(4):661e78. [7] Mishra CB, Bhat AK. Performance study of thermosyphonic circulation solar water heaters using packed bed collectors. Int J Energy Res 1982;6(4):351e5. [8] Budihardjo I, Morrison GL. Performance of water-in-glass evacuated tube solar water heaters. Sol Energy 2009;83:49e56. [9] Dubey S, Tiwari GN. Thermal modeling of a combined system of photovoltaic thermal (PV/T) solar water heater. Sol Energy 2008;82(7):602e12. [10] Anderson TN, Morrison GL. Effect of load pattern on solar-boosted heat pump water heater performance. Sol Energy 2007;81(11):1386e95. [11] Esen M, Esen H. Experimental investigation of a two-phase closed thermosyphon solar water heater. Sol Energy 2005;79(5):459e68. [12] Morrison GL, Budihardjo I, Behnia M. Measurement and simulation of flow rate in a water-in-glass evacuated tube solar water heater. Sol Energy 2005;78:257e67. [13] Morrison GL, Budihardjo I, Behnia M. Water-in-glass evacuated tube solar water heaters. Sol Energy 2004;76:135e40. [14] Dahl SD, Davidson JH. Performance and modeling of thermosyphon heat exchangers for solar water heaters. J Sol Energy Eng 1997;119(3):193e200. [15] Mason AA, Davidson JH. Measured performance and modeling of a evacuatedtube, integral-collector-storage solar water heater. ASME J Solar Energy Eng 1999;117:221e8. [16] Rosengarten G, Morrison G, Behnia M. A second law approach to characterizing thermally stratified hot water storage with application to solar water heaters. J Sol Energy Eng 1999;121(4):194e200. [17] Radhwan AM, Zaki GM, Jamil A. Refrigerant-charged integrated solar water heater. Int J Energy Res 1990;14(4):421e32. [18] Alkhamis AI, Sherif SA. Feasibility study of a solar-assisted heating/cooling system for an aquatic centre in hot and humid climate. Int J Energy Res 1997;21(9):823e39.
618
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
[19] Yamaguchi H, Sawada N, Suzuki H, Ueda H, Zhang XR. Preliminary study on a solar water heater using supercritical carbon dioxide as working fluid. ASME J Solar Energy Eng 2010;132:101e6. [20] Zhang XR. A preliminary experimental investigation on characteristics of natural convection based on solar thermal collection using supercritical carbon dioxide. Int J Energy Res 2013;37:1349e60. [21] Zhang XR, Chen L, Yamaguchi H. Natural convective flow and heat transfer of supercritical CO2 in a rectangular circulation loop. Int J Heat Mass Tran 2010;53:4112e22. [22] Chen L, Zhang XR, Yamaguchi H, (Simon) Liu ZS. Effect of heat transfer on the instabilities and transitions of supercritical CO2 flow in a natural circulation loop. Int J Heat Mass Tran 2010;53:4101e11. [23] Chen L, Zhang XR, Cao S, Bai H. Study of trans-critical CO2 natural convection flow with unsteady heat input and its implications on system control. Int J Heat Mass Tran 2012;55:7119e32. [24] Zhang XR, Yamaguchi H, Fujima K, Enomoto M, Sawada N. A feasibility study of CO2 e based Rankine cycle powered by solar energy. JSME Int J Ser B Fluids Therm Eng 2005;48:540e7.
[25] Zhang XR, Yamaguchi H, Uneno D, Fujima K, Enomoto M, Sawada N. Analysis of a novel solar energy powered Rankine cycle for combined power and heat generation using supercritical carbon dioxide. Renew Energy 2006;31:1839e54. [26] Zhang XR, Yamaguchi H, Uneno D. Experimental study on the performance of solar Rankine system using supercritical CO2. Renew Energy 2007;32:2617e28. [27] Zhang XR, Yamaguchi H. An experimental study on evacuated tube solar collector using supercritical CO2. Appl Therm Eng 2008;28:1225e33. [28] Zhang XR, Yamaguchi H. Forced convection heat transfer of supercritical carbon dioxide in a horizontal circular tube. J Supercrit Fluids 2007;41:412e20. [29] Chen L, Zhang XR. Simulation of heat transfer and system behavior in a supercritical CO2 based thermosyphon: effect of pipe diameter. ASME J Heat Trans 2011;133:2505e13. [30] Chen L, Deng BL, Zhang XR. Experimental study of trans-critical and supercritical CO2 natural circulation flow in a closed loop. Appl Therm Eng 2013;59: 1e13. [31] Chen L, Deng BL, Zhang XR. Experimental investigation of CO2 thermosyphon flow and heat transfer in the supercritical region. Int J Heat Mass Tran 2013;64:202e11.
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
Experimental study on solar thermal conversion based on supercritical natural convection Xin-Rong Zhang a, b, *, Yalong Zhang a, Lin Chen a a b
Department of Energy and Resources Engineering, College of Engineering, Peking University, Beijing 100871, China Beijing Key Laboratory for Solid Waste Utilization and Management, Peking University, Beijing 100871, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 September 2012 Accepted 19 August 2013 Available online 14 September 2013
In this paper, experimental investigation into the basic characteristics of solar thermal conversion using supercritical CO2 natural convection are presented. Natural circulation of supercritical fluids can be easily induced and even a small change in temperature can result in large change in density close to the critical point. The supercritical experimental system carefully designed and operated in this study. It is found that an obvious and continuous long-time drop of solar radiation would not affect the CO2 flow rate, temperature and pressure very much, if the solar radiation is in a relatively high-value level. This continuous drop can induce obvious drops in the CO2 flow rate, temperature and pressure only when the solar radiation is in a low-value level. Furthermore, it is observed that a long-time drop and low-value in the solar radiation may make the flow rate temporarily become zero, which should be paid more attention in future system design and operation. The collecting efficiency increases with the comprehensive coefficient and this pattern is contrary to that of water based system. In addition, it is found that there exist optimal flow rate and CO2 charge amount for system overall performance. This kind of solar thermal conversion has a higher collecting efficiency in spring and winter than summer and autumn; a better performance in cold and low-radiation region than hot and high-radiation region. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Supercritical fluid Natural convection Solar thermal energy Carbon dioxide
1. Introduction Among various solar utilization methods, solar thermal conversion has become one of the most promising kinds. Solar thermal conversion systems can be categorized into low-temperature solar thermal kinds, which usually do not use sunlight concentration, and medium-temperature and high-temperature solar thermal systems. Among those methods, low-temperature solar thermal systems have the potential to supply a significant number of households and commercial buildings with heating and/or cooling. Solar water heater is one of the most promising applications of lowtemperature solar thermal systems [1e7]. In recent years, a lot of studies have been carried out in the field of solar water heater and low temperature solar thermal convection [8e16], which are mainly for the solar water heaters of water-in-glass evacuated tube type [17,18].
* Corresponding author. Department of Energy and Resources Engineering, College of Engineering, Peking University, Beijing 100871, China. Tel.: þ86 10 82529066; fax: þ86 10 82529010. E-mail addresses: [email protected], [email protected] (X.-R. Zhang). 0960-1481/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.renene.2013.08.025
In traditional solar water heaters, both forced convection and natural convection are used. In natural-convective solar heaters, water is usually used as working fluid. Many previous studies were carried out to explore the flow dynamics, heat transfer and performances for natural circulation kind of solar water heaters [5]. From those studies, the natural-convective solar water heater owns a deficient characteristic, because water is an incompressible fluid and its natural convection is relatively weak under the solar heating condition. As an effort of improving the efficiency from solar to thermal energy, supercritical fluid was proposed as working fluid in solar collectors [19]. Supercritical fluid is a substance at a temperature and pressure above its thermodynamic critical point. Fig. 1 shows the thermo-physical properties of supercritical CO2. As shown in Fig. 1, the most important is that, close to the critical point, small changes in pressure or temperature can result in large changes in thermo-physical properties. The data shown in Fig. 1 is obtained from a Program Package for Thermo-physical Properties of Fluids (NIST Fluid Thermodynamic and Transport Properties database, version 8.0). Especially the density of supercritical fluid varies significantly as a function of temperature near the critical temperature. Such characteristics in thermo-physical properties allow a strong natural convection flow to be developed in the solar
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
611
Fig. 1. Variation of thermoproperties with temperature and pressure in the critical region, (a) specific heat; (b) density; (c) thermal conductivity; (d) viscosity.
thermal conversion system [19]. Furthermore, for supercritical CO2, its critical temperature is 31.1 C and that is low enough to be easily reached in the low temperature solar thermal conversion system. Solar water heater is the most widely used conversion unit around the world. Based on previous basic supercritical fluid tests [19e22], the solar water heater using supercritical fluid can be comprised of two loops: one is a supercritical fluid circulation loop and the other is a water loop. The two loops are coupled with each other by a heat exchanger. In the closed loop with supercritical fluid, the natural convective flow develops as a result of the heating process by solar radiation and the cooling process by the heat exchanging process with water. The supercritical fluid flow in the collector can absorb and transport heat, and then it transport the collected thermal energy to water. Therefore conversion from solar to thermal energy can be achieved. Inside the heat exchanger, supercritical fluid is cooled by the cold water flow (in real application, general tap water can be used). Water absorbs heat from the supercritical fluid and can be pipelined for related domestic or other costumer sides. A feasibility study was carried out in order to analyze whether the supercritical state can be achieved at the collector outlet and the Reynolds number was measured higher than 1900 in the previous experimental and numerical investigations [19e22]. Furthermore, some characteristics about this kind of solar water heater by supercritical CO2 were investigated by several experimental tests [20]. The obtained results revealed the supercritical CO2 flow rate is quite stable and will be less affected by the transient variations of the solar radiation. The solar thermal conversion process can be divided into three periods: starting-up, transition and stable period. High solar thermal conversion efficiency was found at a high mass flow rate and under operation pressure near to the critical point. However, in the field of supercritical fluid based solar thermal conversion systems, some basic characteristics of such natural
circulation based solar thermal conversion system design and effective operation are still not clear, such as the influence of solar radiation, supercritical fluid temperature and ambient temperature etc. on solar thermal conversion process and efficiency. Also the parameter description and seasonal performances of supercritical systems are still under development. As an extension of the previous study, in this paper, an experimental set-up is tested and systematic investigations are made to further clarify the basic characteristics of this solar thermal collection by supercritical natural convection flow. 2. Experimental set-ups Based on the concept described above, an experimental prototype of the solar thermal conversion system is tested. This experimental system is specifically designed for using supercritical CO2 natural circulation flow. A schematic diagram of the experimental set-up is presented in Fig. 3. The built experimental machine is mainly comprised of solar thermal collector, heat exchanger, valve 1, valve 2, hot water tank, valve 3, and measurement and data acquisition system. To make sure the solar water heater system is feasible, a solar collector with good heat collecting characteristics is required. Therefore, to effectively heat supercritical CO2 to relatively high temperature state in the experimental set-up, all-glass evacuated solar collector with a U-tube heat removal system. The collector is consisted of a glass envelope over glass tube that is coated with a selective solar absorber coating. The coating has a high solar absorbance of 0.918 and a low emissivity of 0.189. It is applied on the vacuum side of the inner glass tube. The transparence of the glass envelope is 0.92. Such design of the collector allows a maximum operating pressure as high as 15.0 MPa. In the present experimental set-up, evacuated solar collector of 1.69 m2 (gross area) is used which is manufactured basically based on the previous
612
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
ones using the supercritical CO2 forced convection [21e23]. The every U-shaped heat removal fluid tube utilized is 3.5 m long and 0.006 m internal diameter. Related system design and tests experiences in supercritical CO2 fluid systems [24e31] should be referred to as they may provide both experimental and theoretical information for the convenience of novel system development and manufacturing. In the supercritical fluid circulation loop, Valve 1 is operated by hand. Valve 1 can be used to adjust flow rate of the CO2 fluid. A concentric tube heat exchanger is utilized to couple the supercritical fluid circulation loop with the water loop. Supercritical fluid flows inside the inner tube of the heat exchanger and then solar heat is transferred through the wall of a stainless steel to water side. The concentric heat exchanger is a type of counter-flow circular pipe geometry. In this heat exchanger, the outer tube has an internal diameter of 25.0 mm and wall thickness of 1.2 mm and the inner tube is made of stainless steel, having an internal diameter and wall thickness of 6.0 mm and 1.0 mm, respectively. Heat exchanging area of the heat exchanger is 1.0 m2. A mass flow meter was used in this fluid loop to measure the mass flow of liquid CO2. This flow meter also has a maximum permissible operating pressure of 15.0 MPa. The flow meter was installed in the downstream of the heat exchanger in the supercritical fluid loop, as shown in Fig. 2. It provides a measurement range of 0.02e5.0 kg/min with an accuracy of 0.1%. It should be mentioned here that the supercritical fluid loop in the solar thermal conversion system is not symmetric, which can avoid an oscillation flow, based on our previous numerical studies [21e23,29]. As shown in Fig. 2, a hand-operated valve, valve 2, is installed in the water loop side. This valve is adjusted based on the water temperature which is measured at the outlet of the heat exchanger. In the water loop side, cool water flows into the heat exchanger and is heated by the supercritical CO2 fluid flow. Then hot water comes out of the heat exchanger. Hot water from the outlet of the heat exchanger flows into a hot water tank, as shown in Fig. 2. In the present experimental set-up, the hot water tank is made of stainless steel and its volume is 80 L. At the outlet of the hot water tank, there is a valve e valve 3, which can be used to collect hot water from the hot water tank. The valve 3 is also a hand-operated valve. In the measurement system, two T-type thermocouples and two pressure transmitters are respectively mounted at the inlet and outlet of the solar collector to measure CO2 temperatures and
pressures, with an accuracy of 0.1 C for temperature and 0.2% for pressure measurements. In addition, two platinum resistor temperature sensors were mounted to measure the inlet and outlet water temperatures of the heat exchanger. Its accuracy can be expressed as 0.15 þ 0.0002jtj C. Thermal insulation coating applied on all the CO2 and water loops to reduce heat losses from the piping. A meteorological measurement system was also installed. It is mainly comprised of sun radiation sensor, anemometer, and air temperature gauge and so on. In the present study, only measured solar radiation data is presented, and accuracy of the sun radiation sensor is 0.3%. This pyranometer is based on a thermopile sensor. Typically, the pyranometer output signal does not exceed 20 mV. The typical sensitivity is 10 mV/(W/m2). So the accuracy of the sun radiation sensor is less than 0.3%. In addition, a data acquisition system is used in the experimental set-up, which can be used to achieve real-time data measurement, acquisition, processing, and share. For each experiment, the solar collector is adjusted toward the direction of the Sun. Cool water flow is first pumped into the heat exchanger. Then according to the measured CO2 temperature, valve 1 is adjusted in order to achieve the expected extent. During the experiment, the two major sources of experimental error are: (1) temperature, pressure and flow rate measurement accuracy; and (2) errors resulting from data logging and reading by the computer. The error of data logging and reading are analyzed and listed as follows: 0.1 C for CO2 temperature; 0.2% for CO2 pressure; 0.2% for CO2 mass flow rate; 0.2 C for water temperature; 0.2% for water flow rate; 0.2% for solar radiation. Based on the above accuracies, accuracies of all the parameters defined in this paper are calculated to be less than 2.0% using the following equation:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Pn 2 x21 þ x22 þ . þ x2n i ¼ 1 xi ¼ x ¼ n n
(1)
Where x is the average of the parameter x; xi is accuracy of every parameter involved in the calculation equation of x; and n is the number of the parameters in the calculation equation of x.
3. Results and discussion For first stage test, the solar collector is set declined, with an angle of 45 C with respect to the horizontal level in the present
Fig. 2. Flow diagram of the experimental set-up.
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
613
study, which indicates that the U-tubes in the collector has an angle of 45 from the horizontal level. The water inlet temperature is 20 C during all the experimental tests. The CO2 charge amount is 4.4 kg at most tests, except special mention. The following items are mainly measured: the solar radiation, the CO2 fluid temperatures at the inlet and outlet of the heat exchanger, the CO2 fluid pressures at the inlet and outlet of the solar collector, and the mass flow rates of CO2 fluid. Based on these measured quantities, the following parameters are defined to further learn the collecting efficiency, which evaluates the solar thermal energy conversion performance:
C ¼
Tf Ta
I
QC ¼ mCO2 ðhCO hCI Þ
hC ¼
QC IA
(2) (3) (4)
In this paper, CO2 enthalpy values at the different monitoring points were calculated based on the measured temperature and pressure values using a program package for thermo-physical properties of fluids: NIST Fluid Properties database (REFPROP 8.0). The experimental set-up was tested throughout the year of 2011 in the city of Shaoxing, Zhejiang Province of China. As the initial step of investigation into the fundamental characteristics of solar thermal conversion by supercritical CO2 natural convection, it should be noted that optimization structure and operation are not claimed in the present study. In the experimental test in the present study, the opening of valve 1 was adjusted according to the CO2 temperature at the outlet of the collector.
Fig. 3. Variations of the measured parameters with the test time on July 15, 2011. (a) Solar radiation and CO2 pressures; (b) CO2 temperatures and flow rate.
3.1. Transient characteristics of the solar thermal conversion system Based on the operation and measurement principles described above, the first focus point in this study is the transient characteristics of this kind of solar thermal conversion by supercritical natural circulation flow. Understanding transient characteristics is important to build up an optimal and high efficient energy conversion. The best way to obtain transient characteristics is to directly measure the variations of the operating parameters with time. Fig. 3(a) shows the variations of the measured solar radiation and CO2 pressures with the test time on July 15, 2011. Generally speaking, solar radiation in July in Shaoxing area indeed is representative for southeast part of China throughout a year. It is seen from this figure that this day is not sunny all the time and also not cloudy all the day, which represents a typical weather in Shaoxing area of Zhejiang Province. Sometimes in the day is sunny and sometimes is cloudy and its time-averaged solar radiation value is found at 610.0 W/m2, which also represents basic weather conditions seen in this area. From this figure, it is also observed that there is a sudden and obvious drop in the solar radiation around 12:30 and this solar radiation drop lasts shortly, which is marked by one circle in Fig. 3. In addition, a longer time period of obvious solar radiation drop is observed around 10:40, which lasts about 30 min. It is also seen that there are two obvious and continuous drops in the solar radiation, which respectively occur from 13:15 to 14:30 and from 15:00 to 16:30. These two continuous drops last about 75 min and 90 min, respectively. The measured variations of the mass flow rate and CO2 temperatures are presented in Fig. 3(b). From the curves shown in Fig. 3, it is seen that although the solar radiation drops slightly around 12:30, no obvious variation in CO2 temperature, pressures or CO2 mass flow rate happen. That means that the CO2
temperature, pressure and mass flow rate are less affected by a sudden and short-time solar radiation drop. The phenomenon was also partially observed in the previous study [20]. However, if such radiation drop lasts a relative longer time, as observed around 10:40, small drops in CO2 pressure, CO2 temperature at the collector outlet and CO2 mass flow rate are also seen near this time point. But the observed drops in the CO2 pressure, temperature and flow rate are later than the solar radiation drop. The phenomenon can be attributed to this solar thermal conversion being a thermal inertia system. This thermal inertia characteristic in supercritical fluid leads to the variations of the CO2 temperature, pressure and flow rate etc. are after the solar radiation heating. It can be concluded that the CO2 temperature, pressure and flow rate can be influenced by a longer time drop in solar radiation, although the observed drops in the CO2 temperature, pressure and flow rate are small and limited. But if we carefully observe that obvious and continuous drop in solar radiation occurring from 13:15 to 14:30, it is found that the CO2 temperature and flow rate only drop with a small amplitude although the solar radiation drop keeps a long time period almost 75 min. It is also seen no obvious drop in the CO2 pressure is observed due to this long time period drop in the solar radiation, and only two very small drop curves are observed. From Fig. 3, it is observed there are obvious variations in CO2 temperature, pressure and flow rate when a longer time drop in the solar radiation occurs around 15:00e16:30. That is different from the above observation. Although a continuous drop in the solar radiation also occurs for a long time period of 13:15e14:30, no obvious drops are found either in the CO2 temperature, flow rate or the CO2 pressure. The physical mechanism of that difference can be explained as follows: the continuous drop during 13:15e14:30 is in
614
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
a time period with relatively high level of the solar radiation. However, another continuous drop of 15:00e16:30 is close to the end period of the solar radiation heating during that day, and its solar radiation is in a low value level. The time-averaged solar radiation value during 13:15e14:30 is calculated at 550.0 W/m2, and the averaged value during 15:00e16:30 is found at 200.0 W/m2. The long-time and low solar radiation heating during 15:00e16:30 makes this thermal inertia system difficult to maintain its original flow rate, and the temperature and pressure would decrease. Therefore, obvious drops in the CO2 flow rate, temperatures and pressures occur due to this long and low solar radiation heating. Although a long time drop in the solar radiation also occurs from 13:15 to 14:30, the collected heat intensity is high enough to well support the CO2 flow, temperature and pressure. The low viscosity and high circulation rate of supercritical natural loops also contribute to this finding [29e31]. Therefore no obvious drops are observed during this experiment operation. In addition, it is also seen from Fig. 3(b) that during the continuous drop in the solar radiation from 15:00 to 16:30, the CO2 mass flow rate drops to very low values for three times. This observed phenomenon is much different from those obtained from the previous studies [19,20], where the supercritical CO2 flow rate is stable and is less affected by the sudden variation of the solar radiation. When the system reaches stable operation, the CO2 flow rate will keep at a high value even the solar radiation stays at a low level [20]. The stable period is defined as the time period starting from the first peak of the CO2 flow rate. Indeed no continuous solar radiation drop occurs in this period, as observed in Fig. 3, and also the solar radiation is in a low-value level. This long-time drop and relative low-value in the solar radiation lead to transient very low circulation rate in the system. The low-standard solar radiation heating is insufficient to sustain this supercritical natural convection flow. This kind of temporary very low flow rate phenomena is interesting and should be paid more attention in future engineering design and operation. At the same time, it is observed during the continuous drop in the solar radiation from 15:00 to 16:30, after the flow stops and then it can restart. After the transient flow rate drops, flow restarting process happens and such kind of oscillating flow situation are seen when the solar radiation values drops gradually in the late afternoon. When the flow restarts, the mass flow rate is seen to be at relative high standard or even high peaks. Three peaks in the CO2 mass flow rate are seen during this period. The phenomenon of reaching a peak value when the flow is started was also observed and verified in the theoretical analyses [21e23]. Here, it can be simply concluded that a very short time drop in the solar radiation will less affect the CO2 fluid flow rate, temperature and pressure. An obvious and longer time drop in the solar radiation may induce a small scale drop in the CO2 flow rate, temperature and pressure. A continuous long-time drop in the solar radiation would not affect the CO2 flow rate, temperature and pressure much, when the solar radiation is in relatively high-level. The continuous long-time drop in the solar radiation can induce obvious drops in the CO2 flow rate, temperature and pressure only if the solar radiation is in a low-value level, even the flow rate to near-zero values. Furthermore, it is seen that among the CO2 flow rate, temperature and pressure, the most affected parameter by obvious solar radiation variation is the CO2 flow rate, then the CO2 temperature and the minimum is the CO2 pressure. The physical mechanism is due to the sensitivity of supercritical fluids in response to the variations of solar radiation. After that, the basic viscosity drops (as show in Fig. 1) and the fluid flow will be susceptible to instabilities when high flow rate can generally be found in the system. Also, an obvious variation in the solar radiation heating directly affects the driven force and then the flow rate changes. The flow rate change
affects the heat transfer between the CO2 fluid and the collector and then the CO2 temperature vary. The evolution of heat collected from the water side and the transient collector efficiency as defined in Eq. (4) are plotted in Fig. 4 against test time. Also plotted is the evolution of solar radiation on the collector area (1.5 m2) in the same summer day (July 15, 2011). With the increase of solar radiation, the collected heat curve shows similar trend after supercritical CO2 flow started. It can be seen in Fig. 4 that the collector efficiency shows relative stable value around 60%, which indicates the relative stable behavior of the current test system from around 10:00 to 16:00. The detailed seasonal behaviors and efficiency will be discussed later in this paper. In addition, with the fluctuations of solar radiation as shown in Fig. 4, respective fluctuations of collected heat and efficiency are seen, which may due to a sensitive response of the system to the changes of flow rate (see Fig. 3(b)) and temperature curve (Fig. 3(a) and (b)). The collected heat is calculated according to the definition in Eq. (3) which can reflect the transient characteristics directly. First, the measured solar radiation and collected heat with the test time is presented in Fig. 3. The measurement is conducted on July 8, 2011. It is seen that the measurement starts from about 9:30 in the morning to 16:30 in the afternoon. 3.2. Time-averaged daily performance of solar thermal conversion system The collecting efficiency is the most important indicator of this solar thermal conversion performance. The main influencing factors to this solar thermal conversion system are solar radiation, ambient temperature, CO2 temperature, CO2 pressure and CO2 flow rate. The comprehensive coefficient C defined in the Eq. (2) is usually used to consider these influencing factors, which include the solar radiation, CO2 temperature in the collector and ambient temperature. In this session, this coefficient C will be utilized to show the influence of the factors on the solar thermal conversion. In addition, because the CO2 flow rate is an important design and operation parameter, its effect will also be focused on in this session. In order to figure out how the CO2 flow temperature affects the collecting efficiency, Fig. 5 shows the variations of the collecting efficiency with the measured CO2 temperature difference between the collector outlet and inlet. A wide range of the CO2 temperature differences from 2 C to 50 C are shown in Fig. 5, in that figure each point represents a time-averaged value of a day. It is seen from
Fig. 4. Variations of the measured collected heat, solar radiation on the collector area (1.5 m2) and collecting efficiency against test time on July 15, 2011.
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
Fig. 5. Variations of the solar thermal conversion efficiency with the collector temperature difference between outlet and inlet. Every point in this figure represents a time-averaged value per day.
Fig. 5 that the collecting efficiency increases with this CO2 temperature difference. When this CO2 temperature difference between the collector outlet and inlet is small, the collecting efficiency is found small. The collecting efficiency becomes large when the CO2 temperature difference is large. Based on Fig. 5, it is also found the collecting efficiency is at 0.18 when this CO2 temperature difference is 2 C; the collecting efficiency is 0.60 when the temperature difference increases to 27 C; the collecting efficiency increases to 0.83 when the temperature difference 53 C. The averaged collecting efficiency is calculated at 0.62 within the entire tests shown in Fig. 5. A physical explanation is also presented here for the phenomenon observed in Fig. 5. When the CO2 temperature difference increases, the more heat can be collected in the collector, according to the Eq. (3). The more collected heat quantity means increasing the collecting efficiency. Even the solar radiation level is the same, this more collected heat can be achieved by the variations in other parameters, such as ambient temperature. Therefore, the CO2 fluid temperature difference between the collector outlet and inlet should be increased as possible as we can in order to enhance the solar thermal conversion. Based on the previous test [20], the solar thermal efficiency is found to be increased when CO2 mass flow rate increases. Therefore efforts should be made in order to increase the CO2 mass flow rate so as to improve the system efficiency. Fig. 6 represents such an effort, where the measured variation in the CO2 mass flow rate with the solar radiation is presented. In this figure, every data point also represents a time-averaged value per day. Based on these measured daily-averaged values, it is found the CO2 mass flow rate increases
Fig. 6. Variations of the measured mass flow rate with the solar radiation. Every point in this figure represents a time-averaged value per day.
615
with the solar radiation during the entire tests. A good solar radiation heating provides a strong driven force for this supercritical natural convection flow. Then more high flow rate can be made corresponding to this good solar radiation. So, generally speaking, a good weather and high solar radiation are helpful to increase the collecting efficiency and to make an efficient solar thermal conversion. The collecting efficiency evolution trend for major representative experimental tests is shown in Fig. 7. It is found the collecting efficiency will decrease with the increase in the CO2 mass flow rate after this flow rate reaches to some value. This observed phenomenon reveals the previous test and observations are limited in the range. Indeed there exists an optimal CO2 mass flow rate, which is corresponding to the highest collecting efficiency. When the mass flow rate is less this optimal value, the collecting efficiency increases with the flow rate, because a quick flow can enhance the heat transfer and then the collecting efficiency. When the mass flow rate becomes larger than this optimal value, the collecting efficiency decreases with the flow rate increase, because when the flow becomes too quick, the flow cannot carry and transport the thermal energy out of the collector timely. Therefore, the collecting efficiency decreases when the mass flow rate is too high. Fig. 8 shows the evolutions of the collecting efficiency with comprehensive coefficient on August 5, August 19 and August 25, 2011, respectively. It is seen that the collecting efficiency is proportional to the comprehensive coefficient on August 25, 2011. For the measured data on August 5 and 19, though the collecting efficiency is not linearly proportional to the comprehensive coefficient, the general trend shows that the collecting efficiency will increase with the comprehensive coefficient. It is seen that the collecting efficiency points are scattered on both sides of the proportional line. This is much different from the case of water based solar collector. For water, the collecting efficiency is inversely proportional to the comprehensive coefficient, because an increase in the solar radiation or a decrease in the ambient temperature usually brings more heat loss from the solar collector. What physical mechanism leads to this difference found above? The physical explanation may be presented as followings: supercritical CO2 is one very much compressible fluid, and in the solar collector, the mass flow rate varies all the time, which is also different from the water case. For the water collector, the flow rate changes a little and that only depends on the pump power input. However, the flow rate of the supercritical fluid is affected by many factors, such as the solar radiation and ambient temperature etc. This compressibility of the supercritical fluid makes the relationship between the collecting efficiency and the comprehensive
Fig. 7. Variations of the solar thermal conversion efficiency with the measured CO2 mass flow rate. Every point in this figure represents a time-averaged value per day.
616
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
Fig. 9. Variations of the solar thermal conversion efficiency with the comprehensive coefficient under the CO2 charge amounts. Every point in this figure represents a timeaveraged value per day.
flow rate pattern observed in Fig. 6. There also exists an optimal CO2 charge amount, which owns the highest collecting efficiency and the best solar thermal conversion performance. The physical mechanism can be understood that a large CO2 charge amount would produce too quick flow to transport heat from the collector to the ambient. In addition, it is shown in Fig. 9 that the slope decreases when the CO2 charge amount increases, when this amount is less than 4.5 kg. As discussed, there is an optimal CO2 charge amount for this system, whose solar thermal conversion efficiency is the highest. This optimal amount is also considered as a ‘critical point’ or ‘saturation point’ to judge how much CO2 fluid should be charged into the system. When the charge amount is far from this ‘saturation point’, the line slope is larger, which means the collecting efficiency varies faster with C. Closer to the ‘saturation point’ the charge amount is, smaller slope the collecting efficiency has and slower variation will be seen between the collecting efficiency and C. 3.3. Seasonal performance of solar thermal conversion system
Fig. 8. Variations of the collecting efficiency with the comprehensive coefficient. (a) on August 25, 2011; (b) on August 5, 2011; (c) on August 19, 2011.
The measured seasonal variations of the collecting efficiency with C in four seasons are plotted in Fig. 10. It is found that generally the collecting efficiency will increase with C in four seasons. Furthermore, it is seen that the comprehensive coefficient C is
coefficient is contrary to the water case. This relationship observed in Fig. 8 should be pay attentions in the engineering design and operation. Furthermore, the measured variations of the collecting efficiency with C under different CO2 charge amounts are presented in Fig. 9. In this figure, four CO2 charge amounts are included and they are respectively 3.5 kg, 4.1 kg, 4.5 kg and 4.8 kg. For each CO2 charge amount, ten times tests are made. Every point in Fig. 9 represents a time-averaged value per day. Same with the pattern observed in Fig. 8, it is seen from Fig. 9 that the time-averaged collecting efficiency is also proportional to the comprehensive coefficient, under each CO2 charge amount. However, when the CO2 charge amount increases from 4.5 kg to 4.8 kg, a decrease in the collecting efficiency is observed. When the CO2 charge amount is less than 4.5 kg for the present system, the collecting efficiency will increase with the CO2 charge amount. Nevertheless, when the charge amount is larger than 4.5 kg, the collecting efficiency will decrease with the increase of CO2 charge amount. This phenomenon is similar to the
Fig. 10. Variations of the collecting efficiency with the comprehensive coefficient under four different seasons. Every point in this figure represents a time-averaged value per day.
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
generally larger in spring and winter than those in summer and autumn. Also, it should be noted that parameter C is relatively small for summer and autumn, which is due to high solar radiation value and ambient temperature condition in those seasons. Therefore, the collecting efficiencies in summer and autumn are generally smaller than those in spring and winter. The averaged collecting efficiency is calculated at 58% during the whole-year tests. Also it should be noted that the present experimental set-up is not optimized one and it is just based on the basic principle and designed for feasibility considerations. So the solar thermal conversion performance may be enhanced further in future studies, from both optimal system design and better operation strategies. In addition, from the comprehensive observation of Figs. 8e10, it can be seen that the solar thermal conversion performance will increase with C, however the measured data is transient, dailyaveraged or season-averaged. This revealed phenomenon indicates us this solar thermal conversion made by supercritical natural convection is more suitable to be used in the relatively cold or less radiation area. Indeed, Fig. 8 shows the transient behavior data of the collector and due to the sensitive properties of supercritical fluid across the collector, some points may go beyond the general region. However, the major data falls in the normal region. That figure shows the different behavior of efficiency to C parameter evolution trend compared with traditional systems. In Fig. 10, seasonal performances are shown and compared. The comparison in Fig. 10 indicates that relative better performances can also be found in spring and winter season for the current supercritical fluid based system. That is in accordance with the basic conclusion of maintained high performances of such systems in cold and low radiation areas/seasons as discussed in former sections. In these areas, this type of conversion and system are better than the solar collection by water, which generally has a higher efficiency in hot or more radiation area. More investigation and optimization are needed to well design this kind of solar thermal conversion using supercritical natural convection. 4. Concluding remarks In this paper, an experimental work is introduced and conducted to study the basic characteristics of solar thermal conversion made by supercritical CO2 natural convection. Based on the measured and arranged data in the transient, daily and seasonal patterns, the following concluding remarks are made: (1) A very short time drop in the solar radiation could not affect the supercritical CO2 flow rate, temperature and pressure. An obvious and longer time drop in the solar radiation may induce a small scale drop in the CO2 flow rate, temperature and pressure. A continuous long-time drop in the solar radiation will less affect the CO2 flow rate, temperature and pressure very much, if the solar radiation is in a relatively high-value level. This continuous long-time drop can induce obvious drops in the CO2 flow rate, temperature and pressure only if the solar radiation is in a low-value level. (2) Long-time drop and relative low-value in the solar radiation will lead to very low transient drop down and restart process in the system. This temporary flow rate drop phenomena should be paid more attention in the design and operation. (3) It is found that there exist optimal flow rate and CO2 charge amount, which own the highest collecting efficiency and the best solar thermal conversion performance. (4) The collecting efficiency is found to increase with the comprehensive coefficient. This phenomenon is contrary to that using water and compressibility of supercritical fluid may be one main mechanism to this difference. It is also found that
617
the solar thermal conversion supercritical natural convection has a higher efficiency in spring and winter than summer and autumn; better performance in cold and low-radiation area than in hot and high-radiation area is identified. Acknowledgment The support from the National Science Foundation of China (51276001) and Common Development Fund of Beijing are gratefully acknowledged. Nomenclature A C h hCO hCI I mCO2 QC Tf
Ta xi
hC
projected area of the collector (m2) comprehensive coefficient (K/W) enthalpy (J/kg) enthalpy values of the CO2 fluid at the outlet of the solar collector (J/kg) enthalpy values of the CO2 fluid at the inlet of the solar collector (J/kg) solar radiation (W/m2) mass flow rates of CO2 fluid (kg/s) collected heat quantity by the solar collector (W) fluid average temperature in the solar collector, which is calculated based on the inlet fluid temperature and outlet fluid temperature (K) ambient temperature (K) accuracy of every parameter (%) solar collector efficiency (%)
References [1] Davidson JH, Mantell SC, Jorgensen G. Status of the development of polymeric solar water heating system. Adv Sol Energy 2002;15:149e86. [2] Chen G, Dresselhaus MS, Fleurial JP, Caillat T. Recent developments in thermoelectric materials. Int Mater Rev 2003;48:45e66. [3] Greffet JJ, Carminati R, Joulain K, Mulet JP, Mainguy S, Chen Y. Coherent emission of light by thermal sources. Nature 2003;416:61e4. [4] Mancini T, Heller P, Butler B, Osborn B, Wolfgang S, Vernon G, et al. DishStirling systems: an overview of development and status. ASME J Sol Energy Eng 2003;125:135e51. [5] Kalogirou SA. Solar thermal collectors and applications. Prog Energy Combust Sci 2004;30:231e95. [6] Kumar A, Goyal AK, Sodha MS. Technoeconomic optimization of solar natural convection hybrid hot water systems. Int J Energy Res 1988;12(4):661e78. [7] Mishra CB, Bhat AK. Performance study of thermosyphonic circulation solar water heaters using packed bed collectors. Int J Energy Res 1982;6(4):351e5. [8] Budihardjo I, Morrison GL. Performance of water-in-glass evacuated tube solar water heaters. Sol Energy 2009;83:49e56. [9] Dubey S, Tiwari GN. Thermal modeling of a combined system of photovoltaic thermal (PV/T) solar water heater. Sol Energy 2008;82(7):602e12. [10] Anderson TN, Morrison GL. Effect of load pattern on solar-boosted heat pump water heater performance. Sol Energy 2007;81(11):1386e95. [11] Esen M, Esen H. Experimental investigation of a two-phase closed thermosyphon solar water heater. Sol Energy 2005;79(5):459e68. [12] Morrison GL, Budihardjo I, Behnia M. Measurement and simulation of flow rate in a water-in-glass evacuated tube solar water heater. Sol Energy 2005;78:257e67. [13] Morrison GL, Budihardjo I, Behnia M. Water-in-glass evacuated tube solar water heaters. Sol Energy 2004;76:135e40. [14] Dahl SD, Davidson JH. Performance and modeling of thermosyphon heat exchangers for solar water heaters. J Sol Energy Eng 1997;119(3):193e200. [15] Mason AA, Davidson JH. Measured performance and modeling of a evacuatedtube, integral-collector-storage solar water heater. ASME J Solar Energy Eng 1999;117:221e8. [16] Rosengarten G, Morrison G, Behnia M. A second law approach to characterizing thermally stratified hot water storage with application to solar water heaters. J Sol Energy Eng 1999;121(4):194e200. [17] Radhwan AM, Zaki GM, Jamil A. Refrigerant-charged integrated solar water heater. Int J Energy Res 1990;14(4):421e32. [18] Alkhamis AI, Sherif SA. Feasibility study of a solar-assisted heating/cooling system for an aquatic centre in hot and humid climate. Int J Energy Res 1997;21(9):823e39.
618
X.-R. Zhang et al. / Renewable Energy 62 (2014) 610e618
[19] Yamaguchi H, Sawada N, Suzuki H, Ueda H, Zhang XR. Preliminary study on a solar water heater using supercritical carbon dioxide as working fluid. ASME J Solar Energy Eng 2010;132:101e6. [20] Zhang XR. A preliminary experimental investigation on characteristics of natural convection based on solar thermal collection using supercritical carbon dioxide. Int J Energy Res 2013;37:1349e60. [21] Zhang XR, Chen L, Yamaguchi H. Natural convective flow and heat transfer of supercritical CO2 in a rectangular circulation loop. Int J Heat Mass Tran 2010;53:4112e22. [22] Chen L, Zhang XR, Yamaguchi H, (Simon) Liu ZS. Effect of heat transfer on the instabilities and transitions of supercritical CO2 flow in a natural circulation loop. Int J Heat Mass Tran 2010;53:4101e11. [23] Chen L, Zhang XR, Cao S, Bai H. Study of trans-critical CO2 natural convection flow with unsteady heat input and its implications on system control. Int J Heat Mass Tran 2012;55:7119e32. [24] Zhang XR, Yamaguchi H, Fujima K, Enomoto M, Sawada N. A feasibility study of CO2 e based Rankine cycle powered by solar energy. JSME Int J Ser B Fluids Therm Eng 2005;48:540e7.
[25] Zhang XR, Yamaguchi H, Uneno D, Fujima K, Enomoto M, Sawada N. Analysis of a novel solar energy powered Rankine cycle for combined power and heat generation using supercritical carbon dioxide. Renew Energy 2006;31:1839e54. [26] Zhang XR, Yamaguchi H, Uneno D. Experimental study on the performance of solar Rankine system using supercritical CO2. Renew Energy 2007;32:2617e28. [27] Zhang XR, Yamaguchi H. An experimental study on evacuated tube solar collector using supercritical CO2. Appl Therm Eng 2008;28:1225e33. [28] Zhang XR, Yamaguchi H. Forced convection heat transfer of supercritical carbon dioxide in a horizontal circular tube. J Supercrit Fluids 2007;41:412e20. [29] Chen L, Zhang XR. Simulation of heat transfer and system behavior in a supercritical CO2 based thermosyphon: effect of pipe diameter. ASME J Heat Trans 2011;133:2505e13. [30] Chen L, Deng BL, Zhang XR. Experimental study of trans-critical and supercritical CO2 natural circulation flow in a closed loop. Appl Therm Eng 2013;59: 1e13. [31] Chen L, Deng BL, Zhang XR. Experimental investigation of CO2 thermosyphon flow and heat transfer in the supercritical region. Int J Heat Mass Tran 2013;64:202e11.