Analysis of a novel solar energy-powered Rankine cycle for combined power and heat generation using supercritical carbon dioxide

Renewable Energy 31 (2006) 1839 1854 www.elsevier.com/locate/renene Analysis of a novel solar energy-powered Rankine cycle for combined power and heat generation using supercritical carbon dioxide X.R. Zhang a,, H. Yamaguchi a, D. Uneno a, K. Fujima b, M. Enomoto c, N. Sawada d a Department of Mechanical Engineering, Doshisha University, Kyoto 630-0321, Japan b Mayekawa MFG Co., Ltd., 2000 Tatsuzawa Moriya-city, Ibaraki-Pref. 302-0118, Japan c Showa Denko K. K., 1-480, Inuzuka, Oyama-city, Tochigi 323-8679, Japan d Showa Tansan Co., Ltd., 7-1, Ogimachi, Kawasaki-Ku, Kawasaki-city, Kanagawa 210-0867, Japan Received 1 July 2005; accepted 20 September 2005 Available online 13 December 2005 Abstract Theoretical analysis of a solar energy-powered Rankine thermodynamic cycle utilizing an innovative new concept, which uses supercritical carbon dioxide as a working fluid, is presented. In this system, a truly natural working fluid, carbon dioxide, is utilized to generate firstly electricity power and secondly high-grade heat power and low-grade heat power. The uniqueness of the system is in the way in which both solar energy and carbon dioxide, available in abundant quantities in all parts of the world, are simultaneously used to build up a thermodynamic cycle and has the potential to reduce energy shortage and greatly reduce carbon dioxide emissions and global warming, offering environmental and personal safety simultaneously. The system consists of an evacuated solar collector system, a power-generating turbine, a high-grade heat recovery system, a low-grade heat recovery system and a feed pump. The performances of this CO 2 -based Rankine cycle were theoretically investigated and the effects of various design conditions, namely, solar radiation, solar collector area and CO 2 flow rate, were studied. Numerical simulations show that the proposed system may have electricity power efficiency and heat power efficiency as high as 11.4% and 36.2%, respectively. It is also found that the cycle performances strongly depend on climate conditions. Also the electricity power and heat power outputs increase with the collector area and CO 2 flow rate. The estimated COP power and COP heat increase with the CO 2 flow rate, but decrease with the collector area. The CO 2 -based cycle can be optimized Corresponding author. E-mail address: scho@mail.doshisha.ac.jp (X.R. Zhang). 0960-1481/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2005.09.024

1840 X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 Nomenclature A s Efficient area of evacuated solar collector (m 2 ) C p Specific heat (J/kg 1C) f1, f2, f3 Functions used in PROPATH 12.1 for calculating the thermodynamic parameters of carbon dioxide h Convective heat transfer coefficient (W/m 2 K) h Specific enthalpy (J/kg) L Length of metal tubes in evacuated solar collector (m) m c Mass flow rate of carbon dioxide (kg/s) n Number of metal tubes in evacuated solar collector (dimensionless) P Pressure (MPa) q i Incident solar flux (W/m 2 ) Q T Electric power output from the turbine (W) Q H1 Heat power output from the high-temperature heat recovery system (W) Q H2 Heat power output from the low-temperature heat recovery system (W) r Radius (m) s Specific entropy (J/kg K) T Temperature (1C) Greek letters a Solar absorbtance (dimensionless) s Stefan Boltzmann constant (W/m 2 K 4 ) e Thermal emittance (dimensionless) t Solar transmittance (dimensionless) l Thermal conductivity (W/m K) d Thickness of the inner glass tubes (m) Z Turbine efficiency; a value of 0.9 is used in the present study Subscripts a Ambient b Inner glass tube c Selective coating d Metal tube e Carbon dioxide f Absorber coating g Glass o Glass envelope 1 Outlet of solar collector 2 Outlet of turbine 3 Outlet of the CO 2 -loop in the high-temperature heat recovery system 4 Outlet of the CO 2 -loop in the low-temperature heat recovery system 5 Outlet of feed pump

X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 1841 to provide maximum power, maximum heat recovery or a combination of both. The results suggest the potential of this new concept for applications to electricity power and heat power generation. r 2005 Elsevier Ltd. All rights reserved. Keywords: Solar energy; Supercritical carbon dioxide; Rankine cycle; Power generation; Heat collection 1. Introduction Solar thermal power may be defined as the process by which collected solar energy is converted to electricity through the use of some kind of heat-to-electricity conversion device. Solar thermal power probably has the greatest potential of any single renewable energy process. Solar thermal power can be economically competitive with coal-generated electricity if environmental costs are accounted for [1]. Without considering such environmental costs, solar resources are not cost competitive at present because of the high capital expenditures. However, potential exists for reducing the costs by improving the performance of solar thermal power systems. System performance can be improved by employing some new and innovative ideas in thermal power cycles. During the last two decades, there has been rapid development in the basic technology, and prospects for rapid growth now appear to be very bright for newer approaches. For example, hightemperature technologies (including single-axis and two-axis tracking technologies) have been studied by a number of research groups [2 8]. In addition, hybridization approaches such as combined cycle power generation are being advanced by many to improve the overall energy conversion efficiency [9,10]. Another recent improvement in thermal power cycles is based on the use of mixed working fluids. Using a multicomponent working fluid and/or multipressure boiling, the heat transfer-related irreversibilities can be reduced, and thus improving the resource effectiveness. For example, Kalian et al. [11 14] proposed the use of an ammonia water mixture as the working fluid in the bottoming cycle of a combined cycle power plant to improve thermal power cycles. In this paper, we investigate a novel thermodynamic cycle [15], that may have great potential to improve energy conversion efficiency and greatly reduce the global discharge of CO 2 in the world by using solar energy as the energy source of the system, but also by utilizing CO 2 as the working fluid. CO 2 is an old working fluid and the first reported CO 2 system was built in 1866 by the American Thaddeus S.C. Lowe, but it were the efforts in Germany of Windhausen in 1886 and in Great Britain of J. & E. Hall in 1887 that brought CO 2 systems into wider use, mainly in marine systems, airconditioning and general refrigeration applications [16,17]. As the CFC fluids were introduced in the 1930s and 1940s, these safety refrigerants eventually replaced the old working fluid in most applications. However, the interest in CO 2 as a working fluid has increased considerably from the 1990s, and a number of development and co-operation projects were initiated by both industry and research sectors [18,19]. The novel thermodynamic cycle considered in this paper uses CO 2 as the working fluid in an innovative combination of a CO 2 -based Rankine cycle and heat recovery systems based on heat-exchanging systems (hot water supply system or providing heat source for absorption refrigerator, etc.).

1842 X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 Fig. 1. A CO 2 -based combined power/heat generation Rankine cycle powered by solar energy. Fig. 1 shows a schematic of the cycle. An evacuated solar collector is used to heat CO 2 contained in heating channels and increase CO 2 temperature. The heating in the solar collector creates a supercritical CO 2 high-temperature state (Fig. 1, state 1). The supercritical CO 2 drives the engine of the Rankine system, and power output is available from the turbine generator. The lower pressure carbon dioxide, which is expelled from turbine, is cooled in the high-temperature heat recovery system. At the turbine, outlet supercritical CO 2 still has a higher temperature (Fig. 1, state 2), a kind of high-quality energy, which can be easily utilized as a heat source for absorption refrigerating machine, boiling water or other uses, which can be achieved in the high-temperature heat recovery system. The high-temperature heat recovery system is actually a heat exchanger. After leaving the high-temperature heat recovery system (Fig. 1, state 3), CO 2 enters the lowtemperature heat recovery system in which CO 2 can be further cooled and heat can be recovered further. For example, cold water can be heated into a hot water supply for bath and air conditioning and so on in a building. In the low-temperature heat recovery system, CO 2 is cooled into a liquid state (Fig. 1, state 4) and after that, CO 2 is pumped by the feed pump, back into the higher pressure condition (Fig. 1, state 5), and then the cycle recommences. Among various working fluids, CO 2 (R-744) is a non-flammable and non-toxic fluid and has less influence on the environment and personal safety than other working fluids [16]. The critical pressure and temperature of CO 2 are 7.38 MPa and 31.1 1C, respectively. This critical temperature is much lower than other working fluids, so it is easier for CO 2 to transform into a supercritical state among these fluids. And, owing to the low critical

X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 1843 temperature of CO 2, it may be better than other working fluids for use in thermodynamic cycles. In addition, as a waste product, CO 2 is truly inexpensive. The recycling or recovery of CO 2 would not be necessary, either for environmental or for economic reasons. In contrast to the new fluorocarbon alternatives, CO 2 is already available in abundant quantities in all parts of the world, and production capacity or distribution logistics need not be developed. CO 2 is also thermally stable and behaves inertly, thus eliminating material problems or chemical reactions in the system. All of these would simplify work during the life cycle of the proposed new system and may provide an opportunity to further reduce system costs to the point where mass production becomes feasible. This paper gives a theoretical analysis of the proposed CO 2 -based Rankine cycle powered by solar energy, enabling the evaluation of the system performance as a function of design parameters, and to obtain a basic understanding of such a Rankine cycle using CO 2 and suggest modifications to the Rankine cycle design and operation to yield improved useful cycle outputs and energy conversion efficiencies and so on. 2. Thermodynamic analysis of the proposed cycle The use of CO 2 as a working fluid for a solar energy-powered Rankine cycle provides a new field of interest for process modeling and simulation. As the first step towards a fundamental understanding and estimation of the performance and characteristics of the system, a mathematical model that simulates the system behavior of the Rankine cycle considering a steady state was constructed. The thermodynamic properties, illustrated in Fig. 2, indicate critical data and phase envelopes of CO 2 in pressure enthalpy coordinates. This heat transfer processing (in the solar collector) above the critical point results in a transcritical cycle, i.e. with a subcritical low-side and a supercritical high-side pressure. In these proposed transcritical cycle conditions, developments of temperature and pressure of CO 2 were sometimes close to the critical point of CO 2 (T c ¼ 31:1 1C, P c ¼ 7:38 MPa). One characteristic of supercritical carbon dioxide close to its critical point is that its thermodynamic and transport properties exhibit rapid variations with a change in temperature, especially near the pseudo-critical Fig. 2. Pressure enthalpy diagram of carbon dioxide.

1844 X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 point (the temperature at which the specific heat reaches a peak for a given pressure) [20,21]. In this paper, a Program Package for Thermophysical Properties of Fluids database version 12.1 (PROPATH 12.1) was used in order to estimate the thermodynamic and transport properties of CO 2. With the available thermodynamic properties data, we have modeled the components of a system based on the proposed cycle. 2.1. System simulation The thermodynamic state conditions of the proposed combined cycle were evaluated under the processes associated with the flow and heat transfer in the solar collectors and expansion processes in the turbine and so on: 1. In the evacuated solar collector, the working fluid (supercritical CO 2 ) is distributed uniformly in all heat removal tubes. 2. In the evacuated solar collector, thermal resistance among the selective surface, metal tube and fin is neglected. The temperatures of the selective surface, metal tube and fin are considered the same. 3. The solar collector is placed facing south and exposed to solar radiation throughout the day; it is also kept reasonably free from dust. 4. The atmospheric temperature used in the calculation is the monthly averaged one. 5. At point 4 (at the pump outlet), the pressure is 9.0 MPa. 6. The pressure drop in the turbine has a constant value (2.5 MPa). 7. The friction losses in the solar collector and CO 2 /water heat exchangers are neglected. 8. A good thermal insulation is provided for the CO 2 loop, and the thermal loss from the CO 2 loop is neglected. Although the assumptions and simplifications limit the usefulness of this analysis, the results show the potential of using the supercritical CO 2 as the working fluid in the Rankine cycle powered by solar energy. 2.2. Basic equation In this section, a mathematical model that describes the processes in each component of the system is presented. 2.2.1. Evacuated solar collector In this simulation, it is supposed that an all-glass evacuated solar collector with a U-tube heat removal system was used in the CO 2 -based Rankine cycle. Fig. 3 shows cross sections of the evacuated tube collector. It has five essential elements: 1, the glass envelope providing the vacuum; 2, the inner glass tube; 3, the selective solar absorber coating deposited in the interior of the inner glass; 4, the metal U-tube inserted into the inner glass tube with a fin connecting the U-tube to the inner glass tube; and 5, the working fluid CO 2 in contact with the metal tube. It can be seen that solar radiation passes through both the glass tubes to be absorbed by the selective surface. The surface heats up and the heat is transferred by conduction to the metal tube and then by convection to the working fluid. The loss of heat to the ambient air will follow the direction opposite to that of the incident

X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 1845 Fig. 3. (a) Transverse cross-sectional view of the solar collector. (b) Part of a longitudinal section showing the differential control volume used in the analysis of the model. Table 1 Thermophysical properties and geometry used in the mathematical model of the evacuated solar collector Properties of glass: Thermal conductivity, l g (W/m K) 1.25 Solar transmittance, t g 0.90 Solar absorbtance, a g 0.05 Solar reflectance, r g 0.05 (normal incidence) Thermal emittance, e o, e b 0.83 Properties of absorber coating: Solar transmittance, t f 0.04 Solar absorbtance, a f 0.927 Solar reflectance, r f 0.033 Geometry: Length of the tubes, L (m) 1.7 Mean radius of the glass envelope, r o (m) 0.018 Mean radius of the inner glass tube, r b (m) 0.013 Mean radius of the metal tube, r d (m) 0.03 solar radiation. The thermophysical properties and geometry of the five elements used in this simulation are listed in Table 1. Each of the differential control volumes in Fig. 3(b) is an annular region of length dx, the x-axis indicating the direction of flow of the fluid inside the tube. Applying the first law of thermodynamics to each one of the differential control volumes and assuming a steady state with constant thermophysical properties except CO 2, we obtain the following equations in standard notation.

1846 X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 Glass envelope: r b sðt 4 b a g r o q i þ r b h b ðt b T o Þþ T 4 o Þ 1=e g þðr b =r o Þð1=e g 1Þ ¼ r oh a ðt o T a Þþr o se o ðt 4 o T 4 a Þ. (1) Inner glass tube: t g a g r b q i þ l gðt c T b Þ r ln 1 þ d b sðt ¼ r b h b ðt b T o Þþ 4 b T 4 o Þ 1 þ r. (2) b 1 1 2r b e g r o e g Selective coating and metal tube: t 2 g a f r b q i ¼ l gðt c T b Þ ln 1 þ d þ 2r d h c ðt c T e Þ. (3) 2r b CO 2 : m c C p dt e dx ¼ r dh c ðt c T e Þ. (4) 2.2.2. Power-generating turbine The outlet temperature of the turbine and the electric power output from the turbine can be calculated by T 2 ¼ f 1ðs 2 ; P 1 DPÞ, (5) Q T ¼ m c Zðh 1 h 2 Þ, (6) where h 2 ¼ f 2ðs 2 ; P 1 DPÞ: 2.2.3. Heat recovery system The heat recovery systems are intended to recover heat from the CO 2 -based Rankine cycle section, and at the same time to cool CO 2 to a temperature low enough to change into a liquid state to form a complete Rankine cycle. In this simulation, it is supposed that two CO 2 /water heat exchangers (HX1 in the high-temperature heat recovery system and HX2 in the low-temperature heat recovery system) were utilized, respectively, to heat highertemperature water (in the high-temperature heat recovery system) and lower-temperature water (in the low-temperature heat recovery system). In the CO 2 /water heat exchangers, the heat capacity of CO 2 is calculated based on the average temperature of the CO 2 -side of the heat exchangers. The outlet temperatures for CO 2 loop and water loop in the CO 2 / water heat exchangers and heat-exchanging quantity are calculated based on the heat transfer equations and computations of heat balance [22]. For the heat outputs from the heat recovery systems the following are calculated: Q H1 ¼ m c ðh 2 h 3 Þ, (7) Q H2 ¼ m c ðh 3 h 4 Þ. (8)

X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 1847 2.2.4. Feed pump The temperature at the pump outlet and the power consumption of the pump can be calculated by T 5 ¼ f 3ðP 5 ; h 5 Þ, (9) Q P ¼ m c ðh 5 ; h 4 Þ, (10) where h 5 ¼ f 2ðs 5 ; P 5 Þ. The hybrid system has two useful outputs: one is electric power, with a solar efficiency of COP power ¼ Q T R L 0 2r onq i dx (11) and the other is heat recovery achieved in the CO 2 /water heat exchangers, with a solar efficiency of COP heat ¼ Q H1 þ Q H2 R L 0 2r onq i dx, (12) where q i is the solar flux density, and R L 0 2r onq i dx is the total incident solar energy. The performance of the system is determined by solving this set of equations simultaneously. Convergent solutions were obtained based on the method of successive loop iterations in a logic manner. The convective heat transfer coefficient h c of supercritical CO 2 in the Rankine cycle was calculated from the relations given by Ref. [23]. The heat transfer coefficients h a (glass envelope to ambience) was calculated from the expressions reported in Ref. [24]. For the heat transfer coefficient between the two glass (h b ), we used the value 0.026 W/m 2 K for very good vacuum. 3. Results and discussion The mathematical relations presented in the thermodynamic analysis section were employed to determine the performance of the proposed system. Also, in this study, the objective was to analyze the main design parameters that influence the system performance and also to optimize the design of such a CO 2 -based Rankine system. The design conditions were varied with only one parameter changed at a time in order to study the effect of those changes on the cycle performance. For all simulations, the following system specifications were assumed. The inlet temperature and mass flow of water loop in the CO 2 /water heat exchanger in the high-temperature heat recovery system were assumed to be 40.0 1C and 10.0 kg/min, respectively. The inlet temperature and mass flow of water loop of the CO 2 /water heat exchanger of the low-temperature heat recovery system were assumed to be 9.0 1C and 2.0 kg/min, respectively. The heat-exchanging areas of HX1 and HX2 were 0.2 and 0.4 m 2. 3.1. Effect of incident solar radiation The effect of incident solar radiation on the system performance was investigated with the CO 2 flow rate taken as 0.012 kg/s and the efficient collector area as 5.0 m 2. The results are shown in Figs. 4 and 5. Fig. 4 shows the monthly average incident solar radiation (from January to December) in the Kyoto area, Japan. Also, the calculated temperature values of

1848 X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 Temprature ( C) 250 200 150 100 m c =0.012 kg/s; A s =5.0 m 2 outlet temperature of solar collector, T 1 outlet temperature of turbine, T 2 outlet temperature of HX1, T 3 1000 900 800 700 600 500 400 Incident solar radiation, q i 50 300 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Fig. 4. Effect of incident solar radiation on the cycle temperatures. Monthly average solar radiation (W/m 2 ) 1000 m c =0.012 kg/s; A s =5.0 m 2 Electri power output Heat power output of HX1 Heat power output of HX2 0.45 0.40 800 0.35 Output (W) 600 400 COP solar COP heat 0.30 0.25 0.20 0.15 COP 0.10 200 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Fig. 5. Effect of incident solar radiation on the useful cycle outputs and power efficiencies of the CO 2 -based Rankine cycle. some important positions in the CO 2 -based Rankine cycle are plotted. It can be seen from this figure that there is an obvious influence of incident solar radiation on the cycle temperatures. As the average solar radiation increases gradually from January to July, the cycle temperature values increase and vice versa from July to December. The temperature of supercritical CO 2 at the outlet of the solar collector (T 1 ) can reach about 220.0 1C in the summer, in July, and in the winter, for example, in December, the temperature of T 1 is about 125.0 1C. The annual average temperature of T 1 is about 180.0 1C. So the cycle temperatures strongly depend on the seasonal climates, if other design and operating parameters are fixed. It can also be seen that there is a temperature difference of about 30 40 1C between the inlet and outlet of the turbine. And at the outlet of HX1, CO 2 still have a higher temperature, for example, in the summer, in July, the temperature is about 146 1C and the annual average temperature at this point is about 115 1C. This result shows that the CO 2 temperature at the outlet of the high-temperature

X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 1849 heat recovery system (T 3 ) is still relatively high, which can be easily utilized to further recover the heat from the CO 2 -based Rankine cycle. Fig. 5 shows the variations of Rankine cycle outputs and efficiencies with the season (i.e. incident solar radiation). It can be seen that the useful outputs of the Rankine cycle vary with the incident solar radiation, including electric power output, the heat power outputs achieved in the heat recovery systems. In addition, compared with the electric power output, the influence of incident solar radiation on the heat power outputs achieved, respectively, in HX1 and HX2 are more obvious. The reason may be that the solar radiation has a large influence on the cycle temperatures, as shown in Fig. 4. The variations of the cycle temperatures can lead to changes in the useful outputs. However, the influence of the temperatures on the heat outputs seems to be more direct compared to the electric power output, since except for the cycle temperature, the electric power output is decided by many other influencing factors, such as CO 2 pressure. Contrary to the influence of the solar radiation on the useful cycle outputs, the higher electric power efficiencies COP solar and heat power efficiencies COP heat were obtained in winter conditions (as shown in Fig. 5). This means that in summer, the larger cycle outputs and a little lower electricity and heat efficiency are obtained. It can be explained that although the electric power and heat power output increase in summer, the incident solar radiation also increases. Furthermore, the increasing amplitude of the solar radiation is larger than that of the power output. In addition, the temperature difference between the solar collector surface and the surrounding air becomes greater in summer than in winter, and then the thermal loss of the solar collector to the ambient also increases, which may also contribute to the occurrence of this phenomenon. In addition, it can be observed that the annual-averaged electricity output efficiency, heat output efficiency and overall efficiency are 11.4%, 36.2% and 47.6%, respectively. In addition, the solar energy-powered Rankine cycle using CO 2 in the annual average condition is shown by the solid line in the p h diagram in Fig. 6. From the figure, it can be clearly seen that the CO 2 -based Rankine cycles work in the transcritical region at the highpressure side of about 9.0 MPa and the low-pressure side of about 6.5 MPa. The working Fig. 6. A comparison of the Rankine cycles based on CO 2 and other working fluids.

1850 X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 cycle is used for simultaneous heating and heat recovery in the temperature from about 32.0 177.4 1C for the heating process in the solar collector and 143.1 24.2 1C for heat recovery. The heating process in the Rankine cycle is in the supercritical region for CO 2.In other words, under the supercritical state, CO 2 is heated by the absorber surface in the solar collector. A further thermodynamic analysis was done in the example of other working fluids, water, ammonia and HFC-134a. In this thermodynamic estimation for the sake of comparison, the pressures of the high side and low side in the thermodynamic cycles, inlet temperature of the solar collector, the mass flow rate of the working fluid and the quantity of heat absorbed into the fluids in the solar collector are fixed to be the same as the CO 2 -based Rankine cycle in Fig. 6. The results show that the electric power efficiencies were 11.4%, 4.8%, 2.1% and 1.1%, respectively, for CO 2, HFC-134a, ammonia, and water. These results show that there is an obvious advantage of using CO 2 as a working fluid in such a solar energy-powered Rankine cycle, at least in the range of temperature and pressure shown in Fig. 6. Among these working fluids, the critical temperature of CO 2 is 31.1 1C, much lower than those of other fluids, for example, 101.15 1C for HFC-134a, 132.25 1C for ammonia and 374.15 1C for water. And the thermodynamic and transport properties of CO 2 seem to be favorable in terms of heat transfer and pressure drop, compared to other typical working fluids [16]. All these may be contributed to the advantage of using supercritical CO 2 as the working fluid in a Rankine cycle. 3.2. Effect of area of evacuated solar collector Fig. 7 shows the dependence of cycle temperatures as a function of a design parameter, the area of evacuated solar collector for the annual-averaged incident solar radiation of 682.0 W/m 2 and CO 2 flow rate of 0.012 kg/s. It is clear from the curves of Fig. 7 that the cycle temperatures increase with the area of the solar collector. The reason is that the heat quantity absorbed into CO 2 in the solar collector greatly increases when the area of the 300 m c =0.012 kg/s; q 1 =682.0 W/m 2 250 Temprature ( C) 180 120 60 outlet temperature of solar collector, T 1 outlet temperature of turbine, T 2 outlet temperature of HX1, T 3 0 4 8 12 16 Area of evacuated solar collector (m 2 ) Fig. 7. Effect of area of evacuated solar collector on the cycle temperatures.

X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 1851 5000 m c =0.012 kg/s; q 1 =682.0 W/m 2 4000 COP solar COP heat 0.4 Output (W) 3000 2000 Electric power output Heat power output of HX1 Heat power output of HX2 0.3 0.1 COP 1000 0.1 0 0.0 4 8 12 16 Area of evacuated solar collector (m 2 ) Fig. 8. Effect of area of evacuated solar collector on the useful cycle outputs and power efficiencies of the CO 2 - based Rankine cycle. solar collector is enlarged. So when other parameters, including CO 2 flow rate, are fixed, the cycle temperatures increase if the collector area is increased. From Fig. 8, which shows the effect of the solar collector area on the system performances, it is obvious that the electric power output and heat output from the Rankine cycle increase with the collector area. The reason is that the system temperatures can be effectively raised by increasing the collector area (from Fig. 7). It can also be seen from Fig. 8 that the enlarged collector area can increase the heat output more efficiently than the electric power output. Further, from Fig. 8, it is obvious that the electric power efficiency decreases with the increase of the collector area. This is because increasing the collector area increases the overall quantity of solar energy irradiance in the solar collector. Although the electric power output is also increased by enlarging the collector area, the efficiency of electric power output decreases with the collector area. Another important reason is that the heat loss to the ambient environment is increased if the collector area is enlarged. Therefore, when the collector area is enlarged, and that means the cost of the whole cycle system increase, the electric and heat power output can be raised, especially, the heat output can be largely increased by this cost increase, but in contrast, the generation efficiency for electric power decreases with the area increase. 3.3. Effect of CO 2 flow rate How much of CO 2 should be charged into the Rankine cycle loop for providing maximum power, maximum heat recovery or maximum power efficiency? This question is closely related to the optimal flow rate of CO 2 in the cycle. The effects of CO 2 flow rate over the range of 0.004 0.03 kg/s were investigated with the efficient collector area taken as 5.0 m 2. The variations of the cycle temperatures with CO 2 flow rate are shown in Fig. 9. It can be seen from Fig. 9 that increasing the CO 2 flow rate can increase the Rankine cycle temperatures greatly. The outlet temperature of the collector increases from about 120.0 240.0 1C in the range of this study. It may be mainly explained that the heat transfer processes in the metal tubes of the solar collector are enhanced by the increase of CO 2 flow rate and more heat can be removed into CO 2 in the solar collector, although the solar

1852 X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 300 250 A s =5.0 m 2 ; q 1 =682.0 W/m 2 outlet temperature of solar collector, T 1 outlet temperature of turbine, T 2 outlet temperature of HX1, T 3 Temperature ( C) 200 150 100 50 0 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 CO 2 flow rate (kg/s) Fig. 9. Effect of CO 2 flow rate on the cycle temperatures. 2000 A s =5.0 m 2 ; q 1 =682.0 W/m 2 0.8 1600 Electric power output Heat power output of HX1 Heat power output of HX2 0.7 0.6 Output (W) 1200 800 COP solar COP heat 0.5 0.4 0.3 COP 400 0.2 0.1 0 0.000 0.005 0.010 0.015 0.020 0.025 0.030 CO 2 flow rate (kg/s) 0.0 Fig. 10. Effect of CO 2 flow rate on the useful cycle outputs and power efficiencies of the CO 2 -based Rankine cycle. radiation and collector area are the same. From Fig. 10, which shows the effect of CO 2 flow rate on the cycle outputs and power efficiency, it is obvious that the outputs, including electric power output, heat outputs from HX1 and HX2, increase with the CO 2 flow rate. Increasing CO 2 flow rate can enhance the heat transfer processes not only in the solar collector but also in the CO 2 /water heat exchangers in the heat recovery systems, which result in an increase of heat power output from the Rankine cycle. On the other hand, the increase of CO 2 flow rate can be very helpful in not only increasing the temperature of the collector outlet but also enhancing the electric generating process in the turbine, which results in the increase of not only the electric power output but also the efficiency of electric power output, as can be seen from Fig. 10. Thus, the CO 2 flow rate in the Rankine cycle

loop (i.e. the amount of CO 2 charged in the loop) should be raised as much as possible under the condition of a safe operation in order to increase the useful outputs and cycle efficiencies. 4. Conclusions A theoretical investigation has been carried out of a novel solar energy-powered Rankine cycle system using supercritical CO 2 as the working fluid. The cycle is a combination of renewable energy solar energy and truly natural working fluid CO 2 for a combined generation of electric power and heat. Simulations of the proposed cycles for cycle temperatures, useful outputs and cycle efficiencies show that the novel cycle has a high solar electric conversion efficiency and solar thermal conversion efficiency. In the range studied in this paper, the proposed system may have an annual average electricity output efficiency and heat output efficiency as high as 11.4% and 36.2%, respectively. A comparison among the various working fluids shows that the solar energy-powered Rankine cycle using carbon dioxide has a higher power efficiency than other working fluids. A parameter study of such a CO 2 -based solar energy Rankine cycle is performed and the cycle can be optimized to produce maximum power, maximum heat recovery, or maximum power efficiency. The numerical results show that the cycle outputs (electricity output and heat output) can be significantly increased by increasing CO 2 flow rate, by increasing solar collector area and by increasing solar radiation. The cycle efficiencies (electricity output efficiency and heat output efficiency) are found to be increased with CO 2 flow rate, but decreased with solar collector area and solar radiation. The results show that the proposed cycle has great potential to achieve a true green energy generation, relieving energy pressure in the world and greatly reducing CO 2 emission. It is recognized that future studies, especially transient studies and experimental studies, are needed to establish the practical usefulness of this cycle. Acknowledgements This study was supported by the Academic Frontier Research Project on Next Generation Zero-emission Energy Conversion System of Ministry of Education, Culture, Sports, Science and Technology. References X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 1853 [1] Goswami DY. Solar energy the natural solution for energy and environmental problems, heat and mass transfer. In: Proceedings of the second ASME-ISHMT heat and mass transfer conference. New Delhi, India: Tata-McGraw Hill Publishers; 1994. p. 57 60. [2] Frier D, Cable RG. An overview and operation optimisation of the Kramer junction solar electric generating system. Jerusalem: ISES World Congress; 1999. p. 241 6. [3] Zarza E. The DISS project: direct steam generation in parabolic troughs; operation and maintenance experience; update on project status. In: Proceedings of solar forum 2001: solar energy: the power to choose, Washington, DC; 2001. p. 24 5. [4] Francia G. Pilot plants of solar steam generation systems. Sol Energy 1968;12:51 64. [5] Mills DR, Morrison GL. Compact linear fresnel reflector solar thermal powerplants. Sol Energy 1999;68:263 83. [6] Romero M. Distributed power from solar tower systems: a MIUS approach. Jerusalem: ISES World Congress; 1999. p. 286 95.

1854 X.R. Zhang et al. / Renewable Energy 31 (2006) 1839 1854 [7] Schramek P, Mills DR. Potential of the heliostat field of a multi tower solar array. In: Proceedings of the 10th solarpaces international symposium on solar thermal concentrating technologies, Sydney; 2000. p. 157 63. [8] Mills DR. Advances in solar thermal electricity technology. Sol Energy 2004;76:19 31. [9] Bohn MS, Williams TA, Price HW. Combined cycle power tower, solar engineering 1995. In: Proceedings of the 1995 ASME/JSME/JSES international solar energy conference; 1995. p. 597 606. [10] DeLaquil P, Kearney D, Geyer M, Diver R, editors. Solar thermal electric technology. In: Renewable energy sources for fuel and electricity. Washington, DC: Island Press; 1993 [chapter 5]. [11] Kalina AI. Combined cycle and waste-heat recovery power systems based on a novel thermodynamic energy cycle utilizing low-temperature heat for power generation. ASME Paper 83-JPGC-GT-3, 1983. [12] Kalina AI. Combined cycle system with novel bottoming cycle. ASME J Eng Gas Turbines Power 1984;106:737 42. [13] Kalina AI, Tribus M, El-Sayed YM. A theoretical approach to the thermophysical properties of twomiscible-component mixtures for the purpose of power-cycle analysis. ASME Paper 86-WA/HT-54, 1986. [14] Kalina AI, Tribus M. Advances in kalina cycle technology (1980 1991): Part I, development of a partical cycle, energy for the transition age. In: Proceedings of the florence world energy research symposium, vol. 97, Firenze, Italy; 1990. p. 111 24. [15] Zhang XR, Yamaguchi H, Fujima K, Enomoto M, Sawada N. A feasibility study of CO 2 -based Rankine cycle powered by solar energy. JSME International Journal, Series B 2005;48:540 7. [16] Kim MH, Pettersen J, Bullard CW. Fundamental process and system design issues in CO 2 vapor compression systems. Prog Energy Combus Sci 2004;30:119 74. [17] Pettersen J. An efficient new automobile air-conditioning system based on CO 2 vapor compression. ASHRAE Trans: Symp 1994; 657 65. [18] Lorentzen G. Trans-critical vapour compression cycle device. International Patent Publication WO 90/07683, 1990. [19] Lorentzen G, Pettersen J. New possibilities for non-cfc refrigeration. In: IIR international symposium on refrigeration, energy and environment, Trondheim, Norway; 1990. p. 147 63. [20] Bullard CW, Yin JM, Hrnjak PS. Compact counterflow gas cooler for R-744. ASHRAE Trans 2002;108:487 91. [21] Schonfeld H, Krauss WE. Calculation and simulation of a heat exchanger: supercritical carbon dioxidewater. In: Proceedings of IIR international conference on heat transfer issues in natural refrigerants; 1997. p. 72 80. [22] Rohsenow WM, Hartnett JP, Ganic EN. Handbook of heat transfer applications. Washington, DC: Kingsport Press; (2nd ed., New York: McGraw-Hill) 1985. p. 4-1 20. [23] Hashimoto K, Saikawa M. Preliminary experimental result of CO 2 gas-cooling over-all heat transfer coefficient under super critical condition. In: Proceedings of IIR international conference on heat transfer issues in natural refrigerants; 1997. p. 50 8. [24] Sawhney RL, Inderjit, Bansal NK. Steady state thermal analysis of an evacuated tubular collector with a CPC booster. Sol Wind Technol 1984;1:237 44.