- Email: [email protected]
Simulation and experimental study of an air tube-cavity solar receiver
Energy Conversion and Management 103 (2015) 847–858
Contents lists available at ScienceDirect
Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman
Simulation and experimental study of an air tube-cavity solar receiver Kunzan Qiu, Liang Yan, Mingjiang Ni, Cheng Wang, Gang Xiao ⇑, Zhongyang Luo, Kefa Cen State Key Laboratory of Clean Energy Utilization, Zhejiang University, Hangzhou 310027, China
a r t i c l e
i n f o
Article history: Received 20 March 2015 Accepted 4 July 2015 Available online 20 July 2015 Keywords: Solar receiver High temperature Air Tube cavity Simulation model
a b s t r a c t High temperature air is a potential candidate as a heat transfer fluid to transport energy from concentrated solar power to gas turbines. A 15-turn helically coiled tube cavity receiver with an optical splitter at the bottom is designed and fabricated. Its performance is investigated with a five 7-kW Xe-arc lamps array system as heat source. Eight K-type thermocouples are placed from top to bottom with an equal interval. The outlet temperature experimentally ranges from 593 °C to 546 °C when the air flow rate increases from 1 m3/h to 5 m3/h for up-flows, while it ranges from 662 °C to 570 °C for down-flows, when the average flux on aperture is around 120 kW/m2. The Monte-Carlo ray-tracing method and the Lambert testing method with a charge-coupled device (CCD) camera are used to simulate and evaluate the concentrating radiation energy distribution on the cavity’s internal walls, and then the actual flux distribution of each turn of the helically coiled tube is obtained. A comprehensive simulation model is proposed and validated by the experimental results, where the outlet temperature deviations are within 8.0% and 2.5% for down and up-flows, respectively. The model provides a detailed analysis of heat flows at different conditions, and indicates optimization ways to improve the efficiency and reduce heat losses. The simulation results show that the outlet temperature can increase up to around 800 °C at 5 m3/h under an average flux of 300 kW/m2, and the thermal efficiency can be improved from around 56% to around 64% by decreasing the inner radius from 6 mm to 4 mm at the expense of increasing pressure drop of around 56 kPa. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction To alleviate the energy crisis, many countries have considered solar energy as an alternative because it is clean, renewable, and abundant [1]. Concentrating solar power (CSP) has already been applied commercially for power generation [2]. It is well known that high electrical efficiency of the CSP system is highly dependent on being able to achieve a high working temperature of the heat transfer fluid (HTF) [3]. Synthetic oil has been widely utilized in solar trough plants, currently reaching around 4 GW [4] with the maximum working temperature around 400 °C [5]. Molten salt is another material to increase the working temperature up to around 560 °C [6]. Further increase in the temperature of those working fluids is difficult due to the constraints of physical properties. Air receivers have been proposed as an alternative HTF to achieve higher working temperatures in order to obtain higher efficiency cycles for systems such as Brayton cycle [7]. The Institute of Electrical Engineering of Chinese Academy of Science developed an
⇑ Corresponding author. Tel.: +86 571 87953290; fax: +86 571 87951616. E-mail address: [email protected] (G. Xiao). http://dx.doi.org/10.1016/j.enconman.2015.07.013 0196-8904/Ó 2015 Elsevier Ltd. All rights reserved.
air intake tube receiver with bubbling particle. Concentrated solar radiation was mostly absorbed by particles and converted into thermal energy which was carried out by air flow. The maximum air outlet temperature reached 624 °C [8]. AirLight Energy installed an air-based receiver for solar trough concentrators. The receiver is 212 m long, consisting of 4608 cavities made of helically coiled tubes [9]. The receiver can deliver hot air at a temperature around 650 °C under a concentration ratio of 98 when the inlet air temperature is 120 °C. The German Aerospace Center (DLR) developed an air receiver consisting of 40 absorber tubes arranged in a cavity and connected in parallel. The outlet air temperature reached up to 800 °C at 4.5 bar, with the efficiency of the solar-hybrid micro-turbine system reaching 44% [10]. The Sweden Royal Institute of Technology (KTH) designed a gas turbine-receiver unit, in which the inlet air was first compressed by a compressor, preheated by a regenerator, and then heated by a solar receiver. The air outlet temperature reached 950 °C [11]. Abengoa Solar developed a volumetric solar air receiver. The maximum air outlet temperature reached up to 1000 °C [12]. The Swiss Federal Institute of Technology in Zurich (ETH) developed a pressurized solar receiver whose main component was a cylindrical silicon carbide cavity surrounded by a concentric annular reticulate porous ceramic
848
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Nomenclature Q in Q ref Q 1,rd Q 2,rd Q 1,cv Q 2,cv Q 3,rd hcv1 hcv21 hcv22 hcv23 s
q Ti To Ag
incident irradiation on the receiver (W) reflective loss of incident irradiation (W) heat radiation loss between the glass cover and the atmosphere (W) heat radiation loss between the shell wall and the atmosphere (W) heat convection loss between the glass cover and the atmosphere (W) heat convection loss between the shell wall and the atmosphere (W) heat radiation loss between the cavity inner wall and the atmosphere (W) convection heat transfer coefficient between the glass cover and the atmosphere (W/(m2 K)) convection heat transfer coefficient between the shell up surface and the atmosphere (W/(m2 K)) convection heat transfer coefficient between the shell side surface and the atmosphere (W/(m2 K)) convection heat transfer coefficient between the shell bottom surface and the atmosphere (W/(m2 K)) thickness of the glass cover reflectivity of the glass cover inlet air temperature (K) air outlet temperature (K) area of the glass cover (m2)
(RPC) foam. The outlet temperature of the air flowing through the RPC reached up to 1060 °C at 5 bar under an input solar power of 32–38 kW [13]. The Weizmann Institute of Science divided the aperture of the solar receiver into separate stages according to irradiance distribution to minimize heat losses. The air temperature was initially raised up to 750 °C by preheaters, i.e., cavity tubes. Subsequently, a directly irradiated annular pressurized receiver (DIAPR) was used. The receiver was capable of supplying hot air at a temperature of 1300 °C under a pressure of 10–30 bar [14]. The first demonstration of the use of air as HTF in a solar power plant was through a 1.5 MW tower power system with a ceramic volumetric receiver in Jülich, Germany. The receiver produced hot air with a temperature of 680 °C and was in operation successfully in September 2008 [6]. For scaling up and design optimization purposes, a reliable comprehensive predictive model describing performance of the air solar receiver is highly desired. Wang et al. [15] used a 3D numerical model with a uniform flux distribution on tube surface for a coiled tube receiver to analyze temperature distribution and thermal stress of the tube. Paitoonsurikarn et al. [16] numerically investigated the effects of different shapes on the convective losses of solar cavity receivers by assuming an isothermal wall temperature. Prakash et al. [17] computed convective heat losses of a cylindrical cavity receiver under different inclination angles assuming an adiabatic boundary condition on cavity external wall. The foregoing modeling and simulation efforts focused on analyzing the effects of various heat transfer modes. However the incident radiation flux on the cavity wall has often been simplified by assuming a uniform or arbitrarily varying distribution. Since the radiation flux is the most important factor for characterizing the solar receiver performance, a realistic model for the distribution of the radiation heat flux is necessary. In this work, a generic air tube cavity solar receiver is considered. Monte-Carlo ray tracing technique is used to accurately obtain the realistic radiation flux distribution on the air tube external wall in order to build a comprehensive simulation model coupling optical and heat transfer processes of high-temperature
Q absorbed energy carried away by the air (W) L characteristic length (m) Q 21,cv heat convection loss between the shell up surface and the atmosphere (W) Q 22,cv heat convection loss between the shell side surface and the atmosphere (W) Q 23,cv heat convection loss between the shell bottom surface and the atmosphere (W) Q 21,rd heat radiation loss between the shell up surface and the atmosphere (W) Q 22,rd heat radiation loss between the shell side surface and the atmosphere (W) Q 23,rd heat radiation loss between the shell bottom surface and the atmosphere (W) Q 31,rd heat radiation loss between the outer wall of the coils and the atmosphere (W) Q 32,rd heat radiation loss between the optical splitter and the atmosphere (W) Tg average glass temperature (K) Ta average atmosphere temperature (K) eg glass cover emissivity r the Stefan–Boltzman constant
air solar receiver, which considerably affects the distributions of wall temperature and the heat loss at the external walls. A 15-turn helically coiled tube cavity receiver with an optical splitter at the bottom is designed and fabricated in order to assess the efficacy of the comprehensive model. The experiments on the cavity receiver were performed under a five 7-kW Xe-arc lamps array system. Good agreements were obtained between the measurement and the model calculation. The comprehensive simulation model is subsequently used to conduct a parametric investigation to understand the performance of the cavity receiver under different operating conditions and to find ways to optimize solar air receiver performance. 2. Experimental set-up and methods A solar simulator consisting of five 7-kW Xe-arc lamps, as shown in Fig. 1, is employed to heat an air tube-cavity receiver. An experimental flowchart is shown in Fig. 2(a). Air is pumped into
Fig. 1. Solar simulator with five 7 kW Xe-arc lamps.
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
849
Fig. 2. (a) Experimental flowchart; (b) configuration of cavity receiver; (c) surfaces of the receiver (1: surface of the glass cover; 21: shell up surface; 22: shell side surface; 23: shell down surface; 31: the outer wall surfaces of the coils; 32: the outer surfaces of the optical splitter).
a coiled tube by a blower and the flow rate, ranging from 1 m3/h to 5 m3/h, is adjusted by a rotor flow meter (LZB-25), manufactured by Chemical Instrument Company of China [18]. A data acquisition system (Aglient 34970A), produced by Aglient Company of U.S. [19], linked with 8 K-type thermocouples, records air temperatures along the coiled tube every 10 s. A configuration of the air tube-cavity receiver is shown in Fig. 2(b). The solar cavity receiver consists of a 15-turn helically coiled copper tube with inner diameter of 1.2 cm and thickness of 0.1 cm. The inner diameter and height of the cavity are 14 and 25 cm, respectively. Eight K-type thermocouples, produced by Omega Company of U.S. [20], are placed from the top to the bottom with uniform spacing. The aperture of the cavity is covered with a quartz window. At the bottom is an optical splitter, i.e., stainless steel cone with a height of 7.5 cm and a bottom diameter of 14 cm. The tube coils are wrapped with thermal insulating materials. Fig. 2(c) shows different surfaces of the receiver, which will be considered in the calculation of heat losses. 3. Comprehensive modeling of the cavity receiver The comprehensive simulation model of the air tube-cavity receiver consists of three parts: (1) evaluation of the total incident
radiation on the aperture, (2) optical simulation to obtain relative flux distribution, and (3) combination of optical and heat transfer processes using the FLUENT software.
3.1. Evaluation of the total incident radiation The evaluation of the total incident radiation has two important phases: (1) detection of a relative flux distribution of concentrated radiation on the aperture and (2) detection of the absolute radiation intensity at a reference point. The measurement system includes a Lambert plate, a charge-coupled device (CCD) camera, and a radiometer, among others, as shown in Fig. 3. The Lambert plate, a standard diffuse reflection white board produced by Labsphere Company of U.S., is a 25 cm 25 cm square. The Lambert plate is placed at the aperture of the cavity. The CCD camera, manufactured by Hamamatsu Company, Japan, is used to capture the concentrating spots on the Lambert plate. The output video signal has a linear relationship with the received luminance distribution. A radiometer (30 (150) A-BB-18) with a 17 mm-diameter aperture, manufactured by Ophir Company of Israel [21], is used to detect the absolute radiation intensity of reference points on the Lambert plate. The power
850
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Fig. 5. (a) Model of the coiled tube; (b) schematic of ray tracing.
Fig. 3. Detecting devices and method.
sensor has an accuracy of 3.0% at the 0.19–20.00 lm wavelength range [7] (this range is common in most solar lights) [22]. Fig. 4(a) displays a relative flux distribution of the concentrating radiation on the Lambert plate, where the flux intensity increases rapidly toward the center. The calculation area of the concentrating radiation on the Lambert plate is separated into two regions, as shown in Fig. 4(b). A1 and A2 represent the inside and outside areas, respectively, of the glass aperture. The calculation result shows that the average flux intensities on A1 and A2 are 120.63 and 77.62 kW/m2, respectively, and the concentrating radiation on A1 is 1.85 kW, nearly 70% of the total incident energy.
down surfaces, which are marked by different colors, as shown in Fig. 5(a). The thickness of the tube is not considered in the optical simulation model. The solar simulator is modeled according to the actual location of the five Xe-arc lamps, including their heights and orientations, and the model of the lamps has been validated by Ref. [24]. The relative average flux distributions at the walls of each turn are simulated by implementing the Monte Carlo ray-tracing method, as displayed in Fig. 5(b). The absolute radiation energy distributed at each turn of the helically coiled tube is displayed in Fig. 6. Two peaks appear in the absolute radiation energy distribution. The first peak is located at approximately the 5th turn because most incident beams directly hit this area. The second peak is between the 12th and the 13th turns, which is caused by the optical splitter and mainly reflects the beams from the middle Xe-arc lamp to the walls of the bottom five turns. The highest radiation energy distribution is 142.77 W at the 5th turn, whereas the least is 80.04 W at the 11th turn. 3.3. Combination of optical and heat transfer processes
3.2. Optical simulation Optical simulation is conducted using the Advanced System Analysis Program (ASAP) software [23], specifically for the simulation of the flux distributions at each turn of the helically coiled tube. To obtain the relative radiation energy distribution at each turn of the coiled tube, the coiled tube is divided into 15 turns continuously, and the outer surface of each turn is divided into up and
The numerical calculation part of the cavity receiver is built using the FLUENT software to combine the optical and heat transfer processes, with consideration of the total incident radiation flux and the absolute radiation energy distribution, where the computational domain includes the helically coiled tube, the glass window, the splitter, and the insulation. This part of the simulation model focuses on thermodynamic performance, including the
Fig. 4. (a) Radiation flux distribution on the aperture; (b) calculation areas with different radius.
851
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858 Table 1 Optical properties of quartz glass. l
kl (lm)
nl
ql
jl (m1)
1 2 3 4 5 6
0.40–2.65 2.65–2.90 2.90–4.20 4.20–7.00 7.00–11.00 11.00–20.00
1.45 1.434 1.42 1.35 1.3 1.3
0.03 0.03 0.03 0.05 0.05 0.07
1 1000 5 5000 5000 5000
(SIMPLE) algorithm is used for a pressure–velocity coupling scheme [31]. The convergence criteria for the residuals of continuity and velocity equations are at an order of 103, whereas those for the energy and DO intensity are at an order of 106. 3.4. Thermal conversion calculation Steady-state energy conservation is given by Fig. 6. Simulation result of absolute radiation energy distribution at each turn.
heating process of the air, temperature distributions, and heat losses. Boundary conditions of the numerical calculation are set, as shown in Fig. 7. The cavity’s internal walls are interfaces of two computational zones, namely, an air flow zone and a zone enclosed by the coiled tube. A coupled thermal boundary is assumed for the cavity’s internal wall in the model. The light radiation distributed on the cavity’s internal walls is regarded as a volumetric power source based on the experimental data of the real flux distribution. The volume flow rate is assumed to be 1–5 m3/h and the inlet temperature at 27 °C. The pressure–outlet boundary condition is also assumed. A mixed thermal boundary condition is assumed for the shell wall, including a convective heat transfer coefficient of 7.67 W/(m2 K) and an emissivity of 0.3 [25]. A mixed thermal boundary condition is also assumed for the glass cover, where the convective heat transfer coefficient is 9.10 W/(m2 K) and the emissivity is 0.94 [26]. Lastly, a semi-transparent boundary condition is adopted for the glass cover in calculating the radiation heat loss. The thermo-physical properties of the air are evaluated using the piecewise linear interpolation method implemented by Shen et al. [27]. The Navier–Stokes equation and the energy and radiation transport equations are numerically solved using the finite– volume method implemented in FLUENT [28]. The discrete ordinates (DO) model is adopted to calculate radiation heat transfer, under complicated radiation conditions, to surfaces and media [29,30]. The semi-implicit method for pressure-linked equations
Q useful ¼ Q in Q ref Q 1;rd Q 1;cv Q 2;rd Q 2;cv Q 3;rd
ð1Þ
where Qref is the reflective loss of incident irradiation and Qabsorbed is the heat carried by the air, which are respectively expressed by
Q ref ¼ qQ in
ð2Þ
Q absorbed ¼ cmðT o T i Þ
ð3Þ
According to the theory of Yang and Tao [32], heat losses of the glass cover are given by
Q 1;cv ¼ Ag hcv 1 ðT g T a Þ
ð4Þ
¼ Ag eg r T 4g T 4a
ð5Þ
Q 1;rd
where hcv 1 is the convection heat transfer coefficient [32] that is given by
hcv 1 ¼
0:54ðGrg Prg Þ0:25 kg L
ð6Þ
where Grg , Prg , and kg are variables related to a reference temperature T rt1 that is expressed as
T rt1 ¼
Tg þ Ta 2
ð7Þ
The outer surface of the shell includes three parts, i.e., up, side, and bottom surfaces, as shown in Fig. 2(b), and the losses are calculated by
Q 2;cv ¼ Q 21;cv þ Q 22;cv þ Q 23;cv
Fig. 7. Boundary conditions in the numerical model.
ð8Þ
852
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Fig. 8. Schematic of up and down-flows.
Fig. 9. Temperature distributions along flow direction for simulation and experiments (down-flow).
Q 2n;cv ¼ A2n hcv 2n ðT 2n T a Þ ðn ¼ 1; 2; 3Þ
Table 2 The radiation energy distribution at the 15 turns coils. Serial number of turns Radiation energy distribution (W)
1 96
2 130
3 134
4 139
5 143
6 138
7 129
Serial number of turns Radiation energy distribution (W)
9 98
10 81
11 80
12 123
13 130
14 104
15 85
Q 2;rd ¼ Q 21;rd þ Q 22;rd þ Q 23;rd Q 2n;rd ¼ A2n esh r T 42n T 4a ðn ¼ 1; 2; 3Þ
8 115
ð9Þ ð10Þ
ð11Þ
hcv 21 ¼
0:54ðGr21 Pr 21 Þ0:25 k21 L21
ð12Þ
hcv 22 ¼
0:59ðGr22 Pr 22 Þ0:25 k22 L22
ð13Þ
hcv 23 ¼
0:27ðGr23 Pr 23 Þ0:25 k23 L23
ð14Þ
where Gr23 , Pr23 , and k23 are variables related to a reference temperature T rt2n that is expressed as
853
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Fig. 10. Temperature distribution along flow direction for simulation and experiments (up-flow).
Fig. 11. Temperature distribution on the coiled tube at flow rate of 5 m3/h in the (a) down-flow and (b) up-flow.
T rt2n ¼
T 2n þ T a 2
ð15Þ
s¼
6 X
ðF bð0klþ1 Þ F bð0kl Þ Þejl s
ð17Þ
l¼1
The inner surface of the cavity includes the outer wall of the turns and the optical splitter. The energy balance equation of the inner surface of the cavity is given by
Q 3;rd ¼ Q 31;rd þ Q 32;rd
ð16Þ
With the presence of a glass cover, re-radiation of the inner surface of the cavity is reflected, absorbed, and transmitted. The transmitted part is determined by the transmissivity s of the glass cover [32], which is calculated by
where F bð0kl Þ is the black body radiation function, jl is the absorption coefficients of different wave ranges that are identified in Table 1 [33], and l is the corresponding numbers of the different wave ranges.
Q 31;rd ¼
15 X
r T 431;k T 4g
k¼1
ec A k
sk 1ec
þ x 1Ag þ k
1eg eg A g
ð18Þ
854
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Fig. 12. Absorbed heat and heat losses of the surfaces of the receiver at different air flow rates (down-flow).
r T 4sp T 4g
Q 32;rd ¼ ssp 1esp
esp Asp
þ xsp1Ag þ
1eg eg A g
ð19Þ
Table 3 Outlet temperature of the simulation. Simulation results in the down-flow (°C) Average flux (kW/m2)
where sk is the transmissivity of the glass cover for the No. k turn, ssp is the radiation transmissivity of the glass cover for the optical splitter, T 31;k and T sp are the temperatures of the No. k turn and the optical splitter, respectively, ec and esp are the emissivities of the coiled tube and the optical splitter, respectively, xk and xsp are the angle factors from each turn and the optical splitter to the glass, respectively, and Ak and Asp are areas of the No. k turn and the optical splitter, respectively.
4. Results and discussion 4.1. Model validation Experiments were carried out using up and down-flows, as shown in Fig. 8, to study the effects of flow rates and flow directions on the outlet temperature. Simulation work was conducted by coupling optical and thermal processes, and the results of the experiments and simulations are discussed in this section.
Fig. 13. Comparison of heat losses at different parts in the down and up-flows.
120 140 160 180 200 250 300 350
Flow rate (m3/h) 1
2
3
4
5
609 634 669 698 730 796 844 891
602 629 663 694 725 788 839 887
592 626 661 692 721 784 834 880
577 616 653 686 716 780 829 876
564 601 640 674 706 772 818 866
Fig. 9 shows the air temperature along the flow direction of down-flows (1–5 m3/h) under quasi-equilibrium states. The air temperature increased rapidly at the first five turns of the coiled tube and then decreased at the last one turn according to the absolute radiation energy distribution on the coiled tube. The radiation energy distribution at the 15 turns coils are displayed, as in Table 2. The most incident energy was absorbed by the upper half cavity, while less by the last turns. The wall temperature of the last turns was lower than the upstream air, leading to a temperature drop of air. The air flow rate had a significant effect on the thermal performance of the receiver. As the flow rate increased from 1 m3/h to 5 m3/h in the experiments, the air outlet temperature decreased from 662 °C to 570 °C and the thermal efficiency increased from 12.5% to 53.6%. The maximum temperature drop at the last turns was around 64 °C at 1 m3/h, whereas the minimum was around 21 °C at 5 m3/h. The temperature drop of air at the last turns was sensitive to the flow rate, indicating the outlet temperature and efficiency could be optimized by adjusting the flow rate. The temperature change in simulation was in agreement with experiment and the deviations of outlet temperature between the experimental and simulation results ranged from 8% to 2%. Fig. 10 shows the air temperatures along the up-flows. The temperature change trend was similar to the down-flows and the two curves of experimental and simulation results also matched well. The air flow rate plays an important role in the thermal performance of the receiver. As the flow rate increased from 1 m3/h to 5 m3/h in the experiments, the outlet temperature decreased from 593 °C to 546 °C and the thermal efficiency increased from 10.9% to 51.4%. The temperature drop at the last turns along flow direction ranged from 65 °C at 1 m3/h to 14 °C at 5 m3/h. As the flow rate
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
855
Fig. 14. Absolute flux distribution at each turn for different inner radii.
low increase in air temperature rate from the 6th turn. In the up-flows, a rapid increase in air temperature rate occurs at the first 12 turns, and the stationary stable phase is relatively shorter than that in the down-flows. The outlet temperature and the thermal efficiency in the down-flows are always higher than those in the up-flows in the experiments. These results are also demonstrated by the simulation model. The uncertainty of measurement of the air temperature, flow rate, and radiation flux are approximately at 1.0% [20], 2.0% [18] and 3.0% [7], respectively. Because the radiation power on aperture is composed of five lamp sources, the uncertainty of the radiation power on aperture is 6.7% [7]. The uncertainty of the reflectivity involved in the optical simulation is 3.0% [7]. Thus, the experimental uncertainty of the total incident radiation, U Q in , and thermal efficiency, U gt , are estimated to be 7.3% and 7.6%, respectively [7], using the following expressions,
U Q in ¼ Fig. 15. Temperature distribution along flow direction for different inner radii.
U gt ¼ increased from 1 m3/h to 5 m3/h, the outlet temperatures in the simulation decreased from 603 °C to 559 °C. The overall trend of the simulated air temperature coincided with the experimental results as well and the deviations of air outlet temperature between the experiments and the simulation values ranged from 0.3% to 2.3%. The air temperature heating processes are similar under down and up-flow conditions. A rapid increase caused by the relatively low specific heat capacity of the air occurs initially, and then the temperature drops at the end because of the relatively low radiation energy distribution. The different flow directions bring a different flux distribution in the air flow, but the same radiation energy distribution at the internal walls. In the down-flows, air flows through a high-flux area at first, leading to a higher increase in temperature rate at the first turns than in the up-flows, and then a low-flux area is developed at the end of the flow, resulting in a
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi U 2source þ U 2reflectiv ity
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi U 2Q in þ U 2thermocouple þ U 2flow
ð20Þ ð21Þ
where U source is the uncertainty of the radiation power on aperture, U reflectiv ity is the uncertainty of the measured reflectivity, U thermocouple and U flow are the uncertainty of the air temperature and flow rate, respectively. 4.2. Model analysis and parametric investigation 4.2.1. Temperature distribution on the outer wall of the coiled tube The temperature and radiation energy distributions on the cavity internal walls, which are useful to for identifying ‘‘hot spots’’ and evaluating thermal stresses and heat losses. They can be obtained using the comprehensive model. As shown in Fig. 11, the outer wall temperature initially increases along the air flow direction and then decreases at the last turns, in accordance with the change of air temperature in the tube. The wall temperature increased from around 257 °C to 577 °C along down-flow direction
856
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
while it decreased to around 517 °C at the last turns. In the up-flow configuration, the wall temperature increased from around 197 °C to 547 °C along the flow direction, and then dropped around 497 °C finally. Fig. 11 also shows that the hottest region of the coiled tube is close to the glass window in the up-flow configuration; thus, significant heat losses are observed because of conduction and re-radiation through the glass window. This is the main reason for the outlet temperature in the down-flows configuration to be always higher than that in up-flows under similar conditions. 4.2.2. Heat losses and discussion Fig. 12 shows the proportion of absorbed energy by the air and the heat losses of different surfaces of the receiver in the down-flow condition, which are normalized by the total concentrated incident irradiation power Qin = 1846 W. Qabsorbed represents the thermal efficiency, which has a strongly positive correlation with the flow rate. The reflection loss of the glass cover, Qref, is around 2.94%. Generally, all kinds of heat losses increased with the decrease of air flow rate. When the air flow rate reduced from 5 m3/h to 1 m3/h, radiation losses of the shell outer
surface to the environment, i.e. the total of Q21,rd, Q22,rd and Q23,rd, ranged from14.78% to 29.89%; convection losses of the shell outer surface to the environment, i.e. the total of Q21,cv, Q22,cv and Q23,cv, was from 18.11% to 28.57%; radiation losses of the glass cover to the environment, Q1,rd, was from 4.64% to 10.52%; convection losses of the glass cover to the environment, Q1,cv, was from 2.41% to 3.75%, and the radiation loss of the inner surface of receiver to the environment, i.e. the total of Q31,rd and Q32,rd, was from 4.27% to 9.83%. The heat losses of the shell outer surface account for the highest proportion, indicating that the insulation should be improved in future studies. A cooling system is suggested to reduce the heat losses closed to the glass window. The total heat losses increased sharply with the decrease in flow rate, e.g., 44.21% for t = 5 m3/h and 82.51% for t = 1 m3/h. Fig. 13 compares the heat losses at different parts in the down and up-flows. The heat losses of the glass cover and inner surfaces in the down-flows are less than those in the up-flows because the temperature at the region near the glass window is lower in the down-flows, whereas higher in the up-flows, thus more heat losses. Furthermore, heat losses of the outer surface in the
Fig. 16. Absolute flux distributions at each turn for different numbers of turns.
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
857
Fig. 17. Temperature distribution along the flow direction for different total turns: (a) 12–15 turns; (b) 8–11 turns.
down-flows are more than those in the up-flows because the temperature of the cavity’s internal walls in the down-flows are higher than that in the up-flows, resulting in more conductive heat loss through insulation. Generally, the total heat losses in the down-flows are lower by 11 W (minimum) for t = 5 m3/h and by 48 W (maximum) for t = 1 m3/h than those in the up-flows. 4.2.3. Influence of incident radiation flux Table 3 lists the outlet temperatures forecasted by the comprehensive simulation model when the average flux incident on the glass cover ranges from 120 kW/m2 to 350 kW/m2. The outlet temperature can be more than 700 or 800 °C if the average flux incident reaches 200 or 300 kW/m2, respectively, for an air flow rate of less than 5 m3/h. The high air temperature at the expense of high heat losses, i.e., low thermal efficiency. At 200 and 300 kW/m2, the thermal efficiency is 39.93% and 31.01% for outlet temperatures of 706 and 818 °C, respectively, at 5 m3/h flow rate. 4.2.4. Influence of inner radius A smaller inner radius of a tube produces a higher velocity at the same flow rate, increasing the Nu number and strengthening the heat transfer. The effect of the inner radius on the receiver’s thermal performance is studied based on the comprehensive simulation model. Fig. 14 displays the flux distributions on the coiled tube as the inner radius decreases from 5.5 mm to 4.0 mm, where the height of the coiled tube decreased and the optical splitter was moved up, causing more light was distributed by the optical splitter. It is noticed that more incident radiation is reflected onto the last turns, leading higher surface temperatures of the last turns and the air outlet temperature. For example, the radiation energy at the last one turn can be increased from 85 W for 6 mm-tube conditions to 106 W for 4 mm-tube conditions. At an air flow rate of 5 m3/h, as shown in Fig. 15, the temperature increases as the inner radius decreases, regardless of the temperature drop at the bottom tube. The outlet temperature and the thermal efficiency increase with the decrease in the inner radius, and the outlet temperature reaches up to 670 °C with a thermal efficiency of 64.1% when the inner radius is 4 mm. The pressure drop increases sharply from 9.63 kPa for a 5.5 mm inner radius to 56.52 kPa for a 4.0 mm inner radius, while 5.37 kPa for a 6.0 mm inner radius. 4.2.5. Influence of the number of turns The temperature always increases slightly at the 9th to the 13th turns and then significantly drops at the last three turns in the down-flows. To optimize the length of the tube, the effect of the total turns on the receiver’s thermal performance is studied based
on the comprehensive simulation model at an air flow rate of 5 m3/h under the down-flow conditions. The height of the coiled tube decreases and the optical splitter moves upward with the decrease in turns. Fig. 16 shows the absolute radiation energy distributions for the number of turns ranging from 14 to 8. The height of the coiled tube decreased and the optical splitter was moved upward with the decreasing of turns. Thus the flux reflected by the splitter increased and more incident irradiation was absorbed by the bottom turns of air cavity receiver, which may lead to a higher air outlet temperature. The maximum radiation energy at each turn can be increased from 149 W for 14 turns to 243 W for 8 turns, while it is 142 W for 15 turns. However, due to the decreasing of turns, the spillage of incident light increases, which leads to the total radiation energy distribution on the internal walls decreases from 1.82 kW for 14 turns to 1.68 kW for 8 turns, while 1.85 kW for 15 turns. It is also noted the heat transfer area is reduced with the decreasing of turns. Fig. 17(a) shows the air temperature distribution along the air flow direction for 12–15-turn tubes. The temperature changing trends are similar at the first six turns. The air temperature fluctuates at the 7th to the 10th turns because of the flux distribution, and the final outlet temperature is from 559 °C to 572 °C. The pressure drop decreases from 5.37 kPa for 15 turns to 4.62 kPa for 12 turns. Fig. 17(b) shows the air temperature distribution along the air flow direction for 8–11-turn tubes. The outlet temperature reaches up to 568 °C for the 10-turn tube. The outlet temperature of the 9-turn and 8-turn tubes decreases to 554 °C and 542 °C, respectively, because of the decrease in the total flux distribution and the heating time. The pressure drop decreases from 4.16 kPa for 11 turns to 3.61 kPa for 8 turns. Therefore the air outlet temperature and pressure drop should be taken into consideration for optimizing the number of turns of air tube-cavity solar receiver.
5. Conclusions An air tube-cavity solar receiver has been designed and fabricated to generate high temperature air from solar energy. The maximum air outlet temperature can reach up to 662 °C and 593 °C in the down-flow and up-flows configurations, respectively. A comprehensive model coupling the optical and the heat transfer processes has been developed. The model has been validated using the measured data. The difference for the outlet temperature between the measurement and the model prediction is less than 8%. The model predicts that the outlet air temperature can reach up to about 800 °C for an air flow rate of 5 m3/h and average radiation flux of 300 kW/m2. The thermal efficiency of the receiver can reach to 64.1% for an air outlet temperature of 670 °C if the inner
858
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
tube radius is reduced to 4 mm at the cost of the increasing pressure drop. The outlet temperature at 5 m3/h air flow can reach 572 °C if a 12-turn air tube-cavity receiver is utilized, which has a suitable flux distribution under the same operating conditions. The comprehensive model can also provide detailed analysis about thermal performances and point out effective ways to improve performances, including reducing heat loss, optimizing structural parameters. Acknowledgements The authors acknowledge the support from the National Natural Science Foundation of China (No. 51476140), the Program of Introducing Talents of Discipline to University (No. B08026), and Seed Funds of Interdisciplinary Research for Young Teachers of Zhejiang University (JCZZ-2013014). References [1] Behar O, Khellaf A, Mohammedi K. A review of studies on central receiver solar thermal power plants. Renew Sustain Energy Rev 2013;23:12–39. [2] Yang M, Yang X, Yang X, et al. Heat transfer enhancement and performance of the molten salt receiver of a solar power tower. Appl Energy 2010;87(9):2808–11. [3] Muñoz-Anton J, Biencinto M, Zarza E, et al. Theoretical basis and experimental facility for parabolic trough collectors at high temperature using gas as heat transfer fluid. Appl Energy 2014;135:373–81. [4] http://www.nrel.gov/csp/solarpaces/project_detail.cfm/projectID=62. [5] Biencinto M, González L, Zarza E, et al. Performance model and annual yield comparison of parabolic-trough solar thermal power plants with either nitrogen or synthetic oil as heat transfer fluid. Energy Convers Manage 2014;87:238–49. [6] Ávila-Marín AL. Volumetric receivers in solar thermal power plants with central receiver system technology: a review. Solar Energy 2011;85(5):891–910. [7] Xiao G, Guo K, Luo Z, et al. Simulation and experimental study on a spiral solid particle solar receiver. Appl Energy 2014;113:178–88. [8] Bai F, Zhang Y, Zhang X, et al. Thermal performance of a quartz tube solid particle air receiver. Energy Proc 2014;49:284–94. [9] Zavattoni SA, Gaetano A, Montorfano D, et al. A novel CSP receiver based on AirLight energy technology – optimization of the thermal insulation system by means of CFD analysis. Energy Proc 2014;49:589–98. [10] Amsbeck L, Buck R, Heller P, et al. Development of a tube receiver for a solarhybrid microturbine system. In: Proceedings of 14th SolarPACES conference, Las Vegas, NV, March; 2008.
[11] Wang W, Gianmarco R, Lucas A, et al. Integrated design of a hybrid gas turbine-receiver unit for a solar dish system. In: Proceedings of 20th SolarPACES conference, Beijing, China, September; 2014. [12] Korzynietz R, Torreira MG. Soltrec-pressurized volumetric solar air receiver technology. Energy Proc 2014;49:284–94. [13] Hischier I, Hess D, Lipin´ski W, et al. Heat transfer analysis of a novel pressurized air receiver for concentrated solar power via combined cycles. J Therm Sci Eng Appl 2009;1(4):041002. [14] Kribus A, Doron P, Rubin R, et al. A multistage solar receiver: the route to high temperature. Solar Energy 1999;67(1):3–11. [15] Wang F, Shuai Y, Tan H. Thermal analyse of cavity solar receiver. J Eng Thermophys 2011;32(5):843–6. [16] Paitoonsurikarn S, Lovegrove K. On the study of convection loss from open cavity receivers in solar paraboloidal dish applications. In: Proceedings of 41st conference of the Australia and New Zealand Solar Energy Society (ANZSES), Melbourne, Australia; 2003. [17] Prakash M, Kedare SB, Nayak JK. Investigations on heat losses from a solar cavity receiver. Solar Energy 2009;83(2):157–70. [18] http://www.shy-sz.com/index.asp. [19] http://www.agilent.com/home. [20] http://cn.omega.com/section/thermocouples.html?utm_source=baidu&utm_ medium=cpc&utm_term=k%E5%9E%8B%E7%83%AD%E7%94%B5%E5%81%B6& utm_campaign=BSTYWD. [21] http://www.ophiropt.com/laser-measurement. [22] de Jager C. The solar spectrum. The solar spectrum; 1965. [23] http://www.breault.com/software/about-asap. [24] Xiao G, Guo K, Ni M, et al. Optical and thermal performance of a hightemperature spiral solar particle receiver. Solar Energy 2014;109(109):200–13. [25] Price DJ. The emissivity of hot metals in the infra-red. Proc Phys Soc 1947;59(1):118. [26] Dvurechensky A, Petrov V, Reznik VY. Spectral emissivity and absorption coefficient of silica glass at extremely high temperatures in the semitransparent region. Infrared Phys 1979;19(3):465–9. [27] Shen W, Jiang Z, Tong G, et al. Engineering thermodynamics. Higher Education Press; 2001. [28] ANSYS.FLUENT – theory guide; 2009. [29] Raithby GD, Chui EH. Finite-volume method for predicting a radiant heat transfer in enclosures with participating media. J Heat Transfer 1990;112(2):415–23. [30] Wang F. Analysis and calculation of fluid dynamics: principle and application of CFD software. Tsinghua University Press Ltd.; 2004. [31] Ferziger JH, Peric´ M. Computational methods for fluid dynamics. Berlin: Springer; 2002. [32] Yang S, Tao WT. Heat transfer. 4th ed. Beijing: High Education Press; 2006. [33] Tan H. Numerical simulation of infrared radiation characteristics and transmission. Harbin Institute of Technology press; 2006.
Contents lists available at ScienceDirect
Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman
Simulation and experimental study of an air tube-cavity solar receiver Kunzan Qiu, Liang Yan, Mingjiang Ni, Cheng Wang, Gang Xiao ⇑, Zhongyang Luo, Kefa Cen State Key Laboratory of Clean Energy Utilization, Zhejiang University, Hangzhou 310027, China
a r t i c l e
i n f o
Article history: Received 20 March 2015 Accepted 4 July 2015 Available online 20 July 2015 Keywords: Solar receiver High temperature Air Tube cavity Simulation model
a b s t r a c t High temperature air is a potential candidate as a heat transfer fluid to transport energy from concentrated solar power to gas turbines. A 15-turn helically coiled tube cavity receiver with an optical splitter at the bottom is designed and fabricated. Its performance is investigated with a five 7-kW Xe-arc lamps array system as heat source. Eight K-type thermocouples are placed from top to bottom with an equal interval. The outlet temperature experimentally ranges from 593 °C to 546 °C when the air flow rate increases from 1 m3/h to 5 m3/h for up-flows, while it ranges from 662 °C to 570 °C for down-flows, when the average flux on aperture is around 120 kW/m2. The Monte-Carlo ray-tracing method and the Lambert testing method with a charge-coupled device (CCD) camera are used to simulate and evaluate the concentrating radiation energy distribution on the cavity’s internal walls, and then the actual flux distribution of each turn of the helically coiled tube is obtained. A comprehensive simulation model is proposed and validated by the experimental results, where the outlet temperature deviations are within 8.0% and 2.5% for down and up-flows, respectively. The model provides a detailed analysis of heat flows at different conditions, and indicates optimization ways to improve the efficiency and reduce heat losses. The simulation results show that the outlet temperature can increase up to around 800 °C at 5 m3/h under an average flux of 300 kW/m2, and the thermal efficiency can be improved from around 56% to around 64% by decreasing the inner radius from 6 mm to 4 mm at the expense of increasing pressure drop of around 56 kPa. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction To alleviate the energy crisis, many countries have considered solar energy as an alternative because it is clean, renewable, and abundant [1]. Concentrating solar power (CSP) has already been applied commercially for power generation [2]. It is well known that high electrical efficiency of the CSP system is highly dependent on being able to achieve a high working temperature of the heat transfer fluid (HTF) [3]. Synthetic oil has been widely utilized in solar trough plants, currently reaching around 4 GW [4] with the maximum working temperature around 400 °C [5]. Molten salt is another material to increase the working temperature up to around 560 °C [6]. Further increase in the temperature of those working fluids is difficult due to the constraints of physical properties. Air receivers have been proposed as an alternative HTF to achieve higher working temperatures in order to obtain higher efficiency cycles for systems such as Brayton cycle [7]. The Institute of Electrical Engineering of Chinese Academy of Science developed an
⇑ Corresponding author. Tel.: +86 571 87953290; fax: +86 571 87951616. E-mail address: [email protected] (G. Xiao). http://dx.doi.org/10.1016/j.enconman.2015.07.013 0196-8904/Ó 2015 Elsevier Ltd. All rights reserved.
air intake tube receiver with bubbling particle. Concentrated solar radiation was mostly absorbed by particles and converted into thermal energy which was carried out by air flow. The maximum air outlet temperature reached 624 °C [8]. AirLight Energy installed an air-based receiver for solar trough concentrators. The receiver is 212 m long, consisting of 4608 cavities made of helically coiled tubes [9]. The receiver can deliver hot air at a temperature around 650 °C under a concentration ratio of 98 when the inlet air temperature is 120 °C. The German Aerospace Center (DLR) developed an air receiver consisting of 40 absorber tubes arranged in a cavity and connected in parallel. The outlet air temperature reached up to 800 °C at 4.5 bar, with the efficiency of the solar-hybrid micro-turbine system reaching 44% [10]. The Sweden Royal Institute of Technology (KTH) designed a gas turbine-receiver unit, in which the inlet air was first compressed by a compressor, preheated by a regenerator, and then heated by a solar receiver. The air outlet temperature reached 950 °C [11]. Abengoa Solar developed a volumetric solar air receiver. The maximum air outlet temperature reached up to 1000 °C [12]. The Swiss Federal Institute of Technology in Zurich (ETH) developed a pressurized solar receiver whose main component was a cylindrical silicon carbide cavity surrounded by a concentric annular reticulate porous ceramic
848
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Nomenclature Q in Q ref Q 1,rd Q 2,rd Q 1,cv Q 2,cv Q 3,rd hcv1 hcv21 hcv22 hcv23 s
q Ti To Ag
incident irradiation on the receiver (W) reflective loss of incident irradiation (W) heat radiation loss between the glass cover and the atmosphere (W) heat radiation loss between the shell wall and the atmosphere (W) heat convection loss between the glass cover and the atmosphere (W) heat convection loss between the shell wall and the atmosphere (W) heat radiation loss between the cavity inner wall and the atmosphere (W) convection heat transfer coefficient between the glass cover and the atmosphere (W/(m2 K)) convection heat transfer coefficient between the shell up surface and the atmosphere (W/(m2 K)) convection heat transfer coefficient between the shell side surface and the atmosphere (W/(m2 K)) convection heat transfer coefficient between the shell bottom surface and the atmosphere (W/(m2 K)) thickness of the glass cover reflectivity of the glass cover inlet air temperature (K) air outlet temperature (K) area of the glass cover (m2)
(RPC) foam. The outlet temperature of the air flowing through the RPC reached up to 1060 °C at 5 bar under an input solar power of 32–38 kW [13]. The Weizmann Institute of Science divided the aperture of the solar receiver into separate stages according to irradiance distribution to minimize heat losses. The air temperature was initially raised up to 750 °C by preheaters, i.e., cavity tubes. Subsequently, a directly irradiated annular pressurized receiver (DIAPR) was used. The receiver was capable of supplying hot air at a temperature of 1300 °C under a pressure of 10–30 bar [14]. The first demonstration of the use of air as HTF in a solar power plant was through a 1.5 MW tower power system with a ceramic volumetric receiver in Jülich, Germany. The receiver produced hot air with a temperature of 680 °C and was in operation successfully in September 2008 [6]. For scaling up and design optimization purposes, a reliable comprehensive predictive model describing performance of the air solar receiver is highly desired. Wang et al. [15] used a 3D numerical model with a uniform flux distribution on tube surface for a coiled tube receiver to analyze temperature distribution and thermal stress of the tube. Paitoonsurikarn et al. [16] numerically investigated the effects of different shapes on the convective losses of solar cavity receivers by assuming an isothermal wall temperature. Prakash et al. [17] computed convective heat losses of a cylindrical cavity receiver under different inclination angles assuming an adiabatic boundary condition on cavity external wall. The foregoing modeling and simulation efforts focused on analyzing the effects of various heat transfer modes. However the incident radiation flux on the cavity wall has often been simplified by assuming a uniform or arbitrarily varying distribution. Since the radiation flux is the most important factor for characterizing the solar receiver performance, a realistic model for the distribution of the radiation heat flux is necessary. In this work, a generic air tube cavity solar receiver is considered. Monte-Carlo ray tracing technique is used to accurately obtain the realistic radiation flux distribution on the air tube external wall in order to build a comprehensive simulation model coupling optical and heat transfer processes of high-temperature
Q absorbed energy carried away by the air (W) L characteristic length (m) Q 21,cv heat convection loss between the shell up surface and the atmosphere (W) Q 22,cv heat convection loss between the shell side surface and the atmosphere (W) Q 23,cv heat convection loss between the shell bottom surface and the atmosphere (W) Q 21,rd heat radiation loss between the shell up surface and the atmosphere (W) Q 22,rd heat radiation loss between the shell side surface and the atmosphere (W) Q 23,rd heat radiation loss between the shell bottom surface and the atmosphere (W) Q 31,rd heat radiation loss between the outer wall of the coils and the atmosphere (W) Q 32,rd heat radiation loss between the optical splitter and the atmosphere (W) Tg average glass temperature (K) Ta average atmosphere temperature (K) eg glass cover emissivity r the Stefan–Boltzman constant
air solar receiver, which considerably affects the distributions of wall temperature and the heat loss at the external walls. A 15-turn helically coiled tube cavity receiver with an optical splitter at the bottom is designed and fabricated in order to assess the efficacy of the comprehensive model. The experiments on the cavity receiver were performed under a five 7-kW Xe-arc lamps array system. Good agreements were obtained between the measurement and the model calculation. The comprehensive simulation model is subsequently used to conduct a parametric investigation to understand the performance of the cavity receiver under different operating conditions and to find ways to optimize solar air receiver performance. 2. Experimental set-up and methods A solar simulator consisting of five 7-kW Xe-arc lamps, as shown in Fig. 1, is employed to heat an air tube-cavity receiver. An experimental flowchart is shown in Fig. 2(a). Air is pumped into
Fig. 1. Solar simulator with five 7 kW Xe-arc lamps.
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
849
Fig. 2. (a) Experimental flowchart; (b) configuration of cavity receiver; (c) surfaces of the receiver (1: surface of the glass cover; 21: shell up surface; 22: shell side surface; 23: shell down surface; 31: the outer wall surfaces of the coils; 32: the outer surfaces of the optical splitter).
a coiled tube by a blower and the flow rate, ranging from 1 m3/h to 5 m3/h, is adjusted by a rotor flow meter (LZB-25), manufactured by Chemical Instrument Company of China [18]. A data acquisition system (Aglient 34970A), produced by Aglient Company of U.S. [19], linked with 8 K-type thermocouples, records air temperatures along the coiled tube every 10 s. A configuration of the air tube-cavity receiver is shown in Fig. 2(b). The solar cavity receiver consists of a 15-turn helically coiled copper tube with inner diameter of 1.2 cm and thickness of 0.1 cm. The inner diameter and height of the cavity are 14 and 25 cm, respectively. Eight K-type thermocouples, produced by Omega Company of U.S. [20], are placed from the top to the bottom with uniform spacing. The aperture of the cavity is covered with a quartz window. At the bottom is an optical splitter, i.e., stainless steel cone with a height of 7.5 cm and a bottom diameter of 14 cm. The tube coils are wrapped with thermal insulating materials. Fig. 2(c) shows different surfaces of the receiver, which will be considered in the calculation of heat losses. 3. Comprehensive modeling of the cavity receiver The comprehensive simulation model of the air tube-cavity receiver consists of three parts: (1) evaluation of the total incident
radiation on the aperture, (2) optical simulation to obtain relative flux distribution, and (3) combination of optical and heat transfer processes using the FLUENT software.
3.1. Evaluation of the total incident radiation The evaluation of the total incident radiation has two important phases: (1) detection of a relative flux distribution of concentrated radiation on the aperture and (2) detection of the absolute radiation intensity at a reference point. The measurement system includes a Lambert plate, a charge-coupled device (CCD) camera, and a radiometer, among others, as shown in Fig. 3. The Lambert plate, a standard diffuse reflection white board produced by Labsphere Company of U.S., is a 25 cm 25 cm square. The Lambert plate is placed at the aperture of the cavity. The CCD camera, manufactured by Hamamatsu Company, Japan, is used to capture the concentrating spots on the Lambert plate. The output video signal has a linear relationship with the received luminance distribution. A radiometer (30 (150) A-BB-18) with a 17 mm-diameter aperture, manufactured by Ophir Company of Israel [21], is used to detect the absolute radiation intensity of reference points on the Lambert plate. The power
850
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Fig. 5. (a) Model of the coiled tube; (b) schematic of ray tracing.
Fig. 3. Detecting devices and method.
sensor has an accuracy of 3.0% at the 0.19–20.00 lm wavelength range [7] (this range is common in most solar lights) [22]. Fig. 4(a) displays a relative flux distribution of the concentrating radiation on the Lambert plate, where the flux intensity increases rapidly toward the center. The calculation area of the concentrating radiation on the Lambert plate is separated into two regions, as shown in Fig. 4(b). A1 and A2 represent the inside and outside areas, respectively, of the glass aperture. The calculation result shows that the average flux intensities on A1 and A2 are 120.63 and 77.62 kW/m2, respectively, and the concentrating radiation on A1 is 1.85 kW, nearly 70% of the total incident energy.
down surfaces, which are marked by different colors, as shown in Fig. 5(a). The thickness of the tube is not considered in the optical simulation model. The solar simulator is modeled according to the actual location of the five Xe-arc lamps, including their heights and orientations, and the model of the lamps has been validated by Ref. [24]. The relative average flux distributions at the walls of each turn are simulated by implementing the Monte Carlo ray-tracing method, as displayed in Fig. 5(b). The absolute radiation energy distributed at each turn of the helically coiled tube is displayed in Fig. 6. Two peaks appear in the absolute radiation energy distribution. The first peak is located at approximately the 5th turn because most incident beams directly hit this area. The second peak is between the 12th and the 13th turns, which is caused by the optical splitter and mainly reflects the beams from the middle Xe-arc lamp to the walls of the bottom five turns. The highest radiation energy distribution is 142.77 W at the 5th turn, whereas the least is 80.04 W at the 11th turn. 3.3. Combination of optical and heat transfer processes
3.2. Optical simulation Optical simulation is conducted using the Advanced System Analysis Program (ASAP) software [23], specifically for the simulation of the flux distributions at each turn of the helically coiled tube. To obtain the relative radiation energy distribution at each turn of the coiled tube, the coiled tube is divided into 15 turns continuously, and the outer surface of each turn is divided into up and
The numerical calculation part of the cavity receiver is built using the FLUENT software to combine the optical and heat transfer processes, with consideration of the total incident radiation flux and the absolute radiation energy distribution, where the computational domain includes the helically coiled tube, the glass window, the splitter, and the insulation. This part of the simulation model focuses on thermodynamic performance, including the
Fig. 4. (a) Radiation flux distribution on the aperture; (b) calculation areas with different radius.
851
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858 Table 1 Optical properties of quartz glass. l
kl (lm)
nl
ql
jl (m1)
1 2 3 4 5 6
0.40–2.65 2.65–2.90 2.90–4.20 4.20–7.00 7.00–11.00 11.00–20.00
1.45 1.434 1.42 1.35 1.3 1.3
0.03 0.03 0.03 0.05 0.05 0.07
1 1000 5 5000 5000 5000
(SIMPLE) algorithm is used for a pressure–velocity coupling scheme [31]. The convergence criteria for the residuals of continuity and velocity equations are at an order of 103, whereas those for the energy and DO intensity are at an order of 106. 3.4. Thermal conversion calculation Steady-state energy conservation is given by Fig. 6. Simulation result of absolute radiation energy distribution at each turn.
heating process of the air, temperature distributions, and heat losses. Boundary conditions of the numerical calculation are set, as shown in Fig. 7. The cavity’s internal walls are interfaces of two computational zones, namely, an air flow zone and a zone enclosed by the coiled tube. A coupled thermal boundary is assumed for the cavity’s internal wall in the model. The light radiation distributed on the cavity’s internal walls is regarded as a volumetric power source based on the experimental data of the real flux distribution. The volume flow rate is assumed to be 1–5 m3/h and the inlet temperature at 27 °C. The pressure–outlet boundary condition is also assumed. A mixed thermal boundary condition is assumed for the shell wall, including a convective heat transfer coefficient of 7.67 W/(m2 K) and an emissivity of 0.3 [25]. A mixed thermal boundary condition is also assumed for the glass cover, where the convective heat transfer coefficient is 9.10 W/(m2 K) and the emissivity is 0.94 [26]. Lastly, a semi-transparent boundary condition is adopted for the glass cover in calculating the radiation heat loss. The thermo-physical properties of the air are evaluated using the piecewise linear interpolation method implemented by Shen et al. [27]. The Navier–Stokes equation and the energy and radiation transport equations are numerically solved using the finite– volume method implemented in FLUENT [28]. The discrete ordinates (DO) model is adopted to calculate radiation heat transfer, under complicated radiation conditions, to surfaces and media [29,30]. The semi-implicit method for pressure-linked equations
Q useful ¼ Q in Q ref Q 1;rd Q 1;cv Q 2;rd Q 2;cv Q 3;rd
ð1Þ
where Qref is the reflective loss of incident irradiation and Qabsorbed is the heat carried by the air, which are respectively expressed by
Q ref ¼ qQ in
ð2Þ
Q absorbed ¼ cmðT o T i Þ
ð3Þ
According to the theory of Yang and Tao [32], heat losses of the glass cover are given by
Q 1;cv ¼ Ag hcv 1 ðT g T a Þ
ð4Þ
¼ Ag eg r T 4g T 4a
ð5Þ
Q 1;rd
where hcv 1 is the convection heat transfer coefficient [32] that is given by
hcv 1 ¼
0:54ðGrg Prg Þ0:25 kg L
ð6Þ
where Grg , Prg , and kg are variables related to a reference temperature T rt1 that is expressed as
T rt1 ¼
Tg þ Ta 2
ð7Þ
The outer surface of the shell includes three parts, i.e., up, side, and bottom surfaces, as shown in Fig. 2(b), and the losses are calculated by
Q 2;cv ¼ Q 21;cv þ Q 22;cv þ Q 23;cv
Fig. 7. Boundary conditions in the numerical model.
ð8Þ
852
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Fig. 8. Schematic of up and down-flows.
Fig. 9. Temperature distributions along flow direction for simulation and experiments (down-flow).
Q 2n;cv ¼ A2n hcv 2n ðT 2n T a Þ ðn ¼ 1; 2; 3Þ
Table 2 The radiation energy distribution at the 15 turns coils. Serial number of turns Radiation energy distribution (W)
1 96
2 130
3 134
4 139
5 143
6 138
7 129
Serial number of turns Radiation energy distribution (W)
9 98
10 81
11 80
12 123
13 130
14 104
15 85
Q 2;rd ¼ Q 21;rd þ Q 22;rd þ Q 23;rd Q 2n;rd ¼ A2n esh r T 42n T 4a ðn ¼ 1; 2; 3Þ
8 115
ð9Þ ð10Þ
ð11Þ
hcv 21 ¼
0:54ðGr21 Pr 21 Þ0:25 k21 L21
ð12Þ
hcv 22 ¼
0:59ðGr22 Pr 22 Þ0:25 k22 L22
ð13Þ
hcv 23 ¼
0:27ðGr23 Pr 23 Þ0:25 k23 L23
ð14Þ
where Gr23 , Pr23 , and k23 are variables related to a reference temperature T rt2n that is expressed as
853
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Fig. 10. Temperature distribution along flow direction for simulation and experiments (up-flow).
Fig. 11. Temperature distribution on the coiled tube at flow rate of 5 m3/h in the (a) down-flow and (b) up-flow.
T rt2n ¼
T 2n þ T a 2
ð15Þ
s¼
6 X
ðF bð0klþ1 Þ F bð0kl Þ Þejl s
ð17Þ
l¼1
The inner surface of the cavity includes the outer wall of the turns and the optical splitter. The energy balance equation of the inner surface of the cavity is given by
Q 3;rd ¼ Q 31;rd þ Q 32;rd
ð16Þ
With the presence of a glass cover, re-radiation of the inner surface of the cavity is reflected, absorbed, and transmitted. The transmitted part is determined by the transmissivity s of the glass cover [32], which is calculated by
where F bð0kl Þ is the black body radiation function, jl is the absorption coefficients of different wave ranges that are identified in Table 1 [33], and l is the corresponding numbers of the different wave ranges.
Q 31;rd ¼
15 X
r T 431;k T 4g
k¼1
ec A k
sk 1ec
þ x 1Ag þ k
1eg eg A g
ð18Þ
854
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
Fig. 12. Absorbed heat and heat losses of the surfaces of the receiver at different air flow rates (down-flow).
r T 4sp T 4g
Q 32;rd ¼ ssp 1esp
esp Asp
þ xsp1Ag þ
1eg eg A g
ð19Þ
Table 3 Outlet temperature of the simulation. Simulation results in the down-flow (°C) Average flux (kW/m2)
where sk is the transmissivity of the glass cover for the No. k turn, ssp is the radiation transmissivity of the glass cover for the optical splitter, T 31;k and T sp are the temperatures of the No. k turn and the optical splitter, respectively, ec and esp are the emissivities of the coiled tube and the optical splitter, respectively, xk and xsp are the angle factors from each turn and the optical splitter to the glass, respectively, and Ak and Asp are areas of the No. k turn and the optical splitter, respectively.
4. Results and discussion 4.1. Model validation Experiments were carried out using up and down-flows, as shown in Fig. 8, to study the effects of flow rates and flow directions on the outlet temperature. Simulation work was conducted by coupling optical and thermal processes, and the results of the experiments and simulations are discussed in this section.
Fig. 13. Comparison of heat losses at different parts in the down and up-flows.
120 140 160 180 200 250 300 350
Flow rate (m3/h) 1
2
3
4
5
609 634 669 698 730 796 844 891
602 629 663 694 725 788 839 887
592 626 661 692 721 784 834 880
577 616 653 686 716 780 829 876
564 601 640 674 706 772 818 866
Fig. 9 shows the air temperature along the flow direction of down-flows (1–5 m3/h) under quasi-equilibrium states. The air temperature increased rapidly at the first five turns of the coiled tube and then decreased at the last one turn according to the absolute radiation energy distribution on the coiled tube. The radiation energy distribution at the 15 turns coils are displayed, as in Table 2. The most incident energy was absorbed by the upper half cavity, while less by the last turns. The wall temperature of the last turns was lower than the upstream air, leading to a temperature drop of air. The air flow rate had a significant effect on the thermal performance of the receiver. As the flow rate increased from 1 m3/h to 5 m3/h in the experiments, the air outlet temperature decreased from 662 °C to 570 °C and the thermal efficiency increased from 12.5% to 53.6%. The maximum temperature drop at the last turns was around 64 °C at 1 m3/h, whereas the minimum was around 21 °C at 5 m3/h. The temperature drop of air at the last turns was sensitive to the flow rate, indicating the outlet temperature and efficiency could be optimized by adjusting the flow rate. The temperature change in simulation was in agreement with experiment and the deviations of outlet temperature between the experimental and simulation results ranged from 8% to 2%. Fig. 10 shows the air temperatures along the up-flows. The temperature change trend was similar to the down-flows and the two curves of experimental and simulation results also matched well. The air flow rate plays an important role in the thermal performance of the receiver. As the flow rate increased from 1 m3/h to 5 m3/h in the experiments, the outlet temperature decreased from 593 °C to 546 °C and the thermal efficiency increased from 10.9% to 51.4%. The temperature drop at the last turns along flow direction ranged from 65 °C at 1 m3/h to 14 °C at 5 m3/h. As the flow rate
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
855
Fig. 14. Absolute flux distribution at each turn for different inner radii.
low increase in air temperature rate from the 6th turn. In the up-flows, a rapid increase in air temperature rate occurs at the first 12 turns, and the stationary stable phase is relatively shorter than that in the down-flows. The outlet temperature and the thermal efficiency in the down-flows are always higher than those in the up-flows in the experiments. These results are also demonstrated by the simulation model. The uncertainty of measurement of the air temperature, flow rate, and radiation flux are approximately at 1.0% [20], 2.0% [18] and 3.0% [7], respectively. Because the radiation power on aperture is composed of five lamp sources, the uncertainty of the radiation power on aperture is 6.7% [7]. The uncertainty of the reflectivity involved in the optical simulation is 3.0% [7]. Thus, the experimental uncertainty of the total incident radiation, U Q in , and thermal efficiency, U gt , are estimated to be 7.3% and 7.6%, respectively [7], using the following expressions,
U Q in ¼ Fig. 15. Temperature distribution along flow direction for different inner radii.
U gt ¼ increased from 1 m3/h to 5 m3/h, the outlet temperatures in the simulation decreased from 603 °C to 559 °C. The overall trend of the simulated air temperature coincided with the experimental results as well and the deviations of air outlet temperature between the experiments and the simulation values ranged from 0.3% to 2.3%. The air temperature heating processes are similar under down and up-flow conditions. A rapid increase caused by the relatively low specific heat capacity of the air occurs initially, and then the temperature drops at the end because of the relatively low radiation energy distribution. The different flow directions bring a different flux distribution in the air flow, but the same radiation energy distribution at the internal walls. In the down-flows, air flows through a high-flux area at first, leading to a higher increase in temperature rate at the first turns than in the up-flows, and then a low-flux area is developed at the end of the flow, resulting in a
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi U 2source þ U 2reflectiv ity
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi U 2Q in þ U 2thermocouple þ U 2flow
ð20Þ ð21Þ
where U source is the uncertainty of the radiation power on aperture, U reflectiv ity is the uncertainty of the measured reflectivity, U thermocouple and U flow are the uncertainty of the air temperature and flow rate, respectively. 4.2. Model analysis and parametric investigation 4.2.1. Temperature distribution on the outer wall of the coiled tube The temperature and radiation energy distributions on the cavity internal walls, which are useful to for identifying ‘‘hot spots’’ and evaluating thermal stresses and heat losses. They can be obtained using the comprehensive model. As shown in Fig. 11, the outer wall temperature initially increases along the air flow direction and then decreases at the last turns, in accordance with the change of air temperature in the tube. The wall temperature increased from around 257 °C to 577 °C along down-flow direction
856
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
while it decreased to around 517 °C at the last turns. In the up-flow configuration, the wall temperature increased from around 197 °C to 547 °C along the flow direction, and then dropped around 497 °C finally. Fig. 11 also shows that the hottest region of the coiled tube is close to the glass window in the up-flow configuration; thus, significant heat losses are observed because of conduction and re-radiation through the glass window. This is the main reason for the outlet temperature in the down-flows configuration to be always higher than that in up-flows under similar conditions. 4.2.2. Heat losses and discussion Fig. 12 shows the proportion of absorbed energy by the air and the heat losses of different surfaces of the receiver in the down-flow condition, which are normalized by the total concentrated incident irradiation power Qin = 1846 W. Qabsorbed represents the thermal efficiency, which has a strongly positive correlation with the flow rate. The reflection loss of the glass cover, Qref, is around 2.94%. Generally, all kinds of heat losses increased with the decrease of air flow rate. When the air flow rate reduced from 5 m3/h to 1 m3/h, radiation losses of the shell outer
surface to the environment, i.e. the total of Q21,rd, Q22,rd and Q23,rd, ranged from14.78% to 29.89%; convection losses of the shell outer surface to the environment, i.e. the total of Q21,cv, Q22,cv and Q23,cv, was from 18.11% to 28.57%; radiation losses of the glass cover to the environment, Q1,rd, was from 4.64% to 10.52%; convection losses of the glass cover to the environment, Q1,cv, was from 2.41% to 3.75%, and the radiation loss of the inner surface of receiver to the environment, i.e. the total of Q31,rd and Q32,rd, was from 4.27% to 9.83%. The heat losses of the shell outer surface account for the highest proportion, indicating that the insulation should be improved in future studies. A cooling system is suggested to reduce the heat losses closed to the glass window. The total heat losses increased sharply with the decrease in flow rate, e.g., 44.21% for t = 5 m3/h and 82.51% for t = 1 m3/h. Fig. 13 compares the heat losses at different parts in the down and up-flows. The heat losses of the glass cover and inner surfaces in the down-flows are less than those in the up-flows because the temperature at the region near the glass window is lower in the down-flows, whereas higher in the up-flows, thus more heat losses. Furthermore, heat losses of the outer surface in the
Fig. 16. Absolute flux distributions at each turn for different numbers of turns.
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
857
Fig. 17. Temperature distribution along the flow direction for different total turns: (a) 12–15 turns; (b) 8–11 turns.
down-flows are more than those in the up-flows because the temperature of the cavity’s internal walls in the down-flows are higher than that in the up-flows, resulting in more conductive heat loss through insulation. Generally, the total heat losses in the down-flows are lower by 11 W (minimum) for t = 5 m3/h and by 48 W (maximum) for t = 1 m3/h than those in the up-flows. 4.2.3. Influence of incident radiation flux Table 3 lists the outlet temperatures forecasted by the comprehensive simulation model when the average flux incident on the glass cover ranges from 120 kW/m2 to 350 kW/m2. The outlet temperature can be more than 700 or 800 °C if the average flux incident reaches 200 or 300 kW/m2, respectively, for an air flow rate of less than 5 m3/h. The high air temperature at the expense of high heat losses, i.e., low thermal efficiency. At 200 and 300 kW/m2, the thermal efficiency is 39.93% and 31.01% for outlet temperatures of 706 and 818 °C, respectively, at 5 m3/h flow rate. 4.2.4. Influence of inner radius A smaller inner radius of a tube produces a higher velocity at the same flow rate, increasing the Nu number and strengthening the heat transfer. The effect of the inner radius on the receiver’s thermal performance is studied based on the comprehensive simulation model. Fig. 14 displays the flux distributions on the coiled tube as the inner radius decreases from 5.5 mm to 4.0 mm, where the height of the coiled tube decreased and the optical splitter was moved up, causing more light was distributed by the optical splitter. It is noticed that more incident radiation is reflected onto the last turns, leading higher surface temperatures of the last turns and the air outlet temperature. For example, the radiation energy at the last one turn can be increased from 85 W for 6 mm-tube conditions to 106 W for 4 mm-tube conditions. At an air flow rate of 5 m3/h, as shown in Fig. 15, the temperature increases as the inner radius decreases, regardless of the temperature drop at the bottom tube. The outlet temperature and the thermal efficiency increase with the decrease in the inner radius, and the outlet temperature reaches up to 670 °C with a thermal efficiency of 64.1% when the inner radius is 4 mm. The pressure drop increases sharply from 9.63 kPa for a 5.5 mm inner radius to 56.52 kPa for a 4.0 mm inner radius, while 5.37 kPa for a 6.0 mm inner radius. 4.2.5. Influence of the number of turns The temperature always increases slightly at the 9th to the 13th turns and then significantly drops at the last three turns in the down-flows. To optimize the length of the tube, the effect of the total turns on the receiver’s thermal performance is studied based
on the comprehensive simulation model at an air flow rate of 5 m3/h under the down-flow conditions. The height of the coiled tube decreases and the optical splitter moves upward with the decrease in turns. Fig. 16 shows the absolute radiation energy distributions for the number of turns ranging from 14 to 8. The height of the coiled tube decreased and the optical splitter was moved upward with the decreasing of turns. Thus the flux reflected by the splitter increased and more incident irradiation was absorbed by the bottom turns of air cavity receiver, which may lead to a higher air outlet temperature. The maximum radiation energy at each turn can be increased from 149 W for 14 turns to 243 W for 8 turns, while it is 142 W for 15 turns. However, due to the decreasing of turns, the spillage of incident light increases, which leads to the total radiation energy distribution on the internal walls decreases from 1.82 kW for 14 turns to 1.68 kW for 8 turns, while 1.85 kW for 15 turns. It is also noted the heat transfer area is reduced with the decreasing of turns. Fig. 17(a) shows the air temperature distribution along the air flow direction for 12–15-turn tubes. The temperature changing trends are similar at the first six turns. The air temperature fluctuates at the 7th to the 10th turns because of the flux distribution, and the final outlet temperature is from 559 °C to 572 °C. The pressure drop decreases from 5.37 kPa for 15 turns to 4.62 kPa for 12 turns. Fig. 17(b) shows the air temperature distribution along the air flow direction for 8–11-turn tubes. The outlet temperature reaches up to 568 °C for the 10-turn tube. The outlet temperature of the 9-turn and 8-turn tubes decreases to 554 °C and 542 °C, respectively, because of the decrease in the total flux distribution and the heating time. The pressure drop decreases from 4.16 kPa for 11 turns to 3.61 kPa for 8 turns. Therefore the air outlet temperature and pressure drop should be taken into consideration for optimizing the number of turns of air tube-cavity solar receiver.
5. Conclusions An air tube-cavity solar receiver has been designed and fabricated to generate high temperature air from solar energy. The maximum air outlet temperature can reach up to 662 °C and 593 °C in the down-flow and up-flows configurations, respectively. A comprehensive model coupling the optical and the heat transfer processes has been developed. The model has been validated using the measured data. The difference for the outlet temperature between the measurement and the model prediction is less than 8%. The model predicts that the outlet air temperature can reach up to about 800 °C for an air flow rate of 5 m3/h and average radiation flux of 300 kW/m2. The thermal efficiency of the receiver can reach to 64.1% for an air outlet temperature of 670 °C if the inner
858
K. Qiu et al. / Energy Conversion and Management 103 (2015) 847–858
tube radius is reduced to 4 mm at the cost of the increasing pressure drop. The outlet temperature at 5 m3/h air flow can reach 572 °C if a 12-turn air tube-cavity receiver is utilized, which has a suitable flux distribution under the same operating conditions. The comprehensive model can also provide detailed analysis about thermal performances and point out effective ways to improve performances, including reducing heat loss, optimizing structural parameters. Acknowledgements The authors acknowledge the support from the National Natural Science Foundation of China (No. 51476140), the Program of Introducing Talents of Discipline to University (No. B08026), and Seed Funds of Interdisciplinary Research for Young Teachers of Zhejiang University (JCZZ-2013014). References [1] Behar O, Khellaf A, Mohammedi K. A review of studies on central receiver solar thermal power plants. Renew Sustain Energy Rev 2013;23:12–39. [2] Yang M, Yang X, Yang X, et al. Heat transfer enhancement and performance of the molten salt receiver of a solar power tower. Appl Energy 2010;87(9):2808–11. [3] Muñoz-Anton J, Biencinto M, Zarza E, et al. Theoretical basis and experimental facility for parabolic trough collectors at high temperature using gas as heat transfer fluid. Appl Energy 2014;135:373–81. [4] http://www.nrel.gov/csp/solarpaces/project_detail.cfm/projectID=62. [5] Biencinto M, González L, Zarza E, et al. Performance model and annual yield comparison of parabolic-trough solar thermal power plants with either nitrogen or synthetic oil as heat transfer fluid. Energy Convers Manage 2014;87:238–49. [6] Ávila-Marín AL. Volumetric receivers in solar thermal power plants with central receiver system technology: a review. Solar Energy 2011;85(5):891–910. [7] Xiao G, Guo K, Luo Z, et al. Simulation and experimental study on a spiral solid particle solar receiver. Appl Energy 2014;113:178–88. [8] Bai F, Zhang Y, Zhang X, et al. Thermal performance of a quartz tube solid particle air receiver. Energy Proc 2014;49:284–94. [9] Zavattoni SA, Gaetano A, Montorfano D, et al. A novel CSP receiver based on AirLight energy technology – optimization of the thermal insulation system by means of CFD analysis. Energy Proc 2014;49:589–98. [10] Amsbeck L, Buck R, Heller P, et al. Development of a tube receiver for a solarhybrid microturbine system. In: Proceedings of 14th SolarPACES conference, Las Vegas, NV, March; 2008.
[11] Wang W, Gianmarco R, Lucas A, et al. Integrated design of a hybrid gas turbine-receiver unit for a solar dish system. In: Proceedings of 20th SolarPACES conference, Beijing, China, September; 2014. [12] Korzynietz R, Torreira MG. Soltrec-pressurized volumetric solar air receiver technology. Energy Proc 2014;49:284–94. [13] Hischier I, Hess D, Lipin´ski W, et al. Heat transfer analysis of a novel pressurized air receiver for concentrated solar power via combined cycles. J Therm Sci Eng Appl 2009;1(4):041002. [14] Kribus A, Doron P, Rubin R, et al. A multistage solar receiver: the route to high temperature. Solar Energy 1999;67(1):3–11. [15] Wang F, Shuai Y, Tan H. Thermal analyse of cavity solar receiver. J Eng Thermophys 2011;32(5):843–6. [16] Paitoonsurikarn S, Lovegrove K. On the study of convection loss from open cavity receivers in solar paraboloidal dish applications. In: Proceedings of 41st conference of the Australia and New Zealand Solar Energy Society (ANZSES), Melbourne, Australia; 2003. [17] Prakash M, Kedare SB, Nayak JK. Investigations on heat losses from a solar cavity receiver. Solar Energy 2009;83(2):157–70. [18] http://www.shy-sz.com/index.asp. [19] http://www.agilent.com/home. [20] http://cn.omega.com/section/thermocouples.html?utm_source=baidu&utm_ medium=cpc&utm_term=k%E5%9E%8B%E7%83%AD%E7%94%B5%E5%81%B6& utm_campaign=BSTYWD. [21] http://www.ophiropt.com/laser-measurement. [22] de Jager C. The solar spectrum. The solar spectrum; 1965. [23] http://www.breault.com/software/about-asap. [24] Xiao G, Guo K, Ni M, et al. Optical and thermal performance of a hightemperature spiral solar particle receiver. Solar Energy 2014;109(109):200–13. [25] Price DJ. The emissivity of hot metals in the infra-red. Proc Phys Soc 1947;59(1):118. [26] Dvurechensky A, Petrov V, Reznik VY. Spectral emissivity and absorption coefficient of silica glass at extremely high temperatures in the semitransparent region. Infrared Phys 1979;19(3):465–9. [27] Shen W, Jiang Z, Tong G, et al. Engineering thermodynamics. Higher Education Press; 2001. [28] ANSYS.FLUENT – theory guide; 2009. [29] Raithby GD, Chui EH. Finite-volume method for predicting a radiant heat transfer in enclosures with participating media. J Heat Transfer 1990;112(2):415–23. [30] Wang F. Analysis and calculation of fluid dynamics: principle and application of CFD software. Tsinghua University Press Ltd.; 2004. [31] Ferziger JH, Peric´ M. Computational methods for fluid dynamics. Berlin: Springer; 2002. [32] Yang S, Tao WT. Heat transfer. 4th ed. Beijing: High Education Press; 2006. [33] Tan H. Numerical simulation of infrared radiation characteristics and transmission. Harbin Institute of Technology press; 2006.