Renewable Energy 36 (2011) 659e670

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

A procedure to select working uids for Solar Organic Rankine Cycles (ORCs)
R. Rayegan a, *, Y.X. Tao a
a

Mechanical and Materials Engineering Department, Florida International University, 10555 West Flagler St., Miami, FL 33174, USA

a r t i c l e i n f o
Article history: Received 22 February 2010 Accepted 14 July 2010 Available online 5 August 2010 Keywords: Solar energy Organic Rankine cycle Working uid Exergy Regeneration

a b s t r a c t
The selection of working uid and working conditions of the Organic Rankine Cycle (ORC) has a great effect on the system operation, and its energy efciency and impact on the environment. The main purpose of this study is to develop a procedure to compare capabilities of working uids when they are employed in solar Rankine cycles with similar working conditions. The Refprop 8.0 database with 117 organic uids has been considered as the reference in this study. A procedure to compare ORC working uids based on their molecular components, temperatureeentropy diagram and uid effects on the thermal efciency, net power generated, vapor expansion ratio, and exergy efciency of the Rankine cycle has been proposed. Fluids with the best cycle performance have been recognized in two different temperature levels within two different categories of uids: refrigerants and non-refrigerants. Based on categories of solar collectors, 11 uids have been suggested to be employed in solar ORCs that use low or medium temperature solar collectors. Collector efciency improvement and use of the regenerative ORC instead of the basic cycle reduce irreversibility of a solar ORC. Calculation results show that for selected uids, the theoretical limits for irreversibility reduction and exergy efciency enhancement through collector efciency improvement are 35% and 5% respectively, when the collector efciency increases from 70% to 100%. The effect of regeneration on the exergy efciency of the cycle is uid dependent while the effect of collector efciency improvement on the exergy efciency of the cycle is nearly independent of uid type. At the two temperature levels studied, higher molecular complexity results in more effective regenerative cycles except for Cyclohydrocarbons. Published by Elsevier Ltd.

1. Introduction High-temperature thermal power plants that work based on the conventional Rankine cycle are not economic in small-scale applications. The Organic Rankine Cycle (ORC) is a substitutive technology which is applicable for small-scale power generation for use in residential and commercial buildings or in desalination plants. ORC employs low-grade heat from different sources such as biomass, geothermal, solar and waste heat of industrial processes. The main difference between the ORC and the conventional Rankine cycle is the working uid. The boiling point of working uid in the ORC is much lower than steam, hence there is no need to achieve high temperatures to generate vapor for running a micro-turbine or expander. As a result, ORCs can be driven at lower temperatures than the Rankine cycles that use water. Solar radiation has the highest capacity and the lowest replenishment time among sustainable energies [1]. Without considering

* Corresponding author. Tel.: 1 305 3483558; fax: 1 305 3481932. E-mail address: rambod.rayegan@u.edu (R. Rayegan). 0960-1481/$ e see front matter Published by Elsevier Ltd. doi:10.1016/j.renene.2010.07.010

environmental costs, solar thermal power cycles are not cost competitive with thermal power plants with conventional heat sources. The costs can be reduced by improving system performance. The selection of working uid and operation conditions has a great effect on the system operation, and its energy efciency and impact on the environment. Previous works on working uids used in solar ORCs are very limited [2e4]. In previous investigations, working conditions of optimized cycles are different for different working uids making the comparison and analysis questionable. The main purpose of this study is to nd a meaningful procedure to compare capabilities of working uids when they are employed in Rankine cycles with similar working conditions. The procedure is presented based on working uids molecular components, temperatureeentropy diagram and uid effects on thermal efciency, net power generated, vapor expansion ratio, and exergy efciency of the Rankine cycle. Refprop 8.0 [5] database has been considered as the reference in this study. This program, developed by the National Institute of Standards and Technology, provides tables and plots of the thermodynamic and transport properties of industrially important uids and their mixtures with an emphasis on refrigerants and hydrocarbons. Refprop 8.0 consists

660

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

Nomenclature Cg Ex Fe Gb h I K _ m P Ph Q Q* Q0 q R s Sgen DT T T* abs ULs W w geometric concentration ratio of solar collector [-] exergy [W] dirt degree of the collector mirrors [-] direct solar irradiance [W/m2] specic enthalpy [J/kg] irreversibility [W] incidence angle modier [-] mass ow rate [kg/s] pressure [Pa] higher pressure limit of Rankine cycle [Pa] heat transfer rate [W] irradiation rate [W] ambient heat loss rate of solar collector [W] heat transfer per unit mass [J/kg] gas constant [J/kg K] specic entropy [J/kg K] entropy generation rate [W/K] temperatute difference [K] temperature [K] apparent sun temperature [K] thermal loss coefcient per unit area of the absorber tube [W/m2 K] power [W] power per unit mass [J/kg]

hopt;0 s 4

collector optical efciency at a zero incidence angle [%] molecular complexity [-] angle of incidence of the direct solar radiation [rad]

Subscripts a actual abs absorber c collector con condenser cr critical eva evaporator ex exergy f saturated liquid g saturated vapor in inlet ow j cycle component index L low net net output out outlet ow p pump r reduced property s isentropic SV saturation vapor t turbine th thermal wf working uid 0 ambient Subscripts  average

3reg h

Greek symbols regeneration efciency [-] efciency [%]

of 85 pure uids and 55 predened mixtures. Among them 63 pure uids and 54 predened mixtures are organic. R508A and R508B have very low critical temperatures. Therefore they are not proper to be employed in a Rankine cycle. A total of 115 pure uids and predened mixtures from the Refprop 8.0 database are investigated. Irreversibility in solar thermal systems is relatively high because of the high temperature difference between the solar collector and the apparent sun temperature. Important factors that could lead to irreversibility reduction of the solar ORC by collector efciency improvement and using regenerative cycles are investigated at the last section of this paper. 2. Preliminary selection In the preliminary selection Chlorine included uids and wet uids should be discarded. Chlorine containing uids are not Ozone-safe and have been banned by Montreal protocol [6] and thus should be avoided in new systems. Among pure uids of the Refprop 8.0 database, 12 uids are chlorine included. Six of them are CFCs and six uids are HCFCs. Among predened mixtures, there are 28 chlorine-included uids. 23 mixtures are CFC included; three mixtures are HCFC included and two of them include both CFC and HCFC. Depending on the slope of the temperatureeentropy curve to be innity, positive, or negative, working uids can be classied into isentropic, dry, or wet respectively [7]. Dry or isentropic working uids are more appropriate for ORC systems. This is because dry or isentropic uids are superheated after isentropic expansion. Therefore there is no concern for existing liquid droplets at the turbine outlet. The slope of the temperatureeentropy curve for some wet uids is very close to innity. Furthermore the isentropic

efciency of turbine is around 80% in the practical cycle. Subsequently the turbine outlet will be in the dry region that means employing such a uid causes no problem for the turbine. There are four wet uids among the nal preselected uids. Table 1 shows the critical properties of preselected organic uids with their critical properties. 3. Thermodynamic cycle The heat absorption process in an ORC may end in a saturated vapor state or superheated vapor state. Generally, superheating in an ORC increases the thermal efciency of the cycle with a very low slope but decreases the exergy efciency of the cycle [9]. Then superheated cycles are never recommended unless in order to gain more power at the expense of losing efciency. In addition, increasing the maximum temperature of the collector in solar cycles increases the heat loss of it [2]. Because of these reasons the saturated Rankine cycle has been investigated in this study instead of a superheated cycle. Assumptions of the analysis are as follows: steady state condition; no pressure drop in heat exchangers and connecting pipes, and isentropic efciencies of the pump and the turbine are equal to 0.8. During the next step we should determine the practical pressure and temperature limits of the cycle. As the higher pressure ratio leads to a higher efciency, we prefer to expand higher and lower pressure limits of the cycle, but there are always some practical restrictions. Near critical pressure, small changes in temperature are equivalent to large changes in pressure that make the system unstable. Therefore a reasonable distance between the higher limit of the cycle and the critical point of the uid should be considered. But

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670 Table 1 Preselected working uids. Working uid Acetonea Benzene Butane Butene Peruorobutane Perfuoropentane Cis-butene Cyclohexane Decane Dodecane Diuomethane Heptane Hexane Isobutane Isobutene Isohexane Isopentane Neopentane Nonane Octane Pentane Octauoropropane 1,1,1,2,3,3,3-Heptauoropropane 1,1,1,2,3,3-Hexauoropropane 1,1,1,3,3,3-Hexauoropropane 1,1,2,2,3-Pentauoropropane 1,1,1,3,3-Pentauoropropane 1,1,1,3,3-Pentauorobutane Octauorocyclobutane Toluene Trans-butene R-413Aa R-423Aa R-426Aa
a

661

T
Alternative name(s) e e e e C4F10 C5F12 e e e e R32, E134 e e e e e e e e e e R218 R-227ea R-236ea R-236fa R-245ca R-245fa R-365mfc R-C318 e e e e e Pcr (MPa) 4.700 4.894 3.796 4.005 2.323 2.045 4.225 4.075 2.103 1.817 4.228 2.736 3.034 3.64 4.010 3.040 3.396 3.196 2.281 2.497 3.370 2.640 2.999 3.502 3.200 3.925 3.640 3.240 2.777 4.126 4.027 4.022 3.563 4.088 Tcr ( C) 234.9 288.9 152.0 146.1 113.2 147.4 162.6 280.5 344.5 384.9 147.1 267.0 234.7 134.7 144.9 224.5 187.2 160.6 321.4 296.2 196.5 71.9 102.8 139.3 124.9 174.42 154.0 186.75 115.2 318.6 155.5 96.6 99.1 99.8


Ph2 Ph1 C D

B A

S
Fig. 1. Higher pressure limit of the ORC.

across the turbine is restricted to one percent. In this method for both R32 and Cis-butene, higher pressure limit of the cycles are 25  C lower than their own critical temperature. Condensing temperature has been set to 25  C in this study. If necessary, the condenser temperature is raised to make the condenser pressure equal to 5 kPa, the lowest pressure accepted for the condenser [8]. Table 2 shows the practical higher and lower limit of the cycle for each working uid. Decane, Dodecane, Octane, and Nonane are removed from the list because of their high condensing temperatures.

Wet uids. Table 2 Practical limits of the ORC for preselected working uids.

there is no unique interpretation of the reasonable distance from critical point in the literature. Drescher and Bruggemann [8] suggested setting the higher pressure limit of the cycle 0.1 MPa lower than critical pressure. Delgadeo-Torres and Garcia-Rodriguez [3] considered the higher temperature of the cycle to be 10e15  C lower than critical temperature. Because of the difference between the critical properties of working uids, a xed pressure or temperature interval near the critical point of the uid may not be a consistent way to determine the distance between the higher limit of the cycle and the critical point of the uid. For example a 15  C temperature difference next to the critical point of Dodecane is equivalent to a 0.332 MPa pressure difference, while for R-227ea it is equivalent to 0.800 MPa. The slope of the temperatureeentropy diagram has been used to determine the higher limit of the Rankine cycle in this study. To avoid the presence of liquid in every single section of the turbine, the highest input pressure of the turbine is the pressure that the slope of temperatureeentropy diagram is equal to innity at that point (point A in Fig. 1). Calculating the higher pressure and temperature limit of the cycle based on this criterion shows that for most of uids a large capacity of producing power is neglected. For example for R32 and Cis-butene, point A is 45.1  C and 44.6  C lower than their own critical temperature respectively. To modify this criterion, increasing the higher limit of the cycle in expense of liquid droplet presence across a small portion of the turbine process is proposed. As has been shown in Fig. 1, by increasing the higher pressure limit of the cycle from Ph1 to Ph2 liquid droplets present in the turbine across BD. The maximum mass fraction of liquid in this process belongs to point C. In the modied method the highest allowed mass fraction of the liquid

Working uid Acetone Benzene Butane Butene C4F10 C5F12 Cis-butene Cyclohexane Decane Dodecane E134 Heptane Hexane Isobutane Isobutene Isohexane Isopentane Neopentane Nonane Octane Pentane R218 R-227ea R-236ea R-236fa R-245ca R-245fa R-365mfc R-C318 Toluene Trans-butene R-413A R-423A R-426A

Maximum Peva (MPa) 3.379 4.067 3.013 2.808 2.057 1.803 3.035 3.665 1.896 1.723 2.747 2.410 2.680 2.890 2.877 2.682 2.887 2.788 2.059 2.200 2.865 1.899 2.352 2.955 2.288 2.951 2.817 2.712 2.314 3.576 2.906 1.839 2.966 1.562

Maximum Teva ( C) 213 274 138 125 107 141 142 272 337 381 125 258 226 121 125 216 177 152 314 287 186 57 91 132 108 158 140 177 106 307 136 59 90 55

Minimum Pcon (kPa) 30.7 12.7 234.7 297.2 268.3 84.7 213.7 13.0 5.1 5.1 212.8 6.1 20.2 350.5 305.0 28.2 91.8 171.4 5.0 5.0 68.3 867.5 455.2 205.9 272.4 100.8 149.4 53.4 312.5 5.1 234.1 720.2 598.0 687.8

Minimum Tcon ( C) 25 25 25 25 25 25 25 25 85 121 25 25 25 25 25 25 25 25 65 44 25 25 25 25 25 25 25 25 25 31 25 25 25 25

662

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

T 2 1a 1s 4 3s 3a

transfer and power in each component of the cycle are calculated by applying the rst law of thermodynamics on them. Evaporator:

qeva h2 h1a
Turbine/Expander:

(3)

wt h2 h3a
Condenser:

(4)

qcon h3a h4
Pump:

(5)

wp h1a h4

(6)

S
Fig. 2. Actual saturated basic ORC.

where qeva , wt qcon , and wp are absolute values of heat transfer in evaporator, turbine power, heat transfer in condenser and pump power respectively. The thermal efciency of the cycle is:

4. Analysis The equations used to calculate the different parameters to evaluate the performance of the cycle are presented in this section. The rst law of thermodynamics is applied to the individual components of the cycle and the second law of thermodynamics is applied to the whole cycle to determine heat transfer, work input and output, and irreversibility of the cycle. The rst law of thermodynamics for steady-state steady-ow processes when potential and kinetic energy changes are negligible can be expressed as [10]:

hth

h2 h3a h1a h4 h2 h1a

(7)

The procedure to calculate required enthalpies has been depicted in Fig. 3. Heat transfer components to/from a solar ORC have been shown in Fig. 4. The solar collector receives solar radiation at the rate Q * . Q0 and Qcon represent the solar collector ambient heat loss and heat rejection through the condenser respectively. Eq. (2) can be rewritten as:

Sgen

Q0 Qcon Q * * T0 TL T

(8)

_ hout hin Q W m

(1)

 _ , hout and hin are the heat transfer rate, the power where Q, W, m exchange, the mass ow rate and outgoing and incoming ow enthalpies respectively. The irreversibility rate for a cycle in steady state steady ow condition can be expressed as [10]:

_ I T0 Sgen T0 m

X

qj =Tj

Exin Exout

(2)

where T * is the apparent sun temperature as an exergy source, T0 is the ambient temperature, and TL is the temperature of the heat carrying uid in condenser. In this study the value suggested by Petela [11] for T * is adopted that is approximately equal to 3/4 Ts , where Ts is the apparent black body temperature of the sun, which is about 6000 K. Therefore, the T * considered here is 4500 K. Assuming that the temperature difference between the heat carrying uid and the condenser is T yields:

TL Tcon DT
C

(9)

where I, T0 , Sgen , Exin and Exout are the irreversibility, the entropy generation rate and incoming and outgoing exergy ows respectively. qj is the heat transfer per unit mass and Tj is the temperature of the jth component of the cycle. 4.1. Basic cycle Fig. 2 shows a general representation of the actual saturated basic Rankine cycle in the T-s diagram considering assumptions that were mentioned in the previous section. States 1a and 3a are the actual exit states of the pump and the turbine, respectively, and 1s and 3s are the corresponding states for the isentropic case. Heat

T has been considered 15 in this analysis. The net heat absorption by the evaporator (Qeva ) will be the difference between the solar radiation received by the collector (Q * ) and the solar collector ambient heat loss (Q0 ).

Qeva Q * Q0

(10)

The solar collectors efciency hc is dened as that fraction of the solar radiation which reaches the receiver and is absorbed there [12]:

hc

Qeva Q*

(11)

By using eqs. (9)e(11), Eq. (8) can be written as:

Fig. 3. Enthalpy calculation procedure in a saturated basic ORC.

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

663

Evaporator

2 5 3a 6

Pump

Turbine

5 1a

Condenser

S
Fig. 4. Heat transfer components to/from a solar ORC. Fig. 5. Actual saturated regenerative ORC.

Sgen

Qeva 1 hc 1 Qcon * hc T0 TL T

(12)

The only parameter that we should determine to fulll calculations is the solar collector efciency. Every single type of a solar collector has its own formula to calculate its efciency that is a function of the geometry of the collector and the thermophysical properties of the materials that have been used to build it. In this analysis solar collectors suggested by Delgado-Torres and GarciaRodriguez [3] are adopted: LS-3 and IND300. The LS-3 model is one of the biggest collectors within the family of commercial Parabolic Trough Collectors (PTCs) that has been used in some of the largest solar energy plants built in Mojave Desert in California. It was also the PTC chosen to demonstrate the technical feasibility of the direct steam generation process within the scope of Direct Solar Steam (DISS) project [13]. The efciency of the LS-3 PTC is given by the following expression [3]:

the working uid in the collector and the average temperature of the absorbers tube are the same. The IND300 model is a smaller PTC in comparison to the LS-3 made by the Israeli company Solel Solar Systems. The IND300 PTCs efciency is given by Eq. (15).

hIND300 0:733 0:238

T wf T0 Gb

 0:0013Gb

T wf T0 Gb

2 (15)

where T wf is the average temperature of the working uid in the collector. T0 and Gb for IND300 PTC have the same value for the LS-3 model. The exergy efciency of the cycle is dened by [14]:

hex

Exergy recovered Wnet Wnet Wnet Exergy supplied Wrev Wnet I Wnet T0 Sgen

(16)

hLS3 T abs ; Gb ; 4 hopt;0 $K 4$Fe

abs UL T T0 $ abs Gb Cg

where Wnet is the net power output of the cycle.

(13)
4.2. Regenerative cycle To reduce the high irreversibility of the solar ORC, regenerative ORC with regeneration efciency 3reg 0.8 has been investigated in this study. Fig. 5 shows a general representation of the actual saturated regenerative Rankine cycle in the Tes diagram. Regeneration efciency is expressed by:

where hopt;0 is the collector efciency at a zero incidence angle, 4 is the angle of incidence of the direct solar radiation, K(4) is the incidence angle modier, Fe is the dirt degree of the collector abs is the thermal loss coefcient per unit area of the mirrors, UL absorber tube, Cg is the geometric concentration ratio, T abs is the absorbers tube average temperature, T0 is the ambient temperature and Gb is the direct solar irradiance. The thermal loss coefcient per unit area of the tube is given by Eq. (14).

3reg


abs ULs a bT abs T0 cT abs T0 2

W m2 $K

 (14)

h5 h1a h0 5 h1a

(17)

The values for the coefcients a, b, and c are shown in Table 3 [3]. The rest of the parameters in Eq. (13) are set as follows [3]: hopt;0 0.77, K(4) 1, Fe 0.967, Cg 26.2, T0 300 K, and Gb 850 W/m2. In this analysis it has been assumed that the average temperature of
Table 3 Values of the coefcients for the thermal loss coefcient of the LS-3 PTC absorber tube [3]. a T abs < 200  C 200  C < T abs < 300  C T abs > 300  C 0.68726 1.43324 2.89547 b 0.002 0.01 0.016 c 2.6 105 4.6 105 6.5 105

Fig. 6. Supplementary enthalpy calculation procedure in a saturated regenerative ORC.

664

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

a
Vapor Expantion Ratio Wnet(KJ/Kg) 30 25 VER 20 15 10 5 0 40 60 80 100 120 140 Teva (C) 20 10 0 50 Wnet (KJ/Kg) 40 30

Thermal Efficiency Exergy Efficiency-IND300 Exergy Efficiency-LS-3

negligible. For exergy efciency calculations in the regenerative ORC, the same equations of the basic ORC are applicable. 5. Results and discussion In this section four main subjects will be discussed. The rst subject is to choose dominant factors inuencing the performance of an ORC. The second is the calculation of maximum practical thermal efciency of an ORC through employing each working uid. The third subject is to complete the comparing procedure of working uids, started at Section 2, based on their effects on the performance of the ORC. The fourth subject focuses on investigation of different methods to reduce irreversibility of the ORC. 5.1. Dominant factors inuencing the performance of an ORC The rst factor which is always the center of attention among different factors in a Rankine cycle is the thermal efciency or the rst law efciency of the cycle. For a specic working uid and particular amount of input heat rate the higher thermal efciency leads to the higher net power output. As we want to compare different working uids in the Rankine cycle, the net power output should be considered along with the thermal efciency. If for a small amount of work, a high vapor expansion ratio occurs across the turbine (VER), supersonic ow problems, higher turbine size or greater number of stages are inevitable. Thus the high vapor expansion ratio across the turbine is an undesirable factor in an ORC. Exergy efciency or second law efciency is the factor which helps us to choose working uids that recover a greater portion of input exergy of the cycle. Consequently; thermal efciency, net power output, vapor expansion ratio across the turbine, and exergy efciency of the cycle are the most important factors to be considered as the performance improvement of an ORC.

16 12 8 4 0 40 60 80 100 120 140 Teva (C)

Fig. 7. Variation of performance factors with respect to Teva of an ORC employing R236ea as working uid, (a) VER and Wnet (b) hth and hex.

Regeneration has no effect on the turbine and pump power but heat transfer through the evaporator and condenser change in the regenerative cycle in comparison to the basic cycle. Applying the rst law of thermodynamics on the evaporator and condenser processes yields: Evaporator:

qeva h2 h5
Condenser:

(18)

qcon h6 h4

(19)

The procedure to calculate enthalpies of common states with basic cycle (4, 1a, 2, and 3a) are exactly the same. The enthalpy calculation procedure of states 5 and 6 has been shown in Fig. 6. It should be noticed that in this procedure it is assumed that the regenerator is well insulated and changes in kinetic and potential energies are
Table 4 Maximum delivery of an ORC employing different working uids. Working uid R-426A R218 R-413A R-423A R-227ea R-C318 C4F10 R-236fa Isobutane Butene Isobutene E134 R-236ea Trans-butene Butane R-245fa C5F12 Cis-butene Neopentane R-245ca Isopentane R-365mfc Pentane Acetone Isohexane Hexane Heptane Cyclohexane Benzene Toluene Maximum Teva ( C) 55 57 59 90 91 106 107 108 121 125 125 125 132 136 138 140 141 142 152 158 177 177 186 213 216 226 258 272 274 307 Minimum Tcon ( C) 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 31 Maximum hth (%) 6.37 5.73 7.05 10.28 10.11 11.22 10.53 12.31 13.78 14.78 14.65 14.79 14.29 15.84 15.48 15.57 12.37 16.50 15.52 16.96 17.75 17.55 18.51 22.54 19.27 20.08 20.81 23.49 25.79 25.60

Maximum Wnet (KJ/Kg) 11.92 5.03 12.60 16.50 14.06 16.20 13.87 22.93 58.08 66.53 66.27 37.08 30.40 77.05 74.78 40.04 21.31 82.09 71.78 48.20 97.10 56.28 108.12 155.44 123.32 135.54 158.24 170.22 179.00 190.82

Maximum VER 2.35 2.86 2.66 6.93 7.65 11.74 13.47 12.02 11.73 12.08 12.35 16.29 21.60 16.17 17.61 26.50 38.19 25.91 26.90 43.12 50.62 84.27 67.40 128.27 169.34 232.50 720.95 508.92 428.71 1106.28

Maximum

Maximum

hex for IND300 (%)

4.74 4.28 5.25 7.59 7.49 8.31 7.85 9.07 10.10 8.62 10.70 10.80 10.48 11.53 11.30 11.37 9.28 11.98 11.40 12.34 12.94 12.82 13.45 15.87 14.07 14.59 15.14 16.71 17.88 17.87

hex for LS-3 (%)

4.86 4.38 5.38 7.85 7.74 8.62 8.13 9.43 10.55 8.94 11.20 11.31 10.97 12.11 11.86 11.94 9.71 12.62 11.99 13.03 13.75 13.61 14.34 17.24 15.17 15.81 16.64 18.48 19.92 20.09

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

665

Teva= 130 C
20 18 16 14 12 10 8 6 Thermal Efficiency ( ) Exergy Efficiency -LS-3 ( ) Exergy Efficiency -IND300 ( )

Neopentane

R245fa

Butane

R365mfc

Isopentane

Trans-butene

Isohexane

Pentane

Cis-butene

Heptane

Cyclohexane

Acetone
Trans-butene Isohexane

C5F12

R236ea

R245ca

Hexane

Toluene

Fig. 8. Thermal and exergy efciency of the ORC for different working uids at Teva 130  C.

5.2. Maximum thermal efciency of the ORC for different working uids Maximum practical thermal efciency and corresponding performance factors for preselected working uids are calculated in this section. Because of the increasing trend of hth, Wnet , VER, and hex with Teva in all uids, their maximum happen at maximum Teva . Fig. 7 shows this increasing trend for R-236ea as an example. Calculation results for best possible performance of each working uid in an ORC can be seen in Table 4. Results have been sorted from smallest to largest maximum Teva . Results shown in Table 4 conrm conclusions of authors previous study [9]. The higher critical temperature allows setting the evaporation temperature at a higher level that leads to the higher thermal efciency of the cycle. Generally, high thermal efciency

ORCs are achievable by using hydrocarbons rather than refrigerants. This means hydrocarbons have a higher potential to produce power in a Rankine cycle than refrigerants because of their relatively highcritical temperature. On the other hand hydrocarbons are more ammable in comparison to refrigerants. These results add to previous conclusions that the exergy efciency has almost the same trend of thermal efciency with respect to the critical temperature of the uid. In general, the same as thermal efciency, the exergy efciency of refrigerants are lower than hydrocarbons. 5.3. Comparing procedure of preselected working uids A specic solar collector in a region with a denite direct solar irradiance can maintain temperatures within restricted limits. Therefore the highest allowed temperature for a working uid in

120 100
wnet(KJ/Kg)

Teva= 130 C

40 35 30

Teva= 130 C

80 60 40 20 0
C5F12 R236ea Neopentane R245fa Butane R365mfc Isopentane R245ca Trans-butene Isohexane Pentane Hexane Cis-butene Heptane Toluene Cyclohexane Acetone Benzene

25 VER 20 15 10 5 0 Cis-butene Toluene Cyclohexane Isopentane Neopentane Hexane R365mfc Benzene Pentane Heptane Butane Acetone C5F12 R236ea R245fa R245ca

Fig. 9. Net output power of the ORC for different working uids at Teva 130  C.

Fig. 10. Vapor expansion ratio in the ORC for different working uids at Teva 130  C.

Benzene

666

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

Teva= 85 C 13 12 11 10 9 8 7 6 5 C4F10 C5F12 R227ea RC318 R423A R236fa R416A R236ea Neopentane Isobutane Toluene R245fa Butane Isobutene R365mfc Butene Isopentane E134 R245ca Isohexane Pentane Trans-butene Hexane Heptane Cis-butene Cyclohexane Benzene
Fig. 11. Thermal and exergy efciency of the ORC for different working uids at Teva 85  C.

Thermal Efficiency ( ) Exergy Efficiency-LS-3 ( )

Exergy Efficincy -IND300 ( )

the ORC is not necessarily achievable through solar heat source. Thus the capabilities of different working uids should be compared in ORCs with similar collector temperatures. Solar collectors can be categorized based on the temperature level that they can maintain. Generally there are three temperature level solar collectors [2]: (1) Low temperature solar collectors: with the output temperature less than 85  C. Flat plate solar collectors are in this category.

(2) Medium temperature solar collectors: with the output temperature below 130e150  C. Most evacuated tube collectors are in this category. (3) High temperature solar collectors: with the output temperature higher than 150  C. Parabolic trough collectors are mainly in this category. High temperature solar collectors are suitable for large-scale power generation applications. In this section a comparison

70 60 50 wnet (KJ/Kg) 40 30 20 10 0

Teva= 85 C

C4F10 C5F12 R227ea RC318 R423A R236fa R416A R236ea Neopentane Isobutane Toluene R245fa Butane Isobutene R365mfc Butene Isopentane E134 R245ca Isohexane Pentane Trans-butene Hexane Heptane Cis-butene Cyclohexane Benzene
Fig. 12. Net output power of the ORC for different working uids at Teva 85  C.

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

667

12

Teva= 85 C

10

8
VER

0
C4F10 C5F12 R227ea RC318 R423A R236fa R416A R236ea Neopentane Isobutane Toluene R245fa Butane Isobutene R365mfc Butene Isopentane E134 R245ca Isohexane Pentane Trans-butene Hexane Heptane Cis-butene Cyclohexane Benzene
Fig. 13. Vapor expansion ratio in the ORC for different working uids at Teva 85  C.

between performance factors of ORCs employing different working uids at Teva 85  C and 130  C will be made. Figs. 8e13 show variation of performance factors of the ORC for different working uids at two evaporating temperatures. As it can be observed for almost all uids in both evaporating temperatures, the thermal efciency and exergy efciency of the ORC have the same trend with respect to changing working uids. Therefore the thermal efciency and exergy efciency of the cycle play the same role in selecting the proper working uid. The main criterion to select working uid is considered the cycle thermal/exergy efciency in this study. Among uids with the same order of the cycle thermal/exergy efciency, the net power output of the ORC is a determinant factor to select the working uid. The third step will be eliminating uids with a high VER at the close level of efciencies and Wnet . As maximum thermal efciency calculations show and Figs. 8e13 conrm, refrigerants have a lower capacity to produce power

through the ORC. On the other hand refrigerants are less ammable and in some cases less hazardous than non-refrigerant uids. Therefore uids can be analyzed in two different categories: refrigerants and non-refrigerants. Figs. 8 and 9 show that among refrigerants, R-245fa, R-265mfc, and R-245ca provide higher cycle efciency and Wnet . R-365mfc has a higher VER in the cycle but almost the same cycle efciency and Wnet as shown in Fig. 10. Thus the nal refrigerants selected at medium temperature level (Teva 130  C) will be R-245fa and R-245ca. For non-refrigerants, at the medium temperature level, Toluene, Cyclohexane, Acetone, and Benzene provide higher cycle efciency as illustrated in Fig. 8. Among them, Acetone and Benzene produce more Wnet as can be recognized in Fig. 9. Fig. 10 shows that Benzene has higher VER in the cycle, but because of higher cycle efciency with respect to Acetone, we will keep it in our nal uids list. According to Figs. 8 and 9, there are eight non-refrigerant uids, cycle efciency and Wnet of which are about the same level.

a
35
Irreversibility reduction ( )

Teva=130 deg C

Teva=85 deg C

b
) Exergy Efficiency Enhancement (

Teva=130 deg C 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 70 80

Teva=85 deg C

30 25 20 15 10 5 0 70 80 90 ) 100 Collector Efficiency (

90 )

100

Collector Efficiency (

Fig. 14. (a) Irreversibility reduction, (b) Exergy efciency enhancement by increasing collector efciency from 70% to 100% for Isopentane.

668

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

Table 5 Regeneration Effects on thermal efciency, exergy efciency, and irreversibility of a solar ORC employing IND300 and LS-3 solar collectors for different working uids. Working uid Teva 85  C hth (%) NA 0.18 0.91 0.40 0.76 0.52 0.62 1.32 1.05 1.01 0.56 hex (%) for IND300 NA 0.13 0.64 0.29 0.53 0.37 0.44 0.92 0.74 0.72 0.40 I (%) for IND300 NA 1.55 7.96 3.58 6.28 4.71 5.52 10.97 8.95 8.82 5.01 hex (%) for LS-3 NA 0.14 0.68 0.30 0.57 0.39 0.46 0.98 0.78 0.76 0.43 I (%) for LS-3 NA 1.60 8.19 3.70 6.48 4.85 5.69 11.27 9.19 9.06 5.16 Teva 130  C hth (%) 0.02 0.94 1.75 0.73 2.27 NA NA 3.15 2.35 2.04 1.02 hex (%) for IND300 0.02 0.62 1.18 0.54 1.48 NA NA 2.07 1.57 1.37 0.69 I (%) for IND300 0.14 5.35 10.84 5.10 12.23 NA NA 17.47 13.75 12.44 6.45 hex (%) for LS-3 0.02 0.70 1.30 0.60 1.66 NA NA 2.30 1.74 1.51 0.76 I (%) for LS-3 0.15 5.75 11.43 5.41 13.02 NA NA 18.39 14.49 13.09 6.83

Acetone Benzene Butane Cis-butene Cyclohexane E134 Isobutene Isopentane R-245ca R-245fa Trans-butene

However their cycle efciency and power level is lower than cycle efciency and power level of the other four non-refrigerant uids mentioned above, it is good to nd the most suitable uids out of this group to add to the nal uids list. Fig. 10 shows that from this group of eight, Butane, Isopentane, Trans-butene, and Cis-butene have a lower VER in the cycle, so they can be in the nal list as medium performance non-refrigerants. All together, the nal selected uids for a solar ORC at medium temperature level are as follows: (a) R-245fa and R-245ca in the refrigerant group (b) Acetone and Benzene in the high performance non-refrigerant group (c) Butane, Isopentane, Trans-butene, and Cis-butene in the medium performance non-refrigerant group At the low temperature level the procedure is exactly the same as at the medium temperature level. As some uids have lowcritical temperatures (i.e. R-227ea) they are included in the low temperature level analysis but are not included in the medium temperature level analysis. The only liquid that is included in the medium temperature analysis but it is not included in the low temperature analysis is Acetone. Acetone is a wet uid whose slope of temperatureeentropy curve is very close to innity. The isentropic efciency of the turbine is assumed 80% in this study. With this amount of isentropic efciency the uid across the turbine falls into the wet region for Teva less than 124  C. This means that for Acetone in addition to the higher limit, there is lower limit for Teva. Figs. 11 and 12 show that among refrigerants R-245fa, R-365mfc, E134, and R-245ca provide a higher cycle efciency and Wnet . R365mfc has a higher VER but almost the same cycle efciency and Wnet as shown in Fig. 13. Thus the nal refrigerants selected at the low temperature level (Teva 85  C) will be R-245fa, E134, and R245ca.

For non-refrigerants at the low temperature level Toluene, Cyclohexane and Benzene provide a higher cycle efciency as illustrated in Fig. 11. As can be seen in Figs. 12 and 13, their wnet and VER are at the same level. As depicted in Figs. 11 and 12, eleven nonrefrigerant uids have very close cycle efciencies and power outputs. Although their cycle efciencies and power outputs are lower than cycle efciencies and power outputs of non-refrigerants mentioned above, it is good to pick the most suitable uids out of this group as it has been done for the medium temperature level. Fig. 11 shows that from this group of eleven, Butene has a lower exergy efciency; Toluene, Isohexane, Pentane, Hexane, and Heptane have a higher VER as illustrated in Fig. 13. Since Butane,

a
3.5 3 2.5 2 1.5 1 0.5 0 Acetone

Teva=130C Thermal efficiency enhancement (%) Exergy efficiency enhancement -IND300 (%) Exergy efficiency enhancement-LS-3 (%)

Isopentane Isopentane

Benzene

Butane

R245ca

b
20 18 16 14 12 10 8 6 4 2 0 Acetone

Trans-butene

Teva=130C Irreversibility reduction-IND300 (%) Irreversibility reduction-LS-3 (%)

Table 6 Molecular complexity of working uids. Working uid Acetone Benzene Butane Cis-butene Cyclohexane E134 Isobutene Isopentane R-245ca R-245fa Trans-butene

s
2.13 4.22 1.78 1.18 8.97 0.3 0.56 7.2 4.33 2.76 0.14

Benzene

Butane

R245ca

Fig. 15. (a) Thermal and exergy efciency enhancement, (b) Irreversibility reduction by using regenerative ORC based on molecular complexity of working uids (Teva 130  C).

Trans-butene

Cyclohexane

Cis-butene

R245fa

Cyclohexane

Cis-butene

R245fa

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

669

Isobutene, Isopentane, Trans-butene, and Cis-butene have a lower VER in the cycle, they can be in the nal list as medium performance non-refrigerants. All together, the nal selected uids for a solar ORC at the low temperature level are as follows: (a) R-245fa, E134 and R-245ca in the refrigerant group (b) Benzene and Cyclohexane in the high performance nonrefrigerant group (c) Butane, Isobutene, Isopentane, Trans-butene, and Cis-butene in the medium performance non-refrigerant group 5.4. Exergy efciency enhancement in a solar ORC Irreversibility in solar thermal systems is relatively high because of the high temperature difference between the solar collector and the apparent sun temperature. Collector efciency improvement and use of the regenerative ORC instead of the basic cycle are investigated in this section to reduce irreversibility of a solar ORC. Exergy efciency enhancement and irreversibility reduction are calculated for all 11 selected uids when the collector efciency increases from 70% to 100% at low and medium temperature levels. Seventy percent has been selected as a benchmark because the collector efciency for two selected models, IND300 and LS-3, for most analyzed uids in this study is close to 70% for a large interval of Teva . Calculations show that the exergy efciency variation with respect to collector efciency at each temperature level not only

has the same trend but also the exergy efciency enhancement percentage is almost the same for all selected uids. It is also correct for the irreversibility reduction trend and percentage. Fig. 14 shows the variation of irreversibility reduction and exergy efciency enhancement for Isopentane as a representative. As seen in Fig. 14, the theoretical limit for irreversibility reduction through collector efciency improvement for two selected collector models, IND300 and LS-3, is 35%. It also shows this limit is 5% for the exergy efciency enhancement. The second method to reduce irreversibility of a cycle is using the regenerative cycle instead of the basic cycle. Regeneration reduces the absorption heat while keeping the net power output constant. In other words regeneration enhances thermal and exergy efciency simultaneously. For the 11 selected uids regenerative cycles with regeneration efciency 3reg 0.8 have been investigated at low and medium temperature levels. As illustrated in Table 5, in contrast with collector efciency improvement effect on the exergy efciency of the cycle, regenerations effect on the exergy efciency of the ORC is uid dependent. Molecular complexity (s) is proposed as a criterion to nd a trend of regeneration effect on the ORC performance. Molecular complexity is dened as [15]:

Tcr R



vS vT

 (20)
SV;Tg0:7

Teva=85C Thermal efficiency enhancement ( ) Exergy efficiency enhancement -IND300 ( ) Exergy efficiency enhancement-LS-3 ( )

Trans-butene

Cis-butene

Isobutene

Isopentane

Cyclohexane

Benzene

Butane

R245fa

E134

R245ca

1.4 1.2 1 0.8 0.6 0.4 0.2 0

where Pr and Tr are reduced pressure and temperature respectively, R is gas constant and SV stands for saturation vapor. The higher slope of the entropyetemperature diagram results in higher molecular complexity. Table 6 shows the molecular complexity of selected uids. Figs. 15 and 16 show thermal and exergy efciency enhancement and irreversibility reduction of the ORC by using the regenerative cycle for selected working uids. In these gures uids are arranged in the horizontal axis in ascending order of molecular complexity.

Discarding chlorine included fluids Discarding highly wet fluids Setting higher temperature level of the cycle based on the collector type Calculating thermal efficiency, exergy efficiency, net power generated, and vapor expansion ratio Categorizing working fluids into: - Refrigerants - High performance non-refrigerants - Medium performance non-refrigerants

b
12 10 8 6 4 2 Trans-butene Cis-butene Isobutene

Teva=85C Irreversibility reduction-IND300 ( ) Irreversibility reduction-LS-3 ( )

Isopentane

Fig. 16. (a) Thermal and exergy efciency enhancement, (b) Irreversibility reduction by using regenerative ORC based on molecular complexity of working uids (Teva 85  C).

Cyclohexane

Benzene

Butane

R245fa

E134

R245ca

Choosing working fluids in the following order: - Highest thermal/exergy efficiency - Highest net power generated - Lowest vapor expansion ratio
Fig. 17. Proposed selection procedure of the working uid in a solar ORC.

670

R. Rayegan, Y.X. Tao / Renewable Energy 36 (2011) 659e670

At both temperature levels and based on all performance factors discussed in this section higher molecular complexity results in a more effective regenerative cycle. The only exceptions to this rule are Benzene and Cyclohexane. This means that the regeneration will be more effective in ORCs employing high molecular complexity working uids if they are not Cyclohydrocarbons. 6. Conclusions A comprehensive list of working uids has been analyzed to nd the most suitable uids to operate a solar ORC. A procedure to compare working uids capabilities when they are employed in the solar Rankine cycles with similar working conditions has been proposed. This procedure can be summarized and illustrated in Fig.17. At the rst step of the procedure Chlorine included uids and wet uids have been discarded. The slope of the temperatureeentropy curve for some wet uids is very close to innity. These uids have been selected at this step. The maximum practical thermal efciency and corresponding performance factors for preselected working uids conrm that uids with higher critical temperature have better performance in the ORC. Calculation shows that the thermal efciency higher than 25% and the exergy efciency higher than 20% are achievable in ORCs. In the next step a comparison between different ORC working uids based on uids effect on the thermal/exergy efciency, net power generated, and vapor expansion ratio of the Rankine cycle has been accomplished. Thermal efciency and exergy efciency of the ORC have the same trend with respect to changing working uids. Therefore thermal efciency and exergy efciency of the cycle play the same role in selecting a proper working uid. In the investigation, two temperature levels for Teva have been considered which are 85  C and 130  C as representatives of low temperature and medium temperature solar collectors. The main criterion for selecting a working uid in this study was the thermal/exergy efciency. After that among uids with the same order of thermal/exergy efciency, net power output of the ORC is a determinant factor to select the working uid. The thirdstep will be eliminating uids with a high vapor expansion ratio at the close level of cycle efciencies and power outputs. Fluids have been divided into two groups: refrigerants and nonrefrigerants. Fluids with the best performance in the ORC have been recognized in each group. In the non-refrigerants group, two different subdivisions have been considered: high performance uids and medium performance uids. The reason for this subdivision is that most non-refrigerants are in the medium performance group. Then by considering all non-refrigerants as one group, a large group of uids would be omitted from analysis. At medium temperature level the nal selected refrigerants through the introduced procedure are R-245fa and R-245ca. The nal selected non-refrigerants at Teva 130  C are Acetone and Benzene with the high performance and Butane, Isopentane, Transbutene, and Cis-butene with the medium performance. At the low temperature level only a few numbers of uids have been changed in comparison to uids selected at the medium temperature level. At Teva 85  C, E134 has been added to the selected refrigerants at Teva 130  C. In the non-refrigerants group

Acetone has been replaced by Cyclohexane and Isobutene has been added to the uids with the medium performance capability. Collector efciency improvement and use of regenerative ORC instead of the basic cycle to reduce irreversibility of a solar ORC were investigated in the last section. Exergy efciency enhancement and irreversibility reduction have been calculated for all 11 selected uids when the collector efciency increases from 70% to 100% at low and medium temperature levels. Calculation results show that the theoretical limit for irreversibility reduction through collector efciency improvement for two selected collector models, IND300 and LS-3, is 35%. It also shows this limit is 5% for the exergy efciency enhancement. For the 11 selected uids a regenerative cycle with regeneration efciency 3reg 0.8 have been investigated at low and medium temperature levels. In contrast to collector efciency improvement effect on the exergy efciency of the cycle, regenerations effect on the ORC is uid dependent. Calculation results show, at the two temperature levels studied, the regeneration will be more effective in ORCs employing high molecular complexity working uids except for Cyclohydrocarbons. Acknowledgments The nancial support of a Florida International University Doctoral Evidence Acquisition Fellowship is gratefully acknowledged. References
[1] Hermann WA. Quantifying global exergy resources. Energy 2006;31:1349e66. [2] Bruno JC, Lopez-Villada J, Letelier E, Romera S, Coronas A. Modeling and optimization of solar organic Rankine cycle engines for reverse osmosis desalination. Applied Thermal Engineering 2008;28:2212e26. [3] Delgadeo-Torres AM, Garcia-Rodriguez L. Preliminary assessment of solar organic Rankine cycles for driving a desalination system. Desalination 2007;216:252e75. [4] Delgadeo-Torres AM, Garcia-Rodriguez L. Double cascade organic Rankine cycle for solar driven reverse osmosis desalination. Desalination 2007;216:306e13. [5] Lemmon EW, Huber ML, McLinden MO. NIST standard reference database 23: reference uid thermodynamic and transport properties-REFPROP, version 8.0. Gaithersburg: National Institute of Standards and Technology, Standard Reference Data Program; 2007. [6] United Nations Environmental Programme. Montreal protocol on substances that deplete the ozone layer. Website: http://ozone.unep.org/teap/Reports/ TEAP_Reports/Teap-CUN-nal-report-Sept-2006.pdf; September 2006 [accessed 17.08.09]. [7] Hung TC. Waste heat recovery of organic Rankine cycle using dry uids. Energy Conversion and Management 1995;42:539e53. [8] Drescher U, Bruggemann D. Fluid selection for the Organic Rankine Cycle (ORC) in biomass power and heat plants. Applied Thermal Engineering 2007;27:223e8. [9] Rayegan R, Tao YX. A critical review on single component working uids for Organic Rankine Cycles (ORCs). ASME Early Career Technical Journal 2009;8:20.1e8. [10] Cengel YA, Boles MA. Thermodynamics: an engineering approach. 5th ed. McGraw-Hill; 2006. [11] Petela R. Exergy of heat radiation. ASME Journal of Heat Transfer 1964;68:187. [12] Rabl A. Active solar collectors and their applications. New York: Oxford University Press; 1985. [13] Eck M, Steinmann WD. Direct steam generation in parabolic troughs: rst results of the DISS project. ASME Journal of Solar Energy Engineering 2002;124:134e9. [14] Bejan A. Entropy generation through heat and uid ow. John Wiley and Sons Inc.; 1982. [15] Invernizzi C, Iora P, Silva P. Bottoming mico-Rankine cycles for micro-gas turbines. Applied Thermal Engineering 2007;27:100e10.