Applied Thermal Engineering 23 (2003) 2936

www.elsevier.com/locate/apthermeng

A verication study of steam-ejector refrigeration model

National Technical University of Athens, Mechanical Engineering Department, Thermal Section, 9 Heroon Polytechniou Av., Zografou Campus, 15780 Athens, Greece Received 10 April 2002; accepted 9 August 2002

Abstract In this paper a verication study of steam ejector refrigeration model is conducted. A mathematical model for a steam ejector refrigeration cycle is described and the produced results are compared with experimental ones found in the literature. The present work uses the theory which was developed by Munday and Bagster and takes into account the shock phenomena and isentropic eciency in the ejector. 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Steam; Steam ejector; Refrigeration cycle; Coecient of performance (COP)

1. Introduction An improved ejector theory was developed by Munday and Bagster [1]. This theory depends on the assumption that two discrete streams, the motive stream and the secondary stream, maintain their identity down the converging duct of the entrainment region. At some section the secondary ow reaches sonic velocity. Thermodynamic shock and mixing occur at the very end of the converging cone resulting in a transient supersonic mixed stream. There is no supersonic deceleration and a shock takes place immediately on mixing. The mixed stream will be undergo a thermodynamic shock process and becomes subsonic at a point determined by the intersection of the Fanno and Rayleigh lines. After that the combined stream is brought to near zero velocity in the subsonic diuser. In the present work this theory is used to develop a computer model of an ejector with particular reference to steam ejector refrigerator system at various operating conditions and evaluate the variation of the performance of the system and the coecient of performance (COP). The
*

Corresponding author. Tel.: +301-772-3966; fax: +301-772-3670. E-mail address: rogdemma@central.ntua.gr (E.D. Rogdakis).

1359-4311/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S 1 3 5 9 - 4 3 1 1 ( 0 2 ) 0 0 1 1 6 - 3

30

G.K. Alexis, E.D. Rogdakis / Applied Thermal Engineering 23 (2003) 2936

Nomenclature A COP P Pcr Px Py _ Q T V _ W h _ m n s w x area of cross-section of constant area duct, m2 coecient of performance pressure, bar pressure at sonic ow of secondary vapors, bar pressure before the shock, bar pressure after the shock, bar heat rate, kW temperature, C velocity, m/s work rate, kW enthalpy, kJ/kg mass ow, kg/s isentropic eciency entropy, kJ/kg K ow entrainment ratio quality

Greek symbol v specic volume, m3 /kg Subscripts 1,2. . . cycle locations a; b . . . ejector locations c condenser e evaporator f saturated liquid g generator, saturated vapor p pump

performance of the system can be calculated as the evaporation per unit of motive steam. This parameter is used to verify the mathematical results with the experimental results in the literature. Fig. 1 illustrates the operation of the ejector system and a schematic view of ejector. High pressure saturated vapor raised in the generator (1). This vapor passes through a (converging/ diverging) nozzle, drawing stem into the ejector from the evaporator (2) where water remaining there is cooled by evaporation. The two streams mix in the ejector and leave it after a recovery of pressure in the diuser part of the ejector (3). Then, heat is rejected from the uid to the surroundings, resulting in condensate at the exit of the condenser (4). This is divided into two streams. One enters the evaporator after a pressure reduction in the expansion valve (5) and the other enters the generator after a pressure rise in the pump (6). A computer program, based on Munday and Bagster theory, was written in order to calculate the behavior of an ejector and the performance of ejector system for a range of parameters. For

G.K. Alexis, E.D. Rogdakis / Applied Thermal Engineering 23 (2003) 2936

31

Fig. 1. Schematic view of steam ejector refrigeration system and ejector.

thermodynamic properties of water, the equations proposed by Ziegler and Trepp [2] have been used.

2. Ejector and cycle analysis An ejector is a device in which a high pressure jet of uid (motive stream) is used to entrain low pressure uid (secondary stream). The resulting mixture is discharged at a pressure that lies between the driving pressure and the suction pressure. Rao and Singh [3] and Kouremenos et al. [4] showed that the model based on the conservation of uid momentum in mixing process gives results in close agreement with those obtained from available design charts. To better understand how a typical ejector functions, a description of its operation will be given on a T s chart in Fig. 2. Saturated motive steam enters the ejector at a high pressure Pg , temperature Tg and zero velocity (i.e. stagnation condition) corresponding to state (1) and expands to a pressure at state (a1), with eciency n1a1 0:7. The motive stream emerges from the nozzle and remains as an identiable stream for some distance downstream. The saturated secondary vapor enters the ejector at pressure Pe and zero velocity (i.e. stagnation condition) corresponding to state (2) and expand isentropic to a pressure at state (a2). The pressures at states (a1) and (a2) corresponding to pressure at sonic ow of secondary vapor Pcr . In general the mixing zone may be expected to take place in some region (ab) of the converging cone. Munday and Bagster estimated that the resulting velocity of the mixture is always supersonic. Since the supersonic mixed stream is decelerated in a converging section with corresponding raising in pressure Px , region (bc) and an assumed eciency of nbc 0:8. If the stream is still supersonic at the end of the cone (c), a shock will occur in the duct of constant cross-section, resulting in a subsonic stream (d). The intersection of the Fanno and Rayleigh lines determines the pressure Py and temperature Td after shock eect. The stream is then

32

G.K. Alexis, E.D. Rogdakis / Applied Thermal Engineering 23 (2003) 2936

Fig. 2. Ts chart of steam ejector refrigeration cycle.

brought to near zero velocity (i.e. stagnation condition) corresponding to state (3) in the diuser, with an assumed eciency of nd3 0:8. In the present work this theory is used with the assumption that the uid momentum is conserved in the mixing section, i.e., Va1 wVa2 1 wVb Also the overall energy balance equation, can be written as: h1 wh2 1 wh3 h1 hT1 ; P1 h2 hT2 ; P2 2 3 4 1

_ _ where w me =mg the ow entrainment ratio (kg of secondary stream per kg of motive stream). The governing equations for each section in the ejector are: Nozzle: Energy balance equation between points (1) and (a1) is:
2 Va1 =2 h1 ha1

5 6 7 8

The enthalpy ha1 is calculated from the system of equations: s1 sT1 ; P1 sas xas sgas 1 xas sfas has xas hgas 1 xas hfas n1a1 h1 ha1 =h1 has Intake pipe: Energy balance equation between points (2) and (a2) is:

G.K. Alexis, E.D. Rogdakis / Applied Thermal Engineering 23 (2003) 2936

33

9 10 11

The enthalpy ha2 is calculated from the system of equations: s2 sT2 ; P2 sa2 xa2 sga2 1 xa2 sfa2 ha2 xa2 hga2 1 xa2 hfa2

Converging cone: Assumed a value for the w (0 < w < 1), the enthalpy h3 is calculated from the Eq. (2). Energy balance equation between points (3) and (b) is: Vb2 =2 h3 hb 12 The enthalpy hb is calculated from the Eqs. (1) and (12). Assumed a value for the pressure Px (Pb < Px < P3 ), the enthalpy hc and the specic volume vc are calculated from the system of equations: hb xb hgb 1 xb hfb sb xb sgb 1 xb sfb scs sb xcs sgcs 1 xcs sfcs hcs xcs hgcs 1 xcs hfcs nbc hb hcs =hb hc hc xc hgc 1 xc hfc vc xc vgc 1 xc vfc 13 14 15 16 17 18 19

_ _ _ _ Constant cross-section: Assumed a constant value for the ratio m=A (m me mg ) the pressure Pd Py and the temperature Td are calculated from the intersection of Fanno and Rayleigh lines: hd m2 m=A2 =2 c1 d _ _ Pd md m=A2 c2 hd hTd ; Pd md mTd ; Pd The constants c1 , c2 are calculated from the same equations at state (c). Diuser: The temperature T3 is calculated from the system of equations: sd sTd ; Pd s3s sd sT3s ; Pd h3s hT3s ; Pd nd3 hd h3s =hd h3 24 25 26 27 20 21 22 23

34

G.K. Alexis, E.D. Rogdakis / Applied Thermal Engineering 23 (2003) 2936

h3 hT3 ; P3

28

Given the pressures for the states (1)(3) and using iterative calculations, the maximum value of the ratio w can be found. The basic equations are obtained from the conservation law for energy are: Evaporator _ _ Qe me h2 h5 Generator _ _ Qg mg h1 h6 Pump _ _ Wp mg h6 h4
Table 1 Deviations between present work approach and experimental data given by Jackson Pg (bar) 7.9 tc (C) 32.9 40.8 36.8 42.5 49.5 42.5 te (C) 10.0 10.0 15.6 4.4 10.0 15.6 Present work w 0.7978 0.4783 0.9067 0.2688 0.2437 0.5943 _ m=A (kg/sm ) 7.0
2

29 30 31

Jackson COP 0.7223 0.4324 0.8239 0.2420 0.2200 0.5395 wqe (kJ/kg) 1898.8 1122.4 2152.4 626.4 563.1 1396.6 wqe (kJ/kg) 1900 1120 2150 620 560 1400

Deviation (%) )0.06 0.21 0.11 1.02 0.55 )0.24

Table 2 Deviations between present work approach and experimental data given by Stinson Pg (bar) 7.9 tc (C) 33.6 te (C) 4.4 7.2 10.0 12.8 15.6 4.4 7.2 10.0 12.8 15.6 4.4 7.2 10.0 12.8 15.6 Present work w 0.5043 0.5689 0.6694 0.7770 0.9207 0.3811 0.4423 0.5294 0.5696 0.7218 0.2661 0.2656 0.2650 0.2644 0.2638 _ m=A (kg/sm ) COP 7.5 0.4546 0.5139 0.6060 0.7053 0.8371 0.3432 0.3992 0.4788 0.5164 0.6558 0.2604 0.3322 0.3935 0.4604 0.5301
2

Stinson wqe (kJ/kg) 1161.3 1349.4 1591.1 1852.0 2198.2 893.7 1039.6 1246.9 1344.7 1707.7 673.4 859.1 1017.6 1190.6 1370.8 wqe (kJ/kg) 1200 1340 1600 1740 2170 890 1040 1250 1340 1700 680 860 1020 1190 1370

Deviation (%) )3.33 0.69 )0.56 6.04 1.28 0.41 )0.04 )0.25 0.35 0.45 )0.98 )0.10 )0.23 0.05 0.06

7.9

38.8

7.5

7.9

43.1

7.5

G.K. Alexis, E.D. Rogdakis / Applied Thermal Engineering 23 (2003) 2936

35

Thus the COP of the system is determined by its operating conditions and may be calculated from equation: _ _ _ COP Qe =Qg Wp wh2 h5 =h1 h4 3. Comparison with experimental results There are plenty of experimental data available in the literature for steam ejector refrigeration. Some of them have been published by Jackson [5], Stinson [6], Petzold [7] and Eames et al. [8]. The experimental data given by Stinson and Eames are more systematic, covering a wide region of temperatures. The performance of the system can be calculated as the evaporation per unit of motive steam (wqe ). The deviations (%) between present works approach and experimental data given by above scientists are shown in Tables 14.
Table 3 Deviations between present work approach and experimental data given by Petzold Pg (bar) 11.05 tc (C) 37.4 38.5 39.6 te (C) 11.3 17.6 22.3 Present work w 0.6806 0.8467 0.9580 _ m=A (kg/sm2 ) 10.0 COP 0.6132 0.7663 0.8701 wqe (kJ/kg) 1608.7 2007.0 2274.8 Petzold wqe (kJ/kg) 1090 1590 1980 Deviation (%) 32.24 20.77 12.96

32

Table 4 Deviations between present work approach and experimental data given by Eames et al. tg (C) 140 135 130 125 120 140 135 130 125 120 140 135 130 125 120 tc (C) 34.4 33.4 30.8 27.8 26.5 35.3 33.4 31.5 29.5 27.3 36.3 34.0 31.9 30.0 28.3 te (C) 5 Present work w 0.1958 0.2748 0.2996 0.3722 0.4407 0.2610 0.3230 0.3840 0.4507 0.5369 0.3314 0.4246 0.5127 0.5751 0.6275 _ m=A (kg/sm ) 4.5
2

Eames et al. COP 0.1789 0.2518 0.2753 0.3431 0.4074 0.2390 0.2965 0.3535 0.4161 0.4973 0.3039 0.3905 0.4729 0.5320 0.5822 COP 0.1779 0.2513 0.2756 0.3442 0.4044 0.2334 0.2965 0.3553 0.4189 0.5004 0.3093 0.3892 0.4734 0.5374 0.5862

Deviation (%) 0.56 0.20 )0.11 )0.32 0.74 2.34 0.00 )0.51 )0.67 )0.62 )1.77 0.33 )0.10 )1.01 )0.69

7.5

5.0

10

5.5

36

G.K. Alexis, E.D. Rogdakis / Applied Thermal Engineering 23 (2003) 2936

It is obvious that the present model approaches the experimental data given by Stinson and Eames et al. are better than the others.

4. Conclusions Thermodynamic analysis of a steam ejector refrigeration system were presented in this paper. The present model takes into account the shocking phenomena and assumes that the mixing process of the two streams, motive and secondary stream, takes place with constant pressure and the conservation of uid momentum in eect. The comparison between present calculated model and experimental data showed that this model estimates satisfying the experimental data. Consequently this model could be used for parametric study of a steam ejector refrigeration unit and calculation of thermodynamic properties in the refrigeration cycle.

References
[1] J.T. Munday, D.F. Bagster, A new ejector theory applied to steam jet refrigeration, Ind. Eng. Chem. Process. Res. Dev. 16 (4) (1977) 442449. [2] B. Ziegler, Ch. Trepp, Equation of state for ammonia water mixture, Rev. Int. Froid 7 (2) (1984) 101106. [3] S.P.R. Rao, R.P. Singh, Performance characteristics of single-stage jet ejectors using two simple models, Chem. Eng. Commun. 66 (1988) 207219. [4] D.A. Kouremenos, E.D. Rogdakis, G.K. Alexis, Optimization of enhance steam-ejector applied to steam jet refrigeration, Proceedings of ASME, Anaheim, CA, 38, 1998, pp. 1926. [5] D.H. Jackson, Vacuum refrigeration, Ind. Eng. Chem. 28 (1936) 522526. [6] L.S. Stinson, Steam-jet vacuum cooling units, Refrig. Eng. 46 (1943) 316319. [7] M. Petzold, Graphische Methode zer Bestimmung des Dampfverbrauches von Brudenverdichtern, Chem. Ing. Tech. 22 (1950) 147150. [8] I.W. Eames, S. Aphornratana, H. Haider, A theoretical and experimental study of a small-scale steam jet refrigerator, Int. J. Refrig. 18 (6) (1995) 378386.