Accepted Manuscript

Off-design performance analysis of a power-cooling cogeneration system combining a

Kalina cycle with an ejector refrigeration cycle

Yang Du, Yiping Dai

PII: S0360-5442(18)31403-8
DOI: 10.1016/j.energy.2018.07.106
Reference: EGY 13371

To appear in: Energy

Received Date: 23 March 2018

Revised Date: 13 July 2018
Accepted Date: 16 July 2018

Please cite this article as: Du Y, Dai Y, Off-design performance analysis of a power-cooling cogeneration
system combining a Kalina cycle with an ejector refrigeration cycle, Energy (2018), doi: 10.1016/
j.energy.2018.07.106.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

1 Off-design performance analysis of a power-cooling cogeneration system

2 combining a Kalina cycle with an ejector refrigeration cycle

3 Yang Du, Yiping Dai *

4 Institute of Turbomachinery, School of Energy and Power Engineering,

PT
5 Xi'an Jiaotong University, Xi'an, Shaanxi 710049, China

RI
6

7 Abstract

SC
8 This paper conducts the off-design performance analysis of a novel

U
9 power-cooling cogeneration system combining a Kalina cycle and an ejector
AN
10 refrigeration cycle for low-grade hot water. Five plate heat exchangers, a separator, an

11 axial inflow turbine, two pumps, an ejector and two throttle valves are adopted. The
M

12 ejector refrigeration cycle using R134a is driven by the ammonia-poor solution from
D

13 the separator. A novel method for predicting the off-design performance of the
TE

14 power-cooling cogeneration system is proposed. Variable hot water parameters,

15 condensation temperature and evaporator temperature are analyzed by the sliding

EP

16 pressure operation approach. The results indicate that the system shows 619.74 kW
C

17 net power and 71.28 kW cooling at design conditions. As the mass flow rate ratio or
AC

18 the inlet temperature of hot water increases, the net power, thermal efficiency and

19 exergy efficiency increase, while the cooling and cooling exergy decrease. The exergy

20 efficiency reaches the maximum of 39.82% at the saturated evaporator temperature of

21 6 °C. The cooling is more strongly affected by the hot water inlet temperature than the

22 saturated condensation temperature, while the turbine efficiency, net power, thermal
 ACCEPTED MANUSCRIPT

23 efficiency and exergy efficiency are more strongly affected by the saturated

24 condensation temperature than the hot water inlet temperature.

25 Key words: Power and cooling cogeneration; Kalina cycle; Ejector refrigeration

26 cycle; Exergy efficiency; Off-design performance.

PT
27 * Corresponding author. Tel. : +86 029 82668704; Fax: +86 029 82668704.

RI
28 E-mail address: ypdai@mail.xjtu.edu.cn

29 Nomenclature

SC
A heat transfer area (m2) Subscripts

U
Bo boiling number 1-15 state point of working
AN
fluid

c_d channel distance (mm) basic basic ammonia-water

M

D diameter (m) boi boiler

D

E exergy (kW) ce constant section region

TE

exit of ejector

f friction factor cold cold

EP

G mass velocity (kg/m2 s) con condenser

C

h specific enthalpy (J/kg) con I condenser I

AC

H pressure head (m) con II condenser II

J convection heat transfer cooling cooling

coefficient (W/ m2 K)

m mass flow rate (kg/s) cr critical back pressure

N rotational speed (rpm) cw cooling water

 ACCEPTED MANUSCRIPT

Nu Nusselt number de diffuser exit of ejector

P pressure (kPa) des design

P_l plate length (m) eq equivalent

Pr Prandtl number eva evaporator

PT
P_t plate thickness (mm) ex exergy

RI
P_w plate width (m) gas gas

q volumetric flow rate (m3/s) gw geothermal water

SC
Q heat transfer rate (kW) h hydraulic

U
Re Reynolds number hot hot
AN
s specific entropy (J/kg K) hw hot water

T temperature (°C ) in inlet

M

uSL superficial velocity of liquid is isentropic

D

(m/s)
TE

uSV superficial velocity of vapor l liquid

(m/s)
EP

U overall heat transfer coefficient mi location where two

C

(W/m2 K) streams of ejector

AC

finish mixing

W power (kW) ne nozzle exit of ejector

x vapor quality; ammonia mass net net

fraction (%)

XLM Lock-Martinelli (LM) out outlet

 ACCEPTED MANUSCRIPT

parameter

pum I pump I

Greek pum II pump II

letters

PT
η efficiency poor ammonia-poor

RI
solution

θ chevron angle (°) power power

SC
λ thermal conductivity (W/m rich ammonia-rich vapor

U
K)
AN
ρ density (kg/m3) sh shock location

µ viscosity (N s/m2) t ejector nozzle throat

M

ξ ejector entrainment ratio th thermal

D

Φ two phase flow multiplier tur turbine

TE

φchi Chisholm’s constant v vapor

∆h specific enthalpy difference vap vapor generator

EP

(kJ/kg)
C

∆P pressure drop (kPa) wall plate wall

AC

∆Tlmtd log mean temperature y location where two

difference (°C ) streams of ejector

begin to mix

30

31
 ACCEPTED MANUSCRIPT

32 1. Introduction

33 Due to the shortage of fossil fuels and the global warming issue, the utilization of

34 renewable energy or waste heat is a promising solution. In order to recover the

35 low-grade thermal energy, a large number of power cycles, including the Kalina cycle,

PT
36 the organic Rankine cycle and the transcritical CO2 power cycle, have been proposed.

RI
37 During the evaporating process of Kalina cycle, ammonia-water shows variable

38 evaporating temperature, leading to a small temperature difference between the

SC
39 working fluid and the heat source in the evaporator [1]. Compared with the other

U
40 power cycles for low-grade heat source, the Kalina cycle has attracted more attention
AN
41 because of a higher exergy efficiency [2].

42 Apart from the power cycle, the ejector refrigeration cycle is also an efficient
M

43 solution for the utilization of the low-grade thermal energy due to a low maintenance
D

44 cost [3]. Researchers have widely investigated the ejector refrigeration cycle
TE

45 experimentally and theoretically [4-8]. It was found that compared with the ejector

46 refrigeration cycle using R290, the ejector refrigeration cycle using R134a obtained a
EP

47 higher coefficient of performance (COP) at a relatively high condensation temperature

C

48 [9].
AC

49 With diverse energy needs in daily life, the power-cooling cogeneration system

50 driven by the low-temperature heat source is a modern method for the increasing

51 power consumption and cooling energy consumption. Goswami et al. [10] firstly

52 proposed a power-cooling cogeneration system (Goswami cycle), in which an

53 ammonia-water Rankine cycle was integrated with an ammonia-water absorption

 ACCEPTED MANUSCRIPT

54 refrigeration cycle. A superheater was installed between the rectifier and the turbine to

55 improve the inlet temperature of turbine. Kim et al. [11-12] evaluated the effects of

56 thermodynamic parameters, including the turbine inlet pressure, the ammonia

57 concentration and the heat source inlet temperature, on the thermal efficiency of

PT
58 Goswami cycle. An optimal turbine inlet pressure was found when a maximal thermal

RI
59 efficiency was achieved. Fontalvo et al. [13] conducted the exergy analysis of

60 Goswami cycle, and indicated that the absorber and the boiler contributed mostly to

SC
61 the exergy destruction. Besides, a multi-objective optimization method was employed

U
62 by Pouraghaie et al. [14] to obtain optimal thermodynamic parameters of Goswami
AN
63 cycle. Zare et al. [15] analyzed the Goswami cycle in the view of thermo-economics

64 and achieved optimal solutions.

M

65 Various power-cooling cogeneration systems, based on a Kalina cycle, were also

D

66 proposed by researchers. Srinivas et al. [16-17] combined an ammonia-water

TE

67 single-effect absorption refrigeration cycle with a Kalina cycle for low-grade heat

68 source. According to the variable requirement of energy, they designed a facility to

EP

69 control the net power output and cooling output. They found optimal operating
C

70 parameters including the mass split ratio and the vapor fraction of separator. Besides,
AC

71 they [18] presented another power-cooling cogeneration system for solar energy. The

72 system could produce 15 kW cooling and 220 kW power. Zheng et al. [19] proposed a

73 novel system, based on a Kalina cycle, to produce power and cooling by replacing the

74 flash tank with a rectifier and adding heat exchangers. Furthermore, Yu et al. [20]

75 designed a combined power and cooling system by introducing an ammonia-water

 ACCEPTED MANUSCRIPT

76 absorption refrigeration cycle in Zheng’s cycle. The proposed new cycle had a

77 variable cooling to power ratio without the processes of splitting and mixing. Hua et

78 al. [21] proposed a modified Kalina cycle to produce power and cooling by adding a

79 sub-cooler and an evaporator. In order to further improve the energy conversion

PT
80 efficiency, Jing et al. [22] presented a power-cooling cogeneration system based on a

RI
81 Kalina cycle and a double-effect vapor absorption refrigeration cycle using

82 ammonia-water as working fluid. Cao et al. [23-24] coupled an ammonia-water

SC
83 absorption refrigeration cycle with a Kalina cycle for low temperature exhaust gas.

U
84 The results indicated that a higher turbine inlet pressure or a higher concentration of
AN
85 the basic ammonia-water solution led to a higher exergy efficiency. Wang et al. [25]

86 proposed a new power-cooling cogeneration system with a high cooling output by

M

87 combining a Kalina cycle with an absorption refrigeration cycle. They added a

D

88 separator between the condenser and the turbine to obtain the ammonia-water vapor
TE

89 with a higher concentration.

90 The configuration of a vapor absorption refrigeration cycle was complex, so that

EP

91 some researchers began to focus on the power-cooling cogeneration system

C

92 combining a Kalina cycle with an ejector refrigeration cycle [26]. Ghaebi et al. [27]
AC

93 presented a new power-cooling cogeneration system based on a Kalina cycle and an

94 ejector refrigeration cycle. Regarding to the ejector refrigeration cycle, the turbine

95 exhaust, as the ejector primary flow, was applied to suck the secondary flow from the

96 evaporator. Barkhordarian et al. [28] proposed a novel ammonia-water combined

97 power and cooling system, in which a variable power to cooling ratio was achieved by
 ACCEPTED MANUSCRIPT

98 adjusting the reflux ratio of the reboiler.

99 To sum up, power-cooling cogeneration systems, based on a Kalina cycle and an

100 absorption refrigeration cycle, have been widely investigated in the views of

101 thermodynamics and economics. Few researchers have investigated the power-cooling

PT
102 cogeneration system combining a Kalina cycle with an ejector refrigeration cycle.

RI
103 During actual operation process of a power-cooling cogeneration system, the heat

104 source parameters, the environment temperature, and the refrigeration temperature are

SC
105 variable, leading to the degradation of the cogeneration system performance. Thus, it

U
106 is essential to explore the off-design performance of a power-cooling cogeneration
AN
107 system and search for a proper operation method. Few works have been conducted on

108 this issue.

M

109 To fill this gap, this paper proposes a novel power-cooling cogeneration system
D

110 combining a Kalina cycle with an ejector refrigeration cycle for low-grade hot water.
TE

111 The ammonia-poor solution from the separator of Kalina cycle is the heat source of

112 the ejector refrigeration cycle. A combination of five plate heat exchangers, a
EP

113 separator, an axial inflow turbine, two pumps, an ejector and two throttle valves is
C

114 adopted. The particle swarm optimization algorithm is applied to obtain optimal
AC

115 operating parameters in the design phase. A novel method for predicting the

116 off-design performance of the power-cooling cogeneration system is proposed based

117 on the specifically designed ejector and plate heat exchangers. The mature sliding

118 pressure operation approach is adopted to response to variable hot water parameters,

119 saturated condensation temperature and saturated evaporator temperature. Parameters,

 ACCEPTED MANUSCRIPT

120 including the thermal efficiency, exergy efficiency, cooling output, cooling exergy

121 output and net power output, are evaluated at off-design conditions. The results

122 provide significant references for the design and the operation of a power-cooling

123 cogeneration system combining a Kalina cycle with an ejector refrigeration cycle.

PT
124 2. System description

RI
125 Fig.1 shows the schematic of a power-cooling cogeneration system based on a

126 Kalina cycle and an ejector refrigeration cycle. The heat source of the cogeneration

SC
127 system is hot water. The ejector refrigeration cycle is driven by the ammonia-poor

U
128 solution from the separator of Kalina cycle. The cogeneration system consists of a
AN
129 boiler, a separator, a turbine, a throttle valve (throttle valve I), a condenser (condenser

130 I), a pump (pump I) and an ejector refrigeration cycle. The ejector refrigeration cycle
M

131 includes a vapor generator, absorbing heat from the ammonia-poor solution of Kalina
D

132 cycle, an evaporator, an ejector, a throttle valve (throttle valve II), a condenser
TE

133 (condenser II) and a pump (pump II).

134 After being heated by the hot water in the boiler, the basic ammonia-water
EP

135 solution is separated into the saturated vapor stream (ammonia-rich vapor) and the
C

136 saturated liquid stream (ammonia-poor solution) in the separator. The ammonia-rich
AC

137 vapor expands and produces power in the turbine. The ammonia-poor solution rejects

138 heat to the ejector refrigeration cycle and its pressure is reduced in the throttle valve I.

139 Then the turbine exhaust is mixed with the ammonia-poor solution from the throttle

140 valve I. After releasing heat in the condenser I, the basic ammonia-water solution is

141 pumped to the boiler by the pump I.

 ACCEPTED MANUSCRIPT

142 Regarding to the ejector refrigeration cycle, the heat source is the ammonia-poor

143 solution from the separator of Kalina cycle. After absorbing heat from the

144 ammonia-poor solution in the vapor generator, the high pressure and high temperature

145 refrigerant R134a, as the ejector primary flow, enters into the ejector to suck the

PT
146 secondary flow from the evaporator. Fig.2 presents the ejector mechanism. A vacuum

RI
147 region is produced in the suction chamber of the ejector, when the high pressure

148 primary flow expands and its pressure is reduced in the nozzle of the ejector. In this

SC
149 way, the secondary flow could enter into the ejector and then be mixed with the

U
150 primary flow. In the diffuser of the ejector, the mixed flow pressure is increased. After
AN
151 leaving the ejector, the mixed flow rejects heat in the condenser II and condenses to

152 saturated liquid state. Then the saturated liquid refrigerant is separated into two parts:
M

153 one part is pumped to the vapor generator by the pump II; and the other part enters the
D

154 throttle valve II. After leaving the throttle valve II, the low pressure liquid refrigerant
TE

155 flows through the evaporator and absorbs heat from the refrigeration medium (water).

156 Thus, a cooling effect is produced in the ejector refrigeration cycle.

EP

157 For the power-cooling cogeneration system, the following assumptions are
C

158 applied:
AC

159 (1) The cogeneration system is steady-state and one-dimensional.

160 (2) Heat losses and pressure losses in pipelines are ignored.

161 (3) Pressure losses in heat exchangers are not considered during the analysis of

162 the cogeneration system performance.

163 (4) At the outlets of the condenser I and II, the basic ammonia-water solution and
 ACCEPTED MANUSCRIPT

164 refrigerant R134a could condensate to saturated liquid state by controlling the mass

165 flow rate of the abundant cooling water at design and off-design conditions.

166 (5) At the outlets of the separator, the ammonia-rich vapor is saturated vapor and

167 the ammonia-poor solution is saturated liquid at design and off-design conditions

PT
168 because of the full separation of the basic ammonia-water solution in the separator.

RI
169 For the ejector refrigeration cycle, the following assumptions are applied:

170 (1) The heat transfer between the wall of the ejector and the refrigerant R134a is

SC
171 neglected.

U
172 (2) The velocities of primary and secondary flows at the inlets of the ejector and
AN
173 the velocity of the mixed flow at the exit of the ejector are ignored.

174 (3) At the inlet of the ejector effective mixing region, the primary and secondary
M

175 flows start the mixing process with a uniform pressure.

D

176 As shown in Fig.3, plate heat exchangers are adopted as the boiler, the condenser
TE

177 I of Kalina cycle, the vapor generator, the condenser II and the evaporator of the

178 ejector refrigeration cycle because of the compact structure and the high heat transfer
EP

179 performance [29].

C

180
AC

181
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
182
M
D
TE
C EP
AC

183

184 Fig.1 Schematic of a power-cooling cogeneration system based on a Kalina cycle and

185 an ejector refrigeration cycle.

186
 ACCEPTED MANUSCRIPT
Suction chamber Constant area region Diffuser
y mi sh ce de
t ne
Primary flow

t ne
y mi sh ce de

Secondary flow

Shock
Pressure

PT
Mixed flow
Secondary flow

t-t ne-ne y-y mi-mi sh-sh ce-ce de-de

RI
187 Location in ejector

188 Fig.2 Ejector mechanism.

U SC
AN
P_l
M

θ
D

c_d P_t
TE

P_w

189
EP

190 Fig.3 Plate heat exchanger.

191 3. Cogeneration system modeling

C
AC

192 In order to predict the off-design performance of the power-cooling cogeneration

193 system, an off-design model, including the plate heat exchanger, the separator, the

194 axial inflow turbine, the pump, the ejector and the throttle valve, is developed on

195 Matlab software.

196 3.1. Plate heat exchanger

197 For the plate heat exchanger, Eqs.(1) and (2) show the heat transfer rate and the
 ACCEPTED MANUSCRIPT

198 heat transfer coefficient, respectively. Eq.(3) is applied to calculate the pressure drop.

199 Q = UA∆Tlmtd (1)

1 1 P_t 1
200 = + + (2)
U J hot λ J cold

2 fG 2 P _ l

PT
201 ∆P = (3)
ρ Dh

RI
202 In respect to the proposed power-cooling cogeneration system driven by hot

203 water, Kalina cycle uses ammonia-water as working fluid and the ejector refrigeration

SC
204 cycle uses refrigerant R134a as working fluid. Water is selected as the refrigeration

U
205 medium as shown in Fig.1. The correlations of heat transfer and friction factor in plate
AN
206 heat exchangers are listed below:

207 3.1.1 Single-phase region of ammonia-water in the boiler

M

208 When the ammonia-water is at liquid state, the Nusselt number [30] and the
D

209 friction factor [31] are expressed in Eqs.(4) and (5), respectively.
TE

 6θ 
0.646

210 Nu = 0.724   Re0.583 Pr1/3 (4)

π 
EP

14.62 Re
-0.514
(Re > 180)
211 f = (5)
 2.21Re
-0.097
(Re < 180)
C

212 3.1.2 Boiling region of ammonia-water in the boiler

AC

213 For the ammonia-water, the boiling heat transfer coefficient is obtained based on

214 the liquid heat transfer coefficient, as shown in Eqs.(6)-(7) [32].

215 If uSV < −111.88uSL + 11.848

216 J = 5 Bo 0.15 J l (6)

217 If uSV > −111.88uSL + 11.848

 ACCEPTED MANUSCRIPT
5 Bo 0.15 J l
218 J = max  2 0.2 (7)
(φchi ) J l

Gx
219 uSV = (8)
ρv

G (1 − x)
220 uSL = (9)
ρl

PT
221 The Chishom’s constant φchi in Eq.(7) is calculated by Eq.(10).

RI
0.5
1  ρl 
222 φchi = (1 − x ) + x  (10)
1− x  ρv 

SC
223 Eq.(11) is applied to calculate the heat transfer coefficient of single-phase liquid

U
224 Jl . AN
0.14
 λ  0.78 1/3  µ 
225 J l = 0.2092   Re Pr   (11)
 Dh   µ wall 
M

226 Hsieh [33] recommended Eq.(12) to calculate the friction factor during the

227 boiling process of ammonia-water.

D

228 f = 23, 799.8423Re-1.25 (12)

TE

eq

229 3.1.3 Condensation region of ammonia-water in the condenser I and vapor

EP

230 generator

231 Regarding to the condensation process of ammonia-water, the Nusselt number is

C
AC

232 obtained from Eq.(13), in which the hydraulic Reynolds number Reh is expressed in

233 Eq. (14) [34, 35].

234 Nu = 3.77 Re0.43

h Prl1/3 (13)

  ρl  
0.14

G (1 − x ) + x    Dh
  ρ v  
235 Re h =  (14)
µl

236 The pressure drop of ammonia-water during the condensation process is

 ACCEPTED MANUSCRIPT

237 calculated by Eq.(15), in which the multiplier of two-phase flow Φ is given by

238 Eq.(16) [35] and the friction factor of single-phase liquid f is obtained from Eq.(5) .

Φ 2 f [G (1- x)] P _ l
2

239 ∆P = (15)
2 ρl Dh

Φ 2 = 0.5 X LM
-2
(16)

PT
240

0.5 0.1
 1- x   ρ v   µl 
0.9

241 X LM =      (17)
 x   ρl   µv 

RI
242 3.1.4 Single-phase region of water in the boiler, the evaporator, the condenser I

SC
243 and II

U
244 For the liquid water, the Nusselt number is expressed in Eq.(18) and the friction
AN
245 factor is given by Eq.(19) [36].

246 Nu = 0.1368 Re0.7424 Pr 0.35 (18)

M

247 f = 0.572 Re-0.217 (19)

D

248 3.1.5 Single-phase region of refrigerant in the vapor generator, the evaporator
TE

249 and the condenser II

250 In respect to the refrigerant in single-phase region, Eq.(20) is proposed to

EP

251 calculate the Nusselt number [30] and Eq.(21) is proposed to obtain the friction factor
C

252 [37].
AC

 6θ 
0.646

253 Nu = 0.724   Re0.583 Pr1/3 (20)

π 
32
254 f = (21)
Re
255 3.1.6 Two-phase region of refrigerant in the vapor generator and the evaporator

256 For the two-phase refrigerant in the vapor generator and the evaporator, the heat

257 transfer correlation is obtained based on the heat transfer coefficient of refrigerant in
 ACCEPTED MANUSCRIPT

258 single-phase liquid region, as shown in Eqs.(22)-(23). The friction factor of

259 refrigerant in two-phase boiling region is given by Eq.(24) [33].

260 J = J l ( 88 Bo 0.5 ) (22)

0.14
 λ  0.78 1/3  µ 
261 J l = 0.2092   Re Pr   (23)
 µ wall 

PT
 Dh 
−1.25
262 f = 61,000 Re eq (24)

RI
263 3.1.7 Two-phase region of refrigerant in the condenser II

SC
264 Regarding to the two-phase refrigerant in the condenser II, the Nusselt number is

265 calculated by Eq.(25) and the friction factor is given by Eq.(26) [38].

U
Nu = 4.118Re0.4 1/3
266 eq Prl (25)
AN
−1.14
267 f = 21,500 Reeq Bo −0.085 (26)
M

268 In Eqs.(12), (25) and (26), the equivalent Reynolds number Reeq is calculated by

269 Eq.(27). The Reynolds number in Eqs.(4), (5), (18), (19), (20), (21) and (23) is
D

expressed in Eq.(28). The hydraulic diameter of plate heat exchanger is obtained from
TE

270

271 Eq.(29).
EP

  ρl  
0.5

G (1 − x ) + x    Dh
  ρ v  
272 Re eq =  (27)
µl
C

GDh
AC

273 Re = (28)
µ

4c _ dP _ w
274 Dh = (29)
2(c _ d + P _ w)

275 3.2. Separator

276 In the separator, it is assumed that the basic ammonia-water solution is fully

277 separated [2]. Thus, at the outlets of the separator, the ammonia-rich vapor is saturated
 ACCEPTED MANUSCRIPT

278 vapor and the ammonia-poor solution is saturated liquid at design and off-design

279 conditions. The flow in the separator satisfies the mass conversation laws, as shown in

280 Eqs.(30) and (31).

281 m2 = m3 + m4 (30)

PT
282 m2 xbasic = m4 x poor + m3 xrich (31)

RI
283 3.3. Axial turbine

284 The proposed cogeneration system adopts the axial turbine to produce power.

SC
285 The turbine power is expressed in Eq.(32).

Wtur = m3 ∆his ,turηtur

U
286 AN (32)

287 When the flow is sub-critical, the specific volume of ammonia-rich vapor in

288 turbine changes slightly by the sliding pressure operation approach. Thus, the
M

289 off-design mass flow rates of ammonia-rich vapor in the turbine could be obtained
D

290 from Eq.(33) [39].

TE

m3 P32 − P52 T3,des

291 = (33)
m3,des P3,des − P5,des T3
2 2
EP

292 According to the recommendation from Ray [40]，axial turbine stage efficiency

293 is a function of the ratio of the blade tip velocity to the theoretical vapor velocity. The
C
AC

294 blade tip velocity is directly proportional to the rotational speed of turbine. The

295 theoretical vapor velocity is proportional to the square root of the vapor isentropic

296 enthalpy difference. Thus, axial turbine stage efficiency could be given by Eq.(34) at

297 off-design conditions [40]. In Eq.(34), ∆his,des and ∆his are the isentropic specific

298 enthalpy differences at design and off-design conditions, respectively. Furthermore,

299 this equation was validated by Modi [41] in a Kalina cycle.

 ACCEPTED MANUSCRIPT
2
 N ∆his ,des 
300 ηtur = ηtur ,des − 2  − 1 (34)
 N des ∆his 

301 3.4. Pump

302 For the pump I and II, the pump powers are expressed in Eqs.(35)-(36) and the

pump efficiencies are given by Eqs.(37)-(38).

PT
303

m9 ∆his , pum I
304 W pum I = (35)
η pum I

RI
m10 ∆his , pum II
305 Wpum II = (36)

SC
η pum II

∆his, pum I
306 η pum I = (37)

U
∆hpum I
AN
∆his, pum II
307 η pum II = (38)
∆hpum II
M

308 In respect to the pump, the off-design mass flow rates and pressure heads are

309 obtained by adjusting the rotational speed of pump at the sliding pressure operation
D

approach. In this paper, the off-design performance of pump is obtained from the
TE

310

311 affinity laws, as shown in Eqs.(39)-(40), and the performance curves (Fig.4) of a
EP

312 pump at the rated rotational speed [42].

q N
=
C

313 (39)
qdes N des
AC

2
H  N 
314 =  (40)
H des  N des 
 ACCEPTED MANUSCRIPT

1.4 1.4

1.2 1.2

1.0 1.0

0.8 0.8

Η/Ηdes
η/ηdes

0.6 0.6

PT
0.4 0.4
η/ηdes
Η/Ηdes
0.2 0.2

RI
0.0 0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

SC
q/qdes
315

316 Fig.4 Performance curves of a pump at the rated rotational speed.

317 3.5. Ejector

U
AN
318 As shown in Fig.2, when the primary flow expands through the ejector nozzle, its
M

319 velocity is supersonic at the ejector nozzle outlet. In this way, a vacuum region is

320 produced in the ejector suction chamber and the secondary flow could be entrained
D

into the ejector. After increasing velocity in the annular convergent channel, the
TE

321

322 secondary flow is mixed with the primary flow in the mixing region with a uniform
EP

323 pressure [4]. A supersonic and transient mixed flow results in a shock in the constant

324 section region of ejector and a suddenly pressure rise. Then the mixed flow becomes
C

subsonic and its pressure is increased in the diffuser. The ejector entrainment ratio ξ
AC

325

326 is the ratio of the secondary flow mass flow rate to the primary flow mass flow rate,

327 as shown in Eq.(41).

m14
328 ξ= (41)
m11

329 Fig.5 shows the ejector performances at different back pressures and secondary
 ACCEPTED MANUSCRIPT

330 flow pressures. In Fig.5 (a), as the back pressure is lower than the critical back

331 pressure, the ejector entrainment ratio ξ is a constant (named critical mode).

332 However, as the back pressure is higher than the critical back pressure, the ejector

333 entrainment ratio ξ declines with a large gradient and this condition is named

PT
334 sub-critical mode. With the increase of the secondary flow pressure, the gradients of

RI
335 the coefficient of performance (COP) of the ejector refrigeration cycle are different at

336 overall modes (critical/sub-critical modes), as shown in Fig.5 (b).

SC
337 In this paper, the ejector model at overall modes is adopted on the basis of the

U
338 constant pressure mixing theory [5-9]. The detailed description of the ejector model at
AN
339 overall modes is presented in our previous work [43].
M
Entrainment ratio

Sub-critical
Critical mode mode Back flow mode
D
TE
C EP
AC

Pcr Back pressure

340
Back pressure
341 (a)
 ACCEPTED MANUSCRIPT

Back flow Sub-critical

mode Critical mode
mode

COP

PT
RI
SC
Secondary flow pressure
342
Back pressure
343 (b)

344
U
Fig.5 Ejector performance. (a) Entrainment ratio vs. back pressure and (b) COP vs.
AN
345 secondary flow pressure.
M

346 3.6. Throttle valve

347 In this paper, the flow across the throttle valve is considered as an isenthalpic
D

process, as shown in Eqs.(42)-(43).

TE

348

349 h7 = h6 (42)
EP

350 h13 = h14 (43)

351 3.7. Performance criteria

C
AC

352 For the power-cooling cogeneration system, the thermal efficiency and exergy

353 efficiency are selected to evaluate the system performance, as shown in Eqs.(44)-(45).

354 The heat absorbed by the boiler of Kalina cycle is defined as the energy input to the

355 cogeneration system, as shown in Eq.(46). The exergy difference of heat source (hot

356 water) in the boiler is defined as the exergy input to the cogeneration system. The

357 total energy output and exergy output of the cogeneration system are obtained from
 ACCEPTED MANUSCRIPT

358 Eqs.(47) and (48), respectively. The net power output, cooling output and cooling

359 exergy output are expressed in Eqs.(49)-(51), respectively.

360 ηth = Wout / Qboi (44)

361 ηex = Eout / ( Ehw,in − Ehw,out ) (45)

PT
362 Qboi = m2 ( h2 − h1 ) (46)

Wout = Wnet , power + Qcooling

RI
363 (47)

364 Eout = Wnet , power + Ecooling (48)

SC
365 Wnet , power = Wtur − Wpum I −Wpum II (49)

Qcooling = m14 ( h15 − h14 )

U
366
AN (50)

367 Ecooling = m14Tenv ( s15 − s14 ) − Qcooling (51)

368 3.8. Validation

M

369 The studied power-cooling cogeneration system combines a Kalina cycle and an
D

370 ejector refrigeration cycle. Thus, this part validates the proposed Kalina cycle model
TE

371 and the ejector refrigeration cycle model, respectively. For the Kalina cycle, the data

372 from the literature [44] is compared with the predicted results from the proposed
EP

373 Kalina cycle model, as shown in Table 1. It is found that the predicted values are
C

374 consistent with the literature [44].

AC

375 Regarding to the refrigeration cycle, the experimental data of an ejector

376 refrigeration cycle using R11 as refrigerant [45] is compared with the theoretical

377 prediction from the proposed ejector refrigeration cycle model at overall modes. The

378 comparison results are presented in Fig.6. The results show that the predicted results

379 are in line with the experimental data [45]. The maximal error of the coefficient of
 ACCEPTED MANUSCRIPT

380 performance (COP) is 13.4% and most of the error is below 5.0% at overall modes.

381 Table 1 Comparison results of a Kalina cycle driven by hot water

State T (°C) P (MPa) x (%) m (kg/s)

points This Ref. This Ref. This Ref. This Ref.

PT
paper [44] paper [44] paper [44] paper [44]

RI
1 25.6 25.6 2.70 2.70 0.80 0.80 2.966 2.97

2 91.0 91.0 2.70 2.70 0.80 0.80 2.966 2.97

SC
3 91.0 91.0 2.70 2.70 0.99 0.99 1.514 1.52

U
4 91.0 91.0 2.70 2.70 0.60 0.60 1.452 1.45
AN
5 30.54 30.7 0.79 0.79 0.99 0.99 1.514 1.52

6 91.0 91.0 2.70 2.70 0.60 0.60 1.452 1.45

M

7 49.78 49.9 0.79 0.79 0.60 0.60 1.452 1.45

D

8 46.71 46.8 0.79 0.79 0.80 0.80 2.966 2.97

TE

9 25.0 25.0 0.79 0.79 0.80 0.80 2.966 2.97

382
C EP
AC
 ACCEPTED MANUSCRIPT

0.20 Experimental data

Predicted result

Back flow Subcritcial Critcial mode

0.15 mode mode
COP

0.10

PT
Back pressure is fixed at 170 kPa
0.05
Primary flow pressure is fixed at 1020 kPa

RI
0.00
30 35 40 45 50 55 60 65

SC
Secondary flow pressure (kPa)
383

384 Fig.6 Comparison results of an ejector refrigeration cycle using R11 as refrigerant.

385 4. Simulation conditions

U
AN
386 The hot water (120 °C and 34.23 kg/s) from an oil field in China is the heat
M

387 source of the proposed power-cooling cogeneration system. In this paper, the basic

388 ammonia-water solution with the ammonia mass fraction of 80% is selected in Kalina
D

cycle [2]. In respect to the ejector refrigeration cycle, R134a shows a high coefficient
TE

389

390 of performance, so that it is selected as the refrigerant [46, 47]. The studied
EP

391 refrigeration temperature is above 0 °C, thus water is selected as the refrigerating

392 medium. The inlet temperature of refrigerating medium equals to the environment
C
AC

393 temperature. The inlet temperature of cooling water is the environment temperature.

394 Besides, the critical back pressure of the ejector is set as 120% of the condensing

395 pressure of the ejector refrigeration cycle at the design point. The properties of

396 ammonia-water and refrigerant R134a are obtained from REFPROP 9.1 [48]. Table 2

397 shows the operating parameters of the power-cooling cogeneration system at the

398 design phase.

 ACCEPTED MANUSCRIPT

399 This paper investigates the off-design performances of the cogeneration system

400 at different mass flow rate ratios of hot water mhw/mhw,des (60-120%), inlet

401 temperatures of hot water (114-126 °C), saturated condensation temperatures

402 (22-34 °C) and saturated evaporator temperatures (1-9 °C). The mature sliding

PT
403 pressure operation with a fixed turbine inlet temperature is adopted in the

RI
404 cogeneration system at off-design conditions. Specifically speaking, when the turbine

405 inlet temperature is assumed as a constant of 105 °C [2], the boiler pressure of Kalina

SC
406 cycle and the vapor generator pressure of ejector refrigeration cycle are changed by

U
407 adjusting the rotational speeds of the pump I and II at off-design conditions. Fig.7
AN
408 shows the flow chart of the simulation of the power-cooling cogeneration system at

409 the off-design phase. In Fig.7, when the heat transfer area error is less than 10-3, as
M

410 shown in Eqs.(52)-(53), it is assumed that the nominal heat transfer area equals to the
D

411 actual heat transfer area.

TE

A *boi − Aboi
412 | | ≤ 10−3 (52)
Aboi
EP

A *vap − Avap
413 | | ≤ 10−3 (53)
Avap
C

414 Table 2 Operating parameters of the power-cooling cogeneration system at the design
AC

415 phase.

Item Value

Ammonia mass fraction of basic solution (%) 80

Environment temperature (°C) 20

Environment pressure (kPa) 101.325

 ACCEPTED MANUSCRIPT

Hot water inlet temperature (°C) 120

Hot water mass flow rate (kg/s) 34.23

Boiler pressure (kPa) 2000-3000

Vapor generator pressure (kPa) 2000-3000

PT
Turbine inlet temperature (°C) 105

RI
Saturated evaporator temperature (°C) 5

Saturated condensation temperature (°C) 28

SC
Turbine efficiency (%) 80

U
Pump efficiency (%) 70
AN
Pinch point temperature difference (°C) 5

Critical back pressure of ejector (kPa) 872.26

M

416
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC

417

418 (a)
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
C EP
AC

419

420 (b)

421 Fig.7 Flow chart of the simulation of the power-cooling cogeneration system at the

422 off-design phase. (a) Power-cooling cogeneration system simulation and (b) ejector

423 refrigeration cycle simulation.

424
 ACCEPTED MANUSCRIPT

425 5. Results and discussion

426 5.1. Design optimization

427 Fig.8 shows the design process of the proposed power-cooling cogeneration

428 system integrating a Kalina cycle with an ejector refrigeration cycle. Optimal

PT
429 thermodynamic parameters, geometric parameters of plate heat exchangers and an

RI
430 ejector are obtained in this part.

U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
C EP
AC

431

432 Fig.8 Design process of the power-cooling cogeneration system integrating a Kalina

433 cycle with an ejector refrigeration cycle.

 ACCEPTED MANUSCRIPT

434 5.1.1. Optimal thermodynamic parameters and geometric parameters of ejector

435 The design code of the ejector is developed, as presented in Fig.8. For the ejector

436 design code, the inlet pressure of secondary flow, the critical back pressure, the inlet

437 pressure and the mass flow rate of primary flow are given parameters. The control

PT
438 variable is the ratio of the constant section region area to the nozzle throat area Ace /At.

RI
439 In this paper, the condensation pressure Pcon II of the condenser II is the back pressure

440 of the ejector. In Table 2, the design critical back pressure Pcr of the ejector is obtained

SC
441 by controlling the variable Ace /At.

U
442 Fig.9 shows the trends of the power-cooling cogeneration system performance
AN
443 with the boiler pressure and the vapor generator pressure at the design phase. In Fig.9,

444 it is found that an optimal boiler pressure and vapor generator pressure exist when the
M

445 total energy output reaches a maximum. In this part, the objective function is the total
D

446 energy output. The particle swarm optimization (PSO) algorithm is employed to
TE

447 optimize the boiler pressure, the vapor generator pressure, and obtain the maximal

448 total energy output [49]. The PSO algorithm includes a swarm of particles, which
EP

→
449 move in a space of possible solutions. Every particle has a position vector x
C

→
450 meaning a candidate solution and a velocity vector v . Furthermore, every particle
AC

→ →
451 could store a personal best position p (t ) and a global best position g (t ) by

452 communicating with its vicinal particles. The particles could move to better areas in
→
453 the search space at each time step t. The novel particle position x (t + 1) is shown in

454 Eq.(54). The new particle velocity is calculated by Eq.(55), in which U ( y, z ) means a

455 uniformly distributed value between y and z. In Eq.(55), φ1 means the significance of
 ACCEPTED MANUSCRIPT
→ →
456 p (t ) and φ2 means the significance of g (t ) . The inertia weight ϖ could control
→
457 the value of the old velocity v (t ) . The maximal velocity vmax could control the value

458 of the velocity vi at any time step.

→ → →
459 x (t + 1) = x(t ) + v (t + 1) (54)

PT
→ → → → → →
460 v (t + 1) = ϖ v (t ) + U (0, φ1 )( p (t ) − x (t )) + U (0, φ2 )( g (t ) − x (t )) (55)

RI
461 In this paper, the swarm size (the number of particles in the swarm) of 25, the

462 inertia weight ϖ of 0.6 and the maximum velocity vmax of 15% are applied [50]. The

SC
463 parameter for attraction to the personal best φ1 is set as 1.8 and the parameter for

attraction to the global best φ2 is set as 1.8 [50].

U
464
AN
465 The optimization results show that when the boiler pressure is 2645 kPa and the

466 vapor generator pressure is 2720 kPa, the maximal total energy output is 691.02 kW
M

467 (including 619.74 kW net power and 71.28 kW cooling). The performance of the
D

468 cogeneration system at the design point is shown in Table 3. In respect to the ejector,
TE

469 the nozzle throat diameter Dt is 16.54 mm, the nozzle exit diameter Dne is 20.25 mm

470 and the constant section region diameter Dce is 33.39 mm.
C EP
AC

471

472 (a)
 ACCEPTED MANUSCRIPT

PT
RI
473

474 (b)

U SC
AN
M
D

475
TE

476 (c)
C EP
AC

477

478 (d)

479 Fig.9 Power-cooling cogeneration system performance at the design phase vs. boiler
 ACCEPTED MANUSCRIPT

480 pressure and vapor generator pressure. (a) Cooling output, (b) net power output, (c)

481 total energy output and (d) Ace /At .

482 Table 3 Power-cooling cogeneration system performance at the design point.

Item Value

PT
Boiler pressure (kPa) 2645

RI
Vapor generator pressure (kPa) 2720

Ammonia mass fraction of ammonia-rich vapor (%) 97.67

SC
Ammonia mass fraction of ammonia-poor solution (%) 50.47

U
Ammonia-rich vapor mass flow rate (kg/s) 4.91
AN
Ammonia-poor solution mass flow rate (kg/s) 2.94

R134a mass flow rate of the ejector primary flow (kg/s) 2.38
M

Ejector entrainment ratio 0.18

D

Turbine power (kW) 654.10

TE

Pump I power (kW) 28.71

Pump II power (kW) 5.65

EP

Net power output (kW) 619.74

C

Cooling output (kW) 71.28

AC

Cooling exergy (kW) 4.52

Thermal efficiency (%) 8.68

Exergy efficiency (%) 39.87

483

484
 ACCEPTED MANUSCRIPT

485 5.1.2. Geometric parameters of plate heat exchangers

486 In this paper, the off-design performances of plate heat exchangers are predicted

487 based on the geometric parameters of plate heat exchangers. In respect to the passage

488 of the plate heat exchanger, the single-pass flow and the counter-current flow are

PT
489 adopted [42]. The thermodynamic parameters of heat exchangers are obtained from

RI
490 the optimal thermodynamic parameters of the cogeneration system. The pressure

491 drops of plate heat exchangers are restricted to 1% at the design phase [51]. When the

SC
492 channel distance and the plate width of the plate heat exchanger are known, the

U
493 pressure drop could be changed by adjusting the plate length and the number of plates.
AN
494 In this part, five plate heat exchangers in the studied cogeneration system are designed

495 on the basis of the design pressure drop of 1%. Table 4 presents the geometric
M

496 parameters of five plate heat exchangers in the studied power-cooling cogeneration
D

497 system.
TE

498 Table 4 Geometric parameters of five plate heat exchangers

Item Boiler Condenser I Vapor generator Condenser II Evaporator

EP

P_w (m) 0.3 0.3 0.3 0.3 0.3

C

c_d (mm) 1.5 2 1.5 2 7

AC

P_t (mm) 0.6 0.6 0.6 0.6 0.6

θ (°) 60 60 60 60 60

P_l (m) 0.57 0.67 0.66 1.45 0.93

A (m2) 50.20 204.83 4.17 19.61 3.07

499
 ACCEPTED MANUSCRIPT

500 5.2. Sliding pressure operation approach to variable mass flow rate ratio of hot

501 water

502 As the power-cooling cogeneration system is at the rated inlet temperature of hot

503 water (120 °C), saturated condensation temperature (28 °C) and saturated evaporator

PT
504 temperature (5 °C), the off-design performance of the cogeneration system at different

RI
505 mass flow rate ratios of hot water is investigated.

506 5.2.1. Boiler performance

SC
507 Fig.10 shows the trend of the boiler performance with the mass flow rate ratio of

U
508 hot water. It is found that when the mass flow rate ratio of hot water is increased from
AN
509 60% to 120%, the boiler pressure of Kalina cycle goes up from 2019.5 kPa to 2880.5

510 kPa. The reason for this is that the cogeneration system adopts the sliding pressure
M

511 operation approach with a fixed turbine inlet temperature. Furthermore, the increase
D

512 of the boiler pressure results in the increase of the mass flow rate (increasing from
TE

513 5.19 kg/s to 8.97 kg/s) of basic ammonia-water solution and the increase of the heat

514 transfer rate (increasing from 5.81 MW to 8.81 MW) in the boiler. Besides, according
EP

515 to the plate heat exchanger model, the hot water outlet temperature increases as the
C

516 mass flow rate ratio of hot water is increased, resulting in the decrease of the hot
AC

517 water temperature difference in the boiler, as shown in Fig.10.

 ACCEPTED MANUSCRIPT

130 3000 10
9
120
2800 9
110 Thw,in=120 °C (fixed) 8
Tcon=28 °C (fixed) 2600
100 8

mbasic (kg/s)
Teva=5 °C (fixed)

Qboi (MW)
Pboi (kPa)
7
Thw (°C)

90
Hot water inlet temperature 2400
Hot water outlet temperature
7

PT
80 Boiler pressure 6
Heat transfer rate in boiler
Basic ammonia-water solution 2200
70 mass flow rate
5 6

RI
60
2000
50 4 5
60 70 80 90 100 110 120

SC
mhw/mhw,des (%)
518

519 Fig.10 Boiler performance vs. mass flow rate ratio of hot water.

520
U
5.2.2. Turbine performance and ammonia-water pump I performance
AN
521 Fig.11 presents the turbine performance at different mass flow rate ratios of hot
M

522 water. With the increase of the mass flow rate ratio of hot water, the increasing boiler

523 pressure leads to the increasing ammonia mass fraction of ammonia-rich vapor (from
D

96.54% to 97.95%) and the increasing mass flow rate (from 3.57 kg/s to 5.39 kg/s) of
TE

524

525 ammonia-rich vapor flowing across the turbine. The reason for this is that the
EP

526 increasing heat is rejected by hot water in the boiler with the increase of the mass flow

527 rate ratio of hot water. The turbine efficiency declines obviously at a small mass flow
C
AC

528 rate ratio of hot water, while the turbine efficiency is slightly influenced by the mass

529 flow rate ratio of hot water at a large mass flow rate ratio of hot water (100-120%)

530 based on the turbine model. In particular, the turbine efficiency declines to 77.14% at

531 the mass flow rate ratio of hot water of 60%. In Fig.11, it can also be seen that due to

532 the increasing pressure ratio of turbine, the turbine power goes up from 367.20 kW to

533 757.73 kW as the mass flow rate ratio of hot water is increased from 60% to 120%.
 ACCEPTED MANUSCRIPT

534 Fig.12 shows the ammonia-water pump I performance with the mass flow rate

535 ratio of hot water. It is found that the pump I efficiency is above 68% under the

536 variation of the mass flow rate ratio of hot water (60-120%). The power consumed by

537 the pump I goes up from 12.64 kW to 37.54 kW as the mass flow rate ratio of hot

PT
538 water is increased from 60% to 120% because of the increasing mass flow rate of

RI
539 basic ammonia-water solution. However, compared with the turbine power, the power

540 consumed by the pump I is a smaller portion.

SC
82 800
98.0 6.5
Thw,in=120 °C (fixed)
750
81

U
97.8 Tcon=28 °C (fixed)
6.0 700
Teva=5 °C (fixed)
97.6 80
AN
5.5 650
97.4 79 600

Wtur (kW)
mrich (kg/s)
xrich (%)

5.0

ηtur (%)
97.2 550
M

78
4.5
97.0 500
4.0 77
96.8 450
D

Ammonia concentration of ammomia-rich vapor

96.6
Ammomia-rich vapor mass flow rate 3.5 76
Τurbine efficiency 400
Turbine power
TE

96.4 3.0 75 350

60 70 80 90 100 110 120
mhw/mhw,des (%)
541
EP

542 Fig.11 Turbine performance vs. mass flow rate ratio of hot water.
C
AC
 ACCEPTED MANUSCRIPT

40 71.0

Thw,in=120 °C (fixed) 70.5

35
Tcon=28 °C (fixed)
70.0
Teva=5 °C (fixed)
30
69.5
Wpum I (kW)

ηpum I (%)
25 69.0

PT
68.5
20
68.0
15 Ammonia-water pump I power

RI
Ammonia-water pump I efficiency 67.5

10 67.0
60 70 80 90 100 110 120

SC
mhw/mhw,des (%)
543

544 Fig.12 Pump I performance vs. mass flow rate ratio of hot water.

545 5.2.3. Vapor generator performance

U
AN
546 Fig.13 shows the variation of the vapor generator performance at different mass
M

547 flow rate ratios of hot water. For the ejector refrigeration cycle, the ammonia-poor

548 solution from the separator of Kalina cycle is the heat source. In the vapor generator,
D

the inlet temperature of ammonia-poor solution keeps constant at different mass flow
TE

549

550 rate ratios of hot water because it equals to the inlet temperature of the turbine based
EP

551 on the separator model. As the mass flow rate ratio of hot water is increased from

552 60% to 120%, the increasing boiler pressure leads to the increasing ammonia mass
C
AC

553 fraction (from 43.32% to 52.99%) of ammonia-poor solution and the increasing mass

554 flow rate (from 1.61 kg/s to 3.58 kg/s) of ammonia-poor solution flowing across the

555 vapor generator. An increase of the mass flow rate of the ammonia-poor solution

556 results in an increase of the vapor generator pressure (from 2141.1 kPa to 2877.1 kPa)

557 by the sliding pressure operation approach. Meanwhile, the mass flow rate of

558 refrigerant R134a in the vapor generator goes up from 1.85 kg/s to 2.53 kg/s as the
 ACCEPTED MANUSCRIPT

559 mass flow rate ratio of hot water is increased from 60% to 120%. The reason for this

560 is that the increasing heat is released by the ammonia-poor solution in the vapor

561 generator with the increase of the mass flow rate ratio of hot water.

4.0
0.54
2800

PT
100 Thw,in=120 °C (fixed)
3.5
Tcon=28 °C (fixed)
0.51
Teva=5 °C (fixed)
90 3.0 2400

RI
0.48
80 2.5

Pvap (kPa)
Tpoor (°C)

xpoor (%)
m (kg/s)
2000 0.45
70

SC
2.0
0.42
60 1.5 1600
Ammonia-poor solution inlet temperature
Ammonia-poor solution outlet temperature
Ammonia-poor solution mass flow rate 0.39

U
50 Ammonia concentration of ammonia-poor solution 1.0
Vapor generator pressure 1200
R134a mass flow rate in vapor generator 0.36
AN
40 0.5
60 70 80 90 100 110 120
mhw/mhw,des (%)
562
M

563 Fig.13 Vapor generator performance vs. mass flow rate ratio of hot water.

564 5.2.4. Ejector performance

D

Fig.14 shows the ejector performance at different mass flow rate ratios of hot
TE

565

566 water. For the ejector, the condensation pressure of R134a is the back pressure of the
EP

567 ejector, so that the ejector back pressure is a constant of 726.9 kPa under the variation

568 of the mass flow rate ratio of hot water. The primary flow pressure of the ejector is the
C
AC

569 vapor generator pressure. As the mass flow rate ratio of hot water is increased from

570 60% to 120%, the ejector critical back pressure goes up from 741.93 kPa to 905.40

571 kPa because of the increase of the ejector primary flow pressure. In respect to the

572 ejector, it can also be seen that the critical back pressure is larger than the back

573 pressure at different mass flow rate ratios of hot water. Thus, the ejector is at the

574 critical mode at studied conditions. In the constant section region of the ejector, as the
 ACCEPTED MANUSCRIPT

575 mass flow rate ratio of hot water is increased, the mass flow rate of the primary flow

576 goes up, resulting in the decrease of the mass flow rate of the secondary flow. When

577 the mass flow rate ratio of hot water is increased from 60% to 120%, the entraining

578 ratio ξ of the ejector decreases from 0.31 to 0.15.

PT
950
Thw,in=120 °C (fixed)
0.30
Tcon=28 °C (fixed)

RI
900
Teva=5 °C (fixed)
0.27
PR134a (kPa)

SC
850 0.24
Ejector entrainment ratio

ξ
Ejector back pressure
Ejector critical back pressure
800 0.21

U
0.18
750
AN
0.15

700
60 70 80 90 100 110 120
M

mhw/mhw,des (%)
579

580 Fig.14 Ejector performance vs. mass flow rate ratio of hot water.
D

5.2.5. Cycle performance

TE

581

582 Fig.15 shows the cogeneration system performance at different mass flow rate
EP

583 ratios of hot water. In Fig.15(a), it can be seen that the increasing mass flow rate ratio

584 of hot water leads to the increasing net power output (from 351.43 kW to 713.71 kW)
C
AC

585 because of the increasing turbine pressure ratio. However, as the mass flow rate ratio

586 of hot water is increased from 60% to 120%, the cooling output decreases from 95.01

587 kW to 64.66 kW and the cooling exergy output decreases from 6.03 kW to 4.11 kW

588 because of the decreasing entrainment ratio ξ of the ejector.

589 Compared with the net power output, the cooling output is a smaller portion in

590 the total energy output and the cooling exergy output is also a smaller portion in the
 ACCEPTED MANUSCRIPT

591 total exergy output, as shown in Fig.15(a). Thus, as the mass flow rate ratio of hot

592 water is increased from 60% to 120%, the total energy output goes up, leading to the

593 increase of the thermal efficiency (from 7.69% to 8.83%), as shown in Fig.15(b).

594 Besides, due to the combined effects of the total exergy output and the exergy input

PT
595 on exergy efficiency, an increase of the mass flow rate ratio of hot water (from 60% to

RI
596 120%) results in an increase of the exergy efficiency (from 33.74% to 40.41%).

750 6.5

SC
95
700
90 6.0
650

U
85
Wnet,power (kW)

600 5.5

Ecooling (kW)
Qcooling (kW)
Thw,in=120 °C (fixed) Net power output
80
AN
550 Cooling output
Tcon=28 °C (fixed)
5.0
Cooling exergy output
500 Teva=5 °C (fixed) 75
4.5
450
70
M

400 4.0
65
350
D

60 3.5
60 70 80 90 100 110 120
TE

mhw/mhw,des (%)
597

598 (a)
EP

9.0 41

8.8 40
Thw,in=120 °C (fixed)
C

Tcon=28 °C (fixed) 39
8.6 Teva=5 °C (fixed)
AC

38
8.4
ηex (%)
ηth (%)

37
8.2
36
8.0
35
Thermal efficiency
7.8 Exergy efficiency
34

7.6 33
60 70 80 90 100 110 120
mhw/mhw,des (%)
599
 ACCEPTED MANUSCRIPT

600 (b)

601 Fig.15 Cogeneration system performance vs. mass flow rate ratio of hot water. (a) Net

602 power, cooling, cooling exergy and (b) cycle efficiency.

603 5.3. Sliding pressure operation approach to variable inlet temperature of hot

PT
604 water

RI
605 This part investigates the off-design performance of the cogeneration system at

606 different inlet temperatures of hot water, as shown in Fig.16. In particular, the system

SC
607 is fixed at the design mass flow rate of hot water (34.23 kg/s), saturated condensation

U
608 temperature (28 °C) and saturated evaporator temperature (5 °C). The investigated
AN
609 inlet temperatures of hot water are from 114 °C to 126 °C.

610 In Fig.16 (a), it can be seen that the turbine efficiency is above 79.5% at studied
M

611 inlet temperatures of hot water (114-126 °C). Thus, the turbine efficiency is slightly
D

612 influenced by the inlet temperature of hot water. It is also found that an increase of the
TE

613 inlet temperature of hot water means an increase of the vapor generator pressure, an

614 increase of the boiler pressure and an increase of the net power output, as shown in
EP

615 Fig.16(a)-(b). The reason for this is that an increase of the inlet temperature of hot
C

616 water means an increase of the heat released by the hot water, leading to an increase
AC

617 of the log mean temperature difference in the boiler or the vapor generator. For the

618 ejector, the vapor generator pressure is the primary flow pressure. Thus, as the inlet

619 temperature of hot water is increased, the ejector entrainment ratio, cooling output and

620 cooling exergy output decrease because of the increase of the primary flow pressure

621 and the corresponding decrease of the mass flow rate of the secondary flow in the
 ACCEPTED MANUSCRIPT

622 constant section region of the ejector. Specifically speaking, an increase of per unit

623 inlet temperature of hot water means a 2.10% (on the basis of the design net power

624 output) increase of the net power output and a 2.44% (on the basis of the design

625 cooling output) decrease of the cooling output.

PT
626 In Fig.16(c), as the inlet temperature of hot water is increased from 114 °C to

RI
627 126 °C, the thermal efficiency goes up from 8.31% to 8.93% due to the dominant

628 effect of the increasing net power output on thermal efficiency. It means that an

SC
629 increase of per unit inlet temperature of hot water leads to a 0.44% (on the basis of the

U
630 design thermal efficiency) increase of the thermal efficiency. However, the exergy
AN
631 efficiency is slightly affected by the inlet temperature of hot water.

3100 82
0.27
M

mhw=34.23 kg/s (fixed)

3000 Tcon=28 °C (fixed) 81
Teva=5 °C (fixed) 0.24
2900
D

80
2800 0.21
79
TE
P (kPa)

ηtur (%)

2700
ξ

0.18
78
2600
EP

0.15 77
2500
Boiler pressure
Vapor generator pressure 76
2400 Ejector entrainment ratio 0.12
Turbine efficiency
C

2300 75
114 116 118 120 122 124 126
AC

Thw,in (°C)
632

633 (a)
 ACCEPTED MANUSCRIPT

750
85
mhw =34.23 kg/s (fixed) 5.5
700 Tcon=28 °C (fixed)
80
Teva=5 °C (fixed)
5.0
75
Wnet,power (kW)

Qcooling (kW)

Ecooling (kW)
650

70 4.5
600

PT
65
4.0
550 60
Net power output

RI
Cooling output
Cooling exergy output
3.5
55
500
114 116 118 120 122 124 126

SC
Thw,in (°C)
634

635 (b)

8.9

U 40.2
AN
mhw =34.23 kg/s (fixed)
8.8 Tcon=28 °C (fixed)
40.0
Teva=5 °C (fixed)
M

8.7 39.8
ηex (%)
ηth (%)

8.6 39.6
TE

8.5 39.4
Thermal efficiency
Exergy efficiency
EP

8.4 39.2
114 116 118 120 122 124 126
Thw,in (°C)
636
C

637 (c)
AC

638 Fig.16 Cogeneration system performance vs. inlet temperature of hot water. (a) Boiler

639 pressure, vapor generator pressure, turbine efficiency, entrainment ratio of ejector, (b)

640 net power, cooling, cooling exergy and (c) cycle efficiency.

641 5.4. Sliding pressure operation approach to variable saturated condensation

642 temperature
 ACCEPTED MANUSCRIPT

643 When the saturated evaporator temperature is fixed at the design value (5 °C)

644 and the hot water is fixed at the design parameters (120 °C and 34.23 kg/s), the

645 off-design performance of the power-cooling cogeneration system is predicted at

646 different saturated condensation temperatures (22-34 °C), as shown in Fig.17.

PT
647 In Fig.17(a), as the saturated condensation temperature is increased from 22 °C

RI
648 to 34 °C, the boiler pressure goes up slightly from 2609.9 kPa to 2690.9 kPa and the

649 vapor generator pressure goes up slightly from 2698.3 kPa to 2741.3 kPa. It can also

SC
650 be seen that compared with the inlet temperature of hot water, the turbine efficiency is

U
651 more strongly affected by the saturated condensation temperature. The turbine
AN
652 efficiency declines to 78.8% at the saturated condensation temperature of 34 °C.

653 Besides, due to a slight increment of the vapor generator pressure, the entrainment
M

654 ratio of the ejector decreases slightly from 0.183 to 0.176 as the saturated
D

655 condensation temperature goes up from 22 °C to 34 °C.

TE

656 The increase of the saturated condensation temperature means the increase of the

657 condensation pressure, leading to the decrease of the pressure ratio of the turbine.
EP

658 Thus, in Fig.17(b), an increase of per unit saturated condensation temperature means a
C

659 2.50% (on the basis of the design net power output) decrement of the net power
AC

660 output. In addition, an increase of per unit saturated condensation temperature just

661 results in a 1.06% (on the basis of the design cooling output) decrement of the cooling

662 output because of a slight decrease of the entrainment ratio of the ejector. However,

663 the cooling exergy goes up with the increase of the saturated condensation

664 temperature because of the significant effect of the increasing environment

 ACCEPTED MANUSCRIPT

665 temperature on the cooling exergy.

666 In Fig.17(c), it is found that an increase of per unit saturated condensation

667 temperature means a 1.82 % (on the basis of the design thermal efficiency) decrease

668 of the thermal efficiency and a 0.71 % (on the basis of the design exergy efficiency)

PT
669 decrease of the exergy efficiency. The reason for this is that the decreasing net power

RI
670 output has a dominant effect on the thermal efficiency and exergy efficiency as the

671 saturated condensation temperature is increased.

SC
0.188 82
2740
81

U
2720 0.184
80
AN
2700

0.180 79
2680

ηtur (%)
P (kPa)

Boiler pressure

ξ
Vapor generator pressure
M

Ejector entrainment ratio 78

2660
Turbine efficiency
0.176
2640 77
Thw,in=120 °C (fixed)
D

2620 mhw =34.23 kg/s (fixed) 76

0.172
Teva=5 °C (fixed)
TE

2600 75
22 24 26 28 30 32 34
Tcon (°C)
672
EP

673 (a)
C
AC
 ACCEPTED MANUSCRIPT

750 6.5
Thw,in=120 °C (fixed)
76 6.0
mhw =34.23 kg/s (fixed)
700
Teva=5 °C (fixed) 5.5
Wnet,power (kW) 74

650 5.0

Qcooling (kW)

Ecooling (kW)
72
4.5
600 70

PT
4.0

68 3.5
550 Net power output

RI
Cooling output
Cooling exergy output
3.0
66
500 2.5
22 24 26 28 30 32 34

SC
Tcon (°C)
674

675 (b)

U
AN
9.6 Thw,in=120 °C (fixed)
mhw =34.23 kg/s (fixed) 41
Teva=5 °C (fixed)
9.2
M

40
8.8
ηex (%)
ηth (%)

39
8.4
TE

8.0 38
Thermal efficiency
Exergy efficiency

7.6
EP

37
22 24 26 28 30 32 34
Tcon (°C)
676
C

677 (c)
AC

678 Fig.17 Cogeneration system performance vs. saturated condensation temperature. (a)

679 Boiler pressure, vapor generator pressure, turbine efficiency, entrainment ratio of the

680 ejector, (b) net power, cooling, cooling exergy and (c) cycle efficiency.

681 5.5. Sliding pressure operation approach to variable saturated evaporator

682 temperature
 ACCEPTED MANUSCRIPT

683 Fig.18 presents the off-design performance of the cogeneration system under the

684 variation of the saturated evaporator temperature (1-9 °C). The system is fixed at the

685 design inlet temperature of hot water (120 °C), mass flow rate of hot water (34.23

686 kg/s) and saturated condensation temperature (28 °C).

PT
687 In Fig.18(a)-(b), the saturated evaporator temperature has no influence on Kalina

RI
688 cycle, so that the boiler pressure, the vapor generator pressure, the turbine efficiency

689 and the net power output keep constant at different saturated evaporator temperatures.

SC
690 An increase of the saturated evaporator temperature means an increase of the

U
691 secondary flow pressure of the ejector. It can be seen that as the saturated evaporator
AN
692 temperature is increased from 1 °C to 9 °C, the entrainment ratio of the ejector goes

693 up from 0.124 to 0.241. The reason for this is that the ejector primary flow pressure
M

694 (the vapor generator pressure) keeps constant and the ejector secondary flow pressure
D

695 goes up with the increase of the saturated evaporator temperature. Furthermore, due to
TE

696 the increasing entrainment ratio of the ejector, when the saturated evaporator

697 temperature is increased from 1 °C to 9 °C, the cooling output goes up from 48.72 kW
EP

698 to 97.25 kW. However, as the saturated evaporator temperature is increased from 1 °C
C

699 to 9 °C, the cooling exergy output goes up to 4.57 kW at the saturated evaporator
AC

700 temperature of 6 °C before decreasing.

701 In Fig.18(c), owing to the increasing entrainment ratio of the ejector and the

702 increasing cooling output, the thermal efficiency goes up from 8.38% to 8.99% with

703 the increase of the saturated evaporator temperature. As the saturated evaporator

704 temperature is increased from 1 °C to 9 °C, the exergy efficiency goes up to 39.82%
 ACCEPTED MANUSCRIPT

705 at the saturated evaporator temperature of 6 °C before decreasing based on the trend

706 of the cooling exergy with the saturated evaporator temperature.

2800 82
Thw,in=120 °C (fixed)
0.27
mhw =34.23 kg/s (fixed) 81
2700 Tcon=28 °C (fixed) 0.24

PT
80
0.21
2600
79
P (kPa)

ηtur (%)
RI
0.18

ξ
78
2500

0.15 77

SC
Boiler pressure
2400 Vapor generator pressure
Ejector entrainment ratio 76
Turbine efficiency 0.12

2300 75

U
1 2 3 4 5 6 7 8 9
Teva (°C)
AN
707

708 (a)
M

625 110 5.0

100
4.8
D

620
90
Thw,in=120 °C (fixed)
TE

4.6
Ecooling (kW)
Wnet,power (kW)

Qcooling (kW)

mhw =34.23 kg/s (fixed)

615 80
Tcon=28 °C (fixed)

70 4.4
EP

610
60 4.2
Net power output
605 50
C

Cooling output
Cooling exergy output 4.0
40
AC

1 2 3 4 5 6 7 8 9
Teva (°C)
709

710 (b)
 ACCEPTED MANUSCRIPT

39.83
9.0
Thw,in=120 °C (fixed)
8.9 mhw =34.23 kg/s (fixed) 39.82
Tcon=28 °C (fixed)
8.8
39.81

ηex (%)
8.7
ηth (%)

8.6 39.80

PT
8.5
Thermal efficiency 39.79

RI
8.4 Exergy efficiency

8.3 39.78
1 2 3 4 5 6 7 8 9

SC
Teva (°C)
711

712 (c)

713
U
Fig.18 Cogeneration system performance vs. saturated evaporator temperature. (a)
AN
714 Boiler pressure, vapor generator pressure, turbine efficiency, entrainment ratio of
M

715 ejector, (b) net power, cooling, cooling exergy and (c) cycle efficiency.

716 6. Conclusion
D

For low-grade heat source (hot water), the power-cooling cogeneration system,
TE

717

718 integrating a Kalina cycle with an ejector refrigeration cycle, is a promising

EP

719 technology. In the proposed novel cogeneration system, the ejector refrigeration cycle

720 using R134a is driven by the ammonia-poor solution from the separator of Kalina
C
AC

721 cycle. The off-design model of the cogeneration system is established based on five

722 plate heat exchangers, a separator, an axial inflow turbine, two pumps, an ejector and

723 two throttle valves. The optimal boiler pressure of Kalina cycle and vapor generator

724 pressure of ejector refrigeration cycle are obtained by adopting the particle swarm

725 optimization (PSO) algorithm at design conditions. Variable hot water parameters,

726 saturated condensation temperature and saturated evaporator temperature are

 ACCEPTED MANUSCRIPT

727 investigated by the sliding pressure operation approach. Main conclusions are listed as

728 below:

729 (1) At the design phase, when the total energy output of the cogeneration system

730 reaches the maximum of 691.02 kW (including 619.74 kW net power and 71.28 kW

PT
731 cooling), the optimal boiler pressure of Kalina cycle is 2645 kPa and vapor generator

RI
732 pressure of ejector refrigeration cycle is 2720 kPa. The design thermal efficiency and

733 exergy efficiency are 8.68% and 39.87%, respectively. For the ejector, the nozzle

SC
734 throat diameter Dt is 16.54 mm and the constant section region diameter Dce is 33.39

U
735 mm. The design areas for heat transfer of the boiler, the condenser I of Kalina cycle,
AN
736 the vapor generator, the condenser II and the evaporator of ejector refrigeration cycle

737 are 50.20 m2, 204.83 m2, 4.17 m2, 19.61 m2 and 3.07 m2, respectively.
M

738 (2) As the mass flow rate ratio or the inlet temperature of hot water goes up, the
D

739 net power, thermal efficiency and exergy efficiency go up. However, the ejector
TE

740 entrainment ratio, cooling output and cooling exergy output decrease with the increase

741 of the mass flow rate ratio or the inlet temperature of hot water.
EP

742 (3) When the saturated evaporator temperature is increased from 1 °C to 9 °C,
C

743 the thermal efficiency goes up from 8.38% to 8.99%, while the exergy efficiency goes
AC

744 up to 39.82% at the saturated evaporator temperature of 6 °C before decreasing.

745 (4) Compared with the saturated condensation temperature, the ejector

746 entrainment ratio and cooling output are more strongly influenced by the inlet

747 temperature of hot water. An increase of per unit inlet temperature of hot water means

748 a 2.44% (on the basis of the design cooling output) decrease of the cooling output.
 ACCEPTED MANUSCRIPT

749 However, the turbine efficiency, net power output, thermal efficiency and exergy

750 efficiency are more strongly influenced by the saturated condensation temperature

751 than the inlet temperature of hot water. An increase of per unit saturated condensation

752 temperature means a 2.50% (on the basis of the design net power output) decrement

PT
753 of the net power output and a 1.82% (on the basis of the design thermal efficiency)

RI
754 decrease of the thermal efficiency.

755 Acknowledgements

SC
756 The authors gratefully acknowledge the financial support by the National Key

U
757 Research and Development Program of China (Grant No. 2017YFB0603504 and No.
AN
758 2016YFB0600104).

759
M

760
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

761 References
762 [1] Rasool B, Ali B. Thermodynamic optimization and thermoeconomic analysis of

763 four double pressure Kalina cycles driven from Kalina cycle system 11. Energy

764 Convers Manage 2017; 152:110-23.

765 [2] Wang J, Wang J, Dai Y. Assessment of off-design performance of a Kalina cycle

PT
766 driven by low-grade heat source. Energy 2017; 138:459-72.

RI
767 [3] Chunnanond K, Aphornratana S. Ejectors: applications in refrigeration

768 technology. Renew Sustain Energy Re 2004; 8: 129-55.

SC
769 [4] Keenan JH, Neumann EP, Lustwerk F. An investigation of ejector design by

770 analysis and experiment. J Appl Mech Trans ASME 1950;72: 299-309.

U
Huang BJ, Chang JM, Wang CP. A 1-D analysis of ejector performance. Int J
AN
771 [5]

772 Refrig 1999; 22: 354-64.

M

773 [6] Zhu Y, Cai W, Wen C. Shock circle model for ejector performance evaluation.

774 Energy Convers Manage 2007;48: 2533-41.

D

775 [7] Valle JG, Jabardo JMS, Ruiz FC. A one dimensional model for the
TE

776 determination of an ejector entrainment ratio. Int J Refrig 2012; 35: 772-84.

777 [8] Chen WX, Liu M, Chong DT. A 1D model to predict ejector performance at
EP

778 critical and sub-critical operational regimes. Int J Refrig 2013; 36: 1750-61.
C

779 [9] Chen WX, Shi C, Zhang S. Theoretical analysis of ejector refrigeration system
AC

780 performance at overall modes. Applied Energy 2017; 185: 2074-84.

781 [10] Goswami DY, Xu F. Analysis of a new thermodynamic cycle for combined

782 power and cooling using low and mid temperature solar collectors. J Solar

783 Energy Eng 1999;121: 91-7.

784 [11] Kim KH, Kim G, Han C. Performance assessment of ammonia-water based
 ACCEPTED MANUSCRIPT

785 power and refrigeration cogeneration cycle. Int J Mater Mechan Manufactu

786 2013;1: 36-40.

787 [12] Padilla RV, Demirkaya G, Goswami DY. Analysis of power and cooling

788 cogeneration using ammonia-water mixture. Energy 2010;35: 4649-57.

PT
789 [13] Fontalvo A, Pinzon H, Duarte J, et al. Exergy analysis of a combined power and

RI
790 cooling cycle[J]. Appl Therm Eng 2013; 60: 164-71.

791 [14] Pouraghaie M, Atashkari K, Besarati SM. Thermodynamic performance

SC
792 optimization of a combined power/cooling cycle. Energy Convers Manage

U
793 2010;51: 204-11.
AN
794 [15] Zare V, Mahmoudi SMS, Yari M, Amidpour M. Thermoeconomic analysis

795 andoptimization of an ammonia–water power/cooling cogeneration cycle.

M

796 Energy 2012; 47: 271-83.

D

797 [16] Srinivas T, Reddy BV. Thermal optimization of a solar thermal cooling
TE

798 cogeneration plant at low temperature heat recovery. J Energy Res Technol

799 2014;136: 021204.

EP

800 [17] Shankar R, Srinivas T. Coupled cycle with Kalina cycle system and vapor
C

801 absorption refrigeration. Proc Inst Mechan Eng part A-J Power Energy 2014;
AC

802 228: 953-64.

803 [18] Shankar R, Srinivas T. Investigation on operating processes for a new solar

804 cooling cogeneration plant. J Solar Energy Eng 2014; 136: 031016.

805 [19] Zheng D, Chen B, Qi Y. Thermodynamic analysis of a novel absorption

806 power/cooling combined-cycle. Appl Energy 2006; 83:311-23.

 ACCEPTED MANUSCRIPT

807 [20] Yu Z, Han J, Liu H. Theoretical study on a novel ammonia-water cogeneration

808 system with adjustable cooling to power ratios. Appl Energy 2014;122: 53-61.

809 [21] Hua J, Chen Y, Wang Y. Thermodynamic analysis of ammonia-water

810 power/chilling cogeneration cycle with low-grade waste heat. Appl Therm Eng

PT
811 2014; 64: 483-90.

RI
812 [22] Jing X, Zheng D. Effect of cycle coupling-configuration on energy cascade

813 utilization for a new power and cooling cogeneration cycle. Energy Convers

SC
814 Manage 2014; 78: 58-64.

U
815 [23] Cao L, Wang J, Wang H. Thermodynamic analysis of a Kalina-based combined
AN
816 cooling and power cycle driven by low-grade heat source. Appl Therm Eng

817 2017; 111: 8-19.

M

818 [24] Kumar GP, Saravanan R, Coronas A. Experimental studies on combined cooling
D

819 and power system driven by low-grade heat sources. Energy 2017; 126: 801-12.
TE

820 [25] Wang J, Wang J, Zhao P. Thermodynamic analysis of a new combined cooling

821 and power system using ammonia-water mixture. Energy Convers Manage
EP

822 2016;117: 335-42.

C

823 [26] Seckin C. Thermodynamic analysis of a combined power/refrigeration cycle:

AC

824 Combination of Kalina cycle and ejector refrigeration cycle. Energy Convers

825 Manage 2018; 157:631-43.

826 [27] Ghaebi H, Parikhani T, Rostamzadeh H. Thermodynamic and thermoeconomic

827 analysis and optimization of a novel combined cooling and power (CCP) cycle

828 by integrating of ejector refrigeration and Kalina cycles. Energy 2017;139:

 ACCEPTED MANUSCRIPT

829 262-76.

830 [28] Barkhordarian O, Behbahaninia A, Bahrampoury R. A novel ammonia-water

831 combined power and refrigeration cycle with two different cooling temperature

832 levels. Energy 2017; 120: 816-26.

PT
833 [29] Wang J, Yan Z, Wang M, Li M, Dai Y. Multi-objective optimization of an

834 organic Rankine cycle (ORC) for low grade waste heat recovery using

RI
835 evolutionary algorithm. Energy Convers Manage 2013; 71: 146-58.

SC
836 [30] García-Cascales JR, Vera-García F, Corberán-Salvador JM. Assessment of

837 boiling and condensation heat transfer correlations in the modelling of plate heat

U
838 exchangers. Int J Refrig 2007; 30: 1029-41.
AN
839 [31] Jiang J, He G, Liu Y. Flow boiling heat transfer characteristics and pressure drop

840 of ammonia-lithium nitrate solution in a smooth horizontal tube. Int J Heat Mass
M

841 Transfer2017; 108:220-31.

D

842 [32] Taboas F, Valles M, Assessment of boiling heat transfer and pressure drop
TE

843 correlations of ammonia/water mixture in a plate heat exchanger, Int J

844 Refrig.-Revue Int Du Froid 2012; 35; 633-44.

EP

845 [33] Hsieh YY, Lin TF. Saturated flow boiling heat transfer and pressure drop of

846 refrigerant R-410A in a vertical plate heat exchanger. Int J Heat Mass Transf
C

847 2002; 45: 1033-44.

AC

848 [34] Lostec BL, Galanis N, Millette J. Simulation of an ammonia–water absorption

849 chiller. Renew Energy 2013; 60:269-83.

850 [35] Würfel R, Ostrowski N. Experimental investigations of heat transfer and

851 pressure drop during the condensation process within plate heat exchangers of

852 the herringbone-type. Int J Therm Sci 2004; 43:59-68.

853 [36] Khan TS, Khan MS, Chyu MC. Experimental investigation of single phase
 ACCEPTED MANUSCRIPT
854 convective heat transfer coefficient in a corrugated plate heat exchanger for

855 multiple plate configurations. Appl Therm Eng 2010; 30:1058-65.

856 [37] Hewitt GF, Barbosa J. Heat exchanger design handbook. Connecticut: Begell

857 House; 2008.

858 [38] Kuo WS, Lie YM, Hsieh Y . Condensation heat transfer and pressure drop of

PT
859 refrigerant R-410A flow in a vertical plate heat exchanger. Int J Heat Mass

RI
860 Transfer 2005; 48: 5205-20.

861 [39] Cooke DH. Modeling of off-design multistage turbine pressures by Stodola’s

SC
862 ellipse. In: Energy incorporated PEPSE user’s group meeting. Richmond,

863 Virginia, USA: Bechtel Power Corporation; 1983. p. 205-34.

864
U
[40] Ray A. Dynamic modelling of power plant turbines for controller design. Appl
AN
865 Math Model 1980; 4:109-12
M

866 [41] Modi A, Andreasen JG, Kaern MR. Part-load performance of a high temperature

867 Kalina cycle. Energy Convers Manage 2015; 105:453-61.

D

868 [42] Hu D, Li S, Zheng Y. Preliminary design and off-design performance analysis of

TE

869 an organic Rankine cycle for geothermal sources. Energy Convers Manage

870 2015;96: 175-87.

EP

871 [43] Du Y, Han PF, Qiang XC. Off-design performance analysis of a combined
C

872 cooling and power system driven by low-grade heat source. Energy Convers
AC

873 Manage 2018; 159:327-41.

874 [44] Li S, Dai Y. Thermo-economic comparison of Kalina and CO2 transcritical

875 power cycle for low temperature geothermal sources in China. Appl Therm Eng

876 2014, 70:139-52.

877 [45] Aphornratana S, Chungpaibulpatana S, Srikhirin P. Experimental investigation

878 of an ejector refrigerator: effect of mixing chamber geometry on system

 ACCEPTED MANUSCRIPT
879 performance. Int J Energy Res 2001; 25: 397-411.

880 [46] Selvaraju A, Mani A. Analysis of a vapour ejector refrigeration system with

881 environment friendly refrigerants. Int J Therm Sci 2004, 43: 915-21.

882 [47] Varga S, Lebre PS, Oliveira AC. Readdressing working fluid selection with a

883 view to designing a variable geometry ejector. Int J Low-carbon Tech 2013, 10:

PT
884 205-15.

RI
885 [48] 23 NSRD. NIST Thermodynamic and Transport Properties of Refrigerants and

886 Refrigerant Mixtures REFPROP. Version 601. 1998.

SC
887 [49] Kennedy J, Eberhart RC. Particle swarm optimization. IEEE Int Conference on

888 Neural Networks 1995; 4: 1942-48.

889
U
[50] Vesterstrom J, Thomsen R. A comparative study of differential evolution,
AN
890 particle swarm optimization, and evolutionary algorithms on numerical
M

891 benchmark problems. Congress on. IEEE 2013; 2: 1980-87.

892 [51] Li H, Hu D, Wang M. Off-design performance analysis of Kalina cycle for low
D

893 temperature geothermal source. Appl Therm Eng 2016, 107:728-37.

TE
C EP
AC
 ACCEPTED MANUSCRIPT

Highlights

A novel power-cooling cogeneration system is proposed based on a Kalina cycle.

A novel method for predicting the off-design performance of system is proposed.

Particle swarm optimization algorithm is adopted to optimize parameters.

PT

Variable four thermodynamic parameters are considered at the off-design phase.

RI

Thermal efficiency and exergy efficiency of the cogeneration system are

analyzed.

U SC
AN
M
D
TE
C EP
AC