Performance evaluation of a tube-in-tube CO2 gas cooler used in a heat pump water heater

(1)

Performance evaluation of a tube-in-tube CO

2

gas cooler used in a heat

pump water heater

Pei-Yu Yu

a,b

_{, Wei-Keng Lin}

b

_{, Chi-Chuan Wang}

c,⇑ a

Green Energy and Environment Research Laboratories, Industrial Technology Research Institute, Taiwan

b

Department of Engineering and System Science, National Tsing Hua University, Taiwan

c

Department of Mechanical Engineering, National Chiao Tung University, Taiwan

a r t i c l e

i n f o

Article history:

Received 13 November 2013

Received in revised form 9 January 2014 Accepted 13 January 2014

Available online 24 January 2014 Keywords:

Carbon dioxide Supercritical Gas cooler

Tube-in-tube heat exchanger

a b s t r a c t

In this study, investigation of the performance of a tube-in-tube counter-flow water-cooled CO2gas cooler operating above and near critical pressure is presented using a heat pump water heater with CO2flowing in the annulus side. A tube-in-tube heat exchanger model applicable for supercritical fluid CO2and water was also developed and validated. The measured total heat transfer capacity ranged from 1.31 to 4.06 kW at various test conditions. The calculations show good agreement with the experimental results. The results demonstrate that the variation of CO2temperature tends to show very slow decreas-ing near the pseudo-critical region when compared to the inlet region. Yet this phenomenon becomes more pronounced as the inlet pressure is close to the critical pressure (73.8 bar). The calculation also reveals a peculiar phenomenon that the local heat transfer rate of the heat exchanger peaks within the heat exchanger near the pseudo critical region due to the drastic rise of specific heat (CPvalue).

1. Introduction

For the past several decades, revisit of refrigerant CO2 was

strongly taken into consideration as a candidate for replacing the synthetic refrigerants like HFCs. CO2features advantages such as

ecology benign, safety and inexpensive cost[1,2]. However, with its relatively low critical temperature (31.1 °C), CO2is often

oper-ated as a supercritical ﬂuid for a typical vapor compression cycle under normal operating conditions, thereby leading to a so-called transcritical cycle[3–5]. Mobile air conditioning, residential AC/ heat pump systems and military environmental control units (ECUs) are typical examples exploiting CO2 transcritical cycle

applications. Moreover, it was found that CO2 water heat pump

could outperform conventional refrigerants in terms of system COP[6]. The continuous and large temperature glide for CO2in a

transcritical process can contribute to improve the performance of heating tape water.

Gas cooler, in which CO2is cooled with persistent temperature

drop, is different from those constant temperature condensation processes, and it is one of most important device in CO2

transcrit-ical cycle since its ﬂow arrangements and behaviors can greatly affect the optimal operating pressure and system efﬁciency. Many literature had been presented on various gas cooler models for

space heating, cooling and water heating[3,7–10]. Most of studies used the gas cooler model to enhance the system performance of a refrigeration plant or heat pump[11–14]; however, only a few of researches devoted component-level behavior of gas coolers par-ticularly for water heating applications. Sanchez et al.[15] pre-sented a water–CO2 coaxial heat exchanger model with

refrigerant flowing through internal tube bundle using finite vol-ume technique and CO2convective coefficient correlations. Their

results revealed that thermal effectiveness increases with the rise of refrigerant pressure and water mass ﬂow rate, and decreases with the increase of evaporating pressure and water inlet temper-ature. Fronk and Garimella[4,5]carried out an analysis of a water-coupled gas cooler with a compact, multi-pass cross-counter ﬂow of aluminum brazed plate, microchannel CO2gas cooler, and the

model was validated with experimental data. REFPROP[16] pre-sented an analysis for a concentric counter-ﬂow heat exchanger by solving a set of complicated partial differential equations, including conservation of mass, momentum and energy amid CO2

and water and considering the wall conduction in both radial and axial direction. They found that the variation of the local heat ﬂux revealed a local maximum within the heat exchanger due to the tremendous change of speciﬁc heat of CO2.

In a liquid-cooled gas cooler, the contribution of CO2

thermo-physical changes is more important than that in an air-cooled one due to side thermal resistance is dominant in the air-cooled heat exchangers. But, investigations concerning the water-cooled gas cooler are comparatively fewer. Therefore, the objective

http://dx.doi.org/10.1016/j.expthermﬂusci.2014.01.007

⇑Corresponding author. Address: Department of Mechanical Engineering, National Chiao Tung University, EE474, 1001 University Road, Hsinchu 300, Taiwan. Tel.: +886 3 5712121x55105.

E-mail address:ccwang@mail.nctu.edu.tw(C.-C. Wang).

Contents lists available atScienceDirect

Experimental Thermal and Fluid Science

(2)

of this study is to develop a simple counter-ﬂow heat exchanger model capable to analyze the heat transfer behavior of CO2

tube-in-tube water-cooled gas cooler subject to refrigerant ﬂowing in annulus side both numerically and experimentally.

2. Numerical method

The model heat exchanger is a double-pipe heat exchanger with CO2 ﬂowing through the annulus whereas the water is ﬂowing

counter-currently in the tube side.Fig. 1is a schematic of the heat exchanger. Since considerable change of physical properties of CO2

may occur especially nearby pseudo-critical temperatures, the heat exchanger must be subdivided into many small segments. A prior sensitive analysis of the inﬂuence of segments was performed, and a total of 26 segments were used in this simulation. Higher number of control volumes produces smaller errors but increases the calculation time. A number of ﬁnite volumes between 20 and 30 optimize the calculation time and the error[15]. A schematic diagram showing the variation of temperature for CO2and water

is shown inFig. 1(b) where the subscript c denotes CO2and w

rep-resents water. In this regard, the heat balance amid water and cool-ant in each segment i can be written in the following equations:

Qi¼ mcCpc;iðTc;i Tc;iþ1Þ ¼ mwCpw;iðTw;i Tw;iþ1Þ: ð1Þ

Qi¼ ðUAÞi ðLMTDÞi: ð2Þ

The overall heat transfer coefﬁcient is obtained from

1 UA¼ 1 hcAo;iþ lndo di 2

p

kwallLþ 1 hwAi;i : ð3Þ

The physical properties for CO2are a function of local pressure

and temperature and the properties of water are related to local temperature. The relevant properties are obtained from REFPROP 8.0 (2007)[16]. The heat transfer coefﬁcient of CO2is based on

the Dang and Hihara correlation (2004)[7], i.e.

hc¼ Nuckc=dH; ð4Þ Nuc¼ fc 8 ðReb 1000ÞPr 1:07 þ 12:7 ffiffiffifc 8 q Pr2=3 1 ; ð5Þ where Pr ¼ Cpb

l

b=kb; for Cpb Cp; Cpb

l

b=kf; for Cpb<Cp and

l

b=kb

l

f=kf; Cpb

l

f=kf; for Cpb<Cp and

l

b=kb<

l

f=kf; 8 > > < > > : ð6Þ Nomenclature A surface area (m2₎

C heat capacity ﬂow rate (W K1₎

C average heat capacity ﬂow rate (W K1₎

Cp speciﬁc heat (J kg1_K1₎

d diameter (m)

f friction factor

h heat transfer coefﬁcient (W m2_K1₎

i speciﬁc enthalpy (kJ kg1₎

ID inner diameter (m) k conductivity (W m1_K1₎

L tube length (m)

LMTD log mean temperature difference (K) _

m mass ﬂow rate (kg s1₎

Nu Nusselt number (hd/k) OD outer diameter (m)

P pressure (MPa)

Pr Prandtl number

Q heat transfer rate (kW) R thermal resistance (°C W1₎

Re Reynolds number (

q

ud/

l

)

T temperature (°C)

u velocity (m s1₎

U overall heat transfer coefﬁcient (W m2_K1₎

Greek letters

DT temperature difference (K)

D maximum temperature difference of gas cooler, =T

c,i-Tw,i

l

viscosity (kg m1_s1₎

q

density (kg m3₎ Subscripts b bulk c carbon dioxide

c, i ith segment of carbon dioxide

f ﬁlm

H hydraulic diameter

i inner

i ith segment of heat exchanger max larger one

min smaller one

o outer

w water

wall wall

w, i ith segment of water side

Carbon dioxide flow direction Water flow

(a)

(b)

Tc,1 Tc,2 Tc,3 Tw,1 Tw,2 Tw,3 Tc,i Tc,i +1 Tw,i Tw,i +1 CO₂ Water

Fig. 1. (a) Schematic of the tube-in-tube heat exchanger and (b) deﬁnition of the temperature variation for CO2and water alongside the length of the tube-in-tube

(3)

Cp ¼hb hwall Tb Twall ; ð7Þ Reb¼ GdH

l

b ; ð8Þ fc¼ ½1:82 logðRebÞ 1:642; ð9Þ

where the subscript b represents the bulk temperature, and wall is evaluated at the wall temperature and f denotes calculation at the film temperature. The film temperature, Tf, is defined as Tf= (Tb+

-Twall)/2. On the other hand, the heat transfer coefﬁcient for the

water side, hw, is via Gnielinski (1976) correlation[17]:

hw¼ Nuwkw=d; ð10Þ Nuw¼ fw 8 ðRe 1000ÞPr 1:07 þ 12:7 ffiffiffiffifw 8 q Pr2=3 1 ; ð11Þ where fw¼ ½1:82 logðRewÞ 1:64 2 : ð12Þ

In this study, CO2is as a supercritical working ﬂuid which

distin-guishing feature is dramatically rapid variations of its physical properties as the temperature is closed to the pseudocritical point.

The deﬁnition of pseudocritical temperature Tpcis the temperature

at which the speciﬁc heat reaches a peak for a given pressure. The pseudocritical temperature Tpc of CO2 is a function of pressure

and can be best ﬁtted by the following algebraic equation based on the data from NIST Refrigerants Database REFPROP[18].

Tpc¼ 122:6 þ 6:124P 0:1657P2þ 0:01773P2:5 0:0005608P3;

ð13Þ

where the pseudocritical temperature Tpcis in °C and the pressure p

is in bar.

3. Experimental setup

3.1. Experimental apparatus and procedures

Fig. 2shows a schematic diagram of the closed test loop which

is modiﬁed with a water-to-water type of heat pump water heater. A variable speed compressor was used to circulate CO2refrigerant

and maintain the inlet pressure of the test section (gascooler) at as-signed high pressures. Water enters the gascooler to cool down CO2with the counter-ﬂow arrangement and it was controlled at

selected ﬂow rate by a pump and regulated valves. After passing through a Coriolis-type mass ﬂow meter, CO2 was cooled in the

internal heat exchanger and the superheat of CO2in the suction

side is increased before returning to the compressor. The electronic

Compressor Internal HX Evaporator Accumulator Gascooler Electronic Expansion Valve Cooling Water Inlet Water Inlet Test Section

:Volumetric flow meter

:Thermocouple :Pressure transducer :Mass flow meter

(4)

expansion valve was driven by a stepping motor and the stainless steel valve body sustained a maximum working pressure of 150 bar. The opening of the valve ranged is controlled by pulse sig-nal with 480 pulses representing fully opened. Like the gascooler, a double tube type water-coupled heat exchanger served as the evaporator. CO2is then evaporated to complete the full

refrigera-tion cycle. Chilled water ﬂowing through the evaporator is used as the heat source to absorb heat from CO2. To prevent liquid

refrigerant from entering the compressor, an accumulator having U-shape type was installed between the evaporator and the inter-nal heat exchanger.

Fig. 3depicts the conﬁguration of the test section (gascooler)

which is consisted of a 13 m-long spiral tube-in-tube counter-ﬂow heat exchanger. The CO2ﬂows through the outer annular passage

from the top while cooling water ﬂows inside the inner tube. The gascooler was made of smooth copper with inner diameter of 6.34 mm and a thickness of 0.8 mm in the inner tube. The inside diameter of the outer tube is 10 mm and thickness is 1.0 mm with

a subsequent annular gap of 1.03 mm between inner tube and out-er tube. The test section was covout-ered with a 12.7 mm thick rubbout-er insulation to minimize heat dissipation into the ambient environ-ment. The wall temperature of the outer tube was measured at 13 locations is equally distributed along the gascooler with using T-type thermocouples taped on the outside wall. The inlet and outlet temperatures of the cooling water were gauged with Pt 100 sen-sors to determine the total heat transfer capacity in the test section.

3.2. Uncertainty analysis

The uncertainty in this study is determined with ISO/IEC Guide to the expression of uncertainty in measurement[19]. An example of uncertainty analysis also was demonstrated to determine the uncertainty in an experimental statistics[20].

According to the deﬁnition of ISO GUM,

Y ¼ y U; ð14Þ

U ¼ k ucðyÞ; ð15Þ

where Y: measurand; y: estimate of measurand; U: expanded uncertainty; uc(y): combined standard uncertainty;

u2 cðyÞ ¼ Xn i¼1 @f @xi 2 u2_ðx iÞ; ð16Þ

where ucðyÞ is combined standard uncertainty, uðxiÞ is standard

uncertainty and xiare the error sources which affect the estimation

of measurand y, as shown below:

y ¼ f ðxiÞ: ð17Þ

Table 1is a list of the combined standard uncertainties in this

study and instrumentation calibration is the important sources of uncertainty for each parameter of measurement. A coverage factor of 2 multiplied by the values in theTable 2gives a 95% conﬁdence level that the actual uncertainty is less than or equal to the stated uncertainty[20].

3.3. Test conditions

Table 2shows the test conditions in the present work and all

experiment was executed in an ISO 17025 certiﬁed laboratory. Experiments were conducted for three different inlet pressures of carbon dioxide from 76 to 96 bar by tuning a variable speed com-pressor and the opening of an electrical expansion valve. The inlet temperatures of water were adjusted from 15° to 30° centigrade. The water volumetric ﬂow rates were controlled from 1.0 to T1 T5 T3 T7 T11 T9 T0 T2 T6 T4 T8 T10 T12

Fig. 3. Photograph and schematic of the gas cooler.

Table 1

Summary of the combined standard uncertainties for the measured parameters.

Parameter Major source of uncertainty Magnitude of uncertainty Measuring instrumentation

1. Water volume ﬂow rate Instrumentation calibration ±1% (Yokogawa AXF) 2. CO2mass ﬂow rate Instrumentation calibration ±0.35% (Micro Motion)

3. CO2temperature Instrumentation calibration ±1 °C (T-type thermocouple)

4. Water temperature Instrumentation calibration ±0.1 °C (Pt100 RTD) 5. CO2absolute pressure Instrumentation calibration ±0.3% (Yokogawa PT)

Table 2

Test conditions of the gas cooler (CO2side).

Parameter Value

Inlet pressure of CO2(bar) 96, 86, 76

Inlet temperature of water (°C) 15, 20, 25, 30 Volumetric ﬂow rate of water (L min1₎ _{1.0, 1.5, 2.0}

(5)

2.0 L min1 _{by a pump and regulated valves. A total of 36}

experimental test conditions were carried out to investigate the performance of this gas cooler.

4. Results and discussion

The prediction of the total heat transfer capacity of the gas cool-er is compared with the expcool-erimental data ﬁrst. Note that the detailed experimental conditions are tabulated in Table 3. The measured results of the total heat transfer capacity ranged from 1.31 to 4.06 kW subject to various inlet pressures and water ﬂow-rate. The simulation is conducted with the same inlet conditions for CO2and water, and the corresponding comparison of heating

capacity is depicted inFig. 4. Deviation between the calculation and the measured data ranged from 18.85% to 21.73% over the range of refrigerant and water inlet conditions, with the highest relative deviation occurring at the lowest water ﬂow rate and high-est water inlet temperature whose details can be seen inTable 3. As shown inFig. 4, approximate 94% of the 36 data points were predicted within the ±20% accuracy limits, indicating that the mathematical model predicts the total heat transfer capacity reasonably well. Furthermore, the difference between calculations and measurement becomes profound as the total heat transfer capacity is larger than 2.3 kW.

The total heating capacity vs. various cooling water inlet conditions at various CO2inlet pressures are shown inFig. 5. The

corresponding inlet pressures are 76, 86, and 96 bar respectively. The measured total heating capacity ranged from 1.31 to 4.06 kW depending on the test conditions. As seen inFig. 5, the total heating loads are increased when the cooling water flowrate is increased or when the water inlet temperature is reduced. How-ever, the increasing trend subject to water coolant flowrate is not the same pertaining to the rise of water flow rate. Note that the

Table 3

Measured data of the 36 test conditions.

Water inlet condition (Q) Tc,in(°C) Pc,in(bar) Tw,in(°C) m_c(kg s1) m_w(kg s1) Calculated (Q) (Watt) Measured Q (Watt) Deviation (%)

P = 96 bar 15 °C/1.0LPM 91.8 95.9 15.0 0.0135 0.0172 1985.97 2164 8.23 15 °C/1.5LPM 83.3 95.9 15.0 0.0137 0.0249 3109.01 2988 4.07 15 °C/2.0LPM 79.9 96.5 14.9 0.0136 0.0332 3468.41 3662 5.29 20 °C/1.0LPM 92.9 96.2 20.0 0.0135 0.0166 2078.07 1867 11.29 20 °C/1.5LPM 86.9 96.0 20.1 0.0134 0.0245 2644.67 2539 4.16 20 °C/2.0LPM 82.4 96.3 20.1 0.0134 0.0332 2978.29 3112 4.28 25 °C/1.0LPM 94.5 96.4 25.0 0.0136 0.0163 1826.53 1556 17.39 25 °C/1.5LPM 91.9 96.1 25.1 0.0130 0.0247 1978.07 2021 2.13 25 °C/2.0LPM 85.3 96.3 25.0 0.0132 0.0334 2436.62 2380 2.37 30 °C/1.0LPM 92.1 96.0 29.9 0.0139 0.0157 1448.03 1312 10.35 30 °C/1.5LPM 92.1 96.1 30.0 0.0130 0.0246 1284.48 1583 18.85 30 °C/2.0LPM 89.0 96.2 30.0 0.0128 0.0337 1381.36 1671 17.31 P = 86 bar 15 °C/1.0LPM 78.1 86.3 15.0 0.0146 0.0160 3204.83 2895 10.69 15 °C/1.5LPM 76.3 85.9 14.9 0.0138 0.0243 3698.58 3620 2.17 15 °C/2.0LPM 72.9 86.3 15.0 0.0138 0.0345 3635.75 3811 4.60 20 °C/1.0LPM 76.0 85.9 20.0 0.0146 0.0181 3332.00 2790 19.43 20 °C/1.5LPM 79.0 86.3 20.0 0.0134 0.0261 3463.57 3391 2.13 20 °C/2.0LPM 74.9 86.0 20.1 0.0136 0.0350 3448.66 3620 4.74 25 °C/1.0LPM 77.9 85.9 25.0 0.0149 0.0158 2716.02 2233 21.65 25 °C/1.5LPM 76.4 85.9 25.0 0.0143 0.0245 3161.21 2887 9.51 25 °C/2.0LPM 78.9 85.9 25.1 0.0131 0.0334 3211.73 3291 2.40 30 °C/1.0LPM 76.2 86.0 30.0 0.0149 0.0173 2446.94 2010 21.73 30 °C/1.5LPM 76.0 86.4 30.0 0.0145 0.0252 2690.47 2482 8.40 30 °C/2.0LPM 75.1 85.8 30.1 0.0142 0.0336 2948.85 2795 5.52 P = 76 bar 15 °C/1.0LPM 59.3 76.1 15.1 0.0132 0.0157 3787.11 3560 6.38 15 °C/1.5LPM 58.6 75.8 15.1 0.0129 0.0251 3734.60 3907 4.40 15 °C/2.0LPM 65.4 76.0 15.1 0.0130 0.0340 3628.80 4056 10.54 20 °C/1.0LPM 60.2 75.9 20.0 0.0131 0.0159 3503.54 3252 7.75 20 °C/1.5LPM 59.6 76.1 20.1 0.0129 0.0255 3547.41 3733 4.96 20 °C/2.0LPM 60.3 76.4 20.0 0.0125 0.0345 3445.20 3870 10.99 25 °C/1.0LPM 60.9 75.9 24.9 0.0130 0.0157 3190.78 2934 8.74 25 °C/1.5LPM 60.7 76.1 25.0 0.0130 0.0249 3352.63 3454 2.93 25 °C/2.0LPM 59.0 76.1 25.0 0.0129 0.0344 3287.14 3655 10.07 30 °C/1.0LPM 61.2 75.9 30.0 0.0129 0.0171 2867.34 2553 12.32 30 °C/1.5LPM 60.6 76.2 30.1 0.0129 0.0266 3174.02 3185 0.36 30 °C/2.0LPM 59.2 75.7 30.0 0.0128 0.0348 3075.26 3380 9.02 Y= X +20% - 20% 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 Measured Q ( Watt ) Calculated Q ( W att)

(6)

mass ﬂowrate of CO2stays roughly the same in all cases. InFig. 5(a)

where the CO2pressure is maintained at 76 bar, it can be found

that for a lower water inlet temperature of 15 °C the heating capac-ity rises steadily against the water flowrate. On the other hand, only marginal rise of heating capacity vs. water flowrate is seen when the water temperature is increased to 30 °C. The heating capacity is associated with effective overall heat transfer coeffi-cient U and the mean temperature difference between CO2 and

water. Initially, the thermal resistance in the water side is about 40–60% larger than that of CO2side. Therefore, normally one can

expect that an increase of the water flowrate gives rise to an effec-tive drop of total thermal resistance, and accordingly a steady rise of heating capacity. This is actually the case for a lower water inlet temperature of 15 °C. However, it appears that the rise of heating capacity subject to water flowrate is only marginal for a higher in-let water temperature of 30 °C despite the overall heat transfer coefficient still shows a noticeable rise with the water flowrate. The major explanation of this phenomenon can be made clear from

Table 3where an appreciable drop of effective temperature

differ-ence between CO2and water is encountered when the water inlet

temperature is raised to 30 °C. It is found that the effective temper-ature difference is reduced by approximately 40% when the water inlet temperature is raised to 30 °C. Moreover, the maximum heat transfer rate Qmax, represents the maximum heat transfer rate for

an inﬁnite large heat exchanger and is given as Q ¼ CminD (Cmin

represents the average capacity ﬂowrate (¼ _mCp) which can be

either in CO2side or water side depending on the corresponding

effective Cmin value andD= Tc,inTw,in). Normally, Cmin is at the

CO2side. With the enormous change of Cpof CO2alsonside the heat

exchanger, Cpmust be integrated for the whole ﬂow path. An

esti-mation of Qmaxfor the inlet water temperature of 15 °C and 30 °C

yields 4.2 and 1.9 kW respectively. The results suggest that the effective temperature difference plays an imperative role on the heating capacity subject to water inlet ﬂowrate rate and tempera-ture and the corresponding heating capacity is also limited by the Qmax.

Analogous trend for the heating capacity vs. water ﬂowrate is also encountered when the inlet pressure of CO2 is raised to 86

and 96 bar as shown inFig. 5(b) and (c). Note that both the present calculation and Wang and Hihara[21]all indicated that the overall heat transfer coefﬁcient decreases monotonically with the increase of CO2pressure[21]. The decline of overall heat transfer coefﬁcient

is mainly associated with sharp decline with thermal conductivity and heat capacity at the pseudo-critical region which results in an appreciable drop of heat transfer coefﬁcient of CO2. However, from

Table 3, one can see an effective increase in CO2inlet temperature

when the inlet pressure is increased, thereby the considerable rise of mean temperature difference amid water and CO2compensates

the slight decrease in UA value, and accordingly a slight rise of effective heating capacity emerges.

The temperature distribution of CO2and water along the ﬂow

path at three different CO2inlet pressure under 2 L min1 water

ﬂowrate and 15 °C water inlet temperature are shown asFig. 6. The temperature of the CO2drop continuously alongside the CO2

path, yet the temperature may pass through the pseudo-critical temperature for a given pressure and it reaches a lower tempera-ture that is close to the water inlet temperatempera-ture. This is the so-called transcritical operation. A close examination of the tem-perature variation of CO2indicates that the temperature variation

of CO2near the pseudo-critical region is contradictory to those at

the entrance region where the CO2temperature shows a

continu-ous decline. In fact, the temperature drop becomes much less than that in the entrance region. This phenomenon becomes even more pronounced as the inlet pressure is decreased toward the critical point (73.8 bar). The extraordinary phenomenon is associated with the sharp change of CPvalue in the neighborhood of pseudo-critical

temperature as shownFig. 7(a). In fact, it is evitable that the high temperature CO2must pass through the pseudo-critical

tempera-ture during a trans-critical process. In other words, the heat capacity ﬂow rate (C) of the CO2reveals a considerable rise due

to a tremendous rise of CP value around the pseudo-critical

temperature. From a simple energy balance formula, Q = CcDT, it

is not surprised that the variation of CO2temperature tends to be

small adjacent to the pseudo-critical region when compared to that at the inlet region. The phenomenon had been theoretically calculated by Yu et al.[22], yet it is validated from the present

Inlet Pressure of CO2 : P= 76 bar

3.11 2.38 1.67 3.66 2.99 2.16 2.54 1.87 2.02 1.56 1.31 1.58 1.0 2.0 3.0 4.0 5.0 0.5 1 1.5 2 2.5

Cooling Water Flow Rate ( L min-1)

T o ta l H ea t Tra n sfe r C ap ac it y ( kW ) 15°C Inlet Water 20°C Inlet Water 25°C Inlet Water 30°C Inlet Water

3.62 3.29 2.79 2.90 3.62 3.81 2.79 3.39 2.89 2.23 2.01 2.48 1.0 2.0 3.0 4.0 5.0 0.5 1 1.5 2 2.5 T o ta l H eat T ran sf er C apaci ty ( kW ) 15°C Inlet Water 20°C Inlet Water 25°C Inlet Water 30°C Inlet Water

3.56 4.06 3.91 3.25 3.87 3.73 2.93 3.66 3.45 2.55 3.19 3.38 1.0 2.0 3.0 4.0 5.0 T o ta l H eat T ran sf er C ap aci ty ( kW ) 15°C Inlet Water 20°C Inlet Water 25°C Inlet Water 30°C Inlet Water 0.5 1 1.5 2 2.5

(a)

(b)

(c)

Fig. 5. Measured total heat transfer capacity vs. cooling water inlet conditions for three different CO2inlet pressure. (a) P = 76 bar; (b) P = 86 bar and (c) P = 96 bar.

(7)

experimental observation. The present simulations of the CO2

temperature distribution, as shown inFig. 6, are in line with the measurements. The foregoing results imply that the heat transfer characteristics of CO2nearby the pseudo critical region resemble

normal refrigerants which show invariant temperature during condensation process. Hence, it would be beneﬁcial to lengthen the inﬂuence of pseudo-critical region as far as heat transfer augmentation is concerned. Furthermore, the local heat transfer temperature difference between CO2 and water decreases near

the pseudo-critical points[21]. When the pressure is close to the critical value of 73.8 bar, the pinch point obviously occurs at the region where CO2 pseudo-critical temperature takes place. The

results are quite different from the traditional refrigerants to water heat exchangers with constant ﬂuid property.

In addition, the drastic change CPvalue for CO2also affects the

distribution of the local heat transfer rate alongside the heat

exchanger.Fig. 7(b) depicts a schematic showing the local heat transfer rate vs. dimensionless distance counting from the inlet of CO2. It can be clearly seen in the ﬁgure, the heat transfer ﬁrst

de-creases to a local minimum, followed by a rise to a plateau next to the minimum location, and ﬁnally decreased again toward the out-let. This strange phenomenon becomes more and more apparent when the inlet pressure is further reduced. In fact, a signiﬁcant recovering of local heat transfer rate is encountered for P = 76 bar despite the maximum temperature difference still occurs at the CO2inlet.

Fig. 7(c) and (d) represents the variation of local heat transfer

coefﬁcient of CO2and water separately along the heat exchanger.

When CO2passes through the pseudo-critical point, the local heat

transfer coefﬁcient of CO2rises rapidly then descends gradually

to-ward the outlet whereas no detectable changes of heat transfer coefﬁcients are found in the water side. As a consequence, the

Inlet Pressure of CO2 : P = 76 bar, Tpc = 32.2 oC

65.4 54.7 16.2 17.7 18.7 24.0 24.2 30.5 31.3 31.4 32.7 36.8 37.3 0 20 40 60 80 0.00 0.05 0.14 0.23 0.32 0.41 0.50 0.59 0.68 0.77 0.86 0.95 1.00 CO2 Flow Dirction , x/L, L=13m T em p er at u re ( o C ) CO2-Measured CO2-Calculated Water-Calculated T-Calculated

Inlet Pressure of CO2 : P = 86 bar, Tpc = 37.9 oC

72.9 56.8 37.5 25.0 15.4 15.6 15.7 16.1 16.3 38.5 29.1 19.2 19.2 0 20 40 60 80 0.00 0.05 0.14 0.23 0.32 0.41 0.50 0.59 0.68 0.77 0.86 0.95 1.00 CO2 Flow Direction , x/L, L=13m CO2-Measured CO2-Calculated Water-Calculated T-Calculated

Inlet Pressure of CO2 : P = 96 bar, Tpc = 43.0

o C 59.2 16.1 _16.1 _15.5 _15.5 _15.4 39.4 38.2 25.8 22.2 18.1 16.8 79.9 0 20 40 60 80 0.00 0.05 0.14 0.23 0.32 0.41 0.50 0.59 0.68 0.77 0.86 0.95 1.00 CO2 Flow Direction , x/L, L=13 m CO2-Measured CO2-Calculated Water-Calculated T-Calculated T em p er at u re ( o C ) T em p er at u re ( o C )

(a)

(b)

(c)

Fig. 6. Temperature distribution along the CO2refrigerant ﬂow path at three various inlet pressure with a ﬁxed cooling water inlet condition (2 L min1, 15 °C). (a) P = 76 bar;

(8)

dominant thermal resistance of the heat exchanger could switch to CO2side from water for a trans-critical process. Accordingly, near

the pseudo-critical point, there is a signiﬁcant local decrease in refrigerant-side thermal resistance which yields a sharp increase in local heating transfer duty [5,22]. This is quite distinct as compared to those of constant property ﬂuids such as water to water heat exchangers. In summary, the drastic property changes of CO2 occurs near the pseudo-critical region give rise to some

special phenomena like small temperature drop and secondary peak of heat ﬂux. Yet these phenomena may become even more pronounced when the pressure is close to the critical pressure. 5. Concluding remarks

In this study, an experiment is conducted to investigate the per-formance and heat transfer process of a supercritical CO2

water-cooled gas cooler having tube-in-tube counter-flow configuration. The water is flowing in tube side whereas CO2 is flowing at the

annulus. The inlet pressures of the CO2 are 76, 86, and 96 bar

respectively. The temperature variation of the supercritical CO2

alongside the heat exchanger and the corresponding heating capacity subject to water coolant ﬂowrate are also reported. A tube-in-tube heat exchanger model applicable for supercritical ﬂuid CO2is also developed to compare the test results. The

mea-sured total heating capacity ranges from 1.31 to 4.06 kW at various test conditions in this study. The calculations are in line with the experimental results with 94% of the 36 data points being within the ±20% accuracy limits. The results also show that the variation of CO2temperature drop tends to be very small when it is close

to the pseudo-critical region when compared to that at the inlet re-gion. Moreover this phenomenon becomes more pronounced as the inlet pressure is close to the critical pressure (73.8 bar). This is associated with the gigantic rise of heat capacity at the pseu-do-critical region, and it corresponds to significant increase of heat transfer coefficient. The calculation also displays a peculiar phe-nomenon that the local heat transfer rate of the heat exchanger re-veals a local maximum and minimum during the trans-critical process due to the drastic rise of specific heat (CPvalue) of CO2

nearby the pseudo-critical region. Acknowledgements

The authors would like to express gratitude for the Energy R&D foundation funding from the Bureau of Energy of the Ministry of Economic, Taiwan. Part of the supporting from National Science Council under contract 102-ET-E-009-006-ET is also appreciated. References

[1]M.H. Kim, J. Pettersen, C.W. Bullard, Fundamental process and system design issues in CO2vapor compression systems, Prog. Energy Combust. Sci. 30 (2004)

119–174.

[2]A. Pearson, Carbon dioxide new uses for an old refrigerant, Int. J. Refrig. 28 (2005) 1140–1148.

[3]J.M. Yin, C.W. Bullard, P.S. Hrnjak, R-744 gas cooler model development and validation, Int. J. Refrig. 24 (2001) 692–701.

[4]B.M. Fronk, S. Garimella, Water-coupled carbon dioxide microchannel gas cooler for heat pump water heaters: part I. Experiments, Int. J. Refrig. 34 (2011) 7–16.

[5]B.M. Fronk, S. Garimella, Water-coupled carbon dioxide microchannel gas cooler for heat pump water heaters: part II – model development and validation, Int. J. Refrig. 34 (2011) 17–28.

[6]E.A. Groll, J.H. Kim, Review of recent advances toward transcritical CO2cycle

technology, HVAC&R Res. 13 (2007) 499–520.

[7]C. Dang, E. Hihara, In-tube cooling heat transfer of supercritical carbon dioxide part 1: experimental measurement, Int. J. Refrig. 24 (2004) 736–747. [8]C. Dang, E. Hihara, In-tube cooling heat transfer of supercritical carbon dioxide.

Part 2: Comparison of numerical calculation with different turbulence models, Int. J. Refrig. 24 (2004) 748–760.

[9]J. Sarkar, S. Bhattacharyya, M.R. Gopal, Simulation of a transcritical CO2heat

pump cycle for simultaneous cooling and heating applications, Int. J. Refrig. 29 (2006) 735–743.

[10] J. Sarkar, S. Bhattacharyya, M. Ramgopal, A transcritical CO2heat pump for

simultaneous water cooling and heating: test results and model validation, Int. J. Energy Res. 33 (2009) 100–109.

[11]L. Cecchinato, M. Corradi, S. Minetto, A critical approach to the determination of optimal heat rejection pressure in transcritical systems, Int. J. Refrig. 30 (2010) 1812–1823.

[12]P. Neksa, H. Rekstad, G.R. Zakeri, P.A. Schieﬂoe, CO2-heat pump water heater:

characteristics, system design and experimental results, Int. J. Refrig. 21 (1998) 172–179.

[13]P.C. Qi, Y.L. He, X.L. Wang, X.Z. Meng, Experimental investigation of the optimal heat rejection pressure for a transcritical CO2heat pump water heater,

Appl. Therm. Eng. 56 (2013) 120–125.

[14]S.G. Wang, Y.N. He, H.F. Tuo, F. Gao, Z.W. Xing, Effect of heat transfer area and refrigerant mass ﬂux in a gas cooler on heating performance of air-source transcritical CO2heat pump water heater system, Energy Build. 67 (2013) 1–

10.

[15]D. Sanchez, R. Cabello, R. Llopis, E. Torrella, Development and validation of a ﬁnite element model for water–CO2 coaxial gas-coolers, Appl. Energy 93

(2012) 637–647.

[16] REFPROP, 2007. Thermodynamic properties of refrigerants and refrigerant mixtures, version 8.0, Gaithersburg, M.D. National Institute of Standards and Technology, 2007.

[17]V. Gnielinsk, New equation for heat and mass transfer in turbulent pipe and channel ﬂow, Int. J. Chem. Eng. 16 (1976) 359–368.

0 20000 40000 60000 80000 100000 0 0.2 0.4 0.6 0.8 1 x/L C P of C O 2 ( J .kg -1 .K -1 ) P=76bar P=86bar P=96bar 0 100 200 300 400 500 600 700 800 0 0.2 0.4 0.6 0.8 1 x/L

Heat Transfer Rate (W

) P=76bar P=86bar P=96bar 0 4000 8000 12000 16000 20000 0 0.2 0.4 0.6 0.8 1 x/L h of C O2 ( W.m -2 .K -1 ) _P=76bar P=86bar P=96bar 0 4000 8000 12000 16000 20000 0 0.2 0.4 0.6 0.8 1 x/L h of Water (W.m -2 .K -1 ) P=76bar P=86bar P=96bar

(a)

(b)

(c)

(d)

Fig. 7. Effect of CO2 inlet pressure with a ﬁxed cooling water inlet condition

(2 L min1_{, 15 °C) on the (a) variation of Cp vs. dimensionless tube length; (b)}

variation of local heat transfer rate vs. dimensionless tube length; (c) variation of hc

(9)

[18]S.M. Liao, T.S. Zhao, Measurements of heat transfer coefﬁcients from supercritical carbon dioxide ﬂowing in horizontal mini/micro channels, Int. J. Heat Transfer 124 (2002) 413–420.

[19] ISO/IEC Guide 98-3:2008, Uncertainty of measure – Part 3: Guide to the expression of uncertainty in measurement (GUM:1995).

[20] B.N. Taylor, C.E. Kuyatt, Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results, National Institute of Standards and Technology, 1994.

[21]J.F. Wang, E. Hihara, Study on carbon dioxide gas cooler heat transfer process under supercritical pressures, Int. J. Energy Res. 26 (2002) 1237–1251. [22]P.Y. Yu, K.H. Lin, W.K. Lin, C.C. Wang, Performance of a tube-in-tube CO2gas