Effect of partial bypass on the heat transfer performance of dehumidifying coils

(1)

Effect of partial bypass on the heat transfer performance of

dehumidifying coils

☆

Chi-Chuan Wang

a,

⁎

, Yu-Cheng Cheng

a

, Jane-Sunn Liaw

b

, Chih-Yung Tseng

b

a

Department of mechanical engineering, National Chiao Tung University, Hsinchu 300, Taiwan

b_{Green Energy & Environment Research Laboratories, Industrial Technology Research Institute, Hsinchu 310, Taiwan}

a b s t r a c t

a r t i c l e i n f o

Available online 8 September 2014 Keywords:

Air-cooled heat exchanger Partial bypass

Heat transfer Pressure drop

This study extends a so called“partial bypass” concept to dehumidifying fin-and-tube heat exchangers. Tests are performed for a 2-row and 4-rowfin-and-tube heat exchangers having plain fin configuration. The inlet dry bulb temperature isfixed at 25 °C while the inlet relative humidities are 50% and 80%, respectively. Test results showed that for a 2-row coil under RH = 50%, the corresponding heat transfer augmentation ratio (QR) and

pres-sure penalty ratio (PR) decrease with the rise of bypass ratio (BR). At a smaller frontal velocity, the regime of

ap-propriate bypass ratio where QRN 1 and PRb 1 is more apparent. The effect of partial bypass decreases as the

velocity increases. As the velocity gets close to 2 m s−1, it shows that the performance at high bypass ratio be-comes more efﬁcient. At the same time, when bypass ratio becomes lager, the tendency of the decreasing of ΔP is more drastic than the decreasing of Q. Nevertheless, not for all the cases that the performances get better as the bypass ratio increases. For Vfr= 4 m s−1, the experiment results show that the overall effect is less efﬁcient

than at Vfr= 2 m s−1. The major reason is caused by theﬂow pattern. The rise of bypass airﬂow is quite large

which may act as an air curtain to distort the main airﬂow, and result in higher pressure drop. This implies that the design of the bypass design is quite imperative in the practice of real applications. For a 4-row heat exchanger, the experiment data of QRand PRfor the 4-row coil performs better than a 2-row coil.

1. Introduction

The air-cooled heat exchangers are widely used in a variety of in-dustrial, commercial, and residential applications as coolers, heaters, evaporators and condensers. The most common configuration of the air-cooled heat exchangers takes the form asfin-and-tube where air normallyflows along the fin (outside) and the other working fluids flow in the tube side. Depending on the applications and operational conditions, the air side thermal resistance can be as high as 70_–90% of the total thermal resistance. As a consequence, it is imperative to reduce the associated resistance as far as the heat exchanger efficiency is concerned. The most effective reduction of the thermal resistance is via increasing voluminous surface but it also accompanies with gigantic pressure drop penalty. Hence, most studies had focused on increasing the air side heat transfer coefficient via enhanced geometries such as wavy, slit, louver, and vortex generator geometries, and some ingenious designs regarding to thefin patterns were reviewed by Wang[1,2].

Depending on the applications, multiple tube rows can be employed to increase surface area for increasing heat transfer rate. With multiple tube row designs, though the surface area increases linearly with the tube row, the heat transfer rate, unfortunately shows only moderately or only marginally increase with respect to the tube row[3,4]. This phe-nomenon becomes more and more severe when the number of tube row is further increased[5]. To make thing even worse, the pressure drop penalty is proportional to the number of tube row, meaning a lin-ear increase of pumping power. It is quite often to employ air-cooled heat exchangers having deep row configuration. For instance, the commonly used ventilator, heat recovery system, and some large heat exchanging facility take form as deep row coils. Of course, one can always use highly interrupted surfaces to augment airside performance to offset the surface requirement, and accordingly a less number of tube row can be achieved. Nevertheless, highly interrupted surfaces, like louver, offsetfin, slit fin, and the like are prone to fouling and are normally regarded inappropriate for indus-trial applications or large system[6].

In this regard, the present authors extend a newly proposed concept, “partial bypass”, of heat transfer enhancement for this kind of air-cooled heat exchangers[7,8], and the concept can be extended to any highly compact heat exchangers as well. To explain the proposed concept for heat transfer augmentation, let's take the two samples from Wang et al.[8]using two air-cooledﬁnned heat exchangers having a number

☆ Communicated by W.J. Minkowycz.

⁎ Corresponding author at: EE474, 1001 University Rd., National Chiao Tung University, Taiwan.

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

http://dx.doi.org/10.1016/j.icheatmasstransfer.2014.08.029

Contents lists available atScienceDirect

International Communications in Heat and Mass Transfer

(2)

of tube row of 4 and 8 respectively for illustration. A schematic showing the operation of these heat exchangers can be seen inFig. 1(a) and (b). These two heat exchangers have the same longitudinal tube pitch, transverse tube pitch,ﬁn collar outside diameter, and ﬁn surface area and an identical frontal area, the inlet temperature of the air (Tair,in)

and hot water (Twater,in) is also maintained atﬁxed temperatures,

respectively. The heat exchangers are arranged in pure crossﬂow conﬁguration with 4 and 8 circuitries for the tested 4-row and 8-row heat exchangers (see Wang et al.[8]for detailed circuitry ar-rangement). For a frontal velocity (Vfr) of 2 m s−1, the

correspond-ing heat transfer rate and pressure drop for the 4-row heat exchanger is about 4540 W and 43.2 Pa while the heat transfer rate is moderately increased to 5561 W accompanied with a nearly doubled pressure drop of 85 Pa for the 8-row coil. The results impli-cate adding more surfaces to augment heat transfer rate is compar-atively ineffective. The results are not surprising for the maximum heat transfer rate (Qmax= Cmin(Twater,in–Tair,in)) for the heat

ex-changers are the same. Notice that the minimum capacity rate, nor-mally Cmin( m

cp; m

: mass_{ﬂow rate, c}p: speciﬁc heat), is on the

airside for an air-cooled heat exchanger. Raising the number of tube row (N) leads to a rise of the number of transfer unit (NTU = UA/Cmin, U: overall heat transfer coefﬁcient, A: total surface

area) and effectiveness (ε = Q/Qmax, Q: actual heat transfer rate).

The proposed concept is rather simple, focusing on increasing the Qmaxat the rear part of the heat exchanger. For an illustration

of this concept, Wang et al.[8]split the 8-row heat exchanger into two identical 4-row heat exchangers, and some spaces are between these two 4-row heat exchanger (some mixing mechanisms can be provided in the space if needed). However, theﬁrst heat exchanger is not tightlyﬁtted into the duct area. Instead, some bypass area is Nomenclature

A surface area (m2₎

BR bypass ratio, deﬁned in Eq.(1)(−) cp speciﬁc heat (kJ kg−1K−1)

C heat capacity rate (W K−1)

m

massﬂow rate (kg s−1₎

N number of the tube row (−)

NTU number of transfer unit, UA/Cmin(−)

PR ratio of pressure drop (−)

QR ratio of heat transfer rate (−)

Q heat transfer rate (W)

RH relative humidity (%)

T temperature (K)

ε effectiveness (₋₎

ΔT temperature difference (K)

ΔP pressure drop (Pa)

U overall heat transfer coefﬁcient (W m−2_K−1₎

Vfr frontal velocity (m s−1)

Subscript

1 ﬁrst heat exchanger

2 second heat exchanger

air air side

in inlet

max maximum value

min minimum value

water water side

Fig. 1. Schematic of operation of the (1) 4-row heat exchanger; (2) 8-row heat exchanger; and (3) the proposed“partial bypass” heat exchanger.

133 C.-C. Wang et al. / International Communications in Heat and Mass Transfer 58 (2014) 132–137

(3)

(a)

Schematic of the experimental setup

(b)

Photo of the mixing device placed amid the first and the second heat exchanger.

(c)

Schematic of the bypass flow into the bypass device.

(4)

designated as schematically shown inFig. 1(c). The bypass ratio of the airﬂow across the ﬁrst heat exchanger is designated as BR¼_{total flow rate TF}bypass flow rate BF_ð ð_{; evaluated at the second heat exchanger}; bypass from the first heat exchangerÞ_Þ:

ð1Þ With the foregoing design, the performance in theﬁrst part of the heat exchanger is degraded but the performance of the second part of the heat exchanger is appreciably enhanced. Wang et al.[8]had showed that the overall heat transfer performance is about the same but the pressure drop is signiﬁcantly reduced. In this study, the present authors had extended the concept to examine the applicability in dehumidifying heat exchangers.

2. Experimental apparatus and data reduction

The schematic diagram of the experimental air circuit assembly is shown inFig. 2(a). It consists of an open loop wind tunnel in which air is circulated by a variable speed centrifugal fan (3.7 kW, 5 HP). The air duct is made of steel sheet and has a 300 mm × 300 mm cross-section. The temperature and relative humidity of the environment are controlled by an environmental chamber. The airflow rate measure-ment station is set up with multiple nozzles. This setup is based on the ASHRAE 41.2 standard[9]. A differential pressure transducer is used to measure the pressure difference across the nozzles. The air tempera-tures at the inlet and exit zones across the sample heat exchangers are measured by three psychrometric boxes constructed based on the ASHRAE 41.1 standard[10]. Besides, the dry-bulb temperatures of the inlet and outlet air across the sample heat exchangers are also measured by pre-calibrated thermocouples, with a calibrated accuracy of ±1.1%. The pressure drop of the sample heat exchangers is measured by a differential pressure transmitter. The bypass device is made of acrylic plates as seen inFig. 2(b) and is connected with the two heat exchangers. Bypass airflow is provided by another variable speed centrifugal fan (0.75 kW, 1 HP). The schematic of the bypass device is shown inFig. 2(c). Main airflow is from the left side of the bypass device and the bypass airflow is from the top side of the device. Notice that the whole experimental setup is enclosed in an environmental chamber with controllable dry bulb and wet bulb temperatures.

The working medium on the tube side is cold water. A thermo-statically controlled reservoir provides the cold water at selected temperatures. The temperature differences on the water side are measured by pre-calibrated RTDs. The water volumetricﬂow rate is measured by a magneticﬂow meter. All the temperature measur-ing probes are resistance temperature device (Pt100), with a cali-brated accuracy of ± 1.3%. The test conditions of the inlet air are as follows:

Dry-bulb temperatures of the air: 25 ± 0.5 °C; Inlet relative humidity for the incoming: 50 and 80%; Inlet air velocity: From 0.5 to 4.0 ms−1;

Inlet water temperature: 2 ± 0.5 °C; and Water massﬂow rate inside the tube: 1–4 L/min. 3. Results and discussion

To illustrate the proposed partial bypass concept applicable to dehumidifying coils, experiments were conducted for 2- and 4-row heat exchangers subject to a dry bulb temperature of 25 °C and RH = 50 and 80%, respectively. Experiments were made withﬁxed frontal velocities into the second heat exchanger of 0.5, 1, 2, and 4 ms− 1, respectively. The results shown in theﬁgure are in terms of ratios of heat transfer (QR= Qt/Q2or QR= Qt/Q4) and of pressure

drop (PR=ΔPt/ΔP2orΔPt/ΔP4) where the subscripts 2 and 4 denote

experimental results for the 2-row heat exchanger and 4-row heat exchanger having the sameﬁn pitch respectively. Hence enhanced heat transfer is achieved when QRN 1. On the other hand, PRb 1

denotes no pressure drop penalty. As far as the best performance is concerned, it would be better to achieve QRN 1 and PRb 1

simulta-neously. The experimental data of QRand PRfor the 2-row coil are

shown inFig. 3(a) and (b) under RH = 50%. As expected, both QR

and PRdecrease with the rise of bypass ratio (BR). In the meantime,

the region with QRN 1 and PRb 1 depends on the frontal velocity.

At a smaller frontal velocity, the regime of appropriate bypass ratio where QRN 1 and PRb 1 is more apparent. Take Vfr= 0.5 ms−1for

example, the value of QRis higher or almost equal to 1 and in the

meanwhile PRdecreases nearly 6% at BR = 0–0.2. The results can

be explained from the relation ofε-NTU as seen inFig. 4. With a smaller frontal velocity, it is expected that the Cminis reduced. In

the meantime, U is also reduced but the degree of reduction is small-er than that of Cmin. In this sense, the number of transfer unit, NTU,

actually increased, leading to a higher effectivenessε. Therefore, ε can be easily improved via increasing the Qmaxwhen it is operated

at a higher value of NTU. In this regard, the region for QRN 1

increased. However, the effect of partial bypass decreases as the ve-locity increased. As the veve-locity gets close to 2 m s−1, it shows that the performance at high bypass ratio becomes more efﬁcient. At the same time, when the bypass ratio becomes lager, the tendency of the decreasing ofΔP is more drastic than the decreasing of Q. This effect shows that the major beneﬁt of this operation is to reduce

Fig. 3. (a) Experimental results for QRvs. bypass ratio for a 2-row coil under RH = 50%;

(b) experimental results for PRvs. bypass ratio for a 2-row coil under RH = 50%.

(5)

pressure drop. The obvious result for Vfr= 2 m s− 1is that even

though Q decreased by 17%, the corresponding reduction ofΔP is about 40% at BR = 0.8.

Nevertheless, it is not applicable for all the cases that the perfor-mances get better as the bypass ratio increased. For Vfr= 4 m s−1, the

experimental results show that the overall effect is less ef_{ﬁcient than} that at Vfr= 2 m s−1. Moreover, at a low bypass ratio (BR = 0.1–0.3),

ΔP is decreased slowly. This can be explained from the present bypass device which is shown inFig. 2(c). The major reason is caused by the flow pattern being delivered. As shown inFig. 2(c), bypass airflows into the mixing region from the top of the bypass device. Because the bypass airflow is quite large at Vfr= 4 ms−1, it acts as an air curtain

and inhibits the main air stream. Therefore, the main airﬂow is distorted by the bypass airﬂow and results in an uneven decline of pressure drop. This implies that the design of the bypass design is quite essential in the practice of real applications.

The experimental data of QRand PRfor the 2-row coil are shown in

Fig. 5(a) and (b) under RH = 80%. Compared to the conditions with RH = 50%, the overall performances are better under RH = 80%. It's because the latent heat of the heat exchanger increases with the envi-ronmental relative humidity. Meanwhile the pressure drop of the heat exchanger increases as relative humidity increases. As a consequence, the decrease in pressure drop through the concept of partial bypass becomes more pronounced. Hence, it's apparent to see a better improvement under a high relative humidity. As seen inFig. 5(a) and (b), for Vfr= 0.5 and 2 ms−1, QRN 1 as well as PRb 1 can be achieved

at a lower bypass ratio. Especially for Vfr= 0.5 ms−1, the pressure drop

can reduce about 15% with the same heat transfer rate at BR = 0.4. The experimental results also show that there are better performances for Vfr= 2 ms−1at high bypass ratio. Even though Q decreased by 11%,ΔP

Fig. 4. Relation of theε-NTU for unmixed/unmixed cross ﬂow arrangement subject to variation of C*.

(6)

signiﬁcantly decreased about 37% at BR = 0.8. On the other hand, the heat transfer rate can be augmented based on an identical pumping power cri-terion. Nevertheless, for Vfr= 4 ms−1, the experimental results also show

that at a lower bypass ratio (BR = 0.1–0.3), ΔP remains almost the same. This is the same reason as mentioned from the previous explanation where the directed airﬂow may distort the main ﬂow.

For a 4-row heat exchanger, the experimental data of QRand PRfor

the 4-row coil are shown inFig. 6(a) and (b) under RH = 50%. In gener-al, the performances of a 4-row coil are better than a 2-row coil under RH = 50%. At a lower bypass ratio, basically the best situation is at Vfr= 0.5 ms−1. The heat transfer rate just decreased near 5%, but the

pressure drop decreased about 17% at BR = 0.2. On the other hand, at a high bypass ratio, the best situation is at Vfr= 4 ms−1. For Vfr=

4 ms−1, even though Q decreases 16%,ΔP decreases about 42% at BR = 0.8. For a 4-row coil, the percentage of heat transfer rate occurs atﬁrst heat exchanger is more than that at second heat exchanger com-pared to a 2-row coil. According to theﬁgures shown inFig. 7(a) and (b), the overall performances of a 4-row HX under RH = 80% are similar to a 4-row HX under RH = 50% and is also similar to a 2-row HX under RH = 80%. For all the cases, Vfr= 2 ms−1shows the better results than

the other velocities.

4. Conclusions

In this study, the novel_{“partial bypass” concept had been extended} to test its applicability in dehumidifying coils. Tests are performed for a 2-row and 4-rowfin-and-tube heat exchangers having plain fin con-figuration. The inlet dry bulb temperature is fixed at 25 °C while the inlet relative humidities are 50% and 80%, respectively. Based on the foregoing discussion, the following conclusions are made:

(1) For a 2-row coil under RH = 50%, the corresponding QRand PR

decrease with the rise of bypass ratio (BR). At a smaller frontal velocity, the regime of appropriate bypass ratio where QRN 1

and PRb 1 is more apparent. For instance, for Vfr= 0.5 ms−1,

the value of QRis higher or almost equal to 1 and in the

mean-while PRdecreases nearly 6% at BR = 0–0.2.

(2) The effect of partial bypass decreases as the velocity increased. As the velocity gets close to 2 m s−1, it shows that the performance at a high bypass ratio becomes more efﬁcient. At the same time, when the bypass ratio becomes lager, the tendency of the reduction ofΔP is more drastic than the decreasing of Q. (3) Nevertheless, not for all the cases that the performances get

better as the bypass ratio increases. For Vfr= 4 m s−1, the

ex-perimental results show that the overall effect is less efﬁcient than at Vfr= 2 m s−1. The major reason is caused by theﬂow

pattern. The rise of bypass airﬂow is quite large which may act as an air curtain to distort the main airﬂow, and result in higher pressure drop. This implies that the design of the bypass design is quite imperative in the practice of real applications.

(4) For a 4-row heat exchanger, the experimental data of QRand

PRfor the 4-row coil performs better than a 2-row coil.

Acknowledgment

The authors appreciate theﬁnancial support from the Energy Bureau from the Ministry of Economic Affairs and the Ministry of Science and Technology (under grant 103-3113-E-009-002), Taiwan.

References

[1] C.C. Wang, Technology review— a survey of recent patents of ﬁn-and-tube heat exchangers, J. Enhanc. Heat Trans. 7 (2000) 333–345.

[2] C.C. Wang, A survey of recent patents ofﬁn-and-tube heat exchangers from 2001– 2009, Int. J. Air-Cond. Refrig. 18 (2010) 1–13.

[3] G. Xie, Q.W. Wang, B. Sunden, Parametric study and multiple correlations on air-side heat transfer and friction characteristics ofﬁn-and-tube heat exchangers with large number of large-diameter tube rows, Appl. Therm. Eng. 29 (2009) 1–16.

[4] J.M. Choi, Y. Kim, M. Lee, Y. Kim, Air side heat transfer coefficients of discrete plate finned-tube heat exchangers with large fin pitch, Appl. Therm. Eng. 30 (2010) 174–180.

[5] C.C. Wang, J.S. Liaw, B.C. Yang, Air-side performance of herringbone wavy ﬁn-and-tube heat exchangers— data with larger diameter tube, Int. J. Heat Mass Transf. 54 (2011) 1024–1029.

[6] I.H. Bell, E.A. Groll, Air-side particulate fouling of microchannel heat exchangers: experimental comparison of air-side pressure drop and heat transfer with plate-ﬁn heat exchanger, Appl. Therm. Eng. 31 (2011) 742–749.

[7] C.C. Wang, Enhanced heat transfer performance of air-cooled heat exchangers using “Partial Bypass” concept, Heat Transf. Eng. 33 (2012) 1217–1219.

[8] C.C. Wang, K.Y. Chen, J.S. Liaw, C.Y. Tseng, A novel“Partial Bypass” concept to augment performance of air-cooled heat exchangers, Int. J. Heat Mass Transf. 55 (2012) 5367–5372.

[9] ASHRAE Standard 41.2-1987, Standard Methods for Laboratory Air-ﬂow Measure-ment, Atlanta American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc., 1987.

[10] ASHRAE Standard 41.1–1986, Standard Method for Temperature Measurement, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc., Atlanta, 1986.