Documente Academic
Documente Profesional
Documente Cultură
a r t i c l e i n f o a b s t r a c t
Article history: In the present investigation a novel design algorithm is proposed for evaporative uid coolers (EFCs) and
Received 16 January 2010 closed circuit cooling towers. The principal of this study is to derive a relationship among heat transfer
Accepted 26 July 2010 coefcients, pressure drops, heat transfer area, and mass transfer coefcient. Using maximum allowable
Available online 6 August 2010
pressure drops in this algorithm leads to minimize the required heat transfer area and a straight forward
performance appraisal of EFC can be observed. It is revealed that the novel design algorithm can successfully
Keywords:
predicts the design outputs given by the other researchers. In addition, the new algorithm is easily adaptable
Heat transfer
when heat transfer enhancement devices applied inside or outside of the tube bundle of EFCs.
Mass transfer
Evaporative uid cooler
2010 Elsevier Ltd. All rights reserved.
Rapid design algorithm
Pressure Drop
1359-4311/$ e see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2010.07.030
M.R. Jafari Nasr, R. Behfar / Applied Thermal Engineering 30 (2010) 2746e2752 2747
12. Using the heat balance equation between spray water and the Thermal engineers commonly follow up the above steps to
tube wall, conrm the obtained area and the assumed mean design. However, as seen, this needs tedious trial and error calcu-
spray water temperature. lations in the procedure to meet not only to the required heat load
13. Finally, by knowing the total surface area and specic tube duty as an objective but also to satisfy the maximum allowable
dimensions, number of rows, number of pass, and length, pressure drops for airside and process uid as two main constraints.
calculate airside and tube side pressure drop using the Fanning
friction factor with the relevant empirical correlations.
14. Knowing the allowable tube side and airside pressure drops, 3. Rapid design algorithm
conrm the geometry and area of heat exchanger.
Polley et al. have presented design algorithms for the rapid
sizing of shell and tube and compact heat exchangers on a more
objective basis. These algorithms may be used in conjunction with
the detailed rating methods presented in the previous studies as
well as programs for mechanical design and tube bundle vibration
analysis to achieve to an optimal design [2,5]. The philosophy is to
utilize completely of the maximum allowable pressure drops on
both the hot and cold streams rather than use these specications
as mere constraints. Full use of both pressure drops ensures that the
exchanger is designed for the highest possible velocities and
consequently heat transfer coefcients. Thus, the exchanger is the
smallest size for a given service and presumably the most
economical [3].
In previous studies rapid design algorithm (RDA) applied as
a new approach to design shell and tube heat exchangers, compact
heat exchangers, and air-coolers in which the interface among heat
transfer coefcients, pressure drops, and heat transfer area, were
considered.
In this work the new design algorithm is extended to evapora-
tive heat exchangers while it is considered pressure drops as
objective function rather than constraints that are necessary in heat
exchanger design so the outcome is the most economical design in
comparison with the traditional methods. This new algorithm is
also adaptable when heat transfer enhancement technology
Fig. 1. An evaporative uid cooler. applied for tube bundles [4].
2748 M.R. Jafari Nasr, R. Behfar / Applied Thermal Engineering 30 (2010) 2746e2752
To introduce and describe the RDA for evaporative heat hw e4 $m0:333 (10)
w
exchangers, thermal and hydraulic characteristic of airside and tube
side analyzed separately with the following steps: Where,
By considering Equations (1) and (2) and elimination of velocity Since the heat transfer area is given by;
term between them, the relationship between the pressure drop,
heat transfer area, and heat transfer coefcient is obtained: A p$nr do ntr $L
Therefore by expanding the pressure drop equation we reach to:
Dpt s1 Ai ht3y=z (3)
Dpa e5 $A$m0:285
w (14)
Where,
Where,
e1 e2
s1 3y=z
(4)
e3 e5 4:9PL =do $p$L$ntr $ma =Ac 1:94 $2ntr $Pt $L0:285 (15)
1 Cpf DT hw s2 $Dp1:168
a $A1:168 (16)
e2 (6)
4mf 1 4q Where,
between the pressure drop and the mass transfer coefcient and
mass transfer area:
hd s3 $Dp0:487
a $A0:487 (23)
Where,
s3 e6 $e0:487
7 (24)
Consider the interface between the water and the air in Fig. 3 an imasw Cpa Tw wiv wsw wiv (34)
energy balance at the interface yields,
The enthalpy of the airewater vapor mixture per unit mass of
dQ dQm dQc (28) dry air is expressed by;
Where dQm is the enthalpy transfer due to difference in vapor
ima Cpa Ta w ifgwo Cpv Ta (35)
concentration between the saturated air at the interface and the
mean stream air and dQc is the sensible heat transfer due to the Subtracting Equation (35) from (34), the resultant equation can be
difference in temperature. The mass transfer at the interface is simplied if the small differences in specic heats, which are
expressed by; evaluated at different temperatures, are ignored;
dmw hd wsw wdA (29) imasw ima wsw wiv
Tw Ta (36)
The corresponding enthalpy transfer for the mass transfer in Cpma
Equation (29) is:
Substituting Equation (36) into Equation (32) and similarly by
substituting Equation (30) into Equation (28) and nally with
dQm iv dmw iv hd wsw wdA (30)
substitution of the resultant equations follow by a rearrangement,
The enthalpy of the water vapor, iv, at the bulk water tempera- we yield to:
ture, Tw, is given by;
h h
iv ifgwo cpv Tw (31) dQ hd imasw ima 1 iv wsw w dA
Cpma hd Cpma hd
The convective heat transfer from Fig. 3 is given by; (37)
h/Cpmahd in Equation (37) is known as the Lewis factor Lef and is an
indication of the relative rates of heat and mass transfer in an
evaporative process.
The enthalpy transfer to the air stream from Equation (37) is:
1
dima dQ
ma
h dAh i
d Lef imasw ima 1 Lef iv wsw w (38)
ma
To simplify the analysis of an evaporative cooling process Merkel
assumed that the evaporative loss is negligible, i.e. dw 0, From
Equation (27), and that the Lewis factor is equal to one. Equations
(27) and (38) of the counter ow evaporative process simplify
respectively to:
140 3000
100
2000
80
Area(m 2)
60 1500
40
1000
20
500 Traditional Algorithm
0 Rapid Design Algorithm
1 2 3 4 5 6
0
Tubeside mass flow rate(kg/s)
1 2 3 4 5 6
Fig. 4. Variation of heat transfer area vs. tube side mass ow rate for two algorithms. Tubeside mass flow rate (kg/s)
Fig. 6. Variation of spray water lm heat transfer coefcient vs. tube side mass ow
24.81 mm and 28.65 mm, respectively, thus heat transfer area A is rate for two algorithms.
27.17 m2. Nominal operating conditions for the tower are: airow
rate ma 2.07 kg s1, process cooling water ow rate
equations. It should be noted that Mizushina [6] ndings for
mf 2.67 kg s1, spray water ow rate mw 1.845 kg s1, inlet
transfer coefcients has been applied in this algorithm. As
cooling water temperature T 15.6 C, inlet dry bulb temperature
explained, this algorithm is in fourteen stages. The allowable
Tai 10 C, inlet wet bulb temperature Twbi 8.45 C and tube wall
pressure drops were determined arbitrary; however they are close
thermal conductivity kw 45 W m1 K1 (galvanized steel).
to those calculated from Krger sample.
Krger executed rating problem and obtained the following
The last column is determined based on the RDA algorithm.
results for the nominal operating conditions: total rejected heat is
Because of the full usage of maximum allowable pressure drops and
22379 W, outlet cooling water temperature tfo 13.6 C, and mean
consequently maximum uid velocity and heat transfer coefcients
spray water temperature 13.12 C. Although Krger has not
leads to minimum area.
addressed, pressure drops are calculated for Krger problem for
The new design algorithm shows a 26.5% increase in the amount
comparison. Table 1 presents a comparison between the Krger
of mass transfer coefcient, 17.7% increase in the amount of spray
calculations, the results obtained from the traditional algorithm
water lm heat transfer coefcient and 92.2% increase in the
based on Mizushina ndings, and the new design algorithm.
amount of tube side heat transfer coefcient.
First column is related to Krgers results for the rating problem.
It should be noted in traditional design procedure, tube side
It should be noted that he has used transfer coefcients which
pressure drop equation is obtained according to Krger:
presented by Parker and Treybal.
Second column is related to the traditional algorithm for sizing
problem that uses Effectiveness-NTU method and the basic thermal Dp fr L=di rv2 =2
3500
0.35
3000
Tubeside heat transfer coefficient(W/m 2 K)
Traditional Algorithm
0.3 Rapid Design Algorithm
Mass transfer coefficient(kg/sm 2 )
2500
0.25
2000
0.2
1500
0.15
1000
0.1
Traditional Algorithm
500 0.05
Rapid Design Algorithm
0 0
1 2 3 4 5 6 1 2 3 4 5 6
Tubeside mass flow rate(kg/s) Tubeside mass flow rate (kg/s)
Fig. 5. Variation of tube side heat transfer coefcient vs. tube side mass ow rate for Fig. 7. Variation of mass transfer coefcient vs. tube side mass ow rate for two
two algorithms. algorithms.
2752 M.R. Jafari Nasr, R. Behfar / Applied Thermal Engineering 30 (2010) 2746e2752
and airside pressure drop equation is obtained according to Nitsu as EFCs can be optimized straight forward without needing to any
Equation (12). tedious trial and error calculations.
Figs. 4e7 show comparison of the variations between traditional In general by considering maximum allowable pressure drops
algorithm and RDA Algorithm for a range of tube side mass ow for both sides, the frontal area of ows is minimized. Therefore, the
rate variation. velocity of ows can be maximized as possible. This event results
Fig. 4 shows comparison of the area variations between tradi- the heat and mass transfer coefcients maximized so the heat
tional algorithm and RDA Algorithm. As shown the area in the new transfer area can be minimized. The performance was demon-
algorithm has reduced because of considering maximum allowable strated by its application for design of the optimum EFCs in a case
pressure drops, maximum possible velocities and maximum study. The generated optimum EFCs were better than those
transfer coefcients at both sides. Note that pressure drops in the reported in the literature. However, the speed and exibility of the
traditional algorithm are less than allowable ones but in the new proposed procedure is superior to that of traditional algorithms,
algorithm pressure drops are just equal to allowable pressure mainly because it produces the minimum heat transfer area
drops. required in a very short time as mentioned earlier.
The benet of considering maximum allowable tube side pres- The new RDA algorithm that developed for EFC with plain tubes
sure drop in design algorithm is obvious. Fig. 5 presents comparison can also be extended simply to enhanced tube bundles such as
of the tube side heat transfer coefcient variations between tradi- lled tubes by various tube inserts and even twisted tubes just
tional and RDA algorithms. Considering maximum allowable tube simply by replacing the heat transfer and friction factor coefcients
side pressure drop causes maximum possible velocity in tubes that correspondingly. This subject is considered for another paper
makes maximum turbulence in boundary layer (increscent of the publication by authors.
Reynolds number expresses this upheaval) and results the
maximum heat transfer coefcient is obtained.
References
Fig. 6 shows comparison of the spray water lm heat transfer
coefcient variations between traditional algorithm and RDA. By [1] Ala Ali Hasan, Performance Analysis of Heat Transfer Processes from Wet and
considering the maximum allowable airside pressure drop, Dry Surfaces: Cooling Towers and Heat Exchangers, the Department of
maximum possible velocity for spray water lm can be obtained. Mechanical Engineering. Helsinki University of Technology, 2005, Ph.D.
Thesis.
This means the frontal area can be minimized and the spray water [2] G.T. Polley, M.H. Panjeh Shahi, M. Picon Nunez, Rapid Design Algorithms for
mass ow rate per unit length over the tube bundle can be maxi- Shell-and-Tube and Compact Heat Exchangers, Trans. IChemE. Chem. Eng. Res.
mized which causes maximum heat transfer coefcient. Des. 69 (1991) Part A 435e444.
[3] Uday V. Shenoy, Heat Exchanger Network Synthesis. Gulf Publishing Co,
When maximum allowable airside pressure drop is considered, Houston, Texas, USA, 1995, Chapter 6.
the airside velocity increases as much as possible. In other words, [4] M.R. Jafari Nasr, G.T. Polley, Should you use enhanced tubes, Chemical Engi-
air velocity and spray water velocity rise as much as possible. neering Progress 98 (4) (2002) 44e50.
[5] M.R. Jafari Nasr, G.T. Polley, Extension of rapid sizing algorithm for shell-and-
Therefore, air Reynolds number and spray water Reynolds numbers tube heat exchangers with tube-side pressure drop constraint and multi-
increase and consequently the mass transfer coefcient is maxi- passes, Chemical Engineering and Technology, CET, Part B (2000) 141e150.
mized. Fig. 7 indicates this behavior. [6] M.R. Jafari Nasr, S.H. Alaei, A New Algorithm for Design, Simulation, and
Optimization of Enhanced Air Coolers, Journal of Enhanced Heat Transfer 14
(2) (2007) 147e160.
5. Conclusion [7] T. Mizushina, R. Ito, H. Miyashita, Experimental study of an evaporative cooler,
International Chemical Engineering 7 (4) (1967) 727e732.
[8] T. Mizushina, R. Ito, H. Miyashita, Characteristics and methods of thermal
A novel design procedure, based on the Rapid Design Algorithm
design of evaporative cooler, International Chemical Engineering 8 (3) (1968)
(RDA), for the optimum design of evaporative heat exchangers in 532e538.
EFCs and closed circuit cooling towers has been presented. In [9] Y. Nitsu, K. Naito, T. Anzai, Studies on characteristics and design procedure of
contrast with the traditional algorithms, in this work by developing evaporative coolers, Journal Of The Society Of Heating Air-Conditioning
Sanitary Engineering Of Japan 41 12 (13) (1967).
a simple relationship among the pressure drop, area, the heat [10] D.G. Krger, Air-Cooled Heat Exchangers and Cooling Towers, vol. 1, Penn
transfer coefcients, and the mass transfer coefcient, the design of Well Corp, Tulsa, Oklahoma, 2004, pp. 236e298.