y=/EPA
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-82-014 August 1982
Project Summary
Dry/Wet Performance of a
Plate-Fin Air-Cooled Heat
Exchanger with Continuous
Corrugated Fins
S. G. Mauser, D. K. Kreid, and B. M. Johnson
The goal of the project was to contribute
to the development of improved cooling
facilities for power plants that would
help to conserve increasingly scarce
fresh water supples in an environmentaly
compatible and economically viable
manner. Specific objectives of this work
were:
To experimentally determine the
performance and operating charac-
teristics of a plate-fin heat exchanger
in dry/wet or "deluge" operations.
To continue development of the del-
uge heat/mass transfer model.
The experiments were conducted in a
specially designed wind tunnel at Battefe-
Pacrfic Northwest Laboratory (PNL). In
the tests, air that was first heated and
humidified to specified conditions was
circulated at a controlled rate through a
2-ft x 6-ft* heat exchanger module.
The heat exchanger used in the tests
was a wavy surface, plate-fin-on-tube
configuration. Hot water was circulated
through the tubes at high flow rates to
maintain an essentially isothermal condi-
tion on the tube side. Deionized water
sprayed on the top of the vertical plate
fins was collected at the bottom of the
core and recirculated. Instrumentation
'EngH*h engineering, rather than SI. units are used
In this summary; these unit* an convantlonaly used
by designers and users of heat exchangers In the
U.S. Conversion factors between these units are
provided near the and of this summary.
was provided for measurement of flow
rates and thermodynamic conditions In
the air, in the core circulation water, and
in the deluge water.
The air-side pressure drop and heat re-
jection rate were measured as a function
of air flow rate, air inlet temperature and
humidity, deluge water flow rate, and
core inclination from the vertical. The
data were reduced to determine an over-
all heat transfer coefficient and an effec-
tive deluge film convective coefficient.
The deluge model is an approximate
theory for predicting heat transfer from a
wet finned heat exchanger that was
developed in preceding work. The model
was further developed and refined in this
study, and a major extension of the
model was formulated that permits si-
multaneous calculation of both the heat
transfer and evaporation rates from the
wetted surface. The model was used to
reduce and correlate the data and to
evaluate the results. In general, the
analytical predictions were in excellent
agreement with the experiments.
The experiments showed an increase
in the heat rejection rate due to wetting,
accompanied by a proportional increase
in the air-side pressure drop. For opera-
tion at the same air-side pressure drop,
the enhancement ratio, Qwct/Qdry varied
between 2 and 5 for the conditions tested.
Thus, the potential enhancement of heat
transfer due to wetting can be substantial.
-------
However, a number of important trade-
offs exist that must be considered in an
overall assessment of deluge cooling for
a particular application.
This Protect Summary was developed
by EPA'» Industrial Environmental Re-
search Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fuHy docu-
mented In a separate report of the same
title (see Project Report ordering infor-
mation at back).
Introduction
This report provides the experimental
data and supporting theoretical relation-
ships to substantiate a key portion of the
design of an advanced concept for dry/
wet cooling of thermal power generating
plants.
The work was jointly supported by the
EPA and DOE because of the dual incen-
tive that exists for developing improved
cooling systems. Dry cooling has been
the subject of extensive studies by both
agencies because of the growing realiza-
tion that the use of fresh inland water to
provide a heat sink for thermal genera-
tion of power cannot continue to in-
crease indefinitely. Thus the EPA was in
the forefront of early studies to identify
the feasibility and cost of supplementing
the use of freshwater for cooling and
thus reduce the environmental impact of
either consuming large quantities of
freshwater by evaporative cooling or re-
turning an even larger quantity of water
to its original source after being heated
through 20-35 °F.
Except in special situations, it is likely
that the use of freshwater for cooling
will be supplemented by combination
wet and dry systems, because using a
small amount of cooling water reduces
the cost of a dry cooling system far more
than a proportionate difference in its
cost and that of an evaporative system.
Nevertheless, the costs of present dry/
wet cooling systems are so high that util-
ities generally agree that they will be
used only in isolated situations unless
significantly lower cost systems can be
developed. However, because of the
uncertain market for dry cooling, manu-
facturers are reluctant to make large
capital expenditures to develop and
demonstrate radically new approaches.
Public agencies such as DOE and EPA
and the utility-industry-supported re-
search organization, EPRI, have conse-
quently taken the lead in developing ad-
vanced technology for dry/wet cooling.
The Pacific Northwest Laboratory
(PNL), operated for the DOE by Battelle
Memorial Institute, has a major program,
portions of which are funded by these or-
ganizations. The multifaceted work in-
cludes: (1) identifying the need for dry/
wet cooling, (2) assessing the state-of-
the-art and potentials for improvement,
(3) identifying promising advanced con-
cepts, (4) developing technology in sup-
port of selected advanced concepts, (5)
assessing new concepts as they are pro-
posed, and (6) carrying out a large-scale
test of the most promising advanced
concept.
Of many novel concepts proposed by
investigators around the world, a pro-
cess was selected for large-scale testing
which uses ammonia to transport the re-
ject heat from the last stage of the tur-
bine to the air-cooled heat exchanger.
The system also includes the use of
evaporative cooling to augment the dry
cooling in either of two ways:
(1) Deluge cooling in which water is
allowed to flow in excess over the
dry cooling surface.
(2) Parallel condensing of the ammo-
nia in an evaporative condenser
(one in which the bare ammonia
condenser tubes are cooled by
water and air flowing simultane-
ously over the outside surface).
Deluge cooling, in which the dry heat
exchanger surface is covered with a thin
film of water so that evaporative cooling
and sensible heat transfer occur simul-
taneously, appears to be a relatively sim-
ple and inexpensive way of achieving
augmented cooling (i.e., dry/wet cool-
ing). It has been used to some extent in
air conditioning applications in the U.S.
However, for large-scale power plant
use, several uncertainties must first be
overcome in performance prediction and
proper design of the extended surface to
permit good dry performance, together
with proper water distribution to avoid
scaling and corrosion. The concept has
been under study at PNL.
The deluge cooling concept has been
tested on several heat exchangers by
PNL. The performance of the "Forgo"
plate-fin heat exchanger surface, devel-
oped by and manufactured for the HO-
TERV Institute of Hungary (hereafter
identified as the "HOTERV" exchanger)
was determined under both the dry and
the dry/wet operating modes. In addi-
tion, the dry performance of two config-
urations of a chipped fin (or skived) heat
exchanger surface manufactured under
license from the Curtiss-Wright Co. was
tested.
This report provides the data and
theoretical basis for predicting the per-
formance of another plate-fin heat ex-
changer which was manufactured by
the Trane Co. for air conditioning ser-
vice. This was selected for testing in the
Advanced Concepts Test (ACT) facility
because it was more readily adaptable to
ammonia condensation, and procure-
ment was more convenient and less ex-
pensive, due in part to Trane Co.'s manu-
facturing capability in the U.S., than
other candidates' heat exchangers.
The objectives of the work carried on
in the Water Augmentation Test Appara-
tus (WATA) are:
(1) To determine all-dry nonaugment-
ed performance for comparison
with other air-cooled heat ex-
changer surfaces such as the HO-
TERV and Curtiss-Wright surface.
(2) To establish the magnitude of the
potential benefit due to augmenta-
tion.
(3) To measure dry/wet heat transfer
performance and air-side pressure
drop as they are affected by wea-
ther conditions (air temperature
and humidity), air flow rate, and
deluge flow rate.
(4) To compare measured performance
to performance predicted by ana-
lytical models developed at PNL to
verify and help define those
models.
(5) To determine the physical operat-
ing limits of the deluged surface,
particularly the limits of air flow
and deluge flow such that a wet-
ted surface is maintained.
Development and Evaluation of
the Deluge Heat Transfer Model
An important part of earlier and current
test programs has been the development
and testing of an approximate analytical
model (deluge model) for predicting heat
transfer from wetted surfaces. This
model has been very useful in planning
the test program, in reducing and corre-
lating data, and in interpreting the final
results. The model predicted the qualita-
tive aspects of the HOTERV tests very
well. However, because of incomplete
development and inadequate means for
computing two critical parameters in the
model, the predicted heat transfer rates
determined in earlier tests were generally
20-30 percent higher than measure-
ments.
-------
Recent advances in the deluge model
have improved the predictive accuracy
for calculating heat transfer from a wet
surface. In addition, an extension of the
model allows prediction of the rate of
evaporation of deluge water and the re-
sultant air outlet conditions.
A detailed development of the deluge
model that incorporates all of the recent
simplifications and refinements is given
in Appendix A of the full report. A brief
outline of the principal steps in the devel-
opment and a summary of the results are
given here.
TVre Surface Heat/Mass Flux
Analogy
The analysis of the heat and mass
transfer from an element of wetted sur-
face is based on the control volume
shown in Figure 1. Equations 1, 2, and 3
in Table 1 are from the energy and mass
balances for the control volume for dry
and wet operations. For conditions
where the assumption of heat and mass
transfer similarity is valid, Le = 1, where
the convective Lewis number (Le) is de-
fined by
"s
Le=Tc (51)*
Equations 3 and 4 are approximated by
Equations 5 and 6. Additional assump-
tions and approximations employed in
obtaining these results are discussed in
Appendix A of the full report.
Equations 1, 5, and 6 are analogous:
each contains the same heat transfer co-
efficient, hs. The important difference is
that the equations for transport of heat
and mass from a wet surface are written
in terms of the enthalpy difference and
humidity difference instead of the tem-
perature difference. The significance of
these formulations is that dry surface
heat transfer data can be usd to compute
wet heat and mass transfer performance
by merely changing the form of the driv-
ing potential employed. This is the basis
upon which the deluge heat/mass trans-
fer model is formulated.
The mass and energy balances for the
air stream are given in Equations 7, 8,
and 9. When combined with Equations
1, 5, and 6, these yield analogous differ-
ential equations for the distribution of
temperature on a dry surface (Equation
10), and the distributions of enthalpy
Solid Deluge water
surface films
-^Equations 1 through 50 are given in Tables 1, 2,
and 3. Symbols are defined at the end of this sum-
mary.
cMy = dA,
Adiabatic
material
surface
= dA,
A. Control volume, general boundary layer
Fins
Y////////////////////A
V////////////////////X
ft \t.\
B. Control volume, finned surface
Figure 1. Illustration of control volume used for heat/mass balance.
3
-------
Table 1. Summary of Equations for the Surface Mass/Energy Balance
dQ
dAB
= hs(Ts - T J
<1 > TT- = hs + ^"sIHs - HJ
dA.
(2)
dAs
(3
dQ hs
dA ~C
(i' - ico)
(5)
dm
s
dAe
(6
dQ
dAf
= GaCaTa
(7)
dAf
(S)
(9
dT0
(Ts-Tco) = VGaCac5/d
hsD
(11) ,H
dHoo / hsD \
'-HJ = V GaCad/dx M2
and humidity on a wet surface (Equa-
tions 11 and 12). In principle. Equations
10,11, and 12 may be integrated if the
relevant variables can be given as func-
tions of the dimensionless distance x-
However, the information required to
perform the necessary integrations
would seldom if ever be available except
for the simplest heat exchanger configu-
rations (i.e., aflat plate).
Extension to Finned Surfaces
The surface heat/mass transfer ana-
lysis has been extended to the treatment
of heat transfer from finned surfaces
with introduction of the overall heat/
mass transfer coefficients. The results of
the analysis are summarized in Table 2.
Equations 13,14, and 15 are the analo-
gous equations for heat and mass trans-
fer based on overall coefficients and
overall driving potentials from the
primary (tube) side to the free stream
(air-side) conditions. The analogous ex-
pressions assumed for U0, U
-------
Table 2. Summary of Model Development for Extended Surface
dO
dA.
UO
-------
Table 3. Summary of Model Equations for a Deluged Condenser in Cross Flow
«w
Tpi - Tool
00)
(32)
(33)
= e-N*x
(34)
= e-Nmx
(35)
N =
(36) N«
(37)
maCa
(38)
Q0= m
U0AsAT1r
(39) Q0=ma+*(
(42) = U£AS'
(40)
(43)
m=
= Z*AsAH1m
(41)
(44)
AT1m =
(45) 4* = 1 -e-N*
.. _ (iP2 - ioo2>
1n
TP2-
TP1 -
(46)
1n
ip2-
(47)
AH
1m
1n
HP2
terms of the inlet conditions without the
need of the outlet properties.
Equations 13,14, and 1 5 can also be
integrated to obtain Equations 42, 43,
and 44, the alternate equivalent expres-
sions for the heat and mass transfer. The
disadvantage of this approach is that
both inlet and outlet conditions appear in
the log mean property differences de-
fined by Equations 48, 49, and 50.
Thus, the use of these latter equations
generally requires an iterative solution
technique.
Evaluation of £, hd, and £/n
To apply the deluge model for predic-
tion of heat/mass transfer, values of £
and hd must be specified. An explicit ex-
pression (Equation 1 9) has been derived
for £. However, precise evaluation of £
requires knowledge of the fin root tem-
perature, Tr, which varies with core de-
sign and operating conditions in a manner
that is not easily predicted.
Figure 3 illustrates temperature pro-
files in a simple geometry that shows
how Tr can vary. The characteristics of
the three profiles are:
1. (Tco
-------
transfer in terms of the overall enthalpy
difference. If the model is to be internally
consistent, values of hd and £ must be
used that simultaneously satisfy all three
of these expressions. (In fact, only two
of these expressions are independent,
since, for example, the first two can be
used to obtain the third.)
For a given set of operating conditions
defined by values of hp, hs, ma, Tp, Too,
HOD, and md and the corresponding mea-
sured value of Q for that condition, any
two of the above equations constitute a
set of two simultaneous equations with
two unknowns. The solution of these
equations will yield unique values of Tr
and hd for each data set. Because of the
complex interrelationships of the varia-
bles in these equations, an iterative solu-
tion such as that described in Appendix
A of the full report is required.
The values of Tr and hd thus obtained
are empirical results determined from ex-
perimental data according to an assumed
heat transfer formulation in the deluge
model. The values of hd that result are
empirical in exactly the same sense that
hs and hp are. Furthermore, all of these
heat transfer coefficients are "lumped"
parameters in that they account for non-
uniformities in geometry, flow rate, and
surface conditions in some average way
that cannot be precisely defined. The only
difference in hd is that the unknowns
lumped into this parameter are some-
what greater because of the additional
effect of nonuniform wetting.
The procedure used to derive hd values
from the data also yields corresponding
values of Tr (and thereby of {) for each
data set. Although the fin root tempera-
ture, Tr, has a physical interpretation, as
illustrated in Figure 3, the values of Tr
extracted from the data are only approxi-
mately related to any actual tempera-
tures in the heat exchanger. However,
the same would be true of root tempera-
tures calculated for a dry operation of
the same system since the same types
of assumptions and approximations are
involved.
Description of Experimental
System
All testing took place in the Water
Augmentation Test Apparatus (WATA),
an experimental test facility shown in
Figure 4. The WATA consists of three
fluid loops: the air loop, circulation water
loop, and augmentation water loop.
These loops come together in the heat
exchanger test section.
The air loop is an open-ended single-
pass loop providing uniform air flow
through the test section at a desired
temperature and humidity and at ap-
proach velocities from 3 to 16 ft/see.
Outside air is brought in through a centri-
fugal blower whose output is variable
from 2100 to 12000 cfm. After leaving
the blower, the air passes through a
steam heating unit and then through a
steam humidification section to provide
inlet air at the desired wet and dry bulb
Tube Deluge
Wall Water
flow
Figure 3. Simplified schematic of
temperature profiles that
may exist for deluged heat
exchanger operation.
Steam humidifier
Inlet flow Blower H°**r_
damper
temperatures. The air then flows through
a restricted mixing section before pass-
ing through a vaned expansion section
with a 2-ft x 6-ft outlet. A screen pack
at the expansion section outlet helps
maintain flow uniformity. The air then
passes through a vaned 2-ft x 6-ft 90 °
elbow, and another screen pack, and
then through a 4-ft approach section of
the same cross section as the 2-ft x 6-ft
test core.
From the test core section the air flows
through a 3-ft section of 2-ft x 6-ft
duct, through a contraction, through a
flexible duct, and then into an 18-in.
diameter, 20-ft long section of straight
duct before being exhausted to the out-
side. The straight section is equipped
with an Annubar flow sensor to measure
the air mass flow rate through the test
section.
The air loop permits flexibility in core
orientation and airflow direction. Figure
5 illustrates the means provided to vary
the core orientation.
The circulation loop provides the heat
to be rejected by the test core. A centri-
fugal pump capable of up to 365 gpm
flow pulls water from a 400-gal. storage
tank. Part of the flow is passed through
two SCR-controlled electric circulation
heaters providing a total of 135 kW of
heat. The heated water is then mixed
with the remainder of the circulation
water flow and fed to the test core inlet
Test
section
Exhaust
Storage tank
Circulating water loop
Storage/weigh tank -f immersion heater
Augmentation water loop
Figure 4. Schematic of water augmentation test apparatus.
7
-------
manifold. After being cooled in the test
core, the circulation water returns to the
storage tank and is ready for recirculation.
The augmentation loop is used for
evaluating deluged heat exchangers for
integrated dry/wet towers. A centrifugal
pump with a 25-gpm (maximum) capa-
city draws water from a 40-gal. weigh
tank and pumps it to the deluge injection
point at the top of the deluged test core.
After the-deluge water passes over the
air-side surfaces of the core, it is collect-
ed in a catch basin at the base of the test
core. A second pump then returns the
deluge water to the weigh tank. Water
may be added to the weigh tank from a
deluge storage tank when the water in
the weigh tank has been depleted by
evaporation on the test core.
The three loops come together in the
test core. The test core section consists
of a 6-ft high x 2-ft wide x 1 -ft deep
duct section surrounding the specific
heat exchanger core being tested.
Experimental Results
Tests Performed
Prototype tests were performed to in-
vestigate the dependence of heat trans-
fer and pressure drop on several inde-
pendent parameters:
Inlet temperature difference (ITD).
Air-side inlet relative humidity ().
Air-side mass flux (G0).
Deluge flow rate (md).
Core angle (9C).
Dry Heat Transfer Results
Heat transfer tests without deluge
water were done for air velocities rang-
ing from 3 to 15 ft/sec. The results are
shown in Figure 6 in terms of a surface
heat transfer coefficient with and with-
out efficiency (h0 and hs, respectively)
and an overall heat transfer coefficient,
U0. The tube-side convective resistance
and tube wall resistance were subtracted
from the overall resistance (1/U0) to ob-
tain the effective surface resistance, h0.
The fin efficiency model described
earlier was then used in obtaining the dry
surface coefficient, hs. The solid line and
error bar for each variable show a best fit
to the data and the estimated experi-
mental uncertainty. Dry surface tests
run before and, after the wet surface
tests fall within the estimated uncertain-
ty (10 percent) indicating that no appre-
ciable aging or scaling of the surface oc-
curred in the test period.
To outlet duct
& flow sensor
Core exhaust section
Core
Approach section with
deluge trap
Screen pack
|-«- Elbow with vanes
Deluge trap section
Screen pack
Diffuser section with vanes
Transition
Humidifier
section
Heater
Air from
blower
Figure 5. WA TA air ducting upward air flow at any angle up to 45°.
u.
o
I
o
!£
0)
o
c
£
1000
2000
Go fib/hr- ft2)
3000
Figure 6. Plots of U0, h0, andh» versus mass velocity. Go.
8
-------
Wet Heat Transfer Results
Wet mode heat transfer depends on
heat exchanger temperature, air temper-
ature, and air humidity. The parameter
which best correlates this heat transfer
with these meteorological conditions is
the dimensionless driving force, l~, de-
fined by
C.ITP-TJ,
(56)
The heat transfer per unit inlet temper-
ature difference, Q/ITp-TJ, shown in
Figure 7 is for relative humidity values
ranging roughly from 25 to 75 percent
and inlet temperature differences from
1 0 to 50 °F, all at a frontal air velocity,
V0, of 4.5 ft/sec. Single points are also
shown for V0 = 3 ft/sec and 6 ft/sec.
The solid line in each case is the predicted
correlation when an empirically deter-
mined value of the deluge film coeffi-
cient, hj, is used in the heat transfer
model.
To make these predictions from theory,
it is necessary to assume values for both
£ and Hd*. The value of f does not change
substantially with any of the independent
parameters except Tp. Since all of the
tests made on the Trane core were for
the same Tp, a constant value of 9.5 is
used for £ in all of the theoretical calcula-
tions. For the predictions in Figure 7, hj
values are taken from other figures. These
values were: h^ = 18 Btu/hr-ft2- °F at
V0 = 3.0 ft/sec, h£ = 26 at V0 = 4.5,
andh£= 24atV0 = 6.0.
Excellent agreement of the data with
theory is indicated in Figure 7. This good
agreement is not coincidental since the
h,J values were obtained from the same
experiments. The agreement does sub-
stantiate the validity of the model. Un-
certainties in the predicted values of hd
are very difficult to quantify but are rea-
sonably large.
One of the most important parameters
used for characterizing the performance
of a deluged heat exchanger is the ratio
of wet-to-dry heat transfer, Qwet/Qdry
To best evaluate a real operating condi-
tion, the comparison is made for the
same core temperature, the same inlet
air conditions, and the same air-side
pressure drops.
Qwet/Qdry data for the Trane core and
the corresponding predictions are given
in Figure 8. The data correlate very well
with f, and the prediction is in excellent
agreement with the data.
Additional results of Qwet/Qdry for the
Trane core show little dependence of
Qwet/Qdry °n the air mass flux, G0. The
predicted dependence of Qwet/Qdry shows
a very slight reduction in enhancement
at higher air flows, but the effect is well
within the expected uncertainty.
Test data show the dependence of
dry on the mass flow rate of deluge
water, md, to be slight. The predicted
values for Qwet/Qdry show a slight maxi-
mum at md = 4 gpm which is consistent
with apparent test data; however, the
enhancement appears to be essentially
independent of md considering the ex-
pected uncertainty in the predictions.
18
16
14
12
10
* 3.0 Tp = 120°F
o 4.5 oc = 25°
D 6.0 ma = 3 gpm
Predicted Values
V0 = 3.0 fps Go = 750
10 11 12 13 14 15 16
r
Figure 7. Normalized heat transfer versus T.
= 120°F
V0 = 4.5 fps
= 3.0 gpm
ec = 25°
i i i i i i
7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0
r
Figure 8. Dependence of enhancement ratio on ratio of inlet driving potentials, T.
-------
Wet Mass Transfer Results
The model developed for the rate of
evaporation of water from the surface of
the heat exchanger to the air can be de-
scribed in terms of an overall mass trans-
fer coefficient, ££,, that is analogous to
the overall heat transfer coefficient, U£.
Although the experiments were not de-
signed to obtain an accurate measure-
ment of the deluge water evaporation
rate, evaporation rate was estimated
from the measured difference in the air
moisture content across the core.
Because of the extreme data scatter,
no detectable trend of ££, with I" was ob-
served. The predicted values of Z£, ap-
pear to increase slightly with increasing
T, but the effect is small. Ignoring some
of the anomalously low values of ££,, the
deluge model appears to overpradict the
data by about 20-30 percent. It is not
apparent to what extent the fault is in
the data or in the model.
£, is essentially equal to £ at values of
T less than 10 but tends to values less
than £ at higher values of I". However, it
is reasonable to assume £, = £ except
for conditions where the driving poten-
tial for evaporation and heat transfer is
very high.
From the limited data available, the
deluge model appears to overpredict the
rate of evaporation. However, because
of the large uncertainty in the measure-
ments, more and better data are required
before any definite judgments should be
made.
An additional parameter of interest in
evaluating the performance of a deluged
heat exchanger is the fraction of heat
transfer that is attributable to evaporation.
Denoting Q0 at the total heat flux and Qv
as that due to evaporation (the latent
heat component), the ratio QV/Q0 may
be calculated for each operating condi-
tion using:
Q ~
Qo
(57)
Figure 9 shows that the data correlate
quite well with I" and that the prediction
is in good agreement with the data. Fur-
ther, the proportion of the total heat flux
attributable to evaporation increases at
high T (i.e., low humidity, high air flow
rate, low ITD). For values of I" > 20, Qv/
Q0 > 1 for the conditions shown in Figure
9. For these values of f, the air is actually
cooled by evaporation and the sensible
heat flux to the air is negative. None of
the present experiments achieved in this
condition; however, some of the earlier
tests resulted in core outlet air tempera-
tures below the inlet conditions. These
results are relevant to optimizing the
operating conditions to get the maximum
cooling value from the water used.
Conclusions
Conclusions that may be drawn from
the results of this project relate to three
principal areas:
Operating characteristics and po-
tential benefits of the deluge con-
cept for cooling electric power
plants.
Comparisons of the dry/wet perfor-
mance of the two types of plate-fin
heat exchangers that have been
tested.
The applicability and accuracy of
the deluge heat/mass transfer
model.
Characteristics and Benefits of
Deluge Cooling
The notable operating characteristics
observed in the experimental study of
deluge cooling, as compared with dry
cooling, may be summarized as:
The primary parameter used in this
study to characterize the perfor-
mance of a dry/wet cooling system
was the ratio of wet to dry heat
transfer at the same operating con-
ditions and the same air-side pres-
sure drop. This parameter was de-
termined to vary between 2 and 5
for conditions tested in this study.
The size of a dry-cooled system
needed to meet heat rejection re-
quirements at peak ambient temper-
atures could thus be reduced by a
factor of Vi to % by the use of deluge
cooling enhancement. The actual
reduction in size would depend on
the system design and operating
Qv
Qo
1.0
0.9
0.8
0.7
0.6
0.5
Deluge Model
2O°F < ITD < 60°F
0<
rp= 120°F
I
20
0 5 10 15
r
Figure 9. Comparison of predicted and experimental values of QV/Q,
25
10
-------
conditions. In particular it would de-
pend on the amount of water avail-
able for evaporative cooling.
Since water would be used only dur-
ing periods of peak cooling demand,
the water consumption of the
deluged system could be substan-
tially less than in a wet tower of
similar capacity.
The increase in heat transfer due to
deluge must be compared to dry
heat transfer at the same pressure
drop. At a fixed air flow rate, delug-
ing was accompanied by a substan-
tial increase in the air-side pressure
drop. Both the heat transfer and
pressure drop increased with in-
creased air velocity or deluge water
flow rate.
In the anticipated dry/wet opera-
tion, a variable number of heat ex-
changer modules will be deluged op-
erating in parallel with the remainder
of the modules dry. Therefore, all
modules will operate at the same
air-side pressure drop which, for a
given deluge flow rate, will deter-
mine the air flow rate in both wet
and dry sections.
At superficial air velocities greater
than about 6-8 ft/sec, many water
droplets were blown from the back
side of the heat exchanger. Droplet
drift may thus impose an upper
bound on the air flow rate when the
system is being deluged.
The heat rejection rate during deluge
operation was found to be dramatic-
ally dependent on ambient air condi-
tions. The enhancement was great-
est for low inlet temperature differ-
ence (ITD) and low humidity (i.e.,
O-wet/0-dry < 5 at ITD ~ 20 °F, 25
percent RH) and lowest at high ITD
and high humidity (i.e., Qwetd/Qdry
< 2 at ITD ~ 50 °F, 75 percent RH).
Heat transfer enhancement using
deluge is most effective and thus
most attractive where the need is
greatest: in hot dry regions where
water is scarce as in most of the
western U.S.
Deluge cooling is also likely to be at-
tractive in humid regions where the
availability of fresh water for cool-
ing is limited. In all cases, a system
design optimization will have to be
performed for specific sites to eval-
uate the merit of deluge cooling rela-
tive to more conventional cooling
systems.
Dry/Wet Performance
Comparison
From the present and preceding tests,
dry performance data were obtained for
several heat exchanger configurations
that may be compared with the present
dry performance results. In addition,
deluge tests were performed on a plate-
fin heat exchanger of substantially dif-
ferent design. Comparison of these per-
formance data revealed:
For dry performance, the principal
basis of comparison was the heat
transfer per unit ITD and per unit
volume as a function of fan power.
On this basis, the chipped fin Curtiss-
Wright (C-W) design selected for
the ACT facility performed the best
at all fan powers.
The Trane wavy fin design selected
for ACT and a design based on a
five-tube bundle of wrapped helical
fin tubes were next in performance
at about 10 percent lower overall
rating than the top C-W system. The
performances of the Trane and heli-
cal fin designs were essentially the
same.
Comparisons were also made with
two other C-W chipped fin assem-
blies and with a HOTERV perforated
plate fin assembly. All three of these
performed below the preceding three
at all fan powers. The HOTERV per-
formed better than the two C-W as-
semblies at low fan power but sub-
stantially lower than all of the other
assemblies at high fan power.
For wet operation, the primary per-
formance comparison was based on
the ratio of wet to dry heat transfer
rates at equal air-side pressure drop
and equal air inlet superficial veloci-
ties as a function of inlet conditions.
On this basis, the Trane core consis-
tently outperformed the HOTERV
core by a ratio of about 1.2 at com-
parable conditions. The principal
reason for this difference was the
higher pressure drop of the HOTERV
core at the given conditions.
Evaluation of the Deluge Model
A primary objective of this work was
to continue to develop and evaluate an
analytical model for predicting the heat
transfer from a deluged heat exchanger.
This was successfully accomplished,
and the model was also extended to
allow prediction of the rate of evapora-
tion and, thereby, the outlet conditions
of the air passing through the system.
The principal application of the deluge
heat transfer model was to develop cor-
relations used in reducing and presenting
the experimental data. The primary
quantity derived empirically from the
data was the effective deluge film con-
vective coefficient, h£. When experi-
mentally based values of hd were used in
the model equations, the predicted cor-
relations were in excellent agreement
with the data for a large range of operat-
ing conditions. The present experiments
have thus shown that, given suitable
values for hd, the deluge model based on
the enthalpy difference driving potential
will serve as an accurate model for pre-
dicting wet performance of a finned, air-
cooled heat exchanger.
The present study obtained empirical
results for hd as a function of operating
conditions that may be used to predict
the performance of the Trane core. Fur-
thermore, these results for hd are quite
similar to the previous results obtained
for hd for the HOTERV design, which dif-
fered significantly in design and perfor-
mance from the Trane core. Thus, for de-
sign purposes, it is probably safe to use
either of these results for hd for a plate-
fin design similar to, but different from,
either of the above. For a radically differ-
ent design such as a bundle of cylindrical
finned tubes, these results might also suf-
fice for an estimate of performance using
the deluge model. However, the validity
of this approximation cannot be verified
at this time.
The mass transfer extension of the
deluge model could not be extensively
evaluated in this study because accurate
independent measurements of the
deluge water evaporation rates were not
obtained. However, from the approxi-
mate measurements obtained, it ap-
peared that the model correctly predicted
trends, but the rate of evaporation was
overpredicted by about 20 percent. This
result is highly tentative, and additional
measurements are required before a
more definitive assessment can be made
of this aspect of the model.
11
-------
English to Metric Conversion
Table
To Convert from To Multiply by
atm
Btu/hr
Btu/(lbm - °F)
Btu/(hr-ft-°F)
Btu/lhr-ft*- °F)
Btu/lbm
ft
ftz
ft*
ft/sec
ft^/sec
ft3/lb
°F
gal./min.
in.
in. H20
in. Hg
Ib/hr
Ib/ft3
Ib/(hr-ft2)
Pa 1.013E+05
W 0.2929
J/(kg-K) 4184.0
W/(m-K) 0.0120
W/(m2-K) 5.6745
J/kg 2324.4
m 0.3048
m2 9. 290 E -02
m3 2.832E-02
m/8 0.3048
m2/s 9.290 E -02
m3/kg 6. 243 E -02
K TK = n>
459.67)71.8
m3/s 6. 3090 E -05
m 2. 540 E -02
Pa 249.15
Pa 3386.4
kg/s 1.260E-04
kg/m3 16.018
kg/(s-m2) 1.356E-03
Nomenclature
A,
A.
A.,
A«
a,*
at
Bit
Bif
Bi
D
f
Ga
GO
9d
HP
H..
frontal area
total air-side surface area
air-side fin surface area
air-side tube surface area
relative surface area (wet-
mass transfer)
primary side relative area
relative surf ace area
relative surface area (wet-
heat transfer)
relative tube area
Biot number of fin, dry
Biot number of fin, wet
Biot number of fin for mass
transfer
moist air specific heat
diameter or characteristic
length
Fanning friction factor or
function (Equations 24, 25,
and 26)
mass flux of air at the
minimum cross section
mass flux (or velocity) of
free stream air
distance between wetted
fins
humidity ratio of saturated
air at Tp
humidity ratio of un-
saturated air at T,
humidity ratio of free stream
air
log mean humidity ratio dif-
ference
heat transfer coefficient
hd deluge heat transfer coeffi-
cient
h£ effective deluge heat
transfer coefficient = hda£
h0 surface heat transfer coeffi-
cient including fin effective-
ness
hp primary side heat transfer
coefficient
h, surface heat transfer
coefficient
ITD inlet temperature difference,
TP-T«
i«> enthalpy of moist air at To,
ip enthalpy of saturated air at
~P
ir' enthalpy of saturated air at
Tr
i, enthalpy of saturated air at
T.
ii enthalpy of saturated air at
T»
Ai|m log mean enthalpy dif-
ference
k thermal conductivity of tube
wall or fin
If effective circular fin length
ma mass flow rate of air (dry)
md mass flow rate of deluge
water
mw, m0 mass flow rate of
evaporated water
N number of transfer units
(NTU) for dry heat transfer
N * - NTU rating for wet heat
transfer
Nm NTU rating for wet mass
transfer
Pf fin pitch
Q heat transfer
Qdry total heat transferred under
dry operation
QO net rate of heat transfer in
deluge operation
Qy heat flux attributable only to
evaporation
total neat transfer from
primary side to air side dur-
ing wet operation
rb outer tube radius
re outer fin radius
r0 equivalent radius or outer
tube radius
r\ inner tube radius
r0 equivalent radius or outer
tube radius
Tp primary fluid temperature
Tr fin root temperature
T8 surface (air/water interface)
temperature
Ta free stream air temperature
tj deluge water film thickness
wet
A,
X
tf fin thickness
t,, t tube wall thickness
Tim l°9 mean temperature dif-
ference
U0 overall dry heat transfer
coefficient
UQ overall wet heat transfer
coefficient
U£, equivalent coefficient for
mass transfer = C,Z£
V0 frontal velocity
yb half fin thickness
x,y,z, coordinate directions
Greek Letters
r ratio of inlet driving poten-
tials for heat transfer
6 ratio of im Jm/tf also boun-
dary layer thickness
fy dry fin efficiency
wet fin efficiency
fin efficiency for mass
transfer
core angle from vertical
dimensionless temperature
at dimensionless x coor-
dinate, x
dimensionless enthalpy at
dimensionless x coordinate,
X
dimensionless humidity ratio
at dimensionless x coor-
dinate, x
transformation parameter
for heat transfer
transformation parameter
for mass transfer
latent heat of vaporization
of water
dimensionless x coordinate
overall mass transfer coeffi-
cient
surface mass transfer coef-
ficient
relative humidity
relative humidity of ambient
air at Tn
dry exchanger effectiveness
wet exchanger effec-
tiveness for heat transfer
wet exchanger effec-
tiveness for mass transfer
>)m
0C
Q(X>
0*
-------
S. G. Hauser, D. K. Kreid, antJB. M. Johnson are with Battelle/Pacific Northwest
Laboratory. Richland. WA 99352.
Theodore G. Brna is the EPA Project Officer (see below).
The complete report, entitled "Dry/Wet Performance of a Plate-Fin Air-Cooled
Heat Exchanger with Continuous Corrugated Fins," (Order No. PB 82-231
424; Cost: $16.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
oUSGPO: 1982 559-092/0445
-------
n
o
5'
0)
O
ro
a>
oo
o »*
XOOfT'O
O 220
OO
ru
3D »-«
n-i c
nn at
------- |
|