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

-------