-
HEAT TRANSFER PERFORMANCE OF A DRY AND WET/DRY ADVANCED COOLING
TOWER CONDENSER
Hans D. Fricke, David J. Webster, Kenneth McIlroy Union carbide
Corporation - Linde Division, Tonawanda, New York
John A. Bartz Electric Power Research Institute, Palo Alto,
california
ABSTRACT
This paper describes an EPRI-funded experimental evaluation of
advanced air-cooled ammonia condensers for a phase. change dry/wet
cooling system for power plants. 'Two condenser surfaces with
dIfferent air-side augmentation were tested in an ammonia phase
change pilot plant (0.6 MWth) located at UCC/Unde. The first unit
consisted of an integral shaved-fin-extruded aluminum tubing
designed fordry operation. Heat transfer and air-side pressure loss
characteristics were measured under varying air face velocities (1
to 5 m/s) and initial temperature differences, lTD (11 to 33K).
Measured overall heat transfer coefficients, U, ranged between 40
and 49 J /m2 s.K (based on air-side surface). The second
configuration constituted an aluminum plate-fin/tube as sembly,
which was tested in both dry and wet (water deluge) modes at 1 t04
m/s air face velocities and lTD's of 5 to 33K. Deluge rates varied
from 1 to 6 m3/s per meter of core width. In the dry mode, U ranged
from 42 to 63 J/m2 .s.K. Water deluge enhanced the heat rej ection
up to 4.5 times over dry operation.
INTRODUCTION
For over a decade, considerable interest has existed in dry
(nonevaporative) cooling to reject waste heat from steam-electric
power stations. Dry cooling offers the elimination of thermal
pollution of lakes and streams caused by once-through cooling, as
well as a methodio avoid water makeup, blowdown, and fogging
problems associated with wet (evaporativel cooling. In addition,
dry cooling increases siting flexibility, particularly for
locations in arid Western coal fields.
However, dry cooling requires considerable capital investment
for the cooling towers. Hence, the development of effitient (low
cost) heat transfer surfaces in conjunction with neW cooling
processes is very important for this approach to be economica lly
feasible. Extensive studies have indicated that although heat
exchangers contribute about 35% of the cost, the entire heat
transport system, not merely the heat exchangers, need to be
evaluated to achieve an attractive alternative, Ref. (1). ThiS must
include acceptable working fluids to transfer the heat from the
steam condenser to the ambient
414
air, effective methods to transfer heat and methods of achieving
augmented cooling during hot dry weather.
Beginning in 1975 the Electric Power Research Institute (EPR!)
has sponsored a program of dry cooling research proposed by Union
Carbide Corporation/ Linde Division (UCC/Unde) and Westinghouse
Electric Corporation. The primary object of this effort was to
study, select and demonstrate a heat-rejection cycle of
substantially increased thermal performance to reduce the large
penalties commonly associated with conventional dry cooling where
circulating water (the coolant) is heated in the steam condenser
and, subsequently cooled in the cooling tower. The final
recommended concept consists of an isothermal ammonia-phase-change
heat-rejection cycle with heat transfer enhancements on the heat
exchanger surfaces in the loop. The waste heat from the power plant
is rejected to liquid ammonia in a steam condenser/ammonia reboiler
and the ammonia vapor thus generated is condensed in a cooling
tower by heat rejection to the atmosphere.
In order to demonstrate the technical feasibility of ammonia
phase change for dry cooling, a test pilot plant (0.6 MWth) was
designed, built and operated by UCC/Unde under EPRI contract.
Primary obj ectives were to demonstrate the integrated performance
of this cooling concept using UCC/Unde I s heat transfer
enhancements in the steam condenser/ ammonia reboiler and to test
the performance of advanced air-cooled ammonia condensers in the
pilot plant's cooling tower. This paper will address only the
experimental studies relating to the ammonia condenser. Other
efforts of this program are described in Ref. (2). Specifically,
the heat transfer and pressure loss characteristics of two ammonia
condenser configurations 1 with differing air-side augmentation are
discussed: the Curtiss-Wright integral shaved fin-extruded aluminum
tubing, designed for dry operation, and the Hoterv aluminum
platefin-tube condenser which was tested in the dry and wet (water
deluged) mode. Water deluge entails wetting the air-side surface of
the condenser to improve the heat transfer performance of dry
cooling
1 of approximately equal projected cost for a 500 MWe power
station cooling system.
ESL-IE-81-04-71
Proceedings from the Third Industrial Energy Technology
Conference Houston, TX, April 26-29, 1981
-
towers during periods of high ambient air temperature.
Augmentation of the conve ctive and conductive tube-to-air heat
transfer is achieved by the evaporative cooling of the water as it
flows over the external surface. This technique is expected to be
an economical peak shaving approach with minimal water
consumption.
TEST APPARATUS
An ammonia phase change dry/wet cooling pilot plant was built
and tested for demonstrating the concept and for measuring the heat
transfer performance of augmented heat exchangers. A simpli~ fied
schematic is shown in Figure 1.
enters the phase separator. The ammonia vapor from the separator
is then condensed in the air-: cooled condenser from which liquid
is pumped back to the surge tank. Steam is condensed in the
condenser/reboiler on the outside of the tubes which also
incorporate a UCC/Linde condensing enhartcement. The condensate is
collected in the hotw~ll and returned to the steam generator to be
recyded.
The system was designed for a nominal heat load of 0.6 MWth when
condensing steam at 1.7 x 104 Pa absolute pressure (Tsat = 330 K)
with an: ambient air temperature up to 307 K. At design :condition,
this would translate to condensing steam at the rate of 908 kg/hr,
boiling and condensing 2043 kg/hr of ammonia at 2.1 x 106 pa (Tsat
= 327 K) with an air flow rate of 127,440 m3/hr through the ammonia
condenser.
STEAM (FROM STEAM
GENERATOR)
__.... TO ] HOTERV CONDENSER TEST SECTION
___ FROM (SEE FIG.3)
AMMONIA VAPOR
L10UID AMMONIA
AMMONIA CONDENSATE
AMMONIA AMMONIA RECIRCULATION PUMP PUMP
FIG. I - Advanced dry cooling pilot plont flow schematic
There are two primary flow loops: the anhydrous ammonia
isothermal phase change circuit which transports the heat from the
steam condenser to the ammonia air-cooled condenser in the cooling
tower and the steam/condensate loop which provides the source of
low pressure steam to the condenser from a commercia I steam
generator.
Basically, the liquid ammonia is pumped from the surge and phase
separator tanks into the steam condenser/ammonia reboiler where it
boils inside High Flux2 tubes, extracting heat of condensation from
the shell-side condensing steam. Twophase ammonia exits the
condenser/reboiler and
2 Dee/Linde patented porous boiling surface which greatly
enhances heat transfer by typically 5-10 times that of a plain
surface.
Two separate ammonia vapor loops were u;tilized for testing the
ammonia condensers. One: loop (indicated in Figure 1) contained a
cooling tower, housing two Curtiss-Wright tubed condensers which
were tested strictly in the dry mode. The configuration of this
particular condenser is not sUitabl for a wet operation since
water-deluge cannot easily be supplied to each horizontal,
shaved-fin tube. 3 : The heat exchangers consisted of aluminum
shaved-fin tubing (developed by Curtiss-Wright) which was
manifolded horizontally between two headers (a vapor and liquid
channell. The frontal area was 5.95 m2 with a core depth of 0.12 m.
The air-tp
3 Changing the design/orientation of the shavedfin tube
assembly, however, could conceivably accommodate water-deluge.
415
ESL-IE-81-04-71
Proceedings from the Third Industrial Energy Technology
Conference Houston, TX, April 26-29, 1981
-
condensing side area ratio was 7.2. Two such units were
installed at opposite sides of the cooling tower cabinet. A
photograph of this extended surface tubing and geometry is shown in
Figure 2. Noteworthy are the individual rectangular ammonia flow
channels, which assure strength and promote turbulence for improved
heat transfer. As noted in Figure 2, air flow is horizontally acros
s the fins. A variable-pitch fan mounted on top of the cabinet
controls the induced ambient air flow across the heat exchanger
cores. The condenser was tested under air face velocities varying
from 1 to 5 mls and ITO's from 11 to 33 K.
FIG. 2 Curtiss-Wright shoved-- fin lublnQ
The second loop consisted of a separate vapor line, also
originating from the top of the phase separator, an air duct with a
variable speed blower, and a water deluge system for testing the
Hoterv condenser in the wet mode. Figure 3 presents a more detailed
schematic of this branch circuit. Saturated vapor flows from the
pilot-plant phase separator to the Hoterv test unit where it is
condensed. The liquid is then returned to the pilot-plant surge
tank. The ammonia pressure in the loop (and in the pilot plant) can
be varied from 1.03 x 106 to 2.06 x 106
pa in order to control the ITO. 4 An ammonia pump was not
required since the condenser is elevated above the surge tank,
providing sufficient hydrostatic head.
A rectangular duct with the blower was utilized to supply room
temperature air to the test unit. Screens and baffles were
installed inside the duct, upstream of the heat exchanger, to
assure uniform air flow across the face of the unit. The centifugal
fan provided air flow velocities between 0.6 and 4 mls at the inlet
of the heat exchanger. A traversing pitot-static probe connected to
a manometer and a turbine meter monitored the air approa ch ve
locity
A water deluge system was installed above the Hoterv condenser
for testing in the wet mode and was designed to uniformly wet the
heat exchanger up to 6 x 10-4 m3Is per meter of heat exchanger
width. A water softener and heater (not shown in Figure 3) were
employed in the loop upstream of the heat exchanger to condition
the augmentation water. The water softener was used to reduce
variations in the water chemistry and to minimize any corrosion or
deposition tendencies on the external heat exchanger surface during
the tests. Mineral deposition or scaling could cause an increase in
the air-side frictional pressure loss and in the thermal
resistance. The Hoterv condenser was inclined at a 16 angle with
respect to the vertical to assure uniform water coverage. This was
based upon water-deluge experiments which Battelle Pacific
Northwest Laboratories, Ref. (3) had conducted on a similar heat
exchanger. As noted in the figure thermocouples (in the ammonia,
water, and air stream), pressure transducers (absolute and
differential) , and turbine flow meters were provided to measure
the heat transfer and air-side pressure-loss characteristics of the
test condenser in the dry and wet mode.
The Hoterv condenser is an aluminum platefinltube configuration
which was developed and manufactured by Hoterv Institute, a
Hungarian organization. Babcock & Wilcox is now the licensee
and U S. distributor. This heat exchanger (0.46 m2 frontal area)
was nominally rated for a heat duty of 0.06 MWth and condenses
ammonia at a rate of 202 kglhr with an air velocity of 1.8 mls in
the wet mode. As noted in Figure 4, the assembly basica lly
consists of two rectangular headers, manifolding the staggered
horizontal tubing which pass through a closely spaced stack of
plate-fins. A Hoterv heat exchanger is typically fabricated using
aluminum tubes and fins. However I because of the high pressure
used in the tests, carbonsteel liners (1.46 cm I .0. x 1.70 cm
0.0.) were employed inside the aluminum tubes (l.8 cm 0.0.) to
accommodate the 2.4 x 106 Pa operating pressures. The
air-to-condensing side area ratio was 17.3.
4 Initial Temperature Difference = Ammonia Saturation
Temperature - Inlet Air Temperature.
416
ESL-IE-81-04-71
Proceedings from the Third Industrial Energy Technology
Conference Houston, TX, April 26-29, 1981
-
FI
l-d'\I'V,\PU''"1"TEST
AIR ~ ~ IN~ ~ .-v'\.I'\r"".I'-I=: VELOCITY
TURBINE METER
r---
I I I I I I I I I I I
~ AIR VELOCITY INDICATOR
CV
P 5 I E L P o A T R
AP T L o A R N T
DELUGE
DELUGE WATER FILTER SUPPLY
DELUGE WATER DISTRIBUTION HEADER
AIR ....
~~
CONDENSER
AMMONIA CONDENSATE
TE
WATER DRAIN
cv PILOT PLANT SURGE TANK
FIG.3 Pilot plant auxiliary
NH3 VAPOR I I i I I ~LET WATER DELUGE DIRECTION
! ~. ~ ~- -~-- 76 cm -------...1
1 I
I I
RECTANGULAR HEADER
(15cm)( 20cm; I 2cm WALL THICKNESS) :
I
LpLATE FINS (3.5 FINS/em)
+ NH3 CONDENSATE
OUTLET FIG.4 Hoterv ammonia condenser outline
As illustrated in Figure 5 the plate fins of the condenser
feature patterns of slots I and raised turbulators to enhance the
air-side heat transfer.
Condenser performance was measured for air face velocities
varying from 1 to 4 m/s, lTD's between 5 and 33 K and water deluge
changing from 1 to 6.2 m3/s per linear meter of condenser
width.
loop for wet / dry testing of ammonia condenser
6 TUBE ROWS IN DIRECTION OF AIR FLOW ~
0.05 cm TURBULATOR
---15.0cm--~-2.5 cm
~~~~~]ocm -1.8 cm 0.0.
PLATE FIN SLOT
FIG. 5 Hoterv condenser plate fin configuratIon
RESULTS AND DISCUSSION
As previously indicated, the shaved-fin Surface is used for an
all dry ammonia condensing tower 1 5 while the plate-fin/tube core
is designed
5 Possibly augmented with a separate wet tower
for peak-shaving (during high ambient air temperatures)
417
ESL-IE-81-04-71
Proceedings from the Third Industrial Energy Technology
Conference Houston, TX, April 26-29, 1981
-
such that the surface can be operated in the dry mode, or the
heat transfer performance can be enhanced by applying water deluge
when high ambient air temperatures occur. A parameter of key
interest to the designer is the power consumed by the tower fans
employed to draw the air through the heat exchanger core. Hence, in
Figure 6, the measured heat rejection capability of the two
surfaces is compared as a function of fan power.
90 DELUGE RATE HOTERV (3.1 X 104 m1s, m)
80 HOTERV (2.1 X 10-4 m)s m)
70
r
-
ri. 0 t-U (t z Q t !:J U D: \L CURTISS-WRIGHT CORE C) z Z .3
0~ z ~
-
o 10 20 30 40 50 60 70 80 90 100
AIR RELATIVE HUMIDITY
FIG. 10 Deluge enhancement vs. air relative tl.lmidity.
0WET/OORY at equal fan power. Air face velocity ...... 2.03 m/s
Curtiss-Wright surface as a function of air Reynolds N
number.
3 ....
E Water deluging the Hoterv ammonia condenser improves its heat
transfer performance manyfold.
... 200 - 1 r- -r I f DELUGERATE ::0, ITDRANGE~56-112'C (m3/s ml
I The greatest enhancement is achieved at low relaE R H RANGE: 40
-60 or. - 3,.0 It 10-4 tive humidities. The improvement increases
with ~ - : . +. higher deluge rates to a point where a further infZ
crease in deluge offers little or no improvement.w Q Fortunately
the augmentation offered by deluging is lJ
lJ greatest at low lTD's when it is needed most--duringW8 10.0
high ambient air temperatures. 0:: However, it must be emphasized
that even W lJ
-
ACKNOWLEDGEMENTS NOMENCLATURE
The program described in this paper was sponsored by the
Electric Power Research Institute (EPR!); Contract No. RP
422-2.
Special recognition is given to Messrs. C. F. Gottzmann, F.
Notaro and J. B. Wulf (UCC/Linde) , Mr. G. J. Silvestri
(Westinghouse), and Dr. R. W. Zeren and Dr. J. Maulbetsch (EPRI)
for their efforts in originating and developing the study of the
ammonia phase'-change dry-cooling concept.
The authors gratefully acknowledge the assistance of their many
colleagues in various facets of this particular effort; in
particular Messrs. M. H. Gallisdorfer, T. D. Craig, R. A.
McClellan, and Ms. C. E. McHale. The authors also wish to
acknowledge their colleagues at Battelle (PNL) for helpful
information exchange during the test program.
REFERENCES
1 McHale, C. E., etal "New Developments In Dry Cooling of Power
Plants," paper presented at the 41st American Power Conference,
Chicago, Ill, April, 1979.
2 Fricke, H. D., etal "Power Plant Waste Heat Rejection Using
Dry Cooling Towers", ucci Linde Interim Report For RP 422-2 EPRI
contract, Performance Period: December 1976 through December 1979 -
dated December, 1979.
3 Parry, H. L., et al "Augmented Dry Cooling Surface Test
Program: Analysis and Experimental Results",Battelle (PNLl Report
No. PNL - 2746, September, 1979.
ratio of total air-side heat transfer surfa~e to frontal area of
the condenser, dimendonless
Cp specific heat of air, JIkg K f Fanning friction factor,
dimensionless
I. D. inside tube diameter, cm
ITD initial temperature difference, K or 0 C
Colburn air-side heat transfer coefficient, dimensionles s
LMTD logarithmic mean temperature difference, K or 0 C
m air mass flowrate per m2 of frontal area, 2kgls m
O. D. outside tube diameter, cm
P fan power-based on frontal area, W/m 2
O/A heat flux, J/m 2 s q heat rejection rate per m2 of fronta 1
areq,
2J/s mOdry heat rejection rate-dry mode, JIs 0wet heat rejection
rate-with water deluge, J{s R.H. air relative humidity, %
Tsat saturation temperature, K or 0 C
U overall heat transfer coefficient, J1m2 s K U* overall heat
transfer coefficient (wet mope)
J/m2 s K boHin enthalpy di fference of saturated air at t~e
ammonia temperature and the enthalpy of the air at the inlet
temperature and humidity:', Jlkg
bop air-side frictional pressure loss throughi condenser, Pa
421
ESL-IE-81-04-71
Proceedings from the Third Industrial Energy Technology
Conference Houston, TX, April 26-29, 1981
1981 Proceedings Volume II.pdf