Evaporation from Thin Water Films on Horizontal Tubes - Industrial

Evaporation Heat Transfer Coefficients for Thin Sea Water Films on Horizontal Tubes. Industrial & Engineering Chemistry Process Design and Development...
0 downloads 0 Views 3MB Size
Winnick, J.. Marshall, R. D., Schubert, F. H., Ind. Eng. Chem., Process Des. Develop., 13, 59 (1974).

0 = outlet 6 = water Literature Cited Huddleston, J. C.. Aylward, J. R., Final Report, NAS 9-11830, NASA Manned Spacecraft Center, May 1972. Huddleston, J. C., Aylward, J. R., Final Report, NAS-12920, NASA Johnson Soace Center. SeDt 1973. Lin, C. H., Winnick, J., /nd. 'Eng. Chem., Process Des. Develop., 13, 63 (1974).

Received for review September 20. 1973 . Accepted January 7,1974

Part Of this

work was carried

Out

One

Of

the authors

( C . H.L.) was

a N a t i o n a l Research C o u n c i l Postdoctoral Research Associate, a n d t h e other p a r t was under t h e financial support of NASA Contract NAS 9-12526.

Evaporation from Thin Water Films on Horizontal Tubes Leroy S. Fletcher,* Valentinas Sernas, MechanicaL Industrial, and Aerospace Engineering Department, Rutgers University, New Brunswick, New Jersey 08903

and Lawrence S. Galowin Thermal Processes Division, Office of Saline Water, U. S. Department of the Interior, Washington, D. C. 20240

Results are presented for an experimental investigation of evaporation boiling in thin water films flowing over the outside of horizontal tubes. Heat transfer data were obtained for both 1.0-in. and 2.0-in. diameter smooth 90/10 copper-nickel desalination tubes at saturation temperatures ranging from 120°F (1.69 psia) to 260°F (35.43 psia). Heat fluxes were varied up to 20,000 Btu/hr ft2 with feedwater flow rates of up to 1520 Ib,/hr-ft of tube. The experimental results exhibited a slight increase in the film evaporation coefficient as saturation temperature was increased.

Introduction Multiple effect evaporator systems provide a promising technique for supplying large quantities of fresh water. The horizontal tube multiple effect (HTME) process, one of these systems, incorporates horizontal tube bundles with steam condensing on the inside of the tubes and brine vaporizing from a thin film on the outside of the tubes. The condensing steam on the inside of the tubes provides the latent heat for evaporating the brine film. The HTME process may produce potable water more economically than other systems because of the high levels of heat transfer coefficients obtained. In addition, the HTME system design permits the elimination of many of the intereffect pumps and associated equipment. There are many factors that affect the evaporation of thin liquid films on the outside of horizontal tubes. These factors include scaling, tube surface geometry or enhancement, tube material and diameter, the tube location within the bundle, the water or brine distribution system, and the operating pressures. The mechanism of evaporation and condensation in the presence of such conditions is not fully understood. Further, there do not appear to be any useful techniques for the prediction of the associated heat transfer coefficients under these conditions. There is, therefore, a need for a thorough investigation of the basic heat transfer characteristics of thin liquid films on horizontal tubes. The overall heat transfer coefficient of a tube is a function of all the resistances to heat transfer and is a measure of the thermal efficiency or effectiveness of the tube. The thermal resistance of an unscaled smooth tube wall is small compared to the resistance of evaporation and condensation and generally is a known quantity when the heat flux is known. Hence, the resistance due to evapora-

tion and condensation or the film evaporation coefficient and condensation coefficient become the major contributors to the overall heat transfer coefficient. Cox, et al. (1969); and Fletcher and Sernas (1972) indicate that the film evaporation coefficient is the governing parameter in the overall heat transfer coefficient for horizontal tube systems. This paper presents the results of an experimental investigation of the evaporation occurring in thin water films on horizontal tubes. The experimental facility and the techniques for obtaining the experimental data are described in the following sections. The results are presented in tabular and graphical form. The experimental data are compared with previously reported data and with published correlation parameters and are discussed in terms of the parameters involved. Based on this investigation, basic engineering design information will be provided to establish the governing parameters for HTME desalination systems. Experimental Program To separate the evaporation phenomenon from the other parameters involved in the overall heat transfer coefficient, a single horizontal tube test facility was designed and constructed. The water feed rate, tube heat flux, feed superheat, and chamber pressure were designed for independent variation over a wide range. The single tube thin film evaporation test facility utilized in this investigation consisted basically of an evaporation tube and surrounding chamber, a recirculating feedwater system, and the associated control systems. A 7-in. diameter brass cylinder served as the test chamber for a 2 f t length of evaporation tube. A schematic of the feedwater distribution on the evaporation tube is shown in Ind. Eng. Chem., Process Des. Develop., Vol. 13, No. 3,1974

265

Steam to

Condenser

J a c k a l . GIOOVII

Are Filled with

wilh Th.rmosoup1.i Silr.r-8.oring

Solder

Haatad L e n g t h

20”

Figure 2. Schematic of the electrically heated evaporation tube.

Chambrr Wall

I

Return to Pump

Figure 1. Schematic of the feedwater distribution in the evaporation chamber.

Figure 1. The closed recirculating feedwater system for the evaporation chamber consisted of a positive displacement pump, a water filter, flowmeter, preheater, water distribution tray, and associated valves. The steam generated in the evaporation chamber was passed through a condenser, and the resulting condensate was returned to the feedwater system. The evaporation chamber pressure was regulated by a preset reference tank pressure and an automatic pressure control valve. The 7-in, diameter brass cylindrical test chamber was fabricated with a 4-in. diameter steam dome. The electrically heated evaporation tube was centered and cantilevered from one end of the chamber. The feedwater was introduced through a distribution tray which could be adjusted vertically and horizontally to control the water distribution. In addition, the evaporation tube could easily be inclined up to 5” from the horizontal or could be replaced completely by any type of enhanced tube. Further details of the facility construction were reported by Fletcher and Sernas (1972). The feed tray assembly was an open box with a perforated bottom. The bottom plate of the feed tray could be changed to provide different water distribution patterns (See Figure 1). The water level within the tray could be controlled for prescribed flow rates and could be observed through a sight glass at one end of the tray. The excess feedwater left the chamber through a tube in the bottom of the test chamber. The generated vapor passed around the feed tray, collected in the steam dome, and passed out through a tube at the top of the steam dome. The vapor velocities were low enough within the test chamber to pre266

Ind. Eng. Chern., Process Des. Develop., Vol. 13, No. 3, 1974

vent entrainment. As a result, very little vapor shear occurred at the falling film surface to enhance the heat transfer. The experimental facility was designed for operation at saturation temperatures ranging from 115°F (1.47 psia) to 260°F (35.43 psia). Design operating characteristics comparable to those expected in HTME desalination systems were incorporated. These characteristics included capabilities of heat fluxes of up to 20,000 Btu/hr ft2, feedwater flow rates of up to 1520 lb,/hr-ft of tube length, and various tube geometries and feedwater flow patterns. To obtain accuracy and also in the interest of simplicity, the design incorporated an electrically heated tube designed for a Variac-controlled 7-kW power supply. This constant heat flux simulated an idealized test condition for a tube heated by steam flow from within the tube. An electrically heated tube was chosen because it would yield more consistent data than a steam-heated tube. Cox, et al. (1969), and Cannizzaro, et al. (1972), observed that stable filmwise condensation may not necessarily exist over the whole inner surface of a tube. Steam tends to condense in a random fashion which is neither completely filmwise nor completely dropwise in nature. Small changes in the steam chemistry and tube surface have been observed to change the condensation process from day to day. Thus, with steam heating it would not have been possible to obtain a good measure of the local heat flux, but with electric heating the accuracy of the heat flux was as accurate as the measurement of the electrical power. To provide a basis for comparison, two evaporation tubes were assembled and instrumented. A schematic of an evaporation tube prepared for testing is shown in Figure 2. The nichrome heating element for each tube was spirally wound to provide uniform heat generation and, as a result, a uniform surface flux. The evaporation tube surface was prepared by polishing with a medium coarse steel wool (no. 2 grade). A sample section of the 90/10 copper-nickel tube was prepared in the same manner. Photomicrographs of the sample surface “as received” and “after polishing” indicated surface finishes of 25 f 5 and 30 f 5 pin., respectively. The steel wool preparation did not appreciably change the surface finish. Instrumentation was provided for the measurement of temperatures throughout the system, the heat flux ( i e . , the electrical power input to the tube), the evaporation chamber pressure, and the feedwater flow rate. A number of temperature measurement techniques were tried in order to establish reliable temperature measurements for the evaporation tube wall. Techniques for the attachment of thermocouples to the outside of the tube have included soldering, spot welding, capacitor dis-

Table I. Averaged Experimental Evaporation Data 90/10 Copper-Nickel Evaporation Tubes

Tube diameter, in.

r, lb,/hr-ft

1.OQ

532

120

1. o a

532

150

1. o a

532

180

1.o=

532

212

1. o a

532

240

1. o a

532

260

of tube

Taat,

O F

2 .Ob

988

120

2 .Ob

988

150

2 .Ob

988

180

2.05

988

212

2 .Ob

988

240

2 .Ob

988

260

QlA, B t u l hr ft2

20,000 14,920 11,290 20,000 14,920 11,290 20,000 14,920 11,290 20,000 14,920 11,290 20,000 14,920 11,290 20,000 14,920 11,290 16,600 11,180 7,450 16,600 11,180 7,450 16,600 11,180 7,450 16,600 11,180 7,450 16,600 11,180 7,450 16,600 11,180 7,450

AT,,

O F

16.1 12.1 8 .O

12.6 10.0 7.8 12.8 10.1 8.7 12.2 10.2 7.5 10.8 8.1 5.8 9.9 8 .O

5.8 10.7 8.5 7.1 9.5 7.9 7.4 10 .o 7.7 6.1 10.8 8.7 6.4 11.2 9 .O 5.7 10.8 8.4 6.1

he, Btu/ hr f t 2 O F

1242 1233 1411 1587 1492 1447 1563 1477 1298 1639 1463 1505 1852 1842 1947 2020 1865 1947 1551 1315 1049 1747 1415 1007 1660 1452 1221 1537 1285 1164 1482 1242 1307 1537 1331 1221

Wall thickness, 0.041 in.; surface finish, 30 f.5 pin. Wall thickness, 0.032 in.; surface finish, 35 f 5 pin charge, and clamping. Of these methods, soldering provided the most consistent results and was used in this investigation. Techniques for the installation of thermocouples on the inside wall involved soldering the thermocouple junction into a cavity in the wall and the insertion of the thermocouple in a slot in the wall of the tube sleeve. The latter method was preferred, as indicated in Figure 2. The temperature system included 30 AWG copper-constantan thermocouples, thermocouple switches, ice junction, a Leeds and Northrup Model 8686 potentiometer, and a Hewlett-Packard Model 3450 digital voltmeter. Samples-of the thermocouple wire were calibrated over the range of temperatures utilized in this investigation and found to be repeatable and accurate to 0.3"F. The effect introduced by attaching the thermocouples with solder was found to be less than the temperature drop across the tube wall, or approximately 0.4"F.

Results a n d Discussion Tests were conducted in the single-tube thin film evaporation test facility over the range of saturation conditions from 120 to 260°F. The accuracy of the experimental data, along with interpretation of the results,. is discussed in the following sections. Experimental Data. Temperatures obtained along the evaporation tube indicated that the axial temperature variation averaged approximately 1.O"F over the tube length. Circumferential temperatures indicated a variation of up to a maximum of 4.0°F, depending upon the saturation conditions. For the 2.0-in. diameter tube a t a heat flux of 11,180 Btu/hr ft2 and a saturation tempera-

ture of 120"F, the wall superheat (temperature above saturation temperature) was 6.9"F on the top of the tube, 8 5 ° F on the sides of the tube, and 7.7"F on the bottom of the tube. At the same conditions and a saturation temperature of 240"F, the wall superheat was 4.9"F on the top of the tube, 8.7"F on the sides of the tube, and 6 3 ° F on the bottom of the tube. The evaporation coefficients were calculated from the power to the evaporation tube and appropriate temperatures as follows.

The heat flux per unit area is based on the heated portion of the evaporation tube, and the wall superheat (T, Tsat.)is determined from the average tube wall temperature and the feedwater saturation temperature. The experimental evaporation data for the present investigation are listed in Table I. These representative data were obtained for test conditions comparable to those expected for a horizontal tube in a tube bundle of a HTME desalination facility. For each heat flux range, the water feed rate was maintained constant, while the feedwater temperature was varied from 120 to 260°F. The evaporation chamber pressure was always maintained at the saturation pressure corresponding to the feedwater temperature. There was negligible ,difference between the saturation temperature obtained from the evaporation chamber pressure measurement and the saturation temperature obtained from the thermocouples in the feedwater distribution tray. Ind. Eng. Chem., Process Des. Develop., Vol. 13, No. 3, 1974

267

t

He01 Flu*

BI",h,

(12

0 11.290

U 14,920 A 20.000

Figure 5. Photograph of boiling on the thin water film at a saturation temperature of 150°F and a heat flux of 10,080Btu/hr ft2. 1-11

2000-

-. . P

.c

r

=

5

~looo.)

c

olOO

140

180

220

260

300

T.01. O F

Figure 6. Comparison of the present test results averaged over all heat fluxes with reported data far sea water (Cannizzaro, et ol., 1972).

in this investigation is different from that considered by most other investigators. These comparisons, therefore, can be made only in general terms. I t is hoped, however, that they will contribute to the understanding of the heat transfer and fluid mechanics involved in the phenomenon. The experimental evaporation coefficients obtained in this investigation are compared with reported evaporation data of Cannizzaro, et al. (1972), in Figure 6. Data from the present investigation are for a single 2.0-in. diameter smooth 90/10 copper-nickel tube using distilled water. The present data shown are an average of all heat flux test conditions. The reported data, however, are for an average beat flux using a 2.0-in. diameter smooth 90/10copper-nickel tube located in the center of a tube bundle with a sea water film. The present data compare favorably with the reported data at the low saturation temperatures. As saturation temperature increases, however, the present experimental data appear to diverge significantly when compared to the reported data, due to the differences in test conditions. Further, steam shear effects or interactions hetween tubes within a bundle can have a pronounced effect on the heat transfer characteristics, as noted by Dukler (1960). The present experimental data (at high heat fluxes where considerable boiling was occurring in the film) were

used in the dimensionless parameters proposed by Rohsenow (1952, 1972), who used the data of Frost and Li (1971) for subatmospheric pool boiling of water on 0.008-in. diameter platinum wires. The present data do not appear to follow the correlation. In general, the heat transfer coefficients obtained in the present investigation are much higher than those observed by Frost and Li (1971) for pool boiling on thin wires. It appears, then, that the boiling heat transfer occurring on a horizontal tube might be augmented by fluid mechanics effects which are not represented in the parameters proposed by Rohsenow (1952, 1972). The present experimental data (at low heat fluxes where little boiling was occurring in the film) were also compared with the falling film correlation reported by Chun and Seban (1971) as shown in Figures 3 and 4. Their correlation permits the prediction of heat transfer coefficients for water films flowing along the outside surface of vertical tubes. The present data are significantly higher than their correlation which was derived for a fully developed falling film on a vertical tube. The verticai falling film was well developed because it flowed 12 in. down a vertical tube wall before encountering the heated section of the tube. The film on the horizontal tube, however, cannot develop into a “fully developed” profile because the gravity vector changes continuously relative to the velocity vector as the liquid flows over the tube. It appears, then, that both the boiling and the undeveloped nature of the liquid film on the horizontal tube contribute greatly to the large evaporation heat transfer coefficients. Conclusions An experimental investigation of the film evaporation coefficients of thin water films surrounding horizontal 90/10 copper-nickel evaporation tubes has been conducted. Data have been obtained for saturation temperatures ranging from 120 to 260°F with wall superheats of 6 to 16°F. Heat fluxes ranged up to 20,000 Btu/hr ft2 for the 1.0-in. tube, and 16,600 Btu/hr f t 2 for the 2.0-in. tube, with feedwater flow rates of up to 988 lb,/hr-ft of tube. Comparison of the present experimental evaporation data with reported evaporation data indicates that the present data are significantly lower. These differences may be attributed to the diverse characteristics and properties of fresh water and ocean water, the differences between a tube bundle and a single tube, the vapor shear interaction occurring in tube bundles, and the differences in the feedwater distribution system. Further, comparison of the present data with published correlations has indicated that these correlations do not adequately reflect the heat

transfer and fluid mechanics effects of the phenomenon studied in the present investigation. Comparison of the 1.0 and 2.0-in. tube data indicate that the tube diameter definitely influences the heat transfer. Analysis of the experimental results of this investigation suggests that additional studies should be considered, especially in the area of sea water tests. Shrouding the single tube to simulate interior tube bundle vapor shear interaction conditions should be considered. The injection of vapor at various locations in the vicinity of t h e evaporation tube would also be useful in the study of the vapor shear interaction. Evaporative tests without boiling (very low wall superheats) should be conducted to ascertain the contribution of surface evaporation to the total film evaporation coefficient. The investigation provides basic data for evaporation from thin water films on horizontal tubes. Results of further tests under varying conditions may be compared with the present results in order to ascertain the effects of the numerous variables which are involved in the phenomenon of moving thin liquid films. Acknowledgment The authors wish to acknowledge the laboratory assistance of Walter Parken and Terry Clark. Nomenclature he = evaporation heat transfer coefficient, Btu/hr ft2 OF Q / A = heat flux density, Btu/hr ft2 T, = wall temperature, O F T,,,. = saturation temperature, “F AT, = wall superheat ( T , - T,,t.), “F r = mass rate of flow per unit length of tube, lb,/hr f t Literature Cited Cannizzaro, C. J., Karpf, J. Z.,Kosowski, N., Pascale, A . S.,4th HTME Progress Report, Universal Desalting Corporation,July 1972. C h u n , D. R., Seban, R. A., J. Hear Transfer, 93, 391 (1971). Cox, R . B., Matta, G . A., Pascale, A. S., Stromberg, K . G . , OSW R B D Report 492, U. S. Department of the Interior, Oct 1969. Dukler, A. E., Chem. Engr. Progr. Symp. Ser., 56, 1 (1960). Fletcher, L. S.. Sernas, V., Engineering Report RU-TR 139-MAE-H. R u t gers University, June 1972. Frost, C. W., Li, K.W., J . Heat Transfer, 93, 232 (1971). Rohsenow, W. M . , Trans. ASME, 7 4 , 9 6 9 (1952). Rohsenow, W. M., J. Heat Transfer, 94, 255 (1972) Received for revielv October 15, 1973 Accepted March 26, 1974 This paper was presented a t the ASME/AICHE National Heat Transfer Conference, Atlanta, Ga., in August 1973. The authors gratefully acknowledge the financial support of the United States Department of Interior, Office of Saline Water, Contract 14-302907,, and the facilities of the Rutgers University Mechanical, Industrial, and Aerospace Engineering Department.

Ind. Eng. Chem., Process Des. Develop., Vol. 13,No. 3,1974

269