Vacuum Refrigeration - Industrial & Engineering Chemistry (ACS

Vacuum Refrigeration. D. H. Jackson. Ind. Eng. Chem. , 1936, 28 (5), pp 522–526. DOI: 10.1021/ie50317a004. Publication Date: May 1936. ACS Legacy Ar...
0 downloads 0 Views 954KB Size
INDUSTRIAL AND ENGINEERING CHEMISTRY Schestakow, J . Russ. Phys. Chem. SOC.,35, 858 (1903); 37, 5 (1905); German Patent 164,755 (Nov. 2, 1905). Schiff, Ann., 151, 186 (1869). Scholl and Davis, IND. EXG.CHEM., 26, 1299 (1934). Seidell, A., “Solubilities of Inorganic Compounds,” 2nd ed., Vol. I, p. 738, New York, D. Van Nostrand Co., 1919. Ibid., suppl. 2nd ed., Vol. 11, p. 1486 (1928). Shnidman and Sunier, J. Phva. Chem., 36, 1232 (1932). Smolka, Monatsh., 8, 64 (1887). Solvay (Slosse), Bull. acad. YOU. Betg., [3] 35,547 (1898). Speyers, Am. J. Sci., [41 14, 293 (1902). Ibid., 10, 61 (1903). Tanatar, J. Russ. Phys. Chem. Soc., 47, 1283 (1915). Timofeiew, Dissertation, Kharkhov; Seidell, “Solubilities of Inorganic Compounds,” 2nd ed., Vol. I, p. 737, New York, D. Van Nostrand Co., (values derived) 1919. Ullmann, “Ensyklopadie der tech. Chemie,” 2nd ed., Vol. 6, p. 107, Berlin, Urban and Schwarzenberg, 1930.

0

.

0

VOL. 28, NO. 5

(78) Vee and Davis, IND. ENG.CHEM.,Anal. Ed., 7, 259 (1935). (79) Walker and Hambly, J . Chem. Soc., 67, 746 (1895). (80) Walker and Wood, Ibid., 83, 484 (1903). (81) Walton and Wilson, J . Am. Chem. Soc., 47, 320 (1925). (82) Werner, E. A.. J. Chem. SOC.,103, 29 (1913); Ber., 18, 3106 (1885),; 19, 341 (1886); 38, 1010 (1913); Richter, “Organic Chemistry,” 3rd ed., Vol. I, p. 530, tr. by E. N. Allot, P. Blakiston’s Son & Co., Philadelphia, 1934. (83) Werner, E. A., J . Chem. Soc., 103, 1015 (1913). (84) Ibid., 117, 1046 (1920). (85) Ibid.,111, 863 (1917). (86) Werner, “The Chemistry of Urea,” London, Longmans, Green & Co., 1923. (87) Whitaker, Lundstrom, and Hendricks, IND.EKG. CHEM.,23, 1281 (1933). (88) Wohler, An%.Phys. Chem., 12, 253 (1828).

RECEIVED March 14, 1036

VACUUM REFRIGERATION

T

HE chemical engineer and refrigerating engineer have a t their disposal a large variety of industrial refrigerants. New ones have frequently been added to the list and some have been quickly and widely adopted with considerable success. Water as a refrigerant is by no means new, but its possibilities were neglected for many years until recent improvements in the steam-jet type of vacuum producer provided economical equipment for applying this method. Chilling water by partial evaporation a t high vacuum has been known for decades. A few isolated cases are recorded FIGURE 1. REFRIQERATING UNIT OF 24 Tom CAPACITY, WITH VERTICALVACUUM CHAMBER

D. H. JACKSON Croll-ReynoldsCompany, Inc., New York, N. Y. where the method was applied in the early part of the century, but the installations proved impractical and were more in the nature of engineering experiments than of practical or economical installations. During the past few years a large amount of development work has been done by different manufacturers of steam-jet equipment, and numerous installations have been made for air-conditioning, cooling drinking water for large buildings, and cooling a large variety of fluids in chemical process work. A conservative estimate of the total capacity of vacuum refrigerating equipment installed in the past three years is equivalent to 50,000 tons of ice every 24-hour day. The reason for this rather sudden commercial application of an old, well-known principle is interesting. Many of the early attempts to apply this method depended upon reciprocating vacuum pumps to produce the vacuum and draw off the necessary vapor from the high-vacuum evaporating chamber. The problem can be easily understood by considering a typical case of chilling water required for air-conditioning a medium-size restaurant. Assuming that the required capacity for cooling and dehumidifying the air is 50 tons of refrigeration, a total heat absorption of 600,000 B. t. u. per hour is necessary. Assuming that the required chilled water temperature is 45’ F., a vacuum corresponding to an absolute pressure of 0.3 inch of mercury is required. Therefore, sufficient water must be evaporated a t 0.3 inch absolute so that the latent heat of evaporation will absorb the 600,000 B. t. u. The latent heat a t 0.3 inch is 1066 B. t. u. Dividing 600,000 by 1066 gives 562 pounds per hour of water which must be evaporated. The specific volume of water vapor a t 0.3 inch absolute is 2033 cubic feet per pound. The total volume of vapor to be handled, therefore, is 562 X 2033 or 1,142,546 cubic feet per hour (19,042 cubic feet per minute).

Steam-Jet Apparatus Obviously, then, the vacuum cooling method is not practical with conventional reciprocating vacuum pumps. This type of pump has its advantages for low and moderate vacua, but it is generally agreed that its efficiency falls sharply a t high vacuum. A battery of dozens of large-size commercial reciprocating pumps would be required for this duty. On

MAY, 1936

INDUSTRIAL AND ENGINEERING CHEMISTRY

523

condenser is shown in Figure 2. the other hand, t h e s t e a m - j e t A brief history is given of In this case the condenser is operattype of vacuum producer, operatthe process of chilling liquids ing as a low-head jet condenser since ing a t high velocity on an entrainby partial evaporation under it discharges into a pump instead of ment rather than a displacement high vacuum, with reasons into a full-length barometric leg. A principle, can handle this capacity for its recent wide commerlow-head j e t c o n d e n s e r and a in a single medium-size unit which barometric condenser are interweighs, with its condenser, approxicial application. The steamchangeable, and their condensing mately 3000 pounds. jet type of vacuum apparatus efficiency and p e r f o r m a n c e are The reason for the difference is for applying the process is equal. the high v e l o c i t y of t h e steam, described. Curves give actual a v e r a g i n g 4000 feet per second Experimental Data operating data with figures leaving the jets, and the entrainment of the low-density vapor a t for steam and water conIn collecting experimental data, high velocities as the steam exthe vacuum evaporating chamber sumption at different chilled pands. The steam-jet unit also has was tested under its full range of water and condenser water the advantage of no moving parts. working conditions to be sure. that temperatures. The discusFigure 1 shows a c o m m e r c i a l there had been sufficient contact of sion includes : vacuum coolthe water with the vacuum so that unit with a capacity of approxiing of liquids by direct expothe water leaving would be chilled mately 24 tons of refrigeration. completely to the t e m p e r a t u r e The large vertical vessel a t the left sure to vacuum and by an corresponding to the vacuum. The is the vacuum evaporating chamber. indirect method with a heatlarge steam-jet unit was carefully Suitable internal sprays and baffle exchanger bundle submerged designed in each case for the exact equipment are provided in this in chilled water in the vacuum c o n d i t i o n s . The amount of live chamber to give the necessary conchamber ; continuous and steam used in the nozzles was caltact between the water and the vacuum space. The main steamculated from the temperature rise batch cooling, with mention in the condenser water, correcting jet evacuator is bolted onto the of operating economies proconnection on top of the vacuum for the a m o u n t of vapor drawn vided by t h e latter; and chamber. This unit discharges into from the vacuum e v a p o r a t i n g chemical engineering applia surface or barometric condenser chamber. The quantity of live cations, including v a c u u m where an intermediate vacuum of steam was checked by calculating approximately 28 inches is mainthe flow from the size of the orifice crystallization, air-conditiontained. A small two-stage steamusing standard formulas for this ing, chilling mash for disjet evacuator is used to maintain purpose. For each test the design tilleries, simultaneous chillthis 28-inch vacuum on the conof the steam-jet unit was made ing and degasifying of misdenser. This comprises the comaccording to modern improved cellaneous liquids, etc. plete equipment with the exception practice with efficiency as high of pump or barometric leg to draw as possible. the chilled water from the vacuum In testing some of the equipchamber. ment for maximum vacuum The collection of data for different conditions required without water in the chilling tank, absolute pressures slightly considerable time and numerous changes in equipment. lower than 1 mm. were obtained. At any absolute pressure The steam-jet type of vacuum equipment is highly dependable less than 3 mm., ice crystals formed in the flow of steam from when operated within reasonable variations of a given set of the jets. The steam supplied to the jets was a t a pressure of conditions for which it is designed. However, it is also quite 100 pounds per square inch gage and a temperature of 338' sensitive to appreciable changes in certain operating condiF., and the steam velocity from the jets was approximately tions, such as excessive temperatures of condensing water, 4000 feet per second. Considering the steam temperature, low pressure of live steam, etc. The booster steam jet, the exceptionally high velocity, and the fact that the nozzles designed to discharge into a condenser where a vacuum of are only about 4 inches long, the formation of ice crystals 28 inches is maintained, will fail when discharging into a seems somewhat anomalous. It is due to the rapid expansion condenser where warmer cooling water can produce a maxiof the steam and the accompanying heat absorption. Some mum vacuum of only 27.5 inches. In order to obtain good of the steam first condenses into moisture, and, since the efficiency, the internal dimensions of the steam-jet unit have absolute pressure is lower than the vapor pressure of ice a t to be extremely accurate and carefully correlated. Therefore, 32' F., a part of the moisture reevaporates, thereby absorbing in collecting accurate performance data with different temsufficient heat to produce freezing. Without much extra peratures of cooling water and a constant steam-pressure, it manipulation a good imitation of a natural snowfall can be was necessary to change the monel metal steam nozzles produced inside the vacuum vessel. frequently and the combining throat into which these nozzles When operating a t vacuum sufficiently high to form ice, discharge. the steam consumption of the jets is such that operating An entirely different set of data was obtained by making costs are usually higher than for mechanical refrigeration, the main condenser of barometric instead of surface type. In although low-cost steam and relatively cold condenser water this case the discharge from the steam jet evacuator comes in in some cases make the steam-jet unit economical and pracdirect contact with the condenser water and permits a contical even a t temperatures a few degrees below freezing. siderably closer terminal difference between the discharge For refrigeration to intermediate temperatures in the neighcondenses water and the temperature corresponding to the borhood of 40" F., vacuum refrigeration shows excellent vacuum. This means slightly better efficiency than with a operating economy, as Figure 3 indicates. surface condenser where all of the heat must be transferred Operating data from test units and from commercial inthrough tubes, making the terminal difference unavoidably stallations are given as curves for convenience in Figure 3. higher. A typical test. set-up using a barometric type of They are based on steam a t 100 pounds per square inch

524 L

INDUSTRIAL AND ENGINEERING CHEMISTRY

VOL. 28, NO. 5

this exhaust steam even though it has no positive pressure above atmospheric. When the proper vacuum is maintained in the main condenser, this atmospheric-pressure steam develops sufficient velocity through the jets to give practical and economical results for many conditions. Of course, steam eonsumption is greater when using low pressure steam. The quantities required a t pressures down to atmospheric were determined but are not available for publication. Vacuum refrigerating equipment has considerable flexibility. Accurate tests indicate that any unit designed for a definite chilled water temperature can be operated a t a variety of different chilled water temperatures at different capacities; for example, a unit designed to produce 100 tons of refrigeration a t a chilled water temperature of 50" F. will automaticallyproduce 115 tons of refrigeration a t a chilled water temperature of 55" F. If the chilled water temperature is lowered instead of raised, the capacity is also lowered; a t a chilled water temperature of 40" F. the capacity drops to 70 tons, and to 55 tons with a chilled water temperature of 35". Conversely, a unit designed for operation a t a relatively low temperature will provide increased heat absorption capacity when operated a t slightly higher temperatures.

Batch Cooling The foregoing operating data apply to cooling by continuous operation. The vacuum process is also adaptable to batch cooling, usually with improvement in operating FIGURE 2. TYPICAL ARRANGEMENT OF VACUUM REFRIGERATINGefficiency. Batch cooling is not practical for air-conditioning EQUIPMENT or for process work where the temperature range in the chilled water is less than 10" F. However, where the temperature range in the medium to be chilled is more than lo", batch pressure and on the use of barometric or low-head jet concooling frequently offers savings, particularly where liquids densers. Curve A gives steam and water consumption figures must be cooled over a range of 100" or more-for example, when chilling water to 40" F. Curves B, C, and D give the in the cooling of mash in a whisky distillery. Since the desame data for chilled water temperatures of 45", 50", and velopment of vacuum cooling, mash tubs are frequently made 60" F., respectively. The steam consumption figures are to withstand vacuum so that, after the mash has been cooked, expressed in pounds per hour per ton of refrigeration, water it can be quickly cooled in the same vessel by applying the consumption figures in gallons per minute per ton of refrigeraproper vacuum. This is a good illustration of a case where tion. The steam consumption is plotted against the water conthe batch method considerably increases the efficiency, It sumption because one can be compromised against the other. is necessary to provide some form of agitation for the mash Within certain limits the steam consumption can be reduced so that, as it is cooled on the surface, it will be thoroughly by using more water, and vice versa. In using the curves to mixed for homogeneous cooling throughout. The improveobtain performance data, it is necessary only to select the ment in efficiency in the batch method over the continuous proper point somewhere along the curve corresponding to method is obvious when we consider that it permits most of the maximum temperature of the available condenser water. the vapor being removed from the vacuum evaporating chamIn estimating, the maximum summer temperature of conber a t a much lower vacuum than that corresponding to the denser water should always be used; when the weather is final chilled temperature. I n the case of continuous cooling, mild and the water cooler, its quantity can be reduced. it is necessary to maintain maximum vacuum constantly, Many chemical plants have a surplus of exhaust steam at and all the vapor must be handled a t this maximum vacuum. atmospheric pressure or slight positive-pressure, particularly Many data on batch operation of vacuum cooling have been in summer when refrigeration is most necessary. The steamcollected, although it is difficult to put them in concentrated jet type of cooling equipment has been adapted for using

POUNDS STEAM P€R HR, p€R 7VN

FIGURE3. OPERATING DATAFROM TESTUEITSAND COMMERCIAL INSTALLATIONS

MAY. 1936

IUDUSTRIAL AND ENGINEERING CHEMISTHY

525

F I ~ U R4E(Left). EXPERIMENTAL VACUGMREFainmmiNn UNIT WIT= SEIBMERQED-T~BE INTEORAL HEATEXCHANOSR FEUD 5 ( B e h ) . V A C U ~ REFRIOERATINO M UNIT OF Io TONSCAPACITY, NITK STEAH-JET 1:'JECTOH

form since conditions of industrial application vary widely. Initial liquor temperatures vary from 220" to f30"F. or even less, and final temperatures down to 15' F. The saving in steam and water consumption on the batch method over the continuous method varies from 50 per cent for large cooling ra.nges to 10 per cent or less for a cooling range of 10" F.

Indirect Cooling For fluids which must be cooled without direct contact with vacuum, the vacuum method can still be applied by using any convenient type of heat exchanger. Some of the test resulte were obtained by submerging a heat-exchanger bundle in the chilled water inside the vacuum evaporating chamber. Figure 4 sliows an experimental set-up of this type. This heat exchanger built integral with the vacuum chamber indicated a number of advantages over an external-type heat exchanger in aseparate shell: (1) The external heat exchanger requires that water be drawn from the vacuum and pumped through the separate exchanger which involves additional equipment and additional operating cost. ( 2 ) The integral heat exchanger tias uniform temperature of chilled water and therefore a higher mean temperature difTerence than in the case of the external heat exchanger where the chilled m t e r is warmed several degrees before it Aows back to the vacuum chamber. Some suitable form of agitation in the chilled water may be necessary when using an integral heat exchanger; otherwise, the heat transfer rate would be relatively low. However, in cases where oils or other viscous materials a.re being handled inside the tubes, the heat transfer rate in tlie film outside the tubes is not important since there would still be enough thermal circnlation to give a better heat transfer rate in the water film outside than there vould be in the more viscous film inside witli higher velocity. The integral heat exchanger in the vacuum cooling unit was used on some large commercial work such as chilling lubricating oil in the presence of solvents to remove paraffin and wax. This metliod was also applied for producing chilled water for industrial process work where it was desirable to maintain the water under its initial pressure without dissipating this pressure in a V ~ C U U I T I chamher.

Applications Although the largest single application of vacuum cooling is for producing chilled water in air-conditioning, it has also

been applied widely to a number of different chemical engi-

neering processes. The most important is doubtless the Swenson process of vacuum crystallization described by Caldwell (1). In this case the vacuum serves a double purpose in chilling the mother liquor to temperatures as low as 32" F. and a t the same time concentrating it as the chilling progresses, thus increasing the production of crystals. There is further advantage in the fact that the heat is removed from the mother liquor by direct evapenttion rather than by being transferred through tubes or metal surfaces, as would be necessary if mechanical refrigeration were used. In the majority of commereial crystallining operations, the mother liquor is corrosive, which always complicates the heat transfer through metal surfaces. In the vacuum process tlie crystallizing chamber can be rubber-lined or glass-lined without involving any heat trransfer problems, as fong as moderate agitation is maintained in the liquor. Most vaciiiiin crystallizing is done on a batch process, although the continuous process is also successful, even though its operating cost is slightly higher in many cases. Vacuum crystallization is by no mean8 new, although combining it with vacuum refrigeration for subatmospheric temperatures has come only with recent general development in the vacuum refrigerating process. It might also be said that the development of high vacuum crystallizing has in some cases preceded the development of general vacuum refrigeration as the demands of the former process furnished some of the incentive for developing the latter. The vacuum mfrigentting process automatically includes a thorough dedration or degasification of the liquid being cooled. The drop in temperature of the liquid and the accompanying evaporation of water into large volumes of lowdensity vapor take out all dissolved gases down to a trace which is usually negligible. This is a decided advantage for many chemical process operations. Dissolved oxygen is frequently the cause of oxidation of cliemicals or bacteria gr0wt.h in miscellaneous liquids. Any possible concentration

INDUSTRIAL AND ENGINEERING CHEMlSTRY

526

of dissolved oxygen is entirely removed when a liquid is cooled as much as 10" F. by vacuum evaporation. The same is true of carbon dioxide and many other gases. Some previous data on removal of carbon dioxide by vacuum evaporation a t somewhat higher temperatures were published by Jackson and McDermet ( 2 ) . Preliminary testing on hard water and on carbonated soft water indicates that carbonate removal by vacuum evaporation a t low temperatures is practically as efficient as a t high temperatures when the evaporation produces a temperature drop of 10' F. or more. There are numerous cases in chemical process work where chilling of various liquids is being accomplished by mechanical refrigeration and where the automatic degasifying action of vacuum refrigeration would more than justify a change to the latter process. There are other conditions where refrigeration is being applied to liquids that

VOL. 28, NO. 5

must retain certain dissolved gases. In cases of this kind, vacuum cooling can be applied indirectly through tubes or any convenient heat transfer medium. Preliminary experimental work indicates that vacuum cooling of milk improves its flavor by removing traces of dissolved gases. A pilot plant was installed a t Cornel1 University; experimental work is well under way, and considerable data have been collected. Some of these data will doubtless be published in the near future.

Literature Cited (1) Caldwell, Chem. Ce- M e t . Eng., 39, 132 (1932). (2) Jackson and McDermet, IND. ENG.CHEM.,15, 959 (1923). RECEIVED July 11, 1935. Presented before t h e Division of Industrial and Engineering Chemistry a t the 90th Meeting of the American Chemical Society, San Francisco, Calif., August 19 to 23, 1935.

on the apparent plate efficienc y E,, for the case of total reflux a n d a d r y vapor (Murphree) efficiency, E,, = 1.0. Starting from the definition of apparent plate efficiency, E,,

ON

EFFECT OF

PLATE EFFICIENCY IN

Ea = ( Y , - Yn-l),'

DISTILLATION

(y*n

ALLAN P. COLBURN E. I. d u Pont de Nemours & Company, Wilmington, Del.

T

HE cost of a plate column for distillation

- Yn-d

(1)

where Yn and Yn-l are amarent values of the c o m p o s i t i b n of vapor leaving the nth and n - l t h plates, respectively, defined for total reflux, as Y,-l

=

x,, Y, =

xn+1

(2)

and where y/*n is the vapor composition in equilibrium with xn, and x n is the composition of liquid leaving the nth plate, Underwood obtained:

or absorption depends upon both the diameter a n d height required. The limitEa = 1/(1 e ) (for E, = 1, L/V = 1) ing factor in minimizing both of these dimensions by in(3) creasing vapor velocity and reducing plate spacing is often the increase of entrainment to a point where the plate Souders and Brown (6) sought an expression for plate efficiency is considerably reduced. To determine the optimum efficiency for the general case of any reflux ratio and any vapor size of column, it is necessary to know both the variation of efficiency. Their definition of plate efficiency demands a entrainment with velocity and plate spacing and the effect different operating line for each different quantity of entrainof a given amount of entrainment on the efficiency. Conment, so that numerical values are not directly comparable siderable experimental data have recently been obtained as with values of plate efficiency based on an operating line t o the amount of entrainment incurred under various condetermined from a heat balance. That is, Souders and Brown d i t i o n s (1, 2, 3, 5, 7 ) . Several theoretical derivations of the effect of enThe effect of entrainment in a plate column used for distillation trainment on plate effior absorption is to reduce the apparent plate efficiency. This efficiency, ciency have been proposed, E,, can be calculated by the approximate equation: and the purpose of this study is to carry on these attempts toward a simple relationship which might be generally useful.

Previous Work The first a n a l y s i s of the effect of entrainment on plate efficiency was apparently that of Underwood (8). He derived E q u a t i o n 3, expressing the effect of entrainment

+

Employing this equation, a relation is developed for the amount of entrainment occurring when the column is operated at the optimum vapor velocity (assuming that the entrainment increases with the fourth power of the velocity): e (at optimum velocity) = R/3E,. Since this value is so high, the importance of large enough down pipes for liquor and of low back pressure through chimneys and caps is emphasized.