Recovery of Solvents from Air-Vapor Mixtures

KNOEDLER AND C. A. DODGE. Received March 29, 1918. At the present time the great scarcity of solvents and the enormous demand for them for military pu...
0 downloads 0 Views 757KB Size
Aug., 1918

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y

I

I

ORIGINAL PAPERS

RECOVERY O F SOLVENTS F R O M AIR-VAPOR MIXTURES By E. I,. KNOEDLER AND C. A. DODGE Received March 29, 1918

At t h e present time the great scarcity of solvents and the enormous demand for them for military purposes makes necessary the highest possible degree of conservation by those industries in which they are used. Furthermore, the high prices which these solvents now command add t o the other reasons for conservation t h e important inducement of large manufacturing economies. This conservation may be effected either by a reduction in consumption or by the recovery of the solvent vapors from those processes in which they are driven off. The first method in many cases would involve a reduction in volume of business, unthinkable except in cases of dire necessity; the second, on the other hand, actually adds t o the available supply and holds out the prospect of increased business, a t the same time making possible substantial economies in manufacturing cost.

FIG.I

593

their very low concentration; and this led t o the development of tightly closed, steam-heated, drying chambers, efficiently insulated with asbestos and equipped with quick-acting, close-fitting, sliding doors, through which are fed the wheeled carts carrying the product. Tracks running completely through t h e box make it possible t o feed the goods in a t one end of the box and remove them Erom the other end. I n this way a steady stream of wet materials is fed t o the drying chambers, maintaining a reasonably uniform air-vapor mixture. The temperature of each one of the fourteen drying boxes is controlled b y a thermostat which regulates the steam supply, and upon each box is mounted an air compressor of such capacity as t o remove the vapors as generated. This arrangement of independent units permits any dry-box not in use t o be cut out of the system, so as t o deliver t o the recovery equipment a t all times an air-vapor mixture uniform in composition and of a constant temperature.

FIG.I1

General layout of recovery plant showing course of vapors from box through pre-cooler and towers, then back t o pre-cooler and t o atmosphere; also of scrubbing liquors from p u m p t o towers, then t o cooling basins where t h e p u m p takes its suction

Bank of four drying chambers showing delivery end. Individual compressors, temperature regulators and steam supply for each chamber are in plain view, also tracks and wheeled carriages for transporting the product

To those who may have considerable quantities of solvent vapors available, t h e description of a suaessful recovery plant may be of interest. The plant has been in operation several years and has recovered many thousand gallons of these precious materials. The solvents which are being recovered are mixed vapors of methyl alcohol, ethyl alcohol, acetone, and camphor, which are driven off in drying the collodions used in coating gas mantles. The proportions of the several solvents have varied from time t o time, but the outfit has operated with success on any mixture so far attempted. The drying of the collodions was first carried on in the open air, then in chambers with open ends, through which ,$he collodionized mantles were passed as they dried. Under these conditions it was found difficult t o recover much of the vapors because of

As delivered by the compressor t o the recovery plant the mixture of vapors shows about the following: Volume a b o u t . . . . . . . . . . . , . . 450 cu. f t . per M. Pressure.. . . . . . . . . . . . . . , , . . 1 lb. per sq. in. Concentration. . . . . . . . . . . . . 4 per cent solvents (by vol.) Temperature.. . . . . . . . . . . . . . 80' C. After the solvents are removed from the air-vapor mixture, the air, which is discharged t o the atmosphere, shows about the following: Vapor concentration. . . . . . 0.9 per cent solvents (by vol.) Temperature.. . . . . . . . . . . . 14' C. It will be seen from these figures t h a t approximately 80 per cent of the vapors passing through the plant are recovered and these' figures are confirmed by the actual output. The plant consists of a large surface condenser used as a pre-cooler; a group of scrubbing towers,!

594

THE JOURNAL OF INDUSTRIAL A N D ENGINEERING CHEMISTRY

in which the vapors are scrubbed; receiving vessels i n which the liquors from the towers are caught and kept cold by means of brine coils; pumps for re-circulating this liquor until it attains a concentration suitable for distillation; an ice machine; and a column still.

.

FIG.111 Pre-cooler, towers, and catch basins as arranged in solvent recovery plant



As the vapors enter the recovery plant they pass through the tubes of a surface condenser where they are lowered from 80” C. t o nearly the temperature of the washed waste gases (about 10’ C . ) . They then pass into a header and are distributed t o a group of *‘bell-and-seal” scrubbing towers where they bubble through water of a temperature of 5’ C. Here the vapors are removed from t h e mixture, and the air, lowered t o a temperature of about 6 ” t o 7 ” C. from its contact with the refrigerated wash solutions, passes out of the tower and back t o the surface condenser where it serves to cool the vapors just entering the plant. After passing through the pre-cooler, the air, deprived of its vapors and warmed up by the heat absorbed from the entering vapors, is discharged to the atmosphere. Of course, any solvents left in this air are completely lost and it is quite possible t h a t in some industrial operations it would pay t o return this air t o the drying chambers for re-use. The washing of the gases in the towers is carried out by means of cold water which is re-circulated by pumps. From the towers the water returns t o catch basins containing brine coils for keepingait a t a low temperature, the pumps taking their suction from the catch basins, The re-circulation continues for some hours until the solution reaches a concentration of about I 2 per cent solvents, when i t is pumped t o t h e storage tank and held until required b y the still. After passing through t h e still, where the solvents are driven o f f , these waters are permitted t o cool and are then returned ’to the scrubbing system. I n this way small losses in t h e tail-liquors are avoided. Each tower, like each drying chamber, is a complete unit, with its individual pump and catch basin,

Vol.

IO,

No. 8

and can be operated regardless of the other units. The still is a copper, fractionating column, composed of a steam-heating section, 36 in. in diameter, 3 distilling sections, 24 in. in diameter, and 7 distilling sections, 18 in. in diameter, divided into 20 distilling chambers. The 18-in. sections are provided with internal cooling coils, connected in series and regulated by a throttle valve. The vapors from the 24-in. sections are by-passed through a tubular condenser, preheating the feed liquors, then back into the 18-in. sections, and from there through the condenser. The steam is supplied by a closed coil and controlled by a differential steam pressure regulator. The still produces a 97 t o 99 per cent product, according t o t h e varying percentages of acetone, ethyl and methyl alcohols in the vapor mixture being worked. The camphor remains in t h e sections of t h e still where the temperature ranges between 95‘ C. and 7 j ” C. Periodically, thesk trays are drained through a header t o a copper condenser, consisting of a series of trays, where the liquor is cooled and some of t h e camphor separates out on the trays. The remaining camphor passes on t o a sump where i t is precipitated and filtered. The camphor press cake is preserved for further purification, and the filtrate is combined with the still liquors and used over again in the scrubbing towers.

-

-

WLTLR PLeD “ I O Y L I T O R FOR L l F L ” ”

3 T O R q G E FOR

Elevation of still showing arrangement of all elements required in its operation

The cost of operation per 83/*-hr. day, with a production of from 65 t o 7 5 gal. of solvents, runs approximately as follows:

Aug., 1918

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y

Direct labor-one man at 35 cts. p a hr.. ........... $ 3 . 0 7 Indirect labor and supervision.. . . . . . . . . . . . . . . . . . . . 0.50 0.60 Expense and repairs.. Steam (pumps and still). .......................... 5.20 Fixed charges.. 8.75

............................ ..................................

-

Product-70 gals. $18.12 Total cost per gaI.--$0.26 (making no allowance for camphor)

The accompanying cuts give a good idea of the general layout of the plant, also of the appearance of drying chambers, scrubbing towers, and distilling column. WELSBACH COMPANY GLOUCESTRR CITY, NEWJERSEY

DETERMINING THE COMPARATIVE MELTlNG POlNTS OF GLUES AS A MEASURE OF THE JELLY STRENGTH By C. FRANK SAMMET Received March 22, 1918

Methods for testing the comparative jelly strengths of glues have never been entirely satisfactory. Criticisms of the various methods are extensively written into the literature of glue testing, and it would seem t h a t a simple, rapid, and yet accurate method is desirable, as the jelly strength is a n important factor in t h e quality of glue for paper making and other purposes. These features are involved in the following method, which is a comparison of t h e melting points of either the ground glues or their jellies. The melting points may be taken as a measure of t h e jelly strength as the two bear a close relation t o each other. The dried glues are brought t o a ground condition in a hand mill and sieved between 2 0 and 40 mesh screens, and t h a t portion remaining on the 40 mesh sieve is retained for the test. Although in many cases glues are mixtures of several qualities, with different melting points, yet t h e mixture of ground particles has never caused inaccuracies in the comparative melting-point tests. The grades determined by the melting-point test have corresponded exactly ,with grades determined by testing t h e actual jelly strength by other methods. About I g. samples of the glues, so prepared, are placed in small beakers and each stirred with I O cc. t o 1 5 cc. of cold water, not above IO' C. They are allowed t o soak one minute, then a portion of each glue is withdrawn b y a spatula and placed on a thin, smooth surface of metal. This metal should be preferably of brass, having a n approximate length of 6 in., a width of 1 . j in., and a thickness of in. The long edges may be turned over t o give the strip rigidity. The glue particles should be placed about 3 in, from the end of t h e strip, and then a portion, about the area of five pin heads, is separated with the spatula and pushed t o within 2 in. of the end, thereby draining off a certain excess of water which adheres t o the surface of the metal. The little heaps of glue particles are now aligned equally distant from the end of the strip, which is then dipped t o a depth of l/z in. in a beaker half full of water a t 40° C. The heating should be fairly slow, t h a t the initial sign of melting of the glue particles may be noticed, as this is the determining factor.

595

Glues t h a t are a grade apart in jelly strength show a very marked difference in their initial melting points. The poorer grades even slide rapidly down the metal surface, while the better grades melt considerably before a sliding effect occurs. This same procedure may be followed, using the jellies of definite concentrations which have been chilled for a t least 1 2 hrs. Sometimes it is preferable in the case of jellies t o squash them flat on the metal strip with pieces of thin copper, each about in. square, leaving a layer of jelly about 1/32 in. The copper adds weight, and slides quickly a t the first indication of the jelly melting. I n these comparisons, it is essential for accurate results t o keep the operationd on each glue identical, and conditions uniform as t o time, temperature, concentration, etc. The method has proved most satisfactory when conducted with due care. I t 'has the advantage of ease of manipulation, rapidity of determination with ground glues a t least, and only small samples need be used. I t is more positive in its accuracy of the determination of jelly strength t h a n other methods. I n furthering the value of this method for testing jelly strength of glues, i t is possible t o utilize other standards t h a n known grades of glue. Mixtures of petrolatum with paraffin wax having definite initial melting points can be established for glues of higher jelly strength, while mixtures of petrolatum and paraffin oil can be used in the case of lower grade glues. With these mixtures having definite initial melting points, the grades of glue could be more exactly defined as far as their jelly strengths are concerned, and conditions and concentrations standardized for testing, so t h a t results from any analyst would have the same significance. This work should be conducted with enthusiastic cooperation by glue chemists, for the good of all concerned. CRANEAND COMPANY DALTON, MASSACHUSETTS

ON THE INFLUENCE OF THE TEMPERATURE OF BURNING ON THE RATE OF HYDRATION OF MAGNESIUM OXIDE [SECOND PAPER]' By

EDWARD D E MILLECAMPBELL Received April 9, 1918

I n the first paper under the above title1 a series of experiments was described giving the method of burning, a t different temperatures between joo' C. and 1450' C., a sample of pure magnesite and of determining the degree of hydration of the resulting magnesium oxide after treatment with water for periods ranging from I day t o 18 mos. The magnesium oxide resulting from the burning of magnesite had the following composition: Silica, 2 . 53 per cent; alumina and ferric oxide, 2 . 70 per cent; calcium oxide, 3 . 9 6 per cent; magnesium oxide, 9 0 . 7 8 per cent. This burned magnesium oxide required a n . average of 4 4 . 59 per cent of its own weight of water t o com1

THISJOURNAL, 1 (1909), 665-68.