II8
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I i Y E E R I N G C H E M I S T R Y .
is subject t o handling, scraping, and other mechanical injury from being squeezed between the units of the filter press. The filtering requirements of the chemical industry are so many and varied that i t would be absurd t o blindly advocate any specific construction as a cure for all filtering evils. Each problem must be approached with care and each installation properly designed t o suit the needs of the material in hand, and when this is done higher efficiency and lower filtration costs will be realized.
Feb.,
1912
externally by electricity; this apparatus has given such satisfactory results that a description will be of value, i t is believed, to those having t o do with laboratory stills and similar heating devices. D E S C R I P T I O N O F STILL.
The Heater. The heater here described was first furnished t o the writers b y Dr. DavidT. Day, of the United States Geological Survey, and has been used by him in the petroleum laboratory of the Survey a t Washington, D. C. The pattern now used b y the writers has been developed from his spherical pattern, and consists essentially of two pear-shaped halves of heatresisting materials enclosing and closely fitting the retort or distilling flask. The halves hold on their inner or concave surfaces a metal resistance wire, wound back and forth (Fig. z ) for conducting a n electric current. The electrically heated wire warms the flask and its contents, and boiling proceeds quietly, uniformly, and most satisfactorily. Fig. I shows the still in its developed form as used in the petroleum laboratory of the Bureau of Mines. Protectiovt of Still Neck. I
----AN ELECTRIC STILL ADAPTED FOR DIFFICULT DISTILLATIONS.’ By I . C. ALLENAND W. A.
JACOBS
GENERAL STATEMENT.
The process of separating two or more volatile liquids possessing different boiling points b y distillation was one of the first processes employed in the early days of chemistry. With uncomplex mixtures it is comparatively simple. In early attempts a t distillation the volatile substances were heated in earthen bottles over wood or charcoal fires; today they are heated in metal or glass retorts over gas flames. The modern glass distilling flask heated svith a Bunsen flame is too well known to need description here, hut connected with its use are many difficulties that often are not fully appreciated. The continual wavering of the flame, even when it is most carefully screened from air currents, and the variations of the gas pressure subject the retort to uneven and intermittent heating. Such heating causes irregularity in the rate of distillation, and for this reason duplicate distillations do not yield distillates that are alike in quantity and quality. Also, strains in the glass are produced by this uneven heating, and the glass often cracks. With inflammable liquids this is a serious danger. With easily decomposable materials local superheating causes decomposition. Again, the upper parts of the flask are not directly heated, and these chill the vapors and cause them to drip back, thus retarding greatly the progress of the distillation. In the distillation of petroleum, for instance, one has t o do with a material composed of a great number of bodies or rather a number of series of bodies that range from the lightest petroleum ethers t o solid paraffins and asphaltums and form a group, the adjacent members of which so closely resemble one another in their physical and chemical properties, that absolute differentiation is well-nigh impossible with the present methods of procedure. Close adjustment and control of temperature is the prime requisite in effecting the separation of these important series. Dissatisfaction with the ordinarv method of disa heated led the authors to 1 Published from Bulletin of the Director.
19,
Bureau
of
Mines, by
permission
The heaters are wired, that is, the retorts are actually heated, throughout the entire length of the neck; they thus differ from the Day heaters, which are heated only to the base of the neck and have the neck jacketed. The retorts are heated throughout that there may be no cooling or dephlegmatory action in the still neck. If a naked neck is exposed t o the variable air currents of the laboratory, chilling effects are bound to result; consequently, there will he variable condensation in the neck and resultant variations in the distillates. The obtaining of uniform products in the distillation of petroleums depends not so much on rigid adherence t o the exact dimensions of the retort1 and the length and declination of the condenser, though these are important, as on maintaining absolute control of the heating itself. The condenser should be no longer than is necessary for a perfect cooling; a total length of 60 cm. should suffice. K i t h the lighter fractions of petroleums, such as naphthas and gasolines, or with mixtures of alcohols, ethers, or volatile substances in general, a fractionation with a proper dephlegmator gives a better separation than a fractionation without such a dephlegmator, particularly if a long evaporator neck2 is used and is kept at a definite temperature by means of a water jacket or by means of an electrically heated jacket. With the heavier oils, however, the continuous heating that is necessary when a dephlegmator is used causes a degradation in the quality of the distillates. I n distilling such bodies as petroleums consisting of a solution of gasolines, kerosenes, lubri-
*
Redwood. Boverton, “Engler’s Fractionating System.” “Petroleum and Its Products,” Vol. 11, pp 534-6 (1906). Holde, L’lttersuchultg der Milteralole und Fctfe. 1909, pp 20-2. Young, Sydney, “Fractional Distillations.” 1903, pp. 166-8
*
Feb., 1912
T H E JOl -RAYa4 L O F I-YD I-STRI.4 L
cants, waxes, and solid asphaltums, a dephlegmator makes a clean cut separation of the naphthas, but delays the distillation of the kerosenes and absolutely prevents the fractionation of the heavier lubricants. Therefore, to insure uniformity of results, t o remove all oils that boil below their point of decomposition and to compare various oils under like methods of treatment, control of the temperature in the still and the still neck must be maintained in some manner. The thermal influences of variable air currents on the still, particularly if i t is not jacketed (the gas-heated retorts in most laboratories are usually
A simple jacketing of the still neck without special heating gives good results in the case of the lighter fractions-the naphthas and the lighter-burning oils-but does not suffice for the heavy lubricating fractions. As there is little or no lag of the vapors in the still neck-this being of the same temperature throughout and the vapors being driven onward by the continuous and steady generation of more vapors from below-they can not be materially superheated while passing through the neck. Their clean odor, the absence of color, and the non-appearance of carbonaceous deposits a t the base of the still neck, a t least
Vig. 1 .-Petroleum still complete, showing COz generator for driving out water vapors, electric heater, Liebig condenser with water heater. Bruehl receiver x i t h receiving tubes, vacuum pump (Richards) with mercury manometer. electric connections leading to the left to the regulating rheostat (not shown in the cut). To the left is shown a dismantled heater and Bruehl receiver.
without jackets), play an important rble, but are in a measure overcome b y incasing the entire retort with the heater. Moreover, the lag of the distillates in the condenser, particularly noticeable in the case of heavier fractions, must be minimized b y setting the cooler in a nearly vertical position and heating the cooling water as the distillates become viscous and move more sluggishly. I t is therefore considered advisable t o maintain control of the thermal conditions in the retort and the still neck b y warming the neck with the retort itself. Such warming is necessary with the heavier oils in order t h a t the last portion of the vapors may be driven over and not be allowed t o lag in the hot retort under conditions t h a t cause superheating and consequent decomposition.
until the higher temperatures are reached, lead to this conclusion. CONSTRUCTION O F TITE S T I L L .
The Heater. The body of the heater is composed of I O O parts by weight of powdered magnesia, r o o parts b y weight of powdered silica, 40 parts b y weight of finely shredded asbestos, and I O parts b y m-eight of powdered sodium silicate, thoroughly mixed together, dried, and made into a paste with a saturated aqueous solution of magnesium chloride. The use of this paste was suggested b y Mr. A . V. Bleininger, formerly ceramics chemist of the United States Geological Survey. The paste may be molded into any form desired, and after setting over night i t becomes hard and firm enough to endure gentle handling.
I20
T H E JOCRiL'ilL OF I N D G S T R I A L AI1'D E S G I - V E E R I S G C H E M I S T R Y .
To prevent cracking and warping during the subsequent drying, the heater is warmed in an oven a t about 50' C. for 48 hours, after which i t is heated a t 200' C. for several hours t o expel the last traces of moisture. The heater is then hard and sufficiently strong for most laboratory purposes. I t will endure heating, without corrosion of the wire and without injury t o the insulating material, t o the temperature a t which the glass of the retort begins t o soften, the highest temperature a t which the heaters in the Pittsburgh laboratory have as yet been subjected. This heater may be made in any desired form, for heating ovens, water baths, hot plates, viscosimeter
Feb., 1 9 1 2
temperature which is always under the control of t h e operator. M o l d for Heater. The forms for a distilling flask may conveniently be made as follows: Turn from a block of wood EL model the size and form of the distilling flask, with the bulb and neck in one piece, and saw the model into halves from top to bottom. Hollow out the bulb and neck t o form a shallow dipper, making the cavity about I cm. deep, and leaving the rim about 3 mm. thick. Drill holes the size of a small wire brad I cm. apart and 3 mm. below the edge of the rim of the dipper and along the handle or neck as far as i t
Fig. 2.-Electric heater and mold: ( a ) mold after and before wiring; ( b ) completed heater.
jackets, etc., and is most admirably adapted for incasing a still neck or dephlegmator t o maintain a n y desired temperature within, from room temperature up t o the temperatures a t which heavy oils decompose and carbonize.
Wiring the Heater. The manner of wiring the heater for a distilling flask is most important. Wrapping the wires equidistant from one another from the bottom of the flask up to the top of the neck develops too great a heat in the neck and in consequence causes superheating of the heavier oil vapors. Wrapping the wires from the bottom of the flask u p t o the base of the neck results in a dephlegmatory action in the neck t h a t can not be controlled. After repeated trials (shown graphically in Bull. 19, Bureau of Mines), the method of wiring shown has been developed. The method gives a nearly uniform temperature from the bottom of the boiling liquid to the side tube carrying off the vapors-a
is desired t o have the heating wire. Insert brads through these holes from the inside and allow t h e points t o project about 3 mm., thus forming a comb of brad points around the outside of the dipper near the edge (Fig. z ( a ) ) . Then, after greasing well the convex surface of the mold t o prevent the paste from sticking, wrap the heating wire back and forth, the wires being I cm. distant from one another, across the convex surface of the dipper between these points, in t h e manner shown in Fig. 2 ( a ) , tying the wire in place with asbestos cord and making its two ends fast te binding-post projecting a t convenient positions. Lay the wired form, convex side up, on a smooth board covered with a sheet of paper and apply freshly prepared paste t o the desired thickness, about 2 cm. Allow the paste t o harden over night. Remove the brads from the inside of the mold. The heater will then be released and ready for the final drying described above. Do not pass a current through t h e
,
Feb.,
1912
T H E J O U R N A L OF INDLISTRIAL A N D E N G I N E E R I N G C H E M I S T R Y .
wire until the heater has been thoroughly dried, else rapid corrosion of the wire will result. A X/*-liter heater requires about 500 grams of material and 3 . 5 meters of wire, a x/,-liter heater about 1,000 grams of material and 5 . 5 meters of wire, and a I-liter heater about 2 , 0 0 0 grams of material and 1 1 meters of wire. The completed heater is shown a t b in Fig. 2 .
12r
be carefully regulated by means of a n ammeter during the entire distillation. The current required ranges from a very low amperage, a t the beginning of distillation, t o 3 or 3.5 amperes at 2 2 0 volts, as the temperature approaches 3 2 5 ’ C. METHOD OF DISTILLING.
Position of Condenser- Use o j W a r m i n g Oil. The manner of petroleum distillation adopted Wire. in the Pittsburgh laboratory differs from the Engler method‘ in that the condenser is placed in a nearly Any high-resistance wire may be used. In the vertical position instead of a t the usual angle of 7 5 O laboratory of the Bureau of Mines a No. 2 4 (B. & S.) with the vertical. This is done in order t h a t the nickel-chrome wire (the alloy contains about 60 per distillates condensing in the cooler may flow into, cent. nickel, 18 per cent. chromium, and 2 2 per cent. the receivers in the shortest possible time. Care iron) insula.ted with asbestos is used. This wire must be exercised t o prevent the direct falling of t h e has a resistance of about 0.00466 ohm per meter a t drops t o the lower end of the condenser, or the spat2 0 ° c.1 tering will cause loss. When the condensed vapors TEMPERATURE CONTROL. flow quickly into the receivers, the reading of the The current, and therefore the temperature, are thermometers a t the outlet of the retort neck cormost admirably controlled b y a simple water rheostat. responds more nearly t o the temperature of distillnSuch a rheostat hardlyneeds a description as its principle tion of the distillate as collected in the receivers. I n order t o reduce this time factor or lag in the conis so well known. The one used consists simply of a denser, a warming coil is arranged in the water system Io-liter stone jar containing slightly saline water, a of the condenser (Fig. I ) whereby the water can be metal (preferably a carbon) plate a t the bottom of the jar, and a triangular carbon plate t h a t dips into the warmed as the distillates become more and more water and is suspended from a windlass. The plate sluggish. The distillation is continuous2 and not intermittent3 a t the bottom of the jar is connected with one side of the electric circuit in series with the heater, and the and is conducted as nearly as possible a t the rate of triangular plate is connected with the other side of the I drop of distillate per second. This rate is concircuit (Fig. 3 ) . By turning the windlass the upper tinually controlled b y observing the swing of a penduplate may be lowered into the water and theLcurrent lum beating seconds, and carefully regulating t h e current according to this drip. Power
Arrangement of Receivers.
Heater
Ammeter
Fig. 3 . -Arrangement
Rheostat
of electric circuit for heater.
increased. The slightest turn, the lowering of the plate I mm. into the water, effects a change in the temperature of the heater. With a steady current, a n y temperature between the temperature of the room a n d the coking temperature of asphaltum is readily maintained throughout a n y reasonable time of distillation. similarly, a digestion a t any desired temperature may be carried on throughout a period of several hours. Either alternating or direct current may be used, though the former is less destructive t o the rheostat plates. In either case, however, the current must 1
N. J.
This wire may be obtained from the Driver-Hams Wire Co., Newark,
By arranging a Bruehl receiver with I O or more receiving tubes, a fractionation of any ordinary petroleum may be carried on under atmospheric pressure u p t o the desired temperature, say 325’ C., t h a t is, nearly u p t o t h a t temperature a t which t h e petroleum begins t o decompose, thus removing the naphthas and burning oils. Then, b y simply replacing the receivers b y empty ones (the temperature must be dropped t o 1 2 5 ’ C. in order to subject the oil t o vacuum distillation), the distillation m a y be continued under a vacuum up t o near the decomposing point of the petroleum residues, thereby removing the lubricating oils. The distillate cuts are made a t each increment of 25’ C. With a n asphaltum oil this procedure gives 1 5 or 2 0 fractions and leaves a dry, brittle asphaltum residue without causing undue decomposition of the distillates or carbonization of the residue. It is t o b e noted t h a t before the distillation of the lubricating oils under vacuum begins, the receivers containing the naphthas a n d burning oils should be removed, else a considerable percentage of these fractions will 1 Redwood, Boverton, “Petroleum andIts Products,” Vol. 11. pp. 534536 (1906). Holde, Untersuchung der MineralBle und Fette. 1909,pp. 20-22. Ubbelohde. L..“Fixation de la limite d’ebullition du p6trole.” Moni&UT du pdtrole Roumain, March 10, 1908, pp. 279-83. 3 Redwood, Boverton, “Petroleum andIts Products.” Vol 11. pp. 534-36 (1906). Holde. Untersuchung der MznwaZBle und Fette, 1906, pp. 20-22.
*
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERI,VG C H E M I S T R Y .
I22
be volatilized and lost. Also in distilling light oils the receivers should be packed in ice t o prevent volatilization of the lighter portions of the oils during the distillation. Minor Details. I n order that a high vacuum may be maintained successfully, i t is advisable to lute all joints with a thin paste of glycerin and litharge, which rapidly hardens and forms a strong air-tight seal. Very effective litharge, canary-yellow in color, is readily prepared b y heating finely pulverized (200-mesh) commercial litharge or, better, chemically pure lead carbonate, in a n electric furnace a t 600’ C. for 45 minutes. Great care must be exercised in observing the barometric and the vacuum pressures during the distillations, the vacuum should be maintained as near a constant as possible, and corrections should be made for the boiling-point’ if one would duplicate his work. By making use of a good “evaporator neck” either ice cooled or electrically heated, according to the temperature desired, an elaborate refractionation of any desired distillate may be accomplished, and any petroleum may be readily separated not only into its technologic components, but also into bodies having fairly definite boiling points. The rapid removal of the distillable oils from the asphaltum residue is thought to be advisable because the latter is of variable and unstable composition, and continuous heating degrades the quality of the distillates and injures the properties of the asphaltum itself. All distillations are made by weight, the flask and receivers being weighed empty and with contents. This procedure is more convenient and more accurate than measuring b y volume.
Accuracy of Distillatiom. The heaters described are used either for the lightest gasolines or for semisolid bitumens. As an example of the readiness with which difficult distillations can be conducted with the above still, there has been distilled without difficulty a n emulsified crude petroleum, an analysis of which gave the following results : Water.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Distillable oils, atmospheric pressure. ....................... Distillable oils, vacuum pressure of 18 mm. mercury .. , Asphaltum, brittle a t room temperatures. . . . . . . . . . . . . . . . . . . . Sand, etc.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total distillation losses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Per cent. 33.0 9.4 25.8 29.4 1 .O 1.4 100 .o
Again, a petroleum fractionation made in duplicate b y means of this still gave the following results: I. Air fractions (747 mm. pressure): Per cent. 175-200° C.. ...... ;. . . . . . . . . . . . . . 0.9 200-225 O C....................... 14.5 225-250O C.. ..................... 250-275° C....................... 275-300’ C . . .....................
300-325’
C... . . . . . . . . . . . . . . . . . . . .
9 .O 7.6 7.6 8.2
__
Total.. ........................ 47.8 Young, “Fractional Distillation,” 1903, pp. 14-16.
11. Per cent. 0.9 14.1 9.4 7.5 7.6 8.1
__
47.6
Vacuum fractions (31 mm. pressure): 200-225’ C................... 225-250° C . . ..................... 250-275 O C....................... 275-300’ C 300-325’ C.......................
.......................
I. Per cent 2.9 4.2 4.3 5.3 6,3
Total.. ........................ 23 .O Residue.. ........................ 28 .O Distilling loss.. . . . . . . . . . . . . . . . . . . . 1 . 2 Total residue and losses.
...........
-
29.2 100.0
Feb., 1912 11. Per cent 3 .O 4 .O 4.6 4.8 7.0 23.4 27.8 1.2
29 .o 100.0
Prevention of ‘ ’ Buvnping. ’’
In the internally-heated apparatus described by Beckmannx in his account of his boiling-point determinations there is mention of “bumping” when he was boiling anilin (boiling point 184’ C. a t 760 mm.). I n the above externally-heated still there is used in this laboratory, as in many other laboratories, a “boiling stone” of porous porcelain (or of a pumice stone for the lighter oils) weighing about 0 . 5 gram, and no trouble from bumping is experienced even with heavy oils that boil a t temperatures as high as 325’ C. under a vacuum of 2 0 mm. mercury pressure. Richards and Matthews, in describing their internally heated still, remark:2 I n brief, this note (description) shows that while distillation b y means of a wire heated electrically effects a somewhat better separation than the ordinary method and causes much less superheating in the liquid, the gain in efficiency of separation is not always as great as had been a t first inferred from the great constancy of boiling point observed in a special series of cases. This result is to be expected, as conditions are similar t o those of the ordinary method: sharp local internal heating with the small wire coil, as compared with sharp local external heating with the Bunsen burner. In the internally-heated systems of Beckmann, Richards and Matthews, and Bigelow3 only nonelectrolytic materials can be heated. As the authors state regarding them: The short-circuiting through the liquid causes slight decomposition, and this method is not applicable to liquids of this (the electrolytic) type. The internal coil must be wholly immersed in the liquid, else there will be injury to the wire by overheating, with consequent decomposition of the oil, so that distillation t o dryness of even low-boiling liquids is not practicable. The internal coil partially prevents “bumping, ” but a fresh “boiling stone” accomplishes a like result. The boiling stone must be porous and must be fresh. If ebullition once ceases, the boiling stone seems to become less effective. Hence it- is advisable, when practicable, to put a fresh unused boiling stone 1 Beckmann. E., Z . anorp. Chem., 51, 100 (1906); Z. phvsik. Chem., 63, 177-215 (1908);64, 506 (1908). 2 Richards, T. W., and Matthews. J. H . , Proc. A m . Acad. Arts and Sci., June, 1908, pp. 521-24; Z physik. Chem., 64, 120 (1908);J . A m . Chem. S o c . , 30, 1282-84 (1908);Sov., 1909,pp. 1200-2. 3 Bigelow, S. L., A m . Chem. J.,22, 280-7 (1899).
Feb., 1912
T H E J O C R X A L OF I S D I T S T R I A L AiYD E.YGI.\-EERI.YG
in the cooled liquid before again heating the liquid to boiling. The bumping is caused, the authors believe, primarily b y uneven and irregular heating. Once bubbling has started, if i t does not cease b y reason of varying and disturbing external heat conditions, alwa3-s t o be considered in connection with an exposed or a gas-heated flask, or by reason of a diminution of the vacuum when one is used, the boiling will continue quietly. With the residual heavy tars and asphaltums, the internal coils are a source of annoyance in cleaning the apparatus. When oils contain emulsified water, it is advisable during the early stages of the distillation with all forms of stills to slowly pass a dry inert gas, such as CO,, through the still contents, b y means of a capillary tube extending nearly t o the bottom of the distilling flask, until the water is entirely removed. In such case i t is difficult to insert the two internal conductors, a capillary . t u b e and a thermometer through one stopper in the narrow neck of the ordinary distilling flask. To fuse the conductors through the bottom of the flask, though possible, is difficult and hardly practicable for general laboratory usage. I n short, the externally heated still described above is adapted to most, if not all, classes of laboratory work. I n modified form i t may be used for a great variety of heating (devices. The authors rwommend i t to those interested.
CHEMISTRY.
123
in a hundred of the managers, superintendents and engineers of power plants know the true composition of the coal they are using. Not in one plant in a hundred are the products of combustion systematically sampled and analyzed. While i t is undoubtedly true that the majority of power plant managers are aware of the fact that the boiler house is the place where very substantial economies should be realizable, they cannot get themselves to tackle the problem from the chemical side. Yet in the analyses of the flue gas lies the key to fuel economy. Engineers as a class incline to solve their problems b y mechanical means. They prefer t o have something they can get a t with a hammer and cold chisel and monkey wrench. Hence coal weighing machines and water meters, or water weighers, are more readily adopted than CO, recorders.
THE CARBON D I O X I D E R E C O R D E R AS A FACTOR IN FUEL ECONOMY.’ By E. A. UEHLING. Received Dec. 19, 1 9 1 1 .
Vast improvements have been made in power plant equipment during the past quarter century. All these improvements have been along mechanical and physical lines, consequently in the utilization of the steam, which is primarily a mechanical and physical problem, is where the great economies have been achieved. The consumption of steam per h. p. hour has been reduced from that of 2 0 - 3 0 lbs. per h. p., hour less than a generation ago t o 10-15 lbs. a t the present time, whereas in the generation of the steam which is fundamentally a chemical problem, 30-40 per cent. of the heat contained in the fuel continues to go t o waste precisely as i t did 2 5 - 3 0 years ago. The mechanical devices, such as automatic stokers, coal elevators, ash conveyors, forced draft, etc., t o be found in every large and modem boiler house, accomplish a good purpose in so f a r as they increase the capacity and save labor, but do not insure, or even tend to improsre, combustion efficiency. That so little has been accomplished in the reduction of waste in the process of steam generation is due to the fact t h a t the chemist has not been called in t o assist in solving the problem. I t is undoubtedly stating i t very conservatively if we say t h a t not one 1
Abstracted by the author from an address before the Joint Meeting
of A . C . S., S. C . I. and A . E. S . , Xew T o r k , November 10,1911.
Fig. 2 .
Fig
1.
Coal and ‘water weighing machines are a good thing t o have inasmuch as they give definite information as to water evaporated per pound of coal burned. These apparatus are also useful if not absolutely necessary in testing the value of one kind of fuel against another, etc. But since you cannot solve a chemical problem b y mechanical means, coal and water weighers can be of practically no assistance in attaining and maintaining boiler efficiency, which is primarily a chemical problem. Of the heat generated from the coal burned in the average steam plant less then 7 0 per cent., and in
