INDUSTRIAL AND ENGINEERING CHEMISTRY
March 1952
589
gage) over fixed-bed undiluted mixed metal sulfide catalysts to
TABLE VII.
700" F.
PLUS
RESIDUAL Fu EL YIELDAND QUALITY less than 1 weight % ' nitrogen with an estimated catalyst life of
Ht consumption, cu. feet/bbl. orude shale oil Residual fuel, vol. 96 on crude shale oil Residual fuel ins eotions Gravit 'AP? Flash (Pensky), F. Pour F Vis./122"F., SS Furol, sec. Sulfur wt. % Nitroien, wt. % Conradson carbon, wt. % Ash wt. % Sediment, wt. % N.B.T.L. heater fouling test
0
69
600 63
900 56
15.2 24.2 20.1 455 110. 105 106' 210" 0.99 0:09 0:05 2.37 1.59 1.69 7.7 0.04 0.15 Stable
.... ..
.... .. ..
1400 54 26.1 480 110 30" 0.01 0.44 0.7 0.04 0.17
a Extrapolated from viscosity at higher temperatures. Direct determinations not reliable due t o presence of waxy material a t this temperature.
hydrogenation, but was not completely satisfactory even a t the highest level of hydrogen consumption employed. It is believed t h a t this poor color stability of the hydrogenated products was associated with incomplete removal of nitrogen compounds. Rather than go to higher hydrogen consumption levels (with increased degradation in boiling range) for additional improvement i n color stability, this was obtained in the case of the most highly hydrogenated product which was inspected (1400 cubic feet per barrel) by the use of a light acid treatment followed by caustic neutralization and rerunning. This treatment resulted in only about a 4% treating loss due t o the high degree of saturation of the hydrogenated product. By means of this light acid treatment, satisfactory color was obtained for the Diesel fuel fraction (Table VI), but still somewhat poor color for the gasoline and kerosene fractions. More severe acid treatment of these fractions was not investigated, as these presumably could be blended off satisfactorily without further treating. CONCLUSIONS
Total crude shale oil as obtained by retorting of oil shale in KTU-type retorts by the Bureau of Mines, after removal of suspended insoluble ash constituents by filtration, can be continuously hydrogenated at high pressure (3000 pounds per square inch
about 6 months. Total hydrogenated product yield in all cases was over 100 volumetric yo. Sulfur removal was satisfactory (90% or more; sulfur in product down t o 0.1 weight % or less) a t all hydrogen consumptions investigated. Nitrogen elimination appeared t o be more difficult, and improved as the hydrogen consumption increased (30, 60, 80, and 95% nitrogen elimination at 600, 900, 1400, and 1800 cubic feet per barrel of hydrogen consumption, respectively). Residual nitrogen content of the higher boiling fractions appeared generally to be higher than the nitrogen content of the lower boiling fractions. The characteristic odor of the raw shale oil was reduced by hydrogenation, particularly when the hydrogen consumption was above 1000 cubic feet per barrel. Color of the total hydrogenated product improved from black for the raw shale oil t o a deep brown-black a t low hydrogen consumptions and t o a light yellow a t higher hydrogen consumptions (1400 to 1800 cubic feet per barrel). Improved color stability and additional reduction in nitrogen content were obtained by the use of a light acid treatment with little loss in yield (4%) because of the relatively high degree of saturation of t h e hydrogenated products. LITERATURE CITED
(1) Bell, H. S.,"Oil Shales and Shale Oils," New York, D. Van Nostrand Co., 1948. ( 2 ) Blackwood, A. J., and Cloud, G. H., S.A.E. J o u r n d , 46, 49-53 (1940). (3) Brown, C. L., Voorhies, A., Jr., and Smith, W. M., IND.ENG. CEEM.,38, 136-40 (19413). (4) Clark, E. L., Hiteshue, R. W., Ibndiner, €I. J., and Morris, Boyd, Ibid., 43,2173-8 (1951). (5) L a m p , 31,27-31 (January 1949). (6) McKee, R. H., "Shale Oil," New York, Chemical Catalog Co., 1925. (7) Mills, G. A., Boedeker, E. R., and Oblad, A. G., J . Am. Chem. Soc., 72, 1554-60 (1950). RECEIVED for review April 20, 1951. ACCEPTED October 31, 1951. Presented before the Division of Petroleum Chemistry a t the 119th Meeting of the AMERICANCHEMICALSOCIETY,Cleveland, Ohio.
Control of Small Liquid Flows Using Glass Valves CHRISTIE J. GEANKOPLIS' AND A. NORMAN HIXSON University of Pennsylvania, Philadelphia, Pa.
N OBTAINING experimental data on a liquid-liquid extraction tower ( 1 ) 1.5 inches in diameter it was necessary t o con-trol closely the flow rates with l/a-inch needle valves. The liquids employed were aqueous hydrochloric acid solutions and isopropyl ether. Because of the extremely corrosive action of these materials, suitable metallic materials of construction are scarce. Monel and stainless steel valves were tried unsuccessfully. Materials such as Hastelloy, which might be suitable, are difficult to machine and form. As a result glass seemed t o be the most suit.able material. It is desirable t o control the flow by adjustable area devices which are constructed of glass or other corrosive-resistant materials. Present available laboratory devices are stopcocks and pinchcocks. The majority of these small glass valves have very 1
Present address, Ohio State University, Columbus, Ohio.
little throttling action. Stopcocks were unsuitable for a number of reasons: 1. The valves contain grease. 2. They d o not have a sensitive or sufficiently positive throttling action. 3. T h e valve setting often drifts over a period of time. Pinchcocks which are used in conjunction with rubber tubing are also unsatisfactory because of reasons 2 and 3. Hershberg and Southworth (3) designed a needle valve for the microdetermination of nitrogen. The main body of the valve was made of glass and the opening at the seat was a section of capillary tubing. The spindle or needle was a thin stainless steel rod, t a ered t o a point a t one end and fastened t o a narrow knob at t i e other end for rotary movement. The packing gland was a short piece of rubber tubing inside a steel bushing. The bushing was threaded so t h a t
590
INDUSTRIAL AND ENGINEERING CHEMISTRY
turning t h e spindle would throttle t h e flow. Mercury on top of t h e packing was used as a seal t o make the valve stem gastight. Gibson (2)described a small, all-glass valve used in regulating liauid flows of 0.1 t o 10 ml. p& minute or steam flows below 100 grams per hour. The needle tip had a groove that coincided with a groove in t h e seat of t h e valve, allowing flow of liquid through this opening. Rotating the stem and needle tip 180 degrees closed the groove and caused the flow t o stop. The packing was asbestos string packed around the stem. Both these valves used a packing and seal which are unsuitable for use with certain acids or organic solvents. As a result a glass valve was designed which substitutes a removable neoprene diaphragm or tube for the seal or packing. CONSTRUCTION OF VALVE
K h 4
Vol. 44, No. 3
This valve may be used under pressures greatei than atmospheric and in cases where transparency is an asset. Tests showed that a t the operating pressure of 8 pounds per square inch no leakage of liquid occurred through the packing or past the seat of the valve when closed. Flow measurements were taken a t various intervals during runs while using the valve t o maintain a constant rate. The liquid feed to the valve flowed from a b-gallon carboy to a n overflow tank which maintained a constant head on the valve. Over a period of 2 hours when the flow was 300 ml. per minute, the maximum deviation of the flow was *2 ml. per minute. S o drifting or gradual changing of the flow \vas apparent in any of the runs. As it is often desirable to have a valve with no i n t e r n a l surfaces rubbing against each other, the valve in Figure 2 was designed. In this design seizing or accumulation of deposits between adjacent sliding surfaces is avoided.
Figure 1. Glass Valve with Neomene Seal
The glass needle valve, which employs a neoprene seal, is shown in Figure 1. This valve embodies the major characteristics of ordinary 1/8-inch metal needle valves which have satisfactory flowcontrol characteristics. The external parts A , B , C, H , and R, were machined from Bakelite. Turning the valve adjustor, A , which consists of a circular disk, causes t h e threaded portion, R , and the glass stem or spindle, K , t o move up or down. The guide, C, slides up or down along the two guide posts, B, and keeps the spindle in a fixed track. The glass spindle, K , is connected t o t h e valve adjustor assembly by a steel pin, D . A small clearance of 0.0033 inch, E, and the pin, D, form a universal joint uTith limited possible movement. The tube, J , which serves as a guide for the stem, and the valve body, P , are heavy borosilicate glass,. TKOscrews, I , hold the collar, H , firmly t o the valve body. The liquid enters a t L and leaves a t M . The neoprene tubing, *V,serves as the seal or packing joint and forms a tight seal a t the stem, F , and the guide tube, G. Upward movement of the valve stem causw t h e neoprene tubing to stretch a t Ar and downward movement causes a bulge. The flexibility of the neoprene tubing does not affect the flow or cause drifting because the throttling is done only at the seat by the spindle. The spindle, K , wae ground into the glass seat with Carborundum dust by slowly rotating t h e spindle in its track. The hole was drilled in the glass spindle a t D by using a jeweler’s drill and drilling very slowly. The neoprene tubing is easily replaced if it deteriorates from action of solvents. However, in several months of operation replacement wa8 not necessary, even though the tubing did sirell to some extent. Materials such as Tygon and rubber can be used in place of neoprene.
A
I
The supporting guide or track, B , for the stmn is external to the valve body and is not in contact with the liquid, Parts A , B , and C can be made of Bakelite. The neoprene t,ubing, D, is sealed a t E and F. The guide block, B , and the collar, C , a r e attached firmly to the two flat su ports, A , by screws. TRe remainder of the valve is the same as in Figure 1,
Figure 2. Glass Valve with So Internal Surfaces Rubbing against Each Other
I n most cases the liquids being used in the laboratory are clean and relativelv free of Eediment. I n these cases the valve in Figure 1is t o be preferred because of its greater simplicity and greater rigidity in holding the spindle in the track. LITERATURE CITED
(1) Geankopliu. C. J.,and Hixson, A . N.. IND.ENG.CHEX. 42, 1141 (1950). ( 2 ) Gibson, G. P., J . 9 o c . Chem. I d , ,58,317 (1039). ( 3 ) Herahberg, E. B., and Southworth, L., IKD. ENG.C H E MAx.4~. , ED.,11, 404 (1939). . 4 C C E P T E D NO>.eIllher $5, 1951. RECEIVEDfor review May 1, 1951.