Ind. Eng. Chem. Fundam., Vol. 17, No. 3, 1978
1 = axial distance from start of heating to measuring station p = pressure q = heatflux q 2 = heat flux a t outer surface q o = heat flux with no injection Q = volumetric flow rate a t measuring station Qo = volumetric flow rate a t x = 0 r = radius R1 = radius of inner (porous) tube R2 = radius of outer tube R , = radius of position of maximum axial velocity Re = Reynoldsnumber t = temperature t + = dimensionless temperature defined as ( t 2 t)Cp72/q2GG t l = wall temperature of inner (porous) tube t 2 = wall temperature of outer tube t b = bulk temperature tbO = bulk temperature at the start of heating tbr = bulk temperature in the region bounded by R1 and r U = mean axial velocity U,, = average axial velocity UOav= average axial velocity at x = 0 U , = maximum axial velocity V = mean radial velocity
217
V , = radial velocity a t r = R1 V R , = radial velocity at r = R , x = axial distance measured from upstream end of porous tube y = distance from outer wall y + = dimensionless distance from outer wall defined as
( G I 4Y Greek Letters density 71 = shear stress a t inner (porous) wall 7 2 = shear stress at outer wall TO = wall shear stress with no injection; suffix 1for inner wall, and suffix 2 for outer wall v = kinematic viscosity
p =
Literature Cited Brighton, J . A., Jones, J. B., J. Basic Eng., 86, 835 (1964). Deissler, R. G., NACA Rep. 1210(1955). Deissier, R. G., Taylor, M. F., NACA Tech. Memo. 3451 (1955). Dorance, W. H.,Dore, F. J., J. Aeronaut. Sci., 21, 407 (1954). Olson, R . M., Eckert, E . R. G., J. Appl. Mech., Trans. ASME, 35, Ser. E, No 1, 7 (1966). Yuan, S . W., Galowin, L. S., Jet Propul., 28, 178 (1958).
Received f o r review November 22, 1977 Accepted January 26,1978
EXPERIMENTAL TECHNIQUES
Indirect Measurement of Sedimentation Rates at High Temperature and Pressure by X-ray Photography John G. Ondeyka, Joseph D. Henry, Jr.;
and Frank H. Verhoff
Department of Chemical Engineering, West Virginia University, Morgantown, West Virginia 26506
A settling column has been developed which can be used in conjunction with X-rays to follow the settling of mineral particles in opaque liquids at high temperatures and pressures. The mineral particles attenuate the X-rays to a much greater extent than the liquid. This indirect X-ray system produces photographs showing the solids settling front without disturbing the sedimentation processes. This system consists of an X-ray head and a steel settling chamber with aluminum windows in a lead-lined box. The cell is connected to an autoclave from which the sample is transferred after preparation. The apparatus can be used at temperatures up to 315 O C and pressures up to 2.9 X lo6 N/m2. Data are reported for the sedimentation of mineral particles in a coal liquid with and without the addition of solvent. The experiments are reproducible within 10% error. The X-ray measurement provides both the initial settling rate and the height of the compression region.
Introduction In the coal liquefaction process some coal particles and mineral matter are not converted to liquid. T o utilize the product as a fuel, the solids must be eliminated. Many processes for solids removal are being investigated, such as filtration, centrifugation, and solvent precipitation. Solvent precipitation has been studied in detail by various research groups in determining the settling properties of the particles. Continental Goal Development Corp. (Gorin et al., 1975; Burke, 1976), Argonne National Laboratories (Huang and Fischer, 1976), and Oak Ridge Laboratories (Rodgers, 1975, 1976) employed direct sampling techniques to determine the solids concentration as a function of time and/or height in 0019-7874/78/1017-0217$01.00/0
sedimentation columns. This procedure is very time consuming because of the number of chemical analyses which must be performed. In addition, sampling disrupts the settling phenomena to some extent. The indirect X-ray technique avoids the direct sampling problem and permits rapid measurement so that a wide range of chemical additives can be screened. There are numerous techniques to measure the initial settling rate of solids in clear or translucent liquids. Most of the common light-absorption or light-scattering techniques cannot be used in opaque solutions. X-rays have been used through the years in the medical field as well as in industry. X-rays have been previously employed to determine particle 0 1978 American Chemical Society
218
Ind. Eng. Chem. Fundam., Vol. 17, No. 3, 1978
/---- O u t l e t I
-
ettling
Chamber
o l v e n t Addition
Filters
---
Nitrogen
Tank
I
Holes
Side View
Figure Schematic of the high temperature,pressure sedimentation apparatus.
-
OOOl!
Front V i e w
F r o n t and back P l a t e s body and bolts a r e made o f 51eeI
Gasket material
IS
0 08cm garlock
191
Aluminum u s e d
IS
0 16 c m 2 0 2 4 T 3
Top V i e w
Figure 1. Schematic of the sedimentation cell.
size distribution via sedimentation. The Sedigraph 5000 is one such instrument, manufactured by Micromeritics Instrument Corp., that determines particle size distribution from settling rate data by detecting the concentration of particles remaining at decreasing sedimentation depths as a function of time a t room temperature and pressure. This apparatus could not be used for the conditions of sedimentation for coal liquids because of the high pressures and high temperatures which are required. This paper discusses an X-ray system which was devised to meet severe process conditions. The apparatus involved an autoclave for the preparation of the sample and a settling chamber for X-ray detection of the settling solids. The latter was made of steel with aluminum windows for X-ray penetration. Sedimentation can be observed over a range of temperatures (ambient to 315 OC) and pressures (atmospheric to 2.9 X 106 N/m2). The coal liquid used in the experiments which follow is Solvent Refined Coal (SRC) filter feed. The mineral matter content of SRC filter feed is within a range of 2.6 to 3.0 wt %, according to low-temperature ash studies. The SRC filter feed used in a 3:l ratio of solvent to coal. Experimental Equipment The sedimentation cell is shown in Figure 1.The steel parts of the cell are very strong. The thin aluminum windows are of different thickness for the different pressures. Presently, 0.16 cm thick windows are used in all experiments with a maximum pressure of 1.1X 106 N/m2. The X-rays penetrate the thickest windows (0.32 cm) which are required for the maximum pressure of 2.9 X 106 N/m2. Either Garlock gaskets
or Viton O-rings are used to make a seal between the aluminum windows and the steel body of the cell. All dimensions are given in Figure 1. The volume of the cell is 80 cm3. Use of the apparatus is shown in Figure 2. The settling chamber is placed in a lead-lined box and is connected through the box to a 300-mL Parr Instruments autoclave. High-pressure and temperature filter holders as well as the solvent tank and the settling chamber are connected to the autoclave with stainless steel tubing and high-temperature Hoke valves (0.79-cm orifice). The autoclave has a temperature-controlled heating mantle for constant temperature from ambient to 400 "C. An electric motor is used over a continuous range of speeds from zero to 1000 rpm. The stirrer is water-cooled when run at elevated temperatures. Corrosion-resistant pressure gauges (0 to 2.9 X lo6 N/m2) are used on both the autoclave and the settling chamber. The settling chamber is heated either externally with electrical tape or internally with an inconel heater and the temperature is sensed by four type-K thermocouples, in the body of the cell, connected to a multipoint temperature recorder. The settling chamber is insulated such that there is no obstruction to the aluminum windows. A wet test measures the nitrogen flowing from the sedimentation cell during filling. This gas flow is used to calculate the volume of the liquid entering the cell. In addition to sedimentation, this system has the capability of filtering a sample from the autoclave for characterization of the solids. Two high-temperature and high-pressue filter holders from Millipore Corp. are used in the filtration of the treated coal liquid. This filtration system is secondary to the experimental procedures presented in this paper. A Diano Corp. X-ray system is employed which consists of an X-ray head and a power supply with variable voltage (0110 kV) and variable amperage (0-10 mA) controls, and a reset timer from 0 to 5 min. The spread of the X-ray beam from a 0.25-mm beryllium window covers the area of the cell and the film in the 0.46 m distance used. A lead-lined box was constructed to house the X-ray head, the settling chamber, and the film holder. The film holder, shielded by a slotted lead sheet, is movable so that eight exposures can be taken at eight time intervals on one sheet of film. An E-C Corporation automatic densitometer, employing a constant-velocity film support, is used to analyze an intact 5 X 7 in. exposed X-ray film containing the information of one complete experiment. The densitometer is connected to a Sargent-Welch recorder which plots a graph for each exposed portion on the film. Safety requires that the X-ray be shielded to protect workers. The unit is completely contained in a lead box. Pressure relief valves are located on the sedimentation cell and
Ind. Eng. Chem. Fundam., Vol. 17, NO. 3,1978 219
c I
L1 Y I
51
125
173
280
405
772
552
943
Minuter
Figure 3. Photograph of X-ray negative of SRC filterfeed at 260 "C in the sedimentation cell taken at time intervals shown. Settling interfaces and compression interfaces can clearly be seen
T I M E lminl
Figure 5. Corrected settling curve for SRC filter feed at 260 "C "~DeE."e,260'C.
Sequential M i x i n g :
SSOrPm. 2 m i n : 2SOrpm.8mi"
hi?;*, ra,e Icmih.,
aolventioil retlD
11
36 95 100
h., ,
(I
6
12
18
Id
30
16
42
18
I -
51
eo
66
L
12
HOURS
Figure 4. X-ray YS. visual determinationsof interface height of acid mine drainage.
I
I
the autoclave. Care is taken in handling the chemicals and coal liquids used because the materials may be carcinogenic.
Procedure The sedimentation chamber is pressurized with nitrogen and then heated to operating temperature. During this process the autoclave is charged with coal liquid and brought to operating temperature. Solvent is added unheated to the charged autoclave and mixed for a period of time. The resulting temperature, after mixing of the SRC filter feed with a solvent, will sometimes be lowered depending upon the amount of solvent added. Thus, sometimes the initial temperature of the SRC filter feed will he higher such that, after solvent addition, the mixture will be near the desired temperature. After the SRC filter feed has been mixed with the additive for the prescribed time and mixing speed, it is transferred to the sedimentation cell hy using a nitrogen pressure difference between the cell and the autoclave. After the pressure and temperature in both the cell and autoclave are equalized, the wide throated valves between the sedimentation cell and autoclave are opened. Tank nitrogen is opened to the autoclave at the same pressure. Then the flow into the settling chamber is controlled by the gas valve on top of the sedimentation cell thus avoiding high shear between the autoclave and the sedimentation cell. The amount of
1
2
3 'iHo
4 miniCm-i
5
6
Figure 6.Influence ofsolventhi1ratio on initial settling rate for the SRCIn-decane system. material in the cell i s calculated from the nitrogen effluent measured with a wet test meter which is connected to the valve on top of the settling chamber. The nitrogen volume that was measured was not influenced by any solubility losses to the entering coal liquid. The coal liquid had previously been equilibrated under the same nitrogen pressure a t the cell temperature. After the sedimentation cell has been filled (30 s), the valves are closed and the sedimentation process is monitored by X-ray photography. The time of exposure and the intensity of the X-rays is set to yield the maximum sensitivity in the important solids concentration range. Upon completion of the sedimentation measurements, the contents of the sedimentation cell are forced back into the autoclave and parts of the system are dismantled for cleaning. T6e X-ray system also has the capability of chemical additive addition in the ppm range to promote agglomeration of the particles in coal liquids. This can he accomplished by first dissolving the material in a suitahle solvent, or a chemical additive can he placed in a plug and blown in a t the desired
Ind. Eng. Chem.
220
Fundam., Vol. 17, No. 3, 1978
260'C.
n-Decane, 0 25)1, Sequential M i x i n g : 5 5 0 r p m , z min; 2 5 0 r p m . 8 m l n
I I I
initial rate fcm / h r l I"I*rt.cs
01
Settling Solids
y-20
min
I
2 4 6 S 10 2040 60
min
.:tk,H,;;b 12
b
1
%,
of
1"Ydll.
I"
1,111
S.llllnp
velos,ty
64 7'
-
Sludge Fisc1 forno mof
2
Sedimentstlan
0
,
0
0
10
20
30
min (C)
40
50
0
2
4
6
8
1
0
tino /d)
Figure 7. Interpretation of data from X-ray negative to dimensionless plot.
time with nitrogen pressure. The system also will work with gaseous additives by replacing the nitrogen cylinder with one containing the desired gas. X-ray Measurements After the start of settling, X-ray pictures are usually taken at intervals of 2,4,6,8,10,20,40, and 60 min. If slower settling was expected, the time intervals would be lengthened. The time for the X-ray exposure of each picture is 12 s at 60 kV and 3 mA. This is equivalent to 5 R for each exposure and is five orders of magnitude lower than the radiation exposures to cause increased settling in waste water treatment (Grune, 1963). A combination Plexiglas stand and film holder positions the X-ray film for each exposure. A lead sheet with a 1.27-cm slit placed behind the aluminum window of the cell protects the remainder of the film from being exposed. Kodak AA X-ray film is developed in Kodak X-ray developer and fixed for the prescribed period of time. The film is then washed and rinsed in Photo-flo and dried. Usually the density difference of the settling interface is visible to the eye and corresponds to the densitometer readings. An X-ray photograph of SRC filter feed at 260 "C, shown in Figure 3, depicts the interfacial settling zone as well as the compression zone at different time intervals. (The compression zone could not usually be seen with the eye and densitometer readings were used to detect the top of the compression zone.) An independent experiment was conducted to prove that the visual observation of the interface in a Plexiglas cell was identical with the corrected interface from the X-ray film. Neutralized acid mine sludge was used as a sample material. Data points plotted in Figure 4 prove that the visual and densitometer readings are in agreement. Film Interpretation The film was placed as near the sedimentation cell as possible, yet the results needed to be corrected for the spread of
4
3
2
min/,,
5
6
Figure 8. Reproducibility of initial settling rate for SRCln-decane system.
the X-rays between the cell and the film. The correction for the measured height on the film vs. the actual height in the sedimentation cell is found by geometry. An angle 0 from the point source (pointed a t the bottom of the cell) is identical with the angle of the X-rays through the liquid in the cell. A ratio of the height in the liquid to the height on the film can be obtained to get a correction factor of 0.83. The film height is multiplied by this factor to give the height in the cell. Nine data points were obtained for the settling zone and for the compression zone from the eight densitometer plots (the first data point being the initial height). These corrected points were plotted to find the initial settling rate and the height of the compacted solids. Figure 5 represents an interfacial and compression curve for SRC filter feed at 260 "C. Corrected height is plotted against time for the top curve (interfacial height) and the bottom curve (compression height). The height of the compression region increases with time, as would be expected until the compression zone and the interfacial zone meet. The accumulating solids are loosely packed and further compression occurs after the interface meets the compression zone. Figure 6 shows how an antisolvent affects the settling of solids in SRC filter feed at 260 "C. This is one example of the type of data obtained from experiments where the antisolvent, mixing speed, mixing time, temperature, and concentration were varied. A short period of mixing was used a t high speed to distribute the solvent followed by a long period at low speed to promote agglomeration. Since the same amount of liquid was not introduced into the cell each time, the plots of height vs. time were made into dimensionless plots. All interfacial curves start at unity but have the same slope as the nondimensionless plots. The intercept of the initial slope line with the abscissa is equal to the inverse of the initial sedimentation rate. The sequence of events in this new X-ray technique is illustrated in Figure 7. From the X-ray negative (a), four densitometer plots (b) are presented to show how the interface heights change as a function of time. The interface heights of both the settling and compression zones from (b) are plotted vs. time to give plot (c).A dimensionless plot (d) is made from (c) to normalize the data. Reproducibility The SRC filter feedln-decane system was chosen for the reproducibility studies. Three identical runs were made during the three-month period of experimentation. The initial settling rate using SRC filter feed from the same container in a 0.25:l ratio with n-decane at 260 "C is 36 cm/h for the first run,
Ind. Eng. Chsm. Fundam., Vol. 17, No. 3, 1978 221
32 c m h for the second run, and 33.8 c r n h for the third. These slopes are within the estimated error associated with the calculations. The slope of the line is very critical and a slight deviation in drawing the line willlead to an error of 3 c r n h out of 30 c r n h for a gentle slope and as much as 15 c m h out of 150 c r n h for a steep slope. The initial curves are very close in the first 20 min and are difficult to separate, as seen in Figure 8. Separation occurs when the tangents are drawn to the x axis. In spite of the graphical difficulties, the reproducibility of the n-decane experiments is excellent. Conclusion A new experimental technique has been developed to observe the initial settling rate and the height of the compression region in a high-temperature high-pressure sedimentation experiment by X-ray photography. Time-sequenced X-ray photographs are taken to determine the positions of both the settling interface and the compression region. This experimental technique permits the measurement of high-temperature and pressure sedimentation without direct sampling during the course of the sedimentation experiments. This offers the advantage of not disturbing the settling process and eliminates the need for time-consuming chemical analyses of the mineral solids. The X-ray sedimentation measurement provides both the initial settling rate and the height of the compression region. The latter provides information concerning the maximum recovery of the clarified oil. The technique can be used for other settling phenomena in clear or opaque liquids which require elevated temperature and pressure.
Acknowledgments We wish to acknowledge the Electric Power Research Institute (EPRI) for its support on this project. We also wish to acknowledge the staff of the Wilsonville Solvent Refined Coal Pilot Plant for providing coal liquid samples of the SRC filter feed. We wish to thank K. R. .Vaidyanathan for fruitful discussions. L i t e r a t u r e Cited Burke, F. P., "Assessment of Anti-Solvent Requirements for Wilsonville SRC Power Plant: Batch Studies," CONOCO Coal Development Co. Memorandum RM 14224. Nov 1976. Gorin, E., Kuiik, C. J., Lebowitz, H. E., "De-Ashing of Coal Liquefaction Products by Impartial Deasphalting. I. Hydrogen-Donor Extraction Effluents," paper presented at the Division of Fuel Chemistry, 169th National Meeting of the American Chemical Society, Philadelphia, Pa., April 1975a. Gorin, E., Kulik, C. J., Lebowitz, H. E., "De-Ashing of Coal Liquefaction Products by Impartial Deasphalting. 11. Hydrogenationand Hydroextraction Effluents," paper presented at the Division of Fuel Chemistry, 169th National Meeting of the American Chemical Society, Philadelphia, Pa., April 1975b. Grune, N., Ind. Water Wastes, 8 (54,33 (1963). Huang, H., Fischer, J., "Laboratory Studies for Separation of Solvents from Synthoil Goss Product," Argonne National Laboratory Report No. ANL 76-01, 1976. Rodgers, 8. R., Edwards, M. S., Salmon, R., "Supporting Research and Development on Separations Technology," Phase i Report, Oak Ridge National Laboratory Report TM-4801, March 1975. Rodgers, B. R., "Supporting Research and Development in Separations Technology," in Coal Technology Program Progress Report, Oak Ridge National Laboratory Report TM-5321, Feb 1976a. Rodgers, B. R., Hydrocarbonfrocess., 55 (5),191 (1976b).
Receiued for review June 6 , 1977 Accepted April 17,1978
Pulse-Response Measurement of Thermal Properties of Small Catalyst Pellets Warren E. Stewart;
Jan P. Sirensen, and Bruce C. Teeter
Department of Chemical Engineering, University of Wisconsin Madison, Wisconsin 53706
A pulse-response method is given for measuring the thermal conductivity and heat capacity of cylindrical catalyst particles. The method is particularly useful for particles of small diameter. Results are given for a 1/16-in. platinum-alumina extrudate in a hydrogen atmosphere.
Introduction Existing techniques of thermal conductivity measurement are unsuitable for small catalyst particles. The axial-conduction methods (Sehr, 1958; Masamune and Smith, 1963) need a sample several millimeters in diameter to minimize edge effects. The radial-conduction methods (Sehr, 1958; Mischke and Smith, 1962; Cunningham et al., 1965; Miller and Deans, 1967) require imbedded thermocouples and thus can be used only on custom-made or altered pellets. None of the methods have been used for very small particles, though this might be done by working with composite castings of particles and plastic (Sehr, 1958). Measurements on composite samples are difficult to interpret, however, because of uncertain conditions in the catalyst pores. A kinetic study in our laboratory required measurement of the thermal conductivity of a commercial l/Ic-in. diameter 0019-7874/78/1017-0221$01.00/0
extruded platinum-alumina catalyst. The properties of this catalyst are summarized in Table I. Testing in a hydrogen atmosphere was required for application to the kinetic experiments. The following method was developed to measure the radial thermal conductivity k,, (denoted for brevity as k ) and the volumetric heat capacity p e p without altering the catalyst particles. Experimental Section The catalyst particles (mean radius 0.0715 cm) were placed end to end for a length of about 10 cm in a close-fitting thinwalled stainless steel tube (0.0914 cm outer radius; 0.0758 cm inner radius calculated from the tube mass, outer dimensions, and density). The particles were held in place by retaining springs. The tube was mounted horizontally in an isothermal enclosure; it was attached at each end to gas transfer lines and
0 1978 American Chemical Society
