November 1948
INDUSTRIAL AND ENGINEERING CHEMISTRY
pressure considerably lower than the normal running pressure, while the belt dryer maintains a steady operating pressure set by the capacity of the pumping system. Some materials, therefore, which require a final drying period a t pressures lower than those needed in the first part of the cycle, may be more readily handled in the semicontinuous type of dryer. A variety of heat-sensitive materials such as fine chemicals, plastics, and other food products, have been studied with results substantiating the conclusions reached above. PACKAGING FACl L I T 1ES
d n integral part of these pilot plants was the packaging areas. In each case a vapor-tight room which could be maintained at relative humidities of 15% or lower was required. Under these conditions the powder can be stored for several hours, and handled without undue difficulty from moisture reabsorption. No sperial packaging machinery was required, although i t is desirable to use net-weight filling machines because the container often weighs more than the powder. CONCLUSION
The basic difficulties in the vacuum dehydration of liquid materials have been outlined. There are, at present, limitations on the results t h a t can be obtained. However, the probable process requirements for different materials can be predicted from laboratory measurements. Two vacuum dryers of different characteristics were incorporated in pilot plants to test
2033
the feasibility of processes for drying orange juice and coffee. An analysis of their performance indicates t h a t each has a range where i t is best adapted for use. ACKNOWLEDGMENT
The author wishes to acknowledge the valuable assistance of other members of the company in preparing this paper. T h e data presented were based on work done by R. H. Cotton, A. DiNardo, E. G. Hellier, A. L. Schroeder, and G. A. Schroeder. LITERATURE C I T E D (1) Campbell, W. L., Proctor, B. E., and Sluder, J. G., “Research Reports on Quartermaster Contract Project,s,” July 1, 1944, t o Oct. 31, 1945, Massachusetts Institute of Technology, Food
Technology Laboratories, Cambridge, Mass.
d’Arsonval and Bordas, Bull. assoc. SUCT. dist., 24, 917 (1907). Falk, K., Frankel, E. M., and McKee, R. H., IND. ENG.CHEM., 1 1 , 1036-40 (1919). Flosdorf, E. W., FoodInds., 17, 92 (January 1945). Friedman, S. J., IND.ENG.CHEM.,40, 19 (1948). Moore, E. L., Atkins, C . D., Wiederhold, E., MacDowell, L. G., and Heid, J. L., Proc. Inst. Food Technol., 1945, 160-8. Schroeder, A. L., and Cotton, R. H., IND. ENG.CHEM.,40, 803 (1948). Schwarz, H. W., and Penn, F. E., Zbid., 40, 938 (1948). Tucker, W. H., and Sherwood, T. K., Ibid., 40, 832 (1948). Wernimont, Grant, and Hopkinson, F. J., IWD. ENG. CHEM.. ANAL.ED., 15, 272-4 (1943). R E C E I V ~April D 26, 1948.
LANTS
FLUIDIZED SOLIDS PILOT PLANTS A l t h o u g h t h e main commercial application t o date of t h e fluidized solids technique has been in t h e field of t h e catalytic cracking of petroleum oils, this extremely flexible and versatile technique will undoubtedly be applied t o an ever increasing number of processes. T h e technique is particularly adaptable t o mixed phase (solid-gas) processes requiring t h e addition or removal of large amounts of heat, especially where isothermal conditions are desirable. T h e fluidized solids may be catalysts, reactants, inert heat transfer media, or adsorbents. This article presents t h e steps necessary for t h e practical design of a pilot plant utilizing t h e fluidized solids technique. An apparatus which has been used for fluidization studies essential to unit design is illustrated. T h e translation of data obtained in such equipment t o t h e design of reactors, circulating equipment, etc., is outlined. Recommendations are made as t o t h e design of specific features common t o most fluidized solids pilot plants, such as t h e solids-gas distributors and t h e solids recovery equipment. Methods of measurements and control of temperature, pressure, solids flow rate, and space velocity are discussed.
E. W. NICHOLSON
AND
J. E. MOlSE
E S S O LABORATORIES. BATON R O U G E . LA.
R. L. H A R D Y ESSO LABORATORIES. ELIZABETH, N . J
HE use of the fluidized solids technique first became of general
nique to many other processes and fields soon were investigated on pilot plant and commercial scales. One of the most interesting of these new developments is in the field of hydrocarbon synthesis, and commercial plants employing the fluid technique are planned in the near future. As has been pointed out (5,6, 6),the essence of the fluidized solids technique is the suspension of finely divided solid particles in a rising stream of gas. The gas separates and supports the particles and provides mobility and fluidity for the mass of suspended particles. This mass of suspended particles has the appearance, behavior, and many of the properties of a true fluid; these similarities have resulted in the name fluid being given t o the process. I n applying the fluidized solids technique in different fields, the process role of the solids has varied considerably, as may be seen from the following general classification of these applications :
interest during the recent war when i t was applied extensively in the fluid catalytic cracking process. It quickly be-came apparent t h a t this new engineering technique was a n extremely flexible and versatile one, and applications of the tech-
The fluidized solids may be a catalyst, as in catalytic cracking or hydrocarbon synthesis. The fluidized solids may be a reactant, as in the fluid coal gasification process.
T
2034
INDUSTRIAL AND ENGINEERING CHEMISTRY
The fluidized solids may be an inert heat transfer medium used to increase the heat capacity or heat transfer coefficient of a gas stream, as in the case of the recycle heat exchangers on a fluid cracking unit regenerator ( 3 , 5 ) . The fluidized solids may be an adsorbent, as Cor the removal of water vapor from gases, or the recovery of light hvdrocarbons from refinery gas streams. In order t o apply the fluidized solids technique successfully, it is necessary t o havr a clear understanding of the process and design advantages of the technique, and also to appreciate its limitations. Advantages. Perhaps the greatest advantage for fluidized solids operations is the tremendous capacity for heat ti ansfer afforded by the process. Heat transfer between the gas and the suspended finely divided solid particles is extremely rapid because of the enormous solid s u r f a c e a v a i 1a b 1e . Transfer of heat throughout the fluidized solids bed is almost instantaneousbecause of the extiemely rapid rate of cird a t i o n of solid particles t h r o u g h o u t the bed. Finally, transfer of heat between the fluidized solids bed and a metal heat transfer surface 1s at a very high rate, apparently because of Figure 1 . Noncirculating Fluidized Solids System transfer of heat by conduction betm-een part’icles arid metal wall during the innumerable contacts 01 t,he part ides with the wall. Moreover, the fluidized solids bed provides an ideal situation for direct heating or cooling by injection of gas, vaporizable liquid, or combustible gas or liquid. As a result, t.he fluidized solids technique finds its best applications in processes in vhich large heats of reaction arc involved, such as catalytic cracking and hydrocarbon synthesis, or in which large amounts of extraneous hrat h a w to be provided, as in coal gasification or shale ret,orting. This tremendous heat’ transfer capacity of the fluid bed also leads t o another important advantage-the providing of truly isothermal reaction conditions. For example, the 40-foot diameter regenerator of a fluid catalytic cracking unit operates a t e+ temperature level of about 1100” F. burning 15,000 pounds per hour of carbon Kith a maximum variation of temperatures throughout the bed of less than 20 O F. These isothermal reaction conditions despite large heats of reaction are extremely important for reactions having narroTr optimum temperature ranges, such as hydrocarbon synthesis or hydroforming. The violence of agitation of the fluidized solids bed provides excellent gas-solids cont,acting and solids-solids mixing. Gassolids contacting, of course, is of primary importance in all mixedphase react,ions whether catalytic or noncatalyt’ic, a,nd it is of equal importance in operations in which the solids are involved only in the transfer of heat. Aihigh rate of solids-solids mixing is of great importance also in rnaiiy cases because it means that fresh solids introduced into a reaction zone, or incoming solids a t a much different temperature levei than that, of the reaction sone, are almost instantly complett,ly distributed throughout, the fluidized solids bed. One of t.he outstanding features of fluidized solids operations is t.he enormous quantity of solids that, can be handled without difficulty; this fact,or adds great,ly to the flexibility and continuity of fluidized solids applications. Commercial catalj-tic cracking units, for example, circulate cat,algst from one vessel to the other continuously a t a normal rate of 1200 to 1800 tons an hour. It is these high solids circulation rates that make easy t,he providing of tremendous quariLitirs of heal to reaction zones.
Vol. 40, No. 11
Furthermore, spent catalysts can be removed easily from a reaction zone, revivified, and returned for further use. Spent reaction products or low activity catalyst can be withdraxn cont,inuously from the system and fresh fluidized solids reactants or fresh catalyst added continuously. The ease with which large quantities of solids can be handled leads to additional applications of the fluidized solids technique in which heat effects are not a primary factor. Possible Limitations. Although the fluidized solids technique has great versatility and flexibility, there are several limitations that must be borne in mind. The use of the technique is adapted best to processes which require moderate gas contact times, of thc order of 1 t,o 100 seconds. Operat’ions requiring short contact times for gas and solids are possible, howerer, in high velocity dispersed-phase types of reactors. Operations a t almost any system pressure can be carried out in a single vessel cont’aining a fluidized solids bed-for example, hydrocarbon synthesis is conducted in such equipmenl a t pres-. sures of 200 to 600 pounds per square inch gage. If the catalyst, is t o be circulated a t high system pressure, hon-ever, special precautions must be taken to provide adequate pressure differentials between the various parts of the system, because otherwise small percentage variations in total system pressure might produce pressure surges of the same order of magnitude as the pressure differentials between vessels. Possible precautions include taller standpipes and the use of automatic instruments to provide constant pressure diff erent’ials. Losses of solids also must be considered; this factor beconies important in processes where an expeneive catalyst must be used. Recovery of circulated catalyst is exeellent in commercial units-lor example, in commercial catalytic cracking units, operations are carried out with losses of only 0.00591, of the circulated catalyst. At the high circulation rates required in catalytic cracking, these losses may amount to about’ 0.5% of t’he unit inventory per day.
Figure 2.
Two-Vessel Circulating Fluidized Solids System
Because of the large amounts of solids which may be circulated through a fluidized solids system, the possibility of erosion of unit equipment does have t o be allowed for. Through proper design, however, this problem can be so yell handled that erosion is in no way a limitation on unit operation. Commercial catalytic cracking units have operated continuously for over a gear without shutdown for repairs or mechanical failures. GENERAL BASES O F P I L O T PLANT DESIGN
The purpose of this article is to indicate the bases and the steps necessary in preparing a practical design of a pilot unit utilizing the fluidized solids technique. The information required for a process design arid methods for obtaining the needed data are described. Specific information is presented on some features common to many fluidized solids pilot plants.
November 1948
INDUSTRIAL A N D ENGINEERING C'HEMISTRY CYCLONE BAG F I L T E R S
t
, 1 111
METAL SECTION
,c
p-+ P, + PIHI 144 144 144
(4)
and paHa
P z + - 144
+ -~144 >5 H- i
PIHI +PI 144
It is apparent from the above relations and Figure 5 that each standpipe must be designed to build up enough static pressure to take care of the pressure drop through the slide valve and the transfer line following that standpipe. The determination of the required standpipe pressure build-up on this basis is perfectly straightforward for cases wherein the static pressure at the top of each vessel is the same and static head due t o solids level in each vessel is about the same. However, if, for example, the static pressure on one vessel were to be appreciably higher than that on the other, sufficient additional standpipe height would have t o be provided below the lower pressure vessel t o make up the difference in the vessel static pressures, and most of that added pressure head would have to be compensated for by additional pressure drop across the slide valve on the standpipe below the higher pressure vessel. The relations, Equations 1 to 5, apply in all cases, but Equation 3 cannot be used alone without considering also the other relations. In general, making allowances for special cases such as that brought out in the preceding paragraph, transfer lines should be designed for minimum pressure drop, but allowance should be made for the range of pressure drop that might be expected during the various operations of the unit. The slide valves should be designed for a minimum pressure drop of l pound per square inch a t maximum solids flow rate in order t o reduce the possibility of rcversal of flow through the slide valves. Some additional standpipe pressure build-up should be included over these minimum requirements to provide for satisfactory flexibility of the pilot plant; any excess pressure head available during operation can be compensated for by the setting of the slide valve. Further consideration of these relations Indicates: Unless process requiremenls demand otherwise, the unit should be designed so that each vessel operates a t approximately the same static pressure t o avoid making one of the standpipes _ _ excessively long. A flexible system of static pressure control must be provided to compensate for changes in relat,ive soli'ds level between the two vessels when operatcng conditions are changed. A convenient way of doing this is to fix the pressure on one of the vessels by means of an absolute pressure regulator, and t o operate the valve controlling the pressure on the second vessel by means of a differential pressure regulator operating to maintain a constant differential pressure between the bottom sections of the two main vessels. To maintain smooth flow, extremely good pressure regulation of both vessels is necessary. Instruments capable of controlling the pressures, to within a few inches of water should be provided and these instruments should be quickly responsive to sudden changes in pressure.
It is possible now t o proceed to the quantitative design of the circulating system. Calculations must be made on the static head build-up in vessels and standpipes, on the pressure drop across slide valves and distributor plates, on the static head loss in upflowing transfer lines, and on the frictional pressure drop in transfer lines. Each of these items will be considered in turn, The static head build-up in vessels and standpipes is merely the product of the vertical height of the section under consideration
2037
and the average fluidized solids density within that section. D a t a on the fluidized solids density to be expected in vessels and standpipes at the conditions projected for the pilot plant should be obtained from glass fluidization equipment (such as illustrated in Figure 3) or from other experience. Pressure drops through restrictions and lines are calculated using relations basically the same as thoee used for true fluids. For example, the pressure drop experienced by a fluidized solids stream in passing through a slide valve or an opening in a distributor late is calculated using the ordinary orifice formula, as shown bsow. I n the case of the design of slide valves, [mechanical designs of slide valves for fluid solids operations have been discussed in the literature (5, 7)], and frequently for distributor plates, the pressure drop required is determined by other considerations, as discussed previously, and the orifice formula is used to calculate the necessary opening.
Expressing the head, h, as a function of the AP and the flowing density, this can bc simplified to
AP =
0 . 0 0 0 1 0 8 p ( ~~ v:) C2
(7)
I n calculating the area of a slide valve, this fundamental equation may be readjusted for a dense phase system where the weight of the gas is negligible compared to the weight of solids and where o1 is low compared to DO, so that ua
=
144 W -3600 Ao ps
Then
Ao
W
= 0.000417
c
4 3
(8)
In like nianrier, for calculating distributor grid plate openings, where the volume of solids is negligible compared t o the volume of gas and where 01 may be neglected, the area may be expressed as
where
AP = h =
'
P
=
Pz
=
Pa
=
110
=
111
=
9
=
c
=
AB
w = v =
7 drop, pounds per square inch cad loss in feet of fluid flowing
density of flowing solids stream, pounds per cubic foot inlet density of flowing solids stream (weight rate of solids and gas flow per volume rate of gas flow a t conditions), pounds per cubic foot density of flowing solids stream in standpipe above slide valve, pounds per cubic foot velocity through restriction, feet per second upstream velocity, feet per second acceleration of gravity = 32,2 foot-second per second orifice coefficient (0.6 t o 0.8 for fluidized cracking catalysts) area of restriction, square inches solids flow rate, pounds per hour volume of transport gas, cubic feet per hour a t operating conditions
Transfer lines should be designed for minimum total pressure drop, but several factors enter into the determination. Statio head, which is proportional t o the apparent density of the solids in the line, is decreased with increasing gas velocity but, a t high gas velocities, frictional pressure drop becomes more important. The transfer line velocity should be high enough, however, so that good operability is obtained. This is indicated by a steady pressure drop when solids are flowing. At low velocities, slugwise flow of solids is encountered. At very high velocities, not only is pressure drop excessive, but erosion will be more serious. I n order to calculate the static head loss in the transfer lines, the density of the flowing solids stream must be estimated. The density would be equivalent t o the weight flow rate of solids divided by the volume flow rate of gas except for the fact that some slip of the solids in the gas stream occurs. This slip occurs
INDUSTRIAL AND ENGINEERING CHEMISTRY
2038
Vol. 40, No. 11
because the solids are being supported by the gas stream but tend to fall through the strcam. Hence, the concentration of solids in a rising gas stream is always higher than would be expected if slip did not occur. Similarly, in a downflowing gas stream the solids concentration is lower than would exist with no slip. The slip factor, which may be defined as the actual concentration of solids in the gas stream divided by the inlet or entering concentration of solids in the gay, varies with solids characteristicsdensity, particle size distribution, etc.-and velocity in the transfer line. Therefore, it is difficult to predict accurately for new fluid materials. Some information can be obtained on slip factors in small scale fluidization equipment, such as indicated in Figure 3, but the best data probably will be obtained in the completed pilot plant. The static head of solids-in the transfer lines, p3H3/144 and psHs,flM,may be expressed as S TV H (10) 144
amount of aeration gas injected a t the base of the standpipe is adequate for smooth operation. If long standpipes are required, several gas injection points along the standpipe may be provided, and the optimum aeration rates a t each point may be determined by adjustments during actual operation of the pilot plant. It is usually desirable fiom a process standpoint t o strip out vapors entrained with the qolids leaving a reaction vessel. This is normally accomplished in pilot units by injecting stlipping gas into the standpipe below the reaction vessel. The downaard velocity of the solids stream is generally low in pilot plants so that almost all the stripping or aeration gas injected into the standpipe, will flow upward Furthermore, length to diameter ratios of pilot plant standpipes ale generally high. Both these factors contribute to providing efficient s+ripping in pilot plant standpipes, arid it IS not usually necessary to provide a separate stripping vessel or zone, as is the case in commercial unith.
where S = the slip factor: IV = solids flow rate, pound per hour; H = transfer line height in feet; and V = volume of transport gas, cubic foot per hour a t operating conditions. This leaves the friction losses API and APd to be evaluated In general, conventional formulas for friction drop in pipes are used but care must be taken t o insert the correct expression for density. The horizontal section (horizontal runs should be kept to a minimum) and the vertical section of each line must be evaluated separately because the density of solids flowing in each section will be different. The effect of bends should be estimated and the result in equivalent length of pipe should be added to the straight transfer line length. The entrance and exit losses also should be estimated and carried as a separate item in the final calculation. The basic equation for frictional pressure drop is,
O T H E R DESIGN FACTORS
v
2 f L G2 (144) r s u r e drop, pounds per square inch anning friction factor of the transport gas, dctermined using a Reynolds number calculated from the properties of the gas alone equivalent length of line, feet mass velocity in pounds per second per square foot acceleration of gravity = 32.2 foot-second per second diameter of transfer line in feet actual density of flowing solids stream, pounds per cubic foot = p i s inlet density of flowing solids stream (weight rate of solids and gas f l o ~ per volume rate of gas flow a t conditions), pounds per cubic foot slip factor
AP =
where A€' =
f
=
L G
= =
9
=
D
=
Pa
=
Pi
=
s =
PO D
Using the relations and information discussed, the pilot plant circulating system is designed, usually by trial and error methods. Reasonable standpipe lengths are assumed, slide valve openings are calculated, and transfer line lengths and sizes are set up. These values are adjusted until a satisfactory design is achieved. When a tentative design has been established, it should be checked for maximum solids flow rate by use of Equation 3. All the terms in Equation 3 are known or can be expressed in terms of the weight rate of solids flow; hence, the maximum solids flow rate can be estimated. STANDPIPE AERATION AND S T R I P P I N G
Standpipe aeration and stripping are important problems in the design of commercial fluid solids units, but they are generally of minor importance in pilot unit designs. Aeration of a standpipe is required when the pressure build-up in the standpipe is great enough to decrease the specific volume of the carrying gas initially associated with the solids sufficiently t o create a nonfluid gassolids mixture. Ordinarily, pilot plant standpipes are so short that this compacting effect is not significant, and a nominal
Measurements. I n making physical measurements on fluidized qolids equipment, advantage is taken of the considerable similarity b e k e e n the behavior of a fluidized solids stream and that of a true fluid. In many cases the applications are almost identical in the two types of systems, but some special modifications must be made to allow for the fact that the fluidized solids really represent a suspension of individual solid particles in a gas stream. Tho similarities in methods of measurcment, however, are so great that conventional meters and instruments used in connection with true fluids may frequently be applied directly to measurements in fluidized solids systems. Temperature. The highaheattransfer coefficients obtained with fluidized solids streams make the use of conventional therniocouple wells entirely satisfactory. In some pilot plant installations requiring close temperature meawrement or control, Pyods are preferred to thermocouples inserted in thermocouple wells. Vessel Densities and Levels. A gas passing upwards through a bed of fluidized solids supports and suspends the individual particlcs, and hence does work on the solids. This work, of course, results in a loss of pressure head by the gas. It has been found, howcver, that the pressure head lost under low gas velocity conditions corresponds almost exactly to the weight of solids supportedper unit of cross-sectional area; in other words, the pressure head lost due t o friction is generally negligible. Consequently, the weight of solids contained in a vesspl can be determined e a d y by measuring the pressure difference between the top and bottom of the vessel. This pressure differential is measured through the use of two pressure taps. The pressure differential, and hence the weight of solids in the vessel, can be recorded using a conventional pressure differential recording meter, or the quantity of solids desired in the reaction vessel can be set and maintained through the use of a conventional pressure differential controller. The apparent density of the fluidized solids within a vessel or line can be determined by measuring the pressure differential between two points of known spacing. The two points must be located within the dense solids bed if dense bed density is desired, or both in the dispersed phase if measurements on that section of the vessel are of interest. The interface between the dense and dispersed phases in a vessel can be determined by calculation if measurements on the apparent densities of the dense and of the dispersed phases are available as well as the apparent over-all average density in the vessel. Solids Flow Rates. I n pilot plants in which solids are circulated between vessels, it is generally necessary to obtain some measurcment of the flow rates of the circulating solids. Here again, methods applied normally to conventional fluids may be used in connection with the fluidized solids-for example, Venturi meters
November 1948
I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T'R Y
have been applied successfully in fluidized solids systems. Heat balances also can be employed, but they arelimited t o pilot plants wherein heat losses from the part of the system involved are small compared t o the heat contents of the flowing streams. Where impurities, such as carbon, are continuously deposited and removed from the circulated solids, balances can be made on these impurities. For example, in fluid catalytic cracking, catalyst rates &refrequently calculated using the rate of carbon burning in the regenerator and the change in carbon concentration on the catalyst occurring during regeneration. Heat Transfer and Temperature Control. As has been pointed out, heat transfer throughout fluidized solids beds is at a high rate. This is due to the extreme turbulence of the fluidized solids bed which results in rapid transfer of heat between solids and gas, and transfer of solids from one part of the bed to another. Moreover, the tremendous number of contacts of fluidized solids particles with metal walls which occur because of the turbulence of the bed result in high rates of transfer of heat between the particles and the walls. Film coefficients of over 300 B.t.u. per hour per square foot per F. have been obtained under certain conditions with fluidized solids dense beds. These high heat transfer coefficients may be utiliz8d effectively in the addition or removal of heat in a fluidized solids system. A reactor may be provided with a jacket within which a heat transfer medium may be circulated, condensed, or vaporized. Tubes or coils also may be installed withinfluidizedsolidsbeds for transfer of heat. Furthermore, heat transfer can be accomplished effectively in fluidized solids beds by the direct injection of a gas, vaporizable liquid, or combustible gas or liquid. I n circulating solids systems, as illustrated by Figure 2, large quantities of heat are transferred efficiently by circulating solids from one vessel to the other. For example, in fluid catalytic cracking, catalyst is heated t o high temperatures by burning off carbon in the regenerator, and this hot catalyst, transferred t o the reactor inlet, provides the heat necessary for heating and vaporizing the oil fed, and provides the endothermic heat of reaction.
2039
I n many pilot unit operations carried out at high temperatures, it is necessary t o make up for heat losses by providing extraneous heat to the system. This is done frequently by winding sections or all of the unit with insulated electric resistance wire, or by using strip heaters or Calrod heaters. Gas or condensing liquid in jackets also is used frequently. It is possible also t o design for adiabatic operation wherein careful compensation is made for heat losses. I n this type of design, the vessel is covered with about 2 inches of insulation, and electric heating elements are placed on the outside of this insulating layer preferably over a light-gage metal jacket. Additional insulation then is applied on top of the heating elements. Thermocouples are installed in the plane of the heating elements and also directly opposite each of these couples on the metal wall of the vessel. The amount of power provided for the heating elements then is adjusted so that the temperature difference between each pair of thermocouples is essentially zero. Using the methods and information presented in this paper, it should be possible t o design successfully a pilot plant for almost any type of fluidized solids operation. Fluid catalytic cracking plants have been discussed in the literature ( I , 3, 7 ) , and these descriptions should be helpful in arriving at a practical final design. LITERATURE CITED
(1) Anon., Petroleum Processing, 2 , 518 (1947). (2) Blanding, F. H., and Roetheli, B. E., Oil Gas J., 45, No. 41,
84 (1947). (3) Carlsmith, L. E., and Johnson, F. E., IND. ENG.CHEM.,37, 451 (1945). (4) Daniels, L. S., Petroleum Refiner,25, No.9,109 (1946). ( 5 ) Murphree, E. V., Brown, C. L., Gohr, E. J., Jahnig, C. E., Martin, H. Z., and Tyson, c. W., Trans. Amer. Inst. Chem. Engrs., 41, No. I , 19 (1945). (6) Murphree, E. V., Gohr, E. J., and Kaulakis, A. F., tech. paper, Pacific Chemical Exposition (1947). (7) Trainer, R.P., Alexander, N. W., and Kunreuther, F., IND.ENQ. CHEM.,40,175 (1948). RECEIVED April 26, 1948.
LANTS
Polymerization Units for Thermosetting Resins F.E.R E E S E
AND
ELI P E R R Y
MONSANTO CHEMICAL C O M P A N Y , SPRINGFIELD, M A S S .
P i l o t plant units for thermosetting resins (phenolics, aminoplasts, and alkyds) must be capable of handling a wide variety of resins and t h e design must allow flexibility i n operation. Extraneous features are undesirable, however, because t h e installations may become so complicated as t o require highly trained personnel. Most thermosetting resin reactions arecarried out batchwise i n kettles. Separate kettle installations are recommended for liquid and l u m p resins. Kettle design involves size, material of construction, agitation, heat transfer, openings and valves, and safety. T h e proper choice of t h e condenser, receiver, instrumentation, services, gasketing, general layout, and auxiliary equipment will govern t h e over-all usefulnes of the units.
R
ESINS can be classified into two general groups: thermoplastic and thermosetting. Thermoplastics are fusible and soluble whereas thermosetting resins become permanently infusible and insoluble on the application of heat. The production of the two groups of resins depends very much on the different effects of heat. This article deals only with thermosetting resins. I n general, the production of these resins involves the controlled application of heat t o bring about solution, chemical reaction, dehydration, and, finally, rapid cooling t o prevent overreaction. The installation of a pilot plant unit for the production of liquid or lump thermosetting resins presents a unique problem. A pilot plant resin kettle is rarely, if ever, installed for one particular resin with the expectation t h a t it will be dismantled following completion of the project. Rather, it is installed for