INDUSTRIAL AND EA-GINEERING CHEMISTRY
122
Vol. 22. No. 2 ~~
Alkali Recovery from Pulp Liquors' By a Chemical Engineering Process C. L. Wagner J. 0. ROSS E N G I X E E R I N G CORPORATION, 122 EAST42ND S T . , T E W Y O R K , S. Y.
T
H E history of alkali recovery in pulp manufacture is a Operation of Furnace very interesting one. Beginning with flat bed furnaces, where the materials were hand-worked during charring, When the furnace has been preheated by means of oil, this was followed by the rotary roaster system, which, on gas, or wood, liquor is sprayed in a t the top or side adjacent the advent of the so-called sulfate process, was extended to to the top. At first the spray is fine, but the temperature include a smelter, which applied only to sulfate. increases rapidly and the spray is made coarser. I n the The process had remained soda furnace smelt flows out static for many years, but in a few m i n u t e s . A i r , in the late seventies some both primary and secondA new process for the recovery of alkali from woodattempts were made at imary, is regulated as required. pulp manufacture is described. The furnace which provements. T h e s e conIn a sulfate f u r n a c e t h e accomplishes this is a vertical, stationary steel strucsisted chiefly of s p r a y i n g smelt is held in until i t overture, with soapstone or other refractory lining, having a flows the dam, when the flow t h e m o r e or less concenroof formed of boiler tubes, water-cooled tile, or sust r a t e d l i q u o r s into preb e g i n s , u s u a l l y one-half pended tile arch, and is connected to a water-tube hour after starting. The h e a t e d c h a m b e r s , with boiler, with induced draft fan and scrubber, or other c a r b o n dioxide content of a u x i l i a r y fuel. This was device, for removing the small amount of chemical the gases governs the adt h e m e t h o d used by Atcarried out by the gases. mission of air, where reguwood, Blackman, C l o u d This unit is almost identical in design for the soda lation both as to point of man, Dornfeldt and Drewand sulfate processes. Its function is to burn the a d m i s s i o n and volume is sen, and others, and all of volatiles and fixed carbons, securing absolute combusmuch closer than in the soda these attempts proved praction, returning the chemicals for re-use in the most furnace. tical failures. In 1911-12 available form, eliminating wastes, labor, dirt, smoke, The outgoing g a s e s a r e H. K. Moore built and operand odors. passed through the scrubated such a furnace successb e r , w h e r e the scrubbing fullv on sulfate. though reliauor circulates from tank q u i r i n g an auxiliary fuel, and later developed the explosion process in which heat was to scrubber, then through a dust trough to the tank and, when it reaches a certain density, by solution of chemicals, supplied by superheating the liquor under pressure. The work on the Wagner process and furnace was begun is returned to the evaporator black liquor supply or to the in 1923 and is a radical departure from all previous methods solution tank, depending on the process. The whole operation is simple and requires only faithful and apparatus. It is the result of an attempt to improve rotary practice in a soda mill, it being observed that the adherence to the principal details in order to secure the ordinary procedure is analogous to that of a stoker where desired results. the fuel was fed a t the rear end and brought forward through Development of Process an atmosphere of carbon dioxide, with the result that volaThe development of the furnace involved many new chemitile combustibles were lost instead of burned. Boilers attached to such equipment rarely returned the steam value of cal engineering problems, and when it is realized that the class of labor used in this part of the plants was the poorest the extra fuel burned. This furnace can properly be designated as a two-zone fur- in the whole mill and that the process must be almost foolnace, as the peculiar requirements of the operation do not proof, the difficulties will be appreciated. Furthermore, the permit the process to be completed a t one location. With major problems were attacked on commercial installations, minor modifications it can be adapted to waste sulfite liquor, which lengthened the time necessary for the work. The picture was further clouded with numerous attempts distillery slops, and other materials. to burn sprayed liquor in the past, none of which, except that Description of Furnace of Moore, had succeeded in operating commercially, and these attempts had induced a wide skepticism in the industry. The furnace is of the stationary vertical type, circular in The field was wholly unknown in all essential details. form, and consists of a steel shell of two diameters, the upper It can now be definitely stated that the reasons for failure part being larger than the lower, with a refractory lining. A were four in number: roof, which is a water-wall extension of the boiler, covers the (1) Parallel current spraying, the heat traveling away from furnace and gas passage to the boiler. Where the feed-water place where the greatest need existed. conditions do not permit this, a water-cooled refractory roof the(2) Failure to burn volatile combustibles, which could not is used. be done by a single air admission, the ignition temperature of the The complete installation includes a water-tube boiler of gases under existing conditions being approximately 920' C., special design, gas and air fan and blower, scrubber, solution and absence of oxygen for complete combustion. (3) Inability to concentrate the liquor to the required density. tank with agitator, liquor pump and scrubber pump, with (4) Use of boilers or radiant-heat-absorption apparatus, instruments for regulation. which did not permit temperatures for complete combustion.
-
1 Received December 19, 1929. Presented a t the meeting of t h e American Institute df Chemical Engineers, Asheville, X. C . , December, 2 to 4, 1929.
Some of the problems which had to be solved are described in the subsequent paragraphs. Each one of these was acconi-
I.i"i"*
.
.
rte.,ining*etivcever*at 100°C. .. .
. .. - . -
...
. .. ..
I N D U S T R I A L AND ENGINEERISG CHEMISTRY
126
giving intimate mixtures, is more satisfactory-an where practice supplants theory,
instance
Temperatures for Effective Performance
The temperatures in this process have certain limiting factors and, since temperatures give an index of the rate of combustion, the same considerations limit the optimum capacity. The limiting factors are loss of soda salts through volatilisation or sublimation and the effect on linings of fused soda salts in gas form or dispersed suspension. The lining is affected by chemical combination and erosion, as all salts are in a fused state, even though of colloidal dimension. Roughly, the limits are as follows: SODA PROCESS
c.
Upper zone Lower zone Emerging smelt
1000-1260 950-1100 860-1100 SULFATE PROCESS Upper zone 1050-1150 Lower zone 900-1050 Emerging smelt 850-1000
' F. 1832-2300 1742-2012 1580-2012 1922-2102 1652-1922 1562-1832
These temperatures were observed with a total-radiation pyrometer, which is the only one we have found to be suitable on account of the soda flame, or inability of the others t o reach the point desired and protect the thermocouple. In a soda furnace the action is so rapid that a liquor containing approximately 6 per cent free caustic soda will give a smelt containing 4 to 6 per cent caustic, with the remainder carbonate. In the soda process the smelt may contain up to 0.4 per cent carbon in the soda, and in sulfate process up to 2.6 per cent carbon, together with varied amounts of ferrous sulfide, in the form of a double soda and iron sulfide. These temperature ranges are well below the boiling points of the soda salts and only sulfide and sulfate are readily volatile in air currents. It is not thought that the vapor pressures developed can lead to dissociation of carbonates or voIatilization of caustic. Sulfides will be converted to sulfates before passing out with the gases. Sulfates are readily collected. Where sodium chloride exists in variable amounts, a large proportion is carried out, greater by far than its proportion to the other chemicals, due to its lower boiling point,. Heat Values
The heat values of dissolved organic matter in black liquor have been the subject of much controversy. Values given range from 8200 to over 10,000 B. t. u. per pound. I t s determination is very difficult, owing to the carbonizing action of caustic soda and volatilization of methyl and terpene products. Some reports of 11,600 B. t. u. are due to almost total carbonization in drying the samples for determination a t 100" C. The figure of 10,000 B. t. u. is probably nearest the true figure for seasoned spruce wood, but the kind of mood and its age and condition will cause wide variations. We have used 8260 B. t. u. in making heat balances, with usually a higher plant yield than is shown by the heat balance supplied or guaranteed on steam production. The normal heat release per cubic foot of furnace volume is from 8000 to 11,000 B. t. u. per hour and should not exceed this figure more than 20 per cent if serious lining troubles are to be avoided. It is now generally accepted that the limit for air-cooled furnace walls in boiler practice is 15,000B. t. u. per hour for sustained performance, and there is no active slagging agent. The typical heat balance on sulfate liquor, showing the factors entering therein, is given below. No ultimate analysis of liquor has been available for checking this heat balance, which is from jack pine liquor. On western hemlock it is about 60 per cent of the above. On soda liquor from 12 to 14 horsepower continuous may he obtained.
VOl. 22, No. 2
H e a t Balance on S u l f a t e Liquor Based on Organic M a t t e r from 2000 P o u n d s of Pulp, Less 10 Per C e n t Washing Loss ANALYSIS: 455 gallons liquor = 2025 Ibs. organic 1645 lbs. chemical 1560 Ibs. water 455 X 11.5 lbs. = 5230 Ibs. Gxoss HEATVALUE: 2025 lbs. X S2600B. t . U. a230 lbs. X (200 - 60' F.) Y 0 75 sp. ht.
5.1. u . = 16,726,500 = 550,000
Total Losses: Flue gases: 2025 Ibs. X 15 lbs. air = 30,400 Ibs. 30,400 Ibs. X (500' 60' F.) X 0.24 sp. ht. Water: 1560 Ibs. X 1250 B. t. n. = latent and super heat Reduction: 400 Ibs. NazSOd X 3000 B. t. u. Heat of fusion: 1645 Ibs. X 64 B. t. u. Water of combination, 207, (superheating only) Radiation loss, 7.5% (range 5 to 10%)
-
Total loss Gross heat. . . . . . . . . . . . . . . . . . Losses. ....................
N e t . .......................
B . 1. u. 17,276,500 7,878,200
= 17,276,500
= =
3,207,200 1,951,000 1,200,000 105,000 120,000 1,295,000
=
7.878.200
=
= = =
9,398,300
9,398,300 970 B.B t.. u. t . u. = 9660 Ibs. steam % !!-!k= 280 hp, 34.5 lbs. 280 hp. 11.66 hp. continuous 24 -= hours Note-Losses should be about 10 per cent higher, due to sulfur products i n liquor and heating chemicals. No definite data on hand.
Capacity
This item depends on so many factors that it is only possible to give a general statement as to performance under known conditions. Commercially, it is necessary to have a capacity that is safe under all conditions and then allow for overrating operation. Technically the matter is not so simple. During the preliminary work, first on rotaries and then on a small experimental unit, it was found that a certain cubical content would allow a certain number of gallons of liquor t o be burned per hour or minute; when this was exceeded operation difficulties developed. We find that this figure holds good a t present, although an increase in total cubical content allows a slightly greater amount per volume unit to be burned. For the sake of simplicity, capacities are based on burning the liquor from a certain number of tons of pulp. Under best conditions, by the soda process, the organic matter to be burned represents 50 per cent of the bone-dry weight of wood, and in the sulfate process, 45 per cent. This figure varies widely in less modern plants and is dependent on the following: pulp-washing losses; gallons of liquor per ton; density of the liquor; proportions of organic and chemical; allowable (chemical) loss; permissible amount of chemical which can be handled by the boiler scrubber system; life of furnace linings. We advocate the use of a liquor which does not require steam-jacketed piping to keep it fluid. Specific gravities ranging from 1.32 to 1.36 will generally be satisfactory, while too low a density introduces other complications. Results
The most recently completed large installation of this furnace is a t the Union Bag and Paper Power plant a t Tacoma, Wash. It consists of five units, with provisions for a sixth, each unit being entirely independent. The liquor from the pulp washers is concentrated to the required density in a 28,000-square foot quadruple evaporator, with a forced circulation concentrator as a standby. A so-called "trash" pump circulates the concentrated liquor to the feed nozzles. Fuel oil is used for preheating after Sunday shutdowns. Each furnace is shut down in rotation for examination and minor repair, one always being a spare. The blow steam from the digesters is passed through surface condensers, the noncondensable gases being carried to the secondary air supply of the furnaces. Evaporator gases and gases given off by mill effluents are trapped and
Ih-D USTRIAL AA'D ENGINEERIKG CHEMISTRY
February, 1930
127
labor savings, upkeep, and production of chemical very high in sulfide have exceeded all expectations. The plant, designed for 120 tons daily capacity, has consistently exceeded this by from 10 to 20 per cent.
taken to the same source. The result is an odorless pulp mill, which has not been accomplished before in this country. The steani produced has approximated very closely the theoretical shown by a heat balance, while the recovery,
Performance of Economizers in the Drying of Pulp and Paper' Frederick W. Adams and Charles M . Cooper SCH001.
OF
CHEXXCAL ENGISEERISGPRACTICE, hIASSACHUSETTS INSTITUTE
OF
TECHNOLOGY, CAMBRIDGE, MASS
A series of tests has been made to determine the of itself furnish much Inore Ry1n-G with air is One of t h e i m p o r t a n t efficiency of a Briner-type economizer in recovering the than enough heat to preheat chemical engineering heat used in the drying of P U ~ Pand paper. From the the incoming air. I n t h i s results of these experiments it is concluded that a case, unless the economizer processes involved in the production of pulp and paper. recovery 50 to 60 per cent is good practice with this is used to heat air for other economizer when the heat is used only for the drier purposes as well, the heat The heat required to accomr e c o v e r a b l e will be only a system; but this efficiency may be increased greatly plish this drying is a major item in determining manuif there is a demand outside this system for large small fraction of that availfacturing Costs, particularly amounts of tempered air for ventilating purposes. able. The economic savings Possible by the use of this This paper presents the rewhere the product has a low equipment are shown by the calculations of heat results of a series of tests perselling price per unit weight. covery and costs in a hypothetical case. formed by students of the Thus it is necessary to evapoS c h o ol of Chemical Engirate between 1 and 3 pounds neering Practice on a typiral economizer installation. of water per pound of pulp or paper produced. I n drying, relatively large quantities of fresh air displace Procedure for Tests hot. humid air, which is discharged, carrvina away as sensible Hot, humid air is removed by a hood from a 60-ton pulp heat a considerable proportion-of ' the h e 2 supjilied to the process. To recover this heat by preheating the fresh air, drier and passed through a Briner economizer, where heat is cross-current economizers have been commercially employed transferred to fresh, outdoor air, which is blown back onto for nearly twenty years (1, 2 ) . Their application to dry- the drying pulp sheet. When desired the incoming air may ing in the pulp and paper industry has recently become more be further heated by steam coils before reaching the pulp general with the development of the Briner-type economizer. machine. A by-pass is provided which allows some of the The heat-recovery efficiency of an economizer is intimately incoming air to be drawn off for ventilation purposes a t other connected with the process in which it is a part. I n some points in the building. A diagrammatic layout of the econocases a large proportion of heat may be recovered. This mizer installation is shown in Figure 1. The tests were conducted to cover a range of outdoor is true wherever hot air is used directly as a drying agent without other sources of heat. The outgoing air may be temperatures varying between 13" and 74" F. Measureused to preheat entering air and, if in the process it is cooled ments of incoming and outgoing air flows were made nith
D
T a b l e I-Results of T e s t s on T y p i c a l E c o n o m i z e r I n s t a l l a t i o n (Area of heating surface = 10,500 sq. ft.: free area for air flow = 55 sq ft RUN 1 R U N2 R U N3 RUN4 Temperature before, E;. Temperature after, ' F. Humidity, Ibs. water per Ib. air Pounds water-free air per hour Pressure drop, inches of water H e a t picked up, million:; B. t . u . per hour OUTGOING AIR: Temperature before. E . O Temperature after, F. Humidity before, Ibs. wzter per Ib. air" Humidity after. Ibs. water per Ib. air Pounds water-free air Der hour Pounds condensate pet hour Pressure drop, inches of water Sensible heat delivered I J ~humid air. million B. t. u. per hourc H e a t delivered b y condensation. million B. t . u . per hour a Calculated. b Calculated by difference
13
71 0.001
207,000
....
2.8; QI. Q
20 75 0.001 186,000
31 76 0.002 200,000
2.44
2.15
...
....
39 83 0.003 179,000
.... 1.8s
0.0261 236,000 2296
0.0371 0.0303 222,000 1505
112 94 0.0337 0.0290 226,000 1055
0.0396 0.0335 209.000
0.40 2,47
0.83 1.61
1.03 1.12
0.66 1.22
92 0.0358
....
to the entering-air temperatures and a t the same time most of its water burden is condensed, an almost perfect recovery of heat is theoretically possible. Such conditions do not exist, however, in the ordinary drying of pulp and paper, where steam condensing in drier rolls furnishes the principal source of heat. The hot outgoing air carries with it all the mater vaporized in the process, which, if condensed, would 'Received December 11, 1929. Presented a t t h e meeting of t h e American Institute of Chemical Engineers, Asheville, N. C., December 2 t o 4, 1929.
111 96
...
....
111 44
R U N5
_.
OD
ti
74 101 0.014
82 0.009 166,000 0.24
127,000
1.08
0.84
104 91 0.0300 0.0280 203,000 415 0 .li 0.64 0.44 ~~~
....
liu\
....
12s
113 0.0417 0.0413 203.000 79 '0'76
o os
pitot tubes and an anemometer. Air temperatuies and humidities were obtained before and after the economizer from wet- and dry-bulb readings or by determining den points. Stratification rendered impractical direct measurement of the temperature and humidity of the outgoing air before the economizer, necessitating the calculation of these values. The water removed by the drip line was measured to determine the amount of condensate: In run 5 pressure drops resistance Of the economizer to air to flow \yere measured on differential manometers.
