E.T. Guest,

Ass't Technical Services Supt., B. C. Forest Products, Ltd., Crofton, B. C.

Black liquor oxidation has been a subject of interest in the kraft industry for at least 30 years and even now is not a universally accepted practice or a fully proven and perfected part of the process in most mills. The vast volume of work published on this subject through the years and exemplified by the recent excellent review by T. T. Collina (1.), serves to raise the following questions in the mind of the kraft tech nical or operating person.

1. Why do we have so many different techniques employed in obtaining an apparently simple reaction be tween air and liquor?

2. Why have there been such widely different results obtained and published regarding the results and the actual benefits accruing from oxidation of black liquor?

3. Where exactly do we in the coastal area of the Pacific North West fit into the overall picture regarding best types, probable gains, etc?

4. Are we approaching the final development stage in this part of the process?

5. Where does oxidation really fit in the odor control picture?

Discusses Questions

Before going into any details I will attempt to answer these five questions briefly. First, regarding techniques: It would indeed be a massive and confusing process to attempt a description of the pros and cons of all the different oxidation methods. The wide differences stem mainly from the nature of the raw material and to a lesser degree from a tendency toward secrecy fos

Published by permission of the Pacific Coast Branch, Technical Section, CPPA. This won the above group's 1963 Paper Award which was presented on Nov. 14, 1963.

tered by a confused patent situation up to 1959. A program of sampling and testing liquors from mills in the West, Central, East and Southern parts of the continent in 1959 confirmed that aside from the basic inorganic components, black liquor is about as variable a material as can be classed under one name. This then has been one reason for much of the variation in oxidation techniques and explains why a method advantageous in one part of the country can be unsatisfactory in another. Because my detailed experience has been mostly local, and because our wood species gives a liquor more easily handled than most, the emphasis will be on methods shown to be suitable for our area.

The second question, that of vary ing or contradictory results, can partly be attributed to differences in testing techniques and to problems in measurement of chemical relationships through the process cycle. There are obvious problems in obtaining definite answers regarding economic gains from sulphur saving, corrosion reduction, lime saving, or odor abatement. Recent work by the B.C. Research Council and others has helped to clarify the overall picture.

Third, the question of our local position can best be answered by stating that we are certainly in the forefront in practical mill scale application of oxidation. This can largely be attributed to the long term efforts of the B.C. Research Council and the more recent advent of the Troebeck-Ahlen type tower in the Northwest.

Fourth, in regard to the development stage of this process, it is safe to say that we have not seen the ultimate type in mill use, but that with the basic mechanisms becoming better understood it is not likely that properly conceived present day sys

tems will be rendered obsolete. In other words, oxidation is certainly now an integral part in the design of new mills or in mill expansion programs.

The fifth question, i.e. where oxi dation fits in odor control, is one that has been well explored by others. Certainly we do not expect the major gains from oxidation alone that we did 10 years ago. We now know by measurement that where the recos ery furnace in an overloaded condi tion can release several tons of hydrogen sulphide per day, the digester gases may release several hundred pounds. We also know that with fully oxidized liquor followed by a prop erly loaded and operated recovery furnace, it is possible to practically eliminate hydrogen sulphide in the stack gases. When the definite odor reduction in the evaporator condensate is also considered, it is evident that oxidation is an essential link in overall odor control.

Accepted Techniques

Following again the sequence of the introduction we should first look at some of the accepted techniques. Figure I illustrates very simply the configuration of basic types. From the writer's somewhat limited observation the most prevalent (not necessarily in this order) are: the Collins system, the B.C. Research Council system, and the Troebeck-Ahlen. Of these, the Troebeck-Ahlen and the Collins type oxidizes by forcing air through the liquor in a manner to make a controlled volume of foam. while the B.C.R.C. unit operates with air blowing concurrently with the liquor over specially designed plates. Other systems use packed columns with either concurrent or countercurrent air, compressed air in a tank of liquor, or compressed air in a

pipeline; this latter being the newest known technique.

Considering first these less common types, we have the packed tower. This has not been generally successful through the industry for various reasons, mainly channelling and failure of packing. One reportedly successful although undoubtedly high priced installation is reported with stainless steel packing. Performance data has not been released for this type. A new type in the development stage employes free moving polypropolene balls in an upflow column.

Pressure Oxidation

Pressure oxidation in a tank has been successfully practiced for some years at a Southern mill. To our knowledge this method has only been used on strong black liquor, achieving upwards of 95 per cent oxidation. Oxidizing at this location, however, obviates the advantages in reduced evaporator corrosion and sulphur losses in the evaporator circuit. This method does have the advantage, important in the South, of avoiding troublesome foam. It is of interest that in pilot work with Crofton (B.C. Forest Products Ltd.) strong black liquor we simply could not induce enough foam in a Troebeck type unit to successfully oxidize with any real efficiency. Pressure oxidation in a pipeline has been patented in Germany by Schiel (2.) and reported on in a pilot scale by Tirado (3.) in Mexico. We do not know of a mill scale application of this type but from calculated data the horsepower costs for compressed air could be prohibitive. Further, whether the type of foam generated will be easily broken down is yet to be determined.

Collins System

Progressing to the three most prevalent types, Collins, B.C.R.C., and Troebeck-Ahlen, it can be stated that all these do the required job with varying efficiencies depending upon the relationship between size and sulphide loading. Taking these types in order, the first is the Collins. We have had the least experience or contact with this system. Basically it resembles the Troebeck in that air under pressure is blown through liquor, however, there are important differences, one being the use of porous alundum type plates instead of perforated metal. This is intended to give more intimate contact between air and liquor and has been reportedly quite successful in certain applications. Foam handling equipment is generally included

amount of oxidation accomplished for physical installed size because of this characteristic. On the other hand the B.C.R.C. unit has a very low horsepower demand and long term maintenance-free operation. It is unfortunate that without exception, installations of this type on Vancouver Island to date have been under-sized for mill production potential.

The first major improvement put forth by the Council was a change from flat asbestos plate packing to thinner more closely spaced waved surface packing, which has increased the effective surface area per cubic foot of tower to almost double that of the earlier units. A simplification in the liquor feed by a change from stainless steel troughs with crossfeeders, to a shower nozzle system has also been made on the newer units.

The change with the greatest potential, however, is that from parallel operation to series operation. This was first reported by West (4.) in October 1960 in reference to the Elk Falls towers. This is one of the flat plate-type units with a cross-sectional area of 88.4 sq. ft. containing 24,000 sq.ft. per cubicle or per half tower.

In the series operation up to 500 gallons was practically completely oxidized by passing through

(OR TANKS)

FIGURE 2. British Columbia Research Council tower arranged for series operation.

one half of the tower to the sump and then being pumped to feed the second half of the unit as shown in Figure 2. West shows in his data the same phenomenon that most heavily loaded oxidation towers exhibit, namely that in lbs. per hr. oxidizes, a plateau is reached somewhere just beyond the most efficient operation. In this case, an oxidation of 0.765 lbs. Na,S/hr/cu.ft. was the upper limit for both parallel and series operation.

Some Basic Concepts

Before entering into any further detailed description of the two types, B.C.R.C. and Troebeck, it would be advisable to define some basic concepts which are now generally accepted.

1. First, to describe the reaction, would like to quote Dr. F. Murray (5.) as follows "The oxidation of weak black liquor with air cannot be described by any simple chemical or absorption rate theory. The complexity of the oxidation results because the temperature region of practical interest is a transition zone in which the products of the chemical reaction, sulphide concentration, and the oxygen partial pressure act together to influence the oxidation rate." He further elaborates and de

velops a semi empirical relationship to show that in the reaction:

2 NaS+0, No5,0, The rate can be predicted depending upon plate area, feed rate, inlet concentration, air feed rate, and partial pressure of oxygen. This becomes more complex than the intended scope of this paper. A secondary reaction in which methyl mercaptans are oxidized to dimethyl disulphide has also been proven and is believed to result in reduced odor.

2. In regard to test methods for determination of Na,8, several have been used over the years. We cur rently use the B.C. Research Coubcil colorimetric method using paraphenylene diamine for routine use and find it most satisfactory. This test is reasonably specific for sodium sulphide but does not indicate to a predictable degree the presence of polysulphide or mercaptan. For more intensive investigation a potentio metric titration with various refinements has been worked out by Mr. R. Pikner of our staff.

3. The expression percent efficiency should be used with caution as it does not truly define performaner However, we cannot avoid using this term and therefore it should be properly calculated from:

lbs min Na, exit Ibs, min Na,S feed

By eliminating the liquid volume unit in the above we take into account shrinkage due to evaporation. For day to day use in the mill where the evaporation is relatively predictable it is satisfactory to use the commonly accepted:

gm/L in feed - gm/L in liquor out

gm/L in feed

× 100

6. From work by Dr. Murray and other investigators it is established that at temperatures below 160°F. there is a definite tendency to form free sulphur. This can be seen floating on liquor oxidized under these conditions. Murray (5.) found that

4. By far the best yardstick for any oxidation system is lb/min. oxidised at varying feed concentrations and at optimum flow. Starting with this figure various criteria can be developed such as lb/min/cu.ft. of tower, lb/min/lb of air, lb/min/horsepower, lb/min/sq.ft. packing, or lb/ min per dollar investment. Obviously, however, these comparisons should be taken with liquor at a stable inlet concentration, temperature and solids characteristics.

5. The optimum physical flow is generally slightly below that which gives maximum lb/min. oxidized. By optimum in this case we mean the best efficiency obtainable for given capacity. Figure 3, Curve A, is an approximation of what can happen.

efficiency percent.

7. Again quoting Murray, there is a definite relation between feed concentration and oxidation rate, particularly at low concentrations. He shows, for example, at 175° between 2.5 and 5.0 g/l an increase of 25 per cent in oxidation rate. As expected, this gain tapers off as concentration increases. In our mill experience between concentrations of 9 and 12 g/1 we do not observe a predictable change in rate.

8. Within normal mill limits there does not appear to be a relationship between total solids concentration and oxidation rate. However, when

attempting to oxidize at solids above 40 per cent, particularly in a traytype oxidizer (i.e. Troebeck) we did encounter definite problems.

9. It has been demonstrated by Murray (5.) that increase in alkali as NaOH improves the oxidation rate, while increase in thiosulphate inhibits it. This latter is logical considering that thiosulphate is the main reaction product.

10. The question of heat gain ab evaporation work done in oxidation has been widely discussed through the industry. Quantitative data has been presented by Troebeck (6.) Tomlinson (7.), Murray (8.) and Amos (9.) My own conclusion after studying these works is that when the entire liquor cycle including the duplicating the work of West (4.) at furnace is considered we will be

Elk Falls in 1960 but with a different tower packing, a larger tower, and a higher sulphide concentration. Results should prove most useful in finalizing our capacity requirements for current kraft expansion.

Troebeck-Ahlen Unit

Progressing now to the TroebeckAhlen unit I would like to dwell mainly on pilot plant work conducted at Crofton in 1969 and early 1960 with Tore Ahlen. In this co-operative venture we started with a bhe sq.ft. cross-section seven tray conventional design Troebeck tower. We had

measuring devices on air flow, liquor flow, liquid levels in trays and sump, and pressure drops throughout the unit. This was a most interesting project as we had the advantage of using fresh mill liquor and running for 8 to 12 hour periods which enabled us to stabilize conditions and obtain valid information. Figure 4A is a sketch showing the basic arrangement at the beginning of the work and Figure 4B shows schematically the final arrangement which furnished us with the answers being sought. Anyone who has worked on a project of this nature will agree that the work can become both fascinating and frustrating. In the beginning the results were to say the least, discouraging. I will not describe all our problems but attempt to outline the progress of the work. This was in phases according to the type of liquor handled namely: strong black liquor from evaporators, weak black liquor partly oxidized from storage, and raw black liquor from brown stock washers.

In the early trials with strong black liquor we found that an excessive pressure drop across the trays was required before any worthwhile efficiency in oxidation could be obtained. In our first attempts we had the problem of air blow back with the product liquor. This eventually pushed a rain of liquor droplets some 20 feet in the air from the foam box. This problem was overcome by a loop seal with a syphon break. After some weeks of testing and changing feed conditions, etc. we did get oxidation up to 0.3 lbs/min. which amounted to between 60 and 80 per cent. The operation however, was not stable and the air demand and consequent horsepower requirement was high. We simply could not get good countercurrent liquor to air flow, as when all liquor was fed to the top tray even at a low rate, we achieved a negligible oxidation of about .12 lbs/ min. or approximately three per cent. Better results were obtained with feed to the lower trays but here excessive power was used in lifting the liquor. Our conclusion was that this unit in its existing form was not the best arrangement for strong black liquor oxidation.

Next we progressed to reoxidizing liquor from our storage tanks. This product was at 17 to 19 per cent solids and contained 8 to 5 grams/ liter Na,S. Here we had considerably more success. With a flow of 7.5 gal/ min., which was above design for the one sq.ft. tower, and a sulphide feed of .85 to 40 lb/min. we were able to achieve prastically complete oxidation and to maintain stable condi

tions across the trays. Eventually we were able to operate with only one half the pressure drop encountered with the strong black liquor. Problems in achieving balance across the trays were solved by feeding all liquor to the top tray and cutting back the air flow to about 80 c.f.m. Although these results were encouraging we felt that the unit was not being given enough work to do in oxidizing only 0.3 lbs of Na,S per minute.

The final step was to supply weak black which had not been through our mill oxidation system. It was in this stage of development that we first realized the real benefits of recirculation. Figure 4B shows the hook up diagramatically. In its final stages this was a most versatile unit as we could control the amount of recycle to keep the trays properly loaded with liquor and at the same time operate at optimum lbs of Na,S per minute by bleeding in more or less fresh liquor and extracting a like amount of oxidized product. Also at this stage the use of the product liquor to assist in breaking foam was practiced. In actuality foam never a serious problem once the liquor flows were properly balanced and the liquor seal achieved. In this pilot unit the cyclone became almost unnecessary.

was

Summarizing some data from the pilot plant results obtained we have: air at about 5 inches pressure-45 to 50 c.f.m.

Pressure drop total across the trays about 35 inches water column, Pressure drop over each tray about 3 to 4 inches

GPM in and out corrected for evaporation 5.9 ppm

Recirculation-0.6 GPM Solids gain through evaporation 1.0 to 1.3 per cent

Oxidation rate 4 to 55 lb/min or 97 to 99 per cent

In conclusion regarding this project, we did not find any revolutionary short cuts, but I think did establish that this type of tower could be made to perform on West Coast liquor. The oxidation unit now marketed Troebeck-Ahlen bears little resemblance to our final pilot plant arrangement and is also rather different in configuration from the original Troebeck design. Time does not permit going into a full description of the improved model. It is sufficient to say that the following have been incorporated: a more positive pattern of liquor flow in a manner obviating trouble in maintaining pressure drops, an integral foam box with carryover space, a new type of combination cyclone and a foam breaker,

and a simplified square design. Provision for recycle as per our pilot work and/or series operation is also incorporated in most new units.

Black Liquor Oxidation Bonefis

Any article on this subject should repeat the often stated advantages of black liquor oxidation. These then

are:

1. Reduction of odor in evaporator condensate and in recovery stack gas.

2. Reduction in sulphur losses with consequent increase in white liquor sulphidity.

3. Reduction in corrosion of mild steel evaporator tubes and strong black liquor piping.

4. Reduction in recausticizing load due to higher sulphidity giving more sodium sulphide in the active alkali.

Briefly, it can be said that all of these are supported by fact. Only item 1. must be qualified in regard to recovery furnace odor emission by pointing out that this can only be obtained when a furnace is not setiously overloaded and therefore operated with sufficient air. The economic justification for black liquor oxidation depends entirely upon individual mill chemical balance. With mill practices becoming more up-to-date, soda losses have been reduced out of proportion with less easily controlled sulphur losses. In this situation oxidation can at least replace any elemental sulphur and thus show a reasonable return on investment. The justification is further improved by the fact that most methods of sul phur addition are relatively ineffcient, the one exception being the Longview method (as reported by Guide and Graff) but this also requires a sizable investment.

LITERATURE

1. Collins T. T., Paper Trade Journal, 146, No. 30; 39-48 July 23, 1962.

2. Schiel K., German Patent 1,074891 issued January 28, 1980.

3. Tirado A., Guevara V., & Bandun S., Air Pollution Control Assoc. 12, No. 1:34-7, January 1962.

4. West W. B., Tappi 43, No. 10:192A, 194A. October 1960.

5. Murray F. E., Tappi 42, No. 9:761771. September 1969.

6. Troebeck K. G., Pulp Paper Meg Can. 53 No. 8:225-9 (Feb. 1952) "Some Data in Oxidation of Black Liquor."

7. Tomlinson G. H. II, and Douglas H. R., Pulp Paper Mag. Can. 58, No. 4:96-104, (March 1952) "A Progress Report on the Secondary Recovery of Heat and Chemicals in the Alkaline Pulp Mill."

8. Amos, D. La, Can. Pulp & Paper Ind., Vol. 14, No. 11; 35-40. Nov. 1961.