Practical Observations in the Use of Phosphoric Acid and Activated Carbon as an Aid to Colour Removal in Refining

c) Sub-samples of the hourly clear iuice sample were com. . An attempt was made to enhance colour removal in the raw posited over four hours and anaiysed for phosphate conhouse and refinery under the abnormal conditions of the 19831 tent as in (a) and (b) and for colour according to the 84 season so as to maintain an acceptable colour quality in the ICUMSA 420 method. final refined sugar product. Phosphoric acid was added to the (d) Two-hourl~ samples of the and fine liquor were \ mixed juice to improve the defecation process in the raw house analysed for colour according to the ICUMSA 420 method. while activated carbon was used in the carbonatation process (e) Finally, all refined sugars were analysed for colour also to increase colour absorption. according to the ICUMSA 420 method. It must again be stressed that the collection and sampling of Introduction these products were done by the existing laboratory shift staff working under a revised his should be borne in The process used for raw sugar manufacture at ML is the mind when reviewing these results for comparison with similar conventional hot-liming method followed by a 3-boiling system using B-magma as a footing for A-massecuite. The A-sugar trials at other factories. quality is entirely colour orientated as 100°/o refining is practised. This is done using the carbonatation/sulphitation proExperimental Procedure cess, followed by a 4-boiling system The sugars from these The additional analyses were started in week 14 of crushing four stages are blended at the centrifugal station to give a final and by week 20, sufficient data had been gathered on the preproduct of acceptable quality. vailing conditions to be used as a standard for comparison. An important factor of such a refining process is that a morePho~phonc acid was then added to the mixed juice to mainor-less consistent colour removal can be achieved although the tain a 300 to 320 ppm P?O, level in the juice. A diaphragmcolour input may vary considerably. t v ~ e dosine ~ u m ~ was used and the undiluted acid was added Consequently experience has shown that when the A-melt i i io the biffkr tank after the scale. The feed of the juice into colour exceeds a certain point, the fourth sugar will probably this tank caused sufficient turbulence to ensure adequate mixing have to be rejected due to too high a colour, with the resulting, of the acid and the juice. A sample point on the outlet of the loss in recovery and factory thermal efficiency. pumps removing the juice from this tank was used to take catch The 1983184 crushing season brought with it new factors samples to reveal the increased P?O, content. The addition of which greatly increased the possibility of severe colour probacid continued until the end of the SeaSon. lems. Although the major cause was poor cane quality due to The problems of excessive colour in the refined sugar did the drought, a further factor was introduced when ML undernot appear as early as had been expected probably because: took to refine 22 600 tons of Swaziland raw sugar. This contract (a) it is common practice at ML to stop using B ~ ~ ~ ~ ~ for necessitated that the refinery be run at maximum capacity for A-massecuite footing when A-sugar colours start to rise almost the entire season and there was also some doubt as to and may affect the colour of the refined sugar. This was the anticipated quality of this sugar. done in week 21 and for the rest of the'season, B-magma In the light of these developments it was decided to attempt was not used as footing; to maintain optimum colour removal by: (b) ML was under contract to refine raw sugar from Swaziland. (a) improving mixed juice clarification by maintaining optiAlthough it was initially felt that this raw sugar would mum phosphate levels through the addition of phosphoric present a colour problem, it was in fact of excellent quality acid. This was also tried and reported by Carter;' and helped to improve colour in the mill. This contract (b) increasing colour absorption in the carbonatation process was completed in week 25. through the addition of powdered activated carbon to the Discussions with suppliers of activated carbon and conA-melt. sumers in similar industries revealed that a powdered type would be most suitable. It was decided to add this carbon at the raw Analytical Procedure sugar melting stage to ensure adequate retention time for colour absorption prior to filtration. These retention times in practice Many of the analyses required to monitor the above proworked out to be cesses were not done on a regular basis in the mill laboratory. from melting to lime addition 20 minutes Furthermore, colour analyses of the various products were not comparable and this necessitated a standard procedure to be from carbonatation Saturators to filtering 55 minutes. adopted and implemented. To meet these requirements using During filtration, the carbon was removed together with the existing laboratory facilities the following changes were made: calcium carbonate and finally formed part of the wasted refin(a) Sub-samples of the Central Board's hourly mixed juice ery sludge. This had no adverse effect on filter performance. sample were composited over four hours and analysed for The feeding device for the powdered carbon consisted of a phosphate content using the standard SASTA procedure.? supply bin of 0,17 m3 capacity mounted above a variable speed (b) Addition of phosphoric acid was made after weighing and vibrator feeder. An adjustable door fed the desired quantity of hourly catch samples of this juice were analysed for P20, the powder from the bin onto the feeder. After leaving the content as in (a). feeder, the carbon dropped into a tundish fitted with water Proceedings of The South Afiican Sugar Technologists' Association June 1984 55 sprays and was washed into the feed of the A-sugar melter. Two problem areas were revealed: the first being that the angle of the sides of the bin were not steep enough and disrupted the flow and secondly, that any air movement at the feeder resulted in powder being blown away in every direction. After the initial run it was decided to dose the carbon by hand directly onto the wet sugar in the screw conveyor feeding the melter. The advantage was that the powder adhered to the wet sugar and consequently very little was lost in the air. Although not tested, the best method of feeding would probably be an enclosed variable speed screw from a supply bin which would dose the carbon directly into the side of the wet sugar conveyor feeding the melter. Laboratory trials with various carbon powders under simulated conditions indicated that 0,05 to O,lO/o on brix in A-melt would reduce colour to an acceptable level. The carbon chosen was one having a density of 360 to'380 gm/ll and a surface area of 950 to 1 100 m2/gm. Results of Phosphoric Acid Addition The data collected until week 20 and three weeks thereafter, i.e. after all analytical changes had been carried out are shown in Table 1. The improvement which had been brought about in the clarification process by the higher P,O, content, unfortunately did not lead to an improvement in the clear juice colour. However, the improvement in the settling and handling quality of the mud was evident and there was an immediate improvement in the pol % filtercake (see Table 1).