SOME ASPECTS OF FACTORY OPERATION AT THE
Proceedings of The South African Sugar Technologists' Association — June 1970 105
FIGURE 1: Illustration of a Cippoletti weir.
If Q is expressed in metric tons per hour while H and L are given in centimetres the equation for the flow of juice (S.G. 1.06) over a Cippoletti weir is-
O = 0.06795 x L x H1-1
The product 0.06795 x H1-5 was calculated for different values of H and these values used to con-
SCALE A SCALE B
^OUTLETS
w?a
SCALE A
SCALE B
O
WEIRS
QUIESCENT ZONE
o I- o
NLET
FIGURE 2: T.S.C. juice distribution tank.
FIGURE 2A: Details of scales in Figure 2.
struct the (scale A) which indicated the height of juice above the weirs. Hence the product of this reading and the weir opening in centimetres read from another scale (scale B), gave the quantity of juice passing through the weir.
Details of appropriate H/L ratios, free ventilation and chamfered edges, etc., required in the design of such a box can be found in any standard hydraulics or engineering handbook. The arrangement of the Tongaat installation is depicted diagrammatically in Fig. 2.
A check on the accuracy of these weirs was made on a number of occasions by comparing the calcu- lated flow through the weirs with the recorded flow from an orifice plate. The agreement was within
1-2%.
A problem which was found with the weir box was a build up of sand which was such in the latter part of the year that it was necessary to flush it out once per week. To facilitate this operation it is desir- able to have the bottom sloping towards the drain outlet.
Clarifiers
Although all the heated juice at Tongaat passed through a suitable flash tank the Bach clarifiers had flash tanks which were attached to each unit. One of
these had been removed a few years ago when some test work was being carried out on this clarifier and it had functioned satisfactorily since then without the flash tank. The other Bach flash tanks were re- moved during the past season and no problems have been experienced as a result. It should be mentioned that the emergency overflow of the Bach clarifier was located in its flash tank, but in view of the rela- tively large diameter of these clarifiers such an over- flow was considered to be unnecessary under pre- sent circumstances. The main advantages of their removal were the reduction in maintenance and pro- vision of extra space on the clarifier floor. It was noticed thai the juice entered the clarifier at a slightly higher temperature after they had been removed.
T.S.C. operated two 24 ft. Rapidorr clarifiers and in the past it had been found that one of them had a marked tendency to retain caked mud in the mud compartments. A close study of the units was made in order to find the cause of the difference in per- formance. Two features that were noted were—
(a) different rates of rotation of the scraping gear; and
(b) juice take-off pipes in the one clarifier were set about 20 cm too low.
Although these two features were modified to bring the characteristics of the unsatisfactory clarifier into line with the better one it was not successful in achieving the desired improvement.
After seven weeks of operation in July/August the mud in the lower mud compartment had built up to the level of the inspection port. On the other hand, no significant mud build up occurred during the last 16 weeks of operation. Analysis of juice samples from this clarifier after a week-end shut-down showed a greater increase in R.S.R. than did any of the other clarifiers. Some comparative data of the R.S. ratios after a week-end shut-down were:—
5.26 and 7.12 6.36 and 10.16.
Juice Retention in Clarifiers
Trials run in Bach clarifiers at 55 and at 80 tons juice per hour seemed to indicate that there was less loss of sucrose with the lower retention time of 1.7 hours compared with 2.4 hours. This conclusion was based mainly on the change in R.S.R. from clarifier feed to clarified juice. As a guide to what happens in the clarifier it is sug- gested that this comparison is more meaningful than the change in R.S.R. from mixed juice to clarifier juice which is sometimes used as a rough indicator.
Filter Station
The retention of the filter station was known for some time to be low and attempts were made to improve it. These filters had been altered some years ago in such a way that the feed entered through two pipes in the bottom of the tray and the overflow was taken off from the one end instead of along the length of the filter.
Observations showed that the sprays were im- pinging too strongly on the cake and also delivering too much water. This resulted in erosion of the cake and dilution of the bagacillo/mud mixture in the filter tray. Consequently retention was being re- duced in two ways, viz., by—
(a) washing off cake which had been picked up by the filter; and
(b) diluting the feed to the filter which has often been shown to result in lower retention.
New finer sprays were fitted and this did lead to improved results. Using water at a pressure of 3.8 bar absolute, these sprays delivered 200 ml/minute over a circle of diameter 20 cm when the spray was set 20 cm from the filter drum. However, it is still possible for the operator to add too much water on the top of the filter with the above-mentioned effects, but the answer to this is considered to lie in proper training of the operators.
A further aspect of filter operation which was examined was the setting of the cut off and applica- tion of vacuum to the panels of the filter drum. It was found that for a particular filter different panels appeared to cut off at different heights above the mud discharge point while some panels did not appear to break vacuum completely. It was only towards the end of the season that the cause of this unsatisfactory behaviour, viz. eroded or incom- plete brass strips separating the panels, was found.
The leaks occurred between the drum and the strip as well as on the screen side of the strip. Some leakage of vacuum also occurred on the strips seal- ing the ends of the panels. It is expected that the necessary attention to this will result in a further im- provement in filter operation.
One other modification which seemed to have been well worth while was the fitting of steaming pipes along the length of the filter drums. These were set approximately 25 mm from the drum and con- sisted of 1.5 in. n.b. pipes with 1/16 in. diameter holes drilled at | in. centres. These were used to steam-clean the filters using 7 bar gauge steam once every two days and examination has indicated that after a full season's operation the screen and com- partment behind it were still in a satisfactory state of cleanliness" for filter operation. Consequently it was not necessary to use the caustic soda cleaning treatment which had proved so successful on a pre- vious occasion.
Pan Boiling
Attention is drawn here to one aspect of the operation of continuous centrifugals curing C masse- cuite, viz. the effect of irregular size grain. While a continuous machine can cope much more readily than a batch machine with a boiling containing all sizes of grain there can be appreciable loss of small crystals through the screen. Because of this, particu- lar attention must always be paid to the grain regu- larity of the C boilings. At T.S.C. it was found that the regularity of boilings was not satisfactory and investigation showed that a major cause of this was the introduction of fine crystal in the B molasses and more especially in the wash from the double curing
June 1970
Proceedings of The South African Sugar Technologists' Association 107
of C sugar. The presence of the exceptionally large quantity of crystal in the wash was a direct result of crystal breakage in the continuous C fore- curers. A very noticeable improvement was achieved by dissolving this grain through the use of thermostatically controlled silent steam blowers on the pan suction tanks.
Stirrers were also employed so that localized over- heating, which would cause caramelization, was minimized and also the feed to the pans was main- tained at a uniform temperature. By heating the liquor in the suction tanks only, the amount of de- composition attendant on this practice was further minimized.
Discussion
Mr. van Hengel (in the chair): Tongaat uses steam blowers to improve the position regarding crystal in wash.
I think the present system of storage tanks for B-molasses should be reconsidered. These have been replaced at Mount Bdgcombe by one large vertical tank, which pumps continuously to a blow-up tank and continuously overflows so that solution of fine particles is taking place.
It may be that the difference in reducing sugar ratio between limed and clarified juice should be investi- gated before the difference between mixed juice and clarified juice but what do these differences really convey to us?
Dr. Graham: A major reason why Tongaat operates as at present is the belief that decomposition can be minimised by holding molasses in storage at as low a temperature as is practicable.
Hence, only the molasses in the pan feed tanks is heated to dissolve fine grain.
The reducing sugar differences referred to are used as an indicator of chemical changes in the clarifier.
Dr. Matic: Is it the intention to measure pH at the boiling point of the juice?
Dr. Graham: No, because we do not know how to clean electrodes at boiling point with the ultrasonic cleaner.
The other problem is that the electrodes simply do not last, whether they are being cleaned or not.
Dr. Matic: Dr. Parker mentioned a German make of electrode which would stand up to these condi- tions.
Mr. Cargill: Did Tongaat solve the problem of differences in clarifier capacities?
Mount Edgecombe last year installed an automatic flushing system for their electrodes.
Dr. Graham: Although we have modified the clari- fiers and are now using a reliable measuring device which controls the flow of juice, we are still having trouble and are getting an uneven build-up of mud.
Mr. Renton: Empangeni did at one stage introduce mechanical wiping of electrodes but this is no longer carried out.
Mr. Chiazzari: I am very interested in Tongaat's clarification system.
Regarding filter stations and the scaling up of the backing screens we have found that we cannot leave our screens for a week before cleaning them and we try and purge them every eight hours.
In connection with fine grain in wash, our centri- fugals are fitted with a device operated by a timer, to feed water into the monitor casing.
By diluting the molasses in the casing, a smaller pump is required and much of the dissolving of fine crystal is affected.
We put low pressure vapour into our tanks and we seldom have trouble with microscopic grain coming back.
Dr. Graham: Our after-curers were continuous machines but we did not add water and often the wash looked more like massecuite than molasses.
Regarding our filters, an operator on each shift is responsible for two and he normally cleans one machine every second day.
In connection with clarification, the pH of the clari- fied juice is governed by two factors: (1) it should be minimum otherwise the quantity of final molasses is excessive; (2) we do not want an increase in re- ducing sugars from clarified juice to syrup, which is indicative of sucrose hydrolysis.
Mr. Bruijn: This morning we heard that the pH must not be too high because of decomposition of reducing sugars and because more molasses would be made.
Now we are told not to keep the pH too low be- cause we will increase reducing sugars in the syrup.
There is no exact point where the one reaction stops and the other starts.
Dr. Graham: A compromise is necessary when de- ciding the pH at which the juice must be controlled.
This value should be such that the destruction of both sucrose and reducing sugars is minimised.
Mr. van Hengel: The construction of our evapora- tors must be able to cope with the conditions which are optimum both ways and there is need for an intensive investigation.
Mr. Jullienne: Regarding bleeding of filters, it is important that they be properly cleaned at shut-down.
TRIALS ON B.M.A. CONTINUOUS CENTRIFUGALS
By G. G. CARTER
The Tongaat Sugar Company Limited The Tongaat Sugar Company installed 10 new
B M.A. K850 continuous centrifugals at the start of the 1969/70 season as a replacement for 19 obsolete batch type C machine used as forecurers.
The appreciable savings in maintenance and operational labour which were anticipated were realized, as well as lower final molasses purities.
This paper deals with the experience gained while operating these machines during the 1969/70 season and it must be stressed that these results are based on the difficult C massecuite conditions in S.A.
Description of the Machine
The B.M.A. K850 (Fig. 1) is a continuous machine comprising a vertically mounted conical basket spinning at 2,200 r.p.m. The angle of the cone is W and the top and bottom diameters are 850 mm and 360 mm respectively. The gravity factor is 2,265 measured at the top of the basket diameter, at full speed.
The drive is by means of a 30 kW electric motor mounted behind the centrifuge and connected to the basket spindle by means of five vee belts. The basket has no flywheel but requires a 10-minute wait to come to rest should the machine be switched off.
Lubrication of the two main bearings is done by a separate oil pump, which in case of failure switches off the main drive and a pilot lamp (see Fig. 1).
The distribution cone is elaborate and consists of an acceleration cup plus feed cone whereby the massecuite is both fed on to the screen and also lubricated with steam and water prior to the massecuite being purged on the screen.
The type of screen used for the curing of low- grade massecuite has slots 0.06 mm wide and 2.2 mm long and was composed of nickel with a hard chrome surface.
Massecuite was fed from a static finned reheater only capable of reheating 0.86 m3h—l per machine from 40°C to 55°C with water of 60°C. The unit was
KBS0
1199(A)
BMA
TO 201 - 2 0 - 0 2
FIGURE 1
RfffflWPHPHPiWffl' WH*WJPWWSPfPPmP«
Proceedings of The South African Sugar Technologists' Association — June 1970 109
unsuitable for the quantity of massecuite to be re- heated since the heating surface was well below requirements.
Massecuite flow from the reheater could be con- trolled by an automatic diaphragm valve. The con- trolling impulse comes from the power taken by the drive which is proportional to the feed rate. Because of fluctuations in the electrical supply to the machines these valves opened and closed ad lib. and were also electrically unreliable. The net result was that manual operation was adopted.
Wash water could be added through sprays at the apex of the basket as well as mid-basket whilst water could also be added inside the distribution cone to provide lubrication of the feeding massecuite (see Fig. 2).
Steam injection is provided inside the distribution cone to warm the massecuite (see Fig. 2).
Inspection of the sugar is possible on the top of the monitor casing and of the molasses by means of an inspection plate on the molasses drain.
Investigational Aspects
A programme of work was drawn up and the following points were studied:—
(a) The effects of the temperature and quantity of water and steam added in relation to the sugar quality and consequently machine throughput.
(b) The effects of the temperature and quantity of water and steam added in relation to the molasses purity.
(c) The effects of crystal size on the sugar and molasses purities.
(d) The effects of massecuite purity and its vis- cosity effects on machine throughput and molasses purity.
(e) The effects of massecuite temperature and its viscosity effects on machine throughput.
(f) The reliability of the machines with respect to screen life, electrical problems and com- ponent life, etc.
Thermometer thermometer thermometry termtmetro
Durchflu/bnengenme sser flow meter
dibit metre medidor de paso misuratore di quantta
FIGURE 2
Results and Discussion
In compiling the results the average conditions from a series of experiments have been quoted and the various points dealt with in the order mentioned under investigational aspects.
(a) Tests to see the effects of the quantity of water on sugar quality and molasses purity.
1. From visual tests it was clearly seen that water added to the feed cone was essential to serve as a lubrication medium to the molasses film adhering to the sugar crystal and the sugar would not cure with- out it. The water which could be added by nozzles to wash the purging sugar on the screen was shown to wash the sugar through the screen if used in any quantity, or raised the final molasses purity, and consequently was never used.
2. An interesting feature of the tests was that the ratio of lubrication water to massecuite increased in a non-linear fashion as the throughput of massecuite was increased from 0.71 m3 to the maximum which was of the order of 1.23 m3h—\ This is shown in Table I and Graph 1.
TABLE I
The effects of massecuite throughput on the amount of water needed to effect a sugar purity of + t~---
m3 of Mass.
per hour 0.72 0.87 0.99 1.15 1.23
Amt. of water Ratio of water in litre/hour
26.0 37.0 60.0 81.0 94.5
Mass. * 1.0 1.2 1.7 2.0 2.2
Purity of C sugar
86.7 85.8 84.7 84.8 84.3
* The ratio of water/mass, is calculated in this column using 0.72 cubic metres as a base.
The high ratio of water needed at the upper range of throughput resulted in a dilute molasses which was below the target refractometer brix of 86.0 set to prevent the operators adding too much water and thus washing excess sucrose into the final molasses.
1 '"
re IIO s< *
D M
0 - 7 O * - ' ' • •— • ' l-O IZ /••* / • « / • » 2-0 3-2
RATIO litres/ hour m3/hour
G R A P H 1: Showing effects of massecuite throughput on the amount of water needed to produce a sugar of 85.0 purity.
3. The effects of the quantity of water on sugar purity and molasses purity were found to be such that:—
(a) A point can be reached where the sugar purity will not increase much above a certain level however much water is added.
(b) The excess water merely serves to dilute the molasses brix to below an acceptable level and increased the difference between nutsch and final molasses purity.
(c) The increase in difference between final molasses and its nutsch purity became most marked with rising use of wash water when the sugar purity exceeded a level of 85.0.
TABLE ri
The effects at 0.92 mah-1 (constant throughput) of increasing the rate of water addition on sugar purity molasses brix and purity rise.
Rate of water Rise in Nutsch addition in 1/h Sugar Purity Molasses Brix difference
40.0 83.7 87.5 1.4 80.0 85.8 84.3 1.7 120.0 86.3 80.6 2.3 160.0 86.8 76.8 3.1
(b) Tests to see the effects of water temperature on sugar quality and molasses purity.
A series of experiments were conducted to see the effects of water temperature on sugar purity and the tests clearly showed that water at 80° C was better than at 60° C or 40° C in that less water at the higher temperature was needed to produce a superior quality C sugar.
TABLE III
The effects of wash water temperature on the throughput and purity of C sugar
m* of Mass Average Temp, of Purity Amount
per Purity the wash of the of wash Molasses hour of C Mass water °C C sugar water 1 /h Purity 0.99 58.5 80 84.7 46 38.6 0.97 58.6 60 84.1 55 37.9 0.99 58.8 40 82.7 66 38.3
(c) Tests to see the effects of steam on sugar quality.
Steam, at Tongaat, was found to be necessary to improve the quality of sugar produced at through- puts above 0.72 m3h—l. Attempts were made to see what steam pressure was most beneficial. Whereas J. Chen and workers in Peru showed an improve- ment in sugar quality between 1 and 2 atmosphere steam, at Tongaat no difference could be detected and the lower pressure steam was used. The quan- tity of steam added was increased by using a 6 mm copper pipe instead of a 3 mm copper pipe and the generally hotter atmosphere gave visually im- proved sugar quality. It was further seen that the lower the purity of the massecuite the more beneficial was the steam in producing higher purity C sugar by most probably providing better viscosity reduc- tion. This latter observation has not yet been con- firmed by quantitative measurements. With increas-