Proceedings of The SouthAfrican Sugar Technologists' Association - June 1986

CONTINUOUS CENTRIFUGAL SCREEN EVALUATION

By P. L. M. VERMEULEN, C. MACK and R. RAMPHAL
C'G. Smith Sugar Limited - Umzimkulu

Abstract To measure molasses flow rate, a receptacle of known

volume(Figure 1)wasinsertedinto the molasses outletpipe,
The performance of Western States continuous centrifu- and the time taken to fill this was noted.
gals usingvarious open area screens on Band C massecuites
has been monitored over five seasons at Umzimkulu Mill. Volume container 6.920£
Three different screens from three different manufacturers,
witha 10%, 12% and 14% open area were tested.Significant
higherthroughputs(up to 44 %) have beenattained at similar
machine settings. In general final molasses brix was higher
and apparentpuritylowerwithlarge openarea screens. Screen
life was between 6 and 10 weeks with no modifications to
the backing screens. Mol Receptacle

Introduction
Historically, the conventional screen used on C centri-
fugals is the 0.06 mm, 7 % openarea screen and the 0.09 mm,
10% open area on B centrifugals.
The first 12 % openareascreen wasinstalled at Umzimkulu Centrifugal
in July 1981, and since then numerousevaluations withother
makes have been conducted. Because ofthe higherthrough-
put it was decided in 1984 to fit 0.06 mm, 12% open area
screens in all Band C centrifugals. The specifications of the
continuouscentrifugals used at Umzimkulu are the Western
State, 34° basket with 4 mesh, interwoven flat top, backing
screen. Mol Drainage Pipe
!
In 1977, at Empangeni Mill, Princeand Montocchio, test-
inga 14% open area screen found similarincreased through-
puts, at generally higher molasses brixesand equal or lower
molasses purities. To support the thinner screen, a woven
10 mesh, stainless steel screen was fitted in addition to the
existing standard backing screen. It is interesting to note that
the slot length of this 0,06 mm screen is 1.02 mm. FIGURE 1 Receptacle of known volume inserted into molasses out-
let pipe
Large open area screens mentionedin this paper have slot
lengths of up to 2.80 mm. The longer slots could be for
reasons of screen material thickness or improved molasses Procedure
drainage. Before a series of test runs was started, the screens and
the basketswerewashed and inspected. Afterthe basket had
Method gained its maximum speed, the water rotometer was set at
The method adopted for the evaluationwasby comparing a predetermined setting. The water temperature ranged be-
the performance of two centrifugals, keeping all parameters tween 80 and 85T. Steam valves were fully opened. The
similar except for the screen used. The performance of the machines were then slowly fed with massecuite.
centrifugal is evaluated from its capacityand by the quality On a continuous centrifugal the water percent massecuite
of the molasses produced. decreases with increasing massecuite throughput. If the water
Massecuite flow rate was determined by pol balance from is kept consistent and the massecuite throughput is in-
molasses flow rate, correctedfor process water added. How- creased, a point will be reached where the emerging sugar
ever no account could be taken of steam addition or willbe belowstandard. Maximumcapacity was obtainedby
evaporation. maintainingmassecuite throughput as highas possible with-
out reaching this condition.
M = (L - W) (Ps - Pc)/(Ps - Pm) It was imperative to keep the sugar quality the same for
Pc = Pr/(l - W IL) both test machines. (A 1,0 variation in sugarpurity leads to
a 0,1 difference in molasses purity). Afterthe machineshave
Where: M Massecuite flow rate Kglh beenregulated to givesimilarsugarqualityand are operating
L Molasses run off flow rate KgIh at optimum capacity, they are run for a further ten minutes.
W Water flow rate Kglh Thereaftersugar, massecuite and molasses were sampledand
Ps Pol % Sugar molasses flow rate measured. As far as possible sampling
Pc Corrected molasses pol was done simultaneously on both machines.
Pr Measured molasses pol After a series of tests was completed, the screens were
Pm Massecuite pol interchanged between machines and further tests were run.
47
Proceedings of The South African Sugar Technologists' Association - June 1986

Results and Discussion

Test 1
12 % open area screen (0.06 mm) versus 7 % open area screen (0.06 mm)

Screen pattern (1 cm 2) Screen pattern (1 cm 2)

Open Area 12% Open Area 7%

Number of slots: 77 Number of slots: 68
Length of slots: 2,60 mm Length of slots: 1,70 mm
Width of slots: 0,06 mm Width of slots: 0,06 mm
Distance between slots: Distance between slots:
Horizontal: OAO mm Horizontal: 0,56 mm
Longitudinal: 0,30 mm Longitudinal: 0,80 mm
Thickness of screen: 0,225 mm Thickness of screen: 0,250mm

I II IIII
I 1
I I I
I I

IIII II \I,
FIGURE 2 Screen data and slot arrangements for test 1

A summary of results from 35pairsof test runs conducted Table 1

in June 1982 and August 1983 is shown in Table 1. Open area comparison: June 1982 to August 1983

Open Area 7% 12%

Throughput increased by 44,5 %, while the molasses pro-
duced by the larger open area screen was a better quality Water flow rate (1/h) 160 160
Massecuite purity 51,3 51,2
than that produced bythe conventional screen (Brix is higher Sugarpurity 81,2 81,4
and purity is lower). Massecuite temperature ('C) 60-62 60-62
Molasses temperature ('C) 4Q-42 4Q-42
Molasses brix 81,90 84,25
A summary of routine product analysis of the test cen- Purity rise across centrifugal 1,25 0,61
trifugals indicated a similar trend, higher brix (1.12) and Molasses purity 34,3 33,6
Throughput (t/h) 1,348 1,948
lower molasses purity (0.60) in favour of the larger open Water % massecuite 11,87 8,21
area screen.

Test 2
12% open area screen (0.06 mm) versus 10% open area screen (0.06 mm)

Screen pattern (1 cm 2) Screen pattern (1 cm 2 )

Open Area 12% Open Area 10%

Number of slots: 77 Number of slots: 57
Length of slots: 2,60 mm Length of slots: 2,80 mm
Width of slots: 0,06 mm Width of slots: 0,06 mm
Distance between slots: Distance between slots:
Horizontal: OAO mm Horizontal: 0,50 mm
Longitudinal: 0,30 mm Longitudinal: 0,50 mm
Thickness of screen: 0,225 mm Thickness of screen: 0,250mm

48
Proceedings of The South African Sugar Technologists' Association - June 1986

IIIIIIIj III I I II I I I I I
I
I
\ I·

I I
IIIIIII III I j I I I II
1 I I

FIGURE 3 Screen data and slot arrangements for test 2

It was noted that the slots on the 10% screen extended not be found, except that the slots are 0.2 mm longer than
into the unperforated band at the lower portion of the screen the 12 % open area (1.1 mm longer than the 7 % open area
as shown in Figure 4 thus weakening the area of screen, screen).
which is held by the clamp ring. Under a microscope the 12 % open area screen showed a
more regular slot opening. Both screens had very few cracks
in the chrome layer and were very well polished.
A summary ofdaily routine analysis also showed a higher
brix (2.20) and lower molasses purity (0.60) for the 10%
open area screen.
The throughput was marginally lower than the 12 % open
area screen. However, at the present high cost of screens,
the advantage ofa thicker screen with potentially longer life,
together with apparent superior molasses drainage qualities,
justifies further investigations.
FIGURE 4 Lower Section of Screen
Test 3
14% open area screen (0,06 mm) versus 7 % open area screen
The first 10% open area screen was installed in -January (0.06 mm)
1985, at which time a preliminary test was conducted. A The 14% open area screen was installed at Umzimkulu
summary of results of 32 pairs of test runs conducted be- in January 1983. A summary of trials conducted over this
tween 10/9/85 and 19/9/85 is shown in Table 2. period initially produced the results shown in Table 3.

Table 3
Table 2 Open area comparison: January 1983

Open area comparison: 10/9/85 to 19/9/85 Open Area 7% 14%

Open Area 10% 12% Molasses brix 81,45 81,75

Molasses purity 37,18 37,32
Water flow rate (lib) 160 160 Throughput (tIh) 1,332 1,556
Massecuite purity 51,6 51,6
Sugar purity 81,8 81,8
Massecuite temperature ('q 60 60
Molasses temperature ("C) 46 46 However the latter portion of these same trials showed a
Molasses brix 87,23 85,25 completely different picture. Results show a marked de-
Purity rise across centrifugals 1.21 1.70
Molasses purity 35,2 35,5 crease in throughput, this was evident in Tests 6, 7 and 8.
Throughput (tIh) 1,933 2,109 (See Figure 5).
Water % massecuite 8,28 7,59
The 14% screen blinds up after approximately 4 weeks
continuous use. The other test screen did not blind up.
Whenever there was a slack C Massecuite feed, the 14 %
The molasses produced by the 10% screen was of a better screen tended to blind up after only 6 to 8 hours of con-
quality, with a higher brix and a lower purity. tinuous use.
An interesting feature of this 10% open area screen is that Earlier tests with a 7 % open area screen from the same
a reason for the improved molasses brix (drainage) could manufacturer showed a similar problem.
49
Proceedings of The South African Sugar Technologists' Association - June 1986

Table,4 12% open area screens (0,06 mm) have been used on the
Open area comparison after blinding of 14% open area screen. "B" Centrifugals at Umzimkulu since the 1983/84 season.
A summary of results of 16 pairs of test runs conducted
Open Area 7% 14% between 4/7/85 and 11/7/85 is shown in Table 6.
Molasses brix 83,68 82,15
Molasses purity 34,86 35,07
Throughput (t/h) 1,815 1,444 Table 6
Open area comparison between 0.09mm and 0.06mm screen

It appears, that blinding of the screen occurs where the Open Area 10% 12%
screen touches the backing screen, being worse at the lower
part of the screen. This could have been caused by the con- Slow Width 0,09 0,06
Water flow rate (l/h) 260 260
stituents of the molasses and could differ from mill to mill. Massecuite purity 67,1 67,2
The results shown in Table 5 were recorded afterthe screen Sugar purity 89,8 89,6
Massecuite temperature ('C) 40 40
wascleanedwith causticsoda. To exclude the machineeffect Molasses temperature ('C) 43 43
screens were exchanged. Molasses brix 77,23 77,14
Molasses purity 42,2 41,5
Table 5 Throughput (t/h) 3,464 3,964
Open area comparison after cleaning 14 % open area screen Water% massecuite 7,50 6,56

Open Area 7% 14%

Molasses brix 80,62 79,80 These results show a 12,61 % increase in throughput and
Molasses purity 40,92 40,95 a lower molasses purity for the 12% open area screen. The
Throughput (t/h) 1,440 1,800 brix figures remained constant between the two machines,
despite a lower water % massecuite for the 12% open area
Results show again a higher throughput of 25 % for the screen.
14% open area screen, after the screen was cleaned with A summaryof routine analysisof the test centrifugals over
caustic soda, this was evident in Tests 9, 10 and 11. (See this period showed a similar trend, of a slightly lower brix
Figure 5). (0.11) and lower purity molasses (0,5 il) for the 12% open
area (0.06 mm).
Molasses Brix
______ 14% Inspection of the test screens after ten weeks continuous
_ _ 7% USe, showed a slight wear on the 12% open area 0.06 mm
screen and no wear on the conventional 10% open area
85 0.09 mm screen.

Discussion
80 Because ofthe thinner screenit is essential that large open
area screens are fitted very carefully, by selected well trained
fitters, who have a working knowledge of the machine and
are awareofthe importanceof careful fitting of thesescreens.
Molasses Purity' Backing screens and distribution cones should be in-
spectedeach time a screenis changed. The standard 4 mesh
backing screen should be interwoven and have a flat top, to
givethe thinner open area screen better support. Large open
40 area screens are easilydamagedby foreign objects. To over-
comethis, a rigidhousekeeping programme must be adhered
to.
35 Because centrifugal screens tend to blind, especially when
curing viscous massecuites, a stringent washing program,
preferably with high pressurecleaning equipment, shouldbe
carried out at least twice a shift.

Conclusion
These tests have verified that increased open area on cen-
trifugal screens will increase machinecapacity. As the water
% massecuite decreases with increased open area the mo-
lasses quality improves in each case.
However, in the case ofthe 14 % openareascreen, a marked
decrease in throughput was noticed after a period of time.
The cause of this was severe blinding of the screen. The
reason for this is not known but deserves further investi-
gation.
2 3 4 5 6 7 8 9 10 11
Trials with large open area screens (0.06 mm) on B curing
Test Nos also showed an increased throughput and lower molasses
FIGURE 5 Results after the screen was cleaned with caustic soda purity. Trialswith the 10%,0.06 mm large open area screen,
50
Proceedings of The South African Sugar Technologists' Association - June 1986

Test 4
12 %open area screen (0,06 mm) versus 10% open area screen (0,09 mm)

Screen pattern (1 cm 2) Screen pattern (1 cm 2 )

Open Area 12 % Open Area 10%

Number of slots: 77 Number of slots: 68
Length of slots: 2,60 mm Length of slots: 1,70 mm
Width of slots: 0,06 mm Width of slots: 0,09 mm
Distance between slots: Distance between slots:
Horizontal: 0,40 mm Horizontal: 0,53 mm
Longitudinal: 0,30 mm Longitudinal: 0,80 mm
Thickness of screen: 0,225 mm Thickness of screen: 0,250mm

IIIII IIIII f

I I
I

IIIIII1 I II
I I

FIGURE 6 Screen data and slot arrangements for test 4

showed a superiormolasses drainage (higher brix), for which Acknowledgement

no explanation can be givenexcept for the longer slots. Tests
withthis thicker large open area screen will continue. The authors thank the management of C'G. Smith Sugar
Limitedfor permission to publishthis paper. A special word
No modification of the 4 mesh backing screen was found of thanks to the process team at Umzimkulu Mill for the
necessary. With the same screens ~sed on Band C centrifu- contribution which made this paper possible.
gals, inventory costs are lower, WIth only 0.06 mm contm-
uous centrifugal screens kept in stock. REFERENCE
The average screen life of-12 % open area screens during I. Prince PA and Montocchio MRG (1977).. Evaluation of screen mod-
the 1985 crushing season was ten weeks on C massecuites ifications on a continuous centrifugal. Proc S Afr Sug Techno/ Ass 51:
and 6 weeks on B massecuites. 101-106.

51