ate this observation, a series of five leaf tests were
Freezing rate appeared to have little effect on
conducted with cloth 2019 at 100 mm of vacuum
the uniformity coefficient (see Fig. 15). However,
and a 5-minute filtration time. The sludge layer
the plot suggests that the uniformity coefficient
was completely frozen and the filter cloth was not
decreases with increasing initial solids content.
washed between cycles. Results of this test indi-
The average uniformity coefficients are 4.0, 3.3,
cate that the filtrate volume dropped from 85 to
and 3.0 respectively for the 23%, 67%, and 12
60 mL after the second cycle but then stabilized at
13% solids sludges, respectively. A statistical anal-
a constant 60 mL for the remaining three cycles
ysis of this data indicates that these differences are
(see Fig. 13). Concomitantly, the sludge layer
significant at the 95% level. This means that a more
weight gradually decreased during the first four
uniform grain size can be expected from freeze
cycles and then increased slightly during the fifth
thaw conditioned sludges containing a greater ini-
cycle. These results suggest that the filter cloth
tial solids content. The reason for this phenome-
non is not clear but perhaps there is less opportu-
had reached an operational steady state after the
nity for particle movement away from the freezing
first four cycles so the wash cycle could be elimi-
front in sludges with higher initial solids because
nated.
of the close intraparticle proximity. As a result
there would be fewer opportunities to form vari-
Freezing test
The results of the freezing tests show that both
The curing time tests were conducted with
the rate of freezing and the initial solids content
alum sludge containing 23% solids. After freez-
have a significant effect on particle size. As
ing was complete, sludge samples were cured at
shown in Figure 14, the effective grain sizes (D10)
4C for periods of 1 and 24 hours. As shown in
for all three sludges were in the sand size range.
Table 2, curing times of 1 and 24 hours appeared to
However, D10 decreased as the freezing rate in-
have no effect on the grain size or uniformity coef-
creased. Also, the sludges that contained a higher
ficient of freezethaw conditioned alum sludge
initial solids content produced larger grains. The
containing 23% solids. Neither the 67% or the
largest grains were produced by the sludge con-
1213% solids sludges were tested because of the
taining 1213% solids. The grains produced by
apparent lack of any effect on the sludge with the
the sludge containing 23% solids were signifi-
23% solids.
cantly smaller.
Based on the data shown in Figure 14, the
effective grain size (D10) of frozen 23% sludge
DISCUSSION
entered the silt and clay particle size range at a
freezing rate of 6.6 kg/h-m2. For the 67% and the
The results of the filter leaf tests show that vacu-
1213% sludges, the crossover points are 15.5 and
um filtration can remove most of the water from
19.8 kg/h-m2 respectively.
the sludge and, at the same time, produce a uni-
form thin layer for freezing. The optimum opera-
tional conditions for the vacuum filter are a vacu-
Table 2. Data on effect of curing
um level of 100 mm of Hg and a 5.0-minute filtra-
time on effective grain size (D10)
tion time. Generally, vacuum filtration increased
and uniformity coefficient (Cu) of
the total solids content of alum sludge from 2% to
freezethaw conditioned alum sludge
approximately 6%. This represents a 67% reduc-
containing 23% solids.
tion in the volume of sludge to be frozen. After
freezing, thawing, and draining of meltwater, the
Effective
Curing
grain size
Uniformity
remaining granular material contained approxi-
Test
time
(D10)
coefficient
mately 12% solids. Although 12% solids is relative-
number
(hr)
(mm)
(Cu)
ly low, further dewatering will rapidly occur by air
A3
24
0.13
3.9
drying. Our tests show that the sand size particles
B1
0
0.13
3.5
dried to a solids content of 70% or more within a
B2
1
0.13
4.0
few hours.
B3
24
0.14
3.9
Several filter cloths were able to produce a clear
C1
0
0.13
4.0
filtrate even after repeated use. Therefore, selec-
C2
1
0.14
3.2
C3
24
0.12
4.3
tion of the right filter cloth for this application may
D1
0
0.13
3.4
depend more on its durability to withstand several
D2
1
0.11
3.7
freezethaw cycles than its ability to produce a
D3
24
0.13
3.5
13