-------
Figure 7 presents the results of the parallel filtration runs. The first
seven runs represent the data with alum dose of 55 mg/1 in the clarification
system. In comparing runs No. 1 through 3 with runs No. 4 through 7, it is
apparent that the turbidity removals by the filters were essentially the same
at polymer filter aid dosages ranging from 0 to 0.2 mg/1. The headlosses, how-
ever, at filter aid levels of 0.10 to 0.20 mg/1, were definitely higher than
those at 0 to 0.08 mg/1. The effect of polymer filter aid dosages on the head-
loss buildup is further shown by comparing runs No. 8 through No. 11 and run
No. 14 with runs No. 12 and 13. These observations suggest that while polymer
filter aids are desirable in strengthening weak chemical floe thus preventing
premature solids breakthrough, the polymer filter aid doses must be kept as low
as possible to avoid excessive headless buildup.
In evaluating the performance of the filters as shown by the data in
Figure 7, two conclusions could be drawn. First, the turbidity removal per-
formance of the dual-media filter was consistently better than that of mixed-
media filter. The data also indicate patterns of improved turbidity removal
during the early part of the filter run. Secondly, the headless buildup
through the mixed-media filter was definitely and consistently higher than
that observed in the dual-media filter. Moreover, the data show, that with
the exception of run No. 1, the headless buildup through the filter varied
linearly with filtration time which is indicative of an in-depth filtration.
B. With Scheme B Pretreatment
In the comparative evaluation of the filters with in-line coagulation
pretreatment, the alum solution was added to a rapid mixing unit where mixing
for approximately 1.5 minutes was provided. The alum-coagulated secondary
effluent was then pumped directly to both multi-media pressure filters at a
flow rate to maintain a filtration rate of 3.4 1/sec/m2 (5 gpm/ft2). An
anionic polymer (Calgon WT-2700 or WT-3000) was injected directly into the in-
fluent line of each filter. In evaluating the filters, the turbidity removal
in the course of the filter run was used as the primary parameter for compar-
ing the filter performance.
The comparative performance of the filters with an in-line coagulation
pretreatment using various levels of alum and polymer, .is presented in Figure
8. .It is apparent from the data that for the secondary effluent being
treated, high alum dose definitely caused poor filter effluent quality. The
response of the filters with an increase or a decrease in alum dose is clearly
shown in the figure.. The data show that at zero and low alum doses the tur-
bidity removal performance of the dual-media filter was slightly better than
that of the mixed-media filter. Moreover, in the absence of alum, turbidity
removal with or without polymer appears to be the same. The results also
show that the performance of the filters with direct chemical injection of an
anionic polymer WT-2700 was about the same as that obtained with an anionic
polymer WT-3000.
In the course of the evaluation, the filters were also operated using a^
nitrified secondary effluent feedwater (Scheme C) in an effort to determine
what effect, if any, different type of feedwater would have on the perform-
ance of the filters. Figure 9 presents the performance of* the filters with
23 ,
-------
ID
l-
Q
m
en
ID
h-
CO
CO
3
Q
<
UJ
U-
Q
CD
tr
ID
CO
CO
3
Q
<
UJ
X
0
15
10
0
10
OPERATING CONDITIONS:
FEED SOURCE-*NON-NITRIFIED SECONDARY EFFLUENT
FILTRATION RATE-*-5gpm/ft.*
TYPE OF POLYMER-* ANIONIC WT-3000
T= MRS. RUN TO 20 PSI HEADLOSS
(a) ALUM FOR COAGULATION O--O FILTER INFLUENT
(b) POLYMER COAGULANT AID && MIXED-MEDIA FILTER
(C)POLYMER FILTER AID DD DUAL-MEDIA FILTER
T
p-o-o-oN
T
T
T T
UNIT CONVERSIONS:
gpm/ftz x 0.68 = l/sec/m2
psi x 0.0703= kg/cm*
RUN NO. I
(a) 55 mg/l ALUM
(b)0.20 mg/l POLYMER
(c)0.2 mg/l POLYMER
T=8'/2hrs.for DM
hrs.for MM
o-o-
-o-'
.o o
o ฐo.
-o
a
JL
J_
0
468
FILTER RUN, hours
RUN NO. 2
(0)55 mg/l ALUM
(b)0.20 mg/l POLYMER
(c)0.20 mg/l POLYMER
T=23'/2hrs.for DM
= 7'/2hrs. for MM
t-tti
10 "22
24
Figure 7. Performance of filters with chemical coagulation-
sedimentation pretreatment.
24
-------
4
V 3
ซ
m
cc
ฐ- 15
10
5
0
cn
en
3
Q
<
UJ
t 3
Q 2
CQ
a: i
ฐ- 15
*ป
CO
ง 10
o
2 5
0-0-0,
'00
X>-0O O
RUN NO. 3
(a) 55 mg/l
(b) 0.20 mg/l
(c) 0.10 mg/l
T = 23'/2 hrs. for DM
= 7 hrs. for MM
RUN NO. 4
(a) 55 mg/l
(b) 0.20 mg/l
(c) O mg/l
T = 23'/2 hrs. for DM
, ?23'/2hrs.for MM
4 6 8
FILTER RUN, hours
10
24
Figure 7. Continued.
25
-------
li-
fe
CD
0r
u>
a.
CO
CO
3
o
UJ
X
U.
H
g
CO
DC
en
a.
CO
3
O
UJ
X
3
2
I
0
15
10
5
0
4
3
>
4
I
0
15
10
5
0
-tt-
o
RUN NO. 5
(a) 55 mg/l
(b) 0.20 mg/l
(c) 0 mg/l
J=24hrs.for DM
and MM
\
00 O
ฃ~3&53S3ZE===8
am-
-D
RUN NO. 6
(a) 55 mg/l
(b) 0.20 mg/l
(c) 0.08 mg/l
= 24 hrs. for DM
and MM
I
468
FILTER RUN, hours
^H^
/v
10 22
24
Figure 7. Continued.
26
-------
u_
5
CD
or
ID
CO
ง
Q
<
LU
X
LL.
>
O
CO
CC
ID
w
Q.
(O
CO
3
O
<
LU
4
3
2
I
0
15
10
5
0
3
2
I
0
- o'
4.4
4.1 4.2
'""oo-
RUN NO, 7
(a) 55 mg/l
(b)0.20 mg/l
,(c) 0.08 mg/l
23^4hrs. for DM
and MM
O n
RUN NO. 8
(a) 110 mg/l
(b) 0.20 mg/l
(c) O.08 mg/l
T=24hrs.for DM
= 22hrs.for MM
I
4 68
FILTER RUN, hours
10 '22
24
Figure 7. Continued.
27
-------
u_
5
CD
cr
ID
co
CL
ป
1
Q
<
UJ
ID
li.
5
m
cc
u>
Q.
cn
co
3
o
<
UJ
4
3h
2
I
0
15
10
5
0
3
2
I
C
IS
10
5
0
o-o I
\
\
b-oo-o
RUN NO. 9
(a) 110 mg/l
(b) 0.20 mg/l
(c) 0.08 mg/l
T-*- DM and MM
SHUTDOWN AT 71/2 hrs. DUE
TO INFLOW INTERRUPTION
RUN NO. 10
(a) 110 mg/l
(b) 0.20 mg/l
(c) 0.08 mg/l
T= 12 hrs. for DM
= 9 hrs. for MM
_L
468
FILTER RUN, hours
10 22
24
Figure 7. Continued.
28
-------
ID
t
H,
m
oe
h-
8.
ง
o
UJ
X
ID
1-
L_
H
0
fฅ\
CD
cr
ID
H-
Q.
co
CO
o
0
tu
x_
t
3
2
1
0
15
10
5
o
V
3
2
0
15
10
5
1 | | 1
- ' .
-
*^ ^** ^ป S
" ^"Q"" ^Q IHH ~f\^
^F" ^P rf3 -Q p^ A-4 n rp
-
M
o
[JL_
/y RUN NO. II
s'^ ^^"^ (a) HO mg/l ~
./' r^^ (b) Q2ฐ m9/l
J-* ^^ (c) 0.08 mg/l
/^r^^ T = 24
hrs. for DM
-^y^^-^^ =5'/2 hrs. for MM
tr^"^
^ฐ>.
/
1
3,1 [
~ ~s'^Q X> ฐ~-~O p-O o
\j
1
1
RUN NO. 12
(a) 110
mg/l
(b) 0.20 mg/l
(c) 0 mg/l ~
_f^_&= 23*/2 hrs. for DM
/V" ฃj i"f
cr i
i
O2 4 6 8
10 U22 24
FILTER RUN, hours
Figure 7. Continued.
29
-------
u.
Q
CD
CC
1
CO
CO
O
<
Ul
CD
S
to
Ou
*ป
CO
CO
3
O
<
UJ
3 -
2
I
15
10
5
0
I
0
15
10
5
0
O O CL
RUN NO. 13
(a) 110 mg/l
(b)0.20mg/l
(c) 0 mg/l
T=20hrs.forDM
and MM
RUN NO. 14
(a) 110 mg/l
(b) 0.20 mg/l
(c) 0.08 mg/l
T=l3'/2 hrs.forDM
ll'/2hrs.forMM
468
FILTER RUN, hours
10 22
24
Figure 7. Continued.
30
-------
13
H
U.
ป
Q
CO
a:
10
8
6
4
2
0
10
8
6
4
2
0
15
10
5
OPERATING CONDITIONS:
FEED SOURCENON-NITRIFIED SECONDARY EFF
FILTRATION RATE5 gpm/ft.*
POLYMER DOSE*0.23 ma/I WT-3000 (ANIONIC)
0--0 FILTER INFLUENT
&& MIXED-MEDIA FILTER.EFFLUENT
D D DUAL-MEDIA FILTER EFFLUENT
I
59 mg/l ALUM
I i r r
RUN NO. I
94 mg/l
-O
O ฃ&
UNIT CONVERSIONS:
gpm/ft2 x 0.68= I /see/m2
RUN NO. 2
65 mg/l
IM
+ซi 21 mg/l >
r65rr
:/ALU
ALUM
0 mg/l
o-
RUN NO. 3
ALUM: a) 166 mg/l
b) 18.7 mg/l
_L
1 L
46 8 22
FILTER RUN, hours
24
Figure 8. Effect of alum dose on turbidity removal with
non-nitrified secondary effluent feed.
31
-------
TURBIDITY, FTU
4
3
2
1
0
4
3
2
1
0
10
8
6
4
2(
0
C
OPERATING CONDITIONS:
FEED SOURCE-^NON-NITRIFIED SECONDARY EFF
FILTRATION RATE-ป-5gpm/ft.z
POLYMER DOSE-*- 0.23 mg/l WT.-3000 (ANIONIC)
0--0 FILTER INFLUENT
& & MIXED-MEDIA FILTER EFFLUENT
D D DUAL- MEDIA FILTER EFFLUENT
RUN NO. 4
NO ALUM OR
POLYMER
^ o ฐ''ฐ^'X
o o- ^ 0
ฃ==ฃ r&&-T-fr=&3&=&=&
p 1 y | ui TIY Y 4
RUN NO. 5
o^ ^^ฐ~~ฐ-ON
V -^^-.o^0 X
"S? 0
Do tt S SS- D- i>S li tf r-Q I
RUN NO. 6
^ |=: o mn/l Al 1 IM - . .to*
^o
o- -o^ ^QJO o- -o- -o"
1
) 2 4 6 8 22 24
-
-
-
FILTER RUN, hours
Figure 8. Continued.
32
-------
z>
h-
u_
>
Q
CD
CC
Z>
I-
4
3
2
1
0
3
2
1
0
3
2
1
0
3
2
1
OPERATING CONDITIONS:
FEED SOURCE -"NON-NITRIFIED SECONDARY EFFLUENT
FILTRATION RATE 5gpm/ft.2
TYPE OF POLYMER-* ANIONIC, WT.-2700
0 --- 0 FILTER INFLUENT
ฃป - & MIXED-MEDIA FILTER EFFLUENT
D D DUAL MEDIA FILTER EFFLUENT
ALUM = 6.1 mg/l
> POLYMER = 0.36 mg/l
^O--O -- O O--O-'
- -O - O- -6 - -6
RUN NO, 7
V A A ฃj A A A AA & A-ฃrf
+
*
ALUM = 6.1 mg/l RUN NO. 8
POLYMER = 0.22 mg/l
^o - -o - o-o-o- -o QN
ALUM = 18.4 mg/l
POLYMER = 0.25 mg/l
O- O - ^- -o- -O-O - O -O-O- -O--
^-Hitf
RUN NO. 9
-O
ALUM = 18.4 mg/l
POLYMER =0.34 mg/l
_ RUN NO. 10
/ฐ-ฐ-O--O-.o-o
0
468
FILTER RUN, hours
10 22
24
Figure 8. Continued.
33
-------
I-
u.
o
00
o:
4
3
2
I
0
3
2
I
0
3
2
0
OPERATING CONDITIONS:
FEED SOURCE-ป-NON-NITRIFIED SECONDARY EFFLUENT
FILTRATION RATE-*5 gpm/ft.2
TYPE OF POLYMER-*- ANIONIC,WT.-3000
O O FILTER INFLUENT
A A MIXED-MEDIA FILTER EFFLUENT
D '-D DUAL-MEDIA FILTER EFFLUENT
ALUM= Omg/l
POLYMER = 0.2 mg/l
RUN NO. II
ALUM = 9.5 mg/l
POLYMER=0.26 mg/l
ซ~-0--0-o
RUN NO. 12
ALUM = 18.4 mg/l
POLYMER = 0.23mg/l^Q Q_
RUN NO. 13
-OO-
O-O.
ALUM = 24.5 mg/l
POLYMER =0.12 mg/l
X>--O--O--O-O
P O OO
468
FILTER RUN, hours
10 22
Figure 8. Continued.
34
-------
Q
OQ
CC
8
6
4
2
0
8
6
4
2
0
8
6
4
2
0
15
10
5
OPERATING CONDITIONS--
FEED SOURCE-* NITRIFIED SECONDARY EFFLUENT
FILTRATION RATE5gpm/ft.*
POLYMER DOSE* 0.23 mg/l WT-3000 (ANIONIC)
O- --O FILTER INFLUENT
ฃs & MIXED-MEDIA FILTER EFFLUENT
D D DUAL MEDIA FILTER EFFLUENT
RUN NO. I _
6.1 mg/l ALUM-
J0--
I -.1-1p_r t-l-J-t-H-i I I
UNIT CONVERSIONS: RUN NO. 2
gpm/ft2 x 0.68= l/sec/m*
O-O-"ฐ~-O O -O--O--O--O
21 mg/l ALUM
45 mg/l ALUM -*k 21 mg/l
58 mg/l ALUM-4*- 6.1 mg/l
^ I
RUN NO. 4
4 68 10
FILTER RUN, hours
12
Figure 9. Effect of alum dose on turbidity removal
with nitrified secondary effluent feed.
'35
-------
nitrified secondary effluent feedwater. The results shown in the figure dem-
onstrate similar trend as in Figure 8, that is, high effluent turbidity at
high alum dose.
Based on the above observations, which show that the overall performance
of the dual-media filter was.equal to or better than that of the mixed-media
filter, all subsequent filter evaluations were confined to the use of the
dual-media configuration.
Test Series II - Dual Media Filter Performance
A. With Scheme A Pretreatment
As discussed in the Test Series I-A, when the dual-media filter was
evaluated in parallel with the mixed-media filter, the chemical clarification
system was operated at a flow rate of 3.47 I/sec (55 gpm).' After the com-
pletion of the comparative filter evaluation, the flow through the clarifica-
tion system was reduced to 2.42 I/sec (40 gpm) thus providing mean hydraulic
residence times of 4.1, 62.5 and 126 minutes in the rapid mixing, flocculation
and sedimentation tanks, respectively. Alum was continuously added to the
rapid mixing tank at a dosage ranging from 55 to 225 mg/1. An anionic polymer
(Calgon WT-3000) at an average dosage of 0.20 mg/1 was added as a coagulant
aid to the flocculation tank. The .chemically clarified secondary effluent
was pumped to the dual-media filter at a flow rate of 1.58 I/sec (25 gpm)
which was equivalent to a filtration rate of 3.41 1/sec/m2 (5 gpm/ft2). The
remaining clarified effluent flow was then diverted to waste.
The experimental data presented in this section include all the results
obtained with the clarification system operated at 3.47 I/sec (55 gpm) and
2.52 I/sec (40 gpm). Thus, the dual-media filter performance data presented
in Test Series I-A with alum dosage of 55 and 110 mg/1 in the clarification
system are also included in the summary data in this section.
Table 5 presents a summary of the dual-media filter performance with
chemical coagulation-sedimentation pretreatment. In the operation of the
filters, an anionic polymer (Calgon WT-3000) at a dosage varying from 0 to
0.20 mg/1 was injected as a filter aid to the filter influent line. As. in-
dicated by the data, the turbidity removal at a given alum dose remains
essentially the same with or without polymer filter aid. The headloss data
across the filter, however, is definitely higher with the use of a polymer
filter aid.
In Figure 10 is shown a number of dual-media filter runs with the clari-
fication system operated at an alum dose of 110 mg/1 and polymer at 0.20.
mg/1. The figure shows the fluctuation in the effluent turbidity in the
course of the first 7.5 to 10 hours of filter run. Although in most runs
effluent turbidity of 0.5 FTU or less was attained, there were days when
higher effluent turbidity levels were observed. Moreover, the headloss level
also varied from run to run with an observed range of 0.55 to 1.1 Kg/cm2
(7.8 to 15.6 psi) at the end of the first 7.5 hours of filter run. Figures
11 and 12 show a similar plot to Figure 10 except at higher alum doses in the
36
-------
*
t
z.
UJ
s:
ฃ^
UJ
a:
^_
UJ
Cฃ
D.
0
1
>5^5
4.5 ซ>
ปi r
00 TS (O
or JD o
S- E
CO 3 CU
I Q^
I^H
!
LO
r^E -g?
4,5 I1"..
s- s- CZ-Q
r- cu j; s-
r J
CU T-
jr: u_
4->
_t % *~~>
s- p .t
O JiT * '
M- -3 "
ET JQ
en , s>-
> ^ 3
*^C t~~
+
S- S-
a> cu
E 4-5 T3 r
>,r r---.
i -r- =a: en
ou_ E
D.
4-
(O i S-
o ~^ cu
'E Eฐ >,
CU r
J= ซ O
C_J E D.
0
S- !-
0 4J
4- CO
r
CU 3
cn en
ca 10 E
OO O 3
O O r
Q =t
OO
d M- ฃ-
~-^ O fv
00
4-> CD
00 -r- O
> co
O O O O C3
"vf 1^. O * *
CM CO CM CO CM
O CO CO CO
oj o cj o
o
00 00
o o o O o
OJ OJ CM OJ OJ
o o o o o
LO LO LO LO O
LO LO LO IO r
r
CM
oo
*
OJ
o
CO
OJ
CO
0
O '
0
OJ
o
LO
en
*
r--
LO CT> LO
i i i
CO LO CO
i CO r-
CO f^** LO
CM * r
O OJ LO
cn i oj
co cn co
CO CM LO
OJ OJ CO
0 O O
t LO O
oj OJ oj
CO
o
O 0
0
o o o
OJ OJ OJ
o o o
LO LO LO
LO LO OJ
i OJ
CO CM OJ
co cn o
CO
cn
i
CO
CO
LO
cn
CO
ซvf
CO
CO
co
o
r
if>
CvJ
r~ป
o
CD
CD
CO
CD
LO
OJ
OJ
OJ
!_
r""
&_
O
o
cu
oo
-
OJ
LO E
OJ S-
.ซ S.
10 a)
r
O "r~
"Z. M-
JE
en oo
3 S-
O -r-
-E LO
ri
H"
O
o
E
00 CU
cu
r- CU
CU 4J
OO
4, csl fO
to E
CU "--, O)
4-> p cn
CU E
S- oo fO
o -^. s-
"~ CO
p ^ t~
CU 4->
00 CO
^^ cu
l 4-> S-
fC rO
vf- CU oo
1 ^ p
CO (O OO
s~ cu
3 E 4-J
O O E
i -r-^^ CU
M- 4J 0 S-
(O C3 (O
E S- 0 Q.
dJ ^J CO
OO -I 1 'I
OO "O
E CU
E O 00
o i en o
4-> (OP
fO -C CJ E
p cn* cu
r- 3
<4- O S- 00
i- S- CU CU
S- .E E 3
CO 4J >,.
r r lO
P t~~- C >
Q-
r 00
03 O P 00
VZ. <- O
p- E i
^E ^o o "O
CU E -I- to
-E 3 E CU
* +-t
37
-------
15
to
o.
to 10
CO
3
Q
<
UJ -
X 5
J90
080
l*J -7rt
DC 70
t 60
Q
g50
0'
H .8
u.
*
m
oc
H 4
LU J2
u_ -^
Ul
OPERATING CONDITIONS!
FEED SOURCE> NON-NITRIFIED SEC. EFF
FILTRATION RATE-* 5 gpm/ft.2
POLYMER FILTER AID->0.07-.08 mg/l
WT. 3000
UNIT CONVERSIONS:
gpm/ft2 x 0.68 = I /sec/m2
psi x 0.0703 = kg/cmz
4 68
FILTER RUN, hours
10
Figure 10. Dual-media filter performance with 110mg/l
alum and 0.2 mg/i polymer in the clarification
system.
38
-------
15
10
O
O
j- 90
1 80
I 70
ฃ 60
ง 50
cc
P 40
0
e -8
u_
. .6
CD
.4
LU
i i n I
OPERATING CONDITIONS:
FEED SOURCE - NON-NITRIFIED SEC EFF
FILTRATION RATE-* 5 gpm/ft*
7 WITH .07-08 mg/l
JWT-3000 FILTER AID
- o
WITHOUT FILTER AID
UNIT CONVERSIONS-'
gpm/ft2 x 0.68= l/sec/m2
psi x 0.0703 = kg/cm*
4 6
FILTER RUN, hours
8
10
Figure II. Dual-media filter performance with 155 mg/l
alum and 0.2 mg/l polymer in the clarification
system.
39
-------
(O
Q.
CO
CO
o
<
III
15
10
I
LJ
or
9
CD
cc
m
o:
ID
u:
u_
m
90
80
70
60
50
40
r
r
.8
.6
4
.2
0
OPERATING CONDITIONS:
FEED SOURCE-*NON-NITRIFIED SEC. EFF
FILTRATION RATE-* 5 gpm/ft*
WITH .07-.08 mg/l
WT.-3000 FILTER AID ^ซ
A]WITHOUT FILTER^-^
OJAID ^ /
/ /
s x
- xx >x
^ X
-x
UNIT CONVERSIONS:
gpm/ftax0.68= l/sec/m2
psi x 0.0703 = kg/cm*
468
FILTER RUN, hours
10
Figure 12. Dual-media filter performance with 225
mg/l alum and 0.2 mg/l polymer in the
clarification system
40
-------
clarification system. The data in these figures also include the runs with-
out the use of polymer filter aid. The results presented in Figures 10
through 12 show that effluent turbidities ranging from 0.1 to 0.5 FTU were
obtained in all the three alum doses evaluated. It is evident from the
figures, however, that except at the alum dose of 155 mg/1, the filter efflu-
ent turbidities varied from run to run. The data shown in Figures 11 and 12
indicate that the turbidity removal performance of the dual-media filter was
about the same with or without the use of a polymer filter aid. Moreover, it
is apparent that the headless buildup was markedly higher with the use of a
polymer filter aid.
Figure 13 presents the effect of alum dose in the clarification system
on the performance of dual-media filter. Each headloss data point in this
figure represents the observed headloss at the end of the first 7.5 hours of
each filter run. The effluent turbidity and percent turbidity removal data
are the average of hourly values obtained during the course of the first 7.5
to 7.8 hours of each filter run. Based on the experimental data summarized
in Figure 13 along with those presented in Table 5, the following conclusions
about filter performance with Scheme A pretreatment were drawn:
1. A filter effluent turbidity of 0.2-0.4 FTU was achieved at
an optimum alum dose of 150 mg/1 and an anionic polymer
coagulant aid of 0.20 mg/1 (Calgon WT-3000) in the chemical
clarification system.
2. The turbidity removal- efficiency in the dual-media filter
was essentially the same with or without the use of a poly-
mer filter aid.
3. With the use of a polymer filter aid, the headloss across
the filter was higher at higher alum dose in the clarifica-
tion system. In addition, at a given alum dose, the head-
loss across the filter with the use of 0.06 to 0.08 rng/1
polymer filter aid (Calgon WT-3000) was higher than that
without filter aid.
On the basis of the above findings, all subsequent filtration runs' were
conducted with the clarification system operated at an alum and polymer dos-
ages of 150 mg/1 and 0.20 mg/1, respectively. In addition, no polymer fil-
ter aid was used.
B. With Scheme B Pretreatment
In the evaluation of the dual-media filter with an in-line coagulation
pretreatment, a number of experimental test runs were conducted using various
levels of alum in combination with an anionic polymer (Calgon WT-3000). Fig-
ure 14 presents the results of selected filtration runswith alum levels rang-
ing from 0 to 18.4 mg/1 and with an anionic polymer from 0 to 0.23 mg/1. The
test results shown in Figure 14 along with the summary data presented in
Figure 15 showthatwith in-line coagulation pretreatment, filter effluent tur-
bidity levels obtained were greater than 0.5 FTU for all the glum and polymer
41
-------
15
I0
(O
CO
3
o
2
LU
DC.
h
Q
m
oc.
h-
U.
ซ
CD
OC
Z)
U:
U.
LU
0
90
80
70
60
50
40
t
0
.8
.6
.4
.2
0
OPERATING CONDITIONS:
FEED SOURCE-*-NON-NITRIFIED SEC. EFF
FILTRATION RATE-*-5gpm/ft*
- POLYMER COAGULANT AID0.20mg/l -
WT. 3000 (ANIONIC) *
WITH 0.07-.08 mg/l WT.-3000
FILTER AID
o-^WITHOUT FILTER AID
UNIT CONVERSIONS:
gpm/f t2 x 0.68 = I /sec/m2
psi x 0.0703 = kg/cm*
50 100 150 200
ALUM DOSE, mg/l
250
Figure 13. Effect of alum dose in the clarification system
on the dual-media filter performance.
42
-------
15
CO
UJ
10
J90
ง80
o
S50
cc
^40
u.
ป
CD
(T
u:
u_
uj
OPERATING CONDITIONS:
FEED SOURCE -NON-NITRIFIED SEC. EFF
FILTRATION RATE-5 gpm/ft.2
SYMBOL mo/1 ALUM mq/i POLYMER
(WT. 3000, ANIONIC)
UNIT CONVERSIONS:
gpm/ft* x 0.68= I /sec/m*
psi x 0.0703= kg/cm*
468
FILTER RUN, hours
10
Figure 14. Dual-media filter performance with in-line
coagulation pretreatment.
43
-------
15
t 10
(O
3
O
LU
X
OPERATING CONDITIONS^
FEED SOURCENON-NITRIFIED SEC. EFF
FILTRATION RATE5 gpm/ft*
SYMBOL mq/l POLYMER (WT 3000
A o ANIONIC)
O 0.08
x 0.23
/oo
^ ' '
* 9%
J 80i- o 'fiL ~^ฐ"-^-^
o fit w ฐO ^-
ง 7Q
Q; <
>- 6Q
f-
Q 5O
CD
ง 40
H .
ID
H
1 1
LL
*ป
CD
* L
0 h
h-
LU ;
LL '
lฑl t
o
\_ - OO *-^
i ^^ >^^
b V
'
^
UNIT CONVERSIONS:
gpm/ft* x 0.68= 1 /sec/m*
psi x 0.0703 = kg/cmz
1 !
1 '
- 0 ^^^
5r -
-------
doses evaluated. The average filter effluent turbidity in the first 7.5 to
7.8 hours of filter run varied from 0.6 to 1;2 FTU. In the absence of alum,
however, filter effluent turbidities ranging from 0.2 to 0.8 FTU were ob-
served. Moreover, in the absence of alum, turbidity removal with or without
polymer injection was about the same. These results confirmed previous
Pomona filtration data which showed that based on filter effluent quality and
fflter run length, the filter performance with plain filtration (no chemical)
was equal to or better than that with the addition of alum and polymer.
Test Series III - Special Short-Term Filtration Runs with Scheme B
Pretreatment.
In the course of the dual-media filter evaluation with an in-line coagu-
lation pretreatment, two sets of special filtration runs were conducted. In
the first set of runs, the dual-media filter was operated with the use of a
cationic polymer (Calgon Cat-Floe T) alone for pretreatment. Figure 16 pre-
sents the results of selected filtration runs with the cationic polymer dos-
age as high as 2.4 mg/1. A summary of the effects of cationic polymer dose
on the dual-media filter performance is presented in Figure 17. Each head-
loss data in this figure represents the observed headloss at the end of the
first 7.5 hours of each filter run. The effluent turbidity .and percent tur-
bidity removal data represent the average of hourly observations obtained
during the first 7.7 to 8 hours of filter run. As indicated by the data
in Figure 17 the apparent optimum polymer dosage ranged from 0.05 to 0.5
mg/1. At this range of polymer dose, the average filter effluent turbidity
in the first 7.8 to 8 hours of run varied from 0.8 to 0.9 FTU. For the dos-
age range evaluated, the headloss at,the end of 7.5 hours of run was low and
ranged from 0.09 to 0.23 Kg/cm2 (1.3 to 3.3 psi).
The second set of the special filtration runs entailed the use of high
alum dose (40.5 to 182 mg/1) in the in-line coagulation system in combination
with a non-ionic polymer (American Cyanamid Magnifloc 985 N) filter aid at a
dosage of 0.7 to 2.1 mg/1. The results of the filtration runs are presented
in Figure 18. As shown in the figure, for the alum dose of 160 to 182 mg/1
in combination with 1.2 to 2.1 mg/1 non-jonic polymer (Test No. 1 to 4),
filter effluent turbidity levels of 0.3 to 0.4 FTU were attained. At this
high chemical dose, however, the length of the filter run to a terminal
headloss of 1.4 Kg/cm2 was very short and ranged only from 20 to 75 minutes.
Moreover, it required a period of about 15 to 20 minutes (so-called "ripen-
ing period") from the start of the filtration run to. reach the stable filter
effluent turbidity level of 0.3 to 0.4 FTU. Although this ripening period
is short, it constitutes a significant portion of the total filter run.
Thus, the test results show that although low filter effluent turbidity
could be achieved at very high alum and polymer doses in the in-line coagula-
tion pretreatment, the resulting run length was too short to be economically
feasible. The last three runs (Test Nos. 5 to 7) in Figure 18 show the tur-
bidity removal data at alum dose of 40.5 to 111 mg/1 and non-ionic polymer
dose of 0.7 to 1.3 mg/1. As indicated in the figure, the filter effluent
turbidity was high throughout the filter run.
45
-------
OPERATING CONDITIONS^
FEED SOURCE-*NON-NITRIFIED SEC. EFF
FILTRATION RATE-*5gpm/ft.ซ
CAT-FLOG T. mq/l
0.08
o o 0.25
ซ 5 - nn 0.61
ฐ- * * 1.66
2.40
en
CO
3
o
<
LU
X
4
3
2
I
0
ฃ75
I
ฃ50
o
m
0
H 3
u:
Lu
LU
0
UNIT CONVERSIONS:
gpm/ft* x 0.68= I /sec/m2
psi x 0.0703 = kg/cma
46
FILTER RUN, hours
8
10
Figure 16. Dual-media filter performance with direct
cationic polymer injection.
46
-------
5
<0
CO ,
ง3
Q 2
tu
I
0
80
ฃ60
ง40
o:
o
I-
20
t 3
ป
CD ซ
o: 2
ID
I-
u_
LU
0
.01
11
OPERATING CONDITIONS^
FEED SOURCE-^NON-NITRIFIED SEC. EFF
FILTRATION RATE-* 5 gpm/ft.*
4-4++
i i i mm
UNIT CONVERSIONS:
gpm/ft* x 0.68= I /sec/m2
psi x 0.0703 = kg/cm*
i 11 MiniH-++
+-H-+T
2 34 6 0.1 2346 1.0 2346
mg/l CAT-FLOG T
10
Figure 17. Effect of cationic polymer dose on the dual-
media filter performance.
47
-------
h-
Q
CD
cr
ZD
h-
J-
UJ
UJ
(T
UJ
b
LI-
OPERATING CONDITIONS:
FEED SOURCE NON-NITRIFIED SECONDARY EFFLUENT
FILTRATION RATE5 gpm/ft*
TYPE OF POLYMER-*-MAGNIFLOC 985N (NON-IONIC)
TERMINAL HEADLOSS-* 20 psi
TEST NO.
AVG. INF. TURB.5.8 FTU
ALUM 182 mg/l
POLYMER-"-1.7 mg/l
FILTER RUN20-30 min.
UNIT CONVERSIONS:
gpm/ft* x0.68= I /sec/mz
psi x 0.0703 = kg /cm *
TEST NO. 2
AVG. INF TURB.--3.4 FTU
ALUM-^174 mg/l
POLYMER2.1 mg/l
FILTER RUN-ซ-60-77min.
NO. OF RUNS4
TEST NO. 3
AVG. INF TURB.-^4.6 FTU
ALUM-*-163 mg/l
POLYMER2 mg/l
FILTER RUN22-50 min.
NO. OF RUNS 6
TEST NO. 4
AVG. INF TURB.3.5 FTU
ALUM-*-160 mg/l
POLYMER-1.2 mg/l
FILTER RUN 45-57min.
NO. OF RUNS7
20
40 60 80
FILTER RUN, minutes
100
120
Figure 18. Effect of high alum and polymer doses on the dual-
media filter performance.
-------
10
OPERATING CONDITIONS:
FEED SOURCE -NON-NITRIFIED SEC. EFR
FILTRATION RATE5 gpm/ft*
TYPE OF POLYMER --MAGNIFLOC 985N (NON-IONIC)
T
T
-o-o-o
TEST NO. 5
AVG. INF TURB.~3.5FTU
ALUM III mg/l
POLYMER-H.2 mg/l -
FILTER RUN-*2.5 hrs-
3.3 hrs. _
HEADLOSS20 psi
TEST NO. 6
AVG. INF TURBH.7 FTU
ALUM-* 40.5 mg/l
POLYMER-^0.7 mg/l -
HEADLOSS-H3.3 psi
after 6 hrs.
TEST NO. 7
AVG. INF TURB.-2.I FTU
ALUM86 mg/l
POLYMER -H.3 mg/l
HEADLOSS-HQ.I psi
after 5 hrs.
I
I
2 3' 4
FILTER RUN, hours
Figure 18. Continued.
49
-------
PHASE III TEST RESULTS
The data presented in this section of the report include the results of
a long-term filtration run encompassing a total period of about eight months.
During this period, the dual-media pressure filters were operated continuous-
ly in parallel at an identical filtration rate of 3.41 1/sec/m2 (5 gpm/ft2).
One filter was operated with a chemical coagulation-sedimentation pretreat-
ment (Scheme A) using about 150 mg/1 alum and 0.20 mg/1 anionic polymer
(Calgon WT-3000) in the chemical clarification system. The other filter was
operated with an in-line coagulation pretreatment (Scheme B or C) in which
approximately 5 mg/1 alum was added to the rapid mixing unit and 0.05 to 0.08
mg/1 anionic polymer (Calgon WT-3000) injected into the filter influent line
as a filter aid.
Filter Effluent Quality
The average water"quality parameters for each treatment unit of Scheme A
are presented in Table 6. The test results in Table 6 show that the sus-
pended solids level in the clarified effluent was higher than those in the
secondary effluent. This was due to poor solids-liquid separation in the
sedimentation tank resulting in chemical floe being carried-over with the
clarified effluent. Nevertheless, as indicated by the data, solids removal
by the filter was excellent. The average filter effluent suspended solids
and turbidity were 1.3 mg/1 and 0.7 FTU, respectively. This corresponds to
an average removal efficiency of 90.6 percent for suspended solids and 83.3
percent for turbidity. Moreover, in the course of chemical .clarification
and subsequent filtration, total phosphate was reduced about 89 percent, re-
sulting in a filter effluent with total phosphate concentration of 0.9 mg/1 P.
Total COD and color were reduced 48 percent and 38 percent, respectively.
The total dissolved solids (TDS) was slightly increased in the filter efflu-
ent as a result of the,high alum dosage in the chemical clarification sys-
tem.
Tables 7 and 8 present the summary of the average water quality parame-
ters for Schemes B and C. As shown by the data in Table 7, the filter re-
moved 80 percent of the suspended solids and turbidity, resulting in a fil-
ter effluent with average suspended solids of 2.7 mg/1 and turbidity of 1,2
FTU. The suspended solids removal efficiency in Scheme C was about the same
as that in Scheme B. In addition to turbidity and suspended solids removal,
color, total COD and total phosphate were also slightly reduced by the filter
in both Schemes B and C.
In Figure 19 is presented a plot of the filter effluent turbidity and
percent turbidity removal as a function of experimental run for Scheme A.
The turbidity data presented in this figure are based on 113 days run. The
figure shows that the filter effluent turbidity levels remained stable through-
out the pilot plant study. Analysis of the filter effluent turbidity data
show that the median and mean values of the filter effluent turbidity were
0.6 to 0.7 FTU, respectively.
50
-------
h-
z:
LU
s:
'"-C
LU
o:
i ,
LU
D.
=c
111
1-M
LU
T"
O
00
:n
1
1 1
3:
LU
O
-
ce:
^ฃ
^gr
^gr
ZD
00
iJ3
LU
^__$
CO
CD
ซ
1
CO C
> O
O !-
E +J
CU CO
f*V^ 3
p"~ r
r- f_
1 1 1 1 ^
UJ
"^ JJ
+J
1 1
o [j;
0) Itj
oo ^
>,
4-) -|-
r- {/)
i S-
cO ,
CO E "
0 O
i ซ CU
E 0)
CD CO
-ฃ= CO E
C_D CO 3
O i
o =a:
UD CO
'
o co
cn oo
to
CO CO 1--.
ID i CD
*^d"
cn cn CM
iO CO "xh
r
CM
co r~> co to
r-1* co o "^t*
CM r
r-. CM r--ป oo
r-~ co o o
n n T T
'Z. 'Z. 2: 'ZL
^^
CO
E
ซ
to
o -a
o !- ro
r~* 1
ซ> O Lu
aj oo
s-
3 -C >,
4-> C1J 4->
CO "O -i
s- c -o
a> a) T~
Q. CL J2
E to , S-
3T (1) 3 3
a. i oo i
cn
r-- cb
co ซ*
LO
cn cn
i g
to to
o r-.
co co
o to
i CO
V n
2: -z.
r
^^
CO
n
Q
CD
t >
S- i
0 CO
O O
0 h-
CD CO
CM O
CD II
^ f
f %
LO CO
O
CO r
=3- II
i "Z.
CO
CC3
CO
cn
to
CO
CM r-. CM
r LO
(
CD o O cn
cn
LO
CO LO
r^ CTl
cn
\
CM CO
CO ^ tD
^j~ p"""
LO t
LO LO LO tO
o cn co co
T n n n
~ZL "Z. 'Z. "Z.
CL.
,
^^.
D-)
r
ซ\ c^ r '
cu -~~,
CO ^ 1'
_E CO
Q. E
CO >>
O < r -1 1
JC ฃ *^. -r-
O- 3 CO E
E E T-
CO E ซ co
4-> 3 OO -^
0 Q i
r eC r =t
CO
E
O
r-
4->
CO
>
S-
O)
CO
0
M-
O
O
II
2= ,
*
CO
< CQ CO
* +
51
-------
1
1
1
ซ=C
UJ
o:
a.
CO
UJ
z.
UJ
3=
O
-c
lt
~~
UJ
0
s
a:
0
u_
o:
UJ
D.
Cฃ
UJ
1
1 4
u_
u_
o
>-
gy^
^ฃ
;gr
5^
*""^
C/7
PS.
1
CO
I
(.
cd
>
ง ฐ
cu
CtL
.j->
s- c:
cu cu
r i
"i M
U tf-
UJ
^> i ^
S f
ซ" ง
"O ซ
E -T.
o {T
o JT
ซs fc
^J
4*3 -f-
f (/)
'id cu
H3 ปi^
o* cu
p-^
s- to
0) S-
ป-> tO
to a.
3
*
S-
ซ tl)
01 E
co >,
fO i
CO O
a- ฐ-
*^^
i en
to E
O
r-
E
OJ E
.A-~ J
(^"
r O
O O
CO CO
10
co r^ CM
r-. CM i
1^1* r**ป Lfi ^^
r^. co oo ซ3
CNJ r
I**"1* LO P"^ป P*"*
kj' ซsj' Q^ ^^
II II _ll_ II
' -
^~
*^^_
CO
E
f\
in
CJ T3
O - ZD
r- 1
0 U_
"
3 -O >,
4-> Ol -t->
tO T3 -i-
i- E -o
Ol Ol -r-
a. Q. .a
E co S-
3T Cl) 3 3
Q- 1 OO t
O O VD
CTl CM CT>
f^s.
co co LO i en
r-- =* r^ o co i
CM CM LO IO
LO I
o en co to
O i CO ซsf =*
CO -=i- =d- (ฃ)
LO i
o r ^- LO r^ป co
LO LO ซd- * "^ "~f
II II II II II II
- . ^_^
D-
r
CO
pz
^~
r*
0 J= E ^ T-
CJ Q_ 3 CO C
E E -
O tO fO E > to
r- 4-> 4-> 3 OO J^
O O O i O r
0 1 1 eC ( =H
'
f ,
IO CM
0 0
O TT
~y*
'S H
CM
to o
LO IT
in
E
r. O
to
ai
10
-Q
O
q-
o
0
^
II
.
to
Ol
f
Q.
E
fO
in
_a
(O
s-
CO
o
E
3
S-
fO
Ol
S-
. 3
to
O)
Q.
E
0)
1
n
CO
0)
- ^i
o a.
o E
o to
CO CO
i
1 01
3: +->
r-
E CO
O O
Co Q_
"tO 0
c_> o
i-
S- 3
O) O
E -E
>} 1
r IQ
0 r
Q-
0 0 01
r- CO
E ~a to
0 01 i-
r- CO Ol
E fO >
=C ca to
* +
-------
1
sz
LU
s:
LU
LU
Di
Cu
O
LU
^>
LU
rc
0
I
i i
3
LU
0
n
o
Lu
fV^
LU
Q-
01
LU
1
1
1 l
Lu
LU
CD
>
Cฃ
*z3^
J-
^-
=>
oo
CO
LU
I
CO
ซcC *
h-
fซ
>
O ^5
E
O)
ซ
.p
i. E
CU (13
(C Q-
Jc
ซ S-
,
to i
0 0
CD 0_
i i
n3 -
O CD
"i E
E
CD E
.E rs
O i
ซC
CO LO
r^. LO oo
^ . ^ n~
00 CO LO i
r-~ CM i co
CM
=d- LO to LO
.
r^ co CM to CM
CM r CO
OO O CO CO CM
co co ro ' co oo
II II II II 11
f
^
en
gz
A
to
CJ> "O
O -r- ID
r- 1
>. O Lu
ฃ ฐฐ
zs -a >,
4-5 QJ -t->
to -a T-
s- E -a
QJ 0) T- i-
Q. Q. J3 O
E to s- ' i .
"T^ CD 3 Z3 O
Q. | 00 10
to
O
CD
i
to
en
r OO
CO r
,,
CM CO i i
LO to O O =*
CM CO *3-
LO p
LO en r~ป
tO 1^ i LO
oo o =*
to i
CM CM CM CO CM
oo co oo co co
11 II II _N_ II
Q-
l
^^
CD
E
1
r en
en n3 E
E r~ en
Q. E "
"in >,
Q O " r -4->
CD ^: E ~-~~ -i-
00. 3 en c .
E E -
rO CO j-~ ** f^
M -]^ ^5 (/^ _^^
O O i Q i
H- h-
eC CQ
-------
JJ~
Iff
c* U^
Cf) ^f
w
55
LU . . O
^COTfr
OC ^ ti
f*} **
2U IS
O S >-
8
o t
y ฐ ง
o o
oฐ
ฐQ
o 0
o ฐ o ฐ
ft
Ql
o
o
Q
O !
Q
ft
O
- ฐ/
ฐ Q
~ 0 ฐ
*o
o
0ฐ
o
o
ft
- o
0
8
0 0
ฐo 0
Q
Q
O
1 O |
5 10 O
5 Is- 10
1
t* -
.ฃ
^^
o
$
^
n ~"
GO
U>
d
x
Q.
I
10 c
ฃ
^^
Q.
O fn ^
00 ฃ <
S l
OQ ^
N- O o
o: co
UJ
a. .c
S & 'i
H fc
^ *-
o a: =
IO Ul -
o_ a>
O ^
? f
s
^
f~.
ro "o
0
E
Q)
CVJ
^%
;ฃ;
T>
la
2 3
1
gi
O " CD
5 *5
, o>
1VAOIN3H %
54
-------
The daily variations in the filter effluent turbidity and turbidity re-
moval are shown in Figure 20, for Scheme B and Figure 21 for Scheme C. The
turbidity data presented in the figures are based on 112 observations for
Scheme B and 38 observations for Scheme C. The median and mean values of the
filter effluent turbidity for Scheme B were 1.1 and 1.2 FTU, respectively.
For Scheme C, the median and mean filter effluent turbidity values were 1.3
and 1.4 FTU, respectively. The results presented show that for the pretreat-
ment Schemes A, B, and C, the filter operations remained quite stable during
the entire period of pilot plant study.
The daily variations in the filter effluent and effluent turbidity
levels for pretreatment Schemes A, B, and C are presented as frequency curves
in Figure 22. The observed turbidity data are fitted approximately by a geo-
metrically normal distribution and thus the line of best fit is plotted as a
straight line in log-probability paper. The turbidity data in Figure 22 are
based on 113 days run for Scheme A, 112 days run for Scheme B, and 38 days
run for Scheme C. As the frequency curves indicate, the turbidity levels of
the filter influent and effluent with Scheme A pretreatment were consistently
lower than those with Schemes B and C pretreatments. The filter influent
turbidity ranged from 1.0 to 15 FTU for Scheme A, 1.5 to 16 FTU for Scheme B,
and 2.8,to 32 FTU for Scheme C. Turbidity of the filter effluent ranged from
0.2 to 2.4 FTU for Scheme A, 0.3 to 3.7 FTU for Scheme B, and 016 to 3.7 FTU
For Scheme C. Moreoever, the frequency curves show that 50 percent of the
time the filter effluent turbidities were equal to .or less than 0,6, 1.1, and
1.3 FTU for Schemes A, B, and C, respectively. The observed median values'o'f
the filter effluent were about the,same as the geometric means (50 percent
observations) indicated above. The log standard deviations of the filter
effluent were 1.7, 1.6, and 1.4 FTU for Schemes, A, B, and C, respectively.
Figure 23 presents a log-probability plot of suspended solids removal
data through the dual-media filter for the three pretreatment Schemes A, B,
and C. The frequency curves show that although the filter influent suspended
solids concentrations were approximately of the same level in the three pre-
treatment schemes, the filter with a chemical coagulation-sedimentation pre-
treatment (Scheme A) consistently showed lower levels of effluent suspended
solids than those with an in-line coagulation pretreatment (Schemes B and C).
Moreover, the frequency curves show that 50 percent of the time the filter
effluent suspended solids concentrations were equal to or less than 0.95,
2.3, and 2.5 mg/1 for Schemes A, B, and C, respectively. The median values
of the filter effluent suspended solids were approximately the same as the
geometric means indicated above. The log standard deviation of the filter
effluent suspended solids was 2.3 mg/1 for Scheme A, 2.0 mg/1 for Scheme B,
and 1.8 mg/1 for Scheme C.
It is recognized that of the many variables in filtration, the concentra-
tion as well as the physicochemical nature of the influent solids are the pri-
mary determining factors that influence the overall filter performance. Thus,
any pretreatment could drastically alter the physicochemical make-up of the
influent solids which could cause a corresponding change in the filter per-
formance. With this in mind, a regression analysis was performed to determine
55
-------
y
>*
co &
W | T j ป
*-: x. K
| EQ
tO10
UJ
CO
UJ g
J^co
0- Q UJ
ill
<ฃ ซd! 2
F
^00
u_ ^ ^.
c
c
x
1 1 N*
^m ^
- o ฐฐ "o -
oฐฐo H
o* to ~u
ฐo 2 oo _
o O >
) 10 O "C
5 N- IO
1 1
^ . . ซ
0 8
o
o o8
0ฐ
_
Q
o
o
o 0 ฐ _
Q
O
8 I
Q
o
o _
o
- ~ ft
8
O
8
- V* "
"o'ฐ o
o
ฐ ฐ o0
_ o
Q
Q
Q^
ฐo
~ ฐ o ""
0 ฐ
8
ฐ 0
8 :
o
1 1
5 CM C
O
f.f\
JO
O
CVj
O
O
g
o
o v
CO >
4f!
^
C
Is- C
Q
U
Q
O f9
to 1
"
' <
Sa
u,
Q
c
o
M-
O
ro
O
(M
g
*~\
0
0)
o
o.
CD
(D
o
o
o
0>
T3
jQ
C)
OJ
nu -gym
56
-------
100
_, 75
|
LU
CC
S5 50
/
4
0
r>
u.
CD
S 2
U:
IJL
UJ
o:
UJ
5' '
u.
o
pll TDATinM DATP ___ f\nnm/ft'<1
A\/C oni VMPD nncF- ____ ___rtncm/i/t
MVV3. rVJLT mC.r( UUOC. v/.UO (Tig/ 1
1 I 1 "1"
0 o
o ฐฐo 0ฐ
0 o o oฐ
-0 . ฐ 0
0 O rt O ฐO
o
0 O ฐ
o
UNIT CONVERSIONS^
gpm/ft2 x0.68= 1 /sec/m*
/
> (V
' 14
1 1 ฐ 1 1
o 1 1 oo 1 1
3.2 3.7 3.3
o
ฐ 0 0ฐ ฐ ฐ
oฐ <>0oo Q0 o QOO ooo o
"ฐฐ- o oฐ ฐ
0
1 1 1 1
10 20 30
OPERATING PERIOD, DAYS
40
50
Figure 21. Turbidity removal through the filter with Scheme C
pretreatment.
57
-------
O CO
_cg
to
O
CO
TJ
>ป
u
CD
CT
Q)
CVJ
CM
ง
nid
58
-------
100
80
60
40
30
_ 10
w ฎ
* 6
CO
3 4
CO
Q 2
UJ *
Q
uu
a.
CO
CO
1.0
.8
.6
.1
OPERATING CONDITIONS:
FILTRATION RATE *- 5 gpm/ff *
AVG. ALUM DOSE, mg/l
SCHEME A SCHEME B SCHEME C
148.5 5.6 6.1
AVG. POLYMER DOSE, mg/\
SCHEME A SCHEME B SCHEME C
0.20 0.06 0.06
SCHEME
. B
I 8 i
FILTER-
INFLUENT
SCHEME
FILTER-
EFFLUENT
UNIT CONVERSIONS;
gpm/ft2 x0.68= l/sec/m*
SCHEME A
.1 .2 .5 I 2 5 10 20 40 60 80 90 95 98 99.9
PER CENT OF OBSERVATIONS EQUAL TO OR LESS THAN PLOTTED
VALUE
Figure 23. Frequency distributions of filter influent and effluent
suspended solids.
59
-------
the relationship between the filter influent and effluent suspended solids for
the three types of pretreatment schemes. Figure 24 presents the straight line'
of best fit obtained by least squares linear regression analysis. In ex-
amining the plots in Figure 24, two observations are evident. First, for any
of the three pretreatment schemes, the filter effluent suspended solids levels
increase with increasing influent suspended solids concentration. Secondly,
the plots demonstrate the effect of type of filter pretreatment on the filter
performance. For instance, for any given filter influent suspended solids
concentration, the Scheme A pretreatment show consistently better filter per-
formance than those of Schemes B and C. Moreover, for the three pretreatment
schemes, the correlation coefficients exceed the 95 percent confidence coeffi-
cient. This would indicate that, with 5 percent chance of error,,effluent
suspended solids is dependent on influent suspended solids levels.
In Figure 25 are presented the relationship between filter effluent and
effluent turbidities similar to those shown in Figure 24. The'straight line
plots in this figure, which were determined by least square analysis, show
similar trends as those with suspended solids. The results of the regres-
sion analysis indicate that at the 95 percent confidence levels, only Schemes
A and B show significant correlation. Scheme C did not show significant
correlation even at the 90 percent confidence level and this is clearly in-
dicated by the regression line with almost zero slope.
Head!oss Data
The headloss buildup through a granular filter media is influenced by
several factors, the more significant of which are hydraulic surface loading
rate, the nature and concentration of influent solids, media size and fre-
quency and type of filter backwash. In the course of the long-term filter
evaluation, the two dual-media pressure filters were operated in parallel
under identical conditions of hydraulic surface loading rate and backwash
procedure. Thus, the magnitude of pressure drop across the inert media
filters would depend primarily on the nature and concentration of influent
suspended solids. Cognizant of this, a linear regression analysis was per-
formed in an attempt to determine the relationship between the influent
solids concentration and the headloss buildup. In performing the regres-
sion analysis, the influent solids concentration was expressed in three
different parameters; namely, turbidity in FTU, suspended solids in mg/1, and
solids capture in Ibs/ft2/run. For the three pretreatment schemes, attempts
were made to correlate each of the three influent solids parameters with
the total headloss across the filter after 16 hours of filter run. The re-
sults of the regression analysis indicate that only Scheme A showed signi-
ficant correlation, at 95 percent confidence level,of all the three influent
solids parameters with total headloss. In Schemes B and C there was no
correlation found between the headloss and any of the three influent solids
parameters. Figure 26 presents the correlation between filter influent sus-
pended solids and total headloss. A plausible explanation for the absence
of correlation in Schemes B and C between the influent solids parameters
and total headloss could be attributed to the variability of the physico-
chemical characteristics and concentration of the influent solids during
the course of the filter runs. It is important to recognize that because
60
-------
lls
CO
Q
O
CO
Q
UJ
Q
UJ
CL
CO
3
CO
UJ
u.
w
T3
I
3
0)
IE
_3
ซซ-
M
Q>
O
*-
CD
Q>
/**"
'c
fli
Q)
1
.2
B
a>
o
O
si
ZS
o>
IL.
61
-------
CJ
E
u
to
*-*
II
00
CO
o"
X
N
i
<0
CVS
to
O U-
g
m
o:
I-
UJ
CO
o:
UJ
b
u.
Q>
OJ
o
O
Q>
fO
CM
nid 'Aiiaiadni
"cu
(5
O
>d
CM
g>
L,
62
-------
oc woo
2
o>
.ฃ ซ> *R .
IO
O*
E
*ป
CO
o
!O
(M
O
CM
O
UJ
o
z
UJ
Z)
CO
z
UJ
3
!2 _l
u,
Cฃ
UJ
b
o u_
IO
o
CM
isd
"SdH 91
ID 10 o
SS01QV3H 1V10ฑ
to
"O
o
0>
o
CO
w T=;
o
CO
a.
CO
CO
c:
0)
ZJ
03
a>
c
o
Q>
t_ i
t>
O
o
CO
CVJ
ง'
0>
63
-------
of this influent solids variability, a major portion of the resulting headless
buildup could have been triggered by high solid input to the filter during a
certain period in the course of the filter run. The influent solids parame-
ters used in the regression analysis were based on the test of 16-hour com-
posite samples which in effect reflected the average solids concentration
during the 16-hour period. With Scheme A pretreatment, variations in the
secondary influent suspended solids were reduced by the equalizing effect of
the chemical clarification-sedimentation system. Therefore, the influent
solids concentration applied to the filter during the filter run was essen-
tially constant and could be appropriately represented by the test of the
16-hour composite samples.
Figure 27 presents a plot of headless buildup across the dual-media fil-
ter during the course of several selected filter runs for pretreatment
Schemes A, B, and C. As shown in the figure, the headless levels with
coagulation-sedimentation'pretreatment (Scheme A) were considerably lower
than those of in-line coagulation pretreatment (Schemes B and C). For in-
stance, at the end of 16 hours of filter run, the headless across the dual-
media filter ranged from 0.056 to 0.37 Kg/cm2 (0.8 to 5.3 psi) for Scheme A,
and 0.44 to 0.86" Kg/cm2 (6.2 to 12.2 psi) for Schemes B and C. It is inter-
esting to note that the range of headless in Scheme B was essentially the
same as that in Scheme C.
In Figure 28 are presented arithmetic-probability plots of total head-
loss across the filter after 16 hours of filter run. The headloss data pre-
sented in this figure are based on 103 days data for Scheme A, 93 days data
for Scheme B, and 34 days data for Scheme C. As indicated by the frequency
curves, the headloss levels in the filter with Scheme A pretreatment were
appreciably lower than those observed in the filter with Schemes B and C
pretreatments. The headloss across the filter ranged from 0.03 to 0.37
Kg/cm2 (0.4 to 5.3 psi) for Scheme A, 0.32 to 0.99 Kg/cm2 (4.5 to 14.1 psi)
for Scheme B, and 0.32 to 0.93 Kg/cm2 (4.5 to 13.3 psi) for Scheme C. More- .
over, the frequency curves indicate that 50 percent of the time the headloss
levels were equal to or less than 0.13 Kg/cm2 (1.9 psi), 0.6 Kg/cm2 (8.5 psi)
and 0.62 Kg/cm2 (8.5 psi) and 0.62 Kg/cm2 (8.8 psi) for Schemes A, B, and C,
respectively. The median filter headloss was 0.12 Kg/cm2 (1.8 psi) for
Scheme A, 0.58 Kg/cm2 (8.3 psi) for Scheme B, and 0.62 Kg/cm2 (8.8 psi) for
Scheme C.
64
-------
28
24-
20
CO
ง '6
LU
12
8
4
FILTRATION RATE-
o SCHEME A-
A SCHEME B--
D SCHEME C
5gpm/ft.s
ALUM
(mg/1)
148.5
5.6
6.1
POLYMER
(mg/l)
0.20
0.06
0.06
UNIT CONVERSIONS:
gpm/ft*x I/sec/ma
psi x 0.0703s kg/cma
8 12 16 20
FILTER RUN, hours
24
Figure 27. Effect of pretreatment on headless.
65
-------
0}
T3
O
a>
JC
Q)
CD ^ CVJ O 00 (D SfCVJ
*Nna aanid 'SUM 91 ซ3idv ssonavaH
tn
u
c
(U
er
0)
00
66
-------
SECTION 6
ECONOMIC ANALYSIS
The cost estimate presented in this section is based on the treatment of
Pomona activated, sludge plant effluent for an average design flow of 37,850
cu m/day (10 MGD) and a peak design flow of 52,990' cu m/day (14 MGD}. The
process design parameters for sizing the various treatment components are pre-
sented in Table 9. THe unit costs for chemicals and other direct costs for
estimating operation and maintenance (0/M) costs are summarized in Table 10.
In Figure 29 is presented the schematic layout of the proposed filtration sys-
tem with two types of filter pretreatment schemes. The various treatment
units of the overall tertiary system, which are included in the cost estimate,
are indicated ,in the figure. The capital cost estimates include the cost'of
all equipment, installation and construction costs, startup and testing, and a
20 percent allowance for contingencies, plus a 15 percent allowance for engi-
neering costs. The cost of land, sludge treatment facility, chlorination sys-
.tern and interest during construction are not included in the cost estimate.
In addition, the cost of unusual construction requirements such as rock ex-
cavation, site dewatering and extensive demolition work are not included in
the cost estimate.
It must be recognized that the cost estimates presented in this report
are preliminary in nature and are used only as basis to reflect the relative
cost of the filtration system with two types of pretreatment schemes. The
actual construction bid costs of three Sanitation Districts inert media fil-
tration systems varying in size from 47,312 to 141,938 cu m/day (12.5 to 37.5
MGD) as well as data from literature (11,12,13) have served as a major basis
in the preparation of the cost estimate. The estimate of construction costs
presented are based on ENR construction cost index of 2584 for July, 1977.
In Table 11 is shown the complete cost breakdown of the inert media fil-
tration system. The total treatment cost to produce filter effluent with
characteristics similar to those presented in Tables 6 and 7 from a 37,850 cu
m/day (10 MGD) plant is estimated at 4.27<ฃ/m3 (15.99(^/1000 gallons) for Scheme
A, and 2.29<ฃ/m3 (8.58^/1000 gallons) for Scheme B. The capital costs of the
pretreatment system represent about 29 percent of the overall capital cost for
Scheme A compared to only 6 percent for Scheme B. Moreover, in evaluating the
0/M of each scheme, it is shown that, for Scheme A, the chemical cost alone
represents about 63 percent of the 0/M cost and about 36 percent of the total
treatment cost. For Scheme B, on the other hand, the chemical cost represents
only 9 percent of the 0/M cost and about 3 percent of the total treatment
cost. Thus, in comparing the effluent quality from the two schemes in
67
-------
TABLE 9. FILTRATION SYSTEM DESIGN DATA
PRETREATMENT SYSTEM
1. Rapid Mixing:
Detention time, minutes 1.0
Chemical Dosage
Alum, mg/1 150.0* (5.0)+
Polymer, mg/1 0.3* (0.06)+
2. Flpeculation:
Detention time, minutes , 45.0
3. Sedimentation:
Detention time, hours 1.5
Overflow rate, m3/day/m2 36.6
INERT MEDIA FILTRATION SYSTEM
1. Filtration:
Hydraulic Surface Loading, 1/sec/m2 2.7-4.1
Backwash Flow Rate, 1/sec/m2 12.2-13.6
Backwash Volume, % of plant flow 2.5* (5)+
Air scour, 1/sec/m2 ' 15.2-25.4
* for Scheme A + for Scheme B
68
-------
TABLE 10. UNIT COST FOR OPERATION AND MAINTENANCE ESTIMATE
CHEMICALS
Alum, $/Kg Al
Polymer, $/Kg
1.00
4.40
OPERATING COSTS
Power, $/Kwh
Backwash Water, <ฃ/m3 (Backwash/day =
1 for Scheme A and 2 for Scheme B)
Operating and Maintenance Labor,
$/person-yr (4 for Scheme A and
3 for Scheme B)
Laboratory Personel, $/person-yr
Maintenance Materials, $/yr
2.50
0.80,
12,000
15,000
20,000
CAPITAL COSTS
Capital costs were amortized at 7% for 25 years
69
-------
UJ
u.po.<
lUh-Ol-
CQ
LU
2
UJ
X
o
QUJ
ri|
_1
u.
z
o
T3
0)
Q.
O
(O
2
oป
o
1
0)
I
0>
3
70
-------
TABLE 11. ESTIMATED FILTRATION SYSTEM COST*
CAPITAL COSTS, (1000 of $)
Scheme A
Scheme B
1. PRETREATMENT SYSTEM
Chemical Feeding System
Chemical Coagulation-Sedimentation
System
Sub-Total
Contingencies (20%)
Sub-total
Engineering (15%)
Total Pretreatment System Costs
Amortized Cost, (tf/m3)
2. FILTRATION SYSTEM
Pumping Station
Inert Media Filtration System
i Sub^-total
Contingencies (20%)
Sub-total
Engineering (15%)
Total Filtration System Costs
Amortized Cost, (<ฃ/m3)
OPERATING AND MAINTENANCE COSTS (ซฃ/m3)
Chemicals (Alum and Polymer)
Power
Backwash Water
Operating and Maintenance Labor
Maintenance Materials
Total Operating and Maintenance Costs
Total Treatment Cost (<ฃ/m3)
113.70
506.00
619.70
123.90
743.60
111.50
855.10
0.54
171.00
1,330.00
1 ,501 .00
300.20
1,801.20
270.20
2,071.40
1.30
1.54
0.27
0.02
0.46
0.14
2.43
4.27
102.00
102.00
20.40
122.40
18.40
140.80
0.09
171.00
1,330,00
1,501.00
300.20
1 ,801.20
270.20
2,071.40
1 .30
0.08
0.27
0.04
0.37
0.14
0.90
2.29
* Based on ENR construction cost index of 2584 (July, 1977) for a 37,850
cu m/day (10 MGD-) plant.
, 71
-------
addition to the economic analysis, it is apparent that an inert media filtra-
tion system with Scheme B pretreatment is the most practical and economically
feasible choice for the removal of suspended and colloidal materials from an
activated sludge plant effluent.
72
-------
REFERENCES
1. Cookson, J.T., "The Use of Temporary Wastewater Treatment Plants:
Standards and Procedures for Elimination of Health Hazards," report pre-
pared for Montgomery County Council, State of Maryland (Dec. 1972).
2. Melnick, J.L., "Detection of Virus Spread by Water Route," 13th Water
Quality Conference, University of Illinois (1971).
3. "Standard Methods for the Examination of Water and Wastewater," 13th Ed.,
American Public Health Association, New York (1971).
4. "FWPCA Methods for Chemical Analysis of Water and Wastes," Federal Water
Quality Administration, Cincinnati, Ohio (Nov. 1969).
5. Cleasby, J.L., Strangl, E.W., and Rice, G.A., "Developments in Backwashing
of Granular Filters," Proceedings of the American Society of Civil
Engineers, Journal of the Environmental Engineering Division, Vol. 101,
No. EE5, pp 713 (Oct. 1975).
6. Amirtharajah, A., and Cleasby, J.L., "Predicting Expansion of Filters
During Backwashing," Journal of American Water Works Association, Vol.
64, pp 52-59 (1972).
7. O'Welia, C.R., "The Role of Polyelectrolytes in Filtration Process,"
EPA 670/2-74-032, U.S. Environmental Protection Agency (April 1974).
8. Hutchison, W., and Foley, P.O., "Operational and Experimental Results of
Direct Filtration," Journal of American Water Works Association, Vol. 66,
No. 2, pp. 79-87 (Feb. 1974).
9. Adin, A., and Rebhun, M., "High Rate Contact Flocculation-Filtration with
Cationic Polyelectrolytes," Journal of American Water Works Association,
Vol. 66, No. 2, pp 109-117 (Feb. 1974).
10. Robeck, G.G., and Kreissl, J.F., "Multi-media Filtration: Principles and
Pilot Experiments," U.S. Dept. of Interior, FWPCA, Cincinnati, Ohio
,(1967).
11. "Process Design Manual for Suspended Solids Removal," EPA 625/1-75-003a,
U.S. Environmental Protection Agency (Jan. 1975).
73
-------
12. Smith, R., "Electrical Power Consumption for Municipal Wastewater
Treatment," Environmental Protection Agency Technology Series, EPA-
R2-73-281, U.S. Environmental Protection Agency (July, 1973).
13. Patterson, W.L., and Banker, R.F..
Requirements for Conventional Wastewater Treatment
Pollution Control Research Series, 17090 DAN 10/71
Protection Agency (Oct. 1971).
"Estimating Cost and Manpower
Facilities," Hater
U.S. Environmental
74
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-80-148
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
EFFECT OF PRETREATMENT ON THE FILTRATION OF
LOW TURBIDITY SECONDARY EFFLUENT
5. REPORT DATE
August 1980 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Leon S. Directo, Ching-Tin Chen
and Robert P. Miele
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
County Sanitation Districts of Los Angeles County
Whittier, California 90607
10. PROGRAM ELEMENT NO.
1BC611 SOS#5
11. CONTRACT/GRANT NO.
14-12-150
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final - 4/74-5/76
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Irwin J. Kugelman (513) 684-7633
16. ABSTRACT
A pilot study of filtration of secondary effluent was conducted. Turbidity
and solids removal were similar for dual vs. trimedia filters, but headless
was higher across the latter. Coagulation-flocculation and sedimentation
pretreatment resulted in a filter effluent superior to that when in-line
coagulation alone was used and a lower rate of head loss build up. However,
the latter produced acceptable results (s.s = 2.7 mg/1, FTU = 1.2) at much
lower cost. Filtration rate in the range 5-10 gpm/ft2 (3.4-6.8 1/sec/m2)
had no effect on performance except rate of headless build up.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATl Field/Group
Sewage Treatment
Filtration
Coagulation
Filter Media
Suspended Solids
Activated Sludge
Effluent
138
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
85
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
75
U.S. GOVERNMENT PRINTING OFFICE; 1980657-165/0132
-------
-------