v>EPA
t
United States
Environmental Protection
Agency
Technology Transfer
CERI-87-49
Workshop on
Emerging
Technologies for
Drinking Water
Treatment:
Filtration
EPA
CERI
87-
"49
-------
'•
-------
(V,
CERI-87-49
WORKSHOP ON EMERGING TECHNOLOGIES FOR
DRINKING WATER TREATMENT: FILTRATION
September, 1987
Sponsored by:
United States Environmental Protection Agency
Office of Drinking Water and Office of Research and Development
and
Association of State Drinking Water Administrators
Information Resources Center
US EPA (3404)
. 401 M Street SW
Washington, DC 20460
-------
-------
Slide Presentation
-------
^
CO
Uj
<3£
!*rf
CO
LU
a.
Q.
D
CO
LU
O
OC
z>
CO
fe <
O o
= <*
°s
^3
< <
oc 1
LU
ID
OC
H-
Z
LU
LU =
tz
g
i-
UJ^-
<1
u. <
OC
CO
LL
O
LU
CC
z
o
h-
LL
O
o
o
03
u- ~
Q o < g
I 3 § B
So<5
H O (3 U
-------
o
M ^k
JO
UJ
<5
^•F
o
•j
o
5
1:
t\
XM^
UJ
K
e
o
^^F
P:
5
EE
-J
5;
H
2 •
UJ
E
<
UJ
_J
z
o
H-
z
UJ
>
z
o
0
•
ST
DC
LU
K-
E
UJ
DC
CO
CO
UJ
DC
z°-
O 0
iz z
<<
^t
^^J J>^^
c<
, DC
*™ rn
o £
UJ
DC
O
• -
g
uZ
<
DC
H
il
Z
DC
UJ
CO
O
UJ
o
0
<
o
•
z
g
•
<
DC
«J
LL
•M^M
Q
Z
CO
w^
1
0
CO
•
CO
H
z
Q.
UJ
O
^
^
O
<
Q.
•
CM
UJ
O
to
-------
REVISED
DRAFT FINAL
TECHNOLOGIES AND COSTS
FOR
THE TREATMENT OF
MICROBIAL CONTAMINANTS
IN POTABLE WATER SUPPLIES
APRIL 1987
SCIENCE AND TECHNOLOGY BRANCH
CRITERIA AND STANDARDS DIVISION
OFFICE OF DRINKING WATER
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C.
SLIDE 3
-------
CO
O
s
O
t
CM
OCM
W™
eg _L Q.
.. cc o
Q I- ~
F cc o
-t
2 CM Q
Q-.' Z
E O LU
Q.
UJ
UJ
UJ
UJ
I
LU
-------
CONVENTIONAL TREATMENT
• Complete treatment (coagulation, flocculation sedimentation, filtra-
tion, and disinfection)
0 Most widely used
• Process design criteria well established
t Traditional concrete construction
SLIDE 5
-------
O
UJ QCCM
X UJ h-
= t- U.
2 _j
vo
UJ
O
I—t
_l
CO
-i 5
< UJ
ii
3 S
CO
O
O
x 5 Z
I CM O
* i P
goS
°- HI t-
S co uj
01 0°
CO
UJ
U.
-------
DIRECT FILTRATION WITH FLOCCULATION
• Pretreatment Including flocculatlon
• No sedimentation step
• Limited to good quality water supplies
- Turbidity less than 5-10 NTU
- Color less than 20-30
- Coagulant dosage less than 20-30 mg/1
SLIDE 7
-------
O
UJ
X
I
CC
o
§
o
DC CM
UJ H
•H U.
•!!
< o
Q LO
UJ i
5 *
UJ
5
1 _
o< 2
CLZ Q
5t V*
§i ^
zo: z
- i O
4 1 F
CC
CO
ffi
DC h-
X O
CO
AGULANT
O
O
|
X
1
o
&
CC
i
H
2
O
o
I;;
^
o
UJ
DC
5
o
a UJ
LL
5
^
Z
CM O
o ul
UJ I-
CO UJ
0 °
^
2£
u.
o
tL
UJ
CO
o
k «J
p u.
GO
CO
-------
DIRECT FILTRATION WITH CONTACT BASINS
t One hour basin for pretreatment
• Used to level out turbidity spikes
• Provides silt and sand removal
• Smooth out coagulant demands
SLIDE 9
-------
5 s
o <
a- z
o Q
z u.
o
UJ
X
1
QC
III
-I
LL
5
UJ
Q.
o
10
IS
CO
CO
a
t~i
_i
to
oc
O
UJ
oc
S
iu
(0
o
o
oc
? I
O UJ
S
8
UJ
UJ
CO
o
u.
-------
DIRECT "IN-LINE" FILTRATION
• Commonly used with pressure fitters
• Iron and manganese removal groundwater
• Limited to low turbidity surface waters with less than 2-3 NTU sea-
sonal average
• Not permitted in many states
SLIDE 11
-------
S
O
N
(0
S~
_i
DC
LU
Q.
Q.
O
LU
O
Z
LU
«
RECOMIt
oc
§1
O
kl
— ,— ..
OD
••™ i—
fflz
0=5.
3
1-
0)
_,S ^
m
r-
0)
Z
O
L.
oo
ZE
t-
0)
LU
OC
O
o
o
o
CM
V
2 «
V V
^
T- O
N V
V
o o o
o o o
• m co
m y y
V
m
V
in
V
o
in
V
CM
O
o
o
•z
X
o
LU
1-
z
LU
S
h-
LU
OC
h-
H
Z
LU
• s
LU
OC
Z
o
1— .
ONVEN
^
Z
O
5
LU
U.
Z
. (75
S
LU
oc
a
O
H"
g.
• Z
o
LTRAT
LL
O
LU
OC
Q
z
O
p
oc
^
o
^
o
(0
z
o
p
oc
n
X,
oc
LU
(0
D
o
LU
o
o
MM '
o
-------
3
-j
OQ
§
^
Is
oo
u. et
8*.
DC OQ
LU
cc
w
o
cc
UJ
Q.
•
o>
G>
i
10
m
(O
O)
•
o>
o>
1
0)
in
o>
O)
O)
•
IO
o>
o
fs.
i
O
*"
o>
o>
•
O)
o>
1
O)
o>
*
o
I
CO
CO
UJ
o
cc
Q.
H
Z
««
Z
O
t-
<
_J
3
C5
<
O
o
I
H
^^^
5 _
APID FILTRATION
SEDIMENTATION
DC
O
P
<
_i
^
O
<
o
o
H
£
IRECT FILTRATION
Q
Z
WITH FLOCCULAT
i
NO COAGULATION
i
Z
0
P
<
^_
i
UL
H
OC
lATOMACEOUS EA
Q
Z
O
P
<
CC
d
UL
Q
<
CO
o
.J
CO
Z
Occ
P u
<•"
3 d
3 UL
* AIDED BY COAG
* FULLY RIPENED
*
-------
1.0
0.5
B
oc
Ul
u.
u.
Ul
0.1
0.06
0.011 1 L
i i i
i i
40
60
70 80 90 100
CYST REMOVAL -
RELATIONSHIP BETWEEN CYST REMOVAL AND FILTERED
WATER TURBIDITY (AFTER ENCESET)
SLIDE 14
-------
SPIKED GIARDIA PILOT STUDY
LAKE ARROWHEAD, CALIFORNIA
• Adsorption clarification/filtration process
t 20 gpm pilot plant removal groundwater
• 2,100 cysts per liter influent concentration
• 100 percent removal by filtration
t Filtered water turbidity 0.05-0.06 NTU
t Clarified effluent contained cysts at 0.3-0.4 NTU turbidity
SLIDE 15
-------
u.
O
CO
Uj
p
•
•
-j
25
2
2
.j
3
^
o
§
Ul
QC
CO
-J
<
o
5
•
*
CO
Ul
CO
CO
UJ
O
O
oc
CL
O
P
^
£
»J
£
UJ
cc
o
o
_J
5
_|DC
<0
h- UL
0 j
>-o
o
Q CO
rr u
£§
H D
zE
UJ>
-CO
QH
occo
<>
5^
^^k
^
A
CM
i
T-
co
1
CM
CO
CO
UJ
o
o
cc
CL
•
1-
^
UJ
<
Ul
QC
H
_j
<
O
u!
i
- z
Ul
>
z
O'
O
CO
•
,_
CM
i
r»
CO
i
CM
FILTRATION
0
UJ
OC
Q
CM
i
T~
CO
1
T~
CO
1
CM
^^
\ND FILTRATION
CO
O
_j
CO
CO
i
T-
co
1
^
CO
1
CM
Z
P
<
H
i
^v^
LL
CEOUS EARTH
<
O
<
5
z
UT DISINFECTIO
O
I
H
i
*
l/J
I
-------
POTENTIAL REMOVAL EFFICIENCIES
PARAMETER
Coliforms
Turbidity
Enterovirus
Giardia
CLARIFIED
90-99
50-70
70-99
40-75
FILTERED
98-99.99
90-99
90-99.9
96-99.9
FINISHED
>99.99
90-99
98-99.9
96-99.99
SLIDE 17
-------
WHAT IS IT GOING TO COST?
SLIDE 18
-------
WHAT DO COSTS INCLUDE?
• Capital (construction, engineering, administration)
t Operation and maintenance (energy, maintenance, materials and labor)
• Total costs (capital costs at 20 years, 10 percent Interest plus
• Total costs presented as cents per 1,000 gallons
SLIDE 19
-------
ESTIMATED COSTS - FILTRATION PROCESSES
LARGE SYSTEMS
Conventional Treatment
Direct Filtration
Diatomaceous Earth
Total Cost
0/1,000 Gal
30-60
20-50
35-50
SLIDE 20
-------
ESTIMATED COSTS - FILTRATION PROCESSES
MEDIUM SYSTEMS
Conventional Treatment
Direct Filtration
Diatomaceous Earth
Slow-Sand Filter
Total Cost
6/1,000 Gal
70-100
55-90
40-50
25-35
SLIDE 21
-------
op
Ife
Q.
gfe
31
.SB
^2
ojtt
Ul *•*•
o§
M
o£2
LU ><
K UJ
o
z
Q
O
QC
CL
Z
O
<
_J
=)
3-i
°2
of
io
5 o
<3 I
1
o
o
z
X
o
a.
OC
o
PRO
ADDING OR
UJ
o
<
0)
o
a
CONTROL OF COAG
z
O
<
z>
O
o
o
MODIFYING OR ADDING
CM
CM
EFFICIENCY
Z
o
<
H'
z
MG SEDIME
^^^
>
0
QC
a.
^l^H
CO
o
F
<
o
lAULIC MODIFK
c
Q
>•
Z
1
SETTLERS
UJ
CO
3
1
CO
QC
GE COLLECTO
O
3
_J
CO
1
.TRATION MED
.^
EE
h-
Z
UJ
O
u.
u_
LLI
UJ
OC
0
s
-J
<
(/>
-------
K.
^
-J
o
§
o
u.
o
UJ
K.
Ul
CD
2
O
(/)
<
(f>
z
o
DC
O
s
3
•
o
z
O
cc
O
0
^~ III
ZUJ
M
-j >
Is
00
O j
-J
J 111
< 2
LU
•
0
z
o
DC
o
0
r^
H CC
3 j
CD O
o o
^ z
02
J <
-------
CL
<
O
UJ
o
h--
QC
<
Q.
CO
UJ
>
O
QC
s o
LU
O
u.
u.
UJ
CO
UJ
2
o
o
QC
O
LL
QC
<
O
QC
UJ
UJ
CO
UJ
>
UJ
x
Q
^
UJ
UJ
QC
O
ffi
CC
QC
UJ
CO
CO
-------
in
CO O
Z LU
T= LL
DC
LU
H
LU
CO
DC
O
o
5 DC
DC
LU
-------
I
CO
DCO
8*
si
z
O
CM
CO
H :.
Z
_l
O
0
o
LU
O
X
.LONS TO
GALLONS2
^
(9e
§5
"O
^ i^
in
CO
fe
o
CO
LU
H
Zj
0
LL
0
LU
LU
I •
LL
DC
LU
S"
i
^^
o
0.
Q
O
CM
.LONS TO
) GALLONS
<§
OT--
If
^ i^
O
CO
0
-
CO
LU
|_
j
O
LL
Q
LU
LU
LL
Q.
Q
Q
CSJ
O
00
Z«M
— CMOC
C/ QJ
iu
o°,
*^CM
HO
T^ CM
-------
CD
g
Q
5;
2
z
O
co
DC
LU
CL
CO
O
a
CL
^
DC
O
Z
-------
I
§1
-'
•'
..
jf> .
"•—COAGULANT
FEED LINE
^4" CRES PIPE
.14" 14" 14'
< I
X
(i
COAGULANT
DIFFUSER
g*
•*",'*
m
n
*
,
t
MULTI JET '•
SLIDE
GATE "»
"X. _
•J^
J)
;OAGULANT/
k
P
]
]
N
:|
7\
'•it
p
z>
jjjj
Zl
a
•w
/
WATER
.FLOW
gfr
•^*-
FRONT VIEW
SECTION VIEW
COAGULANT DIFFUSER
SLIDE 28
-------
DIFFUSERS
_Q_
RAW WATER
MIXER
SEE DIFFUSER
DETAIL BELOW
[
RAW WATER
EXISTING RAPID MIX
ARRANGEMENT
PROPOSED DIFFUSER
ARRANGEMENT .
REFLECTED PLAN
HOLLOW PLASTIC HUB
CROSS SECTION OF DIFFUSER
COAGULANT
SOLUTION
BRACES FOR
SUPPORT
DIFFUSER FOR COAGULANT APPLICATION
SLIDE 29
-------
MIXING
Q
ill
CO
D
1
O
o
o
CO
o
*6
CO LU
O CO
£°
O o
LU i-
"o
o c
CO
LU
I- >
LU
1-
LU
0
o
CO
LU
a
O
LU
CO
LL
O
OL
X
CM
O
LL I-
O ,-
DC
LU
O
-------
>s
CO
•MM
a.
UJ
O
DC
-1
O
h™
LU
m
o
Q.
Q.
1
••MB
(CT
0
LL
o
UJ
Q.
CO
CO
z
LU
H
Z
O
x
5
^c
UJ
Q
>
O
DC
Q.
O
Q.
Z>
Q.
CO
UJ
CO
D
YI
C>
o
Y™1
O
I-
DC
Z
o
F
Z)
Q
D
d
2
;
X
^
s
1
o
UJ
CO
O
0
o
ci
O
h"
O
O
o^
CO
UJ
^•^
-I
1
o
^
o
Q.
H
O
LU
CO
q
Y-
O
|H^
in
CO
' LU
••••
. §
h
Z
O
H
Z
LU
h-
Ul
Q
to
-------
LU
O
^^^m
o
<
u.
O
x
i
HI
o
DC
LLJ
<
rr
-------
cc
§
§
O
CO
CO
o
UJ
X
o
LL
O
CD
Z
X
s
CO
13
O
UJ
STA
CO
LU
9
O
DC
Q.
o
^
0
o
o^
crT
u.
O
UJ
O
z
DC
EE
CO
UJ
3
-J
<„
^b
O UJ
">~
s§
> in
O
DC
QL
O
UJ
O
Z
UJ
s
o
o
UJ
DC
Q
:
O
o
UJ
CO
in
0
LL,
0
CO
UJ
51
H
UJ
0
Z
LU
Q
CO
UJ
DC
CO
O
o
UJ
>
CO
in
•
o
LL,
O
0.
z
cc
UJ
o
Q.
a,
x
LU
|
CO DC
CO LU
yL3
<
LLY
DC
UJ
<£
tt fc
uj O
z
UJ
o
-------
C
o
2
o>
T>
Ol
8
Ol
u
i
LU
ai
O
cj
y
<
ITS
>•>»
)
VI
^
Q
o
SLIDE 34
-------
I
U. CO
OUJ
coo.
o^
occ
z
o
0)0
?°
O o
J O
o
o
o
in
CO
z
CO
m
x
s
g
Q.
o
D
DC
I-
C/)
O
o
o
Q.
o
o
o
o
CM DC
111
UJ H
O <
z LU
-------
ADDING CHEMICAL DOSAGE CONTROL
• Mandatory for low filtered water turbidity
• Common dosage control techniques
- Jar tests
- Particle charge
- Pilot filters
SLIDE 36
-------
PARTICLE CHARGE DEVICES
t Streaming current detector
• Measures relative particle charge with electrodes
• Electrode output current related to Zeta potential
• Use output to control chemical feed pumps
• Calibrated from jar test results
SLIDE 37
-------
SYNCHRONOUS
MOTOR
LOOSE FITTING PISTON OF "*•
INSULATING PLASTIC
CUTAWAY OF CELL BODY AND PUMP
BORE OF INSULATING PLASTIC
SYNCHRONOUS
RECTIFIER
SIMPLIFIED DIAGRAM OF SCO INSTRUMENT
SLIDE 38
-------
COAGULATION CONTROL WITH PILOT FILTERS
• Direct measurement of fllterability of coagulated plant water
• Determines adequacy of coagulant dosage
0 No correlation to other test procedures required
• Use output signal to control coagulant feed pumps
t Precise coagulant dosage control can save costs
SLIDE 39
-------
o
o
CO
25
S
O
sl
o
o
u. o
.t
(0
j»
U
SLIDE '40
-------
IMPLEMENTING FLOCCULATION IMPROVEMENTS
• Improves clarification efficiency
• Generally required when increasing capacity
• Detention time requirements: 10-40 minutes
t Mixing intensity G values: 10-100 seconds
• Numerous design types widely used .
- Slow-speed paddles
- Vertical turbine
- Oscillating paddles
SLIDE 41
-------
METHODS TO IMPROVE FLOCCULATION
• Install new mixing equipment
t Add baffling
• Improve inlet and outlet conditions
SLIDE 42
-------
Stator beam
stators
not shown
Rotors
Rotors; stators shown only In lower half
Horizontal
shaft
Direction of
displacement
.^ Rotors
^ and
stators
Longitudinal section:
stators not shown In upper half
Cross-section
of channel
Stators
a) Paddle type with rotors and stators.
b) Plate turbine type.
c) Axial flow propeller type with straightening vanes.
TYPICAL FLOCCULATION UNITS
SLIDE 43
-------
Biffles
t 1
i t
s v.
Sitft You
Battvn Ww
BaNtes
Si* Vim
Bottom Vtar
d) Schematic of radial flow
pattern In baffled tank.
e) Schematic of axial flow
pattern in baffled tank.
(CONT.X TYPICAL FLOCCULATION UNITS
SLIDE 43
-------
I
Axial Flow
Turbine
Propallar
Turbin*
f) Basic impeller styles,
g) Walking beam flocculator.
(Cont.). TYPICAL FLOCCULATION- UNITS
SLIDE 43
-------
INSTALLATION OF WALKING BEAM FLOCCULATOR
SLIDE 44
-------
NEW- AND OLD-STYLE FLOCCULATORS
SLIDE 45
-------
IMPROVING INLET CONDITIONS AND BASIN SHORT-CIRCUITING
• Major overlooked problem
o Contributes to unbalanced coagulant dosage
t Breakup of floe
• Reduces effective detention time
SLIDE 46
-------
CORRECTING FLOCCULATION HYDRAULIC PROBLEMS
• Install Inlet diffusers
• Add baffling to compartmentalize basins
0 Enlarge connecting conduits
SLIDE 47
-------
SEDIMENTATION BASIN DEFICIENCIES
o Poor inlet conditions
• Basin turbulence
• Excessive clarification rate
• Improper outlet facilities
• Inadequate or no sludge removal
SLIDE 48
-------
TECHNIQUES TO IMPROVE SETTLING
• Inlet and outlet modifications
• Install sludge collection equipment
• Add tube or plate settling devices
SLIDE 49
-------
TUBE SETTLING
• Established technology
e Widely used
9 Reduce floe carryover In existing basins
e Permit increasing capacity up to 100 percent
o General application rates: 1-3 gpm/sq ft
SLIDE 50
-------
A MODULE OF STEEPLY INCLINED TUBES
(COURTESY OF MICROFLOC PRODUCTS
GROUP, JOHNSON DIVISION, UOP)
SLIDE 51
-------
INFLUENT ZONE
EFFLUENT
COLLECTION
ZONE
EXISTING LAUNDERS NEW LAUNDERS
NEW BAFFLE
y
i
\ *_(FLOWPATTERNl-»— — — -
^ ^ x-vf
-/ ^ ^ \
1(S," " S, I\\\ "V ^ --.^ I/', .'y / - J
" / * """ ^ """ ^ / T
r
TUBE
SUPPORTS
TUBE
MODULES
TYPICAL TUBE SETTLER INSTALLATION
IN RECTANGULAR BASIN
SLIDE 52
-------
,— fxfo AIB - L >i*e
rot* wa
— raouGH
cownrot
SECTIONAL VIEW THROUGH SEDIMENTATION BASIN
SHOWING VACUUM SLUDGE REMOVAL SYSTEM
SLIDE 53
-------
CLOSE-UP VIEW OF VACUUM SLUDGE REMOVAL ASSEMBLY
SLIDE 54
-------
TUBE SETTLING MODULES (TYPICAL
SEGMENTS)
TUBE MODULE
SUPPORT FRAME
LAUNDERS NOT SHOWN
TO PROVIDE CLARITY
NEW VARIABLE SPEED
CIRCULATION PUMP
PLAN
SOLIDS CONTACT CLARIF1ER WITH TUBE SETTLERS
SLIDE 55
-------
VARIABLE SPEED
RECIRCULATION PUMP
•TUBE SETTLING MODULES
•SLUDGE WASTING SUMP
COLLECTOR DRIVE
SLUDGE COLLECTOR
SLUDGE SLOWDOWN
SECTIONAL ELEVATION
(ContJ. SOLIDS CONTACT CLARIFIER WITH TUBE SETTLERS
SLIDE 56
-------
PHOTOGRAPH OF TUBE SETTLERS IN SOLIDS CONTACT CLARIFIER
SLIDE 57
-------
PHOTOGRAPH OF ALUM FLOC CARRIED UP THROUGH TUBE SETTLERS
SLIDE 58
-------
PHOTOGRAPH SHOWING ACCUMULATION OF SOLIDS ON SURFACE OF TUBE SETTLING MODULES
SLIDE 59
-------
TOTAL COSTS FOR ADDING TUBE SETTLING MODULES
Tube Settling Modules
Small plants (less than 100,000 gpd)
Medium plants (less than 6 mgd)
Large plants (greater than 10 mgd)
Total Cost of Modification
$0.009 to $0.027 per 1,000
gallons
$0.004 to $0.009 per 1,000
gallons
$0.004 per 1,000 gallons
SLIDE 60
-------
REPLACE MEDIA t ,
• Improves filtered water quality
• Most rapid sand filters suitable for.conversion
• Hydraulic and structural modifications may be needed
t Doubling plant capacity with no structural additions possible
SLIDE 63
-------
IMPROVING EXISTING DUAL- OR TRI-MIXED MEDIA FILTRATION
• Filter aid polymer prior to filtration
e Polymer application to backwash water
• Install individual filter turbidimeters
t Provide filter to waste
SLIDE 64
-------
PHOTOGRAPH OF TYPICAL GRAVITY FILTERS
SLIDE 65
-------
CUTAWAY VIEW OF TYPICAL MIXED-MEDIA FILTER
SLIDE 66
-------
MIXED-MEDIA FILTER BED
SLIDE 67
-------
ESTIMATED COSTS - PLANT MODIFICATION
Small Systems
Medium Systems
Large Systems
Total Cost
. OOP Gal
0.8-5
0.5-3
0.1-2
SLIDE 68
-------
NEW FILTRATION TECHNOLOGY
• Deep bed mono media
• Popular with ozone pretreatment
• Coarse media: 1.0 to 2.0 mm
SLIDE 69
-------
s
QJ
DC
00
d, O
JO
h* Uj
oS
o
g
H
DC
H
o
<
DC
LU
DC
£0
LU
01
O
O
DC
H
DC
<
O
-------
O
P:
i
2
CQ
CO
z
O
DC
O
LU
IBER MEMBRAt
LL
O
_J
O
X
CNI
O
z
X
cc
0
cc
— r
PARTICLES U
CO
Q
D
-i
O
X
LU
O
LU
CC
O
LU
CC
O
z
z '
O
<
0
^
O
O
UPPLIES
CO
H
> LOW TURBID
VJ
H
o
LU
t
-I
D OPERATION
LU
H
<
S
O
H
=>
<
2
PERATING SKI
O
|
Z™"
i
<*
LU
O
O
LU
X .
O
OCCASIONAL
IBRANES
CO LU
•"• u_
O
LU
OC
-------
Non-Corrosive Piping
System
Clean In Pla
System
Stainless
Steel Skid
Mounted
Frame
Chemical Resistant Membranes
.L ..
Praf lltration and Pretraatment System
When Required
Microprocessor
Controls for Radial
Pulse, CIP System,
and Membrane
Integrity Chock
Stainless Steel
Centrifugal Pumps
SKID-MOUNTED MEMBRANE FILTRATION ASSEMBLY
SLIDE 72
-------
FILTERED
RAW WATER
CLARIFIED
RECYCLE
DISCHARGE
HOLLOW FIBER
MEMBRANES
MEMBRANE
CARTRIDGE
MEMBRANE CLEANING
SOLUTION TO SEWER
AIR INLET FOR
BACKWASHING
BACKFLUSH
WASTEWATER
BACKFLUSH CLARIFIER
FLOW SHEET OF MEMBRANE FILTRATION SYSTEM
SLIDE 73
-------
o
I
o
DC
2
LE MEMBRANES
m
<
z
UJ
1
-J
O
CC
0
UJ
_J
CD
<
CO
0
QL
CO
O
UJ
PROPYLEN
MIC
-J DC
O LU
Q. 0
i i
0.2 MICRONS
CO
UJ
o
cc
a.
CO
UJ
Q
O
X
UJ
Y SUPPLIES
H
9
cc
H
0
o
o
UJ
1
_J
UJ
I
a.
CO
CALLY
^^^^
z
<
X
o
UJ
«E
0
•
OPERA!
_i
<
^
Z
<
!S
CD
I
UJ
Q
^^^
0
O
CO
o
o
CD
<
cc
UJ
Q.
O
-------
PROCESSES APPROPRIATE FOR SMALL SYSTEM
• Slow Sand Filtration
• Package Plants
• Diatomaceous Earth Filtration
SLIDE 75
-------
tr
LJJ
CL
O
<
CO
CM
O O
2 10
o T-
I ?
^ u>
^*
o
6
0
(0
O
<
cc
E
Q
(0
O
CO
I
i
a.
Q
LU
CC
O
UJ
QC
O
I
CC
-------
Ill
5
E
<
HI
.J
Q
Z
o
Ul
c
Ui
a
<
o
G
Z
g
Q
LU
tt
LU
O
u
SLIDE 77
-------
Ill
$
K
HI
.1
O
01
§
CB
CB
O
i
\
O
z
O
i-
«
-------
m
100 r
8.0 -
60
50
4.0
3.0
2.0
10
08
06
0.5
04
0.3
02
0 10
0.08
0.06
0.05
0.04
003
0.02
0.01
11 i i
i i i i i
99.9999.8 9998959080706050403020 10 5 21
PERCENT EXCEEDING SPECIFIED FILTERED WATER TURBIDITY
AVERAGE FILTERED WATER TURBIDITIES AT
SLOW-SAND FILTER PLANTS SURVEYED
SLIDE 79
-------
PHOTOGRAPH OF MECHANICAL SLOW SAND FILTER CLEANER
SLIDE 80
-------
PACKAGE WATER TREATMENT PLANTS
t Capacity range: 10 gpm to 40 mgd
• Factory built and compact
• 650-700 in service
• Simple to operate, designed for unattended operation
• Require basic operator skills
• Cost effective .
SLIDE 81
-------
UJ
CO
CO
Ul
5
<
111
OC
UJ
o
<
u.
O
UJ
CO
1
u.
csi
00
-------
PHOTOGRAPH OF 100 6PM PACKAGE PLANT INSTALLATION
SLIDE 83
-------
a
u
IL
C
o
u
UJ
c
u.
o
UJ
UJ
(0
o
CO
-------
sH
< -
SLIDE 85
-------
FLASH
MIX FLOCCULATION
FILTRATION
INFLUENT
4
s
A/
_L/..
1=
/
/
^ L
r
•*)->
<
<
-e
fa
WASH
SUPPLY
a— ^ EFFLUENT
DESLUDGE BACKWASH- BACKWASH
TO-WASTE SUPPLY
FLOW DIAGRAM OF ALTERNATIVE PACKAGE PLANT
(COURTESY OF WATER TECH, INC., VANCOUVER, WASHINGTON)
SLIDE 86
-------
ADSORPTION CLARIFIER
• Removal mechanism - Adsorption
• Settling not required
• Settleable floe not needed
SLIDE 87
-------
VIEW OF POLYETHYLENE ADSORPTION MEDIA
SLIDE 88
-------
ADSORPTION CLARIFIER
FLOCCULATION
• G = 110/second to 300/second
• Gt = 1 x 104 to 3 x 104
SLIDE 89
-------
RATES
Clarlfier: 10 gpm/sq ft
Filter: 5 gpm/sq ft
SLIDE 90
-------
APPLICATIONS
• At 10/5 gpm/sq ft: Long-term operation
• Turbidity: 100-150 NTU
• Color: 75-125 units
• Iron and Manganese: No limits established
• No lime softening
SLIDE 91
-------
C
fi
•o
.0
CM
-®-
O)
iM^N W
C0
O
UJ
O
UJ
0)
UJ
CD
CO
UJ
il
UJ
-------
ESTIMATED COSTS FILTRATION PROCESSES
SMALL SYSTEMS
Total Cost
4/1.000 gal
Package Plant 100-950
Direct Filtration 150-350
Diatomaceous Earth 50-650
Slow-Sand Filter 50-400
SLIDE 93
-------
Table of Contents
-------
-------
TABLE OF CONTENTS
Section Title Page
I FILTRATION TECHNOLOGIES 1-1
Introduction 1-1
Requirements for Filtration 1-1
Established Filtration Technologies 1-2
Conventional Treatment 1-2
Direct Filtration 1-2
Diatomaceous Earth Filtration 1-3
Slow-Sand Filtration 1-3
Package Plants 1-4
Other Applicable Filtration Technologies 1-4
Effectiveness of Filtration Process 1-4
Basis of Costs . 1-6
General 1-6
Capital Costs 1-7
Operation and Maintenance Costs 1-7
References 1-15
II TECHNIQUES FOR UPGRADING EXISTING WATER II-l
TREATMENT PLANTS
Introduction II-l
Upgrading Techniques H-2
Modifying Chemical Treatment II-2
Alum and Ferric Chloride H-2
Coagulation with Polymers II-3
Coagulant Aids II-4
Providing pH Control II-5
Chemical Feed Costs II-6
Conceptual Design II-6
Operation and Maintenance I1-6
Requirements
Liquid Alum Feed I1-7
Conceptual Design I1-7
Operation and Maintenance 11-7
Requirements
Polymer Feed 11-8
Conceptual Design II-8
Operation and Maintenance 11-8
Requirements
Sodium Hydroxide Feed I1-9
Conceptual Design I1-9
Operation and Maintenance H-9
Requirements
Sulfuric Acid Feed 11-10
Conceptual Design 11-10
Operation and Maintenance 11-10
Requi rements
Modifying or Adding Rapid Coagulant 11-10
Mixing
General 11-10
Changing Point of Coagulant Addition 11-11
-------
TABLE OF CONTENTS (Continued)
Section
II (Continued)
Title Page
Coagulant Diffusers 11-12 '
Mechanical Mixing 11-12
In-Line Blenders 11-13
Jet Injection Blending 11-13
Static Mixers 11-14
Cost of Rapid Mix 11-15
Conceptual Design 11-15
Operation and Maintenance 11-15
Requi rements
Adding Chemical Dosage Control Provisions 11-15
Techniques Available 11-15
Jar Tests 11-17
Techniques Based on Particle Charge 11-17
Zeta Potential 11-17
Streaming Current Detector 11-18
Continuous Filtration Techniques 11-18
Pilot Filters 11-18
Design for Automated Coagulant 11-20
Control
Turbidity Monitoring • 11-21
Particle Counting 11-21
Implementing Flocculation Improvements 11-22
Type of Flocculators 11-23
Flocculation Design Guidelines 11-23
Flocculation Alternatives 11-23
Improving Inlet and Outlet Conditions 11-25
Correcting Short-Circuiting 11-25
Installing High-Energy Flocculators 11-26
Case History 11-27
Cost of Adding Flocculation Basins 11-28
Improving Sedimentation Performance 11-23
General 11-28
Improving Inlet Conditions 11-29
Improving Outlet Conditions 11-29
Installing Sludge Collectors 11-29
Installing Tube Settlers 11-30
General Design Considerations 11-31
Application Requirements 11-31
Horizontal Flow Sedimentation 11-31
Basins
Upflow Clarifiers with Solids 11-32
Contact
Recommended Design Criteria 11-33
Upflow Clarifiers - Loading Rates 11-33
Horizontal Flow Basins - 11-33
Loading Rates
Location of Tube Modules Within 11-33
the Basin
Example Applications 11-34
Horizontal Basins 11-34
ii
-------
TABLE OF CONTENTS (Continued)
Section
II (Continued)
HI
it.
Title Page
Upflow Basins 11-35
Costs of Installing Tube Settlers 11-36
Modifying Filtration Facilities II-36
General 11-36
Filter Upgrading Evaluation 11-37
Regulatory Agency Design Standards 11-40
Filter Design Checklist 11-41
Case Histories 11-42
Sacramento, California 11-42
Erie County, New York 11-44
Corvallis, Oregon 11-46
North Marin County Water District. 11-47
Novato, California
References 11-93
NEW TREATMENT TECHNOLOGY III-l
Introduction III-l
Package Plants III-l
General III-l
Types Available II1-2
Conventional Package Plants II1-2
Tube Type Clarification Package Plants III-2
Adsorption Clarifier Package Plant II1-3
Application Criteria and Requirements III-5
Operational Considerations III-6
Cost of Package Conventional II1-7
Complete Treatment
Conceptual Design II1-7
Operation and Maintenance III-7
Requirements
Chemical Requirements II1-7
Conventional Package Plant Performance - II1-8
Case Histories
EPA Survey of Six Plants III-8
Adsorption Clarifier Package Plants - III-9
Case Histories
Greenfield, Iowa II1-9
Lewisburg, West Virginia 111-10
Philomath, Oregon • I11-10
Harrisburg, Pennsylvania 111-10
Red Lodge, Montana I11-11
Slow-Sand Filtration III-ll
Process Description III-ll
Design Standards I11-12
Operating Characteristics 111-13
Effluent Quality Requirements II1-13
Laboratory and Pilot Plant Studies 111-14
Costs of Slow-Sand Filters 111-15
Conceptual Design 111-15
Operation and Maintenance Requirements 111-16
iii
-------
TABLE OF CONTENTS (Continued)
Section
III (Continued)
Title Page
Case Histories 111-16
Survey of 27 Plants II1-16
New York State II1-17
Mclndoe Falls, Vermont 111-18
Village of 100 Mile, British 111-19
Columbia, Canada
Operating Results II1-20
Cost II1-21
Diatomaceous Earth Filtration 111-21
General 111-21
Design Criteria II1-23
Operating Requirements 111-24
Effluent Quality Requirements II1-24
Treatment Capability 111-24
Costs for Diatomaceous Earth Filtration 111-26
Membrane Filtration II1-26
Introduction 111-26
Membrane Cleaning II1-27
Typical Installation III-28
Performance Capabilities 111-29
Giardia Removal III-29
Coliform Removal II1-29
Turbidity Removal II1-30
Capital Costs 111-30
Application Concerns 111-30
Membrane Failure I11-30
Organics Removal II1-31
Pilot Testing 111-31
Cartridge Filters 111-31
Equipment Description 111-31
Cartridge Cleaning 111-32
Application Considerations 111-32
Removal Capabilities II1-33
Qualifying Filtration Processes 111-33
Need for Pilot Studies II1-33
Regulatory Agency Requirements 111-34
Conducting Pilot Studies II1-35
Flocculation and Sedimentation Studies III-36
Filtration Studies 111-37
Case Study - New Treatment Facilities II1-38
Pilot Study Selection Process
Background 111-38
Quality and Treatment 111-39
Pilot Study Program II1-40
Pilot Study Results 111-40
Recommended Design Criteria 111-41
iv
-------
TABLE OF CONTENTS (Continued)
Section
III (Continued)
APPENDIX
Title
Case Study - Untreated Supply for Small
Community .
Example
Evaluation Procedures
Cost Analysis
References
Chemical Costs
Design Standards for Filtration Plants -
Texas Department of Health
Paje
111-41
111-41
111-42
II1-43
III-84
Pilot Study Report
-------
LIST OF TABLES
Number
SECTION 1
Table 1-1
Table 1-2
Table 1-3
SECTION II
Table II-l
Table II-2
Table II-3
Table II-4
Table II-5
Table 1 1-6
Table 1 1-7
Table 11-8
Table II-9
Table 11-10
Table 11-11
Table 11-12
Table 11-13
Table 11-14
Title
Generalized Capability of Filtration Systems to
Accommodate Raw Water Quality Conditions
Removal Efficiencies of Viruses by Water Treatment
Processes
Removal Efficiencies of Giardia Lamblia by Water
Treatment Processes
Summary of Potential Problems in Water Treatment Unit
Operations
Estimated Costs for Surface Water Treatment by Adding
Alum Feed Facilities (10 mg/1)
Estimated Costs for Surface Water Treatment by Adding
Polymer Feed Facilities (0.3 mg/1)
Estimated Costs for Surface Water Treatment by Adding
Polymer Feed Facilities (0.5 mg/1)
Estimated Costs for Supplementing Surface Water
Treatment by Adding Sodium Hydroxide Facilities
Estimated Costs for Supplementing Surface Water
Treatment by Adding Sulfuric Acid Feed Facilities
Estimated Costs for Supplementing Surface Water
Treatment by Adding Rapid Mix
General Design Criteria for Flocculation Basins
Flocculator Design Guidelines
Estimated Costs for Supplementing Surface Water
Treatment by Adding Flocculation
Upflow Clarifier Loading Rates for Cold Water
Upflow Clarifier Loading Rates for Warm Water
Horizontal Flow Basins Loading Rates for Cold Water !
Horizontal Flow Basins Loading Rates for Warm Water
vi
Page
1-9
1-10
1-10
11-51
11-52
11-53
11-54
11-55
11-56
11-57
11-58
11-59
11-60
11-61
11-61
11-62
11-62
-------
List of Tables (Continued)
Number
SECTION II (Continued)
Title Page
Table 11-15
Table 11-16
SECTION III
Table III-l
Table II1-2
Table 111-3
Table 111-4
Table 111-5
Table 111-6
Table III-7
Table II1-8
Table III-9
Table 111-10
Table 111-11
Table 111-12
Table 111-13
Table 111-14
Table 111-15"
Table 111-16
Table 111-17
Table 111-18
Estimated Costs for Upgrading Surface Water Treatment 11-63
by Adding Tube Settling Modules
Turbidity Ranges During Demonstration Test 11-64
Summary of Results of Adsorption Clarification II1-45
Package Plants (Microfloc Products Group)
Estimated Costs for Surface Water Treatment by 111-46
Complete Treatment Package Plants
Water Treatment Facilities Surveyed in Field Study II1-47
Treatment Process Characteristics 111-48
Plant Turbidity Values (NTU) 111-49
Operating Data - Greenfield, IA II1-50
Operating Data - Lewisburg, WV 111-51
Operating Data - Philomath, OR 111-52
Operating Data - Harrisburg, PA 111-53
Operating Data - Red Lodge, MT II1-54
Estimated Costs for Surface Water Treatment by Slow- II1-55
Sand Filtration
Characteristics of Slow-Sand Filter Installations in 111-56
New York .
Filter Ripening Data - Summary II1-57
Estimated Costs for Surface Water Treatment by Direct II1-58
Filtration Using Diatomaceous Earth
Hollow Membrane Filtration Facility Design Information II1-59
Estimated Costs for Surface Water Treatment by Package II1-60
Ultrafiltration Plants
Clear Lake Water Quality Analysis 111-61
Clear Lake Water Quality at DWR Sampling Station II1-62
No. 1 at Lakeport
vii
-------
Number
SECTION I
Figure 1-1
Figure 1-2
Figure 1-3
Figure 1-4
SECTION II
Figure II-l
Figure 11-2
Figure II-3
Figure I1-4
Figure 11-5
Figure II-6
Figure 11-7
Figure I1-8
Figure II-9
Figure 11-10
Figure II-lla
and II-llb
Figure 11-12
Figure 11-13
Figure 11-14
Figure 11-15
Figure 11-16
LIST OF FIGURES
Title Page
Flow Sheet of a Typical Conventional Water Treatment 1-11
Plant
Flow Sheets for Typical Direct Filtration Plants 1-12
Typical Pressure Diatomaceous Earth Filtration System 1-13
Relationship Between Cyst Removal and Filtered Water 1-14
Turbidity (After Engeset)
Coagulant Diffuser
Diffuser for Coagulant Application
Jet Mixing Facility
Typical Flash Mixing Facility
Motionless (Static) Mixer (Courtesy of Komax)
Simplified Diagram of SCO Instrument
Typical Flocculation Units
A Perforated Baffle with Square-Nozzled Ports
Distributes Flow from the Flocculators at Rio De
Janeiro's Guandu Treatment Plant
New- and Old-Style Flocculators
Compartment Plan for One Flocculation Unit
Flocculator Outlet System and Settling Basin Inlet
System (Side and End Views)
Sectional View Through Sedimentation Basin Showing
Vacuum Sludge Removal System
Close-Up View of Vacuum Sludge Removal Assembly
A Module of Steeply Inclined Tubes (Courtesy of
Microfloc Products Group, Johnson Division, UOP)
Lamella® Separator (Courtesy of Parkson Corp.)
11-65
11-66
11-67
11-68
11-69
11-70
11-71,72,73
11-74
11-75
11-76
11-77
11-78
11-79
11-80
11-81
Typical Tube Settler Installation in Rectangular Basin 11-82
viii
-------
List of Figures (Continued)
Number
SECTION II (Continued)
Title
Figure 11-17
Figure 11-18
Figure 11-19
Figure 11-20
Figure 11-21
Figure 11-22
Figure 11-23
Figure 11-24
SECTION III
Figure III-l
Figure III-2
Figure III-3
Figure 111-4
.Figure 111-5
Figure 111-6
Figure III-7
Figure 111-8
Figure 111-9
Figure 111-10
Partial Tube Settler Installation in Circular
Clarifier
Solids Contact Clarifier with Tube Settlers
Package Solids Contact Tube Clarifier (Courtesy
ERC/Lancy)
Variable Declining Rate Filtration (J. AWWA)
Typical Rate of Filtration Patterns During a Filter
Run (Bauman and Oulman, 1970)
Hypothetical Optimum Filter Headioss/Turbidity
Breakthrough Curve
Flow Diagram of the Pilot Filtration Equipment
Clarifiers Showing Tube Settling Modules
Flow Diagram of a Package Plant
Flow Diagram of Alternative Package Plant (Courtesy
of Water Tech, Inc., Vancouver, Washington)
Operating Cycles of Package Plant (Courtesy of
Microfloc Products)
Pictorial View, Package Plant (Courtesy Neptune
Microfloc, Inc.)
Typical Covered Slow-Sand Filter Installation
Typical Uncovered Slow-Sand Filter Installation
Average Raw Water Turbidities at Slow-Sand Filter
Plants Surveyed
Page
11-83,84
11-85,86
H-87
11-88
11-89
11-90
11-91
11-92
111-63
111-64
111-65
111-66
111-67
II1-68
111-69
Average Filtered Water Turbidities at Slow-Sand Filter 111-70
Plants Surveyed
Average Coliforms in Raw Waters at Slow-Sand Filter 111-71
Plants Surveyed
Average Coliforms in Filtered Waters at Slow-Sand II1-72
Filter Plants Surveyed
ix
-------
List of Figures (Continued)
Number
Title
Section III (Continued)
Figure III-ll
Figure 111-12
Figure 111-13
Figure 111-14
Figure 111-15
Figure 111-16
Figure 111-17
Figure 111-18
Figure 111-19
Figure IH-20
Figure 111-21
Giardia Cysts in the Raw Water (Village of
100 Mile House)
Average Raw and Filtered Water Turbidity
(Village of 100 Mile House)
Typical Pressure Diatomaceous Earth Filtration
System
Hollow Fiber Membrane Operational Description
Skid-Mounted Membrane Filtration Assembly
Flow Sheet of Membrane Filtration System
Tube Settling Test Module
Pilot Filter Assembly
Pilot Filter Schematic
Typical Pilot Filtration Data Log
Flow Schematic 20 gpm Pilot Plant
Page
111-73
111-74
111-75
111-76
111-77
111-78
111-79
II1-80
111-81
111-82
111-83
-------
SECTION I
FILTRATION TECHNOLOGIES
INTRODUCTION
The 1986 Amendments to the Safe Drinking Water Act (SWDA) require that within 18
months of enactment, the EPA must promulgate regulations specifying criteria
under which filtration (including coagulation and sedimentation as appropriate)
is required for surface water sources, and disinfection is required for public
water systems using surface water sources. The EPA must consider the quality of
the source water, protection affected by watershed management, treatment prac-
tices {such as disinfection and length of water storage), and other factors rele-
vant to health. Specific procedures are required to be formulated by the EPA by
which states will determine which water systems shall adopt filtration. The fil-
tration and disinfection requirements are being proposed as treatment techniques
regulations to protect against the potential adverse health effects of exposure
to Giardia Iambiia, viruses, Legionella, and heterotrophic bacteria, as well as
many other pathogenic organisms that may be present.and are removed under these
conditions.
Requirements for Filtration
EPA is in the process of preparing a guidance manual for compliance with the
filtration and disinfection requirements for public.water systems using surface
sources. This document will ensure the proper implementation of the Surface Water
Treatment Rule (SWTR). The SWTR pertains to all public water systems that utilize
a surface water as their source. The SWTR defines a surface water as all waters
which are open to the atmosphere (e.g., rivers, , lakes, streams, reservoirs,
impoundments) and any subsurface sources such as springs, infiltration galleries,
wells or other collectors that are at risk of being contaminated by a surface
water.
1-1
-------
Surface water supplies that are often used as sources of drinking water include
two major classifications, namely running and quiescent waters. Streams, rivers,
and brooks are subdivisions of running water; while lakes, reservoirs, impound-
ments and ponds are subdivisions of quiescent waters. These sources are subject
to the requirements of the SWTR.
ESTABLISHED FILTRATION TECHNOLOGIES
The following methods of filtration are identified as the Most Applicable Tech-
nologies or established filtration technologies and are those most widely used
for removal of turbidity and microbial contaminants:
• Conventional treatment
• Direct filtration {gravity and pressure filters)
t Diatomaceous earth filtration
• Slow-sand filtration
• Package plants
Conventional Treatment
Conventional treatment is the most widely used technology for removing turbidity
and microbial contaminants from surface water supplies. Conventional treatment
includes the pretreatment steps of chemical coagulation, rapid mixing, floccula-
tion and sedimentation followed by filtration. A typical flow schematic for a
conventional treatment plant is shown on Figure 1-1. Section II describes the
application of existing as well as recently developed techniques for the upgrad-
ing of filtration facilities.
Direct Filtration
The direct filtration process can consist of any one of several different process
trains depending upon the application. In its most simple form, the process
includes only filters (oftentimes pressure units) preceded by chemical coagulant
application and mixing. Raw water must be of seasonally uniform quality with
turbidities routinely less than 5 NTU in order to be effectively filtered by an
in-line direct filtration system. A second common, configuration of the process
1-2
-------
Includes flocculation as pretreatment for the filters, in addition to coagulant
application and mixing. Typical flow schematics for direct filtration plants are
shown on Figure 1-2. Direct filtration is an established filtration technology
and will be more widely used on water supplies that up until now received only
chlorination because raw water quality permitted elimination of traditional sedi-
mentation processes prior to filtration.
Diatomaceous Earth Filtration
Diatomaceous earth (DE) filtration, also known as precoat or diatomite filtra-
tion, is applicable to direct treatment of surface waters for removal of rela-
tively low levels of turbidity. Diatomite filters consist of a layer of DE about
1/8-inch thick supported on a septum or filter element. A typical flow schematic
of a diatomaceous earth filtration process is shown on Figure 1-3. The problems
inherent in maintaining a perfect film of DE between filtered and unfiltered
water have restricted the use of diatomite filters for municipal purposes, except
under favorable conditions. Diatomaceous earth filtration is classified as "new
technology" and discussed in Section III because it has received renewed interest
in light of the need to filter surface waters of low turbidity previously only
chlorinated. Since it can remove Giardia cysts it is a viable filtration
technology.
Slow-Sand Filtration
Slow-sand filters are similar to single-media rapid-rate filters in some
respects, yet they differ in a number of important characteristics. In addition
to (1) slower flow rates (by a factor of 50 to 100 versus direct filtration for
example), slow-sand filters also: (2) function using biological mechanisms
instead of physical-chemical mechanisms, (3) have-smaller pores between sand
particles, (4) do not require backwashing, (5) have longer run times between
cleaning, and (6) require a ripening period at the beginning of each run.
Although applicable to medium to large size plants, they are most commonly used
for small communities because of the operational simplicity. Because of the
revived interest in this filtration method it is referred to in this document as
"new technology."
1-3
-------
Package Plants
Package plants are not a separate technology in principle from the preceding
technologies. They are, however, different enough in design criteria, operation,
and maintenance requirements that they are discussed separately in Section III
and are regarded as new technology.
The package plant is designed as a factory-assembled, skid-mounted unit generally
incorporating a single, or at the most, several tanks. A complete treatment pro-
cess typically consists of chemical coagulation, flocculation, settling and fil-
tration. Package plants, for purposes of this document, generally can be applied
to flows ranging from about 25,000 gpd to approximately 6 mgd.
OTHER APPLICABLE FILTRATION TECHNOLOGIES
Technologies classified as Other Applicable Technologies which may have applica-
tion for removal of turbidity and microbial contaminants are listed as follows:
• Membrane (ultrafiltration)
• Cartridge
These technologies have not been widely used for filtration of drinking water
supplies and could therein be classified as truly emerging technologies. Along
with package plants, slow sand filters, and diatomaceous earth filtration, these
filtration technologies are best suited to small systems and are collectively
classified as new technology and are fully discussed in Section III.
EFFECTIVENESS OF FILTRATION PROCESSES
The application of these technologies will depend on certain raw water condi-
tions, as shown in Table 1-1.
Filtration processes provide various levels of microbial contaminant removal.
Tables 1-2 and 1-3 summarize microbial removal efficiencies determined from field
and pilot plant studies completed on a range of filtration processes.
1-4
-------
Table 1-2 includes virus removal results by several filtration processes without
disinfection.1 As shown in the table, all of the processes are capable of remov-
ing 99 percent of viruses without disinfection.
Giardia Iambi i a removal data by conventional treatment, direct filtration, dia-
tomaceous earth filtration, and slow-sand filtration are shown in Table 1-3. 2
Very high levels (>99.9%) of Giardia reduction can be achieved by chemical coagu-
lation followed by settling and filtration, or by direct filtration. The impor-
tance of coagulation to achieve high levels of Giardia removal is noted for both
processes. Diatomaceous earth filtration is also extremely effective in removing
Giardia cysts. Slow-sand filtration which relies on biological as well as physi-
cal mechanisms to remove microbial contaminants is especially effective in remov-
ing GUjrdiA cysts.
In his review of performance data, Logsdon compared slow-sand filtration, diato-
maceous earth filtration, and conventional and direct filtration.1 Using informa-
tion from filtration studies at pilot-scale, full-scale, or both, he showed that
all of the filtration processes, when properly designed and operated, can reduce
the concentration of Gi a rd i a cysts by 99 percent or more, if they are treating a
source water of suitable quality. Many of the studies also contained Giardia
removals of 99.9 percent, agreeing with the values shown in Table 1-3.
Effective particulate removal has been found to be a critical factor in producing
a safe water where the water is at risk for or contaminated with Giardia. Pilot
studies by Engeset et. al. using low turbidity raw water found the relationship
between cyst removal and filtered water turbidity shown in Figure 1-4. 1 Achieving
a 99 percent reduction in cysts required a finished water of 0.1 NTU or less.
Studies at Colorado State using low turbidity raw water found effective cyst
removal to be associated with a 70 percent removal of turbidity.
Filtration processes provide various levels of turbidity and microbial contami-
nant removal. However, when properly designed and operated and when treating
source waters of suitable quality, the above filtration processes are capable of
achieving at least a 2 log (99 percent) removal of Giardia cysts and a 1 log (90
percent) removal of enteric viruses without disinfection. A summary of the
removal capabilities of the various filtration processes is presented in Table
1-3. As indicated, conventional treatment without disinfection is capable of
1-5
-------
achieving up to a 3 log reduction in Giardia cysts and a 2 log removal of enteric
viruses. Direct filtration can achieve up to a 2 log reduction of Giardia cysts
and a 3 log reduction of enteric viruses. Achieving the maximum removal efficien-
cies of those constituents with these treatment processes requires the raw water
be properly coagulated and filtered.
The technologies available to a community searching for the most economical and
effective means to comply with the surface water treatment rule include modifica-
tion of existing treatment plants and installation of new plants. Both filtration
and disinfection processes are capable of achieving high removals of microbial
contaminants. Some examples of their removal efficiences based on data from a
number of; plants are shown below:
Percent Removal
Parameter
Coliforms
Turbidity
Enterovirus
Giardia
*After disinfection.
It is important to note that virus removal (based on mean virus density) in four
of the treatment plants surveyed exceeded 99 percent through prechlorination and
clarification processes only. Very high levels (>99.9 percent) of Giardia reduc-
tion can be achieved by chemical coagulation followed by settling and filtration,
or by direct filtration. The importance of coagulation to achieve high levels of
Giardia removal is noted for both processes.
BASIS OF COSTS
General
Capital and operating costs provided in the following sections for technologies
in this document are based upon updated costs originally presented in several
1-6
Clarified
90-99
50-70
70-99
40-75
Filtered
98-99.99
90-99
90-99.9
96-99.9
Finished*
>99.99
90-99
98-99.99
96-99.99
-------
cost documents prepared for EPA.3!1* Specific details regarding how costs were
calculated for individual.processes and process groups are presented in following
sections. •
Capital Costs
The construction cost data presented in this document are based on late 1986
costs. Methods of calculating construction costs are described in the two EPA
reports mentioned above.3,1* Construction costs are converted to capital costs
by use of typical percentages of construction costs for such factors as contin-
gencies (15%), contractor's overhead and profit (1256), sitework (15%), subsurface
considerations (5%), engineering and technical fees (15%), and interest during
construction (10%). Costs for acquiring new land for treatment sites, however,
are not included in project costs since these costs are extremely site specific.
To estimate annual costs and costs per 1,000 gallons, capital costs were amor-
tized over 20 years at a 10 percent interest rate. • .
Operation and Maintenance Costs
Operation and maintenance requirements are developed for energy, maintenance
material, and labor. The energy category includes process electrical energy,
building electrical energy, and diesel fuel. The operation and maintenance
requirements were determined from operating data at existing plants to the extent
possible, or from assumptions based on the experience of both the author and the
equipment manufacturers.
Electrical energy requirements were developed for both process energy and
building-related energy, and they are presented in terms of kilowatt-hours (kWh)
per year.
Maintenance material costs include the cost of periodic replacement of component
parts necessary to keep the process operable and functioning. Chemical costs are
added separately, and units costs of chemicals used in this document represent
late 1986 conditions. Chemical costs are shown in Table A-l of the Appendix.
Labor requirements include both operation and maintenance labor.
1-7
-------
Total operation and maintenance cost is a composite of the energy, maintenance
material, and labor costs. To determine annual energy costs, unit costs of
$0.068/kWh of electricity and $0.80/gallon of diesel fuel are used. Thei labor
requirements are converted to an annual cost using hourly labor rates of $5.90/
hour for small water systems (less than 150,000 gpd design capacity) and $14.30/
hour for larger systems, based on labor costs found at existing plants during
conduct of this project.
1-8
-------
TABLE 1-1. GENERALIZED CAPABILITY OF FILTRATION SYSTEMS TO
ACCOMMODATE RAW WATER QUALITY CONDITIONS
Treatment Technology
Conventional Treatment
General Constraints
(Indicated values could
occasionally be exceeded)
Total
Conforms Turbidity Color
(I/1QO ml) (NTU) (CU)
<20,000 no 75
restrictions
(with no predisInfection)
<5,000
Direct Filtration
<500
<7-14
<40
Slow Sand Filtration
<800
<5
01atomaceous Earth Filtration
-------
TABLE 1-2. REMOVAL EFFICIENCIES OF VIRUSES BY
WATER TREATMENT PROCESSES
Unit Process Percent Removal
Slow sand filtration 99.9999
99.8
99.8
91
Olatomaceous earth >99.9S
filtration *
Direct filtration 90-99
Conventional treatment 99
*No viruses recovered.
TABLE 1-3. REMOVAL EFFICIENCIES OF GIARDIA LAMBLIA BY
WATER TREATMENT PROCESSES
Unit Process Percent Removal
Rapid filtration with 96.6-99.9
coagulation, sedimentation
Direct filtration with 95.9-99.9
coagulation
- Mo coagulation 48
- With flocculatlon 95-99
- No coagulation 10-70
Olatomaceous earth 99-99.99
filtration
>99.9
Slow sand filtration 100
1-10
-------
1
CM
i °
ax"*
» — N.
«!!
cc o
O >- »
s«i*
2 w <
£Q_,<
ff a < o
I- < D UJ
-I CC O 2
I *
z^
£?
5
UJ
o
O
0
to
*
6
CM
W
I-
z
I
o
o
X
I
O o
2 CMH
uj
m
H
Z
h-
UJ
i
i-
UJ
o
u
J-
<
O
a;
D
oo
UJ
3
U.
Z
1-11
-------
INFLUENT
-COAGULANTS
RAPID MIX
30 sec • 2 min
DETENTION
DUAL OR MIXED
MEDIA FILTER
4-5 gpm/ft2
INFLUENT
-COAGULANTS
RAPID MIX
30 sac - 2 mln
DETENTION
1-HR
CONTACT BASIN
DUAL OR MIXED
MEDIA FILTER
4*5 gpm/ft2
' NONIONIC POLYMER
0.05-0.5 mg/l OR
ACTIVATED SILICA
i-COAGULANTS
INFLUENT
1
p
RAPID MIX
30 sac • 2 mln
DETENTION
FLOCCULATtON
15-30 mln
DUAL OR MIXED
MEDIA FILTER
4-5 gpm/ft2
Figure 1-2. FLOW SHEETS FOR TYPICAL DIRECT FILTRATION PLANTS
1-12
-------
-V'
Filtrate
Clear liquid line
q
Precoat
tank
Raw -j _Q_
(
\
>
q
Body
feed
tank
(
i
(
S '
)
i
<
f ™
)
Backwash
line
Filter
Body feed
pump
x— D_-
f~-
f
JEL,
Septum
Kg
•v«
i
"V.
••••
^
MM
V Filter -X
cake
> (
J
»
€
(
)
>
Precoat
drain
line
source '"t^/Klter" Backwash^/
feed drain line
pump
Figure 1-3. TYPICAL PRESSURE DIATOMACEOUS
EARTH FILTRATION SYSTEM
*-To drain
1-13
-------
1.0
z
I
H
Q
m
DC
3
III
0.5
0.1
u. 0.05
ui
0.01
I ....... I t
40 50 60 70 80 90
CYST REMOVAL - %
100
Figure I-4. RELATIONSHIP BETWEEN CYST REMOVAL AND FILTERED
WATER TURBIDITY (AFTER ENGESET)
1-14
-------
REFERENCES
1. Logsdon, G.S., "Comparison of Some Filtration Processes Appropriate for
Giardia Cyst Removal," presented at the Calgary Giardla Conference, Calgary,
Alberta, Canada, February 23-25, 1987.
2. Lippy, E.C., and S.C. Waltrip, "Waterborne Disease Outbreaks—1946-1980: A
Thirty-Five Year Perspective," J.AWWA, p. 60, February 1984.
3. Culp/Wesner/Culp, "Estimation of Small System Water Treatment Costs," EPA
Contract No. 68-03-3093, U.S. Environmental Protection Agency, Cincinnati,
OH. In publication.
4. Culp/Wesner/Culp, "Estimating Water Treatment Costs, Volume 2: Cost Curves
Applicable to 1 to 200 mgd Treatment Plants," EPA-600/2-79-162b, U.S.
Environmental Protection Agency, Cincinnati, OH, August 1979.
1-15
-------
-------
Section II
Techniques for Upgrading
Existing Water Treatment Plants
-------
-------
SECTION II
TECHNIQUES FOR UPGRADING EXISTING WATER TREATMENT PLANTS
INTRODUCTION
Experience has shown that many existing water treatment plants can be upgraded to
provide significantly increased output by application of newly developed treat-
ment techniques. In many cases these techniques can be applied to improve perfor-
mance while also increasing capacity. If the existing structures are sound, and
if traditional design parameters were applied to the original installation, plant
output often can be increased by as much as 100 to 200 percent without major
plant structural additions, and for substantially less than costs associated with
construction of new facilities.
When an existing water treatment plant does not meet required standards it is
necessary to upgrade the plant performance or provide additional treatment. In
general, upgrading water treatment plants is carried out for one or more of the
following reasons:
0 Water quality improvement
0 Plant capacity increase
0 Reliability improvement
0 Maintenance reduction
0 Cost reduction
The extent of remediation could range from adjustment of a chemical dosage or
loading rate, addition of new chemical feed systems, and baffling to improve
flocculation or sedimentation basin hydraulics, to major structural changes. The
principal objective of these improvements is that of improved finished water
quality. Brief guidance and instructions for evaluating and upgrading the perfor-!
mance of existing treatment plants is furnished in this section. ;
II-l
-------
Treatment facility evaluation should be carefully planned in advance. As it
proceeds, the plant evaluation should be carried out in an organized, logical
fashion that results in consideration of all necessary and important aspects of
the plant and its operation.
Table II-l provides a general overview of some of the problems in various treat-
ment processes that must be corrected in order for the facility to produce accep-
table water quality at manageable cost. Some are of major importance and will be
discussed in the following paragraphs. Others are of lesser importance and relate
primarily to plant operating reliability and maintenance concerns.
UPGRADING TECHNIQUES
The following techniques for upgrading water treatment plants will be described
and discussed in this section:
* Modifying chemical treatment
• Modifying or adding rapid coagulant mixing
• Adding chemical dosage control provisions
t Implementing flocculation improvements
* Improving sedimentation performance
t Modifying filtration facilities
i
Modifying Chemical Treatment
Alum and Ferric Chloride--
Primary coagulants such as aluminum sulfate (alum), ferric chloride, and other
metallic salts are widely used in surface water treatment plants. The specific
choice of coagulant is often determined by cost and availability as well as
effectiveness. Alum, because it is widely available, effective at low dosages,
and easy to store and apply, is used in most treatment facilities. However, „
health concerns related to the aluminum ion in filtered water has caused utili-
ties to consider alternative metallic salt primary coagulants, or in some cases, ^
organic coagulants. Use of ferric chloride, for example, would eliminate the
potential problem of excessive aluminum ions in the filtered water. Likewise,
II-2
-------
organic coagulants of either the cationic or anionic type do not present this
problem.
/
Occasionally, changing from alum to ferric chloride as the primary coagulant may
be appropriate to achieve either improved plant performance or better finished
water quality. For waters containing high concentrations of dissolved color and
of low alkalinity and pH, ferric chloride could prove to provide more effective
removal of color at lower dosage and less cost. Ferric hydroxide floe is heavier
than alum floe, and where raw water contains little turbidity but is highly
colored, improved sedimentation basin performance could be achieved by changing
from alum to ferric chloride as a primary coagulant. However, where trihalo-
methane formation potential (THMFP) of a particular water supply is of concern,
several investigators have found alum to be more effective than ferric chloride
in reducing THMFP and total organic carbon (TOC) on various waters of both acid
and alkaline characteristics.1
To establish which metallic coagulant is best suited for a particular water
supply, use of standard jar testing procedures as a means of determining optimum
dosages in terms of turbidity as well as organic removal is the preferred proce-
dure. Jar test results can then be used to establish appropriate costs for alter-
native metallic coagulants. In general, both metallic coagulants perform quite
similarly on a particular water supply, such factors as cost; safety; ease in
storing, handling and application; and availability influence selection. Alum is
generally preferred because it is superior to ferric chloride in the above major
categories of concern.
Coagulation with Polymers— <•
Synthetic organic polymers have been shown to be effective coagulants or coagu-
lant aids. Polymers are long-chain molecules comprised of many subunits called
monomers. A polymer is called a polyelectrolyte if its monomers consist of ioniz- 't
able groups. Polyelectrolytes having a positive charge upon ionization are1
referred to as cationic polymers. Negatively charged polyelectrolytes are termed
anionic polymers. Finally, polymers that do not contain ionizable groups are
called nonionic polymers.
II-3
-------
Catiom'c polymers can be effective in coagulating negatively charged clay parti-
cles.? Cationic polymers do not require a large molecular weight to be effective
in destabilization.
Anionic particles generally are ineffective coagulants for negatively charged
clay particles. When anionic polymers are used in conjunction with another coagu-
lant such as alum, their coagulating effectiveness is increased.
Dosages of only 1.5 to 10 mg/1 of cationic polymer frequently are sufficient to
achieve coagulation. In contrast, 5 to 150 mg/1 of alum is often needed to obtain
similar results. Other important differences between the use of polymers and
metallic salts are sludge quantities and dosage control. The use of alum or
ferric chloride can result in copious volumes of sludge that must be handled,
whereas the additional sludge quantity is negligible when a polymer is used.
Replacement of alum by either cationic polymers or mixtures of cationic polymers
and alum or sodium aluminate in prepared concentrated solutions is becoming
increasingly popular because these materials can achieve effective coagulation,
but produce less sludge. They are particularly attractive to treatment of rela-
tively low turbidity (less than 5 NTU) surface waters where formation of a
settleable floe is not a prerequisite to effective removal by filtration. Rather,
the relatively small floe settles poorly, but is readily filterable.
Of great benefit is the reduced sludge volume produced through use of the organic
or organic/metallic coagulants. However, because these coagulants are highly
selective, extensive jar testing as well as frequent changes in coagulant types
(anionic or cationic) may be needed as raw water conditions change seasonally.
Also, some waters of high organic content characterized by color imparted by
humic and fulvic acids are not effectively treated because the coagulants do not
impart a pH depression in addition to charge neutralization which is a require-
ment to precipitate organic color.
Coagulant Aids—
The performance of many water treatment plants can be improved by the addition of
a coagulant aid (generally nonionic) to overcome some of the deficiencies of a
11-4
-------
poorly designed flocculation basin. Flocculation is a time-dependent process that
T ' - • ' ' ,• *
can be enhanced and accelerated by the addition of organic coagulant aids. The
resulting floe generally settles more rapidly than unaided alum floe, permitting
the treatment plant to perform more efficiently at higher plant throughput* It
should be noted, however, that application of coagulant aids is no substitute for
adequate upstream mixing of coagulant with incoming raw water.
A coagulant aid dosage is influenced by numerous operational parameters and raw
water characteristics. In general, organic polymer coagulant aids are applied at
an alum-to-polymer ratio ranging from 100:1 to 50:1. Plant jar tests must be run,
however, to establish the proper dosage ratio for optimum results. For a rela-
tively modest investment in capital costs, the application of a coagulant aid can
provide marked improvement in the flocculation and sedimentation process in an
existing plant. The improved sedimentation will reduce turbidity loadings to the
filters, extending filter operation cycles, and in some instances, can provide a
higher-clarity filtered effluent.
Providing pH Control-
Adding artificial alkalinity in the form of soda ash, lime, sodium hydroxide or
sodium bicarbonate may be required in certain instances where conditions for
optimum coagulation cannot be provided. Addition of alum to water forms sulfuric
acid which can depress the pH excessively if the buffering capacity of the water
supply is insufficient. At pH values below optimum, less than complete formation
of aluminum hydroxide, occurs resulting in incomplete coagulation and turbid
filtered water. Additionally, postflocculation occurs in filtered water leading
to high residual filtered water aluminum ion concentration.
In these situations, feeding pH control materials will ensure complete precipita-
tion of the coagulant, and enhance turbidity removal by sedimentation and filtra-
tion processes. Floe formed in proper pH ranges is more dense, settles more
rapidly, and has greater strength to resist shearing forces in the filters.
Occasionally, where raw water pH is artifically high due to the action of algae
or other aquatic plants, the pH may need to be depressed by adding acid to bring
the pH into the effective range for optimum coagulation. In this situation, acid
either in the form of sulfuric or hydrochloric can be added to depress the pH.
II-5
-------
Chemical Feed Costs— '
Capital and operating costs are discussed for addition of coagulants or pH
control chemicals of the following types:
• Basic chemical feed
• Liquid alum feed
t Polymer feed
• Sodium hydroxide feed
• Sulfuric acid feed
The costs reported are derived from those presented in "Technologies and Costs
for the Removal of Microbial Contaminants from Potable Water Supplies," prepared
in February 1987, Contract No. 68-01-6989, under contract to the Science and
Technology Branch, Office of Drinking Water, U.S. Environmental Protection
Agency. Costs are presented for 12 categories of treatment plant capacity ranging
from 0.026 mgd to 1,275 mgd.
Conceptual Design—The costs were derived for the most common type of chemical
feed system consisting of a mixing tank, mixer, and metering pump. Chemical solu-
tions of known concentrations are prepared on a batch basis by manually adding a
chemical and water to the dissolving tank. This system can be used to feed either
liquid or dry chemicals. For dry chemicals, the mixing tank is used for dissolv-
ing and mixing the chemicals with water and for storage of the chemical solution.
For liquid chemicals, the mixing tank {day tank) is used for initial mixing and
solution storage. The costs assume bulk storage of the delivered liquid chemical.
A metering pump is used to accurately feed a solution of known concentration at a
set rate.
Operation and Maintenance Requirements—Operation labor requirements for the
basic chemical feed system include labor for emptying bags of dry chemicals, or
containers of liquid chemicals, into the dissolving tank for turning the mixer on
and off, for calibrating the chemical metering pump, and for occasional preven-
tive maintenance of the mixer and metering pump.
II-6
-------
Electrical requirements are for the mixer and the metering pump. Mixer operation
ranges from 30 minutes/day for the smallest system up to 55 minutes/day for the
largest system. It is assumed that the chemical metering pumps operate continu-
ously 24 hours/day. Maintenance material costs are primarily for the mixer and
metering pump, but maintenance material costs for a hydrated lime feed system are
increased due to the continuous operation of the mixer. Chemical costs used in
determining operation and maintenance costs are included in Table A-l in the
Appendix.
Liquid Alum Feed-
Liquid alum feed systems are assumed to be custom-designed and contractor-
constructed facilities for treatment systems in Categories 5 through 12. For
smaller systems, the basic chemical feed system described above is satisfactory.
Costs for these systems are presented in Table 11-2.
Conceptual Design—The design of these systems is based upon the use of liquid
alum, which has a weight of 10 pounds/gallon. Fifteen days of storage are pro-
vided, using fiberglass reinforced polyester (FRP) tanks. The FRP tanks are
assumed to be uncovered and located indoors for smaller installations, and out-
doors for larger installations. Outdoor tanks are covered and vented, with insu-
lation and heating provided to prevent crystallization, which occurs at tempera-
tures below 30°F.
Operation and Maintenance Requirements—Electrical requirements are for feeder
pump operation, building lighting, ventilation, heating, and heating needs of
outdoor storage tanks. Maintenance materials include repair parts for pumps,
motors, valves, and electrical starters and controls.
Labor requirements consist of time for chemical unloading and routine operation
and maintenance of feeding equipment. Liquid alum unloading requirements are
calculated on the basis of 1.5 hours/bulk truck delivery. Time for routine
inspection and adjustment of feeders is 15 minutes/metering pump/shift. Mainte-
nance requirements are assumed to be 8 hours/year for liquid metering pumps.
II-7
-------
Polymer Feed-
Polymers are used as coagulants or filter aids. Polymer addition in treatment
plants in Categories 1 through 4 is assumed to use a basic polymer feed system
that is similar to the basic chemical feed system previously described. Plants in
Categories 5 through 12 use a polymer feed system like that described below.
Costs for these systems are presented in Tables 11-3 and 11-4.
Conceptual Design—The conceptual design of all polymer feed systems is based
on preparation of a 0.25 percent stock solution.
The basic polymer feed system for Categories 1 through 4 consists of two tanks (a
mixing tank and a storage tank), a mixer, and a metering pump. The mixing tank is
elevated on a stand to just above the top water level in the storage tank to
allow gravity flow of mixed polymer solution into the storage tank. The storage
tank is the same size as the mixing tank, since the transfer of aged polymer
solution to the storage tank should only be required once daily.
The larger polymer feed system for Categories 5 through 12 consists of a dry
chemical storage hopper, a dry chemical feeder, two tanks, and a metering pump.
Operation and Maintenance Requirements—The operation labor requirements for
the basic polymer feed system are primarily for daily measurement of dry polymer,
adding it to the mixing tank, operating the mixer, and transferring the polymer
solution to the storage tank. Additional labor is required for periodically
checking the metering pump, preventive maintenance, and occasional repairs on the
mixer and pump.
Operation and maintenance costs for larger polymer feed systems (Categories 5
through 12} are based on 3 percent of the manufactured equipment and pipe and
valve costs. These costs do not include the cost of polymer. Labor requirements
are for bag unloading (1 hour/ton of bags), the dry chemical feeder (110 hours/
year for routine operation and 24 hours/year for maintenance), and the solution
metering pump (55 hours/year for routine operation and 8 hours/year for
maintenance).
11-8
-------
Sodium Hydroxide Feed-
Sodium hydroxide is added for pH adjustment. Sodium hydroxide is in a dry form
for plants in Categories 1 through 4 (up to 200 pounds/day). A basic chemical
feed system can be used for this purpose. Plants in Categories 5 through 12 use
liquid sodium hydroxide. Costs for these systems are presented in Table II-5.
Conceptual Design—Dry sodium hydroxide (98.9 percent pure) is delivered to the
plant site in drums and then mixed to a 10 percent solution on-site. A volumetric
feeder is utilized to feed sod'ium hydroxide to the mixing tank. Two day-tanks are
necessary: one for mixing a 10 percent solution, and one for feeding. The use of
two tanks is necessary because of the slow rate of sodium hydroxide addition
caused by the high heat of the solution. Each tank is equipped with a mixer, and
a dual-head metering pump is used to convey the 10 percent solution to the point
of application.
For feed rates greater than 200 pounds/day, a 50 percent sodium hydroxide solu-
tion is purchased premixed and delivered by bulk transport. The 50 percent solu-
tion contains 6.38 pounds of sodium hydroxide/gal Ion. For the 50 percent solu-
tion, 15 days of storage are provided in FRP tanks. Dual-head metering pumps are
used to convey solution to the point of application, and a standby metering pump
is provided in each case.
Pipe and valving is required for water conveyance to the dry sodium hydroxide
mixing tanks and between the metering pumps and the point of application. The
storage tanks are located indoors, since 50 percent sodium hydroxide begins to
crystallize at temperatures less than 54°F.
Operation and Maintenance Requirements—Process energy requirements are for the
volumetric feeder and mixer (smaller installations only) and the metering pump. A
maintenance material requirement of 3 percent of equipment cost, excluding the
storage tank cost, is utilized.
Labor requirements are based on unloading time for dry sodium hydroxide in drums,
or the liquid 50 percent sodium hydroxide purchased in bulk for the larger
11-9
-------
installations. For installations using dry sodium hydroxide, additional labor
required for routine operation time for the volumetric feeder is about 10
minutes/day feeder. In addition, for each installation, operation time for the
dual-head metering pump is 15 minutes/day, with an annual maintenance time of 8
hours.
-»*
Sulfuric Add Feed-
Sodium hydroxide -and lime are used to raise pH values. Sulfuric.acid is used to
lower pH. The feed system discussed in this report is capable of metering
concentrated (93 percent) sulfuric acid from a storage tank directly to the point
of application. Costs for these systems are presented in Table II-6.
Conceptual Design—For sulfuric acid feed rates up to 200 gpd, the concentrated
acid is delivered to the plant site in drums, and at larger flow rates it is
delivered in bulk. Acid purchased in bulk is stored outdoors in FRP tanks, and
acid purchased in drums is stored indoors. Fifteen days of sulfuric acid storage
is provided, and a standby metering pump is included for all installations.
Operation and Maintenance Requirements—Process electrical energy requirements
are for the metering pump. Building energy requirements are for indoor storage of
the sulfuri.c acid drums. Maintenance material requirements were estimated at 3
percent of the equipment cost, excluding the cost of storage tanks. Labor
requirements are for chemical unloading and for the metering pumps. Unloading
times of 0.25 hour/drum of.acid and 1.5 hours/bulk truck delivery were utilized.
Metering pump routine operation is 15 minutes/pump/day, and maintenance require-
ments are 8 hours/feeder/year.
Modifying or Adding Rapid Coagulant Mixing
General—
The rapid mixing process usually requires improvement if it is to perform ade-
quately at higher plant throughputs. Effective coagulation requires instantaneous
mixing of the coagulant with the incoming water. If the coagulant does not come ,
in contact with all colloidal turbidity particles, the uncoagulated turbidity
will pass through the filters. Proper design of the rapid mixer can result in Tow
coagulant doses and improved aggregation during flocculation.
11-10
-------
Many devices have been used for rapid mixing, including baffled chambers,
hydraulic jumps, mechanically mixed tanks, and in-line jet blending. The mechani-.
cally mixed tanks, which are typically termed "completely mixed or back-mixed
units," have been the most commonly used devices. The tanks have been designed
for detention times of 10 to 30 seconds and velocity gradients (G) 1n the range
of 200 to 1,000 sec-1. .
The principal mechanisms for alum coagulation, as discussed earlier, are adsorp-
tion and destabilization and sweep-floe coagulation.
The reactions are extremely rapid, occurring in microseconds without the forma-
tion of polymers, and less than 1 second if polymers are formed.3»** However,
sweep-floe coagulation is slower, occurring in the range of 1 to 7 seconds.
The coagulants should be dispersed as rapidly as possible in the raw water stream
for adsorption-destabilization. The suggested dispersion time is 0.1 second. This
will allow the hydrolysis products that develop in 0.001 to 1 second to cause
destabilization of the colloid. The rapid mixer most suited to this requirement
is in an in-line blender.5 »6»7
Changing Point of Coagulant Addition-
Improvements to rapid mixing facilities usually involve adjustments to the means
and point of application of the coagulated chemical. These improvements can be
accomplished at little capital cost, but can reduce chemical dosages and related
operating costs substantially and improve finished water quality. Tests performed
by the Metropolitan Water District of Southern California (MWD) have shown that
filter effluent turbidities could be reduced by as much as 50 percent with the
same coagulant dosage by properly diffusing the coagulant into the flowstream.8
Figure II-l illustrates a coagulant diffuser used successfully by MWD at one of
its facilities.
At a water treatment plant in Rio de Janeiro, the alum solution application point
was moved from 25 feet upstream of a hydraulic jump in the influent channel to
the bottom of the hydraulic jump, which is the zone of maximum turbulence.9
Approximately 5 to 7 tons of alum per day was saved by this improvement. It also
resulted in better settled and filtered water quality. A reported $100,000 (1981
11-11
-------
dollars) in chemical savings and operating costs was realized through reduction
in alum use. These results indicate the importance of applying the coagulant
solution at the point of maximum turbulence in the rapid mix process.
Coagulant Diffusers—
A 1978 report on upgrading of the 200-mgd Potomac filtration plant, prepared by
the Washington Suburban Sanitary Commission, recommended that the mechanical
mixers be replaced with diffusers to provide more effective dispersion of the
coagulant with the incoming raw water.10 In the evaluation of the rapid mix
arrangement at the Potomac plant, it was found that single pipe diffusers placed
in large influent lines would not adequately and completely disperse the coagul-
ant in the incoming raw water. Also, there was excessive travel time between the
point of coagulant addition and rapid mixers that permitted hydrolysis of the
coagulant to occur before the raw water reached the mixer. A third problem was
excessive detention time in the rapid mix chamber that prevented rapid dispersion
of the coagulant. Rapid dispersion is required in order to prevent completion of
the hydrolysis reaction before contact is made with all colloidal matter in the
water.
The modification selected to solve these problems is shown in Figure II-2. The
mechanical mixers were removed and replaced with concentric diffusers located
downstream of the existing mix chambers. The diffusers were positioned with the
orifices facing upstream into the flow to affect complete and almost instanta-
neous dispersion across the entire cross section of the pipe.
Mechanical Mixing—
Mechanical mixing is the most commonly used system for rapid mixers. The system
is effective, has little headloss, and is unaffected by the volume of flows or
flow variations. Typical design practice provides a contact time in the range of
10 to 30 seconds and a G-value in the range of 700 to 1,000 sec-1.
The mechanical mixer has been used in numerous water treatment plants. Because
the mixer speed can be changed by including a variable speed drive, it is amen-
able to operational changes due to changing conditions. Lower speeds result in
lower G-values, which are used with the addition of polyelectrolytes.
11-12
-------
In-Line Blenders—
The adsorption-destabilization reactions are very rapid, and in-line blenders
were developed to approach instantaneous mixing of chemicals. The G-value
suggested is in the range of 3,000 t 5,000 sec-1. In-line blenders with a resi-
dence time of 0.5 seconds and a water hp of 0.5 hp/mgd of flow have been
recommended.11
In-line blenders are preferred for the following reasons:12,13
• In-line blenders provide virtually instantaneous mixing with a minimum of
short-circuiting.
• There is no need to consider headlosses.
• In-line systems are less expensive than more conventional rapid mix units*
Jet Injection Blending-
Other investigators have recognized shortcomings of turbine or impeller mixers in
rapid mixing basins and suggested that this inefficiency may be due to back mix-
ing, which generally results in less efficient use of coagulation chemicals.
Mixing inefficiencies have resulted in a move toward in-line blenders using a
principle known as "in-line jet injection." A jet injection rapid mixing system
is illustrated in Figure II-3. In this system alum solution is introduced to the
incoming raw water through a series of jets positioned radially around a central
injection pipe. The jets discharge in a plane perpendicular to the incoming
water. The dilution water is provided by a pump taking suction from the raw water
line.
In this system the impact of diluting the alum solution with the raw water must
be considered. This dilution (coagulant dilution) effect was evaluated at a
treatment plant in Florida, where it was found that dilution of the delivered
liquid alum solution in ratios of 1:80 had no adverse impact on coagulation
efficiency.11*
11-13
-------
Based on these experiences, the supply pump for the injection jets should be
sized so that a dilution ratio of 1:80 is not exceeded.
A unit used for full-scale applications is shown in Figure II-4. The utilization
of this type of unit is limited .in practice, although it has been shown to have
potential advantages. Disadvantages include plugging of the orifices and the fact
that mixing intensity cannot be varied. The unit shown in Figure II-4 has design
criteria of:13
• G = 750 to 1,000 sec-*
• Dilution ratio at maximum alum dose = 100:1
• Flow velocity at injection nozzle = 20 to 25 ft/sec
• Mixing time = 1 second
The power input, P (ft-lb/sec or watts), for this type of flash mixer can be
computed from:
P - q (AH)p (1)
Where:
P = power input, ft-lb/sec
q = flow rate from orifice hole—jet discharge, cu ft/sec
= v-a = C
-------
edye of one element is perpendicular to the trailing edge of the preceding
element. The mixing element produces a radial (cross-sectional) mixing action
promoting intimate mixing of the process streams thus eliminating possible short-
circuiting. This system is relatively maintenance free but requires headloss
(typically around 1.0 to 6.6 feet). The only disadvantage of this type of mixer
is the dependence of mixing intensity on flow rate, which may vary widely. Such a
device could be installed in influent raw water piping of facilities having no
rapid mix for modest cost and contribute to improved plant performance.
Cost of Rapid Nix-
Costs provided are for newly constructed mechanically mixed basins intended to be
added to facilities not equipped with this process.
Conceptual Design—Construction costs are based on reinforced concrete basins
sized for a 1-minute detention time at plant design capacity. The largest single
basin capacity is 2,500 cu ft. Common-wall construction is used when more than
one basin is required. Mixer costs are for vertical shaft, variable-speed turbine
mixers. Mixing energy is based on a G-value of 900 sec-1. Costs for adding rapid
mix facilities are presented in Table 11-7.
Operation amd Maintenance Requirements—Power requirements are based on the
number of mixers, the G-value of 900 sec-1, a water temperature of 15°C, and an
overall mechanism efficiency of 70 percent. Maintenance material costs consist of
oil for the gearbox drive unit.
Labor requirements are determined using a jar testing time of 1 hour/day for
plants under 50 mgd and 2 hours/day for plants over 50 mgd, 15 minutes/mixer/day
for routine operation and maintenance, and 4 hours/mlxer/6 months for oil
changes. An allowance of 8 hours/basin/year is also Included for draining,
inspection, and cleaning.
Adding Chemical Dosage Control Provisions
Techniques Available—-
An excellent physical treatment plant design and good equipment selection are
worthless unless the chemical coagulation of the raw water being treated Is
11-15
-------
properly carried out. Inadequate doses of coagulant will result in an excessively
turbid finished water. Excessive coagulant doses may also cause this result and
are wasteful of the public's funds. Excessive turbidity makes the water aestheti-
cally unacceptable besides increasing the probability of microorganisms escaping
subsequent disinfection processes.
Since achieving proper ^coagulation has been a universal problem of water treat-
ment operators for many years, a wide variety of techniques have been developed
for controlling the coagulation process. Most of these involve laboratory tests,
the results of which then must be manually transferred to the full-scale plant
operation by the plant operator.
Whether or not the coagulant dose that provides the optimum result in the labora-
tory tests will also provide optimum results on a plant scale depends on whether
the same efficiency of mixing is achieved in both cases (which is unlikely).
Also, the fact that the procedure is a batch procedure results in an inherent
time lag in responding to changes in raw water conditions. This lag may be only
an hour if the operator is on duty and alert to raw water conditions, but it may
be several hours if the operator is off duty or involved in another task, such as
equipment maintenance. This method may give satisfactory—even if not optimum—
results when applied to a raw water of relatively uniform quality that contains
only a moderate amount of organic turbidity. For low-turbidity waters and waters
containing large amounts of organic material, no sharp distinction between coagu-
lant doses near optimum may exist; so it is difficult to determine from the
appearance of the sample whether adequate plant-scale results will be obtained.
Although adequate control can be achieved under certain conditions by visual
monitoring of jar tests, at other times it becomes a guessing game, with each
observer seeing characteristics that may or may not affect plant performance.
Available coagulation control techniques fall into three general categories:
conventional and modified jar tests where visual observations of supernatant,
floe formation time, floe density, and so on are used; techniques based on par-
ticle charge; and techniques based upon filtering the coagulated water and mea-
suring the filtrate turbidity.
11-16
-------
Jar Tests—
A standard jar test procedure is contained within AWWA manual M-12, titled
Simplified Procedures for Mater Examination. A standard testing plan is
required to provide a means for the comparison of coagulants or for coagulation
aids. The procedure basically consists of adding varying coagulant dosages to
several samples of the water in beakers, mixing them simultaneously with a gang
mixer, allowing them to settle, and observing the results.
Techniques Based on Particle Charge—
Zeta Potential—The first technique of this type to receive attention as a
possible means of coagulation control was the zeta potential technique. Zeta
potential is a measure, in millivolts, of the electrical potential between the
bulk liquid and the layer of counterions surrounding the colloids, as described
earlier. Because like charges repel, colloids in water—which are almost always
negatively charged—resist coagulation. When this negative zeta potential is
reduced, the repulsive forces are likewise reduced, and if agitated gently, the
colloids will flocculate. In the treatment process, the reduction of zeta poten-
tial is accomplished by the addition of a positively charged ion or complex from
such coagulants as aluminum sulfate, the iron .salts, and cationic
polyelectrolytes.
The use of this technique is also a batch procedure, as is the jar test. The
general procedure involves varying the coagulant dosage and measuring the result-
ing zeta potential. However, each water requires comparative tests to determine
the correlation between zeta potential and finished water turbidity in the plant.
Organic colloids such as those constituting organic color generally require zeta
potentials near zero, while clay-related turbidity is best removed at somewhat
negative zeta potentials. Typical values for optimum coagulation range from +5 to
-10 millivolts, depending upon the nature of the material to be removed.
Although zeta potential measurements are useful as a research tool, it is gener-
ally agreed that this method is not easily adapted to the typical treatment plant
because of the considerable degree of skill and patience required to make the
measurements, and the amount of interpretation required to make the data useful.
It is subject to the same shortcomings as any other batch test, in that sudden
11-17
-------
changes in raw water conditions may not be detected until a poor-quality finished
water is produced in the full-scale plant.
Streaming Current Detector—A streaming current detector may be used to provide
a continuous measure of the relative charges. This technique involves placing a
sample in a special cylinder containing electronic sensing electrodes at the top
and the bottom. A loose-fitting piston is then partly submerged into the sample,
and is reciprocated along its axis to produce an alternating current between the
electrodes when the cylinder contains moving charges. A synchronous motor drives
the piston and a synchronous rectifier switch, by means of which the alternating
current generated by the alternating fluid motion is made to register on a dc
meter (Figure I1-6). An amplifier with adjustable negative feedback is used to
provide an output proportional to the current collected by the electrodes.
Readout may be by a microammeter, with calibration in arbitrary units. The alter-
nating current is analyzed and related to zeta potential, which can be used to
control coagulant dosage.
The use of this approach for continuous control of coagulation is subject to the
other limitations of the zeta potential approach, including definite limitations
when used to evaluate synthetic polymer coagulants, such as when the polymer is
anionic and is used to coagulate negatively charged particles. Because the tech-
nique involves only the aspects of electronic charge, it may lead to erroneous
conclusions when it is used to study coagulants not following the electrokinetic
theory. As with zeta potential, the data must be correlated with the usual
indices of plant performance, as there may not be any consistent relationship
between charge and filtered water clarity even at a given plant for various
seasons of the year. However, the continuous nature of the streaming current
detector may make it attractive in some instances where cationic coagulants are
used in a water that shows little variation from season to season.
Continuous Filtration Techniques--
Pilot Filters—The goal of the water treatment plant should be to produce the
minimum possible filter effluent turbidity at the minimum chemical cost. The best
measure of the efficiency of the coagulation-filtration steps would be the direct
continuous measurement of the turbidity of coagulated water that has passed
11-18
-------
directly through a pilot granular filter. The application of continuous turbidity
monitoring equipment to the effluent of the plant-scale filters should be a must
for monitoring plant performance, but it has limited value as a control technique
because of the substantial lag time between the point of chemical coagulant addi-
tion and the point of filtrate turbidity monitoring. For example, improper coa-
K
gulation of the incoming raw water will result in a clarifier (with typically
' 1 1/2 to 3 hours of detention time) being filled with improperly coagulated water
before the results become fully apparent at the discharge of the plant filters.
The pilot filter technique is applied by sampling plant-treated coagulated water
from the discharge of the plant rapid mix basin. This sample stream is then
passed through a small {usually 4 1/2-inch-diameter) pilot filter to determine
whether the coagulant dose is proper, by continuously monitoring the pilot filter
effluent turbidity. This technique provides a continuous, direct measurement of
the turbidity, which is achieved by filtration of water that has been coagulated
in the actual plant. Thus, no extrapolation from small-scale laboratory coagula-
tion experiments is required. The only purpose of this test is to determine the
proper coagulant dose; it is not to predict the length of filter run, nor to
determine the optimum filter aid dose, nor to predict the rate of headless
buildup.
The pilot filter technique has the advantages of offering a continuous monitoring
of the plant-scale coagulation process with a minimum lag time. Filtering the
water through the pilot filter yields immediate information.about the adequacy of
the coagulant dose. In a typical situation, correctness of coagulant dose is
determined within 10 to 15 minutes after the raw water enters the plant. Experi-
ence at many locations shows that the pilot filter effluent turbidity is a very
accurate prediction of the turbidity that will be produced by the plant-scale
filter when the corresponding water reaches the plant-scale filter.
> The turbidity of the pilot filter is monitored continuously and recorded. High
turbidity in the filter effluent could result from either an improper coagulant
- dose or a breakthrough of floe from a properly coagulated water. To ensure that
breakthrough does not occur, supplemental doses of polymers are injected into the
pilot filter influent line. These large polymer doses may shorten the pilot
11-19
-------
filter run times, but, to prevent breakthrough, it is desirable to backwash the
filter every 1 to 3 hours in any case.
The pilot-filter system can be used to accurately indicate coagulation conditions
with a minimum lag time so that the operator can make the necessary adjustments
in the plant chemical feed. Alternatively, the coagulant feed can be controlled
automatically through the use of turbidity data from the pilot filter.
Design for Automated Coagulant Control—This system is used in conjunction with
a pilot-filter system and automatically varies the plant coagulant dosage to
maintain the effluent turbidity from the pilot filter at the desired setpoint
value regardless of variations in raw water quality and other related factors.
Other chemicals can be varied automatically by this system in direct relation to
the coagulant dosage. The coagulant control system normally consists of a pilot
filter system, an automatic control unit, and switches and controls for the chem-
ical feed equipment. The output signal from the automatic control unit to the
chemical feeders can be a time duration, current, pneumatic, or other standard
instrumentation control signal.
Basically, this automated control is accomplished by comparing a 0 to 10 NTU
signal from the pilot filter turbidimeter with the desired turbidity value. If
the turbidity signal is greater than the desired turbidity, the alum dosage will
be increased. If the input turbidity is less than the setpoint turbidity, the
alum dosage is automatically decreased. The plant -operator establishes the
desired quality through a setpoint potentiometer, and the unit adjusts the coag-
ulant dosage to maintain the setpoint value. This unit continuously optimizes the
alum dosage for a given water treatment condition and consequently is constantly
changing to establish the precise dosage.
The system may incorporate provisions for automatic plant flow pacing. When the
flow pacing feature is incorporated, the dosage output will be automatically
adjusted for changes in plant flow. At the same time, changes in coagulant feed
dosage requirements will be maintained by the control unit. The system may be
provided with an override means so that the plant coagulant feed system may be
operated in several modes.
11-20
-------
Turbidity Monitoring—
With the currently available, low-cost turbidimeters, which are accurate and
require little maintenance, every municipal water plant in the United States,
regardless of size, should provide a continuous record of the quality of its
filial product. Turbidity is an .important parameter in that it:
o Reflects the efficiency of the coagulation process and the overall treatment
provided.
e Is related to probability of escape of potentially pathogenic organisms from
the treatment plant.
• Is a sensitive indicator of the aesthetic acceptability of the product to
the consumer.
An appalling number of plants provide no continuous monitoring of plant effluent
turbidity. For the measurement of very small amounts of turbidity, the principle
of light scattering (nephelometry) is used. In this process, light is reflected
at right angles to the light beam by the particles of turbidity in the liquid, as
the beam of light passes through the liquid. The amount of reflected (scattered)
light depends directly upon the amount of turbidity present in the liquid.
,If the liquid is entirely free of particles of turbidity, no scattered light will
reach the photocells, and the indicating meter will read zero; thus increasing
turbidity gives an increase in the meter reading.
Turbidimeters is available from numerous manufacturers and their comparatively
modest cost places them within economic reach of the smallest of water systems.
Implementation of the Safe Drinking Water Act Amendments of 1986 will require
continuous effluent turbidity monitoring of all surface water treatment plants.
Particle Counting—By use of recently developed sophisticated electronic equip-
ment for particle counting, it is possible to obtain a more precise description
of particles in water than is afforded by conventional turbidity measurements.15
With instrumentation currently available, the number of particles within given
size ranges can be measured. Particle counters have sensors available in
11-21
-------
different size ranges such as 1.0-60 ym or a 2.5-150 urn sensor. From these data,
particle size distribution and particle volume can be estimated. Such information
can then be analyzed to more thoroughly evaluate treatment process performance.
Particle counting is especially useful in research and pilot-plant work, but also
has potential for improved plant operational control. A total particle counter
provides a direct measurement of the particulate matter present in water, and in
contrast to turbidity measurements, is not influenced by particle size, shape, or
refractive index. For these reasons, the total particle counter offers promise of
being a very useful and sensitive process control tool in water treatment prac-
tice, especially when asbestiform fibers are present in water supplies.16 How-
ever, existing particle counters are designed to be accurate for particles with
at least one dimension greater than 1 ym, and in some waters, submicron particles
may predominate.. In these cases, both turbidimetric and particle size distribu-
tion measurements would be necessary.17
Implementing Flocculation Improvements
Upgrading water treatment plants often requires improvements to the existing
flocculation basins. This is especially true if. the capacity of the existing
facilities is to be expanded greatly. For example, a 100 percent expansion in
plant capacity is possible by changing the existing rapid sand media operated at
2 gpm/sq ft to mixed or dual media operated at 4 to 5 gpm/sq ft. This would
reduce the flocculation detention time from 30 minutes to 15 minutes or less. At
these shortened detention times, it is vitally important that all factors that
might contribute to poor flocculation performance be. identified and corrected
during the modifications.
The objective of flocculation is to provide multiple contacts between coagulated
particles suspended in water by gentle and prolonged agitation. During agitation,
particles collide and produce a larger and more rapidly settling floe.
In designing flocculation facilities, several factors should be considered:
t Characteristics of raw water turbidity and suspended particulates
11-22
-------
• Water temperature
t Downstream treatment process
t Coagulant(s) used
Table II-8 presents general design criteria for flocculation basins.18
Type of Flocculators—
Several types of mixers have been designed for use in the flocculation process.
The most common type of flocculator is the mechanical mixer. Reel or paddle
flocculators are currently used for low- or medium-energy mixing, and propeller
or turbine flocculators are used for medium- to high-energy mixing.
The size of the flocculation basin is determined by the required reaction or
detention time. Typically, reaction times range from 15 to 45 minutes, depending
on the characteristics of the raw water, the type of coagulant used, and the type
of downstream treatment provided. For treatment of low-turbidity raw water in a
cold region, the appropriate design flocculation time could be as much as 30
minutes. A short flocculation time (15 minutes) would be appropriate for direct
filtration in warm regions or where the raw water can be easily flocculated year-
round. Also, in small plants where volumetric mixing efficiency is superior in
small tanks, flocculation times in the 10 to 15 minute range may be acceptable.
Flocculation Design Guidelines-
Table II-9 prepared by James M. Montgomery Consulting Engineers, includes
generalized flocculation design guidelines for various type of flocculation
methods and devices.18 The table also includes a listing of advantages and
disadvantages of each type which must be factored into selection of a proper
design.'
Flocculation Alternatives—
The following types of mechanical mixing devices are typically used in water
treatment flocculation:
0
o
Paddle or reel-type devices
Reciprocating units (walking beam flocculator)
11-23
-------
t Flat blade turbines
• Axial flow propellers or turbines
Typical units are shown in Figure II-7. The paddle or reel-type devices are
mounted horizontally or vertically and rotate at low speeds (2 to 15 rpm). Design
is based on limiting the tip speed of the paddle farthest from the center axis to
1 to 2 ft/sec. When paddle flocculators are chosen, the paddle area should be
kept below 20 to 25 percent of the tank section area to avoid the rotational
movement of water.
Walking beam flocculators are driven in a vertical direction, in a reciprocating
fashion. The unit contains a series of cone-shaped devices or a series of paddles
on a vertical rod. These devices impart energy to the water as they move up and
down, thereby creating velocity gradients,
Turbines are flat-bladed units connected to a disc or shaft. The flat blades are
in the same plane as the drive shaft. The blades can be mounted vertically or
horizontally, and typically operate at 10 to 15 rpm. These plate turbines are
/
effective up to a G-value of 40 sec-1, but produced high-velocity currents at
G-values greater than 45 sec-1.19
Accepted design criterion for these units is to limit the maximum peripheral
velocity to 2 ft/sec for weak floe and 4 ft/sec for strong floe. Because they
impart high velocity gradients, turbines often will not provide effective
flocculation where floe strength is weak.
The axial flow unit "pumps" liquid because the impeller has pitched blades. This
unit may be installed vertically or horizontally. Typically, these units are
high-energy flocculation devices operating at 150 to 1,500 rpm, and there is no
limitation on the tip speed. Some engineers favor these units because they are
simple to install and maintain, and produce uniform turbulence in the
flocculator.11 »19
11-24
-------
Improving Inlet and Outlet Conditions-
Sufficient time and intensity of flocculation are essential to effective opera-
tion. However, a common cause of flocculation basin deficiencies is improper
physical arrangement of the basins, allowing unequal flow distribution and short-
circuiting of the influent flow. Maximum use of the available flocculation basin
detention time must be made . by distributing the coagulated raw water evenly
across the flocculation basin inlet to prevent short-circuiting. This may require
modifying the existing influent flumes, either by adding openings or perhaps by
installing a secondary baffle across the inlet to the flocculation basin. These
inlet baffles are commonly constructed of wood with openings sized and placed to
impart enough headless so that the flow is distributed uniformly across the inlet
end of the flocculation basin. An elaborate perforated flocculation basin inlet
baffle was installed in Rio de Janeiro's Guandu treatment plant, as illustrated
in Figure II-8.9 It should be noted that the velocities through these openings
should not provide a G-value greater than the G-value provided by the floccula-
tion equipment in the first chamber.
Correcting Short-drculting--
Short-circuiting is a major problem with many existing flocculation basins.
Elimination of the short-circuiting can be accomplished by installing either
around-the-end or over-under baffles within the basin. A serious short-circuiting
problem was identified in the Guandu plant in Rio de Janeiro.9 The problem was
solved by installing baffles and changing the flocculator from a single-stage to
a 6-stage flocculator. A secondary improvement involved installation of a perfo-
rated timber baffle with square nozzle ports between the flocculation and set-
tling basin. The headloss imparted by these nozzles resulted in a more even flow
distribution to the settling basins.
Similar techniques were applied to the Potomac filtration plant in Washington,
D.C. At this facility, existing 2-compartment flocculation basins were converted
to 6-compartment basins using a perforated baffle to distribute flow evenly
across the width and depth of the downstream settling basin. It was noted that to
accomplish even distribution, the headloss through the ports must be signifi-
cantly greater than the kinetic energy of the water moving laterally past the
11-25
-------
orifices. Care must be exercised so that the velocities through the ports are
kept low enough to prevent shearing of the floe particles.
Installing High-Energy Flocculators--
Many existing conventional water treatment plants were designed with low-energy,
paddle-style, mechanical flocculators. Often, these were designed with either
variable or constant speed drives. With these original designs, emphasis was on
formation of a large floe, which might or might not settle rapidly. At high-rate
treatment plants, at reduced flocculation and downstream settling detention
times, it is imperative that the floe formed in the flocculation basins settle at
the fastest possible velocity. The existing flocculation equipment must be evalu-
ated with respect to its capability of meeting the new requirements. If the
existing equipment is inadequate, consideration should be given to replacing
existing flocculators with new high-energy turbine-style units. Where turbine
units have been installed and operated side by side with slow-speed, paddle-style
units, they have provided a more dense floe with more rapid settling characteris-
tics than were achieved with the old-style, slow-speed units.
The City of San Diego upgraded two of its major water treatment plants by modifi-
cations to the existing flocculation facilities.20. In both plants, the rapid
mixer and the first two rows of the old-style paddle flocculators were replaced
with new high-energy mixers. Figure 11-9 shows the basin with the new- and old-
style flocculation equipment. The above-described modifications, along with
conversion of the existing rapid sand to dual-media filters, effected a 30 per-
cent increase in the treatment capacity of these two treatment plants.
Before recommending replacement of existing low-intensity, reel-style horizontal
flocculators with new high-energy units, a careful plant evaluation should be
made to determine whether this modification is necessary. The existing equipment
should be inspected t-o determine whether a higher energy gradient could be pro-
duced by simple modifications to the existing equipment. These modifications may
include the addition of paddles or new mechanical drives to impart greater mixing
intensity. The condition as well as the capability of the existing drive equip-
ment will determine whether a modification of this type is physically possible or
economically practical.
11-26
-------
Case History—
The Manatee County, Florida, water treatment plant was designed with an initial
capacity of 27 mgd.11* A laboratory and plant test was carried out to investigate
the feasibility of increasing the capacity to 50 mgd by process changes. Both
laboratory and plant tests were carried out to enable the designers to identify
process deficiencies and develop modifications to permit operation at higher
plant throughput. During the evaluation it was established that flocculation was
inadequate at the expanded flow.
To obtain state approval for increasing the capacity of the flocculation facili-
ties without major construction, a plant test was necessary. Modifications were
made to the existing flocculation basins to minimize short-circuiting at the
higher flows. An existing single-compartment flocculator was modified into an
8-compartment flocculator with alternate connecting top and bottom ports. These
modifications are illustrated in Figure 11-10. In addition, a perforated baffle
wall was installed at the flocculation basin outlet to improve inlet conditions
to the settling basin. The perforated baffle wall is illustrated in Figures
II-lla and lib.
During plant tests at a flow of 63 mgd, the settled water turbidities from the
portion of the plant containing the modified flocculator were slightly better
than those from the unmodified plant at a flow of 37 mgd. On the basis of these
experiments, a state agency granted approval of essentially doubling the flow
through the modified portion of the plant. Doubling of the capacity was achieved
by eliminating two mechanical rapid mixers, dividing the existing single compart-
ment into an 8-compartment flocculator, and installing perforated baffles for
better distribution of flocculated water into the settling basins.
These improvements reduced the average alum dosage from 70 to 50 mg/1. The
settled water color was reduced from the previous 18 to 23 pcu to 10 pcu or less.
The estimated savings in capital costs {1983 dollars) were approximately
$2,000,000, and the estimated power savings were $10,000 per year.'
11-27
-------
Costs of Adding Flocculation Basins-
Table 11-10 provides costs for adding flocculation basins to existing treatment
plants originally constructed without this process. Flocculation basin costs are
for rectangular, reinforced concrete structures 12 feet deep, with a detention
time of 30 minutes at plant design flow. A length-to-width ratio of approximately
4:1 is used for basin sizing, and the maximum individual basin size utilized is
12,500 cu ft. Common-wall construction is used where the total basin volume
exceeded 12,500 cu ft. Costs are calculated for use of horizontal paddle floccu-
lators, since horizontal paddles are less expensive for use in large basins, and
they generally provide more satisfactory operation in the larger basins, parti-
cularly when tapered flocculation is practiced.
A G-value of 80 sec"1 is used to calculate manufactured equipment costs. All
drive units are variable-speed to allow maximum flexibility. Although common
drive for two or more parallel basins is commonly utilized, the estimated costs
were calculated using individual drive for each basin.
The capital cost of installing new mechanical flocculators in existing basins
varies widely depending upon type of equipment, needed basin modification,
required inlet and outlet revisions, etc. In general, the equipment cost will
range from 35 percent for small plants to less than 15 percent for large plants
of the total costs presented in Table 11-10.
Improving Sedimentation Performance
General--
Increasing treatment plant capacity may require modifications and improvements to
the existing settling basins. In most cases, a significant increase in plant
capacity reduces settling basin detention times and increases clarification rates
to the point where they will not perform efficiently at expanded flows. Problems
caused by poor entry and exit conditions as well as inadequate sludge collection
and removal may also have to be corrected when the capacity of existing sedimen-
tation basins is increased.
11-28
-------
Improving Inlet Conditions—
The manner in which flocculated water is delivered to the settling basin influ-
ences the efficiency of the clarification process. Many older plants have sepa-
rate flocculators that require transfer of the flocculated water in open or
closed conduits. The velocity and associated turbulence in these channels break
up floe and thus prevents good clarification. Wherever possible, these channels
should be eliminated and the flocculated water should be introduced directly into
the settling basin through a perforated baffle wall. Such improvements should be
considered if observations during plant tests indicate potential major problems
in obtaining good inlet flow distribution. Care must be exercised in design of
perforated inlet baffles so that velocities through these openings are not high
enough |to break up floe. Proper distribution of the flocculated water at the
settling basin inlet is critical to realizing maximum efficiency from the basin.
t
Improving Outlet Conditions—
A well designed settling basin should have an effluent collection system to uni-
formly withdraw water from the clarification zone. In rectangular basins, finger
launders extending inward from the end wall and located at uniform spacing pro-
vide the best assurance against short-circuiting. Weir loading rates should be
within the range of 10,000 to 20,000 gpd per foot of weir length. These launders
should be designed with adjustable V-notch weir plates, or have orifices placed
in the sides of the launders at uniform centers over the length of the trough.
Circular, radial-flow basins should have a continuous peripheral collection
launder with a sufficient length of weir to match the loading criteria for
rectangular basins.
''f
Installing Sludge Collectors—
The sludge collection equipment should be inspected, and it should be determined
whether it will adequately handle greater quantities of sludge at the increased
plant throughput, sludge piping and pumps should also be inspected to verify
adequate capacity. If a sedimentation basin has no mechanical sludge collection
facilities these should be added to provide complete and continuous sludge
removal at higher plant flows. A vacuum sludge removal system such as illustrated
in Figure 11-12 and shown used with tube settlers can be adapted to many existing
sedimentation basins. Figure 11-13 illustrates the air-driven, track-guided
sludge removal assembly.
11-29
-------
Installing Tube Settlers—
Where clarification rates at expanded flows are too high for adequate clarifica-
tion, tube or plate settlers can be installed to increase the effective surface
settling area and thus increase the capacity of the basin. Tube settlers can be
installed in most conventionally designed settling basins to permit a significant
increase in capacity without loss of clarification efficiency. In many cases,
settling capacity can be more than doubled by introduction of tube settlers. A
number of case histories describing water treatment plants where the capacity was
increased significantly by installing tube settlers and other modifications are
presented later in this section.
Tube settlers as well as plate settlers apply the theory of shallow depth sedi-
mentation. The use of very shallow basins with small-diameter tubes (2 inches in
diameter or less) or plates spaced at close intervals (2 to 4 inches) enables
effective removal of settleable materials in detention times of only several
minutes, in contrast to conventional settling basins designed with 1 to 4 hour
detention times. Various manufacturers have developed alternative tube settler
designs using lightweight plastics fashioned into structural modules which can be
installed in rectangular or circular basins. Application of shallow depth sedi-
mentation theory through installation of tube or plate settlers offers tremendous
potential for minimizing the size 'of newly constructed sedimentation basins or
upgrading performance and capacity of existing basins.
Figure 11-14 is a cut-away drawing of a tube settling module which provides
2-inch-square passageways either 2 or 3 feet in length. The alternating rows of
tube passageways are sloped 60 degrees from the horizontal to permit continuous
gravity draining of settled solids. Other manufacturers construct tube settlers
by bonding continuous sheets of corrugated plastic into a circular module where
all channels slope in the same direction at 60 degrees from the horizontal.
All of the previously mentioned systems are used in configurations in which the
influent is introduced beneath the tubes with the flow passing up through the
tubes but in the Lamella separator, introduced by the Parkson Corporation, the
influent enters at the top of the clarification basin and is then directed
11-30
-------
downward through a series of parallel plates as shown in Figure 11-15. The sludge
is collected at the bottom of the basin with the sludge-water flow being in the
same direction rather than countercurrent as in the other systems. The clarified
water is conveyed to the top of the clarifier by returning tubes, as shown in
Figure 11-15. The plates are typically 5 feet wide by 8 feet long, are spaced
1 1/2 inches apart, are inclined at 25 to 45 degrees to the horizontal, and are
usually constructed of PVC.
General Design Considerations-
Tube settlers can be used in either upflow solids contact clarifiers or hori-
zontal flow basins to improve performance and/or increase capacity. Of course,
they can also be incorporated into the design of new facilities to reduce their
size aricl cost. The overflow rate at which tubes can be operated is dependent upon
the design and type of clarification equipment, character of the water being
treated; and the desired effluent quality. The following paragraphs describe the
most important design and operational variables that affect tube installations in
J!
existing clarifiers.
Application Requirements-
Horizontal Flow Sedimentation Basins—The nature of the existing clarification
equipment • determines to some extent the allowable tube overflow rate and the
physical arrangement of modules in a basin. Ideal flow patterns are rarely expe-
riencedMn practice in clarification basins. Velocities in rectangular horizontal
flow basins vary throughout the basin. Flow lines diverge at the inlet and
converge at the outlet. The velocity gradient across the basin does not remain
uniform due to basin drag, density currents, inlet turbulence, temperature cur-
rents, and so on. In radial flow circular basins, the flow cannot be introduced
to impart velocity components in the horizontal direction only. The use of a
center feedwell imparts downward currents that cause turbulence and produce a
general rolling motion of the contents in an outward and upward direction.
When tube modules are installed in horizontal flow basins, it is best not to
locate them too near entrance areas where possible turbulence could reduce the
effectiveness of the tubes as clarification devices. For example, in a horizontal
basin, often as much as one-third of the basin length at the inlet end may be
11-31
-------
left uncovered by the tubes so that it may be used as a zone for stilling of
hydraulic currents. This is permissible in most basins because the required quan-
tity of tubes to achieve a significant increase in capacity will cover only a
portion of the basin. Figure 11-16 is a typical tube settler installation in a
rectangular basin illustrating arrangement of tube settling modules and utiliza-
tion of collection troughs in the covered area to ensure uniform flow collection.
In radial flow basins, the required quantity of modules can be placed in a ring
around the basin periphery, leaving an inner-ring open area between the modules
and the centerwell to dissipate inlet turbulence. Figure 11-17 illustrates one
method of installing tube settling modules in a circular, horizontal flow basin.
Upflow Clarifiers with Solids Contact—The flow paths in solids contact basins
of the upflow type are in a vertical direction through a layer or blanket of
flocculated material, which is held at a certain level and maintained at a
certain concentration by the controlled removal of sludge. The clarification rate
is governed by the settling velocity of this blanket. The purpose of maintaining
this blanket is to entrap slowly settling, small particles which would otherwise
escape the basin. When the flow is increased, the.level of the blanket will rise.
The efficiency of the tubes is dependent upon both the overflow rate and the
concentration of incoming solids. The allowable loading rate on the tubes in this
situation is dependent upon the average settling velocity of the blanket, the
ability of the clarifier to concentrate solids, and the capacity of the sludge
removal system to maintain an equilibrium solids concentration. If sludge is not
withdrawn quickly enough or if the upward velocity exceeds the average settling
velocity of the blanket, the unit can become solids critical with the result that
the blanket will pass through the tubes with excessive carryover of solids into
the effluent.
In expanding the capacity of an upflow, solids contact clarifier, the ability to
handle increased solids may be the limiting factor. The solids loading of the
basin establishes its maximum capacity. The amount of increased capacity is often
limited to 50 to 100 percent of the original capacity.
A typical tube settler installation in a solids contact clarifier is illustrated
in Figure 11-18. Typically the tube modules are placed 2 to 3 feet below the
11-32
-------
surface In the clarification zone and well above the top of the sludge blanket as
is noted in the sectional view.
Figure 11-19 illustrates a factory-built package tube settling clarifier. This
particular unit, provided in rectangular steel tankage, is designed to operate as
a sludge blanket clarifier and includes special baffling to force incoming flow
containing settleable material through the blanket. Continuous chain and flight
sludge collectors move settled sludge to a central withdrawal point.
Recommended Design Criteria--
Loading rate design criteria for tube settlers for both upflow solids contact and
horizontal flow basins for winter and summer water temperatures have been devel-
oped by one tube settler manufacturer. These are intended to be general guide-
lines only and many other factors previously mentioned can influence performance
of tube settlers. Based upon performance data from many operating facilities they
appear to be conservative.
Upflow Clarlfiers - Loading Rates—In areas where cold water temperatures (less
than 40°F) occur frequently, the guidelines in Table 11-11 apply. In warm water
areas in which temperatures are nearly always above 50°F, the guidelines in Table
11-12 apply.
Of course, these guidelines are based on the assumption that both the chemical
coagulation and flocculation steps have been carried out properly. Also, the
sludge removal equipment has been assumed adequate.
Horizontal Flow Basins - Loading Rates—As indicated in the tables, the raw
water turbidity has a direct influence on allowable tube overflow rates, as does
the raw water temperature. In cold water areas where temperatures are frequently
40°F or less, the guidelines in Table 11-13 apply. In warm water areas (tempera-
ture nearly always above 50° F), the guidelines in Table 11-14 apply.
Location of Tube Modules Within the Basin—The tubes should be located such
that they are not placed in a zone of unstable hydraulic conditions. Thus, they
are frequently placed over one-half to three-quarters of the basin located
11-33
-------
nearest the effluent launders to permit the Inlet portion of the clarifier to
dampen out hydraulic currents. The top of the tubes should be located 2 to 4 feet
below the water surface. In general, the 2 foot minimum is used in shallow
basins. The submergence of 4 feet would be considered only in clarifiers with .a
sidewater depth of 16 to 20 feet. In most basins, where sidewater depths rarely
exceed 13 feet, a submergence of 2 to 3 feet is used. The collection launders
should be placed on 10- to 15-foot centers over the entire area covered by tubes
to ensure uniform flow distribution.
Example Applications-
Horizontal Basins—An existing water treatment plant with a rated capacity of 4
mgd has a horizontal, rectangular settling basin and rapid sand filters. The
basin dimensions are 30 feet wide by 133 feet long. The surface overflow rate at
design capacity is 1,000 gpd/sq ft. The average depth of the basin is 15 feet.
The basin has a single overflow weir across the outlet end of the basin. The raw
water is obtained from a river which has a normal maximum turbidity of 25 to 30
NTU. The water temperature rarely falls below 50°F. The settling basin is
preceded by mechanical flocculation with 40 minutes of detention time at 4 mgd.
Coagulant aids are fed during periods of high turbidity and low water
temperatures to improve coagulation.
A capacity increase from 4 to 8 mgd is desired. At 8 mgd, the overflow rate
increases to 2,000 gpd/sq ft or 1.4 gpm/sq ft. The total basin loading of 1.4
gpm/sq ft is below the values shown in the above guidelines for a horizontal flow
basin under warm water conditions. However, at a higher basin loading of 2 gpm/
sq ft and a tube rate of 3 gpm/sq ft, the expected effluent turbidity is 1 to 5
NTU. This value is compatible with dual- or mixed-media filters. In light of the
moderate raw water, temperature and turbidity, a tube rate of 3 gpm/sq ft and a
basin loading rate of 1.4 gpm/sq ft should give excellent results.
Quantity of tubes required -
Capacity, gpm
(2)
Allowable tube rate, gpm/sq ft
_. 8 mgd^x 700 gpm/mgd
3 gpm/sq ft
= 1,870 sq ft
11-34
-------
The dimensions of the area to be covered by the tube modules are determined as
follows:
Area = length x width
1,870 sq ft = length x 30 feet
Length = 1.870 sq ft = 62.3 feet
30 reet
The length of 62 feet would be rounded off to be a length readily compatible with
the standard module dimensions associated with the specific modules purchased.
The modules would be installed, over an area extending back from the discharge end
of the basin for a distance of 62 feet. A baffle wall would be installed at inner
edge to force all flow through the modules. To improve uniform flow through the
modules, three new effluent launders extending 62 feet back from the existing end
wall launder would be required. The launders would be installed on 10-foot cen-
ters and the tubes would be submerged for a depth of 4 feet because the basin is
deep. The appearance of the basin would be similar to that shown in Figure 11-19.
Upflow
Basins—Assuming a plant has two 42-foot-square upflow clarifiers, each
designed for a flow of 3,000 gpm, with peripheral collection launders. The total
surface area is 1,760 square feet. The influent centerwell reduces the available
settling area by 200 square feet. The peak overflow rate now reaches 1.92 gpm/
sq ft,
which is high enough that the clarifier does not perform well, especially
when water temperatures drop.
It is desired to increase the plant capacity to 4,000 gpm per settling basin. At
this flow, the loading on the total basin settling area is 2.6 gpm/sq ft. The raw
water turbidity is moderate, 30 to 70 NTU, and the water temperature seldom falls
below 50°F.
The guidelines for upflow basins indicate that a maximum total basin loading of
2.5 gpm/sq ft with a corresponding tube rate of 2.5 to 3 gpm/sq ft can be used.
Complete coverage of the settling area would provide a tube rate of 2.6 gpm/sq ft
and would also provide a simplified support problem when compared to only partial
11-35
-------
coverage. Thus, coverage of the settling area with 1,560 square feet of tubes
would be provided. Radial launders would be added to improve the flow distribu-
tion in the basin.
Costs of Installing Tube Settlers—
The conceptual design for which costs in Table 11-15 were developed is based upon
a rise rate of 2.0 gpm/sq ft through.the area covered by the modules. By leaving
a portion of the basin open, a zone is created for inlet turbulence dissipation.
This transition zone is separated by a baffle extending from the bottom of the
modules to 6 inches above the operating water level. Uniform effluent collection
is a requirement for optimum utilization of tube settlers. To meet this require-
ment, effluent launders are spaced at 12-foot centers in all basins. Since the
hydraulic and structural requirements for tube clarification systems are unique,
the costs include tube modules, tube module supports and anchor brackets, a tran-
sition baffle, effluent launders with V-notch weir plates, and installation.
These costs may be added to the conventional basin construction costs to arrive
at a total facility cost.
Modifying Filtration Facilities
General~
Upgrading existing filters to provide improved filtered water quality and for
increased plant capacity can be done by:
* Changing existing rapid-sand media to dual- or tri-mixed media
t Removing 6 to 10 inches of sand and replacing with anthracite coal, known as
"capping"
Improving quality without increasing plant capacity could be achieved on existing
dual-media filters by:
• Adding facilities to feed a polymer filtration aid to clarified water prior
to filtration
11-36
-------
• Adding polymer to backwash water to precoat filter media particles during
backwashing to improve filtered water quality during the "ripening" period
immediately following restoration of filtration
• Installing turbidimeters on each filter to provide continuous indication of
filtered water turbidity
• Providing filter-to-waste provisions on each filter to permit wasting higher
turbidity water immediately following completion of backwashing
Filter Upgrading Evaluation—
To determine the feasibility of expanding plant capacity by filter modifications,
a thorough filter hydraulic study should be completed. Because the plant origi-
nally was probably designed for 2 gpm/sq ft filtration rates if traditional
rapid-sand media was used, settled water collection and distribution systems may
reflect these rates and in some instances, may not have adequate capacity for
expanded throughput. A hydraulic analysis of transfer channels, pipes, opening,
weirs, etc. can determine the maximum amount of water which can be delivered to
the existing filters. In some cases, minor changes in piping, gates, weir open-
ings, etc. will permit these greater flows. In other plants, entire segments of
the settled water distribution piping may have to be modified or replaced in
order to deliver the water to the filters. If possible, field tests should be
carried out to establish the maximum carrying capacity of influent piping and
channels. :
Existing filter boxes should be examined to identify,whether there are any poten-
tial problems in converting the filters to high-rate filtration. In general, most
well designed rapid sand filters are easily converted to high-rate dual- or
mixed-media filters. Backwash rates for dual- and mixed-media are the same as
those for rapid sand filter beds. In some instances, filter wash troughs may have
to be modified in order to provide the required clearance from the surface of the
filter media to the lip of the wash trough. If the clearance is inadequate in the
existing filter, there is a potential for excessive loss of the lighter anthra-
cite coal media. As a general rule where rotary surface wash is used, a filter
should have 24 to 27 inches of clearance between the filter surface and the lip
11-37
-------
of the wash trough and 8 inches between the underside of the wash trough and the
filter surface.
Excessive anthracite coal loss during backwash can also occur if rise velocities
in the area between wash troughs is too high. This problem occurs where wide,
shallow troughs at close spacing were used resulting in as much as 50 percent of
the area being covered by troughs. At velocities of 30 gpm/sq ft {rise velocity
48 inches per minute), coal will be washed out during backwash. New, narrower and
deeper troughs should be considered if this potential problem exists.
The condition of the existing underdrain can often be established by observing
the condition of the filter media. The presence of boils during backwashing or an
uneven mounded appearance of the filter could indicate an underdrain failure.
Only removal of the filter media and gravel and inspection of the underdrain will
establish its physical condition. If the underdrain is found to be damaged or in
poor condition, its replacement must be a part of the filter renovation project.
If the existing filters are not equipped with surface wash facilities, they
should be installed during the rebuilding project. Dual- and mixed-media filters
operating at filter rates of 4 to 6 gpm/sq ft with, poly electrolyte filtration
aids being used to control floe breakthrough are more difficult to clean than
rapid sand filters. A well designed surface wash system is needed to scour the
upper layers of the filter during or just prior to backwashing. This procedure
breaks up surface accumulations and prevents formation of "mud balls." Either a
rotary surface sweep with nozzles which penetrate the surface as the arm rotates
or a fixed jet surface wash system should be installed. Some advocate a 2-arm
rotary surface wash system with one arm placed at the surface and the other
located at the coal-sand interface in a dual-media filter. The theory of
operation of this configuration is that "mud balls" have a tendency to form at
the coal-sand interface and additional agitation is required at this location to
assist in cleaning adequately the filter bed.
Air scour has also been used in a limited number of plants to supplement the
backwashing procedure and to achieve the same purpose as surface wash equipment.
In general, air scour must be used only with underdrain systems that do not
11-38
-------
require supporting gravel layers. Specially designed nozzles with retaining
screens on which the filter media can be placed directly are typically used where
air scour is used. Introducing air beneath the sand and coal in a dual-media fil-
ter thoroughly mixes and scrubs the filter media. Backwashing at a flow of about
15 gpm/sq ft reclassifies the media. Advocates of air-scour systems claim less
backwash water is used with this system.
Although most well designed rapid sand filter plants have adequately sized back-
wash supply and waste piping, rate-of-flow controllers and other flow limiting
devices in filter effluent piping may not pass the higher flows possible with
dual- or mixed-media filters. Replacing undersized rate-of-flow controllers with
new pneumatically or electrically controlled controllers may be required to
handle the higher flows. Frequently, filter effluent piping is also undersized
and will not handle flows associated with filter rates of 5 to 10 gpm/sq ft and
must be replaced. To offset filter gallery space limitations, more compact but-
terfly valves and fabricated steel piping can be used in place of bulky cast iron
gate valves and piping.
A modification of the method of filter rate control can also be a beneficial
technique for increasing the capacity of an existing rapid sand filter. Variable
declining rate filtration can be adapted to an existing plant at minimal expense.
Figure 11-20 shows a desirable design for variable declining rate operations.
This modification, eliminates rate-of-flow controllers which contributes to
troublesome operation and in old plants are often totally inoperative. Another
advantage is that the filter media is submerged at all times eliminating negative
heads which cause air binding. As illustrated in Figure 11-21, the rate on a
clean filter is initially constant and declines as the filter becomes dirty.
Proponents of the variable declining rate method of filter flow control regard
the simplicity of operation and the lack of expensive and troublesome rate-of-
flow controllers as the overriding reason this technique should be seriously
considered in any plant that must be upgraded.
Conversion to dual- or mixed-media filters which operate at a filter rate of 4 to
10 gpm/sq ft should be accompanied by installation of filtration aid application
equipment. Filtration aids, which are typically nonionic polyelectrolytes, are a
useful tool in optimizing the performance of high-rate filters. These materials
11-39
-------
prevent premature filter turbidity breakthrough by controlling floe penetration
into the filter. A dosage sufficient only to retain floe in the bed until the
maximum operating headloss is reached represents optimum conditions. This condi-
tion is illustrated by the hypothetical headloss/turbidity breakthrough curve in
Figure 11-22. Typically, polyelectrolyte dosages ranging from 0.02 to 0.1 mg/1
are appropriate depending upon applied water characteristics. Polyelectrolytes
are especially valuable in preventing premature floe breakthrough where waters
are cold.
By adding a polymer to the water to be used for filter backwashing, a reduction
of the initial turbidity breakthrough period and peak can be obtained in addition
to a much improved settling-thickening rate for the backwash solids.21 Polymer is
added only during the first half of the 10 to 15 minute backwashing period. In
pilot studies it was found that adding 0.1 to 0.15 mg/1 of nonionic polymer to
the backwash water reduced the relative initial turbidity breakthrough (ITB) by
about 50 percent to a 15 to 30 minute period on dual-media filters. Savings in
backwash water would be realized by preconditioning filters with polymer through
this technique.
Regulatory Agency Design Standards-
Design of new facilities as well as upgrading of existing water treatment plants
must conform to regulatory agency requirements. Some state agencies have specific
design criteria for unit processes used in treatment facilities and encourage
design consultants to conform to these regulations. Design standards for filtra-
tion plants established by the Texas Department of Health is included in the
Appendix and is one example of a good standard. Many standards are based upon the
"Ten State Standards for Design of Water Treatment Facilities." Others have been
developed through years of experience in a specific state with treating particu-
lar water supplies. Most have provisions to consider and accept waivers to the
specific requirements and have procedures by which new or alternative technology
can be considered. In Section III policies and procedures involving pilot testing
to demonstrate alternative treatment technology are discussed.
11-40
-------
Filter Design Checklist—
1. Filter media sizing and selection should be based on pilot tests. If this is
not possible, data should be obtained from similar applications to determine
the suitability of the media design.
2. In dual- and mixed-media filter systems, provisions should be made for the
addition of polyelectrolytes directly to the filter influent.
3. The turbidity of each filter unit should be monitored continuously and
recorded.
4. The' flow and headloss through each filter should be monitored continuously
and recorded.
5. Provisions should be made for the addition of disinfectant directly to the
filter influent.
6. Pressure filters must be equipped with pressure and vacuum air release
valves.
7. Provisions should be made to divert any filter effluent of unsatisfactory
quality (i.e., provide a filter to waste cycle).
8. Provisions should be made for automatic initiation and completion of the
filter backwash cycle. The filter controls and pipe galleries should be
housed.
\
/
9. Filter piping should be color coded.
10. The filter system layout must enable easy removal of pumps and valves for
maintenance.
11. The backwash rate selected must be based upon the specific filter media used
and the wastewater temperature variations expected.
11-41
-------
12. Filter backwash supply storage should have a volume at least adequate to
complete 2 filter backwashes.
13. Adequate surface wash or air scour facilities must be provided.
14. There should be adequate backwash and surface wash pump capacity available,
with the largest single pumps out of service.
15. Backwash supply lines must be equipped with air relese valves.
Case Histories-
Sacramento, California—The City of Sacramento has two water treatment plants
with a combined treatment capacity of 140 mgd. The newer of these two water
treatment plants is located on the American River. It was constructed in 1963,
and has a design capacity of 60 mgd. Increased demand, especially during the
summer irrigation season placed a strain on both facilities, and additional water
capacity was needed to meet present as well as future water requirements.
Although both water treatment plants were amenable to expansion, the city
selected the newer American River Plant for the first phase of the project. A
comprehensive study was carried out by the city's consultant to determine how
much additional capacity could be gained from the American River Plant by a
staged expansion program.22 The preliminary review indicated that replacing the
existing rapid sand filter media with either more efficient dual- or mixed-media
would accomplish a significant expansion in capacity and could be achieved with
no major plant structural modifications or additions. The existing filters were
designed at 3 gpm/sq ft and consisted of 24 inches of silica sand over a graded
gravel underdrain. At the expanded flow of 105 mgd, the filter rate was increased
to 5.3 gpm/sq ft, which is well within the capabilities of a well designed dual-
or mixed-media filter.
The existing 60 mgd conventional plant at design flow, provided 25 minutes of
flocculation and 115 minutes of settling prior to filtration. At 105 mgd, the
flocculation and settling times would be reduced to 14 and 65 minutes, respec-
tively. Although the amount of pretreatment would be substantially reduced at the
11-42
-------
expanded flow, either dual- or mixed-media filters could handle anticipated
higher turbidity loadings.
Following preliminary assessment of the feasibility of increasing the capacity of
the American River Plant extensive pilot filtration tests were carried out by the
city. The specific purpose of the tests were to evaluate the performance of
various types of filter media and determine the economics of their use in
expanding the capacity of the plant. These tests were done in February through
May of 1982.
Four types of filter media were evaluated simultaneously in four side-by-side
pilot filters. Each filter was equipped with instrumentation for measuring
turbidity and headloss. Figure 11-23 is a flow diagram showing the general layout
of the pilot filtration equipment. Feedwater to the pilot filters could be drawn
from either of two sources: (1) from the raw river water pipeline (before alum
addition); or (2) from the plant filter influent flume following alum addition,
mixing,!flocculation, and settling.
During the three months of testing, influent turbidities varied from 3 to 33 NTU.
For the controlled variables: (1) alum dosages ranged from 12 to 21 mg/1; (2)
polymer:dosages varied from 0 to 0.3 mg/1; and (3) filter rates were set at 5,
6.5, or'j 10 gpm/sq ft. Test runs were terminated for one of two reasons, either
high headloss or high turbidity. The City of Sacramento, established a maximum
turbidity goal for drinking water of 0.1 NTU. During a particular pilot test run,
an upper tubidity limit of 1 NTU was selected as the point at which a run was
terminated.
All of the pilot filtration tests performed on plant settled water clearly demon-
strated that the turbidity goal of 0.1 NTU could be obtained at all filter rates
examined (5 to 10 gpm/sq ft) by either dual- or mixed-media. Mixed-media
possessed a greater reliability in achieving the turbidity goal.
Based upon the results of the pilot filtration study, it was recommended that the
treatment plant be expanded by replacing the existing sand filter media with
either dual- or mixed-media. It was also recommended that equipment be installed
for preparing and feeding a polyelectrolyte filtration aid.
11-43
-------
The treatment plant has been operated successfully at 101 mgd which is the maxi-
mum capacity of the existing raw water pumps. Before the treatment plant can be
operated at sustained flows beyond 100 mgd, additional raw water pumps and high
service finished water pumps will be required. During qualifying tests effluent
turbidities averaged less than 0.1 NTU (generally 0.06 to 0.08 NTU) and filter
cycles between backwashes exceeded 48 hours.
Erie County, Hew York--The application of mixed-media to this plant in Erie
County, New York, is of special interest because it provided side-by-side compar-
isons of the efficiency of mi-xed-media filters and rapid sand filters in a large
plant. The results have been reported by Westerhoff and are summarized below.23
The Erie County Water Authority serves filtered water to approximately 390,000
persons in all or part of 12 towns and the City of Lackawanna in Erie County, New
York. The average water production in 1969 was 44.3 mgd. Water is taken from Lake
Erie, treated in the Water Authority's two filtration plants (Sturgeon Point and
Woodlawn filter plants), and pumped into the transmission and distribution
systems. Most of the water is treated at the Water Authority's Sturgeon Point
filter plant, located in the town of Evans, approximately 30 miles south of the
City of Buffalo. This plant was originally constructed in 1961.
The water treatment processes at the Sturgeon Point filter plant consist of aera-
tion, chemical addition, rapid mixing, flocculation and sedimentation, filtra-
tion, and chlorination.
In 1965, the water demanded by a burgeoning population in the suburban Buffalo
area brought about expansion requirements. The first stage (to a capacity of 60
mgd) was needed immediately. The Water Authority decided during the first-stage
expansion to install mixed-media in six new filters. The four original filters
contain fine-to-coarse graded-sand media.
Special data were collected from a total of 48 parallel-filter runs during a
12-month study period. The sand filters were operated at the standard 2 gpm/sq ft
filtration rate as a control. The filtration rate on the mixed-media filters was
varied between 2 to 10 gpm/sq ft.
11-44
-------
During the parallel tests, the filter runs were terminated at a headless of
8 feet. Typical turbidities applied to the filters were 2 to 4 NTU with little
variation as the clarifier overflow rate was varied from 625 to 940 gpd/sq ft.
Alum feed rates were typically 14 to 16 mg/1. The following conclusions were
reached:
i
• More than 80 percent of the time, the filtered water turbidity from the sand
filters was 0.10 NTU or less.
• Nearly 88 percent of the time, the filtered water turbidity from the mixed-
media filters was 0.10 NTU or less.
• The mixed-media filters operating at filtration rates of 2 to 10 gpm/sq ft
consistently produced a lower filtered water turbidity than the sand media
filters operating at a filtration rate of 2 gpm/sq ft.
• At filtration rates up to 6 gpm/sq ft, the mixed-media-filter effluent had a
lower total microscopic count than the sand media filters operating at, a
2 gpm/sq ft rate. ' .
* The mixed-media filters operating at 5 gpm/sq ft had an average length of
filter run of 29 hours. At 6 gpm/sq ft they had an average filter run of 20
hours.
• The mixed-media filters operating at 5 to 6 gpm/sq ft used a considerably
lower proportion of wash water than the sand filters operating at 2 gpm/
sq ft (1.8 percent as compared to 2.5 percent on the average).
As a result of these tests, it was concluded that the mixed-media filters at
6 gpm/sq ft produced an end product superior to that from the rapid sand beds at
2 gpm/sq ft with concurrent economic advantage. The finished water goals adopted
for filtered water at this plant were established for the mixed-media filters at
6 gpm/sq ft as follows:
11-45
-------
Turbidity Less than 0.10 NTU average
Less than 0.50 NTU maximum
Total microscopic count Less than 200 su/ml average
Less than 300 su/ml maximum
Color Less than 1 unit
Odo.r Not detectable
Aluminum Less than 0.05 mg/1
Iron Less than 0.05 mg/1
Corvallis, Oregon—Corvallis increased the capacity of its municipal water
treatment plant in the early 1970's by two and one-half times without increasing
the size of the plant. Using the Willamette River as a source of supply, the
original treatment plant was designed and built in 1949 as a conventional rapid
sand filtration plant. It was designed in increments of 4 mgd capacity, each
increment consisting of floccu.lation-sedimentation basins and two rapid-sand
filters.
The initial plant capacity of 4 mgd rose to 8 mgd with the addition of a second
increment in 1961. Under the original plan, the plant was scheduled for expansion
to its ultimate capacity of 16 mgd with construction of the third and fourth
increments in 1968 to 1969. Capacity of the Willamette River treatment plant was
increased without structural addition from 8 mgd to 21 mgd. The expansion of
capacity was achieved by application of: (1) mixed-media filtration, (2) shallow-
depth sedimentation with tube settlers, and (3) coagulation control techniques.
The turbidity of the raw water drawn from the Willamette River normally ranges
from 15 to 30 NTU, with surges up to 1,000 NTU. The water is soft (15 to 30 mg/1
hardness) and exhibits periodic taste and odor problems. Typically, the following
chemical dosages are used: alum, 20 to 40 mg/1; lime, 10 to 20 mg/1; chlorine,
2 mg/1; polymer (as coagulant aid), 0.1 to 0.2 mg/1; activated carbon, 5 to 10
mg/1 (for taste and odor control).
Virtually all of the treatment piping had to be enlarged to handle the increased
plant flow. Settling tube modules were installed over about 60 percent of the
3,500 sq ft of rectangular settling basin area. The tubes were installed on
simple "I" beams spanning the width of the basins and are located at the
11-46
-------
discharge end of the basin. New effluent weirs and launders were also installed
to provide good flow distribution through the area covered by the tubes.
Overflow rate in the sedimentation basin area covered by the tubes is
4.2 gpm/sq ft. This compares with the 1.05 gpm/sq ft design loading on the basin
before they were installed.
Utilization of the existing basins with the tube settlers reduces the turbidity
loading on filters, especially during periods of high river turbidity, and pro-
vides more economical operation by increasing filter runs during normal river
conditions. With an extreme turbidity range of 2 to 1,000 NTU, filter runs have
been increased from 40 hours experienced with the old conventional basins at much
lower plant throughput rates to 60 to 65 hours. Operating on raw water turbidi-
ties of 15 to 30 NTU, the tube effluent has had turbidities of 1 to 2 NTU.
The capacity of the filtration portion of the plant was increased by modifying
the filter piping and by replacing the rapid sand filter media with mixed-media.
At the 21 mgd rate, the filtration rate is 7.5 gpm/sq ft.
The pilot filter coagulation control system is used in conjunction with contin-
uous reading turbidimeters on each plant filter to assure satisfactory filtered
water quality. The full-scale plant continuously produces a finished water tur-
bidity of 0.2 NTU or less. •
The cost to expand the plant from 8 to 21 mgd capacity (including a new 5 MG
reservoir, a new high-service pump station .and a crosstown 16-inch transmission
line to the new reservoir) was $430,000 (1969 prices). It was estimated that the
same expansion through the addition of new settling basins and filters would have
been at least $650,000 (1969 prices). Thus, improved performance and increased
capacity were achieved while realizing a substantial savings in cost.
North Harin County Mater District, Novato, California—In 1973, the Stafford
Water Treatment Plant owned by North Marin County Water District, which provides
water service to Novato, California, and nearby areas, was expanded by installing
tube settlers and mixed-media filtration materials. The turbidity of the Stafford
11-47
-------
Lake supply varies from about 2 NTU to about 35 NTU, with a usual range of about
10 to 20 NTU. Coliform MPN values range from 5 to 72,400 per 100 ml. The lake
water color ranges from 25 to 50 units.
Experience at the Stafford plant had not indicated any operating problems with
high raw water turbidities or coliform densities. However, high color and low
turbidity in the range of 5 to 10 NTU in raw waters produced problems of clarifi-
cation. Further, high plankton populations caused taste and odor problems and
other biological forms had seriously shortened filter runs on occasion.
The basic steps in the existing treatment process prior to upgrading were as
follows:
• Raw water was disinfected with chlorine and coagulated with alum in a down-
flow hydraulic mixing chamber.
• Flocculation was provided using additions of recirculated sludge and lime.
t Flocculated water was clarified and sludge was removed and sprayed on land
for disposal.
• Activated carbon was added immediately prior to filtration in rapid sand
filters.
• Dechlorination to the desired residual was accomplished by the use of sulfur
dioxide in the filtered water clearwell.
• Elevated storage of water for backwash!ng filters was provided.
• Water used in washing filters was stored in a recovery pond and pumped at a
controlled rate back to the head of the plant for reprocessing.
In March, 1973, the district retained a consultant to prepare a report on the
design of. plant improvements to increase the capacity from 3.75 to 6.2 mgd.
11-48
-------
Previous preliminary studies by the district engineering staff had indicated that
this might be possible to accomplish in a short time period without the need for
major plant additions.
The major changes required were modification of the clarifier by adding settling
tube modules, replacement of the existing filter sand with new mixed media and
the installation of a coagulant control and turbidity monitoring center.
To handle the higher flow rates it was necessary to increase the capacity of
certain pumps, pipelines, meters, chemical feeders and valves, and modify
controls.
Cost of the improvements amounted to $337,445 (1974 prices). This made possible
an increase of 2.45 mgd in plant capacity at a unit cost of $137,770 per mgd. The
expanded plant was ready for operation in just eight months after award of the
construction contract in time for the high demand Memorial Day weekend on May 24,
1974.
Other modifications were needed to permit operating at higher plant flows. The
mixing chamber, previously operating hydraulically, was equipped with a vertical
mechanical rapid mixer. Provisions were made to add polymer as an aid to floccu-
lation. Facilities were furnished for storing and feeding sodium hydroxide solu-
tion replacing lime for pH adjustment.
A major requirement was to increase the capacity of the settling basin. This was
done by installing tube settling modules. As a general rule, installing tube
settling modules in an existing basin will permit clarifying two to four times
more water than can be clarified within the same basin volume with no loss in
efficiency. In this particular case, because of the light alum floe present, it
was recommended that the flow rate be increased from 4.0 mgd to not more than 6.5
mgd, or 1.6 times the existing capacity. This would increase the surface overflow
rate from 1.52 to 2.5 gpm/sq ft, decrease the detention time from 1.3 to 0.8 hour
and would increase the weir loading from 7 to 11 gpm/ft. Despite the high over-
flow rates and shortened clarification times, good operation was obtained at
these loadings.
11-49
-------
The tube settling modules were installed in the annular outer settling compart-
ment of the circular clarifier. Figure 11-24 illustrates the arrangements of tube
modules in the existing basin. The existing sludge collection mechanism still has
enough room to operate in the modified basin.
The sand filters were completely rebuilt. The filter bottoms were removed and
replaced. The single-media sand bed was replaced with mixed-media supported on a
gravel bed with an intermediate layer of coarse garnet gravel. New rotary surface
washers were installed. Filter piping and valves were modified as necessary to
accommodate the increased flows and gate valves were replaced with butterfly
valves.
A new coagulant control center with turbidity monitoring was added to the exist-
ing facilities. The coagulant control center was installed to provide the opera-
tors with a precise method for optimizing the coagulant dosage. Continuous filter
effluent turbidity monitoring assures that the California Department of Health
Services mandated standard of 0.5 NTU is never exceeded.
Since 1974, the plant operating experience has been good. The expanded facilities
have been operated routinely at their maximum design capacity. The settling tubes
have functioned well at surface overflow rates of 2.5 gpm/sq ft and the mixed-
media filters have performed very satisfactorily at rates up to 6.5 gpm/sq ft.
Finished water turbidities are consistently less than 0.15 NTU and the bacterio-
logical quality has been excellent.
A demonstration test was conducted for two weeks in June 1975, to show that the
design flow criteria and water quality requirements of the California State
Department of Health Services would be achieved and maintained. The results of
the June 1975 test are shown in Table 11-16.
11-50
-------
TABLE IM. SUMMARY OF POTENTIAL PROBLEMS IN WATER
TREATMENT UNIT OPERATIONS
Operation
Potential Problems
1. Chemical Feed
2. Rapid Mix
3. Flocculation
4. Sedimentation
5. Filtration
- Choice of chemical (s).
- Choice of chemical dose and pH.
- Control of chemical addition. Performance of
chemical pumping equipment.
- Maintenance of chemical feed lines.
- Flexibility in feed system to allow for
changing the point of addition, adding
chemicals at more than one point, etc.
- Sequence of adding different chemicals.
- Degree of dilution of chemicals before
Injection.
- Type of rapid mix. In-line versus mechanical
mix.
- Number of rapid mixers.
- Method of chemical addition.
- Mixing speed/detention time.
- Optimum detention time.
- Optimum mixing intensity.
- Number of stages.
- Adequate baffling to approximate plug flow
conditions.
- Surface loading rates.
- Short circuiting due to wind, temperature,
density differences, inlet and outlet design.
- Amount and rate of accumulation of sludge.
- Sludge removal.
- Filter rate and rate control.
- Hydraulics.
- Chemical pretreatment of water reaching the
filter.
- Inadequate backwashing.
- High filtered water turbidity.
- Excessive headloss.
11-51
-------
IQ
O»
1C 4J O
4-> I/I O
£3J
CO
LU
UJ
M
BO
< en
U_ UJ
^ d
U
j^j ** **
u_ T; is
O. O
« 0
u
Q
UJ
0)
Ol •
CQ
fl oo
»«* M «*> o O r*.
^-i CM m «• UT> vo.r^ GO tfi O >^ Csl
11-52
-------
CO
I-
LU
t-
LU
LU
<
o
LU n!
u S
Q 2
LU **•
to
LU
m
i
±1
CO
4J (A
•*- in e
S^S
at
01 ••
« * -o
t. o ~
^ «^
« o *3
^ «o o>
a. a. €
<»)
CM
o>
VI
01
01
o
OiAeM
^l «•* IM
ei
i
tr>
•^ esi
IA
«O
o o f-
^ * «NJ
8 =
O
0» ^1
R8
11-53
-------
UJ
CD
<
tn
a>
"e
I
CO
I-
Z
S
UJ
Cki
I- in
< O
en
u
* Q
o g
O > g.
LU
i s
I Q *
2 §
< c o &
>C e
|Q O
^^ IB
a. &
CM CO *P
(A
5
2
IT)
•4 o tn tg o 4T
** * !•
VO OO
•w
^- CM m
S S S
-* * CM
O — CM
tfi
3 =
e
U 4J
ai -^
o -
a. —
11-54
-------
h-
Z
u
u_
3S
id
£g
LU Z
<°
I <
LU CO
m
CO
(O
0)
« 4->
4-> M
e e
i
. o •
a u t—
o «»
o>
u •_
&a
U
o.
si
o
O O CO *
i 2?
f- 01
I/) «•»
Ol
CO
co m
s s %
v
«*- VI
o s
C VI
o *•
C 9
« F-
u
IB
csj e esi
f^
^g
S8-
•*- «fl
at
eiAiAO
og m o tet
CM
0)
X (O
tfl O O
w ^^ ••
o >, o
U JE *-
11-55
-------
Ol
UJ
UJ
UJ
U
25
= O
UJ UJ
—I UJ
a. u-
^>O
t- c
§2
£S
So
UJ 00
vO
i
CQ
« fff
a> *
.= s1 °W<0"01D* -
at o
9^^O O
^*0 3
ai e o
o
c
a
OJ|«-» C\l _
O o 01
»- £ >
* , » «
•o
^~. 3% jf| 2L ^^_ J'
^ * o> ^
L.k^ 9144 9
01
9 ^-e
ai . •*•
g e CM
"> _
CM
^»
I,'
1 = 11
s|S|
%o •
_______ +J M C
O> (rt « O
£>B2:'
»— IQ ot vMcsimoofx*
o. OLE »-" * <«
iQ
o ^ ai f
3 ** u ^
^ u e •
O 01 ^ |/>
e b. «« J3
«*§^
SS^£
11-56
-------
LLJ
U
U
F
z
ULJ
o.
CL
D
**
21
^ a.
O<
s <
fe>
LLJ 00
UJ
00
•• ^p^
SJ
t;
S
«
e
ai
d
S-S "•
o
h-
CM
CM m K» o» r* i^ CM
.
^ CM »*> ^« «c
• •
cu to
S S S S
• • • •
oeeecMtni^cMane^
V CM f> O •
lA
r^
CM
11-57
-------
LU C£.
H- LU
< 7
^5
a^
oi (J
^ D
to to
Z Z
t/> O
< u
CO
>
Z ot:
° 2
O
(J
h-
<
_i
D I-
U Z
^°
2^
yj (J
U Q
(J *7
t^ <
LU
Q vn
-_ LU
® _j -J
i <^
LU QS rj
_i LU ^
CQ Z ±
< UU'^
H U Q.
CO
•o
c
2 C/3
•o
Water
rification
u
Reservo
CO
R I* I-1 I
r- — f.
O wi
«nl7rT 7
-~
E
4>
t>
C _ u - C -
- o - § - o
e (5 c « c eo
o b o b o i:
S S 'fs E S S
> u > u > o
c i> c y c «>
o j= o .j= o J=
U Q U O U O
en
.£
.2
E -
.
.S
s
5
<2
,.
.E «
s
-852 s.g
'
di JL* *^
SL^ a
3i
I &
'I. S «~ -" > "i
>* O *•* c S ™
•B o S 8 5 c g
v — "3 eg S 2 —
» c «« *, ^ « ^
S ^ § a 2 g §
•S o -p c o g
•^ *^ . "* iH
- £ 1 "51
S3 5, E o
•& -e "o ^ °
>. 3 Q. O w
.!£ i .^i o o
c O % *p£ u
o * .1 a ^
"5; T> ,~i u X
o.^: w — c*
c <*» g . -S
'>. & 5 L "i
QD MM 2* Iii
j > u IA
e > ° 0
t> i c TI
«S S .2 ta
O S •>
-3 •£
5 .2?
*
M = ^ E
SP o o •=
.S o £ x
fe -s.
^—
— «
.
u e *~ — o — 5
SL « ^ s ^ - *
-8.
rt-o
11-58
-------
RS
»
UJ
_l
i
u
•z
C£
a.
i-
Z
UJ
2
-*• LM
-z
fii =
i-u
s^
wn o«i
UJ UJ
Z 2
30
Qi£
s
•*
CO
<
u
U
dvantages and
Disadvantages
o 'C
'iu
CO
<- o
O B
w
Ko-
g.*2©
«-o e
G. iJ T3 S
"2
£ "5
''SJ
i
li
ill
tt
"3
u
C
v
>
(8
O "O
N T3
'C g
O O-
X
u
e
m
SJ S
.1«
•o
"S t*
t i
s
j^
u
I
"O
.Is
°3
IP
<2
•
&i
o c
U K
a. u
^1
It
— -C
15
u .s
c
«•§
5^
E »
v .£
S ?
• K «
5 £
0= U
S s
o ^
E t-
« 2
ea C
•* «
M 3
.— U
§ 8
si
o U
11-59
-------
1-
1
e*
P"
r*
LU
p"
^^
>
LU
U
u_
Q£
D
CO
(J
2
Z
LU
LU
of
9rZ
5?Q:
o£ ^
oS
LU ~>
gnU
r- U
O\HT
_i
U"-
Q^
LU Z
<5
s?
p >
yj QO
O
M^
v^
1
LU
00
<
a
F— • O
«3 *J O
** W» O
O O *
1-0 —
V)
o —
O IO
0)
U O
Si
c •
OJ f-4
e **
ra
-§ t
to >,
e o
^ o
^^ *
(D •-"
(U
H
To *o
*J 4J O
t- tf> O
a. o *
o **
a> •
(o 5
at ^- a
^ ""
^^
C «^
WOT
«T 0. E
5
>.
t.
c
c
V
4>
<-
^^ *••• f^ f*«* ^o r^* ro Ct ^^ co ^o ^i
\fl ^S CO 1*** ^D CO C^J C^ ^^^ *^ ^^ v^
^ CTt OC 00 tfJ ^f ^T CO CO C\J CSJ
~
-------
TABLE 11-11. UPFLOW CLARIFIER LOADING RATES FOR COLD WATER*
Overflow Rate, Based
on Total Clarifier
Area
gpm/sq ft
1.5
1.5
1.5
2.0
2.0
* Water temperatures
Overflow Rate,
Portion of Basin
Covered by Tubes
gpm/sq ft
2.0
3.0
4.0
2.0
3.0
less than 40°F
TABLE 11-12. UPFLOW CLARIFIER LOADING RATES
Overflow Rate, Based
on Total Clarifier
Area
gptn/sq ft
2.0
2.0
2.0
2.5
2.5
* Water temperatures
Overflow Rate,
Portion of Basin
Covered by Tubes
gpm/sq ft
2.0
3,0
4.0
2.5
3.0
i
above 50°F
Probable Effluent
Turbidity, NTU
1-3
1-5
3-7
1-5
3-7
FOR WARM WATER*
Probable Effluent
Turbidity, NTU
1-3
1-5
3-7
3-7
5-10
i
11-61
-------
TABLE 11-13. HORIZONTAL FLOW BASINS LOADING RATES FOR COLD WATER
Overflow Rate, Based
on Total Clarifier
• Area
gpm/sq ft
Overflow Rate,
Portion of Basin
Covered by Tubes
gpm/sg ft
Probable Effluent
Turbidity. MTU
Raw Water Turbidity = 0-100
2.0
2.0
2.5
3.0
1-5
3-7
Raw Water Turbidity = 100-1,000
2.0
2.0
2.5
3.0
3-7
5-10
* Water temperatures are frequently 40°F or less,
TABLE 11-14. HORIZONTAL FLOW BASINS LOADING RATES FOR WARM WATER
Overflow Rate, Based
on Total Clarifier
Area
Overflow Rate,
Portion of Basin
Covered by Tubes
Probable Effluent
Raw
Raw
gpm/sq ft
Water Turbidity - 0-100
2.0
2.0
2.0
3.0
Water Turbidity = 100-1,000
2.0
2.0
* Water temperatures are frequently
gpm/sq ft
2.5
3.0
4.0
3.5
2.5
3.0
50° F or above.
Turbidity, NTU
1-3
1-5
3^7
1-5
1-5
3-7
•f
11-62
-------
1-
z
tit
UJ
1-
<
UJ
tt
L_
r^
tt
UJ
i^
r~
<
>
>
UJ
U
<
u_
QC
D
CO ^
tit
¥3
Z — '
9o
<^
/•^ ^>
ot ^
( T
Su
D-^
Z
«/• — '
Qt u-
ot
U_ uj
«/l
J/l
j— UJ
cr> £9
o2
BK
QZ
UJ -^
uj CO
ft
to
•~
UJ
^kj
03
1—
«
O)
»- • o
« 4J O
*J wi O
o o «
>— o ^
-**•
4J
Ul
*-
*J ^.
el^to
^^r
n
•o
i «.
>-
o g
Z§
* «
C- ^4
ai M»
1
_
0 * O
** ** o
et e •
10 O •-•
O V*
a»
o» •
«^» •••
S w
fe o o.
51-
•
>>
^ 4J
« O tJ
^- * en
a. o. e
10
•^
u
>
o
en
a
*j
ia
0
r-.
K^^rr^ooiotnooo
OO«^o—
1_
§.
O.
. s
vt
m
>
« -
g
TJ
o
e "-
at .
•a i»
||
"3 H
^ *
^^
in
-------
H
LU
O
h-
1-
cn
Z
0
5
Uj
Q
U
Z
c*
.Q
>
LU
U
Z
c^
^~
1-
Q
CO
2£
H
.
N£
T™
1
"•"•
UJ
•^
00
J-
f—
"oj
s
fO
o>
o
v>
c
o
•r-
• k
<0
c
O)
0)
o
o>
4J
f^M
•r-
U.
CO
CSJ
1— «
0)
cn
c
«J
oc
o
o
_^J
^^^
r*. vo «*• co
evj 0 *-t
O O O 0
IO
<0
.c
4-1
•*^
3;
CVJ tD ^}*
Ol
10
CO O «-i »
* ' • • »
+j
•^
•o
^O VO *^ r*
CO 0 rH C.
* • » 3
0 O 0 4^
c
•r-
"O
o>
CO
to to LO at
CO O »-« f_
• • • a.
O O O X
at
to
at
>
J= C t—
O) 2
-------
:••>•.
P
a
•'**•
.....
-
10
•»— COAGULANT
FEED LINE
r4"
i '
CRES PIPE
OA-'H-H"
1
\
COAGULANT
DIFFUSER
-
?**"
»'*/''
v^
•>£*•;
ll&£ ^
[
[
MULTIJET1
SLIDE
GATE ^
^
J^
A
pOAGULANT/
;•
]
,
•^
N
]
|
J
•«•*«
\
2>
»
&
ea
Np^i
/
WATER
FLOW
:••?:«•
1 i*'^!V:*
FRONT VIEW
SECTION VIEW
Figure 11-1. COAGULANT DIFFUSER
11-65
-------
DIFFUSERS
RAW WATER
~\
-&
M
ok
I&
rJ
h
3
oj
h, ...„, MFfHANTrAI
~— — — nLLnMii lunL i
MIXER I
SEE DIFFUSER
DETAIL BELOW
RAW UATFR : — fc—
/
—
EXISTING RAPID MIX
ARRANGEMENT
PROPOSED DIFFUSER
ARRANGEMENT
REFLECTED PLAN
HOLLOW PLASTIC HUB
COAGULANT
SOLUTION
BRACES FOR
, SUPPORT
CROSS SECTION OF DIFFUSER
Figure 11-2. DIFFUSER FOR COAGULANT APPLICATION
11-66
-------
00
:' •* .'..y. '• • •,-»' ;•' :•«.•/' •'«' .•*'.• •-*•' .».»••• •»•
a.
2
Q.
LLJ
55
DC
D
DC
UJ
U
P
cc
111
11-67
-------
Vertical Turbine Pump—- >
Activated Carbon and Polymer
Raw Water
8 in. Discharge
6 In. Orifice for Homogenizatlon
of Chemicals Added
"^-Baffle Plate
'\C*—-Reaction Chamber
fE5S^^&S&*£i££*it££££i&f£
Type 316 Stainless Steel Nozzle
In. Pump Discharge
Spray Cone \
Angle 100 deg N
Approximately
Figure 11-4. TYPICAL FLASH MIXING FACILITY
11-68
-------
O)
X
c
o
*rt
oJ
O
^u
-------
SYNCHRONOUS
MOTOR
LOOSE FITTING PISTON OF
INSULATING PLASTIC
-SAMPLE RESERVOIR
SILVER
ELECTRODES AMPL H™,
17
CUTAWAY OF CELL BODY AND PUMP
BORE OF INSULATING PLASTIC
SYNCHRONOUS
RECTIFIER
Figure 11-6. SIMPLIFIED DIAGRAM OF SCO INSTRUMENT
11-70
-------
Stator beam
litators Rotors; stators shown only in lower half
not shown
Horizontal
shaft
Direction of
displacement
'Rotors
^ Rotors
and
stators
X"
fe
^
•9 «• n
\ '
1
1 N"
1 ^LJtJ -^*JP"
/
Longitudinal section;
stators not shown In upper half
a) Paddle type with rotors and stators.
Cross-section
of channel
•Stators
b) Plate turbine type.
c) Axial flow propeller type with straightening vanes.
Figure 11-7. TYPICAL FLOCCULATION UNITS
11-71
-------
Sitfln
X.
I
/ /
x
Side View
Bottom Viiw
Battles
Sidt Vltw
Bottom Vta*
d) Schematic of radial flow
pattern in baffled tank.
e) Schematic of axial flow
pattern in baffled tank.
Figure il-7 (CONTJ. -TYPICAL FLOCCULATION UNITS
, 11-72
-------
Axial Flow
Turbine
Prop«ll«r
Turbine
f) Basic impeller styles.
g) Walking beam flocculator.
Figure 11-7 (Cont.). TYPICAL FLOCCULATION UNITS
11-73
-------
Figure 11-8. A PERFORATED BAFFLE WITH SQUARE-NOZZLED PORTS
DISTRIBUTES FLOW FROM THE FLOCCULATORS AT
RIO DE JANEIRO'S GUANDU TREATMENT PLANT
1A'..;*<•."•
11-74
-------
Figure 11-9. NEW- AND OLD-STYLE RQCCULATORS
11-75
-------
PERFORATED
ENTRANCE
BAFFLE
APPLY 0.25 mg/L
PRODUCT B (985 N}
TYPICAL BAFFLE WOULD BE
THE SAME ELEVATION AS
THE PORT EXTENDING 1.22m
(4 ft) INTO THE COMPARTMENT
TYPICAL WOOD WELL
WITH PORT IN
TOP OF WALL
TYPICAL WOOD WELL
WITH PORT
IN BOTTOM
Figure 11-10. COMPARTMENT PLAN FOR ONE FLOCCULATION UNIT
11-76
-------
FloeovMor ourfel1 Mm r4ft> x 1,33in 4t|i
L««r li
compartment
Stillin
2.4m 1811: X 3m .10(1
Figure H-11a. FLOCCULATOR OUTLET SYSTEM AND
SETTLING BASIN INLET SYSTEM (SIDE VIEW)
TolBlot80riol«»«chlScmi«(n )indl«m»t«r
I? hula*
Flap gtttt
Figure ll-11b. FLOCCULATOR OUTLET SYSTEM AND SETTLING
BASIN INLET SYSTEM (END VIEW)
11-77
-------
ac»nic went
OISPC*SION
\
N
\
Tl
&
•FlttCO
fnoM
— TROUGH
£• _,_ -1- i.__^._ ^r- , +. = _ ^
1- •* •• -t' ~T~" "" "^^""^tit""" ~..."..." ——*
•''/'.''/'/'f'/'M1 '';/'/^///%''/'v//v/yl,^/^^J//'///////////'/////'/'//////////
^^WA«%f^^fe?i»M^
SLuoee ;.
HOSt ^
• ruee semen
tiooui.es
raacron
fife
AM IVffLf
•
con nun.
-rrrn
x - *• •» I
Figure 11-12. SECTIONAL-VIEW THROUGH SEDIMENTATION BASIN
SHOWING VACUUM SLUDGE REMOVAL SYSTEM
11-78
-------
Figure IM3. CLOSE-UP VIEW OF VACUUM SLUDGE REMOVAL ASSEMBLY
11-79
-------
Figure 11-14. A MODULE OF STEEPLY INCLINED TUBES
(COURTESY OF MICROFLOC PRODUCTS
GROUP, JOHNSON DIVISION, UOP)
11-80
-------
FLOW DISTRIBUTION ORIFICES
OVERFLOW BOX
OVERFLOW
(EFFLUENT)
DISCHARGE FLUMES
FEED BOX
FEED (INFLUENT)'
SLUDGE HOPPER
(REMOVABLE)
UNDERFLOW (SLUDGE)
Figure 11-15. LAMELLA*'SEPARATOR
(COURTESY OF PARKSON CORP.)
11-81
-------
INFLUENT ZONE
EXISTING LAUNDERS NEW LAUNDERS
NEW BAFFLE
jr
^ »HFLOW PATTERN)-*—
/"^ /^ -J**-
_ ^ ,„,, _
f///*t fS /f I\\\ SV X *xx \?, // .' - - ,
-j-^±^J-
Hf
'*
TUBE
SUPPORTS
TUBE
MODULES
Figure 11-16. TYPICAL TUBE SETTLER INSTALLATION
IN RECTANGULAR BASIN
11-82
-------
<£>
UJ
_l
LL
LL
<
CD
Z
O
p
1
«vj
3
Q
O
2
O
z
_j
H
UJ
co
U
_l
-1
IU
$
O
LU
H 1
L^
cc
UJ
H
Z
LU
/•*
CC
111
Q
Q
_l >
<
£t $
25
(T -1
UJ <
H 5
X
in <>a
\
\^
1 [
^sL
Ill
X
O
<
QC
CO
oc
O
O
z
<
Q
O
CO
LU
S 09
< OC
oc u
w
Qd
UJ
U.
-J
U
|
0
z
z
O
<
_j
<
z
at:
LU
H
H
UJ
UJ
00
cu
L.
3
00
11-83
-------
Q.
o
1-
u
Qi
D
U
Q
CO
^
i-
C
O
U
D
oc
11-84
-------
TUBE SETTLING MODULES (TYPICAL
SEGMENTS)
TUBE MODULE
SUPPORT FRAME
LAUNDERS NOT SHOWN
TO PROVIDE CLARITY
NEW VARIABLE SPEED
^CIRCULATION PUMP
PLAN
Figure H-18. SOLIDS CONTACT CLARIFIER WITH TUBE SETTLERS
11-85
-------
VARIABLE SPEED
REG IRCULATION PUMP
TUBE SETTLING MODULES
COLLECTOR DRIVE
SLUDGE COLLECTOR
•SLUDGE WASTING SUMP
SLUDGE SLOWDOWN
SECTIONAL ELEVATION
Figure 11-18 (ContJ. SOLIDS CONTACT CLARIFIER WITH TUBE SETTLERS
11-86
-------
U
Z
LJJ
cn
LU
QC.
o
U
«_^
at
LU
U
UJ
CD
D
h-
U
<
O
u
tn
O
6
(J
<
^
U
<
ex.
00
11-87
-------
AVAILABLE
HEAD
LOSS
1
••«
WASH
TROUGH -7
LOW — —
WATER LEVEL
MEDIA
UNDERDRAIN
-1_¥ 1
mm^
_1_
~~
INFLUENT VALVE OR
GATE TO EACH FILTER
COMMON INFLUENT
HEADER PIPE OR
CHANNEL
WASTE
f-ijWAbll
BACK-
MASH
DRAIN
RATE INDICATION
AND LIMITATIONS'
J
CLEARWELL
Figure 11-20. VARIABLE DECLINING RATE FILTRATION (J.AWWA)
11-88 :
-------
~>
FILTER
DRIVING
FORCE
CONSTANT
CONSTANT
PRESSURE
FILTRATION
DECLINING RATE
FILTRATION
^-CONSTANT RATE
FILTRATION
DRIVING
FORCE
ONSTANT
AUTOMATIC
OR MANUAL FLOW
CONTROL VALVE
TIME OF FILTRATION
Figure 11-21. TYPICAL RATE OF FILTRATION PATTERNS DURING
A FILTER RUN (BAUMAN AND OULMAN, 1970)
11-89
-------
EFFLUENT
TURBIDITY
MAXIMUM
HEADLOSS
HEADLOSS
FILTER RUN TIME
TURBIDITY
BREAKTHROUGH
Figure 11-22: HYPOTHETICAL OPTIMUM FILTER HEADLOSS/TURBOTY
; BREAKTHROUGH CURVE .
11-90
-------
RAW RIVER WATER
PLUS ALUM
PLANT FILTER INFLUENT WATER
(FOLLOWING ALUM ADDITION, MIX-
ING, FLOCCULATION, AND SETTLING)
-ALTERNATE INFLUENT
v SOURCES—-^
¥
NOTE: FINE FILTER MEDIA
DEPTH IS 24 INCHES
THROUGHOUT
!
PILOT FILTER/""^ INFLUENT
^ ^
TRI -MIXED
MEDIA NO. 1
TRI-MIXED
MEDIA NO. 2
SAND
WATER
DUAL- MIXED
MEDIA
ANTHRACITE-
CAPPED
SAND
*
,
-
_^" _J
i
•
'
i
'
\
I
i
MULTIPOINT TURBIDITY RECORDER
J
}
PILOT FILTERS
EFFLUENT RATE' OF
FLOW CONTROLLERS
AND METERS
TURBIDIMETERS
Figure 11-23. FLOW DIAGRAM OF THE PILOT FILTRATION EQUIPMENT
11-91
-------
Figure lt-24. CLARIFIERS SHOWING TUBE SETTLING MODULES
11-92
-------
REFERENCES
1. Chadik, Paul A., and Gary I. Amy, "Removing Trihalomethane Precursors From
Various Natural Waters By Metal Coagulants," J.AWA, 75:532-536, 1983.
2. Ruehrwein, R.A., and D.W. Ward, "Mechanisms of Clay Aggregation By
Polyelectrolytes," Sell Science, 673:485, 1952.
3. O'Melia, C.R., "Coagulation and Flocculation," in Physlochemical
Processes for Mater Quality Control, by Walter J. Weber, Jr., Wiley-
Interscience, New York,.1972.
4. Hahn, H.H., and W. Stumm, "Kinetics of Coagulation With Hydrolyzed
Al(IIIj," J. Colloid Interface Scl., 28:134-144, 1968.
5. Vrale, Lassi, and Roger M. Jorden, "Rapid Mixing in Water Treatment,"
J.AWA, 63:52-58, 1971.
6. Chao, Junn-Ling, and Brian G. Stone, "Initial Mixing by Jet Injection
Blending," J.AWA, 71:570-573, 1979.
7. Hudson, H.E., and J.P. Wolfner, "Design of Mixing and Flocculation Basins,"
J.AWA, 59:1257-1267, 1967.
8. Bowers, A; Eugene, and James D. Beard II, "New Concepts in Filtration Plant
Design and Rehabilitation," J.ANUA, 73:9:457, 1981.
9. Forbes, Robert E., Gary L. Nickerson, E. Herbert Hudson, Jr., and Edmund G.
Wagner, "Upgrading Water Treatment Plants: An Alternative to New
Construction," J.AWWA, 72:5:254, 1980.
10. "Upgrading Study, Potomac Water Filtration Plant," Washington Suburban
Sanitary Comm., Hyattsville, MD, September 1978.
11-93
-------
11. Hudson, H.E., Jr., "Dynamics of Mixing and Flocculation," in: Proceedings
of the 18th Annual Public Mater Supply Conference, University of Illinois,
Oept of Civil Engineering. *•
12. Kawamura, Susumu, "Coagulation Considerations," J.AWHA, 65:6:417-423,
1973.
13. Kawamura, S., "Considerations on Improving Flocculation," J.ANWA,
68:328-336, 1976. :
14. Brodeur, Timothy P., "Upgrading to Increase Treatment Capacity,"
J.AHHA, 73:9:464, 1981.
15. Tate, C.H., and R.R. Trussell, "Optimization of Turbidity Removal by Use of
Particle Counting for Developing Plant Design Criteria," in: Proceedings
of AWWA Technology Conference, San Diego, CA, p. I, December 1976.
16. Schleppenbach, F.X., "Water Filtration at Duluth." EPA-600/2-84-083, report
prepared, by the Duluth Water and Gas Department under Grant No. S-804221
from U.S. Environmental Protection Agency, MERL, Cincinnati, OH, April
1984. NTIS: PB-84-177807.
17. Kavanaugh, M.C., C.H. Tate, A.R. Trussell, R.R. Trussell, and Gordon
Treweek, "Use of Particle Size Distribution Measurements for Selection and
Control of Solid/Liquid Separation Processes," in: Particles In Hater,
M.C. Kavanaugh and 0.0. Leckie, Eds., Advances in Chemistry Series 189,
American Chemical Society, Washington, D.C., 1980.
18. James M. Montgomery, Consulting Engineers, Inc., Hater Treatment
Principles and Design, John Wiley & Sons, Inc., 1985.
19. Walker, J.D., "High Energy Flocculation," J.AWHA, 60:1271-1279, 1968.
20. King, Richard W., and Eugene I. Crossley, "Upgrading Water Treatment Plants
in San Diego," J.AHWA, 73:9:476, 1981.
11-94
-------
21. Yapijakis, Constantine, "Direct Filtration: Polymer 1n Backwash Serves Dual
Purpose," J.AWIA, 74:8:426, 1982.
22. Sequelra, James A., Lee Harry, Sigurd P. Hansen, and Russell L. Culp,
"Pilot Filtration Tests at the American River Mater Treatment Plant,"
Public Works, 114:1:36, 1983.
23. Westerhoff, Garrett P., "Experiences with Higher Filtration Rates,"
J.AWWA, 63:6:376, 1971.
11-95
-------
-------
Section III
New Treatment Technology
-------
-------
SECTION III
NEW TREATMENT TECHNOLOGY
INTRODUCTION
This section describes new treatment technology that could be regarded as estab-
lished or emerging. Package plants, slow-sand filtration, and diatomaceous earth
filtration are regarded to be new technologies, especially as they relate to
small systems. Membrane and cartridge filtration, because they have not been used
in treatment of raw water supplies for potable use, are defined as emerging
technologies.
Each technology is described and cost estimates provided. Case histories and
performance data from full-scale installations, where available, is also
presented.
Following presentation of the specific treatment technologies, the use of pilot
studies to determine if a process is appropriate for a certain application and
development of design criteria is discussed. Also, case studies involving use of
pilot studies to determine suitability of particular processes for new treatment
facilities are also described. These studies trace the development of a project
from initial feasibility evaluations through pilot studies to qualify a partic-
ular process for design.
PACKAGE PLANTS
General
The package plant is defined as a factory-assembled, skid-mounted unit generally
incorporating a single tank or, at the most, several tanks. A complete treatment
process typically consisting of chemical coagulation, flocculation, settling and
filtration is used in many designs.
III-l
-------
Package plants are most widely used to treat surface supplies for removal of
turbidity, color and coliform organisms prior to disinfection. Presently, it is
estimated that over 650 package plants, ranging in capacity from 5 gpm to 4 mgd,
are in service in the United States. Package plants can be used to treat water
supplies for communities as well as recreational areas, state parks, construction
camps, ski resorts, remote military installations and other locations where pota-
ble water is not available from a municipal supply. Several state agencies have
mounted package plants on trailers for emergency water treatment. Their compact
size, low cost, minimal installation requirements, and.ability to operate virtu-
ally unattended makes them an attractive option in locations where revenues are
not sufficient to pay for a full-time operator.
Types Available
Conventional Package Plants—
Package water treatment plants are available from several manufacturers in a wide
range of capacity, incorporating a complete treatment process (coagulation, floc-
culation, settling, and filtration). Design criteria used for these package
plants varies widely. Some manufacturers adhere closely to accepted conventional
design practices such as 20- to 30-minute flocculation detention time, a 2-hour
sedimentation detention time and rapid sand filters rated at 2 gpm/sq ft. Other
manufacturers have utilized new technology including tube settlers and high-rate
dual- and mixed-media filters to reduce the size of a plant and hence extend the
capacity range of single factory-assembled units. Often, state regulatory agen-
cies dictate the design criteria which must be met by package plant manufacturers
and exclude units using new technology that does not meet arbitrary standards
adopted by these agencies.
Tube Type Clarification Package Plants—
A flow diagram for a package plant incorporating tube settlers is shown in Figure
III-l. Plants of this type are manufactured in two versions: Microfloc Products
"Water Boy" with a treatment capacity range of 10 to 100 gpm; and Microfloc
Products "Aquarius," which generally consists of dual units with a capacity of
200 to 1,400 gpm. The coagulant and disinfectant chemicals are added at the
influent control valve. A polyelectrolyte coagulant aid is applied as the water
enters the flash mix chamber. After the treatment chemicals are added and mixed,
III-2
-------
the water is introduced into a mechanical flocculator. Flocculation detention
time can vary from 10 minutes in small units to 20 minutes in larger units. The
flocculated water is then distributed through a bank of tube settlers, which
consist of many 1-inch-deep, 39-inch-long split-hexagonal-shaped passageways that
provide an overflow rate, related to available settling surface area of less than
150 gpd/sq ft. This overflow rate, together with a settling depth of only 1 inch,
results in effective removal of flocculated turbidity with a detention time of
less than 15 minutes.
After passing through the tube settlers, the clarified water flows to a gravity
mixed-media filter. The filters are designed to operate at a constant flow rate.
Rate control is accomplished with a low-head filter effluent transfer pump dis-
charging through a float-operated valve. The package plant filter is designed to
backwash automatically once a preset filter headloss is reached. The operator may
override the automatic controls and backwash the filter manually. During back-
wash, the material accumulated in the tube settlers is automatically drained from
the unit. Combining backwashing with draining of the tube settlers for sludge
removal eliminates the need for an operator to judge how often or how much sludge
should be wasted from the clarifier. This particular feature simplifies operation
and reduces the required skill level of the operator.
Figure IH-2 is a flow diagram of a package treatment plant of slightly different
design. Again, low-detention time, high-efficiency- tube settlers are used to
reduce the detention time and hence the volume of the clarification portion of
the plant. A dual-media filter rated at 4 gpm/sq ft also contributes to the com-
pactness of this plant.
Adsorption Clarifier Package Plant—
A package plant manufacturer introduced a new concept in package water treatment
plant (Trident by Microfloc Products) design in the early 1980s that utilized an
upflow filter of low-density plastic bead media (termed an adsorption clarifier),
followed by a mixed-media filter for final polishing. The adsorption clarifier
replaces the flocculation and settling processes and results in an extremely
compact unit. The manufacturer claims that it further reduces chemical coagulant
dosage requirements over other systems. Figure III-3 is a flow diagram of the
Trident Water System manufactured by Microfloc Products illustrating the various
III-3
-------
operating cycles. During operation, chemically coagulated water is introduced
into the bottom of the adsorption clarifier compartment where it passes upwards
through a bed of buoyant adsorption media. The adsorption clarifier combines the
processes of coagulation, flocculation, and settling into one unit process.
In passing through the adsorption media, the chemically coagulated water is sub-
jected to: {1} mixing, (2) contact flocculation, and (3) clarification. At opera-
ting flow rates, the mixing intensity defined by the mean temporal velocity gra-
dient value, 6, ranges from 150 to 300 sec-1. Flocculation is accomplished by
turbulence as water passes through the adsorption media and is enhanced by con-
tact with flocculated solids attached to the media.
Turbidity removal in.the adsorption clarifiers is accomplished by adsorption of
the coagulated, flocculated solids on the surfaces of the adsorption media and on
previously attached solids. The adsorption clarifier provides excellent pretreat-
ment, which frequently is better than the performance achievable with complete
flocculation and settling processes. Turbidity removal in this stage ranges up to
95 percent.
Cleaning of the adsorption clarifier is accomplished by flushing. This flush
cycle is initiated by a timer, but the equipment also includes a pressure switch
that monitors headloss across the adsorption media and can automatically initiate
a flushing cycle if required. Figure III-3 illustrates the operation of the unit
during a flushing cycle. When a cycle is initiated, the plant effluent valve
closes, causing the water level to rise in the plant as the influent flow contin-
ues. When the water level reaches a predetermined level, a switch causes the
influent valve to close. Air is distributed through perforated laterals beneath
the adsorption media. This causes an immediate expansion in the adsorption media
and a vigorous scrubbing action takes place. Dislodged solids are then hydrauli-
cally flushed out of the top of the adsorption clarifier to waste. Influent
water is used to flush the adsorption clarifier. Flushing frequency may vary,
depending upon influent water quality. Typically, the controls are set up to
initiate a flushing cycle every 4 to 8 hours. Unlike conventional filters, com-
plete cleaning of the adsorption clarifier is not required, as the majority of
solids are removed by the violent agitation provided during the first minutes of
III-4
-------
the flush cycle. Also, more efficient performance of the adsorption clarifier
occurs if some residual solids are left on the media.
The mixed-media filter is backwashed in a manner similar to a conventional fil-
ter. Although the filter may not necessarily be backwashed each time the adsorp-
tion clarifier is flushed, the equipment is designed to ensure that a backwash
cycle is always preceded by a flushing cycle. The backwash cycle is illustrated
in Figure 1II-3.
Extensive pilot evaluations of the adsorption clarifier package plant design have
been conducted. Table III-l provides a summary of test results obtained at three
sites. More than 40 installations of these type plants are presently in service.
The package plant is available in four models with capacities ranging from 350 to
4,200 gpm. Figure III-4 is a pictorial cut-away view of this type package plant.
Application Criteria and Requirements
Before selecting a package plant for a particular application, it must be deter-
mined that it can produce the required quality and quantity of water from the
proposed raw water supply. Package plants characteristically have limitations
(especially those employing high-rate unit processes) related to the quality
limitations of the raw water supply, which must be recognized. For example, such
factors as low raw water temperature, high or flashy turbidity, excessive color,
or atypical coagulant dosages (higher than expected based upon normal, turbidity
levels) may influence the selection and rating of a. particular package plant.
The manufacturer's nameplate capacity of a package plant may have to be downrated
or a larger unit selected to handle difficult treatment conditions. Water sup-
plies of consistently high turbidity (greater than 200 NTU) may require presedi-
mentation prior to treatment in a package plant.
It is recommended that all records of raw water quality be reviewed to determine
the full range of treatment conditions to be expected before a particular capa-
city package plant is selected. Especially valuable are laboratory analyses of
representative raw water supplies to provide information critical to a proper
application. Under certain conditions, on-site pilot tests may be justified and
III-5
-------
warranted to verify the suitability of a package plant. This is especially impor-
tant because many of the new package plant designs employ high-rate, short-
detention time unit processes which require close control in order to perform
effectively. Advance information on the quality of the proposed raw water supply
and its treatment characteristics helps to ensure a successful installation.
Operational Considerations
Equally important to the success of package plant applications is the quality of
plant operation. Regardless of the size of the facility, if operating personnel
do not possess an adequate understanding of the process and equipment they are
responsible for operating, production of a safe and palatable finished water may
be a hit-or-miss proposition. Some manufacturers have incorporated automatic con-
trols such as effluent turbidity monitors that shut down the plant when the tur-
bidity of the filtered water exceeds a preset limit. This fail-safe device
(assuming that it is not bypassed by the operator) ensures that if the plant
produces any water at all, that it meets a given turbidity standard. An effluent
turbidity control device should be mandatory on all package plants, especially if
the raw water source is contaminated.
Package plants from several manufacturers have an accessory to automate the chem-
ical feed system to maintain a specified finished water turbidity. This is
advantageous where plants do not have full-time operators and raw water condi-
tions change frequently.
No matter what control systems are used, the burden of producing a safe palatable
water supply rests with the operator. In this regard, there is no substitute for
a well-trained operator who has the necessary skills and dedication to operate
the equipment properly.
Most package plant manufacturers' equipment manuals include at least brief sec-
tions on operating principles, methods for establishing proper chemical dosages,
instructions for operating the equipment and troubleshooting guides. An individ-
ual who studies these basic instructions and receives a comprehensive start-up
and operator training session from the manufacturer's start-up technicians should
be able to operate the equipment satisfactorily.
III-6
-------
The engineer designing a package plant facility should specify that start-up and
training services be provided by the manufacturer, and also should consider
requiring the manufacturer to visit the plant at 6-month and 1-year intervals
after start-up to adjust the equipment, review operations, and retrain operating
personnel. Further, this program should be ongoing and funds should be budgeted
every year for at least one revisit by the package plant manufacturer.
e" Cost of Package Conventional Complete Treatment-
Costs for use of package water treatment plants are developed for surface water
treatment plants in Categories 1 through 6. These are presented in Table III-2.
These units include coagulation, flocculation, sedimentation, and filtration, but
exclude raw water pumping, clearwell storage, and finished water pumping.
Conceptual Design—Cost estimates are for standard manufactured units incorpor-
ating 20 minutes of flocculation, tube settlers rated, at 150 gpd/sq ft, mixed-
media filters rated'at 5 gpm/sq ft, and a media depth of 30 inches. The costs
include premanufactured treatment plant components, mixed media, chemical feed
facilities {storage tanks and feed pumps), flow measurement and control devices,
pneumatic air supply {for plants of 200 gpm and larger) for valve and instrument
operation, effluent and backwash pumps, and all necessary controls for a complete
and- operable unit and building. Smaller plants (below 50 gpm) utilize low-head
filter effluent transfer pumps and are used with an above-grade clearwell. Larger
plants gravity discharge to a below-grade clearwell.
Operation and Maintenance Requirements—Process energy requirements are for
chemical mixers, chemical feed pumps, mechanical flocculators, backwash pumping,
and where applicable, surface wash pumping.
Building energy is required for heating, ventilating and- lighting, with the
latter assumed to be 3 hours/day. Maintenance materials are for replacement of
mechanical and electrical components that wear out or break down during normal
*
operation.
<*•
Chemical Requirements—Chemicals most commonly used in package conventional
plants include alum, polymer, chlorine, and soda ash.
III-7
-------
Conventional Package Plant Performance - Case Histories
EPA Survey of Six Plants—
The widespread acceptance of package plants as an economical solution to water]
treatment needs of small systems has resulted in construction of a significant
number of plants during the 1970s. The quality of water produced by these plants
is of concern, and led to an on-site investigation at six selected facilities.1
The six selected plants were in year-round operation, used surface water sources,
and served small populations. Plants were monitored to assess the performance and
ability to supply water meeting the interim primary drinking water regulations.
At each facility, grab samples of the raw water, treated water and water from the
distribution system were collected intermittently over a 2-year period and ana-
lyzed. Data on effluent turbidity, total coliforms, and chlorine residuals were
recorded on all visits. Table III-3 is a description of the results from six
water treatment plants surveyed during the study. Only three of the plants (C, T,
and W) consistently met the 1 NTU effluent standard. Three of these plants
obtained their raw water from a relatively high quality source. The other three
plants (P, V, and R) met the turbidity standard on fewer than one-half the
visits. Table III-4 is a listing of the treatment process characteristics for the
six plants surveyed. Table III-5 compares raw to filtered effluent turbidities
measured during the visitations.
According to the authors of the survey, the performance difficulties of plants P,
V, and R were related to the short detention time inherent in the design of the
treatment units, the lack of skilled operators with sufficient time to devote to
operating the treatment facilities, and {in the cases of plants V and R) the
wide-ranging variability and quality of the raw water source. It is reported that
the raw water turbidity at the site of plant V often exceeded 100 NTU. Later,
improvement in operational techniques and methods resulted in substantial
improvement in effluent quality. After the adjustments were made, plant filters
were capable of producing a filtered water with turbidities less than 1 NTU even
when influent turbidities increased from 17 to 100 NTU within a 2-hour period.
III-8
-------
To Illustrate the importance of raw water quality knowledge prior to the applica-
tion of a package plant, plant M was subjected to raw water turbidities ranging
from 3.5 to greater than 500 NTU for short durations. It was specifically stated
by the manufacturer's representative and supported by the product literature that
a water with turbidity this high could not be treated satisfactorily by the plant
selected at plant site M.
Even though the raw water supply for plant site P was a reservoir where raw water
turbidities seldom exceeded 17 NTU, filtered effluent exceeded the 1 NTU standard
on six of the nine samples taken during the visitation. However, at this site,
equipment was not the source of the problem in obtaining a satisfactory filtered
water quality. The poor record was traced to lack of adequate operator attention.
One of the major conclusions of this survey was that package water treatment
plants with competent operators can consistently remove turbidity and bacteria
from surface waters of a fairly uniform quality. Package plants applied where raw
water turbidities; are variable require a high degree of operational skill and
nearly constant attention by the operators. Further, it was pointed out that
regardless of the quality of the raw water source, all package plants require a
minimum level of maintenance and operational skill if they are to produce satis-
factory water quality.
Adsorption Clarifier Package Plants - Case Histories
The performance record of package plants has improved in recent years, perhaps
due to better equipment, more skilled operators, greater surveillance by regula-
tory personnel, or a combination of these factors. The adsorption clarifier pack-
age plant (Trident), equipped with automatic chemical dosage control appears to
provide consistently higher quality finished water meeting current standards in
almost all instances. Following are operating data from several of these
installations:
Greenfield, Iowa—
This 1.0 mgd plant has been in operation since May 1984.2 It treats a surface
water from Lake Greenfield. The water contains turbidity, taste, odor, and iron
III-9
-------
compounds. The plant Is equipped with powdered activated carbon feed facilities
that are used when taste and odor are a problem. It is fed to the water as a
slurry in the influent line to the treatment plant.
Operating data are summarized in Table 111-6 for the period from July through
December, 1984. Plant operation reflects well the results that were obtained in
pilot studies conducted previously at the site. Alum dosages are somewhat lower
than used in the pilot study, at 7 to 20 mg/1. Filtered effluent quality is typi-
cally in the range of 0.3 to 0.5 NTU, similar to the pilot results.
Lewisburg, West Virginia—
This 2.0 mgd facility started up in December 1983 as the second Trident system
on-line. Records of turbidity removals through the system for three months are
summarized in Table III-7. Turbidity removal through the clarifier is generally
over 90 percent.
Philomath, Oregon—
This 1.0 mgd facility treats a surface water supply, and started up in February
1986. A summary of operating data is shown in Table III-8.
Harrisburg, Pennsylvania—
The Dauphin Consolidated Water Supply Company, owned .by the General Waterworks
Corporation, operates the 6th Street Plant in Harrisburg.3 The original plant was
constructed in 1964, and included a circular upflow sludge blanket clarifier and
stored backwash gravity filters. The filters were subsequently converted to dual
media. Due to the limited capacity of the clarifier, the filters were operated in
the direct filtration mode at a rate of about 4 gpm/sq ft. In a recent expansion
to 9.0 mgd capacity, the company elected, after pilot studies, to install three
adsorption clarifier units for clarification ahead of the existing filters. These
units operate at variable rates between 10 and 15 gpm/sq ft, as shown in the
summary of three months data shown in Table 111-9.
The operating results have been quite satisfactory at this plant, Turbidity
removal through the adsorption clarifiers has been in the range of 70 to 90 per-
cent. The filter run lengths have improved to about 26 hours, from the 13-hour
111-10
-------
length experienced prior to the installation of the adsorption clarifiers, in
spite of flow increases of 20 to 30 percent.
Red Lodge, Montana—
The plant at Red Lodge was started up in January 1984. It is rated at 1.4 mgd.
This was a new treatment facility on a previously unfiltered surface water
supply. The raw water quality is quite high, as can be seen in Table 111-10. Raw
water turbidity levels are commonly below 1 NTU. To aid in treatment of this low
turbidity cold water, the plant is equipped with bentonite feed facilities. This
material is fed into the influent line, just ahead of alum and polymer feed
points. The plant is also equipped with automatic coagulant feed and pH
controls.
The system is routinely producing water with turbidity less than 0.1 NTU, consid-
ered to be a goal to minimize the potential for Giardia contamination. Note that
the clarifier effluent turbidity is sometimes higher than the raw water turbid-
ity. This is explained by the fact that the raw water measurement does not
include the impact of the bentonite and alum feed.
SLOW-SAND FILTRATION
Recently there has been an increased interest in slow-sand filters. This interest
has been expressed principally in small communities that have a protected surface
watershed and provide only chlorination. The need for multiple barriers has been
demonstrated for protection against giardiasis. Slow-sand filters are attractive
to small water systems since they require little operator attention and no chemi-
cal pretreatment. For proper application of slow-sand filters, the raw water must
be of high quality (less than 10 NTU with no color problem).
Process Description
Slow-sand filters are similar to single-media rapid-rate filters in some
respects, yet they differ in a number of important characteristics. In addition
to the obvious difference of flow rate, slow-sand filters: (1) function using
biological mechanisms instead of physical-chemical mechanisms, (2) use smaller
III-ll
-------
sand particles, (3) are not backwashed, (4) have longer run times between clean-
Ing, and (5) require a ripening period at the beginning of each run.
Design Standards
Design filtration rates vary from 1 to 10 million.gallons :per acre per day (mgad)
with 3 to 6 mgad the usual range. Filter sand depth ranges up to 42 inches.
Cleaning is accomplished by scraping off l,or 2 .inches from the surface. Typi-
cally, once the depth is reduced to 24 inches, new sand is added. The sand has an
effective size of 0.25 to 0.35 mm, and a uniformity coefficient of 2 to 3. The
underdrainage system usually is constructed from split tile, with laterals laid
in coarse stone and discharging into a tile or concrete main drain. Slow-sand
filters constructed recently include perforated PVC pipe for laterals. Slow-sand
filters are operated continuously under submerged media conditions, generally
about 4 to 5 feet. Jo avoid air binding,:the maximum head loss across the filter
should not exceed the submergence. Buildup of head loss to maximum is slow, vary-
ing.from 2 or 3 days to 6 months. Manually adjusted, weirs, outlets, and/or valves
are satisfactory for flow control..
Inlet structures may be located at one end or side of the filter. Careful consid-
eration should be given to preventing scouring or erosion of media during filling
of the filter. Gradual filling of the filter may be. necessary until the depth
above the media is sufficient to .allow maximum inlet rates and no disturbance of
the surface. .
The filtered water outlet is a hydraulic control structure designed to maintain
submergence, of the media under all conditions. The difference in level of water
in the collection gallery and the level above the media represents the headloss
through the media. Figure III-5 is a typical covered slow-sand filter installa-
tion and Figure III-6 illustrates an uncovered slow-sand filter.
The initial loss of head is only -about 0.2 feet. When the headloss reaches about
4 feet, the surface, is usually scraped. The length.of run .-between cleanings is
normally 20 to 60 days.4 With varying combinations of-raw water quality, sand
111-12
-------
size, and filtration rate, however, runs may be shorter or longer than are nor-
mally experienced. Slow-sand filters are cleaned by scraping a surface layer
(generally 2 to 3 inches) of sand and washing the removed sand, or washing the
surface sand in place by a traveling washer. Since slow-sand filters must be
removed from service for extended time periods for cleaning, redundant or standby
filters are needed. When .repeated scrapings of the sand have reduced the depth of
the sand bed to approximately one-half of its design depth, the sand should be
replaced. Filter bed depths of less than 12 to 20 inches have been shown to
result in poor filter performance. The replacement procedure should include
removal of the remaining sand down to the gravel support, the addition of the new
sand to one-half of the design depth, and placement of the sand previously
removed on top of the new sand.
Operating Characteristics
As noted above, slow-sand filters produce poorer quality filtrate at the begin-
ning of a run (right after scraping), and require a filter-to-waste (or ripening)
period of 1 to 2 daysvbefore being used to supply the system.5 A ripening period
is an interval of time immediately after a scraped or resanded filter is put back
on-line, in which the turbidity or particle count results are significantly
higher than the corresponding .values for a control filter. More recent work
indicates that scraping does not significantly affect Giardia removal, as long as
the sand bed has developed a mature microbiological population.6'7
Effluent Quality Requirements
•- i =
The SUTR will require utilities using slow-sand filtration treatment to maintain
an effluent turbidity of less than 1 NTU. If more than 5 percent of the effluent
samples contain greater than 1 NTU turbidity, the system is not in compliance
with this requirement of the SWTR, unless the utility reports the dates and
results of total coliform sampling of the filter effluent prior to disinfection
conducted in the same manner and frequency as the Proposed Total Coliform Rule,
and dates and results of sampling to demonstrate that turbidity does not inter-
fere with total coliform measurements. If a turbidity greater than or equal to 5
NTU is detected at any time, the system is not in compliance.
111-13
-------
Laboratory and Pilot Plant Studies
A 14-month pilot study evaluated the effectiveness of a slow-sand filter to
remove turbidity and coliform bacteria from a surface water supply.5 The pilot
filter contained a 37-inch-deep, 0.32 mm effective size sand bed and was operated
at a filtration rate of 0.05 gpm/sq ft (3 mgad). After.an initial 2-day ripening
period, effluent turbidity was consistently near or below 0.1 NTU at applied
water turbidities which averaged 4 to 5 NTU. Coliform bacteria removal was
always 99.4 percent or better' reaching 100 percent in one filter run. Removal of
Giardia cyst-sized particles (7 to 12 ym size range) averaged 96.8 percent or
better for all test runs. Even though Giardia removal was not measured directly,
the results of this study clearly established the suitability of slow-sand filt-
ration as a viable technology for producing high quality filtered water.5 Other
studies, like those described below, have confirmed that the biological action in
a sand bed adds significantly to Giardia removal, so that higher removal rates
than those found for inorganic particles can be expected.
Treatment efficiency of slow-sand filtration was studied under various design and
operating conditions to ascertain removal of Giardia Iambila cysts, total coli-
form bacteria, standard plate count bacteria, particles, and turbidity.6.7
Filter removals were assessed at hydraulic loading rates of 1 mgad, 3 mgad, and
10 mgad; temperatures of 0°, 5°, and 17°C; effective sand sizes of 0.128, 0.278
and 0.615 ym; sand bed depths of 0.48 and 0.97 mm; influent Giardia cyst concen-
trations of 50 to 5,000 cysts/liter; and various conditions of filter biological
maturity and influent bacteria concentrations.
Results showed that slow-sand filtration is effective in removing microbiological
contaminants. Giardia cyst removal was consistently greater than 99.8 percent for
a biologically mature fully ripened filter. Total and fecal coliform removal was
approximately 99 percent. Particle removal averaged 98 percent. Standard plate
count bacteria removal ranged from negative removals to 99 percent, depending on
the influent concentration. Turbidity displayed a unique ability to pass through
the filters, a characteristic not previously reported, and removal ranged from 0
to 40 percent. It is entirely possible that the particles measured as turbidity
111-14
-------
in this water supply were too small to be captured by the filter or perhaps were
charged such that they were repelled by the filter media and passed through in
the effluent. Some of the turbidity could also be due to debris sloughing off the
filter. Operating results from Waverly, New York, tend to substantiate the former
possibilities.8
Changes in process variables resulted in decreased coliform removal efficiency
for increased hydraulic loading rate, increased sand size, decreased bed depth,
and decreased biological activity. Giardia removal was influenced by the biologi-
cal maturity of the filter, but not by the variables mentioned above. During
filter start-up, Giardia removal was 98 percent; and once the filter was mature,
removal was virtually complete.6*7
Costs of Slow-Sand Filters
Conceptual Design—
The slow-sand filter systems are based on use of a cast-in-place, concrete-
covered structures for use in Categories 1 through 3 and uncovered earthen berm
structures for Categories 4 through 7. Slow-sand filter systems are sized on the
basis of a filtration rate of 70 gpd/sq ft (3 mgad). The depth of the sand bed is
3.5 feet over 1 foot of support gravel. A supernatant box height of 4 feet is
provided above the filter surface.
• v .
Covered filters incorporate below-grade concrete structures with beam and slab
covers. All.* piping is either cast iron or steel and is also below grade. Because
slow-sand filters operate most efficiently at a constant filtration rate, flow-
meters for each filter and a below-grade, concrete clearwater reservoir are
included. The rate of flow through the filter is established by an individual
control valve for each filter in an effluent flow control structure.
Uncovered, slow-sand filters include on-grade, clay-lined structures formed by
earthen berms. All piping is PVC and is below grade. A steel tank reservoir, an
effluent flow control structure, and a flowmeter to measure total plant flow are
also included.
111-15
-------
Capital costs for slow-sand filters conforming to the above design concept are
provided in Table III-ll.
Operation and Maintenance Requirements-
Process energy requirements are negligible. Building energy requirements are for
heating, lighting, and ventilation of the control and storage building.
Maintenance material requirements are based upon anticipated costs of replacement
parts and replenishment of the minor amount of consumable supplies involved in
daily operation. The cost of replacement sand is not included. Experience with
operating filters suggests that careful cleaning allows resanding to be necessary
as infrequently as every 10 years.
Labor requirements are developed assuming the filters operate unattended. They
include a short daily inspection and control valve adjustment, daily turbidity
measuranents, quarterly cleanings of each filter by hand, normal repair and
replacement of equipment. Table III-ll provides a summary of operation and
maintenance costs for the seven design flow categories.
Case Histories
Survey of 27 Plants—
A survey of 27 slow-sand filtration plants in the United States indicated that
most of the plants serve fewer than 10,000 persons, are more than 50 years old,
and are effective and economical to operate.9 Most facilities surveyed used lakes
or reservoirs as raw water sources. Filtration rates ranged from less than 0.3
mgad to about 13 mgad. Fifty percent of the plants have filtration rates in the
2.6 to 6.4 mgad range. Filter media depths fall between 15 and 72 inches. Most
installations use sieved sand with effective sizes averaging 0.2 to 0.4 mm. Sand
uniformity coefficients varied from 1,4 to 3.5.
Figure III-7 presents average raw water turbidities for the slow-sand filter
plants surveyed. About 50 percent of the plants process raw water with an average
turbidity exceeding 2 NTU. All plants surveyed treat raw water with turbidities
of less than 10 to 12 NTU.
II.I-16
-------
'-*•
Figure III-8 shows that 85 percent of the plants1 produce average filtered water
with less than 1 NTU, and 50 percent average less than 0.4 NTU. One plant
reported average filtered effluent turbidities of less than 0.1 NTU. Average
annual coliform concentrations in the raw water are shown in Figure 111-9. Figure
111-10 shows for the three plants reporting that 90 percent of the facilities
maintained effluent coliform levels of 1/100 ml or less.
Filter cycle durations vary seasonally. The mean length of filter cycles ranged
from 60 days in winter to 42 days in spring.
New York State—
A study was performed at seven slow-sand filter installations in New York to
assess the impact on filtered water quality of the, cleaning procedure which
involves scraping a thin layer of sand from the surface of the filter.8 The
performance of each filter was monitored before and after the scraping procedure.
Effluent samples were analyzed for turbidity, total particle count, standard
plate count, and total coliforms. Table 111-12 lists the characteristic features
of the seven facilities.
Table 111-13 is a summary of the data collected during plant tests designed to
determine the impact of filter scraping on filtered water quality during the
ripening period.
It is noted in Table 111-13, that even during the ripening period, filtered
effluent turbidities at all but the Waverly installation did not exceed 0.43 NTU.
At the Waverly facility, it appears that the raw water contains submicron par-
ticles which scatter light and increase the turbidity, but are not efficiently
removed by slow-sand filtration. According to the particle count data, Waverly
removes particles larger than 2 pm as efficiently as the other plants visited.
Turbidity data from the control filters which have been in service for at least
one month, typifies the performance, of these facilities.
The study included a comprehensive evaluation of total .particle count, standard
plate count, and coliform removal from both the ripening filter and a control
LI 1-17
-------
filter. In general, removal of total particles in the size range 2 ym to 60 ym
measured by a total particle counter, ranged from 90 to 99.8 percent. Filtration
generally reduced the total plate count, however, results were widely variable.
Removals ranged from greater than 50 percent to no removals. Coliform removals
through the ripening and control filter were only measured at two facilities. One
facility achieved 98.5 percent removal of coliform organisms after completion of
the ripening period..
Mclndoe Falls, Vermont—
A 2-year study at Mclndoe Falls, Vermont, was conducted to evaluate the effec-
tiveness of a slow-sand filter for removal of turbidity, bacteria, coliforms, and
Giardia cysts.10 The community obtains its water from two impoundments which
capture water from two springs. The ponds are shallow and contain several beaver
dams and lodges. Because of the presence of beavers, giardiasis was a major con-
cern. Raw water quality is generally quite good with turbidities ranging from 0.4
to 4.6 NTU with a seasonal average of 2.1 NTU.
The two filters have an individual area of 400 square feet. They contain 42
inches of 0.33 mm effective size silica sand and have a design filtration rate of
2.05 mgad. During the evaluation, the following results were achieved:
t Removed turbidity to less than 1 NTU, 99.19 percent of the time (after the
first 100 days of operation, .the effluent values were below 1 NTU, 99.68
percent of the time) and 72 percent of the time the values were 0.2 NTU or
less (raw water 1.45 NTU or less, 72 percent of the time).
• Reduced total coliform to 10/100 ml or less, 86 percent of the time (raw
water 1300/100 ml or less, 86 percent of the time) and standard plate count
to 10/ml or less, 94 percent of the time (raw water 500/ml or less, 94
percent of the time) under ambient load conditions.
• The average coliform content in the filter effluent under ambient conditions
was 4 organisms/100 ml (raw water 440/100 ml) for total coliform, and 15
heterotrophic organisms per ml (raw water 520/ml) for standard plate count.
111-18
-------
• Removed massive spikes of total coliform and standard plate count bacteria
from raw water at temperature conditions above the range of 5 to 10°C.
t Did not remove bacteria as efficiently at temperatures below 5°C and partic-
ularly around 0 to 1°C. Spiking studies demonstrated the temperature effect
with removal of total coliform deteriorating from 98 to 43 percent, and stan-
dard plate count from 98 to 80 percent in 9 days at 1°C. At 6 to 11DC,
removal remained at an average of 99 percent for total coliform and 97 per-
cent for standard plate count during a 15-day spike.
• Removed Giardia cysts very dependably, 99.98 percent or better, under warm
temperature conditions.
• Did not remove Giardia cysts as completely at low temperatures. Under 7°C,
99.36 to 99.91 percent removal was achieved.
• During cold water conditions (below 5°C and particularly around 0° to 1°C)
the biologic treatment process in the slow-sand filter was less effective in
ranoving bacteria and Giardia cysts. Cyst removal was reduced to 93.7
percent, and total coliform and standard plate count removals dropped to 43
percent and 79 to 82 percent, respectively.
Village of 100 Nile House, British Columbia, Canada-
Outbreaks of waterborne giardiasis occurred at the Village of 100 Mile House,
British Columbia, Canada, in 1981, 1982, and 1983 with 60, 50, and 30 confirmed
cases, respectively, for a service area population of about 2,000 persons.11 The
source was suspected to be beavers and muskrats subsequently confirmed positive
for Giardia cysts and located upstream of the Village's surface water intake. The
addition of chlorine solution, with minimal contact time provided in the distri-
bution system, was the only treatment practiced.
The slow-sand filtration and chlorination processes were chosen after completion
of successful pilot studies by the Village for incorporation in the new water
treatment facilities. Construction on the water treatment plant was completed
111-19
-------
December 1984. A detailed evaluation of the plant was undertaken between November
15, 1985, and November 15, 1986. Of prime concern was whether the slow-sand
filters were effective in removal of Giardia cysts. The plant includes a surface
water intake, raw and treated water pumping stations, chlorine equipment and
contact tank, a clear well, and three slow-sand filters. Each of the three cast-
in-place concrete slow-sand filters are 141-feet-long, 20-feet-wide, and 12 feet
6 inches to 13 feet 3 inches deep (top of wall to top of floor) for a surface
area of 2,820 square feet each cell. The total surface of the three filters is
8,460 square feet. The filters are covered with precast panels and the filter
walls are insulated with rigid insulation and backfilled with soil for full depth
to moderate against cold temperatures.
The filter media was produced by the washing, drying, and sieving of natural
material from one site located in the vicinity of 100 Mile House. The effective
size of filter media, based on samples taken after placement of the materials,
ranges between 0.20 mm and 0.30 mm with an average value of 0.25 mm. The uniform-
ity coefficient ranges between 3.30 and 3.80 with an average value of 3.50. The
depth of media at startup was 45 inches.
The three slow-sand filters are typically operated at rates up to the accepted
maximum unit flow of 640,000 gpd per filter or 1,917,000 gpd for three filters.
The design filtration rate is approximately 0.16 gpm/sq ft or 9.9 mgad.
Operating Results—Removals of a wide range of particles were measured for .the
treatment plant. As illustrated in Figure III-ll, Giardia cysts were detected in
11 of the 21 raw water samples, with counts as high as 300 cysts. Zero cysts were
detected in every sample of the filter effluent.
As illustrated in Figure 111-12, the turbidity in the raw water varied between
0.15 NTU and 3.5 NTU. The filtered water turbidity varied between 0.26 NTU and
1.15 NTU. Generally better reduction was observed in the period June to November
1986 when the raw and filtered water turbidities ranged between 0.4 NTU and 1.17
NTU and 0.15 NTU and 0.62 NTU, respectively. Increased removal seemed to occur
with a maturing filter. Also, an increase in filtered turbidity was noted
immediately after filter cleaning.
111-20
-------
Operating cycles between filter cleanings ranged from a minimum of 52 days to a
maximum of 215 days. The longer cycles were experienced during winter and spring
months. Increased solids loadings during late spring and summer (algae) into
early fall contributed to shorter operating cycles.
Cost—The total construction cost in 1984 for the slow-sand water treatment
plant was $780,000 Canadian. Annual operating costs were determined to be $20,700
including chlorination, power, daily inspection, cleaning, and media replacement.
The plant was inspected every day for 0.5 to 1.0 hour. About 16 person-hours were
required for cleaning of one filter, with a cost of about $225/cleaning. The
total unit cost of water treatment was determined to be $0.96 Canadian per 1,000
gal Ions.
DIATOMACEOUS EARTH FILTRATION
General
Diatomaceous earth filters consist of a layer of diatomaceous earth about 1/8
inch thick supported on a septum or filter element. The thin precoat layer of
diatomaceous earth is subject to cracking and must be supplemented by a continu-
ous body feed of diatomite. If no body feed is added, the particles filtered out
will build up on the surface of the filter cake and cause rapid increased in
headless. Figure 111-13 illustrates a typical pressure diatomaceous earth filtra-
tion system. The problems inherent in maintaining a perfect film of diatomaceous
earth between filtered and unfiltered water have restricted the use of diato-
maceous earth filters for municipal purposes, except under certain favorable
conditions.
An advantage of a diatomaceous earth filtration plant for potable water is the
lower first cost of such a plant. On waters containing low suspended solids, the
diatomaceous earth filter installation cost should be somewhat lower than the
cost of a conventional rapid-sand filtration plant. Diatomaceous earth filters
will thus find application in potable water treatment under the following
conditions:
111-21
-------
• In cases where the diatomaceous earth filtration plant will be found to be
capable of processing the raw water supply and to produce water at a lower
total cost than any practical alternative.
• In cases where financial capacity is tightly circumscribed, when the lower
first cost of a diatomaceous earth filter installation may be the major
factor in the final choice of plant.
• For emergency or standby service at locations experiencing large seasonal
variations in water demand, when the lower first cost of the diatomaceous
earth filter may prove to be economical.
Some of the important operating characteristics of diatomaceous earth filters
have been summarized:12
• A precoat of 0.1 to 0.2 Ib/sq ft is applied to prepare the filter.
• A continuous feed of filter aid as body feed is necessary to prevent the
cake from clogging with the particles being filtered out.
• Acceptable cleaning of the filter will maintain at least 95 percent of the
septum area available for flow after 100 filter run cycles.
• Because of the precoat, diatomaceous earth filters do not require filter-to-
waste upon startup.
• If the flow to the filter is disrupted, the filter cake drops off the sep-
tum. When the filter is restarted, clean diatomaceous earth and filtered
water should be used to recoat the filter to reduce the potential for pass-
age of pathogens.
• It may be necessary to adjust the body feed rate in proportion to the raw
water turbidity to prevent short runs.
111-22
-------
0 Although filter runs can be 2 to 4 days long, decomposition of organic
matter trapped in the filter cake may necessitate shorter runs to avoid
taste and odor problems.
• Vacuum diatomaceous earth filters offer the advantages of not requiring
pressure vessels and being visible during backwash; however, they have the
disadvantage that the increased potential for release of gases, which can
cause shorter runs.
• Diatomaceous earth filters can provide very effective removal of cysts,
algae, and asbestos. In some cases, alum-coating of the diatomaceous earth
improves performance.
• The rate of body feed and size of diatomite used are critical variables
affecting the length of run.
Desiign Criteria
The minimum design criteria presented in the Ten State Standards for diatomaceous
earth filtration are considered sufficient for the purposes of the SWTR with the
following exceptions:
• . It is recommended that the quantity of precoat be increased to 0.2 Ibs/sq ft
of filter area, and that the minimum thickness of the precoat filter cake be
1/8 to 1/5 inch. Studies have shown that a precoat thickness of 0.2 Ibs/sq
ft was most effective in Giardia cyst removal and that the precoat thickness
was more important than the grade size in cyst removal.
• It is recommended that treatment plants be encouraged to provide a coagulant
coating (alum or suitable polymer) of the body feed. Although enhancement of
the diatomaceous earth is not required for Giardia cyst removal, coagulant
coating of the body feed has been found to result in a significant improve-
ment in percent removals for viruses, bacteria and turbidity.
111-23
-------
Operating Requirements
Operating requirements specific to diatomaceous earth filters include:
• Preparation of body feed and precoat
• Verification that dosages are proper
• Periodic backwashing and disposal of spent filter cake
• Periodic inspection of the septum(s) for cleanliness or damage
• Verification that the filter is producing a filtered water that meets the
performance criteria
Effluent Quality Requirements
The SWTR requires that systems using diatomaceous earth filtration maintain an
effluent trubidity of 1 NTU. If more than 5 percent of the monthly effluent
samples contain turbidities greater than 1 NTU, the system is not in compliance
with this requirement. In addition, if any instantaneous filter effluent turbid-
ity exceeds 5 NTU, the system is not in compliance.
Treatment Capability
A study was conducted at Colorado State University to determine the effectiveness
of diatomaceous earth filtration for removal of G i a r d ia cysts.13 In addition,
removals of turbidity, total coliform bacteria, standard plate count bacteria,
and particles were determined. Parameters evaluated included type of diatomaceous
earth, hydraulic loading rates, influent concentrations of bacteria and Giardia
cysts, headloss, run time, temperature, and the use of alum-coated diatomaceous
earth.
Hydraulic loading rates imposed were 1, 2, and 4 gpm/sq ft. Seven grades of dia-
tomaceous earth were used. Temperatures varied from 5° to 19°C; concentrations of
Giardia cysts ranged from 50 to 5,000 cysts/1; and bacteria densities varied from
100 to 10,000/100 ml.
111-24
-------
The results of this study showed that diatomaceous earth filtration 1s an effec-
tive process for water treatment. Glardla cyst removals were greater than 99.9
percent for all grades of diatomaceous earth tested, for hydraulic loading rates
of 1.0 to 4.0 gpm/sq ft, and for all temperatures tested. Percent reduction in
total coliform bacteria, standard plate count bacteria, and turbidity are influ-
enced strongly by the grade of diatomaceous earth. The coarsest grades of diato-
maceous earth recommended for water treatment will remove greater than 99.9 per-
cent of Glardia cysts, 95 percent of cyst-sized particles, 20 to 35 percent of
coliform bacteria, 40 to 70 percent of heterotrophic bacteria, and 12 to 16 per-
cent of the turbidity from Horsetooth Reservoir water. The use of the finest
grade of diatomaceous earth or alum coating on the coarse grades will increase
the effectiveness of the process, resulting in 99.9 percent removals of bacteria
and 98 percent removals of turbidity.
Diatomaceous earth filtration was evaluated at Mclndoe Falls, Vermont, during
parallel studies with slow sand filtration.10 Filtration rates averaged 1 to 1.8
gpm/sq ft on the 10 to 20 gpm pilot pressure unit. The key conclusions from this
study are as follows:
• Pressure diatomaceous earth filtration removed Glardla cysts dependably,
providing 99.97 percent reduction.
• Total coliforms were reduced 86 percent or more in 70 percent of the sam-
ples, and standard plate count bacteria were reduced 80 percent or more in
70 percent of the samples.
• The average bacterial content in the effluent under ambient conditions was
38/100 ml for total coliforms and 6/ml for standard plate count bacteria.
Malina et al, reported that a high percentage of removal could be attained for
pollovirus when coated diatomaceous earth filter aid was used or when cationic
polymer was added to the raw water.114 In one 12-hour filter run, diatomaceous
earth coated with 1 mg of cationic polymer per gram of diatomaceous earth pro-
duced filtered water in which no viruses were recovered from 11 samples (removal
>99,.95%). In another run, one of 12 samples was positive, and in this instance,
111-25
-------
virus removal was 99 percent. In a 12-hour run in which uncoated diatomaceous
earth was used, but 0.14 mg/1 of cationic polymer was added to the raw water, no
viruses were recovered from any of the 12 samples analyzed.
Costs for Diatomaceous Earth Filtration
Costs for diatomaceous earth filtration of surface water in plant Categories 1
through 10 are shown in Table 111-14. Pressure diatomaceous earth filters are
assumed in these samples.
MEMBRANE FILTRATION
Introduction
Membrane or ultrafiltration is a process which utilizes a hollow fiber membrane
to remove undissolved, suspended or emulsified solids from water supplies without
requiring coagulation. There are a variety of hollow fiber membranes of different
size and porosity, and systems wherein the feed water is applied either to the
inside or the outside of the membrane.
The traditional ultrafiltration process applies feed water to the inside and
collects permeate on the outside of the membrane. Systems of this design are most
generally used in specialized applications requiring extremely high purity water,
as pretrestment prior to reverse osmosis or for removal of colloidal silica from
boiler feed water. Application to potable water treatment would be limited to the
production of extremely high quality water supplies of low turbidity (1 NTU or
less) and with Fouling Indexes of less than 10.
In this type of system, the hollow fiber membrane, which operates over a pressure
range of (10 to 100 psi), excludes particles larger than 0.01 v.m; but unlike
reverse osmosis membranes, permits passage of inorganic salts and electrolytes.
Bacteria, Giardja cysts, and some virus particles are removed by the smallest
size hollow fiber membranes. The hollow fiber membranes are contained in a pres-
sure vessel or cartridge. During operation, the flow is introduced to the inside
of the fiber. Filtrate or permeate collects on the outside of the membrane, con-
centrates at the end of the hollow fiber and is discharged to waste. In a typical
111-26
-------
single-pass application, 90 percent of the feed water is recovered as permeate.
The membranes are cleaned by backflushing which restores the original porosity
and allows continuous use for indefinite periods. In some applications a chlorine
solution is added to the backflushing flow.
Hollow fiber membranes are sensitive to the concentration of suspended and col-
loidal solids in the feed water. The flux level {volume of permeate produced per
unit area of membrane per day) and the flux stability are influenced by the feed
water quality, the filtering cycle duration, and the quality of water used for
backflushing the membranes. Where a water supply has a Fouling Index of 10 or
less (typical of most groundwaters and high clarity surface waters), filtering
cycles will range up to 8 hours with about a 10 percent reduction in flux. In
general, capacity can be restored to nearly clean membrane values by a fast-
forward flush of influent to waste. However, after two or three flushings, the
membranes must be backwashed to restore the initial flux level.
To overcome the rapid fouling and loss of flux rate over a relatively short time
period, one manufacturer (Memcor) has introduced a self-cleaning hollow fiber
membrane filtration system that introduces the feed stream tangentially to the
outside of the fiber. The filtrate is collected in the hollow interior of the
fiber and the rejected stream of concentrate is either recycled or discharged.
Continuous recycling permits extremely high filtered .water recovery with the only
water loss from the system being that associated with wasted sludge and cleaning
solution. Recovery rates of 99 percent are claimed by one manufacturer.
The outside-to-inside hollow fiber filtration system uses a more porous (70
percent) membrane with a 0.2 um (nominal) pore size which allows application of
water containing higher suspended solids concentrations. However, it can still
produce a product water meeting present drinking water turbidity standards, will
exclude all Giardia cysts, and can remove coliform and other bacteria. The pore
size of these membranes is. too large to provide effective virus removal.
Membrane Cleaning
In traditional inside-to-outside hollow fiber membranes, cleaning is accomplished
by backflushing, which restores the original porosity and allows continuous use
111-27
-------
for indefinite periods. In some applications a chlorine solution is added to the
backflushing flow.
Fouling has been a major problem, limiting the application of hollow fiber mem-
brane technology. To prevent fouling, one manufacturer has developed a unique,
self-cleaning "radial pulse" system. Gas is periodically injected, under high
pressure, into the center of the hollow fiber. The specially designed membrane
expands as the gas accelerates through the fiber, explosively removing the resid-
ual layer of rejected material. This radial pulse system minimizes flux decay,
eliminates product loss, and does not require auxiliary pumps.
The impact of the radial pulse cleaning procedure on filtration rate is to pro-
long the operation until chemical cleaning of the membrane is required to restore
the original flux rate. Figure 111-14 shows the operating mode and the radial
pulse cleaning sequence.
After several repeated intermittent radial pulse cleaning cycles, chemical clean-
ing of the hollow fiber membrane is needed to restore the original capacity.
This cleaning is done with the aid of a chemical cleaning solution. The cleaning
solution is prepared with a caustic-based detergent and hydrogen peroxide disin-
fectant. Typically, cleaning is practiced once per week or more frequently if the
raw water suspended solids content causes more rapid membrane fouling. In cases
where the supply contains iron, an occasional cleaning with hydrochloric acid may
be necessary. Spent cleaning solution must be disposed of in accordance with
local regulations. -Small quantities could be discharged to a sanitary sewer.
Larger quantities would require accumulation and disposal in accordance with
local regulations for nonhazardous detergents.
Typical Installation
A typical membrane filtration installation for a small water supply system would
include a skid-mounted filtration unit, as illustrated in Figure 111-15, contain-
ing the hollow fiber membranes in cartridges, automatic and manual valves for
backflushing and unit isolation, flowmeters, pressure gauges, integral supply
pump, and control panel. A separate supply pump would be required, as would all
111-28
-------
interconnecting piping serving plants with multiple units. Storage tanks and
chemical feed pumps for membrane cleaning solutions would be needed. Product
water storage with chlorination provisions would also be required.
Table 111-15 provides physical information from one manufacturer on membrane
filtration systems capable of processing up to 300.gpm. The extreme compactness
along with the simplicity of the process makes. it attractive for small
installations. . . *
Figure 111-16 illustrates a complete system including provisions to clarify back-
flush water containing concentrated solids. A small clarifier with facilities to
add a coagulant are needed. Clarified water is either recycled or discharged to
an acceptable disposal location. The concentrated sludge with properties similar
to those experienced in conventional water treatment practice may be dried and
disposed of as landfill or hauled as a semi-solid to an appropriate disposal
site..
Performance Capabilities
Giardia Removal —
An evaluation of the efficiency of the Memcor hollow fiber filtration system for
removal of Giardia lamblia cysts and bacteria was conducted by the Department of
Pathology at Colorado State University.15 The results showed the 0.2 urn membranes
to be 100 percent effective in removing Giardia cysts at applied concentrations
approaching 1,100 cysts/liter. The test also showed that the radial pulse clean-
ing procedure effectively prevented fouling of the membranes and purged cysts
from the system.
Coliform Removal--
A second series of tests were also conducted at Colorado State University to
establish the membrane removal efficiency of E. coli bacteria.15 A tap water sam-
ple was seeded with 2-3 x 10; organisms per 100 ml and exposed to the membrane
filtration unit. The test run, which extended over 130 minutes, produced a fil-
trate with <1 coliform organism per 100 ml. No membrane breakthrough was experi-
enced during this 3-test run evaluation.
111-29
-------
Turbidity Removal —
Prior to introducing the intermittent, air-assisted cleaning procedure, hollow
fiber membranes were subject to rapid loss of operating flux or rates as the
membranes were fouled by suspended solids in other than low turbidity water sup-
plies. Recent testing of the radial pulse system has revealed satisfactory per-
formance on water samples to which bentonite clay was added to impart turbidities
as high as 30 NTU. Under these conditions a filtrate with 0.2 NTU turbidity was
obtained. Actual field tests on raw water with turbidities of 190 NTU yielded
finished water turbidities of 0.6 NTU. Tests in Colorado on natural surface
waters with turbidities in the 2.4 to 3.0 NTU range provided filtrate turbidities
of 0.57 to 0.25 NTU.
Capital Costs
The application of membrane filtration to potable water treatment has not been
extensive. Therefore, capital cost information is based upon estimates with very
little supporting construction cost experience. Table 111-15 lists general design
parameters and information used to develop cost estimates. Table 111-16, using
estimated costs developed by CWC-HDR under contract to US EPA provides estimated
construction costs for systems of the capacities listed and includes: skid-
mounted membrane filtration units containing the hollow-fiber membranes in car-
tridges; automatic and manual valves for backwashing and unit isolation; flow-
meters; pressure gauges; integral backwash pump; and control panel. A separate
supply pump is included, as is all interconnecting piping serving plants with
multiple units.19 The costs also include storage tanks and solution pumps for
membrane cleaning and an air compressor. Housing is provided for equipment and
supporting appurtenances. Product water storage facilities are not included in
the cost estimates. The cost estimate excludes filtrate chlorination facilities.
The costs do include limited site preparations, a backflush clarifier, and a
design cost allowance.
Application Concerns
Membrane Failure--
Membrane filtration systems provide only one barrier to the passage of poten-
tially pathogenic microbial contaminants into the finished water. A membrane
111-30
-------
failure would permit a direct bypass of unfiltered water into the filtered water.
The Memcor system provides a membrane integrity check system that can be used to
manually, or automatically verify that even a single membrane within a cartridge
has broken. The equipment can either be shut down by such a failure or, prefer-
ably, initiate an alarm signaling the need for operator attention.
Organics Removal —
The larger membranes (0.2 iini) will not effectively remove materials contributing
to color in raw water supplies. The smaller (0.01 urn) inside-to-outside membranes
will remove some smaller molecules, but may not totally remove dissolved mate-
rial;; of subcolloidal size. However, where a raw water contains organics contrib-
uting to color, such as humic and fulvic acids, alternative treatment techniques
may be required for removal of these contaminants.
Pilot; Testing-
Experience with hollow fiber filtration of potable water supplies is limited.
Although it is used extensively in the medical and electronics fields for polish-
ing of process water usually obtained from domestic service taps, only one drink-
ing v/ater facility (Keystone ski area) is known to use the newer membrane filtra-
tion process. To establish appropriate design criteria and establish operating
parameters, pilot testing of any surface water supply is recommended.
CARTRIDGE FILTERS
Equipment Description
Cartridge filters using microporous ceramic filter elements with pore sizes as
small as 0.2 ^m may be suitable for producing potable water from raw water
supplies containing moderate levels of turbidity, algae, and microbiological con-
taminants, according to Kutadyne Products Incorporated's literature. Single-
filter elements may be manifolded in a pressurized housing to produce flow capac-
ities approaching 24 gpm (34,000 gpd) from a single assembly. The clean filter
element pressure drop is about 45 psi at maximum capacity. The ceramic elements
are cleaned when the pressure drop reaches 88 psi.
111-31
-------
Cartridge Cleaning
Cleaning is accomplished by opening the filter housing and scrubbing each indi-
vidual vertical filter element with a hand-operated hydraulically driven brush
that fits over each element. An additional feature offered with one manufac-
turer's equipment is the incorporation of finely divided particles of silver
within the ceramic matrix, which prevents the growth of bacteria in the element.
Other manufacturers utilize, disposable polypropylene filter elements in multi-
cartridge stainless steel housings. These units are available in capacities rang-
ing from 2 gpm to 720 gpm. Membrane pore sizes range from 0.2 urn to 1.0 ym in
equipment suitable for producing potable water. A disadvantage of the polypro-
pylene membrane is that it may be cleaned only once and then must be replaced.
However, according to manufacturers' guidelines, filtration of a low turbidity
surface water (2 NTU or less) could be expected to provide service periods
between filter element replacement ranging form 5 to 20 days, depending upon the
pore size of the cartridge selected.
Application Considerations
The application of cartridge filters using either cleanable ceramic or disposable
polypropylene cartridges to small water systems appears to be a feasible method
for removing turbidity and most microbiological contaminants, although data are
needed regarding the ability of cartridge filters to remove viruses. The effi-
ciency and economics of the process must be closely evaluated for each applica-
tion. Pretreatment in the form of roughing filters (rapid-sand or multi-media) or
fine-mesh screens may be needed to remove larger suspended solids that could
quickly foul the cartridges, reducing capacity. Prechlorination is recommended to
prevent microbial growth on the cartridges and to inactivate any organism that
might pass through the filter elements, including viruses.
The advantage to small systems is that with the exception of chlorination, no
other chemicals are required. The process is strictly one of physical removal of
small particles by straining the water as it passesxthrough the porous membranes.
Other than occasional cleaning or membrane.replacement, operational requirements
111-32
-------
are not complex and do not require skilled personnel. Such a system would be
suited ideally to many small systems where, generally, only maintenance personnel
are available for operating water supply facilities.
Removal Capabilities
Long analyzed a variety of cartridge filters for removal efficiency by using
turbidity measurements, particle size analysis, and scanning electron microscope
analysis.16 The filters were challenged with a solution of microspheres averaging
5.7 urn in diameter (smaller than a Giardia cyst), at a concentration of 40,000 to
65,000 spheres/ml. Ten of 17 cartridge filters removed over 99.99 percent of the
microspheres.
In tests conducted at Colorado State University using live infectious cysts from
a human source, cartridge filters were found to be highly efficient in removing
Giardia cysts.17 Each test involved challenging a filter with 300,000 cysts. The
average removal for five tests was 99.86 percent, with removal efficiencies
ranging from 99.5 percent to 99.99 percent.
QUALIFYING FILTRATION PROCESSES
Need for Pilot Studies
The need for piloting a water treatment process prior to design has been stressed
by practitioners of the "art" of water treatment for many years. It has been said
that every water has its own special rate of filtration, which must be determined
by local experiments. The extent of pilot testing and the desired results may
vary, but the basic need to determine if a given technology is acceptable for a
particular raw water source is prudent and. can assist in limiting the possibility
of a design or process application error.
Some of the basic reasons for pilot testing a particular process might be to:
• Determine if the process is appropriate for a given water source.
t Establish operating characteristics
111-33
-------
• Develop design criteria
t Evaluate the effectiveness of alternative design and operating options
• Optimize design and operating criteria
• Uncover unforeseen design or operating problems
t Provide information to assist in determining capital and operating costs
• Demonstrate the process applicability to regulatory authorities
Techniques for upgrading existing water treatment plants such as tube settlers
and dual- or mixed-media high-rate filtration may require either bench- or pilot-
scale evaluation before a satisfactory application to the full-scale plant can be
assured. Section II described these and other technologies and presented the
general limitations and design guidelines for successful application. For the new
treatment technologies discussed in this section, there may be a greater need to
prequalify a specific treatment process because of limited history of successful
performance or the requirement to establish appropriate design criteria for a
full-scale application. Additionally, if the process or equipment has not
received approval of the regulatory agency, a successful demonstration of the
process may be a requirement to gain probationary acceptance for its use.
Regulatory Agency Requirements
The State of California Department of Health Services Sanitary Engineering Branch
has developed guidelines for treatment of surface waters, stating that "all sur-
face waters used for domestic supply should be provided with complete treatment
consisting of coagulation, flocculation, sedimentation, filtration, and
disinfection."
A provision is provided for a waiver from the defined complete treatment process
but requires that the treatment process be equivalent to complete treatment.
Further, the supplier must establish a guaranteed source of funds, such as a
performance bond, that may be used to install additional facilities if the
treated water does not consistently and dependably meet water quality
requirements.
For one such installation of a package plant using new technology, the State of
California Department of Health Services required the following:18
111-34
-------
Engineering Evaluation of Plant Performance
Within 60 days after the first year of operation, the owner must submit
an engineering report prepared by an independent consulting civil engi-
neer, who is experienced in the field of water treatment, registered in
State of California, and acceptable to the Department of Health
Services. It is recommended that the engineer be involved in opera-
tional review of the plant during the'entire twelve month period prior
to submission of the engineering report. The report shall describe the
operating experience of the plant including any problems or failures
encountered with the equipment. The report shall also describe the
results of all water quality tests performed and evaluate the signifi-
cance of these results. Specifically, the plant's performance will be
compared to the water quality goals established for turbidity as well
as compliance with the California Safe Water Drinking Act and regula-
tions. In addition, the engineer's report will include any recommenda-
tions for the need for flocculation and sedimentation facilities prior
to filtration.
Treatment Process Guarantee '"•-..
The Owner will establish a quaranteed source of funds, such as perfor-
mance bond agreement with the manufacturer, that may be used to install
conventional flocculation and sedimentation facilities at the plant if
the treated water does not consistently and reliably meet or exceed the
water quality requirements established. '
Because of these particular verification of satisfactory performance requirements
it is incumbent upon any potential user of new technology to establish the suit-
ability and applicability of a process before a commitment is made to construct a
full-scale facility. For this reason, demonstration pilot studies would indeed be
recommended.
.* ' . - " r
Conducting Pilot Studies :
*? ' '
-, t
Regardless of specific regulatory agency requirements, pilot studies involving
operation of small-scale prototype treatment processes may be necessary in
111-35
-------
Pilot studies are also recommended more frequently where a high-rate
clarification/filtration 'process 1s being considered. Processes using contact
flocculation/clarification such as the Trident package plant manufactured by
Microfloc Products may require pilot-scale qualification on other than low to
moderate turbidity surface water supplies. Although the process has been pilot
studied extensively on a wide variety of water supplies, and many successful
plants are in operation, the process is sensitive to,proper chemical conditioning
and because detention time is extremely short (less than 15 minutes), if removal
of some contaminant requires greater prefiltration contact tinie ^"misapplication
could occur.
A Trident pilot plant has been developed and used extensively to qualify the
process on a wide variety of water supplies. Figure 111-21 is a flow schematic of
a trailer-mounted 20 gpm pilot unit which can be used to qualify the process.
Such a unit was used to qualify the process in the following case study.
Case Study - New Treatment Facilities Pilot Study Selection Process
Background—
The following information was obtained from a water supply feasibility study for
a new treatment and distribution system to serve approximately 2,500 connections
presently on private individual supplies. The only source of water for this large
area is Clear Lake in Lake County, California. Although the lake is used as a
source by several communities, they all experience major difficulties in handling
severe seasonal taste and odor problems. In general, it is classified as a diffi-
cult supply to treat.
The feasibility study recommended complete treatment including preozonation,
chemical coagulation, pH control, flocculation, sedimentation, filtration, and
postfUtration activated carbon contactors followed by disinfection. Because the
capacity was projected to be less than 2 mgd, a package pl'ant of a particular
design was found to be the most cost-effective treatment plant alternative.
However, since package plants had not been used previously on this supply, a
pilot study was suggested to establish the feasibility of the process and to
determine design criteria.
111-38
-------
Quality and Treatment-
Treatment of Clear Lake water to a satisfactory quality drinking water presented
some interesting challenges. The lake is eutrophic and Is plagued by a series of
blue-green algal blooms (Microcystls and Anabaena), which impart serious taste
and odor to the lake water. Taste and odor problems develop as early as May and
last into early November. Taste and odor threshold numbers (TON) as high as 10
have been measured by other users of the supply. Other than the taste and odor
problem, Clear Lake water is of excellent mineral quality as is indicated in the
water analysis presented in Table 111-17. Additional water quality data is pro-
vided in Table 111-18. Turbidity can range from less than 3 NTU to greater than
90 flTU during winter storms. Other than turbidity removal, disinfection and sea-
sonal taste and odor control, no additional treatment of the lake water is
needed.
Treatment experience on Clear Lake water .is relatively extensive. Several water
districts use Clear Lake as their sole supply and have been treating this water
for many years. The City of Lakeport recently installed the most sophisticated
plant which uses both ozone and activated carbon to eliminate taste and odor from
the finished water. During the first year of operation, no taste and odor com-
plaints were registered by the city, attesting to the effectiveness of the
process.
To gain information on available procedures and their success in processing Clear
Lake water, four utilities using lake water were contacted. Detailed information
acquired during this survey is not included in this document, but a general over-
view of their experiences is as follows. In general, lake water users are experi-
encing increasing difficulties controlling taste and odor. Previously, several
utilities were able to use potassium permanganate and powdered activated carbon
with reasonable success. Recently, more expensive postfliter granular activated
carbon contactors have been installed. At least two utilities are using ozona-
tion. Those using ozone have received the lowest percentage of customer
complaints.
111-39
-------
Pilot Study Program-
Other utilities treating Clear Lake water have experienced difficulty with rising
floe due to air bubbles released by algae and generally poor clarification expe-
rience during the algal bloom season. For these reasons it was felt that the
adsorption clarification process using an upflow bed of bouyant media as a pre-
filtration clarification process may perform more effectively. However, there was
still the concern that the adsorption clarifier could not handle the high and
widely fluctuating raw water turbidity conditions.
Consequently, the supplier of the adsorption clarifier package plant, Microfloc
Products was asked to perform a pilot-plant-scale evaluation of their process to
confirm its suitability for this project. In addition to establishing basic tur-
bidity removal capabilities, another objective of the study was to determine
whether potassium permanganate or powdered activated carbon applied prior to
clarification would effectively remove trihalomethane (THM) precursors and taste
and odor substances should they be present. If so, the dependence upon expensive
ozone for taste and odor control could be lessened.
Pilot Study Results-
Data collected by the manufacturer is presented in a pilot study report (see
Appendix). Conduct of the study was witnessed by the design consultant to verify
that results are representative of actual field experience.
The results were favorable. The adsorption clarification/filtration process
removed turbidity to extremely low residuals. Potassium permanganate was success-
ful in removing the low conentrations of taste and odor substances present during
the test period. Unlike previous years, no major algal bloom was experienced in
1986, which made taste and odor removal less difficult during the early September
pilot test. Powdered activated carbon helped remove THM precursors and the
adsorption clarification process performed satisfactorily at dosages up to 25
mg/1 without filter breakthrough or seriously shortened operating cycles.
111-40
-------
Recommended Design Criteria--
As a result of the favorable pilot study experience, the following general
process design criteria is proposed for the new treatment plant:
Item
Nominal Design Capacity
Adsorption Clarifier Package Plant
Clarifier Design Loading
Filtration Rate
Filter Type
Parameter
2 mgd
10 gpm/sq ft
5 gpm/ sq ft
Tri-mixed media
Chemical Feed Systems
Ozone
Al urn
Permanganate
Cationic Polymer (Nalco 8109)
Powdered Activated Carbon
Nonionic Polymer
Sodium Hydroxide
Case Study - Untreated Supply for Small Community
Normal Dosage Range
1 to 2 mg/1
10 to 20 mg/1
0.2 to 1.0 mg/1
1 to 10 mg/1
10 to 20 mg/1
0.1 to 0.5 mg/1
20 to 40 mg/1
Example—
A small community presently uses a surface source (small stream) fed from snow-
melt water from a relatively isolated high mountain watershed. Beavers and musk-
rats are present in the upstream watershed. Presently, the supply is only chlo-
rinated since seasonal average raw water turbidities rarely exceed 1 NTU.
Giardla cyst presence is of concern. The community has about 70 connections and a
peak water usage of about 68,000 gpd. The system is operated by a maintenance man
with entry-level water treatment certification for his state. The community is
currently paying $20 per month for metered water service and has no capital
reserves to build treatment facilities.
111-41
-------
The objective of this case study is to determine what treatment; technology would
be available for this application, considering raw water quality, operator skill
level, and construction and operating costs. ; ,- ,
Evaluation Procedures—
The steps that would be followed to identify an appropriate treatment technology,
establish preliminary process design criteria, prepare a preliminary facility
layout, and establish costs are as follows:
• Review all raw water quality data available for specific supply.
t Establish possible alternatives to be considered.
t Determine if bench scale or field pilot studies are required to qualify
processses under consideration.
t Select alternatives suitable for specific situation.
.*••
9 For this example, consider the following: . .
- Slow-sand filtration
- Package plants
- Membrane filtration
• From literature and bench or pilot studies, establish treatability charac-
teristics and other design considerations of each potential process.
• Basic Process Alternative Concerns
Slow-Sand Filtration
- Is turbidity filterable?
- Is color present at concentrations above standards?
- Will sand sizes available remove Giardia cysts?
- How will filters be cleaned and s'and renewed?
Package Plant .•,.--•
- What treatment chemicals and dosages are needed?
- Will process reliably remove cysts?
111-42
-------
. - What operational skilMevel is needed?
- What regulatory agency design standards must be met?
- Where will sludge be disposed?
Membrane Filtration
- Will it meet effluent turbidity standards?
- How will color, if present, be removed?
- What steps are required to gain regulatory agency acceptance?
- What operational skill level 1s needed?
t Prepare conceptual design criteria and preliminary layouts of selected
treatment alternatives.
• Develop construction costs;
• Compare alternatives in terms of reliability, simplicity, flexibility, and
implementability.
• Select most appropriate process for the application.
• Proceed with final design.
Cost Analysis—
For this example, using cost tables presented in this section, the following
costs are determined:
Process
Package Plant
(Table 111-2)
Estimated Capital Cost, $
295,000
Total Cost,
t/1.000 Gallons
277
Slow-Sand Filtration
(Table III-ll)
273,000
205
II1-43
-------
Process
Membrane Filtration
(Table 111-16)
Estimated CapitalCost, $
269,000
Total Cost,
,000 Gallons
179
All three alternatives are quite close in capital cost. Membrane filtration is
the least-cost alternative both in-capital and total costs. Factors other than
cost will probably-, influence the, decision on what process to adopt for the
particular application.
111-44
-------
uo
L_
r^
Z
^
Q_
UJ
(J
^£
y
^j
^
o_
Z
O
H~
^^
y
u. •
a:
_i
U
Z
0
i-
o_
c*.
QD
2 u.
x "*
i
3
£
Q *• *v "n
§< u.
U UJ
H <
. H
UJ
3
u!
Z
11
u
5 et "o
. H Z £»
< p u.
Q
ul
£ ?
x 6
•B x
a 1 «
u Ss uj
H C —
^ ^
§*
w
o
i
'"**••
T» «s O *r «^
»»» «r> r-i n ri
O O O O O
•* >0
000*^"*"^
— rs
ft rs
8^3 fO ^N oo
— * 2
00000
Sao r~ — —
>o m ^ «r
^ — «N r>«
•
rs ^f m ^ »o
2 S 2 3 3
-------
. +-> C
•r- in
a. o
<0 O rH
o
<0
LU CO
0)
CTI •
(O 3E
t- O
O) r—
4J
C
(O
L Q.
(O
. o
vo oo "
co cv»
er> co t— i r- 1
•M
C
(O
r-»
O.
0)
o>
«
Jt*
VI
(0
^3
0)
3
Trt
•o
co tn cj «± o •-•
» • •
in co cc to
CO
CJ
•o
0)
0) r-
.
a)
x:
•»->
52
O
CM
in CM to
CJ
00 LT> CO CO O CO
r-» CT* tv r»> r~ ir>
CM CJ ^- r- r>- &
Q. o> *J •*•
E «a c -M
O L. 3
O O _=
^_ *- tn
O) |g =
^ rQ |Q
i, * 2
*> ^« +j
a^0
*— _
c Si c
° 3-
+J tt
re » E
_- -w— —^
TO u_ O
T3 T-
^H «• fO O O
o o •-< ^- oo
o o o o «-«
0) -O
o> c
? != °
C TO CT
-r* ^L fl)
ac
CJ VO
C O
O O ID
U1 if) 00
o.c o o co in
CJ CO
tf) VD
•4J Cf> O> V*
I/I (O C C tt>
4-> en
O O 1C
O r- 3: * *J
l_ 4— QJ d> 1X5
D- >—' "O "O O
111-46
-------
Q
Q
_j
^
u_
Z
Q
UJ
>
i
U
HI!
SS2i
<
LLJ
&.
\-
C£
as
(M
OQ
<
e
§
i* 1 I I* I
i
s
6
!i
1 ? 1
§
ia
i
I
••
111-47
-------
to
LU
u
u
u
O
c*.:
a:-:
2'
LU
I
•* a,
HI
£
CO
ME-
EMICA
11
11
j j*
;;i I i II
I a | a a a a
Hliihi ihi
I I . i i
2222
t\
•I
11 i is
£ £ DO
tit
*
i
•5
3s
> 2
j
111-48
-------
ii
ll
H-
>
t
Q
H-
ui
i
ii
i
d_i •
r4 ^ ,
ST i •» ei «i «• - «
v I m M •• m n r>
A A
325
•>M«»OAr«m«kM
deociwdcfodci
-------
a
_i
UJ
u_
LLJ
LU
Qi
u
z
h-
<
Qi
LU
a.
CQ
<
I
2
3
g
0
£
^
a
^
g
mf
t.
z.
3
m
2
Z
—
o
s
at
1
h.
e «*
** °.
•o
it
IM d
it *
• 09
. . e
IF* ^J*
^»
P^
o
S d
0 S
O
O
Al
_ IM
5 "
0 g
d
g
tn e
O IM
d
. o
*
r> •-
in in
CM
s i
£ £
I
I
•o
*
o» m
It rt
e •
•r
ra
it o
e y>
*
e
<>
'IVI It
p
a
• 1
o. o
CM
_ It
•B •
'«• O
o *•
d
*Q •
Mr O
d »•
e
•o
CM •
-
e
o> «o
KI in
CM
I !
•r
S
I
*
o «
K»
K.
« 2J
o it
- e
o>
CM
«» ^
* O
•0
«, 8.
o. o
o ^
*
o
«»
St
o
0 0
d
R*
a
• *
. ,' d
«o
-
it
•* 1C*
«r <«
M
I!
i
i
«
•> *^
M »«.
S ?
• !•%
0 "
£•}
r^ *
^
g
e d
0 o
0
R
M e
o -*
^
•
0 a
S d
• s
e
•
•
IM
9^% ^B '
•o e
^f
i 1
X AC
i
VM
•
S
4
•"* *A
*
IM
IM
•
0 O
e —
d
it
— IM
• (M
*• It
•
^
IM
O
e e
0 o
o
It
It
CM d
O jr
d
^
it d
d £
°*
•
It CM
IM 0
**
e
•» 2
-------
TABLE HI-7. OPERATING DATA - LEWISBURG, WV
Month
Turbidity, MTU
Clarifier
Influent
Clarifier
Effluent
Filter
Effluent
Dae., 1984 Mean
Range
5.0
1.2
0.16
1.8-18 0.6-2.5 0.10-0.20
May, 1985 Mean 9.6
Range 2.0*50
1.0
0.18
0.4-6.0 0.10-0.50
June, 1985 Mean
1.7
0.60
0.16
Range 0.70-5.0 0.1-1.1 0.10-0.50
111-51
-------
TABLE tll-8. OPERATING DATA - PHILOMATH, OR
(Monthly Average Values)
Turbidity. HTU Chemical F«d, mg/L
Month
February, 1986
March, 1986
April, 1986
Nay, 1986
June, 1986
July, 1986
August, 1986
fttu Clarlfler
26.7 3.40
6.50 1.67
6.27 1.49
• 5.82 1.4S
4.25 1.21
3.43
3.02
Miter
• *
0.17
0.20
0.23 ,
0.14
0.22
0.26
Alua
15.5
11.4
• *
11.7
11.3
10.0
16.6
Soda Ash
21.2
4.5
••
1.2
4.8
5.7
• *
Polymer
0.29
0.08
•-
0.08
0.07
0.07
0.07
net
Production
• -
91X
90X
88X
93X
93%
93X
111-52
-------
TABLE 111-9. OPERATING DATA - HARRISBURG, PA
(tenth
Turbidity, MTU
CUriftw Filter
feat feed, mg/l CUrifier
• Ratt;
Aim Polyntr gpn/tq ft
July, 198S Mtan 8.1 . 1.36 0.17 MA NA NA
Rang* 3.7*15.1 0.60-3.4 0.13*0.32
Aug.. 1985 Nmn 8.6 0.94 0.18 11 0.12 14
•HW* 4.8*14.3 0.7*1.3 0.13-0.23 8.1-13.4 0.10-0.13 11.9*16.3
Sept., 1983 ttean 9.3 1.2 0.18 9 0.07 13
Rang* 5.9*37 0.5*4.9 0.09*0.34 4.9*18.7 0-0.31 11.7-14.9
111-53
-------
TABLE 111-10. OPERATING DATA - RED LODGE, MT
Month
Jan., 1985 Hean
Ring*
Turbidity, MTU
Rau CleHffer Filter
Chemical Feed, mg/L '
Water
Temperature;
Aim Polymer Bentonite Oeg. F
0.20 0.29 0.04 2.J 0.48 0.93 36
0.15-0.33 0.21-0.35 0.03-0.05 1.4*3.5 0.48-0.48 0.93-0.93 34-38
ftb., 1985 Hean 0.19 0.28 0.05 2.3 0.48 0.93 34 '
Rang* 0.16-0.21 0.22-0.39 0.30-0.07 1.4-3.6 0.48-0.48 0.93-0.93 33-38
Har., 1985 Mean 0.20 0.29 0.05 3.0 0.48 0.93 36
Range 0.18-0.22 0.25-0.32 0.03-0.06 1.9-3.9 0.48-0.48 0.93-0.93 35-37
Apr., 1985
Hay, 1985
June, 1965
Naan
ft*nga
Hean
Range
Hean
Hang*
1.1
0.22-1.6
2.1
0.16-4.5
3.5
1.5-6.2
0.49
0.27-1.1
0.41
0.16-0.80
0.41
0.21-0.75
0.08
0.04-0.14
0.07
0.03-0.14
0.07
0.04-0.18
5.0
2.3-7.1
7.3
3.6-13.4
10.«
7.4-12.8
0.60 0
0.48-0.75
0.48 0
0.48*0.48
0.60 0
.0.48-0.75
40
38-43
44
43-47
45
43-49
111-54
-------
H-
Z
LU
ex.
h-
LU
u
D Z
a: H
O <
VT ^
O D
P-l
<0
LU 00
LU
oo
IB 4->
.0 O ~*
U
01
<^
Q. O •
IB W »M
01
2S
o» ^>
C f
« O "O
p*' Ifl
00 tO
CV
r^ O> O « »
Ol >•-
U O
£. U
O> 3
O 4J
V. IA
jc e
«•> o
u
O)
c.
O)
CO
o
M e mo
Q,
«-• ID
O «0 CSI <*.
CM m *
rf» w oo « «*> m
•« v <*> Cvj r>-
«g IB
(/) U
f.
It!
IB «B 0»
So
u . a
> u
S2S"
0) -o
— 2 i
vi
l O>
x *5 c
O O) oi
10 oc Vo
O\
VO SO O O _
O C ••» ID i/J OC in
O O C O Co tf> t—
«• if> (O r*
OJ
o»
(A 4-> I
O» U Ol
S £g
QJ *J 1-
o e "i
u o w
o. u e
111-55
-------
z
O
a:
LU
Q
O
u
5,
fljj uJ
-------
So en
-.c.S'eH
l.ll
D
in
H
Q
U
Z
z
LLJ
a.
5
Qd
LJJ
«•- >, «-«—
13~ » ~
X L. «/)
.03.
leiSfe
> ex **
• * IA ^
UJ
co
*
*
^
>!
M ••
« * e —
4J b. «-»•
« 0 U
Q} 4^1 <^ ^»
*j — tn S
w S "C'w
o « *«
*j vi».
^§.
o — c *>
2
CM I I I I IA
• I I I I •
1 CM '.
i
*5 *3 5
CM CM CM
coo
f^ QD
•» CM
e e
• i
S
en
7
»
"7
00
T
e e co
CM
CM
CM
I
CM
CM
«- P- CM
S oo cc
*3 *3 *»
2 S
t i
"I °.
o o
« •
-* CM
« S
i I
£ £ £
i J J
2*2
S g § S
>- s s ^
!fi ^ ° S -.
r^ ^tf ^ N SO
• • • * •
o e o e —
3
5 SS S3
e -e
CM
•r m CM >
J 7 ' I 7
iA *•>• «O CM •
f? CM O •" • •*
I I I I I
CM
I =• - I I c? § s
0)
g
g
<.
01
en
&
u
«
• c * «-
Isi5
4J
^T""*™' *
111-57
-------
•-D
QiO
LU LU
.
OD
tofZ
0<
Uoi
h-
H-o^
en s;
LU UJ
01
»— «o
<0 ** O
*J l/> O
i23 J
VI
o
o
01
u
c
«
01
c
o
•r-
+»
ID
0)
O,
O
•f- l/»
O. O o
ID O >-H
4>
(. O -O
n o -15
CL a.
ID
• «T3 W> WIM-
4J C 2 |
*• 2"oJ>
*- *i Q.w»
O «^«3 fe
W ** * 5
o o o
*£
oo<— i
VO 00 SO
*^" O e c. *^
^ ^ A< C fl*
« ^ I— «
COfMPO I»"SC:>2*S
*-«co P
^ ca to
I-H CM
oj a> o +J 3^
• t_ O>r^ 3 _
«|3-uS
S
111-58
-------
Zj
0
<
LL.
o
t. t_ >*-
5-5-°
t >
O 0. O
VD
0
S. X
•»-» 3
*J ««- r-
V> <*-
>» o-
"* v> m
c o 2>
o «? .5
t_ +J
<-< in o o
o o o o
rH CM CO *d-
O
tfl
£*<
00-5
in o o o
ro o o o
CT» ej <£> r-»
• •> •
cm n ^>
' s s e a * ' B
m r-i r-« r*. i*» r>»
co vo 10 oo oo oo
X X X X X X
S S 8 s
fo r* r^- oo
esj e\j <^ *J-
x x x x x x
e
oo
CM .-« •-•
'£ "S1^
^» ^j
*j «'C
i- 3 > «
4-i p O.
0)
.S!
.o
O O O
-i c«j «*
o E c
R) CL-I-
' O O O
if> O O O O
u> CM in r-i CM ro
VO
4-> t* J* <*•> o>
c
O _frr ^J CU d
.
OJ (O O
Q Ul O
* *
J» O
*- Z
Jr
.111-59
-------
TABLE 111-18. CLEAR LAKE WATER QUALITY AT DWR SAMPLING
STATION NO. 1 AT LAKEPORT
DATE
SAMPLED
1/7/77
2/3/77
3/10/77
4/7/77
5/5/77
6/16/77
7/14/77
8/11/77
9/22/77
10/14/
11/4/77
12/9/77
1/6/78
2/10/78
3/9/78
4/6/78
5/11/78
6/8/78
7/13/78
8/10/78
9/14/78
10/5/78
11/9/78
12/7/78
1/5/79
2/16/79
3/15/79
4/5/79
5/10/79
6/8/79
7/12/79
8/10/79
9/7/79
10/4/79
11/8/79
12/7/79
1/11/80
2/6/80
3/6/80
4/10/80
5/8/80
6/6/80
7/11/80
8/22/80
9/18/80
10/15/80
U/13/80
TURBIDITY
(OTU)
21
9
25
18
14
35
14
7
60
50
33
36
67
49
20
13
6
3
3
7
18
12
22
20
21
18
12
20
8
5
6
4
13
17
28
19
28
17
3
19
3
6
5
5
12
9
13
Sp. Cond.
(UMHOS/cm
307
308
308
316
321
337
346
302
364
385
373
366
193
197
216
223
227
225
238
258
271
275
290
294
299
255
281
260
259
284
291
304
311
320
299
299
260
245
228
235
236
246
263
268
28r.
2!.S
256
V,
8.3
8.1
7.7
8.1
8.0
7,8
8.0
8.0
8.2
8.1
8.3
8.2
7.3
7.6
7.6
7.6
8.2
8.2
8.1
7.8
7.8
8.4
7.5
7.7
7.8
7.7
7.2
8.1
8.2
7.8
8.1
8.0
8.1
8.6
7.7
7.5
8.3
7.9
7.7
7.7
8.1
8.1
8.1
7.9
8.3
8.3
8.4
3
8l
9.9
11.8
9.2
7.7
9.1
4.5
5.0
4.3
2.6
3.6
5.7
6.8
9.7
10.0
9.1
9.1
10.4
6.9
8.1
5.9
7.0
10.6
4.3
9.5
10.4
10.4
6.7
10.2
9.9
7.0
7.0
5.8
5.3
7.0
8.5
7.1
10.2
10.3
9.8
9.6
8.8 ••
8.5
7.7
6.7
4.8
7.1
9rl
|
15
11
13-.
13
12
12
14
14
14
16
17
17
16 .
,16
13
12
13
.
: ;
' •
10
Alkalinity
(«CaC03)
142
144
144
147
156
154
160
138
165
170
174
169
'
138
130
143
. 'i
..
,
Chloride
(ng/D
6.8
7.0
6.6
6.4
6.5
7.9
8.8
9.6
8.8
10
7.7
8.5
.
•
, .
8.1
6;0
6.0
Nitrates
-------
s
fa
NlitAI
1V9IH1N9
111-63
-------
FLASH
MIX FLOCCULATION
FILTRATION
INFLUENT.
0 TUBE SETTLING
u
i
lio
DESLUDGE BACKWASH- BACKWASH
TO-WASTE SUPPLY
SURFACE
WASH
SUPPLY
EFFLUENT
Figure III-2. FLOW DIAGRAM OF ALTERNATIVE PACKAGE PLANT
(COURTESY OF WATER TECH, INC., VANCOUVER, WASHINGTON)
111-64
-------
MEDIA
RETAINER
INFLUENT
AIR
INLET
OPERATING
WATER LEVEL
"ADSORPTION-
>•..-.. MEDIAE.
FILTRATION MODE
FILTER
EFFLUENT
FLUSH CYCLE
WATER LEVEL
MEDIA
RETAINER
' COLLECTION
TROUQH
: t-•*•••••• ••« '-V
•> ADSORPTION '•
«.„' MEDIA -.•«.".
EXPANDED FOR
Tj CLEANING '•;
'AIR AND WATER
oaoeeeooDeeoeoooo
WASTE
INFLUENT
AIR
INLET
BACKWASH
INLET
FILTER
EFFLUENT
ADSORPTION CLARIFIER FLUSH CYCLE
SURFACE WASH
\Muavnriiwn. f
MEDIA'*.'/
.EXPANDED FOR
•4 CLEANING
' \ •' O'
.AIR AND WATER
MEDIA
RETAINER
WASTE AND
OVERFLOW
CONNECTION
SURFACE WASH
AGITATORS
'/I BACKWASH
•Vj
'INLET
BACKWASH CYCLE
FILTER
EFFLUENT
Figure 111-3. OPERATING CYCLES OF PACKAGE PLANT
(COURTESY OF MICROFLOC PRODUCTS)
iII-65
-------
8fiS
&
-------
Ul
$
c
<
lit
u
o
Z
o
Ul
.e
IU
o
o
Ul
o
I
-------
Ul
5
c
Ill
.J
o
Z
o
w
e
ft?
Ul
»-
tu
w
O
o
Ul
d
o
Ul
K-
Z
oe.
D
Z
O
D
LU
DC
UJ
00
si"
111-68
-------
5
m
100.0
80.0
60.0
50.0
40.0
300
20.0
10.0
80
6.0
5.0
40
3.0
2.0
10
0.6
0.6
0.5
04
0.3
0.2
0.1
» i '—I 1 1
9999 99.8 9998 9590 80706050403020 10 5
PERCENT EXCEEDING SPECIFIED RAW WATER TURBIDITY
Figure HI- 7
AVERAGE RAW WATER TURBIDITIES AT
SLOW-SAND FILTER PLANTS SURVEYED
111-69
-------
o
S
a
100
8.0
60
50
4.0
3.0
2.0
1.0
08
06
05
04
0.3
0.2
0.10
0.08
0.06
005
004
003
002
0.01
i i
99.9999.8 9998959080706050403020 10 5 21
PERCENT EXCEEDING SPECIFIED FILTERED WATER TURBIDITY
Figure III-8., AVERAGE FILTERED WATER TURBIDITIES AT
SLOW-SAND FILTER PLANTS SURVEYED
111-70
-------
10000
800.0
6000
500.0
400.0
300.0
200.0
o
o
(C
o
o
o
100.0
80.0
60.0
50.0
400
30.0
20.0
10.0
80
6.0
5.0
4.0
3.0
2.0
1.0
i I I I
9999 998 9998 9590 80706050403020 10 5 2
PERCENT EXCEEDING SPECIFIED RAW WATER COLIFORM CONCENTRATION
it--
Figure III-9. AVERAGE COLiFORMS IN RAW WATERS AT
SLOW-SAND FILTER PLANTS SURVEYED
111-71
-------
10.0
9.0
6.0
7.0
6.0
5.0
4.0
3.0
E
§
2.0
flC
o
u.
Zj
o
u
i.o
0.9
0.8
0.7
0.6
05
0.4
03
0.2
0.1
I I I I i
I 1
I I
9999 998 9998 95 90 80706050403020 10 5
PERCENT EXCEEDING SPECIFIED FINISHED WATER COLIFORM CONCENTRATION
Figure 111-10. AVERAGE COLIFORMS IN FILTERED WATERS AT
' '"' SLOW-SAND FILTER PLANTS SURVEYED
111-72
-------
Ill
CO
o
Z
Ul
_
0
o
u.
o
Ui
o
_J
c
c
tf
xSX^
E
k\V
\SNNN^
V .
K^X^
"OXX^
5
E
kX\\\\\\\\\\\\\\
CO
I
O
z
o
CO
_ o
• o o 6 ' b o
) O O O O
> * CO CM ^«
S1SAO JO
111-73
-------
Ul
(0
3
o
• X
LU
o
o
r»
u.
o
Ul
o
I
«
to
9
a,
ui
CO
i
o
CM
Q
CD
I-
C*.
CD
CO
I
z
" 3
DC
Ul
I-
O
U
K
u
U.
+
a:
LU
I-
D
Z
o
•: J
I Ul
u.
I
10
CO
I
>
O
•4
o
oc
D
em
LU
-------
Filtrate
Clear liquid line
q
Precoat
tank
watpr •- * -*• ii -ii
Wdve =*T* Jh ter
source W feed
pump
(
kj
>
i
(
i
)
Q
Body
feed
tank
(
i
)
r
(
(
J
" U/
Backwash
line
Body feed
pump
/~?— i
,' ^.«
Filter
)
-^
Septum
T /S^
f
\.
^
(
^
X_
\
•H
••
-Filter^
cake
> (
)
€
(
>
Precoat
drain
line
Backwash^/
drain line
Figure III-1.3.-TYPICAL PRESSURE DIATOMACEOUS
EARTH FILTRATION SYSTEM
111-75
-------
OPERATING MODE
The feed stream is pumped into
the cartridge shell and passes
over the fiber walls. Some liquid
filters through the walls and exits
the cartridge as clean filtrate. The
remaining feed and rejected
waste flows across the fiber wall
as concentrate and exits through
the shell outlet.
Concentrated
watte
material It
rejected from
shell
Feed stream
is pumped
into shell and
separated
into two
component*
Clean Illtraia
exits from
end ol shell
RADIAL PULSE Sequence
The filtrate is temporarily shut off
and gas is introduced into the
lumens of the hollow fibers. The
gas explodes through the
microporous fiber walls into the
feed stream causing violent
agitation and purging the
cartridge of any waste buildup.
2. Air pumped in
•I higher than
fMd pressure
5. Concentrated
waste buildup i>
forced out
3. Air explodes
through fiber
walls Into con-
centrate stream
4. Feed stream
Is pumped into
shell
1. Cleen filtrate
I* blocked off
Figure 111-14. HOLLOW FIBER MEMBRANE OPERATIONAL DESCRIPTION
111-76
-------
. Non-Corrosive Piping
System
Chemical Resistant Membranes
Clean In Place
System
Stainless
S.teel Skid
Mounted
Frame
Pref iltration and Pretreatment System
When Required
Microprocessor
Controls lor Radial
Pulse, CIP System,
and Membrane
Integrity Check
Stainless Steel
Centrifugal Pumps
Figure 111-15. SKID-MOUNTED MEMBRANE FILTRATION ASSEMBLY
111-77
-------
FILTERED
RAW WATER
CLARIFIED
RECYCLE
DISCHARGE
UNFILTERED
RECYCLE
HOLLOW FIBER
MEMBRANES
MEMBRANE
CARTRIDGE
MEMBRANE CLEANING
SOLUTION TO SEWER
AIR INLET FOR
BACKWASHING
BACKFLUSH
WASTEWATER
BACKFLUSH CLARIFIER
Figure 111-16. FLOW SHOT OF MF.MBRANH FILTRATION SYSTF.M
111-78
-------
. -Si
FLOW
CONTROL
VALVE —
DISCHARGE
SAMPLE PUMP
DISCHARGE PIPE
SUPPORT BRACKET
2'-6
4'-0
FLOW
INDICATOR
PERFORATED
COLLECTION PIPE
CORRUGATED
FIBERGLASS
SIDE PANELS
TUBE
SETTLING
MODULE
Figure 111-17. TUBE SETTLING TEST MODULE
111-79
-------
HJD1CATOB MlfcM
PIES3SUBC COMUECTIOU
MIMTCK
COKiVJCCTIOM
lOOiOM at
FEMM.C. MOSC
Figure 111-18. PILOT FILTER ASSEMBLY
111-80
-------
MIX! U
r LAGH MIX.
CONSTANT
LEVLL TANK
J"
n
G O A U U L A N T
I
-4—
OVERFLOW
TO DRAIN
DIFFERENTIAL
PRESSURE GAGE
SURFACE
. WASH-
Figure 111-19. PILOT FILTER SCHEMATIC
111-81
-------
CO
cr
D
o
O
<
OJ
I
1? 1C 20 24 28 32 3G 40 44 48 52 56 60
RUN TIME, hours
SAND MEDIA
8 12 16 20 24 28 32 36 40 44 4B 52 56 60
RUN TIME, hours
Moffat Water Treatment Plant - Pilot Filtration Tests
Date: &•-'-' 1985
Mode:
direct filtration Q
convent iona 1 (3
Loading: 5 gpm/sf E3
4 gpm/sf O
gpm/ sf D
Filter T ype:
sand ™
mixed media •
dua 1 media no. 1 ID
dua 1 media no. 2 H
Filter Medi*
Type
Sand
Mixed Media
Dual Media No.l
Dual Media No. 2
Turbidity,
NTU
INF.
1 . (;
1 .1
-
1 .6
Alum Dosage:
m g / 1 20
Coagulant Aid
Dosage:
m5/l U-]J
Coagulant Aid Type:
Nalco 8 1 8 1 C3
Calgon .L-6 52 -E D
n
EFF.
(j.;>3
0.23
-
0.22
Filter
sand: n
Mixed:
Dual n
Dual n
Run
Length.
Hour*
•1 <1
1. u
49
-
T3
Run Ended
. Out To:
HL
x •
X
-
Aid Dosage:
10/1 . 0
m g / L 0
a. l:m j /L
D. 2 :mg/L
-ft—
TUR8
-
X
FPt
ClOOO's)
4S3
793
-
295
FPI
Rank
2
1
-
3
Filter Aid Type:
NP I 0 O
Nalco 8181 O
0
FIGURE NO. D-45
Figure 111-20. TYPICAL PILOT FILTRATION DATA LOG
1II-S2
-------
*«
• v
z
<
O
o.
U
P
<
LLJ
x-
u
1/1
o
CM
I
i
00
11--83
-------
REFERENCES
1. Morand, James M., and Matthew J. Young, "Performance Characteristics of
Package Water -Treatment Plants, Project Summary." Municipal Environmental
Research Laboratory, Cincinnati, OH, EPA Project 600/52-82-101, March 1983.
2. "High-tech Water on the Green Fields of Iowa," Mater Engineering and
Management, January 1985.
3. Microfloc Products, "Dauphin Consolidated Water Supply Company," Case
History No. 50.
4. Culp/Wesner/Culp, "Technical Guidelines for Public Water Systems,"
U.S. Environmental Protection Agency, Cincinnati, OH, EPA Contract
No. 68-01-2971, June 1975
5. Cleasby, J.L., D.J. Hilmoe, and C.J. Dimitracopoulos, "Slow-Sand and Direct
In-Line Filtration of a Surface Water," J.AWWA, p. 44, December 1984.
6. Hendricks, D.W., W.D., Bellamy, and M. Al-Ani, "Removal of Giardia Cysts by
Filtration," in: Proceedings of the 1984 Specialty Conference, Environmental
Engineering, ASCE, June 1984.
7. Bellamy, W.D., G.P. Silverman, and D.W. Hendricks, "Filtration of Giardia
Cysts and Other Substances, Vol. 2: Slow Sand Filtration," EPA-600/2-85-026,
U.S. Environmental Protection Agency, MERL, Cincinnati, OH, April 1985.
NTIS: PB-85-101633/AS.
8. Letter-man, R.D., and T.R. Cullen, Jr., "Slow Sand Filter Maintenance Costs
and Effects on Water Quality," EPA-600/2-85-056, U.S. Environmental
Protection Agency, Cincinnati, OH, May, 1985. NTIS: PB-85-199669/AS.
9. Slezak, Lloyd A., and R.C. Sims, "The Application and Effectiveness of Slow
Sand Filtration in the United States," J.AWHA, 76:12:38, 1984.
MI-84
-------
10. Pyper, Gordon R., "Slow Sand Filter and Package Treatment Plant Evaluation
Operating Costs and Removal of Bacteria, Giardia, and Trihalomethanes,11 EPA-
600/2-85-052, U.S. Environmental Protection Agency, Cincinnati, OH, April
1985.
11. Bryck, 0., B. Walker, and G. Mills, "Giardia Removal by Slow Sand Filtration
- Pilot to Full Scale," presented at the AWWA National Conference, June
1987; .
12. Logsdon, G.S., and K. Fox, "Getting Your Money's Worth from Filtration,"
J.AWHA, p. 249, May 1982.
13. Lange, K.P., W.D. Bellamy, and D.W. Hindricks, "Filtration of Giardia Cysts
and Other Substances, Vol. 1: Diatomaceous Earth Filtration," EPA-600/2-84-
114, U.S. Environmental Protection Agency, MERL, Cincinnati, OH, June 1984.
NTIS: PB-84-212703.
14. Malina, J.F., Jr., B.p. Moore, and J.L. Marshall, "Poliovirus Removal by
Diatomaceous Earth Filtration," Center for Research in Water Resources,
University of Texas, Austin, TX, 1972.
)
15. Hibler, Charles P., and Don L. Monzingo, "Evaluation of the Memcor
Microfiltration System for Removal of Giardia Cysts," Department of
Pathology, Colorado State University, Fort Collins, CO.
16. Long, W.R., "Evaluation of Cartridge Filters for the Removal of Giardia
Lamblia Cyst Models from Drinking Water Systems," Journal of Environmental
Health, 45:5:220-225, March/April 1983.
17. Hibler, C.P., "Evaluation of the 3M Filter 124A in the FS-SR 122 Type 316
s/s #150 Housing for Removal of Giardia Cysts," Department of Pathology,
Colorado State University, unpublished, 1986.
18. State of California Department of Health Services, "Requirements for the
Design and Operation of the Trident Water Treatment Plant," February 1987.
19. CWC-HDR, "Technologies and Costs for the Treatment of Microbial Contaminants
in Potable Water Supplies," U.S. Environmental Protection Agency, Science
and Technology Branch, Office of Drinking Water, Washington D.C., April
1987. .
111-85
-------
-------
CHEMICAL COSTS
-------
TABLE A-l. CHEMICAL COSTS USED TO DETERMINE O&M COSTS
Chemical
Alum (Dry)
Alum (Liquid)
Lime (Quick)
Lime (Hyd rated)
Ferric Chloride
Ferrous Sulfate
Ferric Sulfate
Soda Ash
Sodium Hydroxide
Chlorine
Sodium Hypochlorite
Liquid Carbon Dioxide
Sodium Hexametaphosphate
Zinc Orthophosphate
Ammonia, Aqua
Ammonia, Anhydrous
Sulfuric Acid
Hydrochloric Acid
Activated Carbon Powdered
Activated Carbon Granular
Activated Alumina
Potassium Permanganate
Sodium Bisulfate (Anhydrous)
Sodium Silicate
Sodium Chloride
Polyelectrolyte
Diatomaceous Earth
Hagnesium
Sodium Chlorite
Sodium Hydroxide 76%
Sodium Bicarbonate
Calcium Hypochlorite
Small (<1 mgd)
Systems, $/Ton
$ 500
300
100
150
500
277
200
250
595
500
190
350
1,160
1,520
230
410
140
171
950
1,900
1,694
2,800
909
400
105
1,500
680
650
3,200
590
490
2,700
Large (>1 mgd)
Systems, $/Ton
$ 250
125
75
100
275
250
155
200
316
300
150
100
• 1,100
1,000
200
350
100
166
800
1,600
1,156
2,500
673
200
85
1,000
310
582
2,800
316
380
1,540
A-l
-------
DESIGN STANDARDS FOR
FILTRATION PLANTS
Texas Department of Health
-------
Design Standards for Filtration Plants
Texas Department of Health
(10) Flash nixing and flocculation equipment* capable of adequate
flexibility of adjustment to provide optimum flocculation
under varying raw water characteristics and rates of raw
water treatment, shall be provided.
(A) Where special types of equipment for rapid mechanical
mixing, softening, or sedimentation are proposed, the
manufacturer shall be required to meet the design
criteria in .004(b)(11).
(B) Sufficient facilities for coagulation and sedimentation
must be provided to clarify the water so that the
settled water turbidity is at a level so as to produce
a finished water which meets the turbidity limits
established by the Department's "Drinking Water
Standards."
(i) Settled water turbidities of less than 10
turbidity units are generally required to produce
a filtered water turbidity which meets the
requirements of the "Drinking Water Standards."
(ii) All turbidity measurements must be made in
accordance with the method specified in the
"Drinking Water Standards."
(11) In order to ensure continuous operation, basins for
flocculation and straight-flow sedimentation of coagulated
waters shall be at least two in number,, shall be designed
for parallel operation, and shall provide a total detention
period of at least six hours for clarification plants and
!|.5 hours for softening plants. Where detention periods of
less time are desired, engineering data shall be submitted
to the Department to Justify the lesser approach; however,
in no case shall the detention time in the clarification
zone of any unit be less than 2 hours. The requirement for
two parallel treatment plants (or trains) will be suspended
only with specific approval of the Department after the
review and evaluation of supporting material submitted to
substantiate the suspension.
(A) Facilities for sludge removal shall be provided by
mechanical means or by the provision of hopper
bottomed basins with valves capable of complete
draining of the units.
(B) Basins shall be so designed as to prevent the short
circuiting of flow or the destruction of floe.
Coagulated water or water from flocculators shall be
transported to sedimentation basins in such a manner
as to prevent destruction of floe. Piping, flumes,
troughs, gates, ports, and valves shall be designed at
4 ft/sec velocity in the transfer of water between
units.
(C) Sedimentation basins may be square, rectangular, round,
or other shapes approved by the Department. The length
of rectangular settling basins shall preferably be at
-------
least twice their width, with a side water depth of 10
feet to 12 feet in non-softening water treatment.
Square and round sedimentation may also be used for
clarification and softening plants; however, the
detention times In OOMbXll) Bust be observed.
CD) Surface loading for clarification units shall be 600 to
800 gallons per square foot per day, and for softening
units, 700 to 900 gallons per square foot per day.
(E) Settled water weir overflow rates shall be 15,000 to
22,000 gallons per foot per day.
(F) Sedimentation basins shall be provided with facilities
for draining the basin in a period not in excess of six
hours. In the event that the plant site topography is
such that gravity draining cannot be realized, a
permanently installed electric powered pump station
shall be provided to dewater the basin.
(12) Filters shall be gravity or pressure type.
(A) The design of gravity rapid send filters shall be based
on a maximum design filtration rate of 2 gallons per
square foot per minute. At the beginning of filter
runs for declining rate filters, a maximum filtration
rate of three gallons per square foot per minute is
allowed. The filter discharge piping shall be designed
with an orifice or other permanently installed flow
limiting device to ensure that the maximum filter rate
cannot be exceeded.
(B) Where high rate, dual or multiply media, gravity
filters are used, a maximum design filtration rate of 5
gallons per square foot per minute must be used. At
the beginning of filter runs for declining rate
filters, a maximum filtration rate of 6.5 gallons per
square foot per minute is allowed. The filter
discharge piping shall be designed with an orifice or
other permanently installed limiting device to ensure
that the maximum filter rate cannot be exceeded.
(C) Pressure sand filters shall be subject to the loading
provisions in .004.(b)( 12){A) for gravity sand
filters. When used, the pressure filters shall be
installed such that duplicate capacity is available
to furnish the design capacity with one filter out of
service. In any case, a minimum of two filter units
shall be furnished. The use of pressure filters shall
be limited to installations with less than 0.50 mgd
capacity.
(D) The depth of filter sand, anthracite, or other
filtering materials shall be between 21 inches and 30
inches, and this filtering material shall be free from
clay, dirt, organic-matter, and other impurities. Its
effective size shall range from 0,35 to 0.45 mm for
fine sand, 0.45 to 0.55 mm for medium sand, and 0.55
to 0.65 mm for coarse sand, and its uniformity
-------
coefficient shall not exceed 1.7. The grain size
distribution shall also be as prescribed by AWVA
Standards. Material for Dual or Mixed media filters
shall conform to AWWA Standards. (Refer to rule .001
Glossary of Terms.)
(E) Under the filtering material, at least 12 inches of
gravel shall be placed, varying in size from 1/16 inch
to 2 1/2 inches. The gravel is usually arranged in 3
to 5 layers such that each layer contains material
about twice the size of the material above it. Other
support material may be approved on a case by case
basis.
(F) The rate-of-flow of wash water shall not be less than
20 inches vertical rise per minute and usually not more
than 30 inches vertical rise per minute, which shall
expand the filtering bed 30 to 50 percent. The free
board in inches shall exceed the wash rate in inches of
vertical rise per minute.
(i) The water for backwashing filters shall be of the
same quality as that produced by the plant and may
be supplied by elevated wash water tanks or by
pumps provided for backwashing of filters only,
and take suction from clear wells. For
installations having a treatment capacity no
greater than 150,000 gallons per day, water for
backwashing may be secured directly from the
distribution system if proper controls and rate
of flow limiters are provided.
(ii) Rate of backwashing of filters shall be regulated
by rate-of-flow controllers.
(G) If surface filter wash systems are provided,
atmospheric vacuum breakers shall be installed in the
system supply lines above the overflow level of the
filters such that all-'water passes through them.
(H) Each filter unit shall be equipped with a manually
adjustable rate-of-flow controller, with rate-of-flow
indication or rate-of-flow indicators with control
valves. These devices are in addition to the flow
limiting device required above in .OOH(b)(12)(A) and
(B).
(I) Each filter unit shall be equipped with a device to
indicate loss-of-head through the filter box.
(J) Filter-to-waste connections, if included, shall be
provided with an air gap connection to waste.
(K) Filters shall be so located that common walls will not
exist between them and aerators, mixing, and
sedimentation basins or clear wells. This rule is not
strictly applicable, however, to partitions open to
view and readily accessible for inspection and repair.
-------
PILOT STUDY REPORT
-------
-------
R SYSTEMS
PILOT STUDY REPORT
TRIDENT* PILOT STUDY REPORT
LAKEPORT, CALIFORNIA
by
W. R. Conley
Consultant
March 12, 1987
Study Dates: October 22 through October 29, 1986
Operator: Bob Thompson
Mlcrofloc Products
= mitraflo!
PSR-26
-------
-------
PILOT STUDY REPORT
PAGE 1
TABLE OF CONTENTS
Page
2
2
3
I. INTRODUCTION
II. SUMMARY AND CONCLUSIONS
III. METHODS AND EQUIPMENT
IV. RESULTS AND DISCUSSIONS
A. Study Scope 4
8. Detailed Run Data.... 5
C. Net Mater Production 7
-------
PILOT STUDY REPORT
PAGE 2
I.
INTRODUCTION
Johnson Division conducted pilot tests to treat raw water from Clear Lake
at Lakeport, California for CWC-HDR in October of 1986, using the TRIDENT*
trailer mounted pilot plant. The pilot plant is described in Part III of
this report. The purpose of the test was to determine if the TRIDENT*
process is suitable for producing a low color, iron, manganese and
turbidity effluent and has acceptable levels of THM's while at the same
time giving a net production of over 94% with filter runs of over 24
hours. The results of the pilot tests along with conclusions and
recommendations are the subject of this report.
II. SUMMARY AND CONCLUSIONS
The raw water containing low color, moderate alkalinity and moderate to
high turbidity was treated with alum, permanganate and polymer. The raw
water because of the moderate alkalinity and low color was easily
coagulated. The TRIDENT® AQUARITROL* which automatically adjusts the
chemical feed to a given turbidity setpoint worked well during the test.
The results were uniformly good in that the clarifier effluent turbidity
was usually below 2 NTU even though the raw water turbidity was as high as
130 NTU with an average for the four runs of 33 NTU. The clarifier
effluent filtered well to give an excellent filtered water with long filter
runs. The final filter effluent contained less than 5 color and the
turbidity ranged from 0.07 to 0.10 with an average for the four runs of
0.09 NTU. Filter runs averaged 50 hours for the two runs that were
completed. One incomplete run was for 11+ hours and was terminated when a
storm swamped the inlet pump. The last run was terminated at 20 hours when
the inlet pump quit. As the filter was operated at 5 gpm/ft?, this is
excellent performance. The Adsorption Clarifier" had a flush cycle of 4 to
4.5 hours while operating at 10 gpm/ft^. More detailed data are presented
under details of the runs and in the logs and figures. Alum feed usually
ranged from 12 to 13 mg/L. A cationic polymer {Nalco 8103) was fed at the
rate of 6 to 7 mg/L and a non-ionic polymer (Magnefloc 985N) was fed at the
rate of about 0.3 mg/L. Permanganate was fed at the rate of 0.5 mg/L.
Finally, powdered activated carbon was fed at the rate of 25 mg/L for 6
hours on Run #3. Results in terms of filter run lengths, headless and
water quality were about the same as without carbon. This indicates that
carbon can be fed without operating problems to control tastes and odors
and to remove other organics to some degree.
Five samples (two of the raw water and 3 of the filtered water) were
collected for analysis of trihalomethane potential. This test gives a good
indication of potential problems with THM's in the distribution system,
Fortunately, all of the results were excellent showing that organic
precursors are very low in the raw water and that treatment reduces the
precursors even further. The analytical report from an outside lab is
appended to this report.
Because of the low alum feed used during the study, the alkalinity and pH
were reduced to only a small extent. To bring the water back to the
t
-------
PILOT STUDY REPORT
PAGE 3
alkalinity and pH of the original raw water will require about 5 to 10 mg/L
of soda ash or sodium hydroxide. The pH control chemical should be added
downstream of the filter. Restoring the filtered water to the original pH
before treatment will minimize corrosion problems. A further refinement
will be to maintain the pH and alkalinity in such a way as to hold a
slightly positive langlier index. In order to do this the plant should
have the capability to feed 25 to 50 mg/L of sodium hydroxide or soda ash.
Polymers were fed to reduce alum feed and produce a better floe. Details
of the chemical feeds will be found under details of the runs and attached
log sheets.
Th« data show that the TRIDENT* plant is an excellent treatment device for
the water encountered during the test period.
III. METHODS AND EQUIPMENT
The pilot tests were conducted using the Mi.crofloc trailer mounted TRIDENT*
demonstration'pi ant. Technical data for the plant are listed in Table 1.
The demonstration unit is a small scale replica of a complete treatment
system. The plant includes complete raw water chemical feed capabilities,
is fully instrumented and has an integral clearwell to provide water for
backwash of the filter.
As the influent from the pipeline enters the pilot plant, chemicals are
injected in-line. The first stage Adsorption Clarifier™ combines the
functions of mixing, contact flocculation and solids removal. This system
provides a well conditioned, clarified water to the second stage which is a
mixed-media filter. The clarifier consists of a packed bed of buoyant
media. The flow is upward through the clarifier, and downward through the
mixed-media filter.
The TRIDENT* demonstration plant runs at a constant, but adjustable flow
rate. Automatic flow control devices ensure constant flow throughout a
run. For this study the flow rates were 10 gpm/ft2 through the clarifier
and 5 through the filter;
Because of scale considerations, the clarifier is larger than needed for
flow to the ,filter. Part of the clarifier effluent is split off to waste
and the other part is passed through the mixed-media filter. The bypass
and filter effluent flows are monitored with rotometers. The clarifier and
filter are each equipped with a pressure gauge and recording device to
monitor headloss development. Cleaning of the Adsorption Clarifier1" is
accomplished in a multiple step process:
1. The influent to the clarifier is shut off.
--•--.
2. The clarifier media is fluidized by injecting air into the bottom
of the clarifier chamber.
-------
PILOT STUDY REPORT
TABLE 1
TRIDENT* DEMONSTRATION PLANT
TECHNICAL DATA
I. Chemical Feed System
A. Number of Pumps/Tanks
B. Type Liquid Metronics;
Positive displacement,
variable stroke and speed
II. Adsorption Clarlfier
A. Plan Dimensions:
B. Area, sq. ft.
C. Media Depth, in:
D. Media Type:
III. Filter Design
A. Plan Dimensions:
B. Area, sq. ft.:
C. Media Depth, in:
D. Media Configuration
1. Anthracite Coal (1.2mm)
2. Silica Sand (0.6mm)
3. High Density Sand (0.25
-------
WIfC
<>
.V
**.
^—-
-------
PILOT STUDY REPORT PAGE 4
3. After the media is totally fluidized, about 1 minute, the
influent flow is restarted. The influent water flushes solids
upward and out through a waste trough. This flush cycle was 4
minutes.
4. The flow of air to the clarifier is stopped. The flow of water
through the clarifier to waste was continued for an additional 4
minutes. Upon completion of the flush cycle, the flow from the
clarifier is diverted from waste back to the mixed-media filter.
Air for the flushing process is provided by a small compressor.
The flush cycle can be initiated by any of four means:
1. Manual initiation by pushbutton.
2. Initiation by loss of head across the Adsorption Clarifier"
media. The differential pressure across the plastic media is
sensed by a differential pressure switch. This pressure switch
is normally set to be initiated at a pressure difference of 4
feet of water.
3. Initiation by time clock which will provide flushing at preset
intervals. The time clock resets after the completion of each
flush.
4. Initiation by the mixed-media filter backwash cycle.
The mixed-media filter was backwashed with treated water at a rate of 16
gpm/ft2 for a total of 7 minutes. Supplemental surface agitation if
provided with a fixed surface wash device. A backwash cycle is initiated
either manually or automatically by a high headloss condition as sensed by
a vacuum switch set at 8 feet headloss.
IV. RESULTS AND DISCUSSION
A. Study Scope
There were 4 runs. Turbidity, color, and pH were measured at
frequent intervals and a few samples were run for iron, and manganese
by a field test kit. Other samples were sent to an outside lab for
potential trihalomethane analyses. Treatment chemical feed rates,
clarifier headloss and filter headloss were also recorded at frequent
intervals. All runs were at 10 gpm/ft2 on the clarifier and 5
gpm/ft2 on the filter.
At the 10 gpm/ft2 clarifier rate the flush cycle consisted of 1
minute of air fluidization, 4 minutes of air and water flush and 4
minutes of rinse to waste. The filter wash was for a period of 7
minutes at 16 gpm/ft2.
-------
PILOT STUDY REPORT
PAGE 5
B.
Data were written on log sheets by hand, but headloss data were
recorded continuously for 2 of the runs. The log sheets and recorded
headloss data are presented in this report as appendices.
Details of the Runs
Run #1
For this run the water was treated with approximately 13 mg/L of
alum, 7 mg/L of cationic polymer (Nalco 81093) and 0.3 mg/L of a
non-ionic polymer (American Cyanamid 985-N), and 0.5 mg/L of
permanganate. The alkalinity of the raw water was moderate (137
mg/L). The pH of the raw water was about 8.5 while the treated was
7.7. The raw water turbidity varied from 26 to 130 NTU with an
average of 46. The apparent color was 15 units, the iron 0.2 mg/L,
the manganese was <0.l mg/L and the temperature was about 60°F.
Treatment results were excellent. The filtered water had a turbidity
between 0.08 and 0.10 with an average of 0.09 NTU. The effluent Iron
was <0.1 and the manganese <0.1. The effluent color was <5. The
filter run was 11+ hours and was terminated when a storm caused the
influent flow to stop. The clarifier flush interval was 4 hours. It
is interesting to note that the Adsorption Clarffler1" was doing most
of the work in that the clarifier effluent ranged from 0.6 to 2 NTU.
This of course explains the very long filter run of 47 hours at the
high filter rate of 5 gpm/ft2. See Log Sheet #1 for more details.
Run #2
The raw water turbidity ranged from 17 to 63 with an average of 35
NTU. The alum was about 13 mg/L. Chemical treatment was almost the
same as for Run #1, with essentially the same treatment results. The
final turbidity ranged from 0.07 to 0.11 with an average of 0.09
NTU. The filter run was 47 hours to a final headloss of 6.7 feet.
Two samples were taken for potential trihalomethane formation. The
raw water gave a potential of 31 mg/L. The filtered water gave a
potential of 9.1 mg/L. Both results are excellent and show that
organic precursors in the raw water are very low. The color of the
water gives a rough idea of the organic precursors, but note that all
results for color were apparent color. The true color was no doubt
considerably lower than the 15 units shown on the log sheet. The
filtered THM potential is very low. Probably the organic precursor
was removed by coagulation with alum and polymer. The observed 71%
removal of precursor by coagulation and filtration is higher than
usually observed. This may mean that part of the precursor was
associated with the raw water turbidity which was very efficiently
removed. It may also be we11-to note that analytical results in this
range may have a relatively large experimental error. If the samples
are representative of the raw water throughout the year,
trihalomethanes should not be a problem for future design unless the
EPA changes the current standard. See Log Sheet #2 for more details'.
-------
PILOT STUDY REPORT PAGE 6
Run 13
The raw water turbidity ranged form 23 to 27 NTU with an average of
25 NTU. The chemical feeds were about the same as for Runs #1 and #2
except that powdered activated carbon was fed at the rate of 25 mg/L
and permanganate feed was discontinued for the last 6 hours of the
filter run. Continuous recording of the clarifier and filter
headloss show no effect of the carbon as compared with the earlier
part of the run. Log Sheet #3 shows no difference in clarifier or
filter turbidity as compared with no carbon feed. Treatment results
were about the same as for Runs #1 and #2. The final effluent
turbidity ranged from 0.07 to 0.09 with an average of 0.08 NTU. The
filter run was 53 hours to a headloss of 6.3 feet. The flush cycle
was 4.5 hours, but the clarifier effluent remained good at 0.5 to 0.9
NTU.
Two samples of filtered water and one sample of raw water were taken
for potential trihalomethane formation analysis. The results were
excellent as for Run #2. The raw sample ran 16 mg/L. The filtered
sample that was collected during the period of powdered activated
carbon feed was a very low 4.6. This result is lower than would be
expected and should be viewed with caution. The result could be
correct but more data should be obtained before forming conclusions
about the effectiveness of this rather low active carbon dose. The
filtered sampled collected during the period of the run when carbon
was not fed ran 20 mg/L. Note that for this run the filtered water
sample ran higher than the raw water. This could be caused by swings
in the organic content of the raw water or experimental error.
More detailed data will be found on Log Sheet #3, the headloss
recordings and the analytical report for trihalomethane formation
potential.
Run #4
The raw turbidity ranged from 23 to 26 NTU with an average of 25.
Chemical feeds were similar to the other runs except no carbon was
fed. Permanganate was fed at 0.5 mg/L as in Runs #1 and #2.
Treatment results were excellent as before. The filtered water
turbidity ranged from 0.08 to 0.1 with an average of 0.09 NTU. The
filter run was 20 hours to 3 feet headloss. The run was terminated
by failure of the inlet water pump. The clarifier flush cycle was
4.5 hours and the clarifier effluent ranged from 0.6 to 1 NTU. More
detailed data will be found in the attached Log Sheet #4, and
recorded headloss graphs.
-------
PILOT STUDY REPORT
PAGE 7
C. Net Water Production
Net production is defined as the fraction of the rated filtered water
flow that is available for use. Net production for a flush cycle of
4.25 hours and 50 hour filter runs was 95.5%.
The filter backwash was 0.75% of the rated flow. The out-of-service
time for filter backwash and clarifier flush was 3.8%. The waste
unfiltered water for clarifier flush was 3.1% of the rated, flow.
-------
AnJbb
ANALYTICAL LABORATORY
A DIVISION Or OEWANTE » StOWtU.
I9t4 S STREET. SACRAMENTO. CALIFORNIA 95814 • 916-447-2948
January 30, 1987
Sample Date: 10/24 & 10/28/86
Sample Rec'd. Date: 10/30/86
Report #111089
Culp-Wesner-Culp
3461 Robin Lane
Cameron Park, CA 95682
Attn: Jerry Costah
SAMPLE DESCRIPTION ANLAB ID*
Raw Water Clearlake 111089-1
Filtered Effluent 111089-2
Clearlake Water 10/28 111089-3
Filtered Effluent
Run #3
Clearlake Water 10/28 111089-4
Filter Effluent
Run #3
Clearlake Water 10/28 111089-5
Run S3
TOTAL
rRIHALOMETHANE POTENTIAL
uq/1
31
9.1
20
4.6
16
Data Certified by
Report Approved by
This report is applicant* only to ttw sampla rwMiMd By m« laboratory. Tha liability of th* liooratory u limitMl to ttw Mnount paid lorthto rwort TMa rtport Is tor m«
«xctu»i«* us* of tn« cli«nt to whom it is «tdrtMW« «nd upon th» condition mat ttw client aaaurrMi all liability tor ttw> (Mittwr tflsinbutlon of trw nport or its comanls.
-------
•sir
"
* 0.
Eft O
£• ui
'-O
a
13—
""— *._ .^
,
— -e
-»••-•„
"S-.
f
at
A,
"""
Or
•m
O O
r>
d
L
l-a>
i
L *•
tf|
O
-------
a;
Si <
l/j o
Es
i I ! i
*- K> O) 00
c-j
I'.i
to
rVi
. i
c-4
"&...
i i
o
CO
O
-L
I
I-
M
Ar, t
LJU
•T- O
co
133d)
-------
W
a;
<
CL
o
ag
&3
»-
Ss
*{=>—_._ _
a..._
D-— _
L c J
i tr*
-*G
t •*»
—i
1
r
U»
D
o o
-------
05
o
flJS
1i <
ix
s
""
e-u
CO
U)
a.
O
x
2P
IN
KJ
02
UJ
C'-J
I
>n
«4
/
V
-------
.-1
if
!••
*9
>1
j "IN
^ fl ! J
s i »
a ! «
1 i S.
*
i
i
!1
•a
Sk-
•
X
I
V
^t.
i!
f
«!l
•1.1
&
•• •'•
-
•v i
* a
* • • u
10 . U
\
G
S -3i
•5 1
i §!
a
§<
H
e i
§ i
* M
4 il1
ig'l
M3 '
! »
33
,'B
8ii
«B
*3
Is*
s"
; 81
$ fl *
r*
1 1 Jp'
i !
. . s
! 5
.1'
i -'i
w . ,
! i
i
i>
b
f
it
0
Q
**
.
1
N
i
fte
>
a
5
ff-
«:
o
*
Q
•a-
—
^
5
j
c<
O
IQ
Q
*
1
(
«
r*
•n
V)
'£
r-
j
\>
\
I
P
flO
O)
5
6
Vn
—•
6
Q
6
^
rt
•«
-
Vi
i
«
s
a
U
•4
u
f-
to
*4
•
4
0
«a
^•B
: 9;
O
0
6
- J
5
n
ft
i
t
i
1
.4)
^s,,
f
«
-?
i
X|
^
j
I
U
5
•^
"^
Q
ci
•
1
>v
3
-•
V
•
1
r
t
•
•.
- -
' ^
•
V
•
-------
J»_
i
I
I
I
as
as
i
3
Ba*
» : ss
3
B
*'J
!li
'!!
8
S
88
V*
j
M
c
i
Vi.
V-
u
U
fcft
L
V-
\0
w
O
O
r*
o
r.
in
b
V
L-
N
QO
\
V
v
-------
e
i
I
IV •
HI !
s |SS
= I 5
as
s
IE
*
tt
-------
!
ii-
31 w
S36
in
i
31
fi
a
I
I
«
i
il
IE
6 £
3
'ill
c !
i-s.fc
il
0
K
U
to
.a
cc
V)
-4
v>
on
o-
-------
Information Resources Center
US EPA (3404)
401 M Street SW
\Mashington, DC 20460
-------
'<%
------- |