EPA-430/1-77-004
PHYSICAL-CHEMICAL
TREATMENT TECHNOLOGY
TRAINING MANUAL
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF WATER PROGRAM OPERATIONS
-------
-------
EPA-430/ 1-77-004
August 1977
PHYSICAL-CHEMICAL TREATMENT TECHNOLOGY
This course is offered for professional personnel
engaged in the selection, design and operation of
physical-chemical wastewater treatment facilities.
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Water Program Operations
National Training and Operational Technology Center
-------
DISCLAIMER
Reference to commercial products. trade names. or
manufacturers is for purposes of example and illustration.
Such references do not constitute endorsement by the
Office of Water Program Operations. U. S. Environmental
Protection Agency.
-------
CONTENTS
Title or Description
Outline Number
Basic Principles of Coagulation
1
Solids Removal Processes
2
Methods for Achieving High BOD and Solids Removals from
Conventional Plants
3
Basic Principles of Carbon Adsorption
4
Basic Principles of Ion Exchange
5
In-Depth Filtration
6
Reverse Osmosis
(>
7
Electrod ialysis
8
Status of Chemical Oxidation in Wastewater Treatment
9
Removal of Phosphorus and Colloidal Solids by Coagulation in
Conventional Treatment
10
Current Status of Phosphorus Control Tedlnology for Municipal Wastewater
11
Composition of Sludges
12
Sludge Incineration
13
Reuse of Municipal Wastewater Effluents
14
-------
BASIC PRINCIPLES OF COAGULATION
I
INTRODUCTION
A The natural capacity of our surface waters
to assimilate wastes is being increasingly
taxed in many areas and even exceeded in
some. Methods to treat these wastes to
decrease pollution loads must be developed
to preserve our natural water resources.
A great variety of processes are capable
of reducing this pollution, and all of these
are currently being carefully examined.
An important process is coagulation and
flocculation, the subject of this lesson.
B Coagulation is almost exclusively a process
for the removal of colloidal and suspended
solids. It has application in many areas,
such "as:
1 Removal of BOD and COD associated
with suspended solids from either
primary or secondary waste treatment.
2 Removal of phosphate compounds,
.utilizing both precipitation and coagulation.
3 As a pretreatment process for in-depth
filtration.
4 As a clarification step to precede such
other processes as electrodialysis,
reverse osmosis, carbon adsorption and
possibly others.
5 As a concurrent process with flotation
or micro-air flotation.
C Each of the above applications imposes
different product quality standards which
require individual design, mode of
operation and economics.
II
HISTORY OF COAGULATION
A The principles, methods, and materials
for clarifying water have an ancient
~~ritage dating back to the earliest recorded
date of about 2000 B. C.
SE.TT.3.7.77
B Methods described by Sanskrit and
Egyptian inscriptions demonstrated a
knowledge and use of such processes as,
sedimentation, coagulation by both
organic as well as inorganic compounds,
filtration through sand and porous vessels,
and disinfection obtained from such
metallic containers as copper or bronze.
C The process of coagulation followed by
filtration, as we now recognize them,
really had their inception in the late 19th
Century.
III
CURRENT STA TUS
A Coagulation and flocculation, generally
followed by filtration, is by far the most
widely used process for the production of
potable water from turbid surface water
supplies.
B In the past, domestic wastes were rarely
treated by coagulation. Industrial waste
treatment frequently required coagulation.
C Use of coagulation and flocculation will
predictably be more commonly used in
the future: removal of BOD in suspended
solids, removal of phosphates, pre-
treatment for filtration and other processes.
IV
DEFINITION
Historically, the terms coagulation and
flocculation have been used indiscriminately
to describe the process of removal of
suspended solids from water. There is,
however, a clear distinction between the two
terms.
A Coagulation is the destabilization of a
dispersion by reducing the forces tending
to keep the particles apart. Operationally,
coagulation is achieved by adding the
appropriate chemical which 1) reduces
the forces of repulsion, and 2) increases
the number of particles and hence increases
probability of particle to particle contact.
1-1
-------
Basic Principles of Coagulation
B Flocculation is the formation of large or
settleable particles from the destabilized
colloidal sized particles. In contrast to
coagulation, where the primary forces are
electrostatic or interionic, flocculation is
achieved by particle contact, adsorption,
bridging and physical enmeshment
mechanisms.
v
NA TURE OF PARTICLES
A Particles in wastewater may be broadly
classified according to their origin as
inorganic mineral matter or organic
carbonaceous material.
B Particles may additionally be classified
according to their size which may range
from molecular dimensions to 50 microns
or larger. The fraction;; 10 microns will
settle out given sufficient time. Particles
~ 1 micron, which are classified as
colloidal, will remain suspended for very
long times.
C Colloids possess the distinguishing
characteristic of high ratio of surface area
to mass. For large particles, where this
ratio is low, mass effects such as
sedimentation under gravity predominate.
For colloidal materials, where ratio is
high, the properties associated with the
surface of the particle, such as electric
charges, and ionogenic groups, become
more important.
D Colloids may be further classified as
hydrophobic and hydrophilic.
E A partial list of particles in wastewater
would include:
1 Organics such as polysaccharide~,
proteins, microorganisms, etc.
2 Inorganic minerals such as clays, silica
3 Inorganic precipitates, hydroxides,
metal oxides.
1-2
VI
STABILITY AND INSTABILITY FACTORS
A Stability refers to the inherent property
of colloidal particles to remain dispersed
despite passage of time.
B Instability describes the tendency of
particles to coalesce whenever particle
to partiCle contact is made.
C For hydrophilic colloid systems, stability
is maintained by the phenomenon of hydration
in which water molecules are attracted
to the surface of particles and act as a
barrier to contact between particles.
D Stability of hydrophobic particles is due
to the phenomenon of the electrical double
layer.
E Each particle suspended in water consists
of several components in addition to the
primary particle. At least three distinct
layers can be described.
F A consequence of the electrical double
layer is to create regions of electrical
potential in a bulk solution that nominally
has a zero potential. The most interesting
of the several potentials is the Zeta
potential.
G Electrical double layers prevent close
approach of particles to each other. Both
the thickness of the layer and magnitude.
of the surface charge and density are
important to stability.
VII
INSTABILITY FORCES
A Forces which operate in the opposite
direction to cause particles to be attracted
to each other are called instability forces.
B Brownian motion, inherent to all colloidal
particles ,promotes instability by causing
particle contacts. Brownian motion is
important only for particles with diameters
100 millimicrons or less.
C A force which is always one of attraction
between particles is Van. der Waals force.
-------
Basic Principles of Coagulation
This force is operative only within short
distances of the order of atomic dimensions
but may be greater for colloidal particles. '
VIII
MECHANISM OF COAGULA TION
A Objective of coagulation is to destabilize
the colloids by:
1 Addition or presence of electrolytes
which tends to decrease the thickness
of the double layer.
2 Addition of ions of opposite charge which
tends to neutralize the charge on the
particle. The coagulating power of an ion
increases sharply with increase of
valence.
3 Colloids of opposite charge will cause
. a coagulation termed mutual coagulation.
4 Coagulation may be obtained by "bridging"
such as occurs with polymeric flocculants
(polyelectrolytes) .
5 Addition of metal coagulants such as
alum or iron salts.
IX
CHEMISTRY OF METAL COAGULANTS
A Coagulation by aluminum and iron salts is
not obtained by ionized tripositive aluminum
or iron ions but by their hydrolysis products.
B Aluminum salts react with water in a
series of chemical reactions described
as hydrolysis.
[AI (H20)6] 3+ +H20 = [AI (H20)5 OH ] 2++ H30 +
[AI (H20)5 OH] 2++H20= [Al(H204)4 (OH)2] 1++ H30 +
C Hydrolysis products undergo polymerization
. reactions such as the following:
[ ("20)4 .
-------
Basic Principles of Coagulation
a Broadening pH zone of coagulation, for
example SO 4 =
b Shifting the ptI zone of coagulation, for
example PO 4 =, shift to more acid pH's
D Nature of Particles
1 In general, coagulation is controlled more
by the coagulant and chemical composition
of the water than by the nature of particles. XII
2 A certain minimum of coagulant must be
added for any suspension to provide
enmeshing mass of floc.
3 Dosage of coagulant does not increase
linearly with increase of turbidity.
4 Paradoxically, very high turbidities are
easier to coagulate than very low
turbidities.
5 Hydrophobic particles are easier to
coagulate than hydrophilic.
E Effect of Coagulant
1 A lum not widely used.
2 Iron salts have broader pH range for
good coagulation.
3 Iron salts preferable for color removal
at acid pH's and for coagulation at
softening pH's.
F Effect of Physical Factors
1 Temperature is only physical factor to
consider.
2 Temperature affects viscosity of water
and rate of chemical reactions.
3 Very low temperatures, < 50 C, adversely
affects coagulation - requiring more
coagulant and longer time.
XI
MIXING
A Two stages of mixing or other means to
create turbulence are generally used.
1-4
B A rapid mixing to uniformly mix the
coagulant and promote contact with
particles. Thirty to 60 seconds are
usually provided.
C A slower and longer stirring to obtain
flocculation by floc growth. Detention
times as little as 10 minutes but more
frequently 30-60 minutes are used.
FLOCCULA TION
A The rate of flocculation is proportional to
the mean velocity gradient (G), the
detention time, the concentration of
particles and the particle size.
~1
C A 3 '2
Gt~vv
where
-1
G = mean velocity gradient - sec
CD = drag coefficient - for flat blades
made of wood CD is taken to be 1. 8, for
metal blades 1. 2
2
A = area of paddles - ft
v = mean velocity of the paddles relative
to the water ,ft/ sec, generally taken as
Vw = 0.75 Vp
iJ = kinematic viscosity of the water -
ft2/sec
3
V = volume of flocculator - it
C Time for flocculation can be expressed
by product of "G" and "t" which yields a
dimensionless value Gt.
1 Gt values maS range within the limits
of 104 and 10
D Design Criteria
1 Rapid mix compartment
a Detention time - 30-60 sec
. "., -1
b Deslgn G - 500 sec
-------
----
Basic Principles of Coagula0-0n
c Gt - 30, 000
2 Flocculation chamber
a Detention time - 10-30 min
b Design "G" - 20-30 sec-l
c Gt - 54, 000
Gt values up to 200,000 have been
suggested to produce a dense sludge.
REFERENCES
1 Baker, M. N., Sketch of Histroy of Water
Treatment, JA WWA, 26, 902 (1934).
2 LaMer, V. K., Coagulation Symposium -
Introduction, J. Colloid Sci., 19, 291
. (1964). -
3 Matijevic. E.. Detection of Metal Ion
Hydrolysis by Coagulation, J. Colloid
Sci., 20, 322 (1965).
4 Matijevic, E., Mathai, K. G., Ottewill, R. H.
and Kerker, M., Detection of Metal Ion
Hydrolysis by Coagulation. III. Aluminum
J. Phys. Chem. 65, 826 (1961).
5 Packham, R. F., The Theory of the
Coagulation Process - A Survey of the
Literature, I The Stability of Colloids,
Proc. Soc. Wtr. Treatment and
Examination, ..!:.!, 50 (1962).
6 Priesing, C. P., A Theory of Coagulation
Useful for Design, Ind. Eng. Chem.,
54, 38 (1962).
7 Hudson, H. E., Jr., Physical Aspects of
Flocculation, JAWWA, 57, 885(1965).
8 Packham, R. F., The Theory of the
Coagulation Process - A Survey of the
Literature. II Coagulation as a Water
Treatment Process. Proc. Soc. Wtr.
Treat. & Exam., ..!:.!, 106 (1962).
9 Marion, S. P. and Thomas, A. W., Effect
of Diverse Anions on the pH of Maximum
Precipitation of "A luminum Hydroxides ",
J. Colloid Sci., .!., 221 (1946).
10 Packham, R. F., The Coagulation Process.
A Review of Some Recent Investigations,
Proc. Soc. Water. Treat. and Exam.,
g, 15 (1963).
11 Tenney, M. W. and Stumm, W., Chemical
Flocculation of Microorganisms in
Biological Waste Treatment, J. Wat.
Poll. Cont. Fed., 37, 1370 (1965).
12 Camp, T. R. and Stein, P. C., Velocity
Gradients and Internal Work in Fluid
Motion, J. Boston Soc. Civ. Engrs.,
30, 219 (1943).
13 Rich, Linvil G., Unit Operations of
Sanitary Engineering, John Wiley &
Sons Inc., New York (1961).
14 Nordell, E.. Water Treatment for
Industrial and Other Uses. 2nd Ed. ,
Reinhold Publishing Corp., New York
City (1961).
15 Powell, S. T., Water Conditioning for
Industry, McGraw-Hill Book Co. Inc.,
New York City (1954).
This outline was prepared by Jesse M.
Cohen, Chief, Supv. Res. Chemist,
Municipal Environmental Research
Laboratory, USEPA, Cincinnati, Ohio
4526:i
Descriptors: Wastewater Treatment,
Waste Treatment, Water Pollution
Treatment, Coagulation, Flocculation,
Te rtiary Treatment
\
1-5
-------
SOLIDS REMOVAL PROCESSES
I
INTRODUCTION
If one were to rank wastewater treatment
operations in order of overall importance,
separation of solids from liquids would
probably assume major importance. This
operation may be required to remove
objectionable components from a waste,
to render the waste more amenable to
subsequent treatment or disposal, or
to support another treatment operation.
II
SEDIMENTATION
Sedimentation tanks are employed for
various applications which have one
common objective - the removal of solid
matter from a flowing liquid. The solid
matter may have been present in direct
suspension, such as in munic::ipal sewage,
or may alternatively be a precipitate
resulting from prior chemical treatment,
such as lime precipitation, alum, or iron
coagulation.
A Horizontal Flow Design
1
The principal basis of the earliest
designs of horizontal-flow sedimen-
tation tank was the "retention time" .
The objective was to move water so
slowly through the tank that there would
be ample time for settlement. The
nominal detention period allowed was
usually more than four hours.
a
However, each drop of water did
not take the same time to travel
from inlet to outlet, since some
degree of "short-circuiting" was
bound to occur.
b
A typical pattern of flow under
ideal conditions in an elementary
design of a horizontal-flow tank
is represented by Fig. 1.
Diagram (a), solid line, rep-
resents the most favored flow-
path, while the dotted lines
SE. AE. s1. 5. 7.77
indicate typical eddies which are
induced in the remainder of the tank
volume. Diagrams (b) and (c)
represent the flows which result,
respectively, to entering water
being warmer or colder than the
water in the tank.
FIGURE 1.
-~
'.- -r ,-- ) .
...........---.......---"
. .
(A)
~-
--- RiJ"t------------.----."
'"'-- -~. ----.-- -.--.-1
... .. .. ..
(B)
-~-
'--..--- - -.- - ----
- - -
(C)
FIGURE 1 SEDIMENTATION TANK OF HORIZONTAL
- FLOW TYPE UNDER DIFFERENT CONDITIONS
OF RELATIVE TEMPERATURE
2-1
-------
Solids Removal Processes
2
The basic design which has been most
widely used for sewage treatment is
a tank circular in plan. The inflow
is introduced at the center of the
tank and the outflow is collected at
a peripheral launder . As the general
direction of flow is mainly a radial
spread from the center, it is reasonable
to regard this design as in the hor-
izontal-flow category as in Fig. 1.
3
Vertical-Upflow Design
a
The improved performance
obtained in upflow tanks has led
to a variety of new designs, in-
corporating variants of the upflow
principle. These new designs
are designated by various proprietary
names - Accelator, Clarifow,
Flocsettler, Reactor - Clarifier,
etc. One of the earliest of the
upflow tanks was the Spaulding
Precipitator, shown in Fig. 2.
In this design, the flow is in-
troduced into the center of the
tank and flows upward through a
blanket of previously formed
solids.
APPROXIMATE
- TOP LEVEL
OF SLUDGE
BLANKET
........-.
SLUDGE
FIGURE 2 SPAULDING PRECIPITATOR
2
b The principal advantages of the
upflow versus the horizontal tanks
are (1) improved flow control, and
(2) sludge blanket effect. Salt-
injection tests performed on a variety
of tank designs clearly show the
superior efficiency of the upflow
principle in the matter of flow control.
The radial-flow still largely used in
. sewage treatment is distinctly in-
ferior, being subject to short-
circuiting.
4 Tube and Lamella Settlers
In an ideal settling basin, as defined
by Camp, the paths of all discrete
particles will be straight lines, and all
particles with the same settling velocity
will move in parallel paths.
a The settling pattern shown in Fig. 3
would be the same for all longitudinal
sections. It is apparent from this
that as the interval (h) is reduced,
the size of the basin required to
remove a given percentage of the
incoming settleable material
decreases.
SURFACE AREA A
o
J::
-......:
""-
"
"
"
..........
v
----- - ---.,
"" I vo
~
"
1
I:
DIRECTION OF FLOW a
---
..I
~I
L
Lo
FIGURE 3
IDEALIZED SETTLING PATHS OF DISCRETE
PARTICLES IN A HORIZONTAL FLOW TANK.
-------
Solids Removal Processes
b Many devices have been proposea
since this principle was first proposed
by Hazen in 1904 and further de-
veloped by Camp in 1946.
c None of the proposed devices were
accepted commercially until the
recent introduction of the device
called the tube settler. These con-
sist essentially of closely packed
small diameter tubes, 1-4 inches
in diameter and 2-4 feet in length,
inclined at some angle to provide for
removal of sludge as the water flows
upwar.d through the tubes. Detention
times are in the order of 6 minutes
and less. The tubes provide as much
as 24 hours of sludge storage,
depending of course, on the amount
of suspended soFds, and sludge is
readily removed by gravity drainage.
1 A schematic diagram of the
tube settler integrated into a
complete clarification device
which consists of coagulation-
flocculation, a tube settler
unit and a mixed me dia
filter is shown in Fig. 4.
ALUM
RAW WATER
--
FIGURE 4 SCHEMATIC DIAGRAM OF API'ARA TUS USED IN
FIELD TESTS OF 1\J8E SETTLER
2
More recently, a device
consisting of parallel plates.
rather than tubes has become'
available. The device is .
called a lamella. Flow is
co-current in contrast to the
countercurrent flow of the
tube settlers. The plate
arrangement in a tank is
shown in Fig. 5.
FIGURE 5 LAMELLA SEPARATOR
III
DISSOLVED AIR FLOTATION
A
The use of dissolved air to float suspended
solids was first used in industrial
operations. In recent years, the process
has been adapted to domestic wastewater
and particularly, to sludge thickening.
The process achieves the separation of
suspended particles by attachment of
gas bub.b.les to the suspended particles,
thereby reducing the effective specific
gravity of the particles to less than
that of the water.
3
-------
~olids Removal Processes
B
A flotation system using the pressurization
and de-pressurization sequence consists
of the following elements (Fig. 6).
1
2
3
4
5
6
Pressurizing pump
Air injection facilities
Retention tank or contact vessel
Back-pressure regulating device
Flotation device
Facility for addition of chemicals
if needed
!.PPLU!.NT
UNPREeOURIZED
PORTION 01" pe.eD
"
e.PPLueNT
UNPR OURlzeD peeD
PAR'TIALPRESBURIZATION 01" FeeD
PARTIAL PReS8URIZATION 01" I!PPLUeNT
MeTHODO eMPLOyeD POR PARTIAL AND TOTAL PRESSURIZATIONS
PIGURE e
4
C Advantages and Disadvantages
1
Much reduced retention time -
10-20 minutes
Greater solids concentration in float
than in settled sludge
Greater efficiency of solids recovery
Offers mechanical control over
the process
Increased cost of operation for
pumping, etc.
Need to remove top and bottom solids
2
3
4
5
6
IV SCREENING DEVICES
A Microscreening is a form of simple
filtration by straining (Fig. 7).
1
These mechanical filters consist of a
rotary drum which revolves on a
horizontal axis. The peripheral surface
of the drum is covered with a stainless
steel fabric. The effectiveness of the
woven mesh screen for retaining fine
particles is dependent on the size of
the openings in the screen and on the
pattern of the weave.
2
Influent enters the open end of the drum
and is filtered through the fabric with
the intercepted solids being retained on
the inside surface of the fabric. As
the drum rotates, the solids are trans-
ported and continuously moved at the
top of the drum by pumping strained
effluent, under pressure, through a
series of spray nozzles which extend
the length of the drum. The solids and
wash water are collected in a central
trough within the drum and discharged
through a hollow axle.
3
The microstraining device is available
in several unit sizes ranging from 5'
in diameter and l' width with a capacity
of 0.05 to 0.5 mgd to 10' diameter and
10' wide with a capacity of 3-10 mgd.
B
Microscreening devices have found their
greatest application to treatment of
river waters, and information on their
performance on wastewaters is quite
scarce.
-------
Solids Removal Processes
1
The effect of aperture size on re-
moval efficiencies .!U1d flow rate
is shown below.
Fabric
Removal Efficiency
Flow
Solids
2
BOD gals/hr/ft
Mark 0
(23 microns) 70-80% 60-70%
400
Mark 1
(35 microns) 50-600/., 40-50%
600
2
The advantages of microstraining
are the low initial capital cost and
ease of operation. The disadvantages
are the incomplete solids removal
and the inability to handle solids
fluctuations.
3
The capital and operating costs
associated with microstraining have
been prepared by Smith and are
shown in Fig. 8. For a 10 mgd
plant, total cost for microstraining
is calculated to cost about 1. 5~/1000
gallon.
FIGURE 7
TYPICAL MICROSTRAINER UNIT
2
7
NOT SHOWN
WASTEWATER HOPPER
ULTRA - VIOLET LAMP
WASH. WATER PUMP
t, Dfllrva UNIT
II. ftOTAT1N8 DRUM
O. WASH. WATEft "STS
.. "'1e1M). ~"'811tIC
D. tNPLUI!HT CHAMBER
0. II'I'LUI!HT CHAM8ER
T. I!I'JILUI!HT WAn..
MICROSTRAINING OF SECONDARY EFFLUENT
CAPrTAL COST OPERATING. MAINTENANCE COST. DEBT SERViCe:
va.
DESIGN CAPACITY
m
g 10.0
~
.
..
~
0:'
r'
'~
o
u
~ 1.0
~
~
~
T
A
- -0
./
tr1e~R~ ~3
to
m
~
~
o
o
L
o
m
z
2
0.10 j
i
0.10
~
o
u
~
.
t:
~
u
to
0.D1
100
10.0
DESIGN CAPACITY MILLIONS OPGALLON8 PER DAV
C .. CAPITAL COST, MILLIONS OP DOLLARS
A co DEBT SERVICe.. CENTS PER 1000 CSALLONS(4 1/2,.. 2e VR)
o . M .. OPERATING AND MAINTENANce COST,CENTS PER 1000 C9ALLON8
T II TOTAL TReATMENT COST, CENTS peR 1000 GALLONS
FIGURE 8
v
IN-DEPTH FILTRATION
A
. In-depth filtration is the passage of a
a fluid through a bed of granular media
designed to permit the captured
particles to be retained within the
filter. The degree of penetration and
solids removal efficiency can be
altered by changing the size and
character of the particulates to be
removed as well as the size and com-
position of the filter media itself.
1 A basic prerequisite for the operation
of rapid sand filters is that good
coagulation and flocculation must be
obtained. Barring adequate pre-
treatment of the wastewater,
filtration efficiency is decreased as
evidenced by "breakthrough" of
floc. With good pretreatment by
coagulation, higher filtration rates
are attainable while still maintaining
clarity of the effluent.
.5
-------
Solids Removal Processes
2
It is beyond the scope of this talk
to discuss the subject of coagulation
and flocculation. A variety of
coagulants and flocculants are
available including various salts of
aluminum and iron, lime and organic
polymeric flocculants. The former
inorganic salts, in the proper dosages
can also provide for precipitation
of phosphate in addition to clarification.
3
In the past 70 or 80 years, there has
been a gradual improvement in the
basic process of media filtration.
Some of the earlier work centered
on increasing the filtration rates of
slow sand filters to the modern
rapid sand filter rates. Various
cleaning methods were also developed
including backwashing and surface
scouring. More recently, the
engineering advances have been
concerned with modifications of
filter media that woUld allow greater
production of high quality water from
a given filter area.
4
The evolution of filter design is
illustrated in Fig. 9. The cross-
section shown at the top represents
the rapid- sand filter which is in
common use in many filter plants
today.
a
Typical effective size of the media
used 18 0.5 although effective sizes
from 0.35 to 1. 0 mm have been
used.
b
During filter backwashing, the sand
grades hydraulically with the finest
particles rising to the top of the
bed. As a result, most of the
material removed by the filter is
removed at very near the surface
of the bed.
c
Only a small part of the total voids
in the bed are used to store partic-
ulates and headloss increases very
rapidly. When secondary effluent
is being processed, the high solids
concentration will blind the surface
2-6
in a very short time. As much as
75-95% of the headloss, under
these. conditions, will occur at
the uppe r 1- inch laye r of the filter.
Filter runs will be so short as to
be prohibitive. Further, floc
breaking through the topmost layers,
have increased opportunity to pass
through the entire filter since voids
become increasingly large r with
increase in depth.
FIGURE 9.
FIGURE 1
CROSS - seCTION THROUGH
SING LE . MeOlA BED
SUCH AS CONveNTIONAL
RAPID SAND FILTER
FIGURE '2
CROBS - SECTION THROUGH
DUAL - MEDIA BED
COARSE COAL ABove
FINE SAND
FIGURE 3
CRoes - SECTION THROUGH
. IDEAL PIL TER
UNIFORMLY GRADED FROM
COARse TO FINE
FROM TOP TO BOTTTOM
N vv
y
y
FIGURE 9
GRAIN SIZE
GRAIN SIZE
GRAIN SIZE
One approach to increasing the effective
filter depth is the use of a dual media
bed using a discrete layer of coarse
coal above a layer of find sand. The
filter provides basically a two-layer
effect to achieve increased penetration
of particles.
B
-------
Solids Removal Processes
1
The amount of sand is reduced to
afford lower headlosses at the higher
throughput rates used. Normally,
such a bed is designed so that 24
inches of anthracite coal, with a
nominal size of about 1 mm, overlays
a 6 inch sand layer with a size of
about 0.45 mm.
2
Hydraulic stratification still occurs
following backwashing but the difference
of specific gravity is such that the
larger coal remains on top of the sand.
The bulk of filtration is accomplished
in the upper layers of the coal and at
the top inch or two of the sand bed.
3
With low applied turbidities and
constant rate operation, the coal- sand
media bed has demonstrated an ability
to?perate in the range of 4-5 gal/minI
sq ft of filter surface area.
4 A defect in this design is that if a
flow change occurs, the particles
held in the relatively large void
volume of a coal bed, can become
dislodged and will be captured by the
fine layer of sand and the filtration
run would have to be terminated
either because of high head loss at the
coal-sand interface or because of
breakthrough of particles through the
relatively shallow sand bed. This
design, then, presents a serious
inconsistency.
5
Ideally, the effluent should pass
through as fine a filter material as
is feasible. This ideal design is
illustrated in the bottom cross-
section of a filter uniformly graded
from coarse to fine from top to bottom.
6
One solution is the use of three
materials of differing specific
gravity of such size gradation that
some intermixing of the materials
occur at the interfaces of the bed
layers. The third media material
is garnet which has a specific
gravity of about 4.2. An ideal
filter bed, then, would consist of
about 60% of anthracite with a size
of 1. 0 to 1.5 mm at the top, 30%
sand with a size of 0.4 to Q. 5 mm
at the middle and 10% of garnet with
a size of about O. 15 mm. The
materials are so sized that inter-
mixing occurs at the interfaces.
In this ideal filter, the effluent is
passed through increasingly finer
media. The uniform decrease in
media particle size with filter depth
allows the entire filter depth to be
used for floc removal and storage.
7
Cost of the filtration step is shown
in Fig. 10. For a 10 mgd plant
this cost amounts to about 3. 5r;/ 1000
gallon. when operating the filter at
4 gpm/ sq ft. Further economy. of
course, can be obtained at higher
rates. Rates as high as 6-8 gpm/ sq
ft have been shown to be feasible
when proper pretreatment of
coagulation-flocculation followed
by sedimentation is practiced.
FIGURE 10.
FIGURE 10
FILTRATION THROUGH SAND
OR GRADED MEDIA - 4GPM/Sa FT
CAPITAL COST. OPERATING.. MAINTENANCE COST. DEBT seRVICe.
va.
DESIGN CAPACITY
~
z
~
.
G
10
COST A ",USTEO TO JUNE. 196'7
,
V
... 7'
...
-- .,.
......
./ .. ... .._.~&!,f'"
./ ...
..
.4
...
FIGURe: ~
.01
1.0
~
~
~
~
o
o
~
~
z
o
j
2
~
..
~
u
~
o
u
..
z
~
~
~
..
1.0
"
0.1 ~
u
-'
.
t:
~
.
u
0.1
1.0
10.0
100
DESlaN CAPACITY, MILLIONS OF GALLONS PER DAY
C D CAPITAL COST. MILLIONS OP DOLLARS
A = DEBT SERViCE. CENTS PER 1000 GALLONS(4 112~ . '25 YR.)
t.I .. Jot D OPERATING AND MAINTENANCe:. COST, CENTS PER ,OOC GALLONS
T = TOTAL TREATMENT COST, CENTS PER 1000 GALLONS
2-7
-------
Solids Removal Processes
c
Moving Bed Filter Technique
1 A new filtering technique has been
evaluated by Johns-Manville under
contract with FWQA. The technique
is designed to overcome the problem
of surface clogging and to achieve
what is obtainable by multi-media
filtration. A schematic diagram of
the moving bed filter is shown in
Fig. 11.
BASIC CONCEPT OF MOVING BED FILTER
INFLUENT
SAND RECYCLE
FIGURE 11
2
The unit is basically a sand 'filter.
Particulate matter is removed as
the water passes through the sand
(0.6 to 0.8 mm). As the f~lter surface
becomes clogged, the filter media is
moved forward by means of a mechan-
ical diaphragm. The clogged filter
surface is removed either hydraul-
ically or, as shown, mechanically
thereby exposing a clean filter surface.
The sand and accumulated sludge is
collected and washed. The sand is
returned via a hopper to the base of
the bed. The unit is thus a form of
countercurrent filtration device
feeding sand countercurrent to the
wat er being filtered. The moving
bed filter has a renewable filter'
surface analogous tb the microstrainer
8
and the advantage of depth filtration
comparable to the coarse media
filter. The unit does not have to be
taken off-stream for backwashing.
In theory, 1% of the filter is being
backwashed 100% of the time compared
to the conventional practice of back-
washing 100% of the filter 1% of the
time.
3
Several pilot MBF units have been
built to date and used to treat settled
and non-settled trickling filter
effluents and primary effluent. The
system lends itself well to the use of
chemical aids ahead of filtration
because of designed flexibility to
handle high solids loadings.
Fig. 12 shows some data with the MBF
treating non-settled trickling filter effluent
with various dosages of alum and coagulant
aid. Product quality was maintained at
about 10 BOD while being fed an influent with
a widely varying BOD content of 40 to as
high as 180 BOD.
It is premature to talk about cost of this.
method of filtration. Design and performance
,information developed from these pilot units
~ill be used to obtain these costs.
FIGURE 12
200
160
" 120
co
E
0
0
II) 60
40
20
00 0
o 0
~ "
N
BOD REMOV ALS BY MBF SYSTEM
--
--
o
o
'"
o
o
o
"
N
o
o
~
o
o
~
o
o
'"
o
SAMPLING TIME
-------
Solids Removal Processes
VI
ULTRAFILTRATION
A
One of the newest unit processes to
separate solids from liquids is the
operation known as ultrafiltration. This
method has been under development over
the past 10 years, with commercial
applications over the last 2 or 3 years.
1 Ultrafiltration is closely related to
reverse osmosis with the distinction
generally made on the basis of size
of particle separated. Reverse osmosis
removes molecular sizes including
inor ganic salts. Ultrafiltration
generally will not separate molecules
smaller than 500-1000 MW, thus
inorganic salts are not separated from
solution. Ultrafiltration uses pressures
of about 50 psi in contrast to reverse
osmosis where pressures in excess
of 500 psi are generally used.
2 Membrane ultrafiltration is a pressure
activated process using semi-permeable
membranes which act as molecular
screens to separate molecular and
colloidal materials dissolved or sus-
pended in a liquid phase.
3
Thus far, the principal commercial
applications of the process have
included:
a industrial operations where valuable
products can be recovered by
separation from a bulk solution,
b analytical application which provides
a new method to separate molecules
according to size and molecular
weight, and
c in a package waste treatment plant
which separates mixed-liquor
solids from a biological reactor.
Solids are returned to the aerator
and clarified product water is
discharged.
d A plant employing ultrafiltration
has been installed on Pikes Peak
to provide waste treatment and
water reuse. The essentials of the
process are shown in Fig. 13.
The 15,000 gpd plant treats
wastewater by high-solids
activated sludge. The solids
are separated in an ultrafiltration
unit. Solids are returned to the
aerator. Product water is
excellent. Some typical re-
movals are shown in the follow-
ing:
ENZYME REACTOR
CONCENTRATE
FEED
RECYCLE LINE
REACTOR
PUMP
FIGURE 13
REACT ION SYSTEM
REMOVALS BY ACTIVATED SLUDGE
AND ULTRAFILTRA TION
Influent Effluent
BOD 382 < 1
COD 678 20
Turbidity < 0.1
SS 323 0
PO -p* 12.2 7.7
4 *No added coagulants for P
0/0 Removal
> 99
97
100
37
.9
-------
Solids Removal Processes
REFERENCES
Evans, G. R., "Microstraining Tests on
Trickling Filter Effluents in the Clear Creek
Watershed Area, Texas," Public Works,
October 1965).
Bodien, D. G., and Stenburg, R. L.,
"Microscreening of Effluent from a Municipal
Activated Sludge Treatment Plant, " Water
and Wastes Eng., (September 1966).~
Truesdale, G. A., Birkbeck, A. E., and
Shaw, D., "A Critical Examination of Some
Methods of Further Treatment of Effluents
from Percolating Filters, " Conference
Paper No.4, Water Pollution Research
Laboratory, Stevenage, England (July 1963).
Diaper, E. W., "Microstraining and Ozonation
of Sewage Effluents, " Presented at the 41st
Annual Conference of the WPCF, Chicago,
Illinois (September 1968).
Truesdale, G. A., and Birkbeck, A. E.,
"Tertiary Treatment of Activated Sludge
Effluent, " Reprint No. 520, Water Pollution
Research Laboratory, Stevenage, Herts,
England (1967).
Smith, R., "Cost of Conventional and Advancec
Treatment of Wastewaters, " Jour. Water
Pollution Control Federation, 40, 1546-; 574
(September 1968). -
Rich, Linvil G., Unit Operations of Sanitary
Engineering, John Wiley and Sons, Inc.,
New York, 1961.
Camp, T. R., and Stein, P. C., "Velocity
Gradients and Internal Works in Fluid
Motion, " Jour. Boston Soc. Civ. Engrs.,
30, 219 (1943).
Robeck, G. G., "High Rate Filtration Study
at Gaffney, South Carolina, Water Plant, "
USPHS, R. A. Taft Sanitary Engineering
Center, Cincinnati, Ohio (1963).
Craft, T. F., "Review of Rapid Sand
Filtration Theory, " JAWWA, p. 428-439,
April 1966.
O'Melia, Charles R., and Crapps, David
K., "Some Chemical Aspects of Rapid
Sand Filtration, JAWWA, 56, 1326 (1964).
2-10
Hudson, H. E., Jr., "Physical Aspects
of Flocculation, "JAWWA, July 1965.
Gurnham, C. F., Industrial Wastewater
Control, Academic Press, Inc.,
New York, 1965.
Katz, W. J., and Wullschleger, R.,
"Studies of Some Variables Which Affect
Chemical Flocculation When Used with
Dissolved-Air Flotation, "Proc. 12th
Purdue Industrial Waste Conference (1957).
van Vuuren,L. R. J., et al., "Dispersed
Air Flocculationl Flotation for Stripping of
Organic Pollutants from Effluents, " Water
Research, Pergamon Press, 2, 177-~
(1968). -
Vrablik, E. R., "Fundamental Principles
of Dissolved-Air Flotation of Industrial
Wastes, "Proc. 14th Purdue Industrial
Waste Conference (1959).
Michaels, A. S., "New Separation Tech-
nique for the CPI, " Chemical Engineering
Pro~ress, 64, December 1968.
Weissman, B. J., Smith, C. V. Jr., and
Okey, R. W., "Performance of Membrane
Systems in Treating Water and Sewage, "
Water - 1968, 64, Chemical Engineering
Progress Symposium Series.
Mohanka, S. S., "Multilayer Filtration, "
JAWWA, 61, 504-511 (Oct. 1969).
Ives, K. J., and Gregory, J., "Basic
Concepts of Filtration, "Proc. Society for
Water Treatment & Examination 16
Part 3 (1967). ' ,
Boby, W. M. T., and Alpe, G., "Practical
Experiences Using Upward Flow Filtration, "
Proc. Society for Water Treatment &
Examination, 16, Part 3 (1967).
Rimer, A. E., "Filtration Through a
Trimedia Filter, " Jour. San Eng. Div.,
ASCE, SA 3, June 1968.
This outline was prepared by Dr. S. A.
Hannah, Supervisory Research Chemist,
Municipal Environmental Research
Laboratory, USEPA, Cinci.ru1ati, Ohio 45268
Descriptors: Wastewater Treatment,
Waste Treatment, Water Pollution Treatment
Suspended Solids
-------
METHODS FOR ACHIEVING HIGH BOD AND SOLIDS REMOVALS FROM
CONVENTIONAL PLANTS
I
INTRODUCTION
It is an established fact that many of our
biological treatment plants produce poor
quality effluent. Often, the treatment pro-
cess itself is mistakenly cited as being
inadequate, whereas the true fault lies in not
applying known principles and technology
efficiently. In other cases, plants perform
poorly because they are overstressed out of
all proportion to design conditions.
Can we upgrade secondary treatment in
general and the activated sludge process in
particular? Certainly at many plants we
can. In many cases, pollution can be greatly
reduced almost overnight by maximum use of
existing treatment technology. A properly
designed and well-operated activated sludge
plant n'eeds no new technology to achieve
90-950/, carbon, suspended solids, and coli-
form removals; what is needed is an aware-
ness of the fundamentals involved and a
willingness to pay for and apply all we know.
II
PROPER PLANT OPERA TION
A Since the inception of secondary treatment,
the major emphasis has been on reducing
costs, not reducing pollution. For a
weakly concentrated process stream such
as wastewater, it is difficult to conceive
of any scheme that would be more efficient
and economical than aerobic microorganisms.
However, there is simply no substitution
for time and aeration, and we must be
willing to pay for the required treatment
volume and blower capacity to allow the
biological system to perform efficiently.
When we skimp on these two necessities,
we have cheaper treatment, but also
poorer treatment.
B All too often, poor treatment is caused
due to the failure of city administrators
to realiz e the importance of providing
competent, certified plant operators. A
plant will not run itself. The difference
between a poor and a good operator can
often mean 25-300/, difference in overall
effectiveness. It really defies explanation
to invest huge sums in designing ahd con-
structing a plant and then turn it over to a
chief operator who has no conception of what
is involved in waste treatment and what steps
can be taken to meet varying problems.
Clearly, local, state, and Federal officials
SE. TT. 6. 7. 77
have a responsibility to train, attract, and
hold good technicians and operators. This
implies, again, that we must be willing to
pay a fair salary for value received.
C Once a competent operating staff is
assembled, adequate maintenance and
replacement budgets must be provided to
keep the hardware in good condition.
Continuing good operation is not possible
when key equipment is constantly down for
repairs or becomes antiquated.
D Proper plant operation should include
around-the-clock operation and frequent
inspection and adjustment, where necessary,
of all controllable process flow rates.
Many new instruments are on the market
which are applicable to sewage treatment.
These instruments can assist the operator
not only in monitoring his plant, but also,
in some cases, automatically controlling
certain key parameters such as sludge
recycle rates. Magnetic flow meters,
variable speed pumps, rate controllers,
sludge blanket indicators, and on-stream
pH, dissolved oxygen, and temperature
detectors and recorders are types of equip-
ment which are being requested in new
plant designs.
E And last, but certainly not least, proper
operation must include adequate laboratory
facilities and staff to make the necessary'
measurements and analyses to document
plant performance. Continuous performance
records not only assist in the day-to-day
operation and control of the plant, but they
are valuable reference information when
plant . expansion is planned. The day is
past, too, when BOD and suspended solids
measurements will suffice. The BOD test
is strictly a procedure to measure the rate
at which oxygen is depleted as organic
matter is decomposed. It was never intended
to indicate the total quantity of organic .
matter present in a sample nor to measure
treatment efficiency. Yet, we use the
procedure in both these conte~ts today.
1 Figure 1 shows the daily oxygen depletion
of an influent and an effluent. One curve
is concave and one is conves which means
that the reaction rates of the two bear no
relation to one another. Comparing the
BOD's of these two samples to obtain a
measure of plant efficiency is like sub-
tracting apples from oranges.
3-1
-------
Methods for Achieving High BOD and Solids Removals
Figure 1
6
5
.J 4
.....
C'
:E
,
I
Z 3
a
I-
w
-.J
a.
ltJ
0 2
-
d
a
o
I.
2
J.
o
3
4
T - DAYS
5
6
7
8
2 A better estimate of total organic matter
and the oxygen demand present is given
by COD or TOe, and unoxidized nitrogen.
Figure 2 indicates how these tests can
be used to predict the ultimate oxygen
demand of waste.
3 In addition to carbon, suspended solids,
and nitrogen, phosphorus measurements
should become routine practice for
laboratory staffs. The subject of
phosphorus will be dealt with in greater
detail later on today. I should note
3-2
-------
Methods for Achieving High BOD and Solids Removals
Figure 2
Ultimate Potential Oxygen Demand of an Effluent
UOD = 2. 67C + 4.57N
UOD
C
Ultimate Oxygen Demand
N
=
Organic Carbon Content
Total Oxidizable Nitrogen
(incl. Organic and Ammoniacal)
1.
Recommended Tests
TOC to determine C
2.
TKN to determine N
COD may be roughly substituted for
2. 67C if TOC Analyzer not available.
before leaving the subject of analytical
procedures that the preceding comments
on the BOD test are strictly my opinion.
III
AIDING OVERLOADED PLANTS
Let's assume we have a competent operator,
a good laboratory staff, and modern equipment
and controls, yet overall treatment efficiency
is still unsatisfactory. If there are no un-
usual toxic or grossly atypical wastes present,
the impairment can usually be traced to one
or more of three factors: hydraulic overload,
organic overload, and poor final clarifier
liquid-solids separation.
A Hydraulic Overload
Figure 3 shows several techniques which
may be tried to combat hydraulic overload
in certain situations. They are simple
but they work.
1 The most obvious place to start,
particularly in small towns, is to
locate and reduce needless sources.
Industrial cooling water, for example,
should never enter the sanitary sewer
system. Many industrial rinsing
operations can reduce the volume of
rinse water if asked. Ground water
infiltration is another possible area
which may be attacked.
2
Very large interceptors quite possibly
can serve as temporary holding tanks
during the day and the backup processed
through the plant during low nighttime
flows.
3 Perhaps the most direct and efficient
way is to install a surge or equalization
tank ahead of the plant. The flow to the
plant can be smoothed considerably. If
an outpumping scheme is used as shown
in Figure 4, the flow can be kept virtually
constant over a given 24- hour period. It
is well to provide some minimal aera-
tion to keep the waste from going septic.
In certain instances by utilizing a floating
takeoff and outpumping, a primary settler
can also serve as a surge tank, though
generally they are smaller than needed.
When all else fails, additional capacity
must be provided.
B Organic Overload
Possible remedies for this problem are
given in Figure 5.
1 A surge tank is useful in diluting strong
wastes and evening out the organic load-
ing to the secondary system.
2 Industries in smaller towns are usually
willing to program their more noxious
3-3
-------
Methods for Achieving High BOD and Solids Removals
\50,000
w
\5
~ 1001°00
o
~
u..
o
I/')
2
3
j 50,000
~
3-4
Figure 3
Control of Hydraulic Overload
1.
2.
Find and reduce needless sources
Consider using large interceptors as
holding tanks
3.
Construct surge tank
Fancy
Plain
4.
Enlarge plant
Figure 4
Surge Tank Sizing
MA)(. 5ToRA6E
"O~uMe .
57 qA~.
\Ai-JIL C.O\o.l'TEt-JT5
p..., \ '2..' '" 00 \J
D e-re.t-JTIOt.J
T\ME A-r
\'2.. NOO~
/
/ ~ OlJ,PUMP,t.Jc,
MI\~'5 D'A~RA.M
MItJ. STO!:2,!>..qE: VOl-uME:
\0 000 ~Al-.
I
/
o
,2. "':.\~
" AM
, PM
\'L MID
\'2.. tJool-.{
T\M~
OF D/:>..'<
-------
Methods for Achieving High BOD and Solids Removals
Figure 5
Control of Organic Overload
Surge tank to smooth load
Have industry program load
Program digester supernatant
recycle or eliminate
Install roughing filter-in line or
loop operation
Consider two-stage treatment
1.
2.
3.
4.
5.
waste discharges to coincide with lower
loading periods of the day.
3 Probably the most frequent and upsetting
shock load plants have to deal wi th is
digester supernatant. COD and ammonia
concentrations of 10,000 - 20,000 and
1,000 mg/l, respectively, are common
and together constitute a tremendously
high oxygen demand. If the supernatant
recycle cannot be broken by separate
treatment, it should be returned slowly
to the main plant flow stream, not in a
slug in a one-hour period in the middle
of the day as is often done.
4 The recently developed plastic media
have a good potential for taking out the
bulk of incoming organics when used as
a roughing filter ahead of the normal
secondary process. If used strictly
to control organic overloads, a roughing
filter does not require a separate settler.
5 Two-stage treatment is worth consider-
ing when a high rate system is to be
enlarged. The two sludge systems
will enhance the possibility of achieving
nitrification, and it is unlikely that
both sludges will be adversely affected
by toxic spills.
C Liquid-Solids Separation
1 Physical features that will improve
liquid-solids separation are suction-
type sludge removal mechanisms,
variable speed pumps tied to a sludge
blanket indicator or the incoming raw
flow meter signal, scum removal
equipment, and proper baffling to
prevent short circuiting and strong
velocity currents.
2 Polymers can temporarily keep a
bulking sludge from "washing out" gross
mixed liquor solids. A drum of polymer
and a small dosing pump, if kept in
reserve, can get a plant through an
upsetting shock load of several days
duration. Polymers are stable for long
periods of time and are available in
powder and liquid form.
IV A RECENT DEVELOPMENT FOR UPGRAD-
ING OVERLOADED PLANTS
Union Carbide Corporation has recently
completed a research contract with FWPCA
which successfully substituted oxygen for air
in the activated sludge process. A closed
system is employed to achieve high oxygen
utilization. Dissolved oxygen is readily
maintained in the system even with mixed
liquor suspended solids concentrations up to
8,000 mg/l. The necessary aeration volume
to achieve efficient organic synthesis and
oxidation is reduced to 2-3 hours. Figure 6
shows a schematic diagram of the system
and Figures 7 and 8 give treatment data at
two nominal aeration periods of 3 and 1-1/2
hours, respectively, which were studied
during the contract. This process has very
good potential for increasing the capacity and
improving efficiency of many overloaded plants
in this country without enlarging aeration
tank volumes.
V
SUMMARY
A summary of anticipated effluent quality and
cost data for conventional processes is given
in Figure 9. Note that as effluent quality
improves, the cost associated with the process
rises. Process 6 as shown i. e., primary
3-5
-------
Figure 6
~
(1)
.....
::r
o
0-
m
....,
o
Ii
~
(')
::r
,...
(1)
~.
::3
'"
I
en
SCHEMATIC DIAGRAM OF UNION CARBIDE
OXYGENATION SYSTEM
::r:
,...
::r
tJ:I
o
t:J
II'
::3
0-
en
o
I::i
0-
m
~
(1)
!3
o
<:
II'
'-'
m
AGITATOR
GAS RECIRCULATION
COMPRESSORS
. .
..
. .
WASTE
LIQUOR
FEED' .
t:: .\ .f.... \. :.
. . ..
. -.: . : .' ..~ ... . :'
.. . . .. .' STAGE
.. . .. . BAFFLE
.\::::"i (0..: f:~
. . .
. . . ..8 .....
..~.. ... ~.'.
MIXED
LIQUOR
EFFLUENT
c1
c1
~
---"
.~
----./
-------
Methods for Achievin~ Hi~h BOD and Solids Removals
Figure 7
PHASE NO.3 OPERATION
SUMMARY OF DAILY AVERAGE PERFORMANCE (PRELIMINARY)
Oxy~en Train
Parameter
Air Train
Raw Sewage Feed Rate (Mgd)
Return Sludge Recycle Ratio
Haw Sewage Temperature (oF)
MLSS Concentration (ppm)
ML VSS Concentration (ppm)
Recycle Sludge TSS (ppm)
1. 30
0.26
69
3400
2410
15400
1. 30
0.45
69
6800
4375
22050
Mixed Liquor D. O. Concentration (ppm) 0.9
Clarifier Effluent D. O. Concentration (ppm) 0.0
8.6
6.9
TSS of Clarifier Effluent (ppm)
o/c TSS Removal
31
92.5
16
96
BOD of Clarifier Effluent (ppm)
o/c BOD Removal
42
85
19
93
COD of Clarifier Effluent (ppm)
o/c COD Removal
147
72
69
87
Ft3 Air/ Gal. Raw Sewage
Lbs. Oxygen Utilized/ Lb. BOD Consumed
o/c of Feed Oxygen Utilized
Lbs. Dry Solids Wasted
Total Oxygenation System Power (HP)
Nominal Aeration Time (HI's. )
3.91
3
1.0
93.5
650
18.3
3
2900
3-7
-------
Methods for Achieving High BOD and Solids Removals
Figure 8
PHASE NO.2 OPERATION
SUMMARY OF DAILY AVERAGE PERFORMANCE
Parameter
Raw Sewage Feed Rate (mgd)
Return Sludge Recycle Ratio
Aeration Detention Time (Hrs.)
Raw Sewage Temperature (oF)
MLSS Concentration (ppm)
MLVSS Concentration (ppm)
Recycle Sludge TSS (ppm)
Sludge Volume Index
1. 5 Nominal
Oxygen Train
2.62
0.33
1. 17 Actual
66
7142
4575
29600
35.0
Mixed Liquor D. O. Concentration (ppm)
Clarifier Effluent D. O. Concentration (ppm)
9.8
1.0
TSS of Clarifier Effluent (ppm)
o/c TSS Removal
18
90.6
BOD of Clarifier Effluent (ppm)
o/c BOD Removal
22.9
90.1
COD of Clarifier Effluent (ppm)
o/c COD Removal
99.0
70.0
o/c Phosphate Removal
Lbs. Oxygen Utilized/ Lb. BOD Consumed
o/c of Feed Oxygen Utilized
Lbs. Dry Solids Wasted
Total Oxygenation System Power (HP)
22.2
1. 03
92.6
1804
16.8
3-8
-------
Methods for Achievinf{ Hif{h BOD and Solids Removals
Figure 9
EFFLUENT CHARACTERISTICS AND COST DATA FOR CONVENTIONAL PROCESSES
Process
Total operating
cost: $/ mil. gal.
10 mgd plant
Expected Ranges of Effluent Characteristics in mg/l
1. PS
2. PS + TF (high rate)
3. PS + TF (low rate)
4. PS + AS (high rate)
5. PS + AS (conventional)
6. PS + AS (operated to nitrify)
60-65
85-90
90-95
95-105
110-115
120-130
BOD5 COD TOC SS TKN TP
115-150 200-275 65-90 75-100 25-35 8-12
40-60 100-150 30-50 30-50 20-30 7-11
20-35 55-90 20-30 20-35 15-20 8-12
30-40 75-100 25-35 30-35 20-30 7-11
15-25 40-70 15-25 20-25 15-20 7-11
10-20 30-60 10-20 15-20 2-4 8-12
Assumed Raw Wastewater Quality
BOD5: 175-225 mg/l
COD : 325-450 "
TOC : 100-130 "
SS 200-250 II
TKN : 30-40 II
TP 10-15 II
Legend
PS - Primary sedimentation
TF - Trickling filtration
AS - Adivated sludge
SS - Suspended solids
TKN - Total Kjeldahl nitrogen as N
TP - Total phosphorus as P
settling followed by activated sludge operated
to nitrify, represents about the maximum
efficiency that can be extracted from biologi-
cal treatment alone.
This outline was prepared by R. C. Brenner,
Research Sanitary Engineer, Municipal
Environmental Research Laboratory,
USEPA. Cincinnati, Ohio 45268
Descriptors: Wastewater Treatment, Waste
Treatment, Water Pollution Treatment
3-9
-------
BASIC PRINCIPLES OF CARBON ADSORPTION
I
INTRODUCTION
A dsorption can be described as the condensation
of gases, liquids, or dissolved substances on
the surfaces of solids. Almost any material
has some adsorptive capacity, but the effects
are not evident unless the adsorbing material
(adsorber) is porous and has a very large area
for a given mass.
II
FACTORS TO BE CONSIDERED
A Contact Time
Adequate time must be allowed to obtain the
desired removal.
1 Adsorption usually consists of a rapid
initial rate of uptake followed by a marked
decrease in rate. The latter rate can be
interpreted as intraparticle diffusion and
the uptake varies almost proportionately
with the half-power of time. Linear 1/2
variation of the quantity adsorbed with t
can be expected.
B pH
The effect of pH is significant in adsorption.
An inverse relationship between capacity and
pH has been observed. Acidic materials
adsorb best at low pH. A free acid adsorbs
better than a salt of that same acid.
C Temperature
Temperature effects on adsorption equilibria
generally are not felt to be significant,
particularly over the range of temperatures
encountered in water and wastewater.
Increasing temperature should result in
decreasing capacity because physical
adsorption is ordinarily an exothermic
process, however, if viscosity is a factor,
the benefits of a decrease in viscosity may
override other temperature effects.
SEe TT. 2. 7. 77
D Concentration of Solute
The mechanism, as previously stated,
probably includes rapid formation of an
equilibrium concentration at the liquid-
particle interface followed by slow diffusion
into the pores of the particle. A simple
relationship between adsorption rate and
the solution concentration cannot be
expected. A further consideration is the
proportion of contaminants that are
adsorbable in the solution. Higher con-
centrations should give more rapid initial
adsorption, but this has no relation to the
amount of contaminant adsorbed at
equilibrium. Tests using a single solute
indicate the adsorbed film obtained at very
high concentrations of solute is different
from that obtained in the lower concentra-
tions common to wastewaters. Low
concentration equilibrium data correlate
best with a Langmuir equation. Relatively
high concentrations are best described by
a Brunauer-Emmett-Teller equation.
E Molecular Size of Solute Molecules
If the diffusion of solute molecules is a
rate-determining step in adsorption,the
size of the diffusing molecule should be a
factor. The larger the molecule, the
lower the rate of diffusion, and, therefore,
the rate of adsorption. This has proven
to be the case.
F Viscosity
A highly viscous solvent would decrease
the rate of diffusion;
G Carbon Particle Size
A decrease in carbon particle size should
give an increase in rate and quantity of
adsorption. More surface area is exposed
and more pores should be opened for
diffusion.
4-1
-------
Basic Principles of Carbon Adsorption
H Carbon Concentration
Data from experiments using the high carbon
concentration (powdered carbon) necessary
for wastewater treatment are not available.
It is expected that the carbon concentration
would not affect the rate of adsorption,
except for a slight decrease in the rate due
to the decreased concentration of the solute.
An increase in the amount of carbon would,
of course, increase the quantity of con-
taminant adsorbed.
I
Carbon Pore Structure
Large pores (macropores) completely
permea te each carbon particle and act as
access ways for diffusion of materials.
The macropores contribute very little to
the total surface area. The micropores are
largely responsible for the adsorptive
action. These pores are many orders of
magnitude smaller than the macropores
and can be thought of as tributaries to the
macropores. The diameter of the micro-
pores are in the order of 10 to 100 Angstrom
units and, although the micropores con-
stitute only one-third to one - half of the
volume of the carbon particle, they con-
tribute as much as 1,000 square meters
of surface per gram of carbon, representing
virtually the entire surface area. A carbon
with very fine pores will adsorb small
molecules while leaving larger ones in
solution. A carbon with large pores will
preferentially remove large molecules
while the smaller ones are displaced from
the surface.
J Active Surface Area
The total surface area that has specific
attractive power for the wastewater con-
taminants and the accessibility of such
surface is an important factor. Comparing
two carbons, the one with more active,
accessible surface area will remove more
of the contaminants.
4-2
III
ADSORPTION THEORY
Various theories have been proposed for the
phenomenon of adsorption, but due to the
many factors involved the concept is very
complex. No single mathematical equation
other than a purely thermodynamic one can
cover all cases.
Adsorption is termed reversible if the same
equilibrium conditions are reached when the
solute concentration is increased to a definite
level from a lower concentration as when it
is decreased from a higher concentration to
the same level. At equilibrium, the rate of
adsorption is just equal to the rate of
desorption. Reversibility is observed with
many solutes although it is not a universal
phenomenon; many adsorptions are not
reversible or are incompletely reversible.
The equilibrium relation between the con-
centration of a substance in the adsorbed
state and the concentration remaining in
solution forms the basis of most equations
that have been developed to represent
adsorption. Many of these equations are
useful in specific cases, but none are
universally applicable. It is for this reason
that the simplest relation is commonly used.
A Freundlich A dsorption Isotherm
A simple relationship is found by conducting
adsorption experiments through a range of
equilibrium concentrations and determining
the amount of substance adsorbed at each
equilibrium concentration.
The mass of material adsorbed is related
to the concentration of the solution by the
equation
1
n
~ = kC
m
where x = units of material adsorbed
m = weight of adsorbent
C = equilibrium concentration of
materials remaining unadsorbed
in solution
-------
k & n = constants which have different
values for each solute and
adsorbent.
If we take logarithms of both sides of the
equation, we get:
1
log x/m = log k +- log C
n
If we now plot log x/m against log C we
should get a straight line. This expression
is usually known as the Freundlich adsorption
isotherm. It has no thepretical basis, but
is simply empirical. The curve is termed
an adsorption isotherm because the curve
changes with temperature, but is fixed for
one given temperature.
B Langmuir's Formula
Langmuir's formula has a better theoretical
basis than that of Freundlich's.
Suppose that, for 1 sq. cm. of a given
surface, there are N available sites for
adsorption. At equilibrium, let n of these
sites be occupied; there will then be (N -n)
free sites. As stated before, the rate of
adsorption is just equal to the rate of
desorption at equilibrium. The rate of
adsorption is proportional to the con-
centration of the solute and to the number
of free sites, that is,
rate of adsorption = k (N -n)C
a
The symbols x, m, and C have the same
definitions as in the Freundlich isotherm.
The rate of desorption depends only on the
number of occupied places, that is,
rate of desorption = kdn
Thus at equilibrium
k (N -n)C = kdn
a .
and
kaNC k1C
k +k C = 1 + k2C
d a
Basic Principles of Carbon Adsorption
where k1 = k N/kd and k2 = k /kd. But
x/m is direcfiy proportional t8 the number
of occupied places, n per sq. cm. From
which Langnuir' s adsorption isotherm
follows
x/m = k1C
1 + k2 C
An important difference between the two
theories is that Langmuir's formula takes
into account the fact that, for very large
values of C, the surface becomes
saturated, and can take up no more. For
large values of C, k2 C greatly exceeds 1,
so the latter can be neglected and the
equation then becomes
x/m = k1
k2
that is, the mass of substance adsorbed
reaches a constant value which is not
exceeded although C may be increased.
Also, for small values of C, k2C may be
neglected in comparison with unity. The
equation then becomes
x/m = k1C
showing that the amount adsorbed is
proportional to the concentration when
the surface is nearly bare.
The Langmuir theory assumes that the
maximum adsorption corresponds to a
single layer of solute molecules on the
adsorbent surface.
C Physical-Chemical Adsorption
In many early studies of gas adsorption,
it was observed that the more condensable
gases are usually adsorbed in larger
quantities. Because of this, adsorption
was explained as resulting from forces
similar to those which produce liquefaction.
According to this view, adsorption will
occur when the physical attraction of a
solid surface is greater than the attraction
of the solvent and other solute molecules
for the affected molecules. The cohesive
4-3
-------
Basic Principles of Carbon Adsorption
physical forces that cause the molecules to
be attracted to each other are called van
del' Waal's forces.
There are adsorptions, however, in which
the behavior indicates that a chemical
change is involved. A dsorption of this type
is termed chemical adsorption or
chemisorption.
Physical adsorption usually involves a
smaller energy change than does
chemisorption. Chemisorption is specific
and dep(mds on the chemical nature of both
the adsorbent and adsorbate. The selectivity
in physical adsorption is usually due to
purely physical properties. Physical
adsorption decreases at higher temperatures,
while higher temperatures are often needed
for chemisorption.
IV
SUMMARY
In most adsorption systems, a high initial rate
of adsorption is followed by a rapid decrease
in the rate. Days or weeks may be required
to obtain equilibrium. The rate after the
initial adsorption appears to be controlled by
diffusion processes. The diffusion process
has been shown to take place in accordance
with the half power of time. If the rate of
adsorption is wholly dependent upon the surface
area, then it should be inversely proportional
to the first power of the diameter, but, for
intraparticle diffusion, the variation seems to
be with the reciprocal of the square of the
dia.meter of the particle. The intraparticle
diffusion, following initial adsorption, seems
to be the controlling factor. Many factors
affect the adsorption process, temperature,
concentration, pH, etc., all being important.
Using simple formulae to explain the
adsorption of the mixtures of materials found
in natural and wastewaters is impossible.
Some materials hinder the adsorption of other
constituents or may aid in the process. In
general, the net effect of mixed materials.
is not detrimental. Research has shown the
following generalities can be made:
1 Weak electrolytes are adsorbed better
than strong electrolytes.
4-4
2 The more ionized a material is, the
more difficult it is to adsorb it out of
solution.
3 Slightly soluble materials are generally
adsorbed better than highly soluble
materials.
4 High-molecular-weight materials may
be adsorbed better than those of low
molecular weight.
There are, of course, exceptions to any
generalities, including those given above.
REFERENCES
1 Atlas Chemical Industries, Inc.,
Measuring A dsorptive Capacity of
Activated Carbons for Liquid
Purification. D-87-4M-3:66. 1963.
2 Bishop, D. F., et al., Studies on
Activated Carbon Treatment. JWPCF,
Vol. 39, No.2, pp. 188-203. Feb. 1967.
3 Fornwalt, H.J. and Hutchins, R.A.,
Purifying Liquids with Activated Carbon,
Chemical Engineering. April 11 and
May 9, 1966.
4 Hassler, J. W., Active Carbon, Chemical
Publishing Co., Inc., Brooklyn, N. Y.
1961.
5 Kruyt, H. R., et al., Physical Chemistry,
Holt, Rinehart and Winston, Inc., N. Y.
1964.
6 Middleton, F. M., Use of Adsorption in
Treatment of Waste Waters, presented
at Florida Pollution Control Association
Meeting, Panama City, Florida.
September 12, 1963.
7 Weber, W.J., Jr. and Morris, J.C.,
Kinetics of A dsorption on Carbon from
Solution, Jour. San. Eng. Div., ASCE.
April 1963.
8 Weber, W. J., Jr. and Morris, J. C.,
Equilibria and Capacities for Adsorption
on Carbon, Jour. San. Eng. Div., ASCE.
June 1964.
-------
Basic Principles of Carbon Adsorption
This outline was prepared by Eugene F.
Harris, Chief, Extraction Technology
Branch, Industrial Environmental
Research Laboratory, USEPA.
Cincinnati, Ohio 45268
9 West Virginia Pulp and Paper, Nuchar for
Purification and Reclamation.
Descriptors: Wastewater Treatment,
Waste Treatment, Water Pollution
Treatment, Carbon Filters, Activated
Carbon, Adsorption, Tertiary Treatment
4-5
-------
BASIC PRINCIPLES OF ION EXCHANGE
I
INTRODUCTION
Ion exchange is a well known method for
softening water and for producing deionized
water. It may also be practical for removing
from wastewater the inorganic materials
added during use. Removal of this mineral
increment is necessary if the quality of the
water is to be preserved. It may also be
applicable to the removal of nutrient materials
that are becoming increasingly serious as a
pollution problem.
II
MECHANISM OF ION EXCHANGE
Ion exchange materials are those containing
ions that can be replaced by other ions from
solution. They may be liquids, but are
usually solids with a particle size range that
makes them convenient for use in a column.
A typical solid ion exchange material can be
looked upon as a microporous matrix con-
taining many ionic sites - either all positive
or all negative - as shown in Figure 1. To
maintain electroneutrality each ionic site
must have associated with it an ion of opposite
sign (counter ion). This counter ion must
be mobile enough so that it may be replaced
by, or exchanged for, another ion of the same
sign upon contact with a solution of this new
ion. When all the original counter ion is
replaced, the ion exchange material is
exhausted. A concentrated solution of the
original counter ion must be used to regen-
erate the ion exchanger. If the counter ion
is a cation, the ion exchange material is a
cation exchanger. If the counter ion is an
anion, the ion exchange material is an anion
exchanger. All the common cations and
anions in wastewater may take part in ion
exchange. Hydrogen and hydroxyl ion are
also often involved.
III
MA TERIA LS HAVING ION EXCHANGE
PROPERTIES
Proper design of ion exchange systems
requires considerable knowledge concerning
SE. TT. cpo 2. 7.77
the physical and chemical nature of exchange
materials. Common types of ion exchange
materials include:
A Minerals
The most common of these are the
zeolites. They are crystalline alumino-
silicates with cation exchange properties.
They have been used often to soften water
by exchanging sodium ion for calcium and
magnesium.
B Coa Is
These contain carboxylic and possibly
other groups that give cation exchange
properties. Chemical treatment,
especially by sulfonation, improves the
chemical and physical properties.
C Synthetic Organic Resins
Since their development, these resins
have become the most important of all ion
exchange materials. They can be divided
generally into the following four groups:
1 Strong acid resins
These are characterized by containing
strongly ionized anionic groups, such
as sulfonate, in the organic matrix.
A typical structure is shown in Figure 2.
Since the counter ions are cations, the
resins are cation exchangers. They
have the ability of "salt splitting" or
exchanging hydrogen ion for the cation
of a salt such as NaCl. A typical
reaction is
- + + - + +
RS03 H + Na -.-. RS03 Na + H
2 Weak acid resins
These are cation resins containing
groups that are only partly ionized,
such as carb.oxyl. They are not effective
for salt splitting,. but in the free acid
form will remove cations from basic
5-1
-------
Basic Principles of Ion Exchange
FIGURE I.
TYPICAL EXCHANGE MATERIAL
2
-------
Basic Principles of Ion Exchange
-CH -CH2 - CH -CH2-
H+503
503 H+
-CH-CH2-
FIGURE 2. STRONG ACID RESIN
solutions. A typical reaction is
+
RCOOH + NaOH ~ RCOO Na + H20
3 Strong base resins
These anion exchange resins are often
quaternary ammonium compounds.
They are strongly ionized and can
accomplish salt splitting. A typical
reaction is
+ - - + -
RN(CH3)3 OH + Cl ~ RN(CH3)3 Cl + OH
4 Weak base resins
These anion resins are not highly ionized
in the hydroxyl or free base form. They
are actually acid absorbers and are not
usually considered to be salt splitting.
A typical reaction is
+ -
RNH2 + HCl ~ RNH3 Cl
D Liquid Ion Exchangers
Ion exchange can take place between two
immiscible liquids. Liquids that are
immiscible with water, may for example,
function as ion exchangers if they contain
ionogenic compounds. Materials are
available that can be dissolved in organic
solvents to give cation or anion exchangers.
IV
SELECTIVITY
A Non Specific Exchangers
Although a general-purpose cation exchange
material will remove any of the common
cations from a wastewater, there is
competition between the various cations.
Similarly, there is competition between
the various anions for the anion exchanger.
If a sample of resin is allowed to remain
in a solution containing two counter ions,
at equilibrium there may be a much
3
-------
Basic Principles of Ion Exchange
larger amount of one counter ion on the
resin than the other. There are a number
of rules of thumb for determining what ion
wil11Je preferred. Two that are likely
to apply to wastewater state that the counter
ion of higher valence and the counter ion
of smaller solvated equivalent volume
will be preferred. For cations the order
of preference for some common examples
++ + + +
is Ca > K > NH4 > Na. For strong
acid resins, hydrogen ion would be to the
right of sodium. Weak acid resins have
strong affinity for hydrogen ion. For
anions the order of preference for two
examples is SO - - > Cl-. For strong base
re::!ins, hydroxj]. would be to the right of
Cl. Weak base resins have strong
affinity for hydroxyl.
B Exchangers with Specific Selectivity
There are applications for ion exchangers
highly specific for certain ions. In
domestic wastewater treatment, there is
presently a demand for methods to remove
nutrients. The most important of these
are nitrate, phosphate, and ammonium ion.
Of the three, only an ammonia selective
material is presently being investigated.
Resins specific for the other nutrient ions
could be of value.
v
EXCHANGE CAPACITY
A Total Capacity
Ion exchange materials have a total exchange
capacity determined by the number of
active sites on the matrix. The following
discussion pertains primarily to the solid
synthetic organic resins. The total
capacity is usually expressed by manu-
facturers as meq/ dry gram or meq/ ml
where volume is bed volume, not resin
particle volume. The latter method of
expressing capacity is convenient for the
engineer who must deal with total bed
volume. Care must be used to define the
counter ion associated with the ion exchange
material ("form" of the resin) when giving
capacity. On a weight basis the different
equivalent weight of different counter ions
5-4
will obviously affect resin weight. On a
volume basis, the effect of the counter
ion is even greater since different counter
ions swell the resin to different degrees.
B Useful Capacity
The total capacity of ion exchangers is
never used completely in practice.
Consider a column of cation exchange
resins having hydrogen as the counter
ion. A sodium chloride solution passed
through the column will exchange sodium
for hydrogen. Because of diffusion
resistance in the liquid and in the resin
particles, exchange is not instantaneous.
If it were, complete exhaustion of each
resin particle would be possible. Instead,
the resin will be exhausted as shown in
Figure 3. As exhaustion proceeds, some
sodium will begin to breakthrough in the
product water. The ion exchange resin
must then be regenerated with an acid
solution. As Figure 3 shows, the resin
capacity will not all be used by the time
regeneration is required because the
resin at the end of the column has not
become completely exhausted. The use-
ful capacity is further decreased after
regeneration. Regeneration would not be
continued until all the counter ions were
replaced by hydrogen. That resin
remaining in the sodium form would be
lost for further sodium removal.
Utilization of resin capacity can usually
be improved by increasing the column
length or by changing conditions so that
diffusion rate is increased.
VI
REGENERA TION
A Cocurrent
In the usual ion exchange installation, the
ion exchange resin is contained in a column
that is fed downward. Upon exhaustion,
the bed is backwashed and regenerated in
a downward direction. This cocurrent
method of regeneration makes poor use
of both resin capacity and regenerant. In
Figure 3, for example, passing acid in the
same direction as the sodium chloride
-------
Basic Principles of Ion Exchange
INITIAL CONDITION
1.0
z!
ii)O
w II.
~
u..Z
OW
zC>
g~
~~
u..z
o
DISTANCE ALONG COLUMN
Figure 3. Effect of passing NaCl Solution through
Cation Resin in the Hydrogen Form.
solution results in the movement of sodium
into the resin that had not been completely
exhausted. A large excess of acid is
required to remove the sodium. Unless
essentially all the sodium is removed it
will contribute to early breakthrough during
the next ion exchange cycle. This is
shown in the data of Table 1 which were
taken by a manufacturer using a common
cation resin.
Table 1
Regenerant Level
(meg/ ml resin)
0.76
4.56
7.6
Useful Capacity
(meg/ml resin)
0.48
1. 47
1. 60
Data from Rohm & Hass, "Technical Notes,
AMBERLITE IR-120" revised March 1960.
Sodium chloride was fed under the same
conditions for each run and the breakthrough
criterion was the same. Acid was then
used for regeneration. Obviously., a
compromise must be made between capacity'
and amount of regenerant. Economics
determine the regenerant level chosen.
B Countercurrent Regeneration
Although operating problems have pre-
vented the extensive use of countercurrent
regeneration, this method may be a
necessity to make ion exchange competitive
with other processes that are being
developed for treatment of wastewater.
In countercurrent regeneration the least
exhausted resin contacts the most con-
centrated regenerant. The driving force
for regeneration of this resin is greater
than with cocurrent regeneration. The
overall driving force for the whole column
is also increased. During ion exchange
the resin at the outlet of the column is
highly regenerated so that early break-
through is prevented.
This outline was prepared by Dr. Carl A.
Brunner. Research Chemical Engineer.
Municipal Environmental Research
Laboratory. USEPA. Cincinnati. OH 45268
Descriptors: Wastewater Treatment, Waste
Treatment. Water Pollution Treatment, Ion
Exchange. Demineralization, Zeolites,
Tertiary Treatment
5-5
-------
IN-DEPTH FILTRATION
I
DEFINITIONS
A In - Depth Filtration
In-depth filtration is the passage of a fluid
through a bed of granular media designed
to permit the particulates to be removed
to penetrate the media surface and yet be
retained within the filter. The degree of
penetration and solids removal efficiency
can be altered by changing the size and
gradation of the filter media or the size
and character of the particulates to be
removed. In the treatment of wastewater
the size and character of the particulates
or floc can be changed by pretreating the
water prior to filtration. Conventional
pretreatment includes chemical coagulation,
flocculation and sedimentation.
B Coagulation
Coagulation is the destabilization of a
conoidal dispersion by reducing the electro-
static and interionic forces tending to
keep the particles apart. In the treatment
of wastewater the destabilization is
accomplished by the addition of an appro-
priate chemical in an environment of high
turbulence achieved by rapid mixing. This
turbulence assures good dispersion of the
chemical and increases the opportunity for
particle to particle contact.
C Flocculation
Flocculation is the growth in particle size
achieved by gentle mixing which provides
the opportunity for additional particle to
particle contact, bridging, adsorption and
physical enmeshment.
D Sedimentation
Sedimentation is the separation of a
suspension into a clarified fluid and a more
concentrated suspension. In the clarifica-
tion of wastewaters we are primarily
concerned with the settling of flocculat.ing
particles as opposed to descrete non-
flocculating particles. Secondary considera-
tion is given to thickening within the
clarifier.
II
THEORETICA LA SPECTS OF CHEMICA L
COAGULATION, FLOCCULATION,
SEDIMENTATION AND FILTRATION*.
A Coagulation
1 Stability of colloidal dispersions
Colloids can be broadly classified as
hydrophillic (water-loving) and hydrophobic
(water-hating). Hydrophillic colloids
attract water to their surface which
engenders stability by preventing
particle to particle contact. Hydrophobic
colloids achieve stability according to
to the classical double layer theory
advanced by Helmholtz. The hydrophobic
colloids acquire an electrical charge by
ionization or adsorption of ions. This
charge is usuaUy negative for particles
in wastewater. This layer of adsorped
ions is tightly bound to the particle and
attracts another layer of oppositely
charged ions which is also tightly bound.
A diffuse layer of positive and negative
ions of varying composition forms around
the particle. The double-layer prevents
particle agglomeration. Since the
*It is beyond the scope of this manual to give a complete treatise on the theoretical aspects of
chemical coagulation, sedimentation and filtration. For those who desire a more extensive
review of the subject matter an extensive bibliography on chemical cdagulation can be found
in reference 1.
SE. TT. pp. 5. 7.77
6"1
-------
In - Depth Filtration
thickness of the double layer and the
surface charge density are dependent
upon the concentration and valence of
the ions in solution the stability of the
suspension can be reduced by adding
appropriate ions. Coagulation is
achieved by the introduction of ions
having a charge opposite to that of the
colloid. Coagulation results not because
of a neutralization of charges, but
rather the thickness of the double layer
is reduced so that the cohesive van der
Waals forces are operative. Brownian
movement causes particle to particle
contact but the probability of contact is
greatly improved with mixing. The
coagulating power of an ion is dependent
upon its valency with a bivalent ion
30-60 times as effective as a monovalent
ion and a trivalent ion 600-1000 times
as effective.
2 Reactions of primary coagulants
a General
Originally the metal ions were thought
to be the effective coagulating agents
but it is now known that the hydrolysis
products of the metal ions are the
coagulating agents. The reactions of
aluminum and iron hydrolysis products
are very complex. The reactions
presented in this manual are over-
simplified. A thorough treatment
of the hydrolytic chemistry of alumi-
num and iron is given in reference 1.
b Reactions of aluminum salts
1) Alum
Alum is the most widely used
coagulant. Aluminum hydroxide
can be formed from the reaction
of commercial alum and the bi-
carbonate alkalinity of the water.
A12(S04)3' 14H20 + 6HC03 --. 2A I(OH)3 + 3S0: +6C02
6~2
2) Sodium aluminate
Sodium aluminate is another
aluminum salt from which alumi-
num hydroxide can be formed.
NaA102+HCO; +H20~ Al (OH)3 +CO; + Na +
There appears to be no difference
in the coagulating efficiency of the
aluminum hydroxide formed from
either salt at equivalent final pH.
A combination of the two salts
can be used to give precise pH
control in unbuffered waters.
Al2 (S04)3' 14H20+ 6 NaAl02 ~ 8Al(OH)3
+ 3Na2S04 + 2H20
c Reactions of iron salts 2
The reactions of iron salts are very
similar to those of aluminum salts.
1) Ferrous sulphate (Copperas)
Usually there is insufficient
alkalinity in natural waters to
rf'act with the ferrous sulfate.
Lime is usually added to produce
a floc and avoid soluble compounds
of iron remaining in the treated
water. The chemical reactions
of ferrous sulphate and lime depend
somewhat upon the order of
addition. The following reaction
occurs when lime is added first.
FeS04 + Ca(OH)2 --+ Fe(OH)2 + CaS04
4Fe (OH)2 + 2H20 + 02 -. 4Fe (OH)3
2) Ferric chloride
Ferric chloride is extremely
corrosive to metal containers and
cannot be dry fed because of its
hygroscopic characteristics.
FeCl3 + 3H20 ---+ Fe'(OH)3 + 3H+ + 3CI-
-------
In - Depth Filtration
d Reactions of lime
Lime is always applied as slaked lime
Ca (OH) . Some of the lime may not
be availa~le because of encrustation
in precipitates of calcium phosphate,
carbonate or sulfate.
- ++
Ca (OH)2+2HC03+ Ca ~ 2CaC03+ 2H20(pH > 9.5)
++ ++
Mg + Ca (OH)2 ---. Mg(OH)2 +Ca (pH> 11)
The gelatinous precipitate Mg (OH)2
gives excellent clarification. In
waters low in magnesium clarification
can be improved by adding magnesium
salts such as MgCl2 or dolomite lime-
stone CaMg( C03) . Dolomite lime
CaO. MgO is less2effective because
the MgO is resistant to slaking and
does not dissolve at high pH to be
precipitated as the gelatinous
Mg(OH)2'
B Flocculation
The rate of flocculation is proportional to
the mean velocity gradient (G), the detention
time, the concentration of particles and
the particle size.
G rcDAV~J1/2
Evv
where: -1
G= mean velocity gradient - sec
CD= drag coefficient
2
A = area of paddles - ft
v = mean velocity of the paddles
relative to the water ftl sec
v = kinematic viscosity of the water
- ft2 I sec
v= volume of flocculator - ft3
The most difficult task in calculating the
velocity gradient is determining CD and
v. For flat blades made of wood CD is
taken to be 1. 8. For metal blades the
value of CD is 1.2. The velocity of the
paddles relative to the water is taken as
Vw = 0.75 vp' Although this value will
depend upon the geometric configuration
of the flocculator and relative area of the
blades to the flocculation basin. Without
the use of stator blades the area of the
paddles should not be greater than 15% to
20% of the cross-sectional area of th~
basin to prevent rolling of the water.
The flocculation process was characterized
by Camp and Stein5 in the following
equation.
G 3
N =6 N1 N2 (d1 + d2)
where:
N = number of collisions per unit
volume per unit time.
-1
G = mean velocity gradient - sec
N 1 = number of particles originally
in the water
d1 = diameter of these particles
N2 = number of floc particles in
the water
d2 = diameter of floc particles
Studies by Robeck 6 indicate N is in the
order of 106 particles per milliliter, and
d1 is in the range of < 10~. Floc particles
on the other hand are often 100-200 IJ..
Therefore, d1 can be omitted from the
equation with little error.
C Sedimentation
The separation of dilute suspensions of
flocculating particles is a function not
only of the settling properties of the
particles but also of the flocculating
characteristics of the suspension. A
fraction of the subsiding particles overtake
and coalesce with smaller particles to
form particles which settle at rates higher
than the parent 'particles. The opportunity
for contact increases with liquid depth.
The removal efficiency is therefore
6-3
-------
In - Depth Filtration
dependent on both the clarification rate
and depth. There is no satisfactory
formulation available for evaluating the
flocculation effect on sedimentation. This
effect has to be evaluated by performing
a settling-column analysis. Samples are
withdrawn from several depths at different
time intervals from the column. The
solids concentration of each sample is
determined and the fraction removed
(relative to the initial concentration) is
plotted on coordinates of time and depth to
correspond to the sampling. Isoconcentra-
tion lines are then drawn to describe a
depth-time ratio equal to the minimum
average settling velocity of each fraction.
Graphical summation of the removed
fractions at the detention time and depth
values compatible with the design clari-
fication rate will give the overall removal
efficiency to be expected. An illustrated
example .is presented in reference 3.
D Filtration 7
1 Flow determination
The frictional headloss through a clear
bed of media can be calculated by the
following relationship which was
developed experimentally by Rose.
L
h =["'.
L tHD
P
VS2
g
1
4
e
where:
hL = headloss, ft
f'" = bed depth, ft.
D = particle diameter, ft.
p
I = coefficient of nonshericity
e = porosity
Vs = average superficial velocity
ft/ sec
g = gravitational constant ft/ sec2
The dimensionless friction factor f' " is'
related to the drag coefficient Cd'
6-4
f'" = 1.067 Cd
When flow is nonlaminar C is related
to the Reynolds number as ~ollows:
C = 24 + 3
d N N
re re
+ 0.34
For laminar flow, Cd approaches a value
24
ofN'
re
This expression applies strictly to flow
through a clean bed. Once particulate
matter has been deposited in the medium
the relationship is not validbecuase
particulate matter changes both the'
porosity of the bed and the shape of the
bed particles. A valid expression to
calculate the headloss across a filter in
operation has not been developed. In
practice, it is simply measured.
2 Removal mechanisms
The removal mechanisms described
below have all been advanced as a
possible mechanism of particle removal
from a fluid during passage through a
porous medium. Just as no exact
mathematical model has been developed
to describe filtration, no single mech-
anism can explain all the observed
phenomena. It is a number of interacting
mechanisms.
a Straining
The most logical mechanism for the
removal of particulate matter from
water is the deposition of particles
too large to enter the openings of
much smaller particles, which
actually occurs. In addition, this
mechanism implies that the size of
particles removed is independent of
the particle composition, which is
not the case. Removal is dependent
on the nature of the particle.
-------
In-Depth Filtration
b Sedimentation
f Diffusion
This theory considers each opening
in the filter medium as a tiny
sedimentation basin. If this were
the case, the removal efficiency
should be size dependent which is not
actually observed. In addition, it
offers no explanation why particles
remain settled out, particularly
since higher velocities will develop
as the openings become constricted.
The particulate matter has a greater
density than the fluid transporting it.
Consequently, as the streamlines of
the fluid bend around an obstacle,
the greater inertia of the denser
particles forces them to the outside
of the turning radius where they may
inpinge on another obstacle. The
removal efficiency of such a mechanism
should increase with hydraulic loading
which in practice does not occur.
Soluble and small suspended particles
will migrate toward areas of low
concentration which accounts for
particulate removal into regions of
the filter where flow is essentially
zero. Colloidal particles will
accumulate in regions of low shear
even if they have to migrate against
a concentration gradient. An equili-
brium with the opposing shear forces
probably develops. These forces are
definitely operative but are they of
sufficient magnitude to be of
significance?
c Inertial impaction
g Van der Waals Forces
Small colloidal particles could be
forced out of the moving streams and
into contact with a solid surface by
random molecular motion but the
effect is appreciable only for particles
less than 2 ~ in diameter.
These attractive forces vary inversely
as the seventh power of their distance
of separation and become significant
only when the distance approaches
the molecular diameter. Repulsion
forces exist at these distances but
are of a considerably smaller order
of magnitude. Since cohesive forces
are additive, the force is inversely
proportional to the cube of the
separation distance for large aggre-
gates. Van der Waals Forces are
not considered a dominant removal
mechanism because of the rapid
decrease of effect as distance
increases. They may be significant,
however, in preventing the redispersion
of removed particulates.
d Brownian movement
e Chance contact
The chance contact of suspended
particles with each other or with
surfaces of the medium caused by
flow through a constriction has been
advanced as a logical mechanism,
particularly if some other force
preventing redispersal can be postu-
1ated. The probability of removal
is thought to be directly proportional
to the square of the particle diameter
and inversely proportional to the cube
of the diameter of the bed particles.
h Electrokinetic effects
Dispersed particles in an aqueous
environment with similar surface
charges will exhibit a repulsive
force. The presence of ions in the
aqueous phase greatly influences
these repulsive forces. Surface
charges on the filter medium can
determine the stability of the colloidal
dispersion. A preferential removal
of smaller negatively charged parti-
cles has b~en observed by investigators
6-5
-------
In-Depth Filtration
-------
when a positive charge was imparted
to the filter medium. The electro-
kinetic effect is opposite from what
would be expected from a purely.
physical mechanism.
Flocculation
The chance contact of suspended
particles with each other and the
filter medium provides an opportunity
for floc formation and/ or growth.
These larger particles are mechani-
cally held in the crevices of the filter
until they are sheared off and driven
deeper into the bed. It has been
observed that filtration efficiency
decreases with temperature due to
the higher viscosity of the water and
decreased flocculation.
j Biological activity
This is the predominant removal
mechanism for slow sand filters. It
is not significant in rapid sand fiUra-
tion because of the velocity and
frequency of backwash. Highly
polluted waters with growth nutrients
enhances biological activity and its
influence on particulate removal.
k Chemical influences 8
Recent work has shown that the
chemical composition of the water
had a significant effect on the re-
moval of ferric floc. The phosphate
ion was particularly influential.
More work is required in this area.
The physical, chemical and biological
influences on filtration have not been
sufficiently defined; nor has a mathe-
matical or empirical formula been
developed to account for the changes
occurring during filtration to permit
the scientific design of a filter.
Practical experience with successful
installations and field investigation
of new applications is still required
to design an effective and economical
filter plant.
6-6
III
DESIGN CRITERIA
A Rapid Mix Compartment
1 Detention time 30-60 sec.
-1
2 Design "G" 500 sec.
3 Gt 30, 000
B Flocculation Chamber
1 Detention time 10-30 min.
-1
2 Design "G" 20-30 sec.
3 Gt 54, 000
9
Note: Hudson advocates Gt values of
up to 200,000 to produce a dense sludge.
C Sedimentation
1 Horizontal-flow basins
2
a Overflow rate - 700 GPD/FT
b Detention time 1-3 hrs.
c Used primarily for coagulants forming
light hydroxide floc.
2 Up-flow clarifier
a Rise rate 0.5 - 1. 5 GPM/FT2
b Detention time 1-3 hrs.
c Sludge recirculation rate - up to 10%
of design flow
d Used primarily for coagulants forming
heavy precipitates capable of with-
standing the high shear forces during
recirculation.
D Filtration
1 Media
a Types
1) Sand
2) Anthracite coal
3) Garnet
-------
In-Depth Filtration
b Effective size
1) Sand 0.35 - 0.55 mm
2) Anthracite 0.75 - 1. 80 mm
3) Garnet 0.15 - 0.30 mm
c Depth
1) Sand - 6 in.
2) Anthracite coal - 18 in.
3) Garnet 1. 5 - 3.0"
4) Total 24 - 30 in.
d Configuration
1) Single
2) Dual media
3) Multi-media
2 Underdrains
a Pipe lateral use 1.8 in. gravel to
support media
b Commercial underdrain systems, use
minimum of 12 in. of gravel to support
media.
3 Rate of filtration
2
a Range 2 - 10 GPM/FT
2
b Mean 3 GPM/FT
4 Backwash rate
a Water
2
15 - 25 GPM/FT
b Air - 4 cfm/FT2
5 Mode of operation
a Gravity
b Pressure
IV
OPERATING PARAMETERS
A Coagulation
1 Chemical addition sequence
a pH or alkalinity adjustment
b Primary coagulant
c Coagulant aids
d Filter aids
2 Point of chemical addition
a Prior to rapid mix
b During rapid mix
c After rapid mix, prior to flocculation
d After setting, prior to filtration
3 Optimum chemical dose determination
a Empirical methods
1) Jar tests
a) Floc size
b) Supernatant clarity
c) Appearance of first floc
d) Rate of settling floc
e) Filterability characteristics
f) Zeta potential
2) Pilot filters to predict end of run
conditions
3) Production filters performance
a) Length of run
b) Product quality
6-7
-------
In - Depth Filtration
b Analytical methods
1) Residual coagulant
2) Suspended solids
c Monitoring methods
1) Turbidity
a) Settled
b) Filtered
2) Conductivity
3) pH
B Filtration
1 Headloss termination criteria (usually
8 ft. o.f water for a gravity system)
2 Product quality criteria for termination
(1.0 J.T.U.)
v
PERFORMANCE IN A TERTIARY
TREATMENT APPLICATION
A A lum Coagulation
1 Lebanon, Ohio
The Federal Water Pollution Control
A dministration has conducted a study
of alum coagulation and filtration at
Lebanon, Ohio. The final effluent
from the activated sludge treatment
plant had an average suspended solids
concentration of 17 mg/l during the
study period. The tertiary treatment
consisted of chemical coagulation,
flocculation, sedimentation in a
horizontal-flow basin and gravity
filtration. Table 1 indicates the range
of variables investigated and the optimum
value for each.
Under the optimum operating conditions
it was possible to consistently produce
a high quality water removing essenti;illy
all suspended solids and acid hydrolyzable
phosphate. Filter runs have exceeded
30 hours to a terminal headloss of 8 feet
6-8
at a filtration rate of 3 gpm/ft2. Floc
strengtheners were required to produce
a filterable floc and prevent premature
turbidity breakthrough with the dual
media filters. Typical results achieved
with optimum operating conditions are
shown in Table 2.
2 Nassau County, New York
Pilot plant experience at Nassau County
has also demonstrated the effectiveness
of alum coagulation and filtration in a
tertiary treatment application. Two
desirable aspects of alum coagulation
are the reduction in phosphate and pH,
both are treatment objectj.ves. Lime
and iron salts were tried but found
ineffective at Nassau, possibly because
of the low alkalinity of the effluent.
Alum dosages of 50-300 mg/l are used
with a cationic polymer as a floc
strengthener. A high-rate sludge
blanket clarifier with sludge removal
and a high-rate up flow clarifier with
sludge recycle were found equally
effective for clarification with a 2
maximum rise rate of 0.5 gpm/ft and
detention time of 2 hours. Dual media
filters consisting of 30" of 0.9 mm coal
and 6" of 0.35 mm sand produced water
with turbidities less than 0.4 J. T. U.
when a small amount of alum (8-10 mg/l)
and polymer (0. 5 mg/l) were added
after clarification and prior to filtration.
The optimum filtration rate was found
to be 4 gpm/ft2 with filter runs of 8 to
24 hours.
B Lime Clarification
1 Lebanon, Ohio
Recently a 75 gpm single stage, up-flow
sludge blanket clarifier was put on
stream at the FWPCA Pilot Plant in
Lebanon, Ohio. The lime addition of
approximately 300 mg/l as Ca(OH)2 is
controlled by pH.
-------
--------------------
----
In- Depth Filtration
TA BLE 1.
OPTIMUM OPERATING CONDITIONS FOR TERTIARY
CLARIFICATION USING ALUM COAGULATION
FILTRATION AT LEBANON, OHIO
--------- ---------
Variable
Range Optimum Value
---- -----------r.-'--"'--'
300-900 700
( 1) Overflow Rate
GPD/FT2
(2) Alum Dose
mg/l
(a) Clarification
(b) Phosphate removal
(3) Type of Media
0-500
0-500
150
300
(a) Sand
(b) Coal
(c) Sand and coal
(c) Sand and coal
dual media
2-5
1. 4 @ 4 GPMfFT2
1. 3 @ 3 GPM/FT2
4
(4) Effective Size of Coal mm.
Sand Size 0.45 mm.
2
(5) Filtration Rate GPM/FT
(6) Floc Strengtheners
0.75-1.80
(a) Activated silica
(b) Cationic polyelectrolyte
C-7*
+++
(7) Source of A 1 on Phosphate
Removal
0-15 mg/l
0-1. 5
3 mg/l
1. 5 mg/l
(a) Alum
(b) Aluminate
(c) Both
0-100%
0-1000;.
No difference until
pH exceeded 8. O. This
occ~.f.fed when 90% of
A I came from the
aluminate
*A product of Rohm and Haas Company, Philadelphia, Pa.
1.700 GPD/FT2 based on economic consi~erations; optimum solids removal
efficiency occurred at the 300 GP D/ FT rate.
.9
-------
In - Depth Filtration
TABLE 2.
TYPICAL RESULTS UNDER OPTIMUM
OPERA TING CONDITIONS
Process Stream
Suspended
Solids, mg/l
Acid Hydrolyzable =-
Phosphate, mg/l PO ~
Turbidity
J.T.U.
Secondary Effluent 45.6 22.4 12.2
Chemically Treated *
Settled Effluent 11. 0(760/0) 2.2(900/0) 1.5(880/0)
Filtered Product 1. 3(970/0) 0.9(960/0) 0.5(960/0)
Operatmg Conditions:
Overflow Rate:
Alum Dose:
2
700 GPD/FT
82 mg/l
68 mg/l
3 mg/l
Aluminate Dose:
Silica Dose:
* Denotes Removal Efficiency as 0/0
The clarifier is operated at a rise rate
of 1 gpm/ft2. Dual media filters;
consisting of 18 inches of 0.75 mm coal
over 6 inches of 0.45 mm sand polish
the clarifier effluent2 The filters are
operated at 2 gpm/ ft to a terminal
headloss of 9 feet of water. Some
preliminary operating results are shown
in Table 3.
The suspended solids concentration of
the settled sludge ranged from 1. 5 to
2. 5% solids by weight and was further
concentrated to 12-15% solids by weight
in a gravity thickener and additional
drying on sand beds prior to disposal
to a land fill.
C Jar Test Results
When effluent quality standards require the
removal by phosphate in addition to clari-
fication, significantly higher coagulant
doses can be anticipated. A rou~
stoichiometry of 2 moles of A I + to 1
mole of P has been found empirically, to
exist in tertiary treatment applications.
10
Filtration Rate: 3GPM/FT2
Effective Size of Coal: 1.325 mm
Length of Run: 25 hours
Sludge Concentration: 1.2% by weight
Solids Concentration in
Filter Backwash Water: 475 mg/l
In addition, as shown in Figure 1 all the
primary coagulants are equally effective
in removing phosphate. Thesi{esults were
obtained from jar test studies on a
pilot plant activated sludge effluent.
VI
INDUSTRIA LAPP LICA TIONS 12
A Food Industry
1 Meat
Grit removal, flocculation and sedi-
mentation will remove 50% of the total
volatile solids and BOD and 75% of the
suspended solids in a manure-bearing
wastewater. Alum coagulation removes
an additional 2. 5 mg/l of BOD and
1. 5 mg/l of grease per mg/l of alum
added. The chemical costs for 50%
and 78.5% increased removal efficiency
is $0.64 to $2.05 per 100 lb. BOD
removal, respectively.
2 Poultry process wastes
Plain sedimentation can remove 17-28%
of the BOD and 30-65% of the S. S.
-------
In - Depth Filtration
TABLE ~- PRELIMINARY OPERATING RESULTS FOR LIME
CLARIFICATION OF SECONDARY EFFLUENT
TURBIDITY, J. T . U . SUSPENDED SOLIDS -;j LENGTH OF
PHOSPHATE,P04
FILTER RUN
pH mg/l mgfl (Hrs.)
Influent Settled Filtered Influent Settled Influent Settled
9.0J 9.7 9.9 0.27 - - - - 26
9.66 11.7 8.5 0.2 30.2 13.2 25.8 1.74 53.5
0.0 14.3 10.2 0.17 47.6 36.1 '20.8 0.7 -
1
I
dt
Go
III
0
..
~
at
E
..
'U
t-
~
5;
0
i
....
~
e
....
~
9
II)
'U
QI: 5
35
-
-e- ALUM
-0- FERRIC SULFA TE
--0-- LIME, as Ca (OH~
--t.-- LIME + 20 mg/L
ALUM
o
o
100 200 300 400
COAGULANTDOSAGE,mgA
500
600
FGURE 1 PHOSPHATE REMOVAL BY VARIOUS COAGULANTS
11
-------
In - Depth Filtration
3 Canning wastes
Lime followed by either ferrous sulfate
or alum can remove 50% of the BOD of
pea and corn wastes and less than 20%
of the BOD in fruit wastes. The
treatment is considered a pretreatment
to biological secondary treatment.
B Manufacturing Industries
1 Organic chemical wastes
Chemical coagulation and sedimentation
is used to treat colloidal and oily wastes.
2 Pulp and paper industry
Without dispersing agents present,
70-80% of the solids in a paper making
waste can be removed by plain sedi-
mentation. Alum, activated silica and
polyelectrolytes can increase the removal
of suspended solids to better than 90%.
3 Textile industry
Chemical coagulation gives good results
but is more expensive than biological
treatment methods.
C Mineral Industries
1 Iron and steel industry
a Flume water clarification
Treatment consists of primary
clarification in a scale pit followed
by secondary clarification and oil
removal. Lime, ferric sulfate and
polyelectrolytes are used as coagu-
lating agents. The water is usually
suitable for reuse.
VII. CAPITA L AND OPERA TING COSTS OF
CHEMICAL COAGULATION, SEDIMENTATION
AND FILTRA TION
A Capital Costs for a 10 MGD Chemical
Coagulation, Sedimentation and Filtration
Plant
6"12
2
Design Basis: Overflow Rate 700 gpd/ft
2
Filtration Rate 4 gpm/ ft
1 Sedimentation basin:
(including sludge collection
system and appurtenances)
$194,300
2 Influent pump:
3 Chemical feed equipment
4 Land at $10, OOO/acre
5,400
48,000
15,000
462,500
725,800
5 Filters:
6 Total capital costs
B Operating and Maintenance Costs
1 Amortization, 20 years at 4%:
2 Labor:
53,500
26,000
172,300
9,125
3 Chemicals:
4 Power:
5 Maintenance:
6 General overhead and
Administration:
3,630
5,820
7 Total annual cost:
270,375
C Unit Cost for Clarification and Phosphate
Removal
Based on annual water production of
3.65 X 109 gallons the unit cost of water
is $0.074/1, 000 gal.
VIII
ULTIMATE DISPOSAL OF CHEMICAL
SLUDGES
A General Disposal Techniques
1 Lagooning
One of the most widely practiced methods
of disposal. It is to be considered a
stopgap method because it creates both
a nuisance and a liability.
2 Sludge drying beds
Must be evaluated ~n terms of local
climatic conditions such as evaporation
rate and mean annual rainfall. A s much
as 3 -5 acres per mgd of capacity may
be required.
-------
In-Depth Filtration
3 Disposal to a waste sump
This is often a more "ultimate" type of
disposal depending upon future land use
and degree of isolation from the surround-
ing environment.
a Costs
Mode of Transportation
Rail
Cost in Mills Per Ton Mile
34-40
15
Truck
Bar ge
4
2
Pipe
Pumping a distance of 250 miles is
considered reasonable and a minimum
scour velocity of 5 ft/ sec is recom-
mended.
Power costs are 12% of the total cost
per year if the optimum hydraulic
pipe size is installed.
4 Vacuum filtration
Sludges from coagulation facilities are
often thin and generally difficult to
dewater. Intermediate dewatering
operations such as thickening should be
considered to reduce the volumetric
load on the filter. Significant savings
can be achieved by altering the process
conditions to yield a sludge of superior.
dewatering characteristics.
Vacuum filtration is attractive because
of small space requirements, all year
operation, widespread applicability, and
satisfactory properties of the cake.
5 Recovery and reuse
Recovery and reuse of chemical sludges
offers the most promising solution to
the disposal problem because significant
savings in chemical costs can be realized
and the ultimate disposal problem is
minimized.
B Recovery and Reuse of Aluminum
Hydroxide Sludge
Since aluminum hydroxide is amphoteric
it will ionize in both acid and base. The
acid recovery scheme is not practical for
wastewater treatment because of the
difficulty associated with separating the
phosphates from the recovered aluminum.
The alkaline method of alum recovery is
the most promising solution and merits
further investigation. The recovery
scheme involves dissolving the aluminum
hydroxide and adsorbed phosphates by
raising the pH to 11. 9 with NaOH. The
reaction converts the aluminum hydroxide
to sodium aluminate. Calcium chloride is
then added which reacts with the phosphates
to form the insoluble precipitate of
tricalcium phosphate. The calcium phos-
phate is separated by sedimentation and/ or
filtration and the comparatively phosphate-
free sodium aluminate solution is reused.
The reactions for this process are as
follows:
pH = 11. 9
Al (OH). [PO-j+4NaOH--.NaAI0 +Na3PO +2H20+30H
3 4 2 4
NaAl 02 + Na3PO 4 + 3CaC12 -. NaAl 02 + Ca3 (PO 4)2 + 6NaCl
(13) .
Culp, et aI, 10 laboratory tests
achieved 75% recovery of the feed alum by
sedimentation and 85% by filtration compared
to the 93% recovery reported by Lea,
et al( 14). Lime can also be used to recover
alum. The lime not only raises the pH
but also supplies the necessary calcium
to precipitate the phosphates. Although
the percent recovery is less with lime
(35%); the cost of chemicals to regenerate
a ton of alum is lower. If alum recovery
proves successful, treatment costs can
be reduced to $0.05/1, 000 gal. The
problem with alum recovery to date has
been the decreased efficiency of the
recycled sodium aluminate in removing
phosphate.
13
-------
In-Depth Filtration
C Recovery and Reuse of Lime Sludges
A two stage lime precipitation treatment
system including facilities to recover and
reuse the lime is being installed at the
FWPCA Pilot Plant in Washington, D. C.
Lime is added with or without a magnesium
salt or other coagulant aid as required to
the first stage precipitator. At a pH of
11.3 the calcium and magnesium phosphates,
sulfates and carbonates are precipitated
together with the organic sludge. Com-
pressed stack gases which are rich in
carbon dioxide are usually used to
recarbonate the waste stream as it enters
the second stage clarifier. The excess
calcium is precipitated as almost pure
calcium carbonate to a pH of 9. O. Sludges
from the two stages are dewatered by
centrifugation or vacuum filtration and then
calcined in a multiple hearth furnace at
18000F to produce calcium oxide. Calcining
destroys the organic matter, but phosphates
accumulate as inert tricalcium phosphate
along with other inert minerals which must
be wasted. The calcium oxide is then
slaked and reused. The chemical cost
savings realized by recalcining versus
purchase of fresh lime is negligible. The
significant cost savings are the reduced
disposal costs. Cost estimates for the
Washington, D. C. Plant at the 250 mgd
size indicate that savings of up to
$0.02/1, 000 gal. can be realized.
REFERENCES
1 Cohen, Jesse M. and Hannah, Sidney A.,
"Coagulation and Flocculation" A
Chapter in Water Quality and Treatment,
American Water Works Assn., New York,
N. Y., to be published.
2 Babbitt, Harold E. and Doland, James J.,
Water Supply Engineering, 5th Ed.
McGraw-Hill, New York (1955).
3 Rich, Linvil G., Unit Operations of Sanitary
Engineering, John Wiley and Sons, Inc.,
New York, 1961.
4 Beam, Elwood L., "Study of Physical
Factors Affecting Flocculation, " Water
Works Engineering, January 1953, p. 33.
6~14
5 Camp, T.R. and Stein, P.C., "Velocity
Gradients and Internal Works in Fluid
Motion." J. Boston Soc. Civ. Engrs.,
30:219 (1943).
6 Robeck, G. G., "High Rate Filtration Study
at Gaffney, South Carolina, Water
Plant, "USPHS, R.A. Taft San. Engr.
Center, Cincinnati (1963).
7 Craft, T. F., "Review of Rapid Sand
Filtration Theory, " JA WWA, April 1966,
p. 428-439.
8 O'Media, Charles R. and Crapps, David K.,
"Some Chemical Aspects of Rapid Sand
Filtration;' JAWWA, 56, 1326 (1964).
9 Hudson, H. E., Jr., "Physical Aspects of
Flocculation, JA WWA, July 1965,
p. 885-892.
10 Kreissl, J. F., Robeck, G. G. and
Sommerville, G.A., "Use of Pilot
Filters to Predict Optimum Chemical
Feeds." Presented 87th Annual
Conference A WWA, June 1967, Atlantic
City, N. J.
11 Garland, C. F. and Shell, G. L., "Integrated
Biological-Chemical Wastewater Treat-
ment, " Final Report on FWPCA Contract
No. PH 86-63-220 from Infilco/GATC,
November 1966.
12 Gurnham, C. F., Industrial Wastewater
Control, Academic Press, Inc.,
New York, 1965.
13 Culp, G. and Slechta, A., "Recovery and
Reuse of Coagulant from Treated Sewage. "
Final Progress Report USPHS Demon-
stration Grant 85-01, South Tahoe
Public Utility District, Feb. 1966.
14 Lea, W. L., Rohlich, G.A. and Katz, W. J.,
"Removal of Phosphates from Treated
Sewage." Sewage and Industrial Wastes,
. .
This outline was prepared by John J. Convery,
Sanitary Engineer, Municipal Environmental
Research Laboratory, USEPA, Cincinnati,
Ohio 45268
Descriptors: Wastewater Treatment,
Waste Treatment, Water Pollution
Treatment. Coagulation, Flocculation.
Tertiary Treatment
-------
REVERSE OSMOSIS
I
INTRODUCTION
A The phenomenon of osmosis was first
observed in the mid-18th century when
studies were being made on biological
membranes.
B It was noted that when two solutions of
different concentration were separated by
a semi-permeable membrane, the solvent
would flow from the dilute to the concen-
trated side. A semi-permeable membrane
is defined as one that will permit passage
of some materials (usually a solvent)
while rejecting others.
C The first experiments with artifically-
prepared membranes were conducted in
the mid-19th century when a membrane
was made by precipitating copper ferro-
cyanide in the pores of porcelain.
D By 1920, interest in osmosis waned and,
except for its biological importance, it
was considered a laboratory curiosity.
E Developments in the last ten years by
C. E. Reid and others of the University of
Florida and S. Loeb and others at UCLA
showed that the process can be reversed
by applying pressure to the higher con-
centration side and that it has potential
as a process for recovering reuseable
water from a contaminated or saline
source.
II
THEORY
A Figure 1 illustrates the principle of
osmosis and reverse osmosis. At equil-
ibrium, the pressure on the more
concentrated solution is known as the
osmotic pressure. This pressure is
dependent entirely on the difference in
concentration between the two solutions
and is not a function of the type of mem-
brane' provided it is semi-permeable.
SE. TT. pp. 10.7.77
B It is well known that the "activity" or
"chemical potential" of a solvent decreases
when in a solution. Thus, the activity
of water in distilled water is greater than
the activity of water in a salt solution.
This greater activity is thought to be the
driving force that, in normal osmosis,
causes water to flow through the membrane.
In brackish and sea water, this driving
force is about 1 psi for every 100 ppm
salt in solution.
C The actual mechanism of transfer of
water through a reverse osmosis mem-
brane is not completely understood. The
most generally-accepted theory is that
water passes through the membrane by
successive transfer from one adsorption
site to the next. In the case of cellulose
acetate, the most common membrane
material, the adsorption forces arise from
hydrogen bonding.
D The rate of transfer of water through a
membrane in reverse osmosis is directly
proportional to the applied pressure minus
the osmotic (back) pressure and inversely
proportional to the thickness of the
membrane. Increasing temperature,
since it raises the activity of the water,
will result in a greater water flux through
the membrane.
E As stated before, the flux of water is
directly proportional to the pressure
driving force. The flux of salt, however,
is a function only of the salt concentration.
Raising the applied pressure, therefore,
reduces the salt content of the product.
III
MEMBRANES AND THEIR MANUFACTURE
A Reverse osmosis was first made practical
by the discovery (by Dr. Loeb and co-
workers at UCLA) of a process for making
a cellulose acetate membrane with a very
thin active layer. These membrane are
about 100 microns in total thickness. The
active, dense layer that transmits the
7-1
-------
t....:
Nonoal Osmosis
fresh
vater
A
I
fresh
water
08lllot:1o Equ1libr1ua
vaste
vater
J
THE PRINCIPLE OF REVERSE OSMOSIS
(WASTE WATER)
Figure 1
Reveree OBDI08U
...........,.
.".",",','."'.',',',' ,','
........... ..
........... ..
it
fresh
water
p
vaste
water
C
I~
'1
UJ
(1)
o
UJ
S
o
UJ
.....
UJ
VI
....
VI
$
~
-------
Reverse Osmosis
water and rejects the salt is only 0.25
microns thick - the rest is a spongy back-
up material that has little, if any,
desalination properties.
B A typical formulation for making cellulose
acetate membranes is shown in Table 1.
TABLE 1
CASTING SOLUTION FOR CELLULOSE
ACETA TE REVERSE OSMOSIS MEMBRANES
Material
Weight
Cellulose Acetate
Magnesium Perchlorate
22.2
1.1
Acetone
66.7
10.0
Water
This casting solution is poured on a glass
plate and its thickness is adjusted by a
doctor blade. The plate is then immersed
in ice water for 30 to 60 minutes during
which essentially all of the magnesium
perchlorate is removed from the membranes.
The membrane is then immersed in hot
water for 3-10 minutes. The water tem-
perature is critical - from 550 C on up,
the salt rejection properties are improved
at the expense of flux through the mem-
brane. Continuous casting procedures
have been developed but they are
proprietary and nothing has been published
on them.
C The degree of acetylation of the cellulose
is an important factor. The optimum
membrane is produced from cellulose
acetate containing about 2.5 acetate groups
per cellulose molecule. The tri-acetate
is produced and it is hydrolyzed to the
2. 5 level. Further hydrolysis degrades the
membrane - this is controlled in service
by maintaining the pH of the feed water
at about 5.5.
D Compression of the spongy under-layer
at the elevated pressures necessary
during operation results in flux decline.
Efforts are underway to produce a thin
(0.25 micron) dense layer that could be
put on a porous, non-compressible
substructure.
E Considerable effort has been and is being
expended on the development of new
membrane materials. As yet, none of
those investigated has proven as effective
as cellulose acetate. Some of the more
promising materials studied to date
include poly (vinylene carbonate),
polyviny lpyrollidone and graphitic oxide.
F A recent development in this field has been
made at the Oak Ridge National Labora-
tories of the Atomic Energy Commission.
Dr. Kraus and co-workers have developed
a dynamically-formed membrane having
permeation rates of 100-300 gallons/
day ft2 (gfd) compared to a typical rate
for cellulose acetate of 20-30 gfd.
1 The rejecting layers are formed by
pumping water containing a hydrous
oxide (zinconium or thorium) or a
polyelectrolyte [ poly (vinylbenzyl-
trimethyl-ammonium chloride)] over
a porous body.
2 The porous body can have a pore size
up to 5 microns in diameter.
3 Salts containing po!yvalent counter-
ions (Mg +++ or SO;) are poorly rejected
and destroy the reJection capacity of
the membranes for mono-valent salts.
IV
ENGINEERING OF REVERSE OSMOSIS
UNITS
A Several different designs have been
proposed to meet the design objective of
producing the maximum amount of quality
product per unit cost.
7-3
-------
Reverse Osmosis
1 The "flat plate" membrane configuration
is similar to a plate and frame filter
press. These units have been tested
extensively on brackish- and sea-water
but little work has been done on waste-
waters.
2 The spiral-wound membrane is illustrated
in Figures 2 and 3. The objective of
this design is to increase the surface
area to volume ratio. This configuration
has been evaluated on brackish- and
sea-water and is being evaluated on
wastewater at the FWPCA Pomona.
California pilot plant.
3 The tubular unit is shown in Figure 4.
Another form this can take is that of a
metal tube with weep-holes drilled for
delivery of the product water.
4 The newest configuration consists of
hollow fibers with water passing from
the outside to the inside of the fibers.
Figure 5 shows a schematic of such a
unit.
B Two different concepts have been employed
in the design of these units. The first
three described above depend on relatively
high permeation rates (20 gfd). The
hollow fiber units rely more on very high
surface area per unit volume with low
permeation rates (0.15 gfd).
C Equipment must be designed to minimize
the "boundary layer" effect - that is.
the existence of a high concentration layer
at the membrane surface. This problem
is related to the permeation rate. becoming
more severe as the rate increases.
D Rapid membrane replacement and
inexpensive pressure construction are
also goals of the designers in this field.
. Membrane life is still an unknown quantity.
particularly in wastewater treatment. so
configurations in which membrane replace-
ment is costly could be impractical.
7-4
v
LA BORA TORY INVESTIGA TIONS ON
WASTEWATER
A Aerojet-General conducted a 1-year
study on the application of reverse osmosis
to wastewater. The equipment consisted
of two 0.5 ft2 membranes in series (flat-
plate). The concentrate stream was
returned to the feed tank to determine
the effect of increasing concentration.
The principal findings are listed below.
1 The product quality was excellent.
Table 2 shows the product quality from
two test runs.
2 Water flux through the membrane
decreased rapidly. A cid addition to
the feed to maintain a pH of 5 minimized
this problem but flux reduction with
time was still severe.
3 After a period of several hours. the
flux at 1500 psi feed pressure was
about the same as that obtained at
750 psi.
4 The flux through "loose" membranes.
though high originally. rapidly declined
to the same level as the "tight" (low
flux. high salt rejection) membranes.
5 Removal of organics from the feed by
carbon adsorption was helpful in -
reducing the rate of flux decline.
6 Test results showed that biological
attack of the membrane could be a
severe problem.
B The General Atomic Division of General
Dynamics ran several spiral-wound
modules on wastewater at the Pomona
Water Renovation Plant of the Los Angeles
County Sanitation Districts. This was a
continuous flow-through operation
recovering about 5% of the feed as product.
1 Water quality was again excellent as
can be seen in Table 3. The "A"
modules contained "tight" membranes.
the, B - "loose" membranes and the "c" -
intermediate. '
-------
Reverse Osmosis
ROLL TO ....."':">
ASSEMBLE......""""'" ....."
~""""f
"""......""" I
/ .....
BRINE-SIDE ,,/ ..........."""
SEPARATOR" / ...... /
SCREEN ,,::'.-......
/
I
I
I
I
\
\
/ "
PRODUCT WATER" "-
"-
......
......
......
"-
......
......
PRODUCT-WATER- SI DE BACKING'
MATERIAL WITH MEMBRANE ON
EACH SIDE, GLUED AROUND EDGES
AND TO CENTER TUBE
MEMBRANE
PRODUCT-WATER-
SIDE BACKING
MATERIAL
MEMBRANE
BRINE-SIDE SPACER
Figure 2. A Spiral-Wound Reverse Osmosis Module
Reprinted with permission from General Dynamics,General Atomic Division
5
-------
CP
GENERAL DYNAMICS
General Atomic Division
::0
(1)
<
(1)
'1
en
(1)
o
en
8
o
en
....
en
100,000 GALLONS PER DAY WATER TREATMENT PLANT
PRETREATMENT
EQUIPMENT
AIR COMPRESSOR
PRESSURE
VESSEL SKID
/'
Figure 3
5005 -MD-GO
-------
Reverse Osmosis
(1) FIBERGLASS TUBE
(2) OSMOTIC MEMBRANE
(3) END FITTING
(4) PVC SHROUD
to collect product water
(5) PRODUCT WATER
(6) FEED SOLUTION
(7) EFFLUENT
r Igure 4. A Tubular Reverse Osmosis Unit
Reprinted wit!1 permission of Havens Industries
2 As bt '.,)re, the "loose" (high flux-low
rejection) membranes plugged fastest
and were soon delivering no more than
the "tight" membranes.
3 Flux decrease was still a problem,
ranging from 25% in 1500 hours for the
tight membranes to 86% in 290 hours
for the loose membranes.
C Similar work was Jone on a spiral unit by
the New Jersey Department of Health and
others on a sewage plant effluent and on
water from the Hackensack River. Results
were similar. Bacteriological removals
were determined during this study and the
MPN reduction/ range was from 80 to 99.9%.
VI
PILOT INVESTIGATION ON WASTEWATER
A The only pilot investigation on wastewater
of reasonable duration has been conducted
at the Pomona pilot plant on a 5000 gpd unit
utilizing the spiral membrane configuration.
B The first year of operation was plagued
by operating difficulties and no encouraging
results were obtained. The more signif-
icant problems are listed below.
1 Some of the feed by-passed the mem-
brane module, going between the outside
of the module and the pressure casing.
This caused a flow decrease through
the feed channel, thus accentuating the
boundary-layer effect.
7
-------
~ ~
ro ro
CP '0 <:
I-j ro
..... I-j
~ [/J
CD ro
Q. 0
:;1 [/J
..... S
<+
::r" 0
'0 [/J
.....
ro [/J
I-j
s
.....
[/J
[/J
.....
0
::s
0
....,
tr1
~
0..
~
'U
0
::s
<+
0..
ro
Z
ro
S
0
~
I-j
[/J
p)
::s
0..
()
0
S
'0
p)
::s
«:
Figure 5. A Hollow-Fiber Reverse Osmosis Unit
-------
Reverse Osmosis
TABLE 2
VA'l'EB. Ift11PAT. ABALYSIS
feed Vater Prod)1ct Product
(Typical) (Test 18) (Test 28)
MID 1)18 'D!IR
TDS 550 15 28
ABS 4.5 0.1 0.1
COD 95 2 6.0
pH 51 6., 5.5
CatiODS
.+ 85 4.6 6.1
K++ 40 4.9 2.7
BIt 25 0 '.9
4
ea++ 125 0 2.6
JIg++ 50 0.7 0
'h++ B. D. 0 0
AI!HH
Cl- 65 20.' 22.1
10; 2 0 0
~iCO; 260 8.1 '.5
0'. 0 0 0
,
,.. 200 1.7 ,.,
504
810; :50 2.9 2.9
PO; (total) 25 0 0
B.D. - not dete:naiD8d
lpH a4jute4 4uriq test
7-9
-------
Reverse Osmosis
TABLE 3. PRODUCT QUALITY - REVERSE OSMOSIS TREATMENT OF WASTEWATER
Product From Product From Product From
Feed Module 1-6-3 Module 3-24-1 Module 1-22-1
Water* 5A 1 ** 3B2 ** 5C2 **
Alkalinity, ppm CaC03 213 < 25 88 37
Ammonia Nitrogen, ppm N 11.2 2.8 7.0 3.5
Total Nitrogen, ppm N 12.6 2.8 7.4 3.5
Specific Conductance, J.lmhos/ cm 933 60 399 215
Chloride, ppm Cl 92 4 68 32
Hardness Total, ppm CaC03 195 3.8 27.4 6.1
Phosphate Total, ppm PO 4 5.6 0.2 2.2 0.4
Potassium, ppm K 17.8 1.2 10.4 5.0
Sodium, ppm Na 107.2 7.5 63.0 29.5
Calcium, ppm CaC03 138 2.8 20.3 3.8
Magnesium, ppm CaC03 58 1.4 7.1 2.3
Dissolved Solids, ppm 531 28 224 118
Sulfate, ppm SO 4 71.4 5.3 2.1 3.9
Fe, ppm Fe 0.10 0 0 0
ABS, ppm 3.8 < 0.1 1.1 < 0.1
COD, ppm ° 41 4 8 4
pH 7.2 6.6 7.1 6.7
*Chlorinated and diatomaceous earth filtered secondary effluent.
**Pressure =
Temperature =
Time in Use =
205 psi
700F
1464 hours for 1-6-3 (5A1)
160 hours for 3-24-1 (3B2)
975 hours for 1-22-1 (5C2)
2 The modules tended to "telescope" in
the direction of flow due to pressure
drop through the feed channels.
4 Solids build-up in the feed channels
raised the pressure drop through the
unit. Calcium sulfate and biological
solids were the principal sources of
trouble.
. 3 pH control was erratic and the control
set-up was such that the feed pump
would shut down when pH control was
lost. The pressure surge caused by this
sudden loss of pressure accentuated the
telescoping and ripped apart some of .
the taped joints holding the module
together.
C From late June until late September of
1967, the unit has been operating satis-
factorily. The feed water was carbon-
treated secondary effluent rather than
untreated secondary effluent as before.
7-10
-------
Reverse Osmosis
Table 4. WATER QUALITY DATA FROM GENERAL ATOMIC REVERSE OSMOSIS UNIT
P04 (mg 1) COD (mg 1) NH1-N (mg 1) N01-N (mil: 1) TDS
Tube No. Ave. Range Ave. Range Ave. Ranl!e Ave. Ranl!e Ave. Ranl!e
Feed 30.9 21. 6 - 37.8 10.8 6.7 -14.4 9.2 2.5 - 26.0 2.4 0.8 - 5.0 623 530 - 748
lA 0.22 .09 - .53 1.2 0.0 - 2.3 1.3 0.6 - 2.0 0.4 0.0 - 0.8 51 0 - 133
IB 0.16 .05 - .35 1.1 0.0 - 2.6 1.2 0.5 - 2.2 0.5 0.0 - 0.8 36 9 - 85
2C 1.4 1.3 - 1.5 2.4 0.4 - 6.2 2.8 0.7 - 7.3 0.9 0.4 - 1.7 60
3C 1.1 .07 - 3.4 1.6 0.0 - 3.6 1.8 1. 0 - 2.6 0.7 0.0 - 1.6 70 14 - 173
4C 0.43 .07 - 1.1 1.8 0.6 - 4.0 1.7 0.9 - 2.6 0.9 0.4 - 1.6 67 22 - 120
5C 0.48 .08 - .90 2.0 0.0 - 5.0 1.7 1.1 - 2.6 0.9 0.4 - 1.3 53 35 - 83
6C 0.41 .08 - 1.0 1.6 0.0 - 3.6 1.8 1. 0 - 3. 1 1.0 0.7 - 1.6 53 6 - 113
7C 0.37 .04 - 1.8 2.3 0.0 - 6. 1 2.6 1. 3 - 4.6 1.0 0.5 - 2.4 67 34 - 186
8C 0.77 .07 - 1.8 2.4 0.0 - 5.0 3.2 1. 8 - 5.3 1.6 0.6 - 2.6 95 36 - 196
Total Product 0.57 . 13 - 1.2 1.7 0.0 - 3.4 1.7 0.9 - 3.4 0.8 0.4 - 1.2 73 30 - 113
Brine 177. 73. -240. 43.8 26.6 -60.4 94. 19. -240. 7.5 2.8 -13.7 3402 2861 -3889
0/0 Reduction 98.2 84 82 67 88
In addition, pH control was switched from
sulfuric to hydrochloric acid and 5 ppm
chlorine was added to the feed. The
results have been most satisfactory, as
listed below.
1 We are operating at a feed rate of about
4800 gallons per day and recovering
830/0 of the feed as product. The flux
rate is 5 gfd at a pressure of 400 psi
and there has been no significant decline
in flux in 2000 hours of operation.
2 A special anti-telescoping device
installed by General Atomic has been
successful in overcoming this problem.
3 Product quality has remained high as
shown by Table 4. Table 5 shows
gradual degradation of product quality
between start-up and after 2000 hours
on-stream. These increases are
probably due to slight pin-hole leaks
developing in the membranes.
4 Daily flushing with an air-tap water
mixture has been successful in removing
any particulate material that accumu-
lates or forms in the modules.
D The study program for the future at
Pomona calls for: 1) another attempt at
feeding secondary effluent instead of
carbon-treated water, 2) using sulfuric
rather than hydrochloric acid for pH
control, and 3) increasing the operating
pressure to obtain 90% recovery of feed.
11
-------
Reverse Osmosis
TABLE 5. PRODUCT WA TER CONDUCTIVITY -
GENERA LA TOMIC REVERSE OSMOSIS UNIT
Tube No. Initial Present
Condo (I-Lmhos/ cm) Condo (i-lmhos/ cm)
1A 50 80
1B 50 80
2C 100* 140
3C 70 200
4C 70 110
5C 75 110
6C 85 120
7C 90 220
8C 110 250
Feed Conductivity: Before acidification: 800
After acidification: 950
Brine Conductivity: 4000
*High due to use of several modules with slight leaks
VII
COSTS
A Reverse Osmosis is in an early stage of
development and, as could be expected in
this stage, there is a wide range on
estimated costs. Different manufacturers
have estimated costs for a 10 mgd plant
ranging from $0.25 to over $1. 00/ 1000
gallons. Some of the factors that will
determine the cost of a reverse osmosis
system for wastewater renovation are
listed below:
3 The degree of pre-treatment necessary
prior to entering the reverse osmosis
cell will influence application of the
process. If organic materials have to
be removed before the membranes,
then reverse osmosis is simply a
demineralization process and must
compete economically with electro-
dialysis and ion exchange.
1 Flux through the membranes is the
prime factor affecting costs. Most
cost estimates have been based on an
optimistic flux of 20-30 gfd.
4 The product to waste ratio obtainable
with reverse osmosis is an important
consideration. If 90% of the feed is
recovered as product, the cost of
disposing of the concentrate stream by
injection into deep wells (providing the
geology is appropriate for this) would
be about 2~/ 1000 gal. of water produced.
If only 80% of the feed can be recovered,
the cost of disposal will be 5-6~/1000
gallons.
. 2 Membrane life is also an important
cost consideration. If the membrane
life is short, the reverse osmosis unit
will have to be designed so that mem-
brane replacement is rapid and
inexpensive. This is an approach
being taken by some manufacturers.
5 The production rate of membrane
materials will affect the cost of the
7-12
-------
Reverse Osmosis
operation. The raw material costs are
almost negligible but the manufacturing
costs are high. Membrane consumption
must be high enough to support continuous.
automated production facilities.
B Despite all these considerations. reverse
osmosis could have widespread application
in the future in locations where demineral-
ization of wastewater is required for
reuse or pollution control purposes. It
must be remembered that reverse osmosis
is still an "infant" when compared to
processes such as distillation and carbon
adsorption. The field is wide open and
significant developments are occurring
rapidly.
VIII
FUTURE PLANS
A The FWPCA has plans for continuing
support of development work in the reverse
osmosis field. both in testing of new
membrane materials and in improvements
in engineering design.
B A major undertaking under consideration
is the funding of work to determine the
minimum pretreatment necessary to
prepare a raw sewage for treatment by
reverse osmosis. It seems most unlikely
that conventional primary and secondary
treatment provide the ideal conditioning
of the wastewater for the membrane
process.
C A major study funded jointly by the FWPCA
and the Eastern Municipal Water District
of Hemet, California is just" getting under-
way. Starting with secondary effluent,
several different types of reverse osmosis
units will be run in parallel. Pretreat-
ment of the secondary effluent by any
combination of clarification. filtration
and carbon adsorption will be possible.
IX
INDUSTRIAL APPLlCATIONS OF REVERSE
OSMOSIS
A The Pulp Manufacturers Research League
in Appleton. Wisconsin has investigated
reverse osmosis as a method for solving
both their water pollution and water
supply problems. They hope to operate
a 50.000 gpd plant that could be evaluated
on several waste streams at a number of
pulp mills. Laboratory investigations
have been limited to short-term batch
runs but the results appear promising.
B The Aerojet-General Corporation has
received a grant for a study of the appli-
cation of reverse osmosis to vario!Js
industrial waste streams at the Odessa.
Texas petrochemical complex. A 560
square foot unit will be mounted on a
trailer that will also include a clarifier
and diatomaceous earth filter to serve as
pretreatment units. The objectives of
the study will be to investigate the follow-
ing factors:
1 Reliability of reverse osmosis as an
industrial waste treatment process.
2 Sensitivity of the membrane performance
to input water quality.
3 Membrane life.
4 Design data for large-scale units.
5 Economics of the process
Table 6 presents characteristics of some
of the waste streams that will be fed to
the unit.
REFERENCES
1 "Desalination by Reverse Osmosis. "
Ulrich Merten. editor. The M. 1. T.
Press. Cambridge, Massachusetts,
1966.
2 Wilford. J. and Perkins. F. R. "Test of
G.A. Reverse Osmosis Unit in
New Jersey. 1965." New Jersey State
Department of Health and New Jersey
Department of Conservation and
Economic Development. January. 1966.
3 Loeb. S. and Johnson. J. S. "Fouling.
Problems Encountered in a Reverse
Osmosis Desalimition Pilot Plant. "
7-13
-------
-J ::tJ
I TABLE 6. ODESSA INDUSTRIA L CaMP LEX ro
...... <
~ WASTE WATER STREAMS ro
'1
UJ
ro
o
UJ
S
o
en
,....
UJ
Combined
Stream A B C D E F Waste
- - - - - -
Total Hardness 80-200 10-100 0-30 130-200 150-320 0-625 60-1700
(130) (40) (15) (150) (200) (150)
Ca 21 25 0-8 45 109 20-150
Mg 6 3 0-2 3 8 2-30
Fe 2.5 1 3-5 7 2 1 <1.0-2.5
(4)
NH3 2-15 1-15 40-125 0-50 0-13 4 1-25
(60) (10) (3) (10)
Total Dissolved Solids 2250-2500 700-1150 4400 10, 800 1550-1810 2300 3,500-25,000
(5000)
pH 8.5-9.5 8.5-11.5 9-12 6.5-7.5 6.5-8.0 8.5-12 8.5-10.0
(9.6) (10) (7.3) (9.5)
( COD) 2300 770 2440-3715 140 45-80 1500-7500 650-800
Temperature, F 100-120 96-130 110-135 68-85 70-95 75-110 85-110
(110) (85) (90)
-------
Reverse Osmosis
Preprint 21A - Presented at the
Symposium on Desalination: Part II.
Sixtieth National Meeting, American
Institute of Chemical Engineers.
September 1966.
4 Marcinkowsky, A. E. et. al. Hyperfiltration
Studies - IV. Salt Rejection by
Dynamically-Formed Hydrous Oxide
Membranes, Journal American
Chemical Society, 88, 5744-46, 1966.
5 Kraus, K.A. et. al. "Hyperfiltration
Studies - VI. Salt Rejection by
Dynamically- Formed Polyelectrolyte
Membranes." Desalination, .!., 225-230,
1966.
6 Bray, D. T. et. al. "Reverse Osmosis
for Water Reclamation, " Report No.
GA -6337. General Atomic Division,
General Dynamics Corporation, San
Diego, California.
7 Okey, R. W. and Stavenger, P. L.
"Industrial Waste Treatment with
Ultrafiltration Processes." Dorr-
Oliver, Incorporated, Stamford, Conn.
8 Wiley, A. J. et. al. "Application of
Reverse Osmosis to Processing of
Spent Liquors from the Pulp and Paper
Industry~' Tappi, 50, 9, 455-60.
September 1967. -
This outline was prepared by Arthur N.
Masse, Office of Enforcement, USEPA,
Denver, Colorado
Descriptors: Wastewater Treatment,
Waste Treatment. Water Pollution
Treatment. Reverse Osmosis. Tertiary
Treatment
7-15
-------
ELECTRODIA LYSIS
I
INTRODUCTION
Usually when water is used, there are added
varying amounts of inorganic materials. In
the case of domestic wastes, the increment
amounts to about 300 to 400 mg/l. Most of
the inorganic material is composed of ions
common to natural waters. Also added are
phosphate and nitrogen as ammonia or nitrate.
Removal of the added material is obviously
necessary to maintain the quality of the
water. With industrial wastes, especially
from industries dealing with metals, the
inorganic materials added are heavy metal
ions that may be very toxic. Their removal
is even more important than the removal of
the ions usually found in natural waters.
Electrodialysis is useful for the partial
demineralization of fairly dilute waters. For
practical application usually the total dissolved
solids (TDS) concentration is about 2000 mg/l
or less. Since municipal wastewaters nearly
always have a TDS less than 1000 mg/l they
are well within the range where the process
should be economically feasible. Many
industrial wastes also undoubtedly fall into
the practical concentration range. With
electrodialysis sufficient demineralization is
usually carried out to reduce the TDS to
500 mg/l or slightly less. Further deminer-
alization results in appreciable added cost.
II
PRINCIPLES
A direct electric voltage impressed across a
cell containing mineralized water will cause
positively charged ions or cations to migrate
to the negative electrode and negatively charged
ions or anions to migrate to the positive
electrode. If cation- and anion-permeable
membranes are placed alternately between the
electrodes as shown in Figure 1, alternate
compartments become more concentrated
while the intervening compartments become
diluted. Many membranes can be placed
between the electrodes forming many dilute
and many concentrate compartments.
SE. TT. pp. 9. 7.77
Manifolds can then be added so that the
membrane stack can be fed continuously and
demineralized product and waste concentrate
streams can be removed continuously.
The cation(+) and anion( -) permeable mem-
branes that make electrodialysis possible are
composed of ion exchange materials. Cation
membranes are likely to be sulfonated
styrene with divinylbenzene added for cross
linking and strengthening of the polymer.
Anion membranes are also often styrene
base with various amine groups being used to
give the ion exchange property. The
sulfonate and amine groups make the mem-
branes very hydrophilic. Membranes absorb
appreciable water when in contact with
aqueous solutions.
Although the structure and function of the
membranes is complicated, a rough
quantitative explanation of the reason for
their ability to transport ions of one charge
can be obtained from consideration of the
Donnan principle. This principle, in an
incompletely rigorous form, states that the
ion concentration product of a compound will
be equal for water in the membrane and for
the surrounding solution. A cation membrane
in contact with very dilute sodium chloride
can be considered as an example. The
product of the concentrations of the sodium
and chloride ions will be very small in dilute
solution. In the structure of the membrane
there will be a high concentration of fixed
negative charges which, to maintain
electroneutrality, will attract an equal number
of positive charges or in this case sodium ions.
The concentration of sodium ions associated
with the water in the membrane will then be
very high in relation to the original solution
concentration. By the Donnan principle the
chloride ion concentration in the membrane
must be very small. If a voltage is applied
across the membrane, sodium and chloride
ions will move in opposite directions through
the membrane. Because of the preponderance
of sodium ions, however, most of the current
will be carried by that ,ion. The cation
8-1
-------
Electrodialysis
A
DILUTING
COMP ARTMEN
+
CONCENTRAT-
ING COMPART
MENT
e ANION £ CATION A-ANION PERMEABLE MEMBRANE
C-CATION PERMEABLE MEMBRANE
Figure 1. Electrodialysis Principle
membrane will, therefore, have a high
selectivity for sodium and for cations in
general. A similar argument can be used to
show why anion membranes are selective to
anions. As the Donnan principle indicates,
selectivity decreases as solution concentra-
tion increases.
One of the important factors to be considered
in operating an electrodialysis stack is the
electrical power requirements. In treatment
2
of wastewater, fortunately, this is not as
important as in demineralization of a more
highly mineralized water. Since direct
current is used for electrodialysis, the
stack power is simply the product of the
current and voltage. The demineralizing
current can be calculated from the product
of feed rate per diluting compartment,
concentration change from feed to product
water expressed as normality, and Faraday's
constant, 96,500 cou1ombs/ equivalent. The
-------
Electrodialysis
required voltage is more difficult to deter-
mine. Classical thermodynamics allows the
minimum voltage.to be calculated (1) for a
membrane separating solutions of different
concentrations. In the usual electrodialysis
application this minimum voltage, which only
applies strictly at a zero current, is a small
fraction of the total. The usual approach is
to consider the stack as a series of resistances
for which the voltage can be calculated from
Ohm I slaw. Resistances that should be
included are diluting and concentrating liquid
streams, membranes, and the thin stagnant
liquid films adjacent to the membranes.
Usually, however, the resistances other than
those of liquid streams are simply combined
as a single value which must be determined
experimentally. This empirical resistance
can then be used to estimate the approximate
voltage for an electrodialysis stack using the
same kind of membranes. To calculate the
total power requirements the voltage require-
ments of the electrodes and electrode com-
partment streams must be added to the
voltage for the remainder of the stack.
III
OPERA TING PROBLEMS
The stagnant liquid films that occur at the
membranes are responsible for significant
problems during operation. Serious con-
centration gradients can develop across these
films because of the higher electrical trans-
ference numbers in the membranes compared
to transference numbers in solution. As a
result there is a decrease in concentration
from bulk dilute stream to the membrane
surface and an increase in concentration from
bulk concentrate stream to the membrane
surface. The high concentration at the face
of the membranes on the concentrating side
leads to scaling from precipitation of com-
pounds with low solubility. The low mineral
ion concentration at the dilute side of the
membranes causes transfer of hydrogen and
hydroxyl ions from the water through the
membranes. Polarization is the term used
to describe operation when water decom-
position becomes significant. The
decomposition of water into ions wastes
electric power. A more important effect,
however, is the increased scaling potential
at the concentrate side of the anion mem-
branes. Hydroxyl ion increases the
precipitation of magnesium hydroxide and
calcium carbonate.
An effect that is common to electrodialysis
is membrane fouling. This phenomenon is
different from scaling since it involves
deposition of materials on the dilute side of
the membranes. Although there can be a
number of causes in wastewater treatment
fouling appears to be due to electrophoretic
movement of negatively charged colloidal
particles to the anion membranes. Because
of the size of the particles they cannot move
through the membrane. Instead they form a
layer at the membrane surface that interferes
with demineralization.
IV
APPLICATION TO MUNICIPAL
WASTEWA TERS
A Laboratory Investigation
Bench scale study (2) of electrodialysis
for treatment of municipal secondary
effluent indicated that good removal of
soluble and insoluble organic materials
was necessary for satisfactory operation.
Cellulose cartridge filters were used for
removal of the suspended organic solids.
These filters, however, would not be
practical on a large scale and would have
to be replaced by another form of sus-
pended solids removal. Granular activated
carbon was used for soluble organic
removal. Using adequate pretreatment
the cost of electrodialysis, exclusive of
pretreatment and waste concentrate
disposal, was estimated at less than
10 cents/lOOO gal.
B Pilot Plant Investigation
As a result of encouraging results from
the laboratory study a pilot scale investi-
gation was undertaken (3). Equipment was
installed at the Lebanon, Ohio Sewage
Treatment Plant. Secondary effluent was
the feed to the system. Based upon the
experience gained from the laboratory
work, clarification and soluble organic
8-3
-------
Electrodialysis
removal were considered necessary
pretreatment for the electrodialysis stack
feed. Diatomaceous earth filtration, using
for the most part an alum treated water
grade filter aid, was chosen as the method
of clarification and granular carbon in
fixed beds was chosen for adsorption of
soluble organics. The pilot system is
shown in Figure 2. The electrodialysis
stack is an Ionics Mark III capable of
holding up to 150 cell pairs. It is designed
to remove about 40 percent of the TDS from
the feed. Membranes for this stack are
18 X 40 in. and have about 3000 m2
active area. Spacer material and edge
gasket are combined in one piece. Spacer
thickness is 0.032 in. Using all 150 pairs
the nominal capacity of the stack is 50 gpm
product water. A s a result of difficulties
with the diatomaceous earth filter, it was
necessary to reduce the feed water rate.
The number of cell pairs was reduced to
125 and the product rate to 42 gpm. The
concentrate waste rate at reduced feed
rate has been 4 gpm. Internal recirculation
of concentrate at a rate about equal to the
product rate is necessary, however, to
maintain approximately equal pressures
on both sides of the membranes. The pH
of the concentrate stream is held at 5 or
less to prevent calcium carbonate scale
formation.
Operation of the pilot plant has resulted
in only one serious problem, fouling of
anion membranes. Membrane fouling
increases stack resistance and necessitates
higher voltage to maintain the desired
degree of demineralization. Figure 3
shows how demineralization at constant
voltage was affected by fouling. Fouling
rate depends to some extent upon feed
turbidity as might be expected. The
unexpectedly low fouling rate for the run
made after the stack was acid rinsed at a
low enough pH to kill organisms suggests
biological growth on the membranes as a
contributing factor. Fortunately fouling
has not been permanent. Shutdown over a
weekend has always restored the fraction
demineralization to near normal. Even
with some fouling, operation is considered
practical since excess stack capacity can
8-4
be provided to take care of down time for
self-cleaning of membranes. Control of
feed turbidity is necessary for practical
run lengths before shutdown is required.
Scaling of anion membranes, a problem
that is serious in many electrodialysis
installations, has not been particularly
troublesome in treatment of municipal
wastewater. Calcium carbonate scale
formation has occurred on occasions,
but has not done serious damage to mem-
branes. Failure to control the pH to 5 or
less can cause serious scaling that ruins
the anion membranes. In the pilot work
concentrate blowdown has been held at
9. 5 percent of the product. . Scaling
potential would be reduced if the amount
of blowdown were increased. Such an
increase would, however, increase the
problem of brine disposal. From a
pollution control standpoint, therefore,
the concentrated waste must be kept as
low in volume as possible.
Permanent damage to membranes during
pilot runs over a period of 1500 hours has
not been great. Membrane pair resistance
at the beginning of runs, for example,
only increased about 5 percent even though
fouling conditions during some runs were
severe. Current efficiency, that is the
fraction of the current that results in
demineralization, has not changed
significantly. Serious membrane damage
would result in decreased current efficiency.
To obtain better control of turbidity and
more dependable operation than was
possible with the diatomaceous earth
filter, a chemical clarification system
has recently been installed before the
carbon columns. This system consists
of an up flow clarifier and dual media
filters. Lime has been chosen for use
in the clarifier because of the softening,
phosphate removal, and alkalinity removal
tha t this material offers. This change
in pretreatment has not eliminated mem-
brane fouling. The fouling rate is roughly
the same as was obtainable with the
diatomaceous earth filter when it was
producing water of less than 0.1 JTU water.
-------
:J' ~J1 i'~1
I I I I I I
I I I
I I
I I I I I I
I I I I I I
I I I I I I
I I I I I I
I I I 1 I I
I I I I I I
~1 i~: !~I i
I I I I I I
L ...J L J L ...J
--
FILTER
96 S. F. AREA
60 TO 70 G.PM.
-
2800 GALLON
HOLDING TANK
DIATOMACEOUS EARTH FILTRATION UNIT
--..
LOWERS pH TO 4 IN CONC. STREAM
..--
ax 30
MESH
CARBON
t
8 X 30
MESH
CARBON
8X30
MESH
CARBON
ELECTRODIALYSIS
UNIT
300 GAlLO N
HOLDING TANK
t
40% DISSOLVED SAlTS
REMOVAL
40-50 G.P. M.
FINAL PRODUCT WATER !
WASTE CONCENTRATE
<10% OF FEED
CARBON ADSORPTION COLUMN
I:Ij
~
ro
n
....
'1
o
Q.
,.....
PJ
«'
en
,.....
en
FIGURE 2. ELECTRODIALYSIS PILOT PLANT
WITH DIATOMACEOUS EARTH FILTER PRE-TREATMENT
LEBANON, OHIO
CJ1
-------
Electrodialysis
.
0.45
.
. . . .
-lFEED TURBIDITY=O.03 J:~
.
.
.
0.40
z
o
~
N
:::; 0.35
«
e:::
w
z
~
w
o
z
o
I- 0.30'
u
«
e:::
u..
020
o
.
FEED TURBIDITY= 1.3 JTU
STACK ACID RINSED
.
0.25
FEED TURBIDITY=O.2 JTU
5
25
30
fIGURE 3. EffECT Of TURBIDITY ON
DEMINERALIZA TION .
6
-------
Electrodialysis
Further work with chemical clarification
will be carried out.
C Other Investigations
Ionics Incorporated has operated a small
pilot electrodialysis stack at Orange
County on trickling filter effluent that had
been alum clarified, chlorinated ,and treated
with granular activated carbon. In addition
the stack had before it 10 and 3 micron
cartridge filters and an ultraviolet light
for sterilization. The fouling rate
apparently was considerably lower than
observed during operation of the Lebanon
stack. Operation was halted after more
than 800 hours of run time when a serious
stack resistance increases occurred.
More recently this same electrodialysis
system, including cartridge filters and
ultraviolet light, has been operated on
carbon treated, chlorinated secondary
effluent at Pomona California. Here
carbon treatment is carried out without
any clarification. Frequent replacement
of the cartridge filters is necessary.
Preliminary results suggest that fouling
will be at a lower rate than observed on the
Lebanon stack. Further work is required
to verify the early results and to determine
the effect of the cartridge filters and
ultraviolet light.
v
APPLICATION TO BRACKISH WATER
The principal use for electrodialysis has
been in treatment of brackish waters. These
are inland waters containing enough salts to
prevent or limit direct use. The mineral
content is considerably less than for sea
water, often about 2000 mg/l, and their
composition~ can vary widely from one
location to another. Electrodialysis has
been found practical for demineralizing these
waters to 500 mg/l TDS or slightly less.
There are problems associated with brackish
water treatment, some of which are different
from those observed in wastewater treatment.
Fouling is sometimes troublesome, and is
often not of organic origin. Instead inorganic
materials such as traces of iron and man-
ganese can cause the difficulty. It is usually
considered necessary to remove essentially
all the iron and manganese from ground
water to be treated by electrodialysis.
Scaling is often a problem because of high
concentration of materials of low solubility
such as calcium sulfate and calcium carbonate.
To avoid scale formation the amount of con-
centrate waste is kept considerably higher
than is possible in wastewater treatment.
Often the volume of concentrated waste is
from 33 to 100 percent of the volume of
treated water.
There is an increasing number of electro-
dialysis installations using brackish water
in the United States. Two of these are
particularly well known. They are at
Webster, South Dakota and Buckeye, Arizona.
The Webster plant began operation in 1962
as a Demonstration Plant by the Office of
Saline water and has a nominal capacity of
250,000 gpd. Electrodialysis equipment was
supplied by the Asahi Chemical Industry
Company of Japan. The feed water is not
unusually high in TDS for brackish water
since it contains only about 1700 mg/I. It
is, however, a very difficult water to treat.
It is high in calcium, sulfate, and bicarbonate
so that scale formation is a serious problem.
The water also contains significant amounts
of iron and manganese that must be removed
by using potassium permanganate injection
and filtration through manganese zeolites
before electrodialysis treatment. One of the
wells that was to provide part of the feed
water contains organic materials that cause
a rapid increase in stack resistance. Many
problems have arisen, therefore, during
operation of this plant. These have been
described in detail in a number of Office
of Saline Water publications and summarized
in a 1965 paper (4). As a result of the prob-
lems the Webster site has been used to test
a number of pretreatment methods, both for
iron and manganese removal and softening.
An interesting feature of the plant as now
operated is the use of periodic polarity
reversal to reduce scale formation. Improved
operation is obtained at a slight loss in
current efficiency. Successful use of polarity
:reversal could represent a significant break-
8-7
-------
Electrodialysis
through in the application of electrodialysis
to high-hardness, badly-scaling waters.
Results at Buckeye, Arizona have been more
optimistic than results at Webster. The
water, although higher in TDS, has a relatively
large amount of sodium and chloride ions.
Scaling is, therefore, much less likely to be
a problem. The water contains no manganese
and little iron. The equipment installed at
Buckeye was supplied by Ionics, Inc. and
has a design capacity of 650, 000 gpd. Over
the period October 1962, through June 1965,
it is reported (5) that there were no major
malfunctions. During that time the average
load factor was only 32 percent. This
unexpectedly low figure resulted partly from
the recirculation of water through evaporative
coolers. Recirculation was not possible with
the more highly mineralized brackish water
available before installation of the electro-
dialysis equipment and, therefore, the daily
water consumption was higher. The low
load factor has resulted in a higher than
expected water cost when amortization of
equipment is included, but the installation is,
nevertheless, considered a success.
VI
INDUSTRIA L APPLICA TIONS
Use of electrodialysis for treatment of
industrial wastes has been very limited. Two
possible applications that have a pollution
control aspect are treatment of pickle liquor
and treatment of spent sulfite liquor. The
first application is actually more nearly
electrolysis than electrodialysis and results
in recovery of iron and sulfuric acid from
the waste (6). There is more interest in this
process in Europe than in the United States.
The second process was developed by the
Sulfite Pulp Manufacturers Research League
(7) and results in recovery of cations from
the spent liquor. The principle of this
modified electrodialysis process is shown
in 'Figure 4. Spent liquor enters compartments
3 and 7 where cations and organics such as
small ligno- sulf onate ions are removed. New
liquor is made up in compartments 4 and 8.
Compartments 5 and 9 serve as sources for
sulfite ion. The spent liquor with low mole-
cular weight organics removed may have a
number of uses. The process operates at
8~8
high current efficiencies and high current
densities. Because of the arrangement of
membranes, the anion membranes are
protected from fouling by organics in the
spent liquor.
Although electrodialysis has not yet been
used extensively for treatment of industrial
wastes, it may have wider application in the
future. Its practicability is improved if
useful materials can be recovered from the
treatment. With the present cost picture
for electrodialysis, recovery is probably a
necessity. It is accomplished in the two
examples cited above. There are undoubtedly
many industrial waste streams where chemi-
cal recovery is a possibility.
In its basic form of alternating dilute and
concentrate streams the process would
ordinarily only be applicable to streams free
of fouling soluble organics and suspended
materials. By using special membrane
arrangements such as in the sulfite liquor
treatment, fouling problems can be reduced.
It should be recognized that in applications
where useful materials are recovered the
added maintenance resulting from membrane
fouling may not be prohibitive in cost.
The maximum concentration that can be
economically treated would be limited to the
brackish water range if by-product recovery
were not possible. Electric power consump-
tion for removing a large concentration of
ions, ordinarily becomes prohibitive. Where
useful materials are recovered the maximum
concentration could be increased until the
cost for electrodialysis exceeds the cost of
other processes such as distillation. A
detailed study of each particular application
would have to be made to determine whether
electrodialysis is practical.
There are two limitations to electrodialysis
that may prevent its use in some applications.
One limitation is the inability to practically
remove ions below a concentration of several
hundred mg/l. Where waste are very dilute,
ion exchange should be more appropriate.
The other limitation is the inability for the
process to be very selective for anyone
material in a mixture. At present a valuable
metal ion could not in general be removed
-------
02
11
C
N
c
A
C
N
C
A
H2
J
2
3
.4
5
6
7
8
H+ NH+ ~ H+ NH~ ~ NA+7
.4 .4
+
~S04 L LSOJ L 50- L LSO; L SO-
3 3
~- - "'SUGAR
f-.
.. SUGAR
llRGE
LULSO -
3
~ARGE
LS03
C-CATION SELECTIVE MEMBRANE
A-ANION SELECTIVE MEMBRANE
N-NON-SELECTIVE MEMBRANE
Figure 4. Electrodialysis Process for Treating Spent Sulfite Liquor
ttj
......
(1)
o
....
'1
o
a.
.....
!1>
-<
en
.....
en
CD
-------
Electrodialysis
with high selectivity from common ions such
as sodium and calcium. Relatively little
work has been done to develop selective
systems, however, and improvements are
likely if a strong need arises. Again ion
exchange may be a more practical form of
treatment.
VII
ULTIMATE DISPOSAL OF CONCENTRATED
WASTE
Experience in the use of electrodialysis on
municipal wastewater indicates that the volume
of the concentrated waste will be 5 to 10
percent of the product volume. To prevent
pollution of surface waters by this highly
mineralized waste it must either be trans-
ported to the sea or in some other way be
treated for disposal. Disposal, no matter
what the method, will not be insignificant in
cost except near the sea. For brackish
water the potential disposal problem is also
great. Presently the treatment of brackish
water is limited enough that disposal of brine
has not been a serious problem. Industrial
waste treatment would not prevent a
significant disposal problem if recovery of
by-products from the concentrated waste
were practiced. In that case the waste
would be further treated anyway. If recovery
were not possible, disposal or treatment of
the concentrate to make it innocuous would be
necessary. The volume of concentrate would
vary widely depending upon the particular
waste. If it is low in membrane scaling
materials ,the volume may be only a few
percent of the product. Each case must be
considered separately.
VIII
EQUIPMENT DESIGN
Since electrodialysis is a recently developed
process, design of equipment is not yet highly
s~phisticated. The major American manu-
facturer, for example, has available a limited
number of stack configurations to be used
primarily for brackish water. Minor
modifications are made in the equipment to
adapt it to other uses. It would be a fortun,ate
coincidence if obtainable equipment were the
optimum for any given application. Because
8-10
of the limited market, manufacturers have
not been able to justify the extensive research
and development work that is required for
optimized design.
Several discussions of equipment design have
been published (8,9). Application of the
methods were also demonstrated for the case
of municipal wastewater (2). The reader is
referred to these publications for details.
In principle, these methods depend upon Ohm's
Law and Faraday's Law and material balances.
The resulting equations allow membrane area
and power requirements to be calculated.
Before making design calculations, it is
necessary to know membrane resistances
and limiting current density. These are
empirical quantities,. but manufacturer.s
should have the necessary data available for
their particular stack configurations and for
brackish water feeds. Experimental data
might be required for design of equipment to
treat an industrial waste.
One serious fault with published design
methods is that they do not emphasize the
design of membrane spacers. After the
designer chooses an available spacer and
has the necessary empirical data it is
relatively easy to size the equipment and
determine power requirements. Optimum
design of a spacer is a much more difficult
problem. Presently, spacers range in form
from rather open types that give .low velocity
and pressure drop to tortuous path types that
give high velocity and pressure drop. To'
reduce production costs, manufacturers tend
to pick one type and use it for all applications.
The Office of Saline Water is presently
sponsoring work on spacer design. Prelim-
inary results suggest that even the brackish
water spacers may be improved significantly.
Further work is obviously justified.
At the present time, the user of electro-
dialysis equipment usually depends heavily
upon the limited number of manufacturers for
the ultimate design of a system. The number
of types of stack components available is
small. Use of the equipment is not yet so
widespread that consulting engineers are
experienced enough to undertake design them-
selves. If the market increases substantially,
-------
Electrodialysis
the situation will change. Not only will a
greater variety of components become
available, resulting in more flexibility of
design, but the designing itself will be done
by a greater number of people.
IX
ELECTRODIA LYSIS COSTS
A Capital Costs
When discussing the capital cost of
electrodialysis, distinction must be made
between present and projected costs. The
market for this equipment has not expanded
to the point where mass production of
components is possible. The costs of
membranes and spacers especially are
very high. Capital costs do not represent,
therefore, the minimum that would be
possible just from an increased market
and the competition that should develop.
It is difficult to make capital cost estimates
without consulting a manufacturer. There
is not sufficient information available in
published form to make the precise
estimates possible. Ionics, Inc. has at
various times made available in company
literature plant costs for their equipment
as a function of capacity and feed concen-
tration assuming the feed to be brackish
water and product to contain 500 mg/l TDS.
Table 1 shows some typical costs. These
figures do not include pretreatment beyond
a protective cartridge filter. Cost is
strongly affected by both feed concentration
and capacity. An independent cost
estimate has been made for treatment of
secondary effluent based partly upon
experience from pilot operation. For a
10-mgd plant and using Ionics equipment,
the estimated installed cost is $0. 34/ gpd.
This is significantly higher than the Ionics
estimate for 900 mg/l brackish water to
which it should most closely compare.
Much of the difference can be accounted
for in the conservative feed rate per cell
pair that was chosen for the wastewater
estimate. A small amount was contributed
by the inclusion of pH control equipment
on the concentrate streams. This equip-
ment is not usually supplied for brackish
water treatment, but is desirable on
wastewater treatment where there can be
frequent changes in alkalinity.
The Office of Saline Water has, in the
past, supported work on a procedure for
estimating the cost of electrodialysis
applied to brackish water. More recently,
work was done on optimization of the
process. This work will appear in
publications from that organization.
B Operating Cost
Table 1
Long term operating costs for electro-
Investment Cost for Electrodialysis Plants
Plant Capacity
(mgd)
Feed Concentration
(mg/l)
900
2
2
3,000
10
900
10
3,000
100
900
100
3,000
Installed Plant Cost
($/gpd)
0.33
0.69
0.19
0.39
0.11
0..25
8-11
-------
Electrodialysis
dialysis are available from only a few
brackish water ~nstallations. Estimates
are usually based upon experience combined
with assumptions about long term main-
tenance requirements. Membrane life is
probably the most questionable factor.
Usually a 20 percent replacement per year
is assumed. Pilot experience on waste-
water suggests that membrane life is
greatly affected by the care that is taken
in handling the membranes and in operating
the equipment. Although definite proof is
not yet available, there is reason to believe
that an annual 20 percent replacement may
not be necessary. Ionics has estimated the
cost of electrodialysis operation over a
wide range of capacities and feed water
concentrations. Some examples are given
in Table 2. The estimates are for brackish
water and assume 90 percent load factor
and 20 percent annual membrane replace-
ment. They include amortization of
equipment. An independent estimate for
the treatment of 10 mgd of wastewater was
16~/l, 000 gal. This is somewhat higher
than the 12~/ 1,000 gal for the nearly
comparable 900 mg/l brackish water. The
difference can be accounted for in the
conservatively high capital cost estimate
for the wastewater plant.
The effect of load factor on operating cost
is significant. A t the Buckeye, Arizona,
brackish water plant, for example, it is
estimated that at full load, the total cost
Table 2
of producing water should be 32~/ 1,000
gal. The plant was actually designed to
operate at a 48 percent load factor and
produce water for 52~/ 1,000 gal. For a
number of reasons the load factor declined
below the design value. For the first 32
months of operation it was only 32 percent.
The result was a water cost of about
70~/ 1,000 gal. This is an extreme
example, but it points out the need to
consider ways for minimizing fluctuations
in flow. For industrial wastes, it may be
possible to make extensive use of relatively
cheap water storage to allow operation at
a high load factor.
REFERENCES
1 Wilson, J. R., "Demineralization by
Electrodialysis ", Butterworths
Scientific Publications, London, 1960.
2 Smith, J. D., Eisenmann, J. L.,
"Electrodialysis in Advanced Waste
Treatment, " Water Pollution Control
Research Series Publication No.
WP-20-A WTR-18.
3 Brunner, C. A., "Pilot Plant Experiences
in Demineralization of Secondary Effluent
Using Electrodialysis ", presented at
the 39th Annual Conference of the Water
Pollution Control Federation, Kansas
City, Missouri, Sept. 1966.
Operating Cost for Electrodialysis
Plant Capacity
(mgd )
Feed Concentration
(mg/l)
900
Total Operating Cost
($/1.000 gal)
16
2
2
3,000
900
10
10
3,000
100
100
3,000
21
8-12
39
12
29
900
8
-------
4 Calvit, B. W., Sloan, J. J., "Operation
Experience of the Webster, South Dakota,
Electrodialysis Plant", presented at
the First International Symposium on
Water Desalination, Washington, D. C. ,
Oct. 1965.
5 Gillilard, E. R., "The Current Economics
of Electrodialysis ", presented at the
First International Symposium on Water
Desalination, Washington, D. C.,
Oct. 1965.
6 Farrell, J.B.. Smith, R.N., Ind. Eng.
Chern.. 54, No.6, 29-35 (1962).
Electrodialysis
7 Dubey, G.A., et al., TAPPI, 48, 95 (1965).
8 Mason, E.A., Kirkham, T.A., Chern.
Eng. Prog. Symposium Series, 55,
No. 24, 173 (1959). -
9 Mintz, M. S., Ind. Eng. Chern., 55, No.6,
19 (1963).
This outline was prepared by Carl A.
Brunner, Research Chemical Engineer,
Municipal Environmental Research
Laboratory, USEPA. Cincinnati, Ohio
45268
Descriptors: Wastewater Treatment,
Waste Treatment, Water Pollution
Treatment, Electrodialysis, Tertiary
Treatment
8-13
-------
STATUS OF CHEMICAL OXIDATION IN WASTEWATER TREATMENT
I Oxidation is a spontaneous and continuous
process in nature. It occurs as slowly as in
the natural purification (oxidation) process in
a river, or as rapidly as in spontaneous com-
bustion. Just as activated sludge waste treat-
ment is a man-made accelera\ion of a natural
biological purification process, chemical
oxidation in waste treatment is a man-made
acceleration of oxidation in nature. Physical
and biological waste treatment processes have
an upper limit for the removal of undesirable
constituents from wastewaters. Beyond that
point, no further reductions are realized unless
process variables are strained to the point
where the treatment method is no longer
practical nor feasible.
n Any chemical change in which an element
becomes more electropositive is oxidation.
A species loses electrons upon being oxidized.
Conversely, the oxidizing agent (oxidant) gains
electrons and is reduced. The oxidation of
organic compounds may be considered as the
process of covalent bond fission in which one
or both of the electrons of the disrupted bond
are transferred or partially transferred to
the oxidant.
A Homolytic oxidations are those reactions
in which only one of the electrons of the
disrupted covalent bond is transferred to
the oxidizing agent.
B Heterolytic oxidations are those reactions
in which both of the electrons of the disrupted
covalent bond are transferred to the oxidiz-
ing agent.
C Halo-oxidants
1 Halogen compounds
2
Chlorine free-radical
D Oxy acids
1 Permanganate
2
Ferrate
TABLE 1: POTENTIAL OXIDATION SYSTEMS
System
Oxidant
Source
Environmental oxygen
°2
02 + catalyst
A ctive oxygen
OH'~
°3
°
Halo -oxidants
C12
HOCl(OCl-)
Cl*
Oxy acids
Mn04
FeO=
4
liquid or gaseous oxygen
02 + Mn02
02 + ultra-sound
Fentons reagent
ozonator, corona
ozone photolysis
C12
hypochlorite
chlorine photolysis
KMn04
Na2Fe04
SE. TT. cpo 4.7.77
9-1
-------
Status of Chemical Oxidation in Wastewater Treatment
III Effective oxidation presupposes that the
oxidant has been selected with respect to the
type of contaminant to be controlled and
applied under conditions favoring its effective-
ness. Condition variables include:
A The pH of the oxidation system may
greatly influence behavior.
1
Chlorine application in a surface water
reacts to form HCl + HOC!. Reaction
of HOCl with oxidizable materials forms
more HC!. If alkalinity is insufficient
to neutralize the accumulated acid the
pH may be reduced to the point where
ammonia or cyanide cannot be destroyed
by chlorination.
2 Permanganate oxidation of organic
amines to form ammonia and CO is
effective over a wide range in pIt
Dichromate appears more effective in
highly acid media.
B Temperature has a large effect upon
reaction rate and may alter the nature of
the oxidizing agent or its availability in
the aqueous system.
1
Chlorine, ozone, H202' react very
slowly near freezing temperatures.
Near the boiling point, the oxidant may
be driven from the system or decomposed
in undesirable side reactions.
2 Oxy acids are much more effective and
show higher oxidative efficiencies at
elevated temperatures.
C Oxidative energy may be favored by
appropriate catalysts.
1 Permanganate oxidation at ambient
temperatures is favored by the presence
of ferrous or ferric iron.
2
The presence of ionic silver favors
oxidation by dichromate.
D Adequate mixing of oxidant and aqueous,
media affect both oxidation efficiency and
nature of the process.
9-2'
1
Chlorine application necessitates
vigorous mixing to dilute the chlorine
gas to a point where it will not be lost
in gaseous form and to prevent unde-
sirable side reactions due to a tremend-
0us excess of chlorine at local sites.
2 H202' ozone, and hydroxyl free-radicals
are notoriously unstable. They generally
are applied in relatively low concentrations
or formed in situ, to favor contact with
undesired contaminants prior to uncon-
trolled side reactions.
E Time is essential to permit desired
oxidations to be completed.. Time require-
ments vary with the particular oxidant
and the nature of the substances to be
treated.
1
Chlorination of a relatively stable
water may favor formation of free
available residual chlorine favoring
high germicidal effectiveness in a
short interval of time. Grossly con-
taminated water chlorination favors
production of combined available
chlorine which is likely to require a
much larger application dosage and
more time for disinfection.
2 Permanganate oxidation is character-
ized by a latent period followed by an
auto catalytic effect once oxidation
has started. Catalysis reduces this
effect but does not eliminate it.
F The concentration of contaminants controls
the type of oxidant feasible for economic
use and the sequence in which a combin-
ation of oxidants may be applied.
1 The concentration of combustible
organic materials determines whether
high temperature incineration is feasible
or not. Heat recovery during oxidation
vs heat requirements for drying are
major criteria.
2 Economics of the over-all system
largely determine which oxidants will
be employed for each situation.
-------
Status of Chemical Oxidation in Wastewater Treatment
a Phenol may be oxidized by chlorine,
ozone, and chlorine dioxide. Large
scale continuous operations commonly
employ biochemical stabilization
because capital and operational costs
of aeration are likely to be lower than
for chemical oxidation.
b Conversely, oxidation may be more
economical for discontinuous or low
concentrations of phenol.
This outline was prepared by Francis L.
Evans, III, Sanitary Engineer, Municipal
Environmental Research Laboratory
USEPA, Cincinnati. Ohio 45268
Descriptors: Wastewater Treatment.
Waste Treatment. Water Pollution
Treatment
9-3
-------
REMOVAL OF PHOSPHORUS AND COLLOIDAL SOUDS
BY COAGULATION IN CONVENTIONAL TREATMENT
I
INTRODUCTION
A
Why do we consider chemical coagulation?
1 Obviously any operation can be improved
with better attention, understanding and
input effort. This is particularly
applicable in many wastewater treatment
plants.
2 These plants are faced with effluent
upgrading requirements, tired equipment,
increased and more difficult loading.
This calls for more effective use of
what we have on a da.y- by- day basis.
(1 2 3 4, 5)
~ Recent reports' ,. show that
progress has occurred in applying addi-
tional known techniques in existing
facilities to improve cost-benefit ratios.
4 Chemical coagulation has almost
universal application in water supply.
Why not in wastewater treatment?
5
Coagulation can provide more days of
better effluents to gain time and needed
facilities as a trade off for chemical and
control cost.
6
Dawn comes up slowly at the treatment
works. This is n::>t only because the
plant usually is located at the lowest
elevation. Mental fog sometimes is
more difficult to dispel. An aggressive
approach with positive pragmatism is
required.
7 Designers must consider modification
of existing facilities and provision for
coagulation in new facilities. Designers
should be part of start-up and initial
operations until the personnel and facility
function smoothly. The designer and
operational control personnel must be a
working team for mutual benefit.
B
Should you use these processes? When
and Where?
SE. TT. 17.7.77
1 This is not intended as part of the great
phosphorus debate. Look at your local
Standards. BOD and solids are
inseparable. Phosphorus cannot be
removed without efficient solids removal.
2 Adapting existing plants for coagulation
usually is simple. Inclusion in n.ew
plants is more simple. Capital costs
for equipment are relatively low and
usually can be traded off in excess
capacity for peak loading. Operating
costs are higher for chemicals and
controls.
3 Upgraded effluent quality necessarily
costs more. There are no free-rides.
Going from 80 to 95% phosphorus (P)
removal may increase cost by 50% or
more. The increased cost provides more
complete control of general and specific
cleanup with improved contingency con-
trol as an added benefit. Major upgrading
costs are for applied chemicals once
equipment and control are available.
4 The owner must make the choice of
routes to meet requirements. Reduction
in solids and BOD may be pivotal. Success
demands the owners commitment to:
a Provide capable operation 24 hours/
day, 365 days/year
b Provide capable laboratory support
for operational control not just "for
the record"
c Provide a budget of about 5r;/ 1,000
gallon of treated flow.
5 The degree of treatment dilemma forces an
irrevocable design decision in many treat-
ment processes. Coagulation is a very
flexible means of providing a high degree
of upgrading capabilities. The identical
facilities can be used over a wide range
of degrees of treatment desired. Use
what you need today and adjust for
tomorrow's needs.
10-1
-------
Removal of Phosphorus and Colloidal Solids
II
COAGULANT APP IlCA TION SITES
A
Many possibilities exist. Choice depends
upon facility, equipment, layout, and
requirements to be met. Ingenuity is
required to pick the best location for a
given situation on a cost-effective basis.
B
Coagulant addition during primary clarifi-
cation provides the highest yield of solids
separation per unit disage at a reasonable
percentage removal.
1 Increased oxygen demand removal
from about 30 to 65% may be a clear
choice in an overloaded facility.
2 Primary sludge is easier to handle
and much more concentrated than
equivalent solids as a secondary sludge.
3 Remaining turbidity in the primary
effluent is more readily cleared in
subsequent units.
4 The raw influent has the lowest fraction
of ortho/total P of any liquid in process.
Phosphorus or solids removal may
not be as complete as at later stages
of operation without high coagulant
dosage.
C
Coagulant Addition in a Biological Unit
is Attractive
1
Coagulation during trickling filtration
may cause blotchy appearance and
increased sloughing. Plugging does
not appear to be a problem. Mixing
characteristics and other factors
decrease the cost/benefit ratio of
coagulation in the trickling filter.
2
The mixing, detention, and floc
characteristics of activated sludge
aerators favor coagulation benefits.
a Coagulants may be added at the
inlet, middle or outlet zone of
aeration or en route to the final
clarifier. The last is most popular.
10-2
b Tenny and Stumm(6)proposed
coagulation during contact
stabilization.
c Effed of metalsdf return sludge and
biota is unclear, but biological re-
activity is probably unimpaired. The
floc characteristics may permit de-
creased chemical usage but may also
impede contact of substrate and critters,
oxygen transfer, and off gas release.
The control of pH in the cell environment
during coagulation is imperative to avoid
biological problems, especially when
nitrification is intended. This generally
means in place measurement in a
dynamic situation where C02 equilibria
is involved.
d Coagulation of activated sludge tends to
increase sludge concentration, provide
greater solids capture, speed up sedi-
mentation and decrease phosphorus
recycle. Coagulation is the most direct
and feasible correction for "bulking. "
D Coagulation in the final clarifier is a
last chance situation, but, quite effective.
1 Prio: stabilization decreases
surfactants and general resistance
to liquid- solids separation.
2 Phosphorus generally exists mainly
in the ortho form the only form that
can be precipitated.
3 If the underflow is returned to the
primary clarifier it enhances liquid-
solids separation in the primary.
4 Producing a clear well-stabilized
overflow is obvious proof of good
operation. Correction of earlier poor
guesses is good public relations too.
E Multi-point application of coagulants
makes it possible to remove most of the
loading at the most favorable stage and to
polish the discharge at its most favorable
stage with lower chemical requirements
and improved control.
-------
F
Emphasis on upgrading existing facilities
precludes consideration of filtration and
other advanced treatment in this outline.
III
TRIAL EFFORTS--HOW MUCH?
LONG?
HOW
A
The vagaries of wastewaters, situations
and conditions make it hazardous to
generalize the response of any particular
"improvement." Joe may swear by it--
Tom, at it. Try it--preferably on small
scale under conditions anticipated for
use. Become familiar with likely events
in response to your particular situation
and its changes. If the results include a
horrible mess--hopefully--it's a small
one.
B
Jar tests are an ever present help in
time of trouble--a vital but treacherous
ally. Don't just "follow the book" - - THINK.
1
The "stare and compare" technique
between jar test and full scale operations
is invaluable. Don't expect jar test
information to be transferable to full-
scale if the situations are incomparable.
2
Coagulation occurs within seconds of
contact. If it isn't mixed, it hasn't
contacted, and the coagulant addition
benefit is reduced. An auxiliary flash
mix is highly desirable to promote
desired reactions including substrate
rather than interactions of coagulant
and water.
3
Flocculation or gathering together of
small particulates takes place under
lower energy input situations. Time
and energy are important--you can
favor agglomeration or break it up.
4
Settling is a dynamic situation. Some
movement in the agglomerating mixture
is beneficial to pick up stragglers, too
much produces stragglers. For jar tests
a paddle speed of 5 to 10 rpm during the
"settling" period is advisable to improve
clarification and comparison with plant
performance. Sludge compaction tests
Removal of Phosphorus and Colloidal Solids
require stopping the paddle.
5
Practice, practice, practice.
C Other laboratory tests are described. (8,9)
Many of these techniques are new;
experience has shown them valuable in
water plants but less effective in wastewater.
Unfortunately information in the book isn't
much good until it is applied.
D Full plant (or an isolated module) trials
generate confidence in results. If
carefully planned and executed in line with
pre-tests and other information. the trial
is not likely to bankrupt the organization
or create an intolerable mess. Figure 1
desc(ii~es the full scale test at Richardson.
TX.
TIaU88
"''''..
01"" Dopth Circu:l Area Volumo
(Ft) (Ft) (Ft) (Sq P~) (Cu Ft) (Gal)
Primary Clarifior No.1 40 8 126 1257 10,054 75,200
2 40 10 126 1257 12,570 94,000
3 40 10 126 1257 12.570 94.000
All Primary Clar1flC!ra 378 3771 35.194 263,200
Pinal Clarifier 70 220 1848 23.088 173.000
PUter No. 1 84 6.5 5542(1) 36,000
2 120 6.5 11310(1) 73.500
FUters Combined 16852(1) 109.500
Diluter No. 1 40 14.3(2) -- 12S7 13,000 135,000
2 40 14.3(2) --- 1257 13,000 135,000
3 40 14.3(2) 1251 13 ,000 135.000
DiSCI ters Combined 39,000 404 .000
Sludse DrJina Beds
12.000 Square Feot
(1) Area ia. acres: 0.127, 0.260 and 0.387, ro.apoctivoly
(2) 14.3 Ufecttve. 18.0 SWD. 15.8 Clear 0 Center
Figure 1. Treatment Plant with Chemical Precipitation Facilities(10)
10-3
-------
Removal of Phosphorus and Colloidal Solids
IV
CHEMICALS: WHICH ONES? WHAT
CAN THEY DO? HOW?
A
Coagulation for use in conventional
treatment usually means an iron or
aluminum salt. Lime is an excellent
coagulant but it's generally more
valuable in physical chemical treatment
rather than for supplementation of
conventional treatment. Lime may be
required along with iron or alum salts
to maintain the biological systems in
a favorable pH range for growth.
1 Iron salts as the chloride or sulfate
are+more effective in the oxidized form
(Fe 3). Waste acid from scale
removal containing reduced iron may
be neutralized and aerated to provide
an effective coagulant.
2 Aluminum salts may be applied as
alum or sodium aluminate. Alum or
iron addition tends to reduce pH upon
hydrolysis. Sodium .aluminate is more
expensive but it includes alkalinity
and tends to raise mixture pH. Its
use may relieve neutralization cost in
low alkalinity water.
B
Coagulant aids, as the name implies,
may be used to favor agglomeration of
destabilized particles or for the same
purpose after coagulation. Many forms
are available and many more are under
development for specific needs.
Available Forms of coagulants(l1)
3
1
Liquid solutions of iron or alum are
easiest to use. Solution preparation
is a tedious and time-consuming job.
Cost, effectiveness, flexibility, and ease
01' handling all point toward use of the
dissolved form. Transportation cost of
water for long distances may rule
them out..
2 Polymers generally are shipped in dry
form.
10-4
C Technical Details on Coagulants
1 Alum as a 48.5% solution is 4.37%
Al and weighs 11. 1 lbs I gal. See
the suppliers data for other preparations.
2 Ferric chloride is available in 35% to
45% solutions. (The 45% solution
freezes at 450F and 37% solution freezes
at 150F.) The 40% solution weighs 11.9
lbsl gal and contains 16.4% Fe.
D Fundions
1 Calcium, iron, and aluminum salts
hydrolyze to form insoluble precipitates
with solublf1~rho phosphate and poly
phosphates
2 Metal salts act upon stable suspensoids
to destabilize the forces maintaining
dispersion such as the reduction of the
surface charge on hydrophobic dispersions
and dehydration of the water layer on
hydrophillic dispersed components513,14; 15)
E Coagulation Reactions and Kinetics
1 Hydrolysis and reactions of metals in
coagulation are very rapid--probably
less than one second.
2 Metal radicals are complicatei 16) .
3 Effects of biologically produced polymers
are obscure(17,18).
F Flocculation Reactions and Kinetics
1 Flocculation primarily is a gathering
together of fine particles produced by
precipitation or coagulation into
aggregates large enough to separate
effectively.
2 Flocculation should provide a decreasing
energy situation and time required is a
matter of 10 to 30 minutes.
3
Coagulant aids, natural polymeric
materials and coagulants function to
-------
glue particles together, to provide a
skeleton for a dense floc and to
"sweep up" straggler particles. Particle
contact is essential at an energy level
favoring agglomeration rather than
dispersion.
G
Inferred Physical Arrangement
1 Intense flash mix for 2-30 seconds
while adding coagulants.
2 High energy flocculation for 1-5
minutes (add polyelectrolytes near
midpoint).
v
3 Low e:1ergy flocculation - 5 to 20
minutes.
HARDWARE: TYPE, SIZE, USAGE(19)
A
Coagulant Feeding Equipment
1 Many types o~ variable discharge pumps
are available. Positive displacement
pumps are effective. Diaphragm,
plunger, gear, and progressing cavity
types and typical alternates.
2 Use pumps designed for 500 psi service,
install a 40 psi back pressure valve and
neglect head loss in following piping.
3 Backpressure helps seat check values.
Double check values are good.
4 Pump sizes that are too large makes low
range control difficult. Some pumps have
interchangeable heads for selection of
suitable control ranges.
5 Ratio control is helpful in addition to
percentage output control.
6 Proper selection of pump materials
permits interchange of coagulants and
polymers.
7 Control units
a Pumpage record from operators log
is one method of manual control.
Removal of Phosphorus and Colloidal Solids
b A magnetic flow meter from storage.
gravity flow, a totalizing indicator
recorder (TIR) and a setpoint
controller on a valve ahead of the
feed point are shown in Figure 2.
Will not work on viscous polymer
solutions.
STORAGE
TANk
--,
I
FEED
PO!HT
MAG METER
Figure 2. Gravity Chemical Feed System
c A compound loop control is indicated
in Figure 3 including a flow meter,
phosphorus analyzer and an integrator
and set point selector and chemical
pump controller.
CHEMICAL PUMP CONTROLLIER
I
I
I
I fLOW METEit I
I
I
I
C:=
~-
Figure 3. Chemical Feed Control by
Compound Loop
10-5
-------
Removal of Phosphorus and Colloidal
d Automatic controls permit closer
trimming of dosage and flow demand
but require close control of sensors,
setpoints, instrumentation and mechanics.
It takes more know-how to keep them
automating.
e Other alternates are available.
B
Flash Mixing
1
This is a most grevious short coming
of many chemical operations. (20, 21r
2
Contact of coagulant, water and substrate
is necessary before the reaction is com-
plete or coagulant hydrolysis products
may not efficiently collect substrate.
3
Hydraulic jump boxes, drop box or man-
hole, pump intake or baffled basins pro-
vide mixing but not flash mixing. However
along with air agitation, these are about
all we have for raw wastewater application.
4 A propeller or turbine mixer, jet, vortex
or other devices(22, 23, 24) can be very
good. The in-line unit is excellent.
Providing an input G of 800- 1000 for up
to 30 seconds is important.
5
For an electrically driven mixer:
G:: [ (WHP) (550)
u V
(5
where:
(WHP)
:: delivered water HP
A Motor Power)
(KV )(Efficiency)( Factor
0.746
absolute viscositY6
(0.2 X 10-4) at 70 F
or
u
V
mixed volume (cu. ft. )
6
For a baffled basin or other head loss
unit where
H ::
1 ft. head loss
T ::
31. 2 sec. detention time O. 5
[ (62.4) (1. 0)] -1
G:: (31. 2)(0. 2 X 10 4) :: 316 sec
then:
10-6
This is much less than the recommended
800 to 1000 G. The chemical reaction
should be complete in 5% of the time
allowed.
a
To produce a G of 1000 the preceding
detention time would be 3. 12 sec and
box volume would be 4.5 cu. ft. /MGD--
an interesting design problem.
b An effective device is outlined in
Figure 4. This is an open flash- mix
box. Mix at the inlet end leaving at
least three fold volume for future
needs. A junction box may serve if
an appropriate mixer is provided.
./
....
./'
Plow Rate. Coaaulat10D Flocculation Time (Minut..)
~ "PI! ~ ~ Low Enerav !!!!!!
1 100 1.42 5.71 28.6 34.3
1., 1.050 0.95 3.81 19.1 22.,
2 1.400 0.71 2.86 14.3 17.1
2.5 1.750 0.57 2.28 11.4 13.7
3 2.100 0.48 1.91 '.5 11.4
Figure.. Junction Box ModUted to Flasb MiJr:(10)
c
Proper mixing in an activated sludge
aeration tank depends on its mixing
characteristics. Try adding coagulants
at the point of maximum energy, and/ or
just as flow proceeds to final clarification.
C Flocculation
1
(25 26 27)
Water plant technology' , serves
for wastewater use.
-------
Removal of Phosphorus and Colloidal Solids
2 High energy flocculation occurs at the
outlet end of the mix box, piping, or
clarifier inlets. Clarifier inlet hydraulics
are critical, and additional baffling may
be needed.
3
Low energy flocculation occurs in the
sludge blanket.
4 Extensive baffles, paddles, etc., probably
will not be needed; can be added later.
5 An activated sludge aeration tank may
be a good high energy flocculator; don't
overexpose at this energy level,
however.
D
Piping
1 Polyvinyl chloride (PVC) and Fiberglass
Reinforced Plastic (FRP), protected from
freezing, is required for coagulants.
2 Provide flushing tees following pumps.
3 Accumulators are not necessary; manual
air blow- offs are recommended. These are
also useful as sampling and calibrating
ports.
4 Use straiI1€rs on pump suctions and
provide oversized suction lines.
5 Install a dilution line in the polyelectrolyte
pump discharge; provide about 1/1 dilution
water by volume. A 50 ft. run of pipe,
a jet or turbulence section in a shorter
pipe run, will provide dilution of the
small feed of polymer.
6 Include hose stations for spillage washdown.
E
Storage Tanks
1 Filament wound or layup fiberglass tanks
from a good supplier are recommended.
Natural amber or colored tanks are
available.
2 Size for a 7-10 day supply on a 2/1 mole
ratio. About 3,000 gal/MGD for alum
(equals three weeks for FeCl3 or aluminate)
is conservative. A 6000 gal. tank will
accept 5000 gal. tank truck lots.
3 Use a polyurethane foam pad between
foundation slab and tank unless weather
dictates a heated pad to keep coagulant
warm.
4 Specify a sight gauge, manhole, fill inlet
with snap coupling, pump suction,
gooseneck vent, flush-bottom drain.
5 Polymer tanks usually are sized for 2-3
day supply of 0.5% solution. This would
be 400 gal/MGD at a 2 mg/l dose. Can
and probably will want to dilute below
0.5% before introduction. Plan for it
on pipe, size and connections.
6 Polymer tank fittings are the same as
for coagulant (E4) plus an over powered
mixer for 1000 cp viscosity dissolving
requirements. Water connectors for
fill inlets and dispenser funnel required.
7 Consider shelter for polymer tanks--
the operator will appreciate it.
8 It is advisable to provide an emer gency
reserve tank for coagulant and
polymer--a 4 to 6 hour supply.
VI Dosage selection control is the key to
success; this controls both cost and
performance.
A In General
1 Inject coagulants at full strength (to
prevent premature hydrolysis) into
an intense dispersion zone. Use one
inlet pipe per mixing unit up to 8 to 10
MGD of treated flow.
2
Inject ploymers in a diluted stream with
thorough dispersion at a moderate
energy level. Have several available
inlet points. Use a multiport header
at each point for treated flows above
3 MGD.
B The Key parameters in dosage control are
the mole ratio of coagulant or metal to
phosphorus (C/p or M/P) on a reacting
equivalent basis, . and also the residual
phosphorus after treatment is completed.
10-7
-------
Removal of phosphorus and Colloidal Solids
1
The coagulant expressed as metallic
aluminum (AI) or iron (Fe) in a pound
molar form makes it possible to change
coagulants or coagulant mixtures and
maintain effective dosage by correcting
for the particular mixture on the basis
of its metallic percentage or decimal
fraction. One pound mole for Al = 27
pounds; Fe=56 pounds.
2
For purposes of this outline, phosphorus
(P) is used as the index of material to
be removed. One pound mole of Al or
Fe is equivalent to 1 pound mole of P.
The molar weight of P is 31 or 31
pounds - 1 pound mole.
3 If the treated flow contained reactable
phosphorus only, it would be possible to
remove P predictably by adding one
pound mole of Al solution for each pound
mo~e of P. Municipal wastewaters are
complex mixtures of different components;
incomplete or competing reactions always
require more coagulant than is indicated
by anyone measured index such as P.
Experience shows an effective ratio to
be 1. 511. 0 to 2. 011. 0 in coagulant metal
(M) to P load. (This may be expressed as
a C/P or M/p ratio.) Since M is used to
represent the term "million" a C/P ratio
is used herein.
4 For each 1 MGD of flow containing 10
mg pi liter (equivalent to 10 lbs. p/M
lbs. of flow)
a Calculate P load in terms of pound
moles I day
10 lbs. (P)x1 MGx8. 34 lbs. x 1 lb. mole
M lbs. ---rr- G 31lbs.
Cancelling common terms and
computation
83.4 = 2.69 or 2.7 lb. moles of P
-31 D
b Knowing the phosphorus per MGD
entering the plant in pound moles
provides an entry for coagulant
dosage, i.e., for a 2.0/1.0 C/P
ratio to produce a satisfactory
effluent--use it.
10-8
2 x 2.7Ib. moles = 5.4Ib. moles in
terms of required metal dosage/MGD.
c This metal may be added as a dry
salt or as a water solution. Suppose
we decide to add it as an alum solution
containing 48.5% alum, and 4.37%
Al weighing 11. 1 lbsl gal. Dosage
becomes:
Alum
5. 4 lb. mole Al (C) 27 lbs. x 100% soln. x 1 gal.
lb. mole 4.37% Al 11. 1 ]b.
= 302 gal. alum solution/MGD coagulant
dose rate.
d The same number of lb. moles of iron
coagulant could be used but the volume
would vary with the iron content of the
applied coagulant. For a 40% FeCl3
solution containing 16.4% Fe and
weighing 11. 9 lbsl gal:
soln.
5. 4 lb. moles Fe( C)x 56 lbs. x 100%FeC13 x 1 gal.
lb. mole 16.4% Fe 11.9Ibs.
= 155 gal. FeCl3 soln/MGD coagulant dose rate.
Hourly dose rate = 155 x.Jr or 6.5
gal/hr. for the average dose rate of
40% Fecl3 solution.
5 The data in Item 4 were based on average
flow and concentration basis. The
average doesn It hold for either flow or
concentration on mupicipal wastewaters.
a Figure 5 shows the typical wide
distribution of influent P by the hour.
Note that P is expressed in lb. per day
basis. Relationship of soluble P and
total P are indicated in the same
graphic. This means coagulant
dosage adjustment according to the
time of day to prevent overdosage
at night and underdosage during the
peak loading.
b Adjusting coagulant dose rate 3 to 5
timesl day is more likely to match
coagulant dose and requirements.
-------
Removal of Phosphorus and Colloidal Solids
500
PHOSPHOROUS LOAD
IN PLANT INFLUENT
~200
!
.!
~ 180
!
o
8A
110-
NIGHT
Figure 5. Phosphorous Income Plot(lO)
..
-
.,
c
Measurement of soluble P requires
coagulant dosage adjustment to meet
total P requirements. In Figure 5
the soluble P was about 70"!o of total
P; therefore, a sensor detecting
soluble P only requires a coagulant
dosage correction of 100/70.
d Daily income of P also may vary.
Experience( 10) showed high P on
Saturday, low P on Sunday and fairly
regular loading on weekdays. Check
your loading pattern.
e
Automatic control relieves the opera-
tor from manual adjustment. (28) The
operator is obligated to adjust set-
points for control in line with effluent
quality and to validate control
operation.
6
Effluent P is a direct result of coagulant
dosage control. Figure 6 shows the
effluent P concentration by the hour.
A twice daily dosage adjustment was
insufficient to maintain effluent P below
1 mg P /liter.
5.0
EFFLUENT PHOSPHOROUS
ALUM IN 'INAL A"p. 1.8"
MOfC)AY. SEPT 21, 1870
PLANT INfWENT
'.9 COMPOSITE
i
.!
..
!1t.O
"
AT 10.8 0
I
UI.80P
LI ID
i
1.0
..
~
..
o
KIA NOON
8A
..
..
lID-
NIQHT
Fillure 6. Too Few Chemical Pump Settings Give Poor Control(lO)
7
Dosage corrections must be varied
until the significant peaks are
eliminated. Figure 7 indicates dosage
changes and effluent P response. Stay
on a given schedule at least five days--
the first day may not show adjustment
effects meaningfully. Don't over-
compensate. Figure 8 shows effects of
split dosage to raw and final clarifiers.
The dosage is the same but better
control is evident.
8 Figure 9 shows a series of iron addition
results indicating effluent P versus mole
ratio of C/P. These results are
characteristic of the influent, the opera-
tion and the situation. Some change may
be expected in other situations.
C Polymer coagulant aids are intended to
improve solids capture and may reduce
coagulant requirements. If total cost is
reduced enough to pay for coagulant aids or
if necessary to meet effluent requirements--
use them.
1
Acquire information on recommended
available polymers.
10-9
-------
Removal of Phosphorus and Colloidal Solids
2
Try several by jar test with your
influent and operations.
'.0
EFFLUENT PHOSPHOROUS
ALUM IN FINAL A.IIP. 2.111
TUESDAY, SEPT 22, 8870
PLANT INFLUENT
..a COMPOSITE
u
"
i:
!2.0
..
"
o
..
o
:0:
:1;
o
:0:
a. 1.0
..
g
..
LIQUID ALUM FED
0.11 C.O~~':"E}
o
IDA NOON
IP
1110-
HIGtrT
Figure 7. Effluent Phosphorous (Tuesday, 9/22/70)
IA
9A
i:
PLANT INFLUENT EFFLUENT PHOSPHOROUS
9.2 COMPOSITE SPLIT FEED AI/P. 211
WEDNESDAY. NOV 4, 1970
(20'" 0' ALUM TO RAW
... 3.8GPH I 4.70'" 4.1 I 2.7GPtt ...
,., I5.8OPH ,I" 18.8GPH .... 10.1- .....
\80" CW ALUM TO FINAL
.
-=- . . . . - ~
./ . -- ~~ COMPOSITE
. . ---:--r-
'.0
r
'.0
"
.!!
g:
~ 2.0
fi
:0:
..
..
:! 1.0
o
..
o
lOA NOOII
SA
IA
IP
180-
NIGHT
Figure 8. Effluent Phosphorous (Wednesday, 11/4/70)
10-10
VII
EFFLUENT PHOSPHOROUS VS IRON DOSAGE
E 1 .
..
z
...
:>
~ 5
...
1
TAE AT MENT 0'
PLANT INFLOW
\
TREATM E NT IN
FINAL CLARIFIER
G.'
!
. ~ 4
..
i! .
..
o
f
..
~ 2
..
..
c
..
I:! I
G
8
2
MOLE RATtO'
I 2
TO PHOSPHOROUS (PI
IRON (m)
Figure 9. Study of Results Anows Selection of Desired Male RaUl"
3
Try the more promising- items in
plant trial.
4
A typical dosage is 1 mg/liter or
less. The proper dose and dispersion
is indicated by floc behavior, "Quicky"
test like effluent turbidity, floc
appearance, and "slick fingers" are
useful. If the plant effluent feels slick
to the touch, you are not getting proper
dilution or. dispersion, are overdosing,
or both. .
5
Feed rate usually is constant, possibly,
reduced at night. Some plants interlock
raw inflow to the polymer pump; this
works well.
SAMPLING AND ANALYSIS
A Coagulation during plant operations
requires more laboratory control than the
same operation without coagulation.
Whatever has been done will continue and
extras are inevitable. As in all upgrading
operations the tests for the record may not
be the same as those for control but both
are essential. A good look at what you are
doing (and why) may lead to significant
improvements.
-------
Hemoval of Phosphorus and Colloidal Solids
B
Figure 10 lists possible tests for
conventional treatment and additional
tests for phosphorus removal according
to stage of treatment or sample.
ANALYSES FOR
CONVENTIONAL TREATMENT
R P F F R S S
A R I I E L U
W I L N C U P
M T A I D E
L R G R
E E C E N
F F E A
F F F T
F
FLOW . . . .
TOT SOL . . . . . . .
TO T VOL SOl . . . . . . .
SUS SOL . . . . . .
SUS YDL SOL . . . . . .
SET SOL . . . .
BOD . . . .
DO . .
COD . . . . . .
PH . . . . . .
TEMP . . . . . .
ADDITIONAL ANALYSES FO R
CHEMICAL TREATMENT
R P F F R S S
A R I I E L U
W I L N C U P
M T A I D E
L R G R
E E C E N
F F E A
F F F T
F
PHOS . . . . . .
ALK . . . . .
FE . . . . . .
AL . . . . . .
504 . . . . . .
CL . . .
TURB . . .
Figure 10. Chemical Precipitation Involved
Added Analyses
1 Some of these are more important
than others. Control tests may be
spot checks on grab samples by the
most rapid available routine. Record
tests may be based upon composite
proportional samples by a referee
procedure. Good control means more
tests and more meaningful testing
with rapid feedback to process control.
2 Coagulant analysis means analysis
of both cation and anion concentration.
Alkalinity may not be necessary if pH
problems are unlikely. When alkalinity
decreases to less than 50 mg/liter watch
it carefully and readjust if it declines
below that range.
3 Effluent turbidity is a rapid and handy
test for dosage control. Sludge blanket
level provides an early warning of
impending problems. Watch for
dispersed floes escaping over the
final weir; stop them with polymer
or a change in metal addition rate.
VIII
SLUDGE HARVESTING AND DISPOSAL
A Clarifiers: Make it drop- - and stay down.
1 Nine feet SWD for Trickling Filter,
final clarifiers, 12 feet for activated
sludge. 500gpd/sq.ft. surface
overflow rate avg. flow, «1000 peak
flow), good inlet and outlet systems
are required.
2 Preserve planket, keep recirculation
low in trickling filters. Watch
operations, particularly, floc behavior.
3 Modification by addition of new "bundles"
can improve performance(29).
B Sludge Treatment and Disposal--Perennial
Headache, Doubled
1 Sludge digests well, percent solids
usually higher than conventional.
2 Be courageous with digesters; Provide
heat and mtx.' Consider thickening,
blit'treat ~flow carefully- - colloidal
P strays.
10-11
-------
Removal of Fhosphorus and Colloidal Solids
3
The volatile solids may increase
50 to 100% by weight- -and carries
added minerals with it.
4
Fe sludge on drying beds better than Al
>conventional. Drying time 1/2. Don't
add too much at one time, maintain
>6" of sand for a good underflow.
5
Chemical cost reduced for vacuum
filter or centrifuge dewatering step.
C
Supernatant, Rogue Recycle, Other Happy
Thoughts
1
Supernatnat treatment before recycle:
a Fill and draw tanks, or flow through
b
250 mg alum/liter plus 20 min air
then settle, or assess P and ~dd
metal at 2/1 mole ratio
c
Sludge to dewatering or digester;
return clear water to head of plant.
It works.
d
Consider aeration, 200- 400 mg
lime/liter, aerate settle. Can
strip NH3 at high pH.
2
Iron leakage: (Alum does too but it
isn't red. )
a Worst on final settler treatment--
use split dose
b
Polymer reduces escaping colloids,
Figure 11.
c
Use most of the iron in the primary.
Give it time to be gathered.
Aeration time helps.
3
Other radicals have an impact too:
a
lIb. AI+3 adds 5.35 lb. sulfate
+3 .
1 lb. Al ,asalummate, adds
0.85 lb. sodium
b
c
+2 .
1 lb. Fe ,as chlonde, adds
1. 26 lb. of chloride
10-12
d
+3
1 lb. Fe ,as chloride, adds
1. 91 lb. chlorides
+3
1 lb. Fe ,as sulfate, adds
2.58 lbs. sulfate
e
IX COSTS
A Capital Investment, Some Gross
Approximations:
1
FRP Tanks: $1/ gal up to 1, 000 gal;
$0.60 above.
2
Pipe, fittings, valves: $l,OOO/MGD,
Don't scrimp here.
3
3 HP Mixer and starter $1, 000.
4
Chemical feeders, starters, with
automatic capability, $1, 500 each.
5
Concrete, baffles, samplers, laboratory
equipment, controls, electric gear,
shelter: depends on the situation.
6
Capital costs are a small part of the total.
IRON LEAKAGE IN PLANT EFFLUENT
10
o
i
- 8
~
..
:>
~
::: 8
..
!
~ 4
~ I
o WHEN TREATING
o 0 (PLANT INFLOW
~~.. 0
0----- 0 ~ 0 0
o 0
Q8
OJ!
1.4
18
u
2.4
18
ZJ)
ID
MOLE RATIO' F. (Drl/P
Figure 11. Iron Leakage in Plant Effluent
-------
Removal of Phosphorus and Colloidal Solids
B
Chemical Costs: This is where the
money goes.
1 Liquid alum 48.50/0 Solution costs
about $0. 24/lb Al in 1972; plus
freight: typical 250 mile haul adds
$0. 12/lb Al.
2 Sodium Aluminate, near $0. 35/lb.
AI, plus freight.
3 Ferric chloride $0. 12/lb. plus freight
(about $0. 05/lb/250 mi).
4 Spent pickle liquor: No guesses,
check local situation.
5 Polymer costs vary widely. About
1~/ 1, 000 gal. is a typical upper limit.
C
Other Costs:
1 Power nominal, about $200/yr/MGD
2 Man power: . With intelligent 24 hour
per day operation the same personnel
numbers can be trained to handle
coagulation. If 24 hour operation wasn't
available it should be. For auto.
control a good instrument man is
needed or on call.
3 Laboratory Support: If lab staff is
available possibly one more. Usually
a good lab man can handle both
conventional and coagulation tests
with technician help by operating staff.
REFERENCES
1
Phosphorus Removal - The State of the Art.
Nesbitt, J. B. Jour. WPCF. May 1961.
p. 701-13.
2
Phosphorus Removal - Past, Present, and
Future. Hall, M. W., and Engelbrecht,
R. S. Water and Wastes Engrg.
Aug. 1969. p. 50-3.
3
Phosphate Removal: Summary of Papers.
Scalf, M. R., et. al. Jour. SED.
ASCE. Oct. 1969. p. 817-27.
4 Wastewater Treatment and Renovation--
Status of Process Development.
Stephan, D. G.. and Schaffer, R. B.
Jour. WPCF. Mar. 1970. p. 399-410.
5 Soluble Phosphate Removal in the Activated
Sludge Process- - A Two Year Plant Scale
Study. Long, D. A., et. al. 26th Purdue
Ind. Waste Conf. 1971.
6 Chemical Flocculation of Microorganisms
in Biological Waste Treatment. Tenney,
M. W., and Stumm, W. Jour. WPCF.
Oct. 1965. p. 1320.
7 The Microbiology of an Activated Sludge
Waste- Water Treatment Plant Chemi-
cally Treated for Phosphorus Removal.
Ung, R. F., and Davis, J. A. 26th
Purdue Ind. Waste Conf. 1971.
8 Coagulation Testing: A Comparison of
Techniques - Part 1. Te Kippe, R. J.,
and Ham, R. K. Jour. AWWA.
Sept. 1970. p. 594-602.
9 op. ciL, "... - Part 2." Oct. 1970.
p. 620-628.
10 Modification of a Trickling Filter Plant
to Allow Chemical Precipitation.
Laughlin, J. Proc. Adv. Waste Treat.
and Water Reuse Symp. EPA.
Dallas, Texas. Jan. 1971. Full report
to be presented at WPCF Meeting.
San Francisco. Oct. 1971.
11 Coagulants for Waste Water Treatment.
Chern. Engrg. Prog. January 1970.
p. 36.
12 Kinetics and Mechanism of Precipitation
and Nature of the Precipitate Obtained
in Phosphate ~emoval from Wastewater
Using Aluminum (III) and Iron (III)
Salts. Recht, H. L., and Ghassemi, M.
Water Poll. Control Res. Rep. 17010EKI
04/70, FWQA. April 1970.
13 Chemistry of Nitrogen and Phosphorous
in Water. AWWA Committee. Jour.
AWWA. Feb. 1970. p. 127-140.
10-13
-------
Removal of Phosphorus and Colloidal Solids
14
Phosphate Removal: Summary of Papers.
discussion by Theis, T. L., et. al.
Jour. SED ASCE. Aug. 1970. p. 1004-9.
15
State of the Art of Coagulation. AWWA
Committee. Jour. AWWA. Feb. 1971.
p. 99-108.
16
Aluminum and Iron (III) Hydrolysis.
Bilinski, H., and Tyree, S. Y., Jr.
Jour. AWWA. June 1971. p. 391-2.
17
Colloids Complicate Treatment Processes.
Dean, R. B. Envir. Sc. and Technol.
Sept. 1969. p. 820-4.
18
Chemical Interactions in the Aggregation
of Bacteria Bioflocculation in Wastewater.
Busch, P. L., and Stumm, W. Envir.
Sci. and Technol. Jan. 1968. p. 49-53.
19
Water Quality and Treatment. AWWA.
3rd ed. McGraw-Hill Book Co. New York.
1971. p. 66-159.
20
Rapid Mixing in Water Treatment. Vrale, L.
and Jorden, R. M. Jour. AWWA. Jan.
1971. p. 52-8.
21
Turbulence in Aeration Basins. Kalinske,
A. A. Ind. Water Engrg. Mar. 1971.
p. 35- 8.
22
Mixing Theory and Practice. ed. by Uhl,
Vincent W., and Gray, Jos. B. Vol. 1
(1966), Vol. 2 (1967). Academic Press.
New York. 1966-7.
23
Liquid Mixing and Processing in Stirred
Tanks. Holland, F. A., and Chapman,
F. S., Reinhold, New York. 1966.
10-14
24 Guide to Trouble-Free Mixers. Penney, W.
Roy. Chem. Engrg. June I, 1970.
p. 171-180.
25 Design of Mixing and Flocculating Basins.
Hudson, H. E., Jr., and Wolfner,
J. P. Jour. AWWA. Oct. 1967.
p. 1257-67.
26 Determination of Optimum Velocity
Gradients for Water Coagulated with
Polyelectrolytes. Hemenway, D. R.,
and Keshaven, K. Water and Sewage
Works. Dec. 1968. p. 554-9.
27 Turbulence and Flocculation. Argaman,
Y., and Kaufman, W.J. Jour. SED
ASCE. Apr. 1970. p. 223-241.
28 Phosphate Removal from Municipal
Sewage. McAchran, G. E., and
Hogue, R. D. Water & Sewage Works.
Feb. 1971. p. 36-9.
29 Improved Settling Tank Efficiency by
Upflow Clarification. Sparham, V. R.
Jour. WPCF. May 1970. p. 801-11.
This outline was prepared from material
written by James E. Laughlin, P. E.,
Shimek, Roming, Jacobs & Finklea
Construction Engineering, 2118 Adolphus
Tower, Dallas, TX 75202.
Descriptors: Wastewater Treatment,
Waste Treatment, Water Pollution Treat-
ment, Phosphorus, Phosphorus Compounds,
Nutrient Removal, Colloids, Coagulation,
Flocculation, Tertiary Treatment
-------
CURRENT STATUS OF PHOSPHORUS CONTROL
TECHNOLOGY FOR MUNICIPAL WASTEWATER
I
INTRODUCTION
The basic chemi stry and engineering factors
needed to provide for a phosphorus removal
process are well documented in "Process
Design Manual for Phosophorus Removal" (1).
This text illustrates the chemical reactions
of iron, aluminum and calcium salts when they
are applied to wastewater, and discusses the
role of polymers for solids capture. It also
documents the need for pacing chemical dosage
to pro sphorus content, and provides guidelines
for handling and storage of chemical inventories.
Engineering techniques for providing adequate
mixing and quisecent gravity clarification are
given. Application considerations for different
municipal treatment processes are highlighted.
Case history reports are presented to define
cost. operational control, and efficiency at
operating facilities.
II
CURRENT IMPLEMENTATION OF
PHOSPHORUS CONTROL TECHNOLOGY
A In the United States, an excess of 300 facili-
ties are providing for some type phosphorus
control. Canada, predominantly in Ontario
Province, has 200 plants on line. Western
Europe removes phosphorus at about 100
locations. Japan is starting towards nutrient
control and has two large facilities in the
process of implementing chemical dosing.
The size of these plants range from 0.2 mgd
to 300 mgd.
B The following list shows the effluent residual
total phosphorus obtainable with different
treatment methods;
SE. TT. 20. 7.77
Type of
Treatment
Residual Total
Phosphorus
Primary Settling
2.5 to 2 mg/l
Biological Secondary:
Trickling Filter
Activated Sludge
Rotating Bio Discs
1 to 0.5 mg/l
Media Filtration
O. 5 to O. 2 mg / 1
Two Stage Lime
0.2 to 0.05 mg/l
C A biological phosphorus removal process
called PhoStrip has recently been shown
to achieve a residual total phosphorus of
1 to 0.5 mg/1. In this process part of
the phosphorus content of active micro-
organisms is leached into a side stream.
Lime is then used to precipitate the phos-
phorus in this small volume.
III
COST OF TECHNOLOGY
A Table 1 provides some general considera-
tions in regard to the cost of phosphorus
removal on the basis of the effluent residual
requirement. The cost is in relative terms
for guidance in the planning of alternate
technology or regulatory decisions. The
three levels of residuals: 1 mg/1. 0.5 mg/l
or O. 1 mg/ 1; are the levels most commonly
applied in existing regulations. The great
majority of facilities are required to meet
only the 1 mg/l value. Similar technology
can be applied for both 1 mg and O. 5 mg/l
requirements. The latter requires the
additional unit process of filtration, without
any increase in chemical additions. Entirely
different technology is required for a residual
of O. 1 mg/l or less, and the total cost will
be about 200 pe rcent more than the 1 mg/ 1
requiremen t.
11-1
-------
Current Status of Phosphorus Control Technology For Municipal Wastewater
B Table 2 shows actual co sts for operating
plants employing various technologies to
meet different residual effluent require-
ments. The range of costs is from
90 cents to 0.5 cents per 1000 gallons
for Ely, Minnesota, and Grand Haven,
Michigan, respectively. Both these costs
are typical. Ely has the most stringent
effluent phosphorus residual of any facility
in the world due to the fact that the dis-
charge is to a lake that is the focus of a
eutrophication recovery study. Grand
Haven, Michigan, is a joint industrial
municipal facility that receives phosphorus
insolubilizing chemical in the influent, and
only a small amount of extra precipitant
is required. The precipitant used is
ferrous sulfate from an industrial waste
source.
C The other cases on Table 2 indicate the
usual magnitude of cost associated with
chemical addition. It is very difficult to
obtain total treatment cost at any location
because internal accounting procedures or
sampling points in most facilities do not
give sufficient refinement to break out
these costs.
D The usual data available is related to the
cost of chemical purchase only. For
chemical addition, the capital cost asso-
ciated with dosing equipment is only a
minor fraction of total capital cost. In
the matter of sludge handling actual costs
are almost impossible to come by. If
phosphorus removal is added to an exist-
ing facility with adequate sludge processing
no real capital cost can be applied. If a
new facility is constructed, the capital
requirements can be obtained; but, in
either case operating costs associated
with the sludge processing become very
vague. It is the general experience at
operating phosphorus control plants with
chemical additions, that operating costs
for sludge handling do not increase greatly(3).
11-2
E The one plant that does have detailed
breakdown of the operating cost of
phosphorus removal is Ely (4). Table 3
gives the cost in cents per 1000 gallons
for several categories and the fractional
percent that each category is of the total
operating cost. In the case of Ely, since
it is a tertiary plant and increased labo-
ratory analytical capability was required
for the lake study, the personnel category
is associated with vacuum filtration of
the total plant sludge and hauling to a
land fill. Actual chemical costs are
13.6 cents per 1000 gallons. The rest
of the costs are about evenly split between
utilities, supplies, and repair of equipment.
IV RELIABILITY OF TECHNOLOGY
A A subject we should now discuss is what
do these costs buy in terms of efficiency
and reliability. Figure 1 is a frequency
distribution plot of 365 daily values for
the residual total phosphorus in the Ely
final effluent. The geometric mean con-
centration is 0.059 mg/1. Certainly the
cost of 90 cents per 1000 gallons is high
compared to other phosphorus removal
processes, but there is no other plant in
existence which achieves such a low resi-
dual.
B Additionally, the entire cost is charged
to phosphorus removal and does not con-
sider the greatly enhanced quality of the
effluent in regard to other pollutants.
This facility has been on line several years
and Lake Shagawa is recovering from a
eutrophic condition.
C The ratio of the mean and standard devia-
tion on Figure 1 is calculated as 1. 7 and
represents a term called the Spread
Factor (5). This factor relates to the
reliability of a process in regard to
residual concentration. A Spread Factor
of 1.0 would indicate the residual con-
centration was absolutely consistent day
in and day out. The higher the Spread
Factor, the less consistent the effluent
concentration would be. The Spread
Factor indicates the slope of the line on
a frequency distribution curve. As the
-------
Current Status Of Phosphorus Control Technology For Municipal Wastewater
TABLE 1
GENERAL CONSIDERATIONS FOR PHOSPHORUS REMOVAL
PHOSPHORUS
RESIDUAL
REQUIREMENT.
CONTROL
TECHNOLOGY
NEEDED
RElATIVE
TOT AL
COST
1 mg/I, or less
1.0
05 mg/I, or le55
0.1 mg/ I, or less
METAL SALT ADDITION TO
CONVENTIONAL SECONDARY
PROCESS
METAL SALT ADDITION TO
CONVENTIONAL SECONDARY
PROCESS, PLUS FINAL
EFFLUENT FILTRATION
TERTIARY TWO-STAGE LIME
COAGULATION AND FINAL
EFFLUENT FILTRATION, WITH
POLISH DOSE OF METAL SALT
.30 Day Average Val\le
TABLE 2
COST OF PHOSPHORUS REMOVAL
1.25
3.0
EFflUENT
PLANT SIZE COST PHOSPHORUS, CHEMICAL
LOCATION mgd (/l,OOOgol. mg/I USED
Ely, Minn. 1.5 90 0.05 TWO STAGE
LIME
EI Lago, Tx. 0.5 13 0.5 FeCI3
River Oaks, FI. 1.1 12 1.0 ALUM INA TE
Hatfield Twp., Pa. 3 6 0.5 LIME + ALUM
Maumee River, 2.5 4. 1.0
(Lucus Co.), OH ALUM
Columbu 5, Ind. 8 2. 1.0 ALUM
Kaukauna, Wi. 2.1 2. 1.0 ALUM
Grand Haven, Mi 3.9 0.5 1.0 i WASTE FERROUS
.Data from Allied Chemical
IRON
11-3
-------
Current Status Of Phosphorus Control Technolop;y For Municipal Wastewater
TABLE 3
TOTAL TREATMENT COST, TERTIARY PHOSPHORU$ REMOVAL
ELY, MINNESOTA, APRil, 1973 - MARCH, 1974
CAPACITY 5700 m3/d (1.5 mgd)
CA TEGORY (/m3 (/1,0009al PERCENT OF TOTAL
Personnel 14.7 55.7 61.1
Chemicals 3.6 13.6 15.1
Utilities 2.3 8.8 9.7
Supplies 1.4 5.3 5.8
Equipment
Operation 2.0 7.5 8.3
and Repairs
TOTAL 24 90.9 100
COST
0.2
ELY, MINNESOTA
. ........
CJ) 0.05
E
0.035
0.059
SPREAD FACTOR =- = 1.7
0.035
0.1
0.059
0.02
MEAN
STANDARD DEVIATION
0.1%
1%
10%
50%
90%
99% 99.9%
PERCENT OF OBSERVATIONS
-------
Current Status Of Phosph~!us Control Technology For Municipal Wastewater
slope increases the Spread Factor becomes
larger.
D The Spread Factor can be used to compare
the reliability of various plants regardless
of the absolute magnitude of the residual
concentration. Table 4 compares several
phosphorus removal plants, both in the
United States and Northern Europe. Differ-
ent chemicals for phosphorus control are
employed at different points in the treatment
process at the various facilities. In general
the plants employing the tertiary process
report the lowest arithmetic average residual.
Even though different dosing sites are employ-
ed a frequency distribution plot shows the
Spread Factor was about the sarre for each
case. This indicates that chemical control
tends to lessen variability and increase reli-
ability of obtaining target effluent residuals.
E Table 5 substantiates this reasoning. Various
residual pollutants in the South Lake Tahoe
final effluent are given. The Tahoe plant
employs tertiary lime, filtration and carbon
adsorption. All the residuals show about
the same Spread Factor, with the exception
of the coliform data which is based on a biologi-
cal assay.
V UNIT PROCESS CONTROL
A In general discussions of phosphorus control
we tend to think only of influent and effluent
values. However, for reliability of operation
to obtain low Spread Factors the fate of
phosphorus through each unit process in a
system must be established. Figure 2 shows
the efficiency of various unit processes at
Hatfield Township, Pennsylvania(6). Of
. special importance is the impact of recycle
streams on unit process operation. At this
plant all recycle streams such as, thickener
overflow, vacuum filtrate, alum floc, incin-
erator scrubber water and filter backwash
enter a surge tank. The effluent from this
surge tank contains more than twice the
phosphorus concentration of the influent
wastewater. At this facility lime addition
to pH 10 is used to remove the bulk of the
phosphorus at the primary solids-contacting
process. Incremental removal is then
achieved via cell synthesis in the activated
sludge process and by tertiary alum coagu-
lation. The alum clarified effluent is of
such high quality that additional multimedia
filtration has little effect on overall effici-
ency. The slight difference between tertiary
effluent and final effluent values in probably
related more to sampling and analytical
variations then actual efficiencies. The main
function of the filter is reserve capacity for
shock alum floc carry-over.
VI EFFECT OF LOWER INFLUENT PHOSPHORUS
CONCEN TRA TION ON TECHNOLOGY
A The above discussion shows that phosphorus
removal at municipal facilities can be
achieved to meet various effluent residual
requirements. The cost to obtain a 1 mg/l
residual is not prohibitive and the processes
employed are reliable.
B Presently, two major process choices for
phosphorus control are: metallic salt addition
to conventional secondary plants (with or
without terminal filtration), and tertiary lime
coagulation and filtration. An emerging
technology is biological control via an anoxic
stripping process with chemical treatment
of a side stream concentrate. (The previously
mentioned PhoStrip process.)
C In several locations in the United States a
ban on phosphates in detergents exists or is
under consideration. While this approach
wi 11 reduce the amount of phosphorus reaching
receiving waters, it will not greatly reduce
the cost of phosphorus removal at wastewater
treatment facilities.
D To estimate the impact of a lower influent
phosphorus concentration on capital and
operating cost we will have to assume the
effluent requirements remain the same as in
existence today because these present re-
quirements are based on water quality con-
siderations. Municipal wastewaters are of
such a composition that influent phosphorus
concentrations in the range of 5 mg/l will
still produce a residual of greater than 1 mg/l
unless some typical situation exists. Add-
itionally, as has been note d above, internal
recycle streams contribute measurable
phosphorus loads to the treatment facility.
11-5
-------
Current Status Of Phosphorus Control Technology For Municipal Wastewater
TABLE 4
PHOSPHORUS RESIDUALS AND SPREAD FACTORS AT VARIOUS LOCATIONS
B
PHOSPHORUS
NUMBER OF AVERAGE SPREAD
PLANT CHEMICALa DATA POINTS (mg/I) FACTOR
5jalevad AI III 51 0.48 1.85
leksand Co IIIb 21 0.36 1.62
orlange Fe+3111 52 0.91 1.87
anshytten Fe+211 52 0.61 1.51
Eoishall All 52 0.65 1.64
nd Haven Fe+21 285 0.85 1.54
Ely Co III 365 0.043 1.70
Tahoe Co III 1145 0.20 2.11
Vikm
Gra
a
I, II, and III reler to chemical additions at the primary, secondarv,
and tertia ry stages, respectively
b
Co = Ca(OH)
2
TABLE 5
SPREAD FACTORS FOR SOUTH LAKE TAHOE EFflUENT. 1968 - 1974
PARAMETER MEDIAN SPREAD FACTOR
MBAS mg/I 0.18
BOD mg/I 1.3
COD mg/I 9.6
Suspended Solids 0
Turbidity JTU 0.30
Phosphorus mg/I 0.20
Chlorine Residual mg/I 0.90
Coliform MPN/100 ml (0.025)
11-6
1.72
2.05
1.57
1.94
.2.11
1.77
13
-------
Current Status Of Phosphorus Control TechnoloRY For Municipal Wastewater
TABLE 6
REDUCTION IN COST ~ REDUCTION OF PHOSPHORUS CONCENTRATION
Percent Reduction In Cost If
Phosphorus Control Influent P Were 5 mg/I ~ 10 mg/I
Process Capital Operating
Metal Salt to 0% 30%
Secondary System
Metal Salt to 0% 30% - 40%
Secondary System
Plus Fi Itration
Tertiary Lime 0% 0%
Biological Removal 15% 20%
with Anoxic Stripper
£L
.....
:! 2 50
o
~
V)
~ 200
230
HATFIELD TOWNSHIP, PA.
12,000 m3/d
C)
Z
~ 150
c(
~
w
all:
V) 100
;:)
all:
o
J:
5; 50
o
J:
Ea..
12.5
9.9
FINAL
~
o
RA W SURGE PRIM.
AER.
SEC.
TERT.
FlU.
FIGURE 2. PERCENT PHOSPHORUS REMAINING. AVERAGE VALUES
APRIL.1973 - MARCH. 1974
11-7
-------
Current Status Of Phosphorus Control Technology For Municipal Wastewater
A fifty percent reduction in inn uent
phosphorus as a result of a phosphate
ban does not translate to a fifty percent
reduction in chemical demand. This is
due to the fact that a portion of the
added chemicals are used up by non-
specific reactions with extraneous ions.
enough chemical must be added to aid
coagulation not simply enough to insolu-
bilize the phosphorus. and finally a
slight excess is necessary to insure the
solubility product of the metal/ phosphorus
compound is under control.
References
1.
"Process Design Manual for Phosphorus
Removal." U. S. Environmental Protection
Agency. Technology Transfer. EPA 625/
1-76-001a (April 1976). 2nd Ed.
2. Lukins. B. W.. and Smith. J.. "Interim
Report on the Impact of P. L. 92-500
on Municipal Pollution Control Technology:
EPA-600/2-76-018. January 1976.
E Within this framework. Table 6 can be
constructed. The basis for the table is
the fact that municipal wastewater has
about a 50/50 split between detergent "
formulation derived phosphorus and humas
derived phosphorus.
3. Braasch. D. A.. "A Survey of Phosphorus
Removal in Wisconsin." Water and Sewage
Works. July 1976 (pages 70-73).
4. Sheehy. J. W.. and "Evans. F. L..
"Tertiary Treatment for Phosphorus
Removal at Ely. Minnesota.: EPA-600/2-
76-082. March 1976.
F Table 6 shows that capital cost would not
be substantially reduced if the effluent
residuals requirements remained the same
as Table 1. Except in the case of tertiary
lime. treatment operating cost would be
reduced. but not in proportion to the
magnitude of the phosphorus reduction.
5. Dean. R. B.. and Forsythe. S.. L. .
"The Reliability of Advanced Waste
Treatment" to be published in Water
and Sewage Works (submitted Dec. 1975).
G The estimates for the biological process
are conceptual and not based on any opera-
tional data.
6. Greenlund. T. W. and Gaines. F. R..
"Hatfield Township. Pennsylvania
Advanced Waste Treatment Plant. "
EPA-600/2-75-030. September. 1975.
This outline was prepared by E. Barth. Chief
Biological Treatment Section. Wastewater
Research Division. Municipal Environmental
Research Laboratory
Descriptors: Phosphates. Phosphorus.
Phosphorus Compounds. Nutrient Removal.
Organophosphorus Compounds. Wastewater
Treatment. Waste Treatment. Water
Pollution Treatment
11-8
-------
COMPOSITION OF SLUDGES
INTRODUCTION
A Definition of sludge
B Reasons why we need to know the composi-
tion of sludges
II
PRODUCTION AND PHYSICAL
CHARACTERISTICS OF SLUDGES
A Sewage treatment plant
B Types of sludges produced
C Amounts presently produced and future
trends
D Some physical characteristics of sludges
III
CHEMICAL CHARACTERISTICS OF
SLUDGES
A Organic content
B Inorganic content
C Nitrogen and phosphorus - nutrients
D Metal content
IV SOME HAZARDS CONNECTED WITH THE
DISPOSAL OF SLUDGE
A Metal content
B Pathogens in sludge
C Nitrate to ground water
SE. AE. s1. 7.7.77
~
This outline was prepared by B. V.
Salotto. Research Chemist. Municipal
Environmental Research Laboratory.
USEPA. Cincinnati. Ohio 45268
Descriptors: Wastewater Treatment.
Waste Treatment. Water Pollution
Treatment. Sludge. Sludge Treatment.
Sludge Disposal
12-1
-------
SLUDGE INC INERA TION
INTRODUCTION
A Incineration of sewage sludge started in
Dearborn. Michigan in 1935. Since that
time incineration has expanded because
landfill areas and lagoons for sludge m ve
become harder to find farther from treat-
ment plants and more expensive. A large
community has sufficient sludge to support
the continuous operation of an incinerator.
and often finds incineration the best means
of ultimate sludge disposal. Right now
incineration processes are being looked at
as one alternate to ocean dumping which will
come to an end in 1981.
B Incineration reduces the sludge to a sterile
ash having about five percent of the original
volume. Originally. incineration did cause
air pollution. but high pressure drop water
scrubbers or electrostatic precipitators have
been added to take care of air pollution
problems.
C "Pyrolysis" as it is called is starved oxygen
(air) incineration. At the present time this
method of incineration is being demonstrated
in a multiple-hearth furnace pilot plant. and
could become the method of incineration most
used in the future.
II SLUDGE CHARACTERISTICS
A The parameters of sludge Which are most
important to the incineration process are:
1
Moisture content
2 Combustable content
3 Calorific value
4 Ine rts
Moisture is the characteristic over which
the treatment plant operator has no control.
It is generally reduced by mechanical sludge
dewatering techniques before incineration.
SE. AE. s1. 8. 7.77
B Sludge is normally less than one percent
solids at the start of processing and is
dewatered by some form of filtration
until it is 18-25 percent solids. It is
important to remove as much water from
the sludge as possible in order to reduce
the amount of energy (fuel) used in the
incineration process. When the sludge
has 28-30 percent solids or more. it
usually has enough heating value to dry
and burn itself. and fuel is only necessary
to heat up the incinerator to start burning.
Dewatering has become an important step
in sludge incineration processes. Formerly
fuel did not cost as much and/ or was not
in short supply. so it was not as important
to dewater as much as possible as it is now.
C The combustible material (often called
volatile) material in the sludge. determines
its BOU value. heating value or calorific
value. The combustibles can be paper.
animal or petroleum grease and/ or oils.
ground garbage or about anything that can
be put into sewers.
Table I by Owens (1). describes the types
of solids and percentage combustibles.
the ash content and BTU content of the
more important materials normally in-
volved in sludge incineration at a municipal
waste treatment plant.
D The calorific value of sludge is considered
to be 10. 000 BOU per pound of combustible
material in the sludge. Experience indicates
this usually ranges between 5000 and 7500
BOU per pound of dry sludge depending upon
the sludge source. A given plant no rmally
does not have an appreciable variation in
sludge calorific value. so once the calorific
value is determined it remains the same
summer and winter. The calorific value
will vary from plant to plant depending on
the sources of sludge and the method of
processing. if the sludge is digested. this
lowers the heat content. however digestion
creates a fuel gas ,which can be used for
burning the sludge. Table IV by Owen gives
an example of the variation of calorific value
of sludges at some different plants.
13-1
-------
Sludge Incineration
TABLE I
. Combustible Matter and Heating Values
of Municipal Wastes and Low Rank Fuels
Average B. t. u. per pound
Moisture-free basis
ITEM
COMBUSTIBLE
0/0
88.5
ASH
0/0
11. 5
Grease & scum
Lignite
89.0 11. 0
74.0 26.0
86.4 13.6
89.5 10.5
86.0 14.0
84.8 15.2
97.5 2.5
49.6 50.4
91. 8 8. 2
89.0 11. 0
59.6 40.4
Fresh sewage solids
Fine screenings
Wood
Peat
Ground garbage
Rags
Digested sewage & garbage solids
Newsprint
Rubbish, less tins and glass
Digested sludge
B.t.u.
16,750
10,850
10,285
8,990
8,67'5"
8,275
8,245
8.050
8.020
7,825
5,900
5,290
9!!~___--------------------------------------------~~~~---------~~~~-----~'-~~~--'"----
TABLE IV
Calorific Value of Sludges
B. t. u. per pound
PLANT
TYPE OF SLUDGE
DRY SOLIDS
COMBUSTIBLE SOLIDS
CS
DB
PE
AA
D,
DT
AS
MS
Digested
Digested
Digested
Digested
Fresh
Fresh
Fresh
Fresh
5,975
5,830
5,290
4,980
9,485
10,285
9,125
7,010
12,480
10,750
8,980
9,500
13,855
13,905
11,415
10,155
-------------------------------------------------------------------------------------
13-2
-------
Sludge Incineration
E Inerts are always present in the sludge
and reduce the calorific value of the
sludge, because they dilute th€ combustible
content of the dry solids. Lime and ferric,
or other solids are often added to aid filtra-
tion and tend to decrease the percentage of
combustibles of the sludge.
III THEORY
A The incineration process involves two steps;
drying and combustion. Also required are
fuel and air for combustion, along with
time, temperature and turbulence of the
air. The sludge requires a certain amount
of mixing, agitation or dispersion to break
it into small enough masses to assure con-
tact with the air for combustion. Drying
should not be confused with dewatering.
Dewatering is a mechanical means of water
. removal before the sludge is introduced to
the incineration process. A sludge having
about 20-25 percent solids is delivered to
most types of incinerators. It is necessary
to use fuel to evaporate (dry) another five to
ten pounds of water to get the solids content
up to 28- 30 percent, before the sludge will
have enough fuel value to burn. At about
28 to 30 percent solids the sludge has a
high enough heating to burn without using
another source of fuel.
B Drying and combustion are usually in the
same piece of equipment. This is not
necessary, and there are a few processes
where the drying is done in one piece of
equipment and then the dry sludge conveyed
to another piece of equipment where it is
burned. There are several types of equip-
ment available for incineration. These
include; multiple-hearth furnaces, fluidized
bed furnaces, rotary-kiln type furnaces,
flash drying units, atomized spray units,
wet oxidation units and traveling grate units,
to mention a few. The drying and combustion
process consists of tre following phases:
1
Riising the temperature of the sludge to
2120F.
2 Evaporating the water from the sludge
at 2120F.
3 Increasing the temperature of the
water vapor to the temperature.
at which it is exhausted (800 to
15000F).
4 Increasing the temperature of the
sludge combustibles to the ignition
point.
C A very important part of any burning opera-
tion is the excess air required. There is
usually a trade off of how much excess air
is needed to get better burning efficiency,
and how much the energy will cost to heat
this excess air. Normally, 50 percent
excess air is used for burning operations.
The excess air has the effect of reducing
the burning temperature and increasing
heat losses from the furnace.
D The dry sludge burned in a furnace emits
heat. The heat is used for heating the in-
coming sludge and evaporating the water
and heating the resultant water vapor (steam).
Also, the heat increases the temperature
of the excess air going through the furnaces,
as well as the gases created from the com-
bustion process. Heat is lost by radiation,
hot flue gas and ash, leaving the furnace.
Fuel must be added to the furnace to bring
it up to combustion temperatures for starting,
then enough fuel must be added to the furnace
to make up any for the heat used for drying
and furnace heat losses.
E The products of incineration are considered
to be water vapor, carbon dioxide and ash.
F Pyrolysis is a starved air (oxygen) made of
combustion. The basic differences between
incineration and starved air incineration is
better control of the operation, because
burning must take place where the air with
oxygen is present. Pyrolyses can limit the
amount of burning taking place, and form a
carbon char instead of ash. It can also form
combustible gases of sufficient quantity so
that they can be used as fuel. If they are not
used as fuel, these gases can be afterburned
to prevent emissions to the air.
G Air cleaning systems are not a part of the
incineration process, but must be considered
with it. It is necessary to remove the parti-
culates from the offgas and have a clean,
odorless stack to have a good incineration
process. This normally means a high
pressure drop wet scrubbing system, or
an electrostatic precipitator.
13-3
-------
Sludge Incineration
IV TYPES OF FURNACES
A There are three types of furnaces
generally used for municipal sludge
incineration. They are;
1
multiple-hearth furnaces.
2 fluidized bed furnaces. and
3 rotary drum furnaces.
B Multiple-hearth furnaces (Figures 1 & 2).
1
The multiple-hearth furnace consists
of a vertical cylinder. steel plate shell.
lined with firebrick and having firebrick
hearths. one above the other throughout
the height of the furnace. In most furnaces.
successive hearth levels will alternate
between an in-feed design and an out-feed
design. Figure 2.
2 The in-feed design is one where sludge
moves from the hearths outer edge and
drops through a large opening or part
near the central shaft. Sludge travels
the opposite direction in an out-feed
hearth. As sludge reaches the outer
edge. it falls through a number of ports
or drop holes around the periphery onto
the next hearth which is an in-feed design.
3 The alternating hearth designs and
selection of the number and position
of the ports give optimum distribution
and flow. and help to regulate gas velo-
cities. A central hollow air-cooled shaft.
with air cooled radial arms and plow s .
(rabble arms and teeth) located above
each hearth moves the sludge toward
the inside port. or to the periphery of
the hearth. Cool air enters through the
inner tube within the shaft. and travels
the full length of the shaft and by branch-
ing inner tubes to the extreme end of
each arm. Air then returns through the
space between the inner tube and shaft wall.
either to the bottom hearth as pre-heated
combustion air. or to a duct for outside
venting. Burners and their control systems.
are arranged from the top to the bottom of
the furnace. The burners supply energy
to heat up the furnace to operating tempera-
ture. evaporate moisture and to sustain
burning where necessary.
13-4
a Furnaces are built up to
26 ft diameter. and have
up to 12 hearths. Usually.
for capacities under
10.000 Ibs/hour. the furnace
has 5 to 7 hearths and a
diameter of 6 ft to 19 ft.
Above 10.000 Ibs/hour. the
number of hearths will in-
crease up to 12 and the
diameters up to 26 ft.
b Because of the phenomenon
known as "thermal jump".
offensive odors do not exist
in the vent from the furnace.
Volatile matter will not escape
from sludge until it has 80
percent solids content. The
sludge in the furnace is held
at 1600F by evaporation until
the solids increase to 40
percent. then the sludge enters
the flame area where the
volatiles (odor) are burned as
they are formed.
c
The exhaust gas contains about
10 percent ash as it leaves the
furnace. This ash is normally
removed by a wet scrubber
system which is a relatively
low cost method for gas cleaning.
C Fluidized Bed
1 Fluid-bed manufacturers produce a
compact form of furnace suitable for
burning almost anything that can be
fed to it. Figure 3. This furnace
consists of a firebrick lined. cylind-
rical steel shell bed support. or air
distributing grating near the bottom.
A bed of selected silica sand is
maintained in a highly turbulant
state (fluidized bed). and appears
about the same as sand suspended
in water. To fluidize the bed. air
is compressed. passed through a
heat exchanger to recover heat from
the vent gas and then fed to the bottom
of the furnace. Combustion is accomp-
lished by spraying both sludge and
supplementary fuel into the fluidized
bed where the material burns in about
two seconds or less at 17200F (950oC).
-------
Sludge Incineration
FIGURE I
Ib
bJ
.
~ Waste Cooling Air to Atmosphere
f:
Clean Gases
to Atmosphere
Fan
Protection
System
\
f
r----- ----"l : I
I
t
To Storage
Hopper
I
III
I
I
t
I
I~
Filter Cake, .
Screen i ngs
& Grit
I~
f:
o
NICHOLS/HERRESHOFF SLUDGE FURNACE - BURNING FLOW DIAGRAM
13-5
-------
Sludge Incineration
FIGURE
'0 ...
........
Furnace under construction-an Out-feed heanh. steel shell .nd
central shaft In place.
In-feed t..rth In place.
13-6
l
2
Central shaft sections and main gear under erection.
-------
Sludge Incineration
FIGURE
3
SECONDARY
SCRUBBfR
FIGURE 4
FEED
ASH
Fig. 4
Diagram of Rotary Furnace
13-7
-------
Sludge Incineration
2
This furnace discharges all the ash
in the flue gas. After discharge.
the flue gas travels through a
recuperative heat exchanger to
heat the incoming fluidizing air.
This heat exchanger must handle
gas at 9000 to 10000C which makes
it rather costly initially. and at
those temperatures it is a high
maintenance price of equipment.
If the temperatures for this heat
recovery unit are lowered then the
heat balance around the system
indicates the overall system is not
economical compared with a multiple-
hearth furnace.
a This type furnace is normally
used to burn sludges with a
~ignificant content of highly
volatile material. such as those
from an oil refinery.
b The size of this type furnace
is li~ited by the ability to
distribute the feed uniformly
over the fluidized bed. and
because of this it is not
possible to build one with
as large a capacity as some
multiple hearths.
c Since all the ash leaves in
the vent gas stream. there
is a considerable amount of
air cleaning necessary to
prevent air pollution. This
is accomplished by first using
a dry cyclone dust removal
system followed by a high
pressure drop water scrubbing
system.
D Rotary Drum Furnace
A rotary drum furnace consists of a huge.
slightly inclined cylinder. supported on
rollers at each end and with a large gear
ring around the cylinder near the mid
point. The gear is used for rotating the
cylinder. A hood at each end acts as .
duct' work and an air seal for the rotating
cylinder. The inside of the cylinder is
lined with fire brick which are also arrang-
ed to form baffles within the furnace.
13-8
Sludge and heated air are normally
introduced at the upper end of the
furnace and flow together to the
lower end. Due to the difficulty
in lining the drum with fire brick.
the heat losses are high and a well
dewatered cake is necessary to achieve
incineration. A rotary drum incinerator
can be used for incineration of grit and
skimmings. Also. it can be used for
co-incineration of sludge and solid
wastes with countercurrent air to
dry the material before it burns. The
rotary kiln sludge incinerator has about
20 percent by weight ash in the flue gas
which makes the clp.ani.ng up the flue gas
expensive when compared to a multiple-
hearth furnace. There is also a problem
of air sealing the cylinder at each end
which increases the flue gas quantity
to be cleaned.
REFERENCES
Owen. Mark B.. Journal of the Sanitary
Engineering Division. Proceedings of
the American Society of Civil Engineers.
Vol. 83. No. SA1 paper 1172. Feb 1957.
Sludge Incineration.
This outline was prepared by Howard Wall.
Chemical Engineer. Municipal Environmental
Research Laboratory. USEPA. Cincinnati.
Ohio 45268
Descriptors: Wastewater Treatment. Waste
Treatment. Water Pollution Treatment. Sludge.
Sludge Disposal, Sludge Treatment
-------
REUSE OF MUNICIPAL WASTEWATER EFFLUENTS
NEED FOR REUSE
A Contamination of surface and groundwater
B Alternate for meeting future water demands
C National mandate to conserve water
1 P. L. 92-500 Federal Water Pollution
Control Act Amendments of 1972
2 P. L. 93-523 Safe Drinking Water Act of
1974
II
PRESENT EXTENT OF PLANNED REUSE
A Types
1 Irrigation and agriculture.
2 Industrial
3 Recreational
4 Groundwater recharge
5 Miscellaneous
B Treatment Requirements
C Volume - less than 2 percent of available
municipal wastewater
D Geographical distribution
III PAST REUSE RESEARCH
A Treatment technology development
1 AWT research since 1960
2 Exists for most reuse purpose
B Socio-economic
C Health effects
W. RE.lu. 3. 7. 77
IV NEAR TERM REUSE CONCEPTS
A Public is favorable where health effects
are not an issue
B Economics is driving force
C Source substitution
1 High quality water not used where
lower quality is adequate
2 Dual distribution.
3 Industrial quality requirements
D
Water quality management tool - inter-basin
transfer
E
Reuse in 208 area wide planning
V
POTABLE RE USE AS A LONG-TERM GOAL
A
Health effects are main issue
B
AWWA and WPCF resolution.
C
Groundwater recharge
1 California
2 Chicago
3 Long Island
D
Denver program
VI
DIRECTIONS OF REUSE RESEARCH
A
Near term non-controversial projects
1 Identify applications and institutional
arrangements
2 Adaptability to various qualities
3 Dual distribution systems
4
Water conservation
14-1
-------
REUSE OF MUNICIPAL WASTEWATER EFFLUENTS
1 Characterize residues in effluents
and Recreational Impoundments, "
"State of California Department
of Public Health, May 1968.
B Potable Reuse
3 Pollutant behavior in underground
environment
2 "Report on the Consulting Panel
on Health Aspects of Wastewater
Reclamation for Groundwater Recharge, "
State of California Water Resources
Control Board, June 1976.
2 Large scale demonstrations
4 Toxicology
5 Monitoring and regulation
3 U. S. Environmental Protection Agency,
"Research Needs for the Potable Reuse of
Municipal Wastewater, " EP A Tech. Ser.,
EPA-600/9-75-007 (Dec. 1975)
6 Epidemiology
7 Socio-economic aspects
4 "Water Policies for the Future, " Final
Report of the National Water Commission.
June 1973.
VII
PROJECTS UNDERWAY
A AWWA research foundation coordination
5 Schmidt, C. J., et aI., "Municipal Wastewater
Reuse in the U. S. " Journal Water Pollution
Control Federation, 47, 2229 (1975).
B Blue Plains potable reuse results - slide
presentation
6 Stone, Ralph, "Water Reclamation: Technology
and Public Acceptance, " Journal of the
Environmental Engineering Division, ASCE,
Vol. 102, No. EE3, Proc. Paper 12193, June
1976, pp. 581-594...
C Orange County Water District
D Non-volatile organics
E Wastewater in drinking supplies
7 "Wastewater Reclamation Project, St. Croix,
U. S. Virgin Islands, " EPA Tech. Ser..
EPA-600/2-76-134 (June 1976)
F Status of virus removal
G Guidelines for source substitution
feasibility
8
"Appollo County Park Wastewater Reclamation
Project, Antelope Valley, California," EPA-
600/ 2-76-022 (March 1976)
H Quality of A WT effluents
I
Dallas, Texas pilot plant
9 "Renovated Wastewater as a Supplementary
Source for Municipal Water Supply: An
Economic Evaluation, " EPA Environmental
Health Effects Res. Series, EPA-600/1-76-
033 (Oct. 1976)
VIII SUMMARY
A Industries, municipalities, and public
will make decision
10
Warner, H. P., et al., "The Influence of
Municipal Wastewater Treatment on Used
Water Withdrawn for Domestic Supplies"
presented at the A WW A Convention in
Anaheim, California. (M ay, 1977)
B EP A is providing research
REFERENCES
1
"Statewide Standards for the Direct Use
of Reclaimed Wastewater for Irrigation
14-2
-------
REUSE OF MUNICIPAL WASTEWATER EFFLUENTS
11 Doran, R. B., "Estimating the Reliability
of Advanced Waste Treatment", Water &
Sewage Works (June 1976)
This outline was prepared by John N. English,
Sanitary Engineer, Municipal Environmental
Research Laboratory, USEPA, Cincinnati,
Ohio 45268
Descriptors: Wastewater Treatment, Waste
Treatment, Water Pollution Treatment,
Effluents, Reclaimed Water
14-3
------- |