XEA1
'VATI IJ
WATER POLLUTION CONTROL RESEARCH SERIES • ORD-17O7OEHEO7/7O
ELECTROOSMOTIC PUMPING
FOR DEWATERING SEWAGE SLUDGE
U.S. DEPARTMENT OF THE INTERIOR • FEDERAL WATER QUALITY ADMINISTRATION
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe
the results and progress in the control and abatement
of pollution in our Nation's waters. They provide a
central source of information on the research, develop-
ment, and demonstration activities in the Federal Water
Quality Administration, in the U. S. Department of the
Interior, through inhouse research and grants and con-
tracts with Federal, State, and local agencies, research
institutions, and industrial organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Planning and Resources Office, Office of Research
and Development, Department of the Interior, Federal Water
Quality Administration, Room 1108, Washington, D. C. 20242.
-------
ELECTROOSMOTIC PUMPING FOR DEWATERING SEWAGE SLUDGE
by
Jerome Greyson
Rocketdyne, A Division of North American
Rockwell Corporation
Canoga Park, California 91304
for the
FEDERAL WATER QUALITY ADMINISTRATION
DEPARTMENT OF THE INTERIOR
Program #17070 EHE
Contract #14-12-568
FWQA Project Officer, Dr. R. B. Dean
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio
July, 1970
-------
FWQA Review Notice
.This report has been reviewed by the Federal
Water Quality Administration and approved for
publication. Approval does not signify that
the contents necessarily reflect the views
and policies of the Federal Water Quality
Administration, nor does mention of trade
names or commercial products constitute
endorsement or recommendation for use.
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price 65 cents
-------
ABSTRACT
Electroosmotic pumping derives from the electrokinetic phenomenon of
electroosmosis, i.e., the flow of water through porous matrices or
constrained suspensions of fine particles that is produced by an
applied electric field. Electroosmotic pumping is theoretically inde-
pendent of the cross-sectional area of the pore structure of the
filtering media. Thus, the inherent limitations of conventional
filtration resulting from pore clogging or filter cake compression
are minimized, and electroosmotic pumping provides a technique for
dewatering hard-to-filter materials. An investigation of electro-
osmosis for application to dewatering waste sludges to a solids level
sufficient for auto-incineration has been carried out.
The feasibility program consisted of investigation of electrooamotic
flow in sewage sludge suspensions and identification of parameters
critical to efficient electroosmotic pumping under varying conditions
of applied voltage, current flow, and variability of the surface
properties of the sludge solids. Appropriate configurations for
pumping apparatus were investigated through design and testing of
prototype electroosmotic pumps, and new electrode materials were
tested for anode service in electroosmosis.
It has been determined that electroosmosis can be applied to dewatering
waste sludges of various types. For efficient dewatering from very
wet sludge (2$) to burnable solids, it is necessary to orient the
pumping apparatus with a cathode screen beneath the anode so that an
initial sludge film can be formed by gravity settling at the cathode
surface. Sludge cannot be thickened to a solids level exceeding 20$
if the cake is beneath the liquid level of the sludge suspension; for
drying beyond the 20$ value, the cake must be brought out of the liquid.
Electroosmotic pumping is slow. Thus, large electrode surface areas
are required for pumping productivity to keep pace with a typical
treatment plant's sludge production. However, the development of
stannic oxide for anode service is expected to reduce projected elec-
trode costs. Furthermore, in tandem with a pre-thickening process,
the rate of dewatering by electroosmosis becomes less critical.
Operating costs for electroosmotic dewatering appear to be promising
for sludges with solids contents of 6$ or greater, further suggesting
that electroosmosis may be more practically used as a process in
tandem with a pre-thickening stage.
This report was submitted in fulfillment of Contract 14-12-568 between
the Federal Water Quality Administration and the Rocketdyne Division of
North American Rockwell Corporation. The content of the report is a
summary of work carried out under Contracts 14-12-406 and 14-12-568
during the period 24 June 1968 through 23 July 1970.
The Rocketdyne assigned report number is R-8360
ii
-------
CONTENTS
Abstract ii
Conclusions vi
Recommendations •
Introduction 1
Experimental Program 4
Sludge Collection and Water Flow Patterns 4
Interelectrode Distance, Voltage, Current, and Power 8
Electrode Distance 8
Controlled Voltage and Current 9
Constant Power Operation 9
Sewage .Sludges ' 11
Water Treatment Sludges 19
Prototype Apparatus 19
Multiple Disc Electroosmotic Pump 22
Inclined Plane Electroosmotic Pump 24
Dewatering in the Inclined Plane Apparatus 24
Wet Sludges 27
Pre-concentrated Sludges 29
Electrode Materials 29
Estimates of Electrical Costs for Electroosmotic Pumping 38
Acknowledgement 40
References 41
Publications and Patents 42
Appendix I - Source and Types of Sludges Tested for
Efficacy of Electroosmotic Dewatering 43
Appendix II - Accelerated Dewatering Through Tailoring of
Surface Properties 44
Adjustment of Zeta Potential to Increase Rates of
Electroosmosis 44
Accelerating Electroosmosis Through Selected
Filtering Media
111
-------
FIGURES
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5«
Figure 6.
Figure 7«
Figure 8.
Figure 9.
Figure 10.
Figure 11.
Figure 12.
Figure 13.
Figure 14.
Figure 15.
Figure 16.
Figure 17.
Open-Top Plexiglass Test Cell 6
Equipotential Maps 7
Relation of Water Pumped to Power Consumed (Wet) 12
Relation of Water Pumped to Power Consumed (Dry) 13
Electroosmosis - JWPCP Raw Sewage 15
Electroosmosis - Aerobically Digested Sludge 16
Electroosmosis - Hyperion Raw and Raw/Activated
Sludge 17
Rinconada Plant Alum Sludge 1.4$ Suspend Solids 20
LaVerne Plant Alum Sludge 4.95$ Suspended Solids 21
Multiple Disc Electroosmotic Pump
Prototype Apparatus
Schematic of Prototype Electroosmotic Pump —
Pumping Energy as a Function of Power Level
Electrical Costs for Electroosmotic Dewatering
Electrophoretic Mobility Dependence on Ionic
Strength
Electroosmotic Membrane Filter Cell
Membrane Location in Composite System
23
25
26
31
39
46
54
55
IV
-------
TABLES
Table I. Effect of Anode-Cathode Distance on Electroosmosis 9
Table II. Effect of Current and Voltage on Electroosmosis 10
Table III. Electroosmotic Flowrates and Energy Requirements 18
Table IV. Results of Sludge Dewatering by Electroosmosis in
Prototype Sludge Dewatering Apparatus 28
Table V. Electroosmotic Dewatering of Preconcentrated
Sludges 30
Table VI, Anode Test Results for Hyperion Digested Sludge,
Constant Current, 19 ma/cm of Anode Surface -
Table VII.
Table VIII.
Table IX.
Table X.
Table XI.
Table XII.
Table XIII.
Table XIV.
Table XV.
Table XVI.
Table XVII.
Seven Hour Tests
Stannic Oxide Electrodes Pressed at 14,000 psi,
Sintered at 1400° C '
Stannic Oxide Compositions Pressed 14,000 psi,
Sintered at 1400° C, Sb 0, Atmosphere
Anode Life Tests Hyperion Digested Sludge,
Constant Current of 19 ma/cm2
Electrophoretic Mobility - pH
Electrophoretic Mobility - Conductivity
Surface Active Agents Tested
Electrophoretic Mobilities in Sludge Supernatant
Electrophoretic Mobilities in Water
Electrophoretic Mobilities in NH.HCO-^ Solution
(w 2600 ppm)
Electroosmotic Dewatering Data in Composite
Systems
Electroosmotic Dewatering Data for Nad
Solutions
— 33
— 34
— 35
— 37
— 45
— 45
— 48
— 50
— 50
— 51
.... 57
•— 59
-------
CONCLUSIONS
General conclusions that may be drawn from the results of the investi-
gation are as follows:
1. Electroosmosis can be applied to dewatering sewage treatment and
water treatment sludges of various types.
2. For efficient dewatering from very wet sludge (pfo) to burnable
solids, it is necessary to orient the pumping apparatus with the
cathode screen beneath the anode so that an initial sludge film
can be formed by gravity filtering at the cathode surface.
However, starting with pre-concentrated sludges that are less
liquid in character, the formation of an initial film is not
necessary.
3. Sludges cannot be thickened to a solids content exceeding 20$ if
the cake is beneath the liquid level of the sludge suspension.
For drying beyond the 20$ values, the cake must be brought out
of the liquid.
4. The operating costs for electroosmotic dewatering appear to be
promising for sludges with solids contents of 6$ or greater,
suggesting that the use of electroosmosis in tandem with a pre-
thickening process may be more attractive than using it alone.
5. Experiments with prototype apparatus indicated that electro-
osmotic pumping is slow. Average pumping rates of 3«5 ml/hr/cm
(0.8 gal/hr/ft^) of electrode surface were observed. Thus, large
electrode surface areas are required if pumping productivity is
to keep pace with a typical treatment plant's sludge production.
However, preconcentration of sludge suspensions can reduce the
electrode area requirements.
6. Stannic oxide ceramics are oxidation resistant and can be used
to advantage in anode service in electroosmotic pumping.
VI
-------
RECOMMENDATIONS
The major purpose of the investigation described herein was to
assess the feasibility of applying the phenomenon of electroosmosis
to dewatering sewage treatment sludges. It was not within the scope
of the program to attempt scale-up of apparatus to pilot plant sized
units. Sufficient information has been obtained to indicate that
electroosmosis holds promise as a sewage treatment unit process, par-
ticularly in view of the discovery that anodes for electroosmotic
pumping could be constructed from conductive stannic oxide ceramics,
thus reducing projected equipment costs. However, the practicality
of the process, in terms of operating economics and efficiency,
remains to be properly assessed. Such assessment can be made only
through continued study of the process in pilot plant sized equipment.
It is recommended that such a study be carried out.
VI1
-------
INTRODUCTION
The ultimate disposal of waste sludge is a problem of major propor-
tions with one of the more significant contributors being dewatering.
With convenient and economic dewatering techniques, the intrinsic
fuel value of the sludge can often be used to carry out final disposal
by incineration. Unfortunately, few sludges lend themselves to con-
venient or economic dewatering by conventional techniques, primarily
because these techniques are generally dependent upon the hydraulic
permeability or the settling characteristics of the sludge suspensions.
It is the objective of this report, therefore, to describe an investi-
gation of electroosmotic pumping, a sludge dewatering method with
promise for more general water removing capability.
Electroosmosis is a physical chemical phenomenon discovered over 150
years ago by Reuss (1809) who observed that if an electrical poten-
tial is applied to a porous diaphragm, water moves through the dia-
phragm capillaries toward the cathode; and that as soon as the
electric current is stopped, the flow of water stops. The basis for
the flow is the presence of a "double layer" at the walls of the
capillary. That is, a distinction can be made between free water and
a boundary film of water immediately adjacent to the capillary walls
that is thin relative to the capillary diameter. Surface chemists
call the boundary film the double layer because it is assumed that
there is a separation of electric charges on the two surfaces of the
film. One surface normally carries a charge that is rigidly attached
to the wall, and the other surface carries a diffuse distribution of
opposite charge that is movable. Thus, if an electric potential is
applied to the ends of the capillary, the diffuse charges can move
toward one pole. In doing so, they can drag along the water molecules
of the movable part of the film and the cylinder of free water that
the film encloses.
A similar phenomenon occurs in the liquid phase of suspensions of
small particles. When the movement of the solid phase is constrained,
water flows under the influence of an applied field because of the
existence of the double layer at the surfaces of the particles.
Furthermore, as will be seen, the velocity of the flow is dependent
only on the magnitude of the applied field and is essentially inde-
pendent of the state of compression of the suspension. It is the
latter property that makes electroosmosis a potentially attractive
sludge dewatering technique.
-------
It can be shown that the quantity of liquid moved in unit time by
electroosmosis in a single capillary is given by the equation^
EDr Z
qe - TT^
where E is the electrical potential, D the dielectric constant of the
liquid, r the radius of the capillary, Z the electrokinetic or zeta
potential, which is defined as the potential existing between the
rigid and the moval parts of the double layer, L the length of the
capillary between the electrodes, and v the viscosity of the liquid.
The electrokinetic potential Z is characteristic of a specific surface
and solution. Thus, variations in solution pH or ionic strength can
affect Z. For example, an increase in pH would tend to neutralize a
negatively charged surface and reduce Z^. However, for a given system,
one can consider the quantity DZ/4Tfv constant. Also, one can define
the potential gradient E/L = e , and T! r = a = cross sectional area
of the capillary. Multiplying and dividing Eq. 1 by'rr then, one
obtains
f - Ve - Ce « (2)
where V , the flow velocity, is seen to be independent of the capil-
lary size for a fixed value of Z.
In contrast to electroosmotic flow, ordinary hydraulic flow is depen-
dent on capillary diameter. That is, since flow in a hydraulic filter
is almost always laminar, it can be described by the Hagen-Poiseuille
equation (McCabe, 1956). By analogy to the example above then, the
velocity in a long capillary is given by
Vh = Ch $ a (3)
where $ is the hydraulic pressure gradient and is analogous to the
electric gradient e. C, is a constant equal to g/8rr and a, again, is
the cross section of the capillary. Thus, for laminar hydraulic flow,
the capillary cross section appears in the flow velocity equation
whereas it does not in the equations for electroosmotic flow.
Derivation of Eq. 1, which can be found in any standard text on
Physical Chemistry (Glasstone, 1952) is based on the assumption that
the capillary diameter is large relative to the thickness of the
double layer.
2
See Appendix II.
-------
The significance of capillary size independence is illustrated by
consideration of the nature of a sewage sludge cake which, in general,
behaves as a compressible filter medium, i.e. , the cross sectional
area of the capillaries within the cake decreases under filtration
pressure, and hydraulic flow through the cake diminishes in time.
Thus, the ultimate hydraulic dewatering capability is limited by the
driving pressure and by the physical nature of the sludge. However,
because electroosmotic flow velocity is independent of cross section,
flow remains essentially constant regardless of the compressed state
of the cake. No filter-aid chemicals are required, and electroosmosis
will occur, in principle, even in gelled cakes that ordinarily might
be impossible to handle with pressure filtration.
A search of the literature has yielded two attempted applications of
electroosmosis to waste sludge treatment. In 19^3 > Beaudoin (19^3)
reported some success in decreasing water content in flocculated
activated sludges by combining electroosmosis with rotary filtration.
Somewhat later, electroosmosis was investigated (Cooling, 1952) as a
method for accelerating water drainage from sewage sludge drying beds.
As a result of the 1952 effort, sludge solids were increased from
about 5 to 60$ over a period of 21 days. With air drying and gravity
drainage, sludge beds serving as controls dried to the extent of only
21$ solids in the same time period. However, in the configurations
of both investigations, i.e., as a technique supplementary to a
prime dewatering method, electroosmosis appeared uneconomical. Thus,
it was the purpose of the program reported here to investigate the
phenomenon as an independent unit process in the treatment of sewage.
-------
EXPERIMENTAL PROGRAM
The laboratory program that has been underway has been directed toward
several objectives! First, an investigation of electroosmotic flow
patterns in sewage sludge suspensions was carried out to determine
equipment configurations for efficient solids removal. Secondly,
efforts were made to identify parameters critical to efficient
electroosmotic pumping under varying conditions of applied voltage
and current flow. In addition, two prototype electroosmotic pumps
were designed and tested, and a search for economical oxidation
resistant electrodes was carried out.
As supplementary objectives, accelerated dewatering by adjustment of
the surface properties of sludge suspensions with additions of surface
active agents was investigated, and attempts were made to achieve more
rapid dewatering by means of filtering media through which fast electro-
osmotic flow might be effected. However, little success was achieved
in increasing flowrates with these methods, and the supplementary
aspects of the program were not pursued^.
In the following discussion, the results of the main experimental
program are presented. As will be seen, the technique selected for
removing solids from electroosmotically dewatered suspensions was
critical for efficient pumping operation. In addition, although
elevated driving voltages resulted in accelerated water flow, they
decreased the electrical efficiency of electroosmosis. Furthermore,
the energy requirements for dewatering were found to depend on the
amount of bulk water in the sludge, all sludges behaving similarily
in energy consumption up to a specific solids level apparently
characteristic of the sludge type. Of two prototype electro-
osmotic pumps tested, one in the configuration of an inclined plane
was able to efficiently dewater waste sludges to solids levels that
would auto-incinerate. And,finally, two relatively inexpensive
materials, magnetite and stannic oxide, were found to be useful for
anode application.
SLUDGE COLLECTION AND WATER FLOW PATTERNS
Determination of electroosmotic water flow patterns allowed an
assessment to be made of the most efficient method for removing solids
from sludge suspensions. The technique for obtaining flow patterns
was based on Eq. 1, which indicates that flow velocity is proportional
to applied voltage. Thus, flux pattern analysis could be achieved by
measuring potential differences between a fixed anode and a probe
A discussion of these experiments is presented in Appendix II.
-------
electrode as a function of probe position in an experimental cell.
From such measurements, lines of equipotential could be constructed
for varying cell or electrode geometries, and potential gradients
could be determined throughout. Since, from Eq. 1, water flux in any
area of the cell would be directly proportional to the gradient of
the potential in that area, a semi-quantitative assessment of flux
patterns could be extracted from the equipotential lines. The
direction of water flow would, of course, depend on the sign of the
zeta potential. However, for all the suspensions used in this study,
water was found to move toward the cathode, indicating that the
solids carried negative surface charges.
An open top plexiglass cell of approximately 500 cc capacity was
used for pattern analysis (Fig. l). The cell is composed of two
chambers separated by a fine mesh stainless steel screen cathode.
One chamber of the cell served to collect effluent from the other
chamber in which the sludge to be dewatered was placed. The open top
of the cell allowed the position and geometry of an anode to be varied
with respect to the screen cathode. Platinum, platinized-titanium,
and graphite were used as anode materials to minimize corrosion from
anodic oxidation. Potential measurements were made with a platinum
tipped probe electrode.
Typical equipotential maps are illustrated in Fig. 2. The maps
shown are for samples of anaerobically digested sludge from the Los
Angeles Hyperion Treatment Plant. However, they are characteristic
of all the sludge suspensions examined. The equipotential lines are
shown as if one were looking down into the cell at a plane located
at one-half the total sludge depth. Variations in position, geometry,
or the nature of the anode used did not significantly change the
overall patterns that are illustrated.
Comparison of Fig. 2A, 2B, and 2C shows the most dramatic effect to
be a progressive increase in potential gradient near the cathode and
a decreasing gradient near the anode with continued electroosmotic
pumping. This was found to be generally associated with the formation
of an increasingly thick, dense cathode cake (20 wt. % solids ) No
significant anode cake formed, despite the electrophoretic drift
toward the anode that one would have expected for negatively charged
solids.
Accumulation of negatively charged sludge solids at the cathode
could result only from mechanical drag by water moving toward the
cathode. Apparently, a sludge cake, formed at the cathode surface
when the cell is filled, results in an initial elevated potential
The cake was also self-filtering, yielding relatively clear effluent
with a very low level of suspended fine solids.
5
-------
3
5DZ41-8/8/68-C1
Figure 1. Open-Top Plexiglass Test Cell
-------
CATHODE 15 VOLTS AT 0.3 HOURS, 0.15 AMP. ANODE 0 VOLTS
CATHODE 15 VOLTS
AT 2.0 HOURS, 0.125 AMP.
ANODE 0 VOLTS
t
B
CATHODE 15 VOLTS
AT 21 HOURS, 0.057 AMP.
ANODE 0 VOLTS
\
2*8
2.3
Figure 2. Equipotential Maps
-------
gradient relative to the bulk of the sludge and a higher electroosmotic
flowrate in the neighborhood of the cathode. Water in the bulk sus-
pension is then forced to move with the solids by gravity to replace
that which has been electroosmotically pumped through the cathode
screen and, in so doing, continually builds a thicker, denser cake.
In supplementary experiments, it was determined that the presence of
a thin sludge cake at the cathode surface was actually required for
electroosmotic water flow to start. In general, such an initial
cake was formed by gravity filtration through the clean cathode
screen at the outset of an experiment. With the formation of the
initial cake, the sludge solids accumulated efficiently at the
cathode screen.
That no significant solids accumulation occurred near the anode
probably resulted because of the low potential gradient in its
vicinity and because of partial neutralization of the zeta potential
by hydrogen ions formed at the anode. Measurement of pH indicated
that the area near the anode became acidic (pH about 2) with con-
tinued electroosmotic pumping^.
Based on the results obtained from the flow analysis experiments
and the location of the resultant sludge cakes, it was concluded
that the optimum configuration for an electroosmotic pump was one in
which solids were collected from the neighborhood of the cathode.
Furthermore, it was concluded that electrophoretic deposition of
solids at the anode could not lead to efficient solids collection
in a practical electroosmotic dewatering system.
INTERELECTRODE DISTANCE, VOLTAGE, CURRENT, AND POWER
To assess the effects of variation in electrical parameters, a
series of experiments was conducted in which interelectrode distance
was changed, and electroosmosis measurements were carried out under
conditions of controlled voltage, current, and power dissipation.
Interelectrode Distance
The effect of varying interelectrode spacing in electroosmotic
pumping is illustrated by the data in Table I. It can be seen that
the energy consumed per unit volume of water pumped is directly
proportional to the interelectrode distance. Apparently, ohmic
losses in the bulk of the sludge increase with increasing electrode
spacing resulting in a loss of electrical efficiency. Thus, in a
practical electroosmotic pump, the minimum interelectrode distance,
c
The effluent pH was elevated and ranged around pH 11.
8
-------
in keeping with reasonable sludge flow between the electrodes, would
yield highest efficiency.
TABLE I. EFFECT OF ANODE-CATHODE DISTANCE ON ELECTEOOSMOSIS
(5-cm Diameter Electrodes, Hyperion Digested Sludge)
Distance
Anode-
Cathode ,
cm
12.0
5-0
1.7
Applied
Voltage j
volts
15-8
12.1
7-7
Initial
Current ,
amperes
0.30
0.30
0.24
Final
Current,
amperes
0.18
0.12
0.09
1/4 hr
0.13
0.090
0.06?
3.0 hr
0.13
0.082
0.058
5.0 hr
0.11
0.070
0.041
Total
Water
Pumped ,
ml
149
143
100
Controlled Voltage and Current
Typical changes in pumping energy requirements as a result of varying
electrical operating parameters are illustrated in Table II. Experi-
ments A and B were carried out at constant voltage. In experiment C,
current was held constant. The applied voltage in experiment B was
selected to yield approximately twice the initial voltage gradient
across the sludge suspension than in experiment A. It is to be
noted that nearly twice as much water was pumped in experiment B as
in A, in accord with Eq. 1, i.e., electroosmotic flow velocity being
proportional to applied voltage.
Despite the large differences in voltage and current conditions, the
average pumping efficiency for experiments B and C is seen to be
similar, and nearly the same quantity of water was pumped. Apparently,
pumping efficiency is independent of the operational mode. Either
controlled voltage or controlled current sources may be used.
As can be seen by re-examination of the data for experiments A and
B, however, the efficiency of electroosmotic pumping will depend on
the level of applied voltage or the current control point selected
because part of the energy consumed is used in electrode reactions
and in resistive heat losses in the system. Thus, the higher the
voltage-current levels, the greater will be losses not associated
with moving water.
CONSTANT POWER OPERATION
Because pumping efficiency appeared to be independent of operational
mode, investigation of electrical operating parameters was continued
under conditions of constant power. Power was controlled with a
-------
TABLE II
EFFECT OF CtJREENT AND VOLTAGE ON ELECTROOSMOSIS
(Hyperion Digested Sludge, 2.1 cm Interelectrode Distance,
5~cm Diameter Electrodes)
Exp.
No.
A
B
C
Applied
Voltage ,
volts
9.4
16.8
10.0-
20.0
Initial
Current ,
amperes
0.29
0.61
0.30
Final
Current ,
amperes
0.09
0.18
0.25
Pumping Energy Expended ,
watt-hr/ml
1/4 hr
0.077
0.163
0.075
3.0 hr
0.074
0.094
0.12
5.0 hr
0.046
0.084
0.11
Average
0.065
0.10
0.10
Weight
Dry,
grams
1.56
2.61
2.63
Percent
Solids
18.6
17.6
19.3
Total
Water
Pumped
ml
124
230
221
-------
device in which current and voltage values were electronically multi-
plied and the product signal used to control the voltage output of
the electroosmosis power supply. Constant power experiments were
carried out in new test cells fabricated with a fixed cathode as a
bottom surface and a movable anode oriented above the cathode ("over-
under cells"). The anode was attached to a servo-device that main-
tained it in continuous contact with the sludge cake surface regard-
less of the height of the cake above the cathode.
The purpose of the new test cell configuration was two-fold; first,
the over-under configuration provided a convenient way to establish
the initial thin sludge cake required to start electroosmotic pump-
ing and second, the servo-device maintained the anode in contact with
the sludge cake regardless of the amount of shrinkage the cake might
suffer as a result of dewatering.
Sewage Sludges
With these cells, and starting with well-digested anaerobic sludges
of about 2-jfe solids, sludge cake accumulated at the cathode surface
beneath the liquid level of the- sludge suspension with maximum achiev-
able solids content of 2Qfc. Apparently, a steady-state flow of water
from the liquid portion of the sludge suspension through the accumu-
lated cake was established. If the liquid suspension above the cake
was removed, however, the solids content of the remaining cake could
be increased beyond the 2Qfo value. It was also determined that solids
accumulation was proportional to the amount of water pumped from the
cell, and that the water pumped from the digested sludge was directly
proportional to the power consumption up to about JOfo solids. Above
JOfo solids, the power requirements per unit volume of water removed
increased.
The results are illustrated graphically in Fig. 3 and 4. Figure 3
represents results for a cathode cake immersed beneath a liquid
sludge suspension. Figure 4 represents results for a sludge cake re-
moved from beneath liquid sludge. Although the pumping efficiencies
illustrated in the two graphs are somewhat different, the accumulated
water removal is seen to be essentially linearly dependent on power
consumption up to a level near 30$ where the power required to remove
water increases. The increase in power consumption has been attri-
buted to a transition between bulk and bound water in the sludge, the
former being removed in the early linear portion of the dewatering
process and the latter occurring later at a higher energy cost. An-
other possibility, however, may be that compression of the filter
cake has reduced the pore diameter to a value low enough to preclude
the applicability of Zeta potential theory described in the introduction.
The difference in pumping efficiencies arises because different
applied voltages were used for the two measurements.
11
-------
20
40 60 80 100 120 li(0 160 180 220 220 2^0260 280 300 320 3^0 360
ELAPSED TIME, MINUTES AT 12 WATTS
Figure 3. Relation of Water Pumped to Power Consumed (Wet)
12
-------
16
\k
12
10
SLUDGE SOLIDS CONTENT AS
INDICATED
20 *0 60 80 100 120
ELAPSED TIKE. HINUTES, I WATT LEVEL
140
Figure 4. Relation of Water Pumped to Power Consumed (Dry)
13
-------
The pattern in Fig. 4 for anaerobically treated sludge is seen, in
Fig. r), 6, and 7 to be repeated with several other sludges. However,
the point at which the dewatering efficiency diminishes falls at some-
what different water content levels for each of the sludges, and the
transition does not appear to be as sharp as that obtained for the
anaerobically digested type. In Fig. 7, the transition point appar-
ently had not yet been reached. These differences are believed to
be caused by differences in the water binding properties of the various
sludges.
In Table III, results of measurements of a series of sludges from
several different sources are summarized. Included in the table are
values for flowrates in ml/hr/cm of electrode area and efficiencies
in power consumed per unit volume of water removed for both the early
linear portion of the dewatering process (the "bulk water" part) and
the later, less efficient, portion of the water removal process (the
"bound water" part). It is interesting to note that the electro-
osmotic flowrates for all the sludges are of about the same order of
magnitude during the earlier portion of the process. However, after
the transition, significant differences are observed in electroosmotic
flowrates among the various sludges. Pumping energy requirements are
also seen to be about the same, the widest variation being about a
factor of two in the early part of the cycle and a factor of four at
the end. Specific filter cake resistances for vacuum filtration,
measured as recommended by Adrian, et al. (1968), for each of the
sludges are included in the table for an indication of the difficulty
one might experience in filtering these materials by conventional
methods.
Perhaps the most important conclusion to be drawn from the data in
Table III is that electroosmotic dewatering occurs in all the sewage
sludges regardless of their processing or geographical source, i.e.,
successful dewatering was achieved with primary sludges, mixed acti-
vated-primary sludges, aerobic or anaerobically digested sludges, and
from locations that represented extremes from industrialized areas
to areas that were primarily residential. Thus, one would expect
that dewatering could be accomplished by electroosmosis with any
sewage sludge. It is also important to note that the flowrates and
energy costs per unit volume of water removed appear to fall in about
the same range for all of the materials tested, the variability being
about a factor of two or three. It is expected, in view of the di-
versity of the sludges tested, that the same range of flowrates and
energy costs per unit volume of water removed would be found for any
other sludge selected for electroosmotic dewatering. Thus, it should
be reasonable to extrapolate any cost estimates based on a single
sludge type to any other sludge with the qualification that, ulti-
mately, the total cost of dewatering will depend on the particular
sludge's initial solids content and its fuel value.
14
-------
20 30 40 50
ELAPSED TIME, «NOTES, 10.5 WATTS
Figure 5. Electroosmosis—J¥PCP Raw Sludge
15
-------
i!
I
SOLIDS CONTENT AS INDICATED
I I I I I I I I
°0 20 do 60 80 100 120 \kO 160 180 ZOO 220 21.0
TIHE, MINUTES. 4.5 WATTS
Figure 6. Electroosmosis--Aerobically Digested Sludge
16
-------
SOLID CONTENT
INITIAL FINAL
O 100* 0*
D 90* 10*
A 70* 30*
20 'tO
ELAPSED TIME, MINUTES
Figure 7. Electroosmosis—Hyperion Raw and Raw/Activated Sludge
17
-------
TABLE III
ELECTROOSMOTIC FLOWRATES & ENERGY REQUIREMENTS
Sludge 'd'
Digested Hyperion
Raw Hyperion
I c )
Raw Hyperion
Raw JWPCP
Raw/Activated
9:1 Hyperion(c)
Raw/Act i vat ed
7:3 Hyperion(c)
Aerobic Filter
Cake, Tapia
Aerobic Digested,
Tapia
Activated, Tapia
c\
Electroosmotic Flowrates, ml/hr/ctn
Early Portion/ \_
Rate
2.2
1.4
1.4
1.4
1.4
1.3
1.7
1.4
1.2
% Solidsva/
33
23
27
21
26
27
22
17
16
Later Portion/, \
Rate
0.09
0.51
0.14
_
-
0.56
0.13
0.07
% Solidsvu/
42
25
2k
_
-
34
35
26
Avg.
Rate
0.79
1-3
1.4
0.84
—
_
-
0.82
0.82
Pumping Energy
watt-hr/ml
Early
Portion
0.060
0.14
0.13
0.15
0.14
0.14
0.12
0.063
0.066
Later
Portion
1.5
0.38
1.5
_
—
0.36
0.65
1.5
Avg.
0.16
0.15
0.13
0.24
0.14
0.14
0.16
0.10
0.10
Specific Cake
Resistance
o /
sec'Vgram
x 10-10
1.3
1.3
1.7
0.36
2.5
2.3
-
9.5
0.73
00
(a)
00
(c)
(d)
Before transition
After transition
Stopped before transition reached
A list of sources and sludge types used in this study is presented in Appendix I
-------
Water Treatment Sludges
Because the disposal of alum sludges from reactor clarifiers in water
treatment plants is a problem as severe as disposal of sewage sludges,
it was appropriate to test the efficacy of electroosmotic dewatering
for application to water treatment sludges. Samples of sludge were
obtained from the Rinconada Water Treatment Plant near San Jose,
California, and from the Metropolitan Water District Treatment Plant
at LaVerne, California. Sludge from the Rinconada Plant is con-
ventional reactor clarifier underflow. The sludge from the LaVerne
Plant, however, results from the precipitation of backwashings from
sand filters. The vacuum filtration characteristics of the two
sludges are reported to be significantly different.'
Typical electroosmosis data are presented in Figs. 8 and 9. It is
noteworthy that both sludges exhibit water removal characteristics
similar to those of the sewage sludges, i.e., a nearly linear water
removal-energy consumption relationship at the early stages of de-
watering and a sharp increase in energy requirements in later stages
of dewatering. Both sludges also exhibit lower energy expenditure
per volume of water removed in the early stages of dewatering than
do sewage sludges. The point at which the water becomes more diffi-
cult to remove, however, appears at a much lower solids level for
the Rinconada sludges than for the LaVerne sludge, correlating with
the information that the former is more difficult to vacuum filter
than the latter.7
The quality factor for the disposal of water treatment sludges is
the mechanical state of the dewatered solids, i.e., whether or not
the resulting dewatered solids lend themselves to removal by shovel-
ing and trucking. At the 10$ solids level, the water treatment
sludges are mud-like in consistency and could not be easily handled
by shoveling. At the 19$ level, however, the material is a thick,
highly viscous paste which could be shoveled conveniently. It is
apparent, then, based on the curves shown in Figs. 8 and 9, that
the Rinconada sludge would be more expensive to dewater by electro-
osmosis than the back-wash sludge obtained from the LaVerne Plant.
PROTOTYPE APPARATUS
The experiments and results summarized in the foregoing discussion
suggested several different configurations for an electroosmotic
pump. However, time permitted investigation of only two configur-
ations. These included a multiple disc device and an inclined plane
device. Both devices incorporated the necessary provisions for
carrying accumulated filter cake out from beneath the sludge sus-
pension for final drying, and for the formation by gravity of an in-
itial thin filter cake at the cathode surface before electroosmosis
^Personal communication from the technical staff of the
Eiraco Corporation, Salt Lake City, Utah.
19
-------
_ 120
o
o
UJ
QC
ENERGY CONSUMPTION - 0.015
20
10 15 20 25
TIME IN MINUTES AT 10 WATTS
Figure 8. Rlnconada Plant Alum Sludge
\.k% Suspend Solids
20
-------
17 PERCENT
SOLIDS
ENERGY CONSUMPTION - 0.033
10
20 30 40 50
TIME IN MINUTES AT 10 WATTS
Figure 9. taVerne Plant Alum Sludge
4.95% Suspended Solids
21
-------
started. The multiple disc device provided a configuration with a
large amount of electrode surface area, and, thus, a high electro-
osmotic volume flow could be incorporated into a small volume. The
inclined plane device, as will be seen, although lover in possible
electrode surface to apparatus volume ratio, provided a more con-
venient way to remove sludge cake from beneath the sludge suspension
for final drying. Both configurations lent themselves to eventual
adaptation to automatic and continuous operation. Of the two, the
inclined plane device gave promise for application in an operational
sewage treatment plant.
Multiple Pise Electroosmotic Pump
The model of the multiple disc electroosmotic pump is illustrated in
Fig. 10. The device consists of a pair of hollow disc chambers
faced with 100 mesh stainless steel screen cathodes. One disc is
seen with one face removed to display the interior portion. Each
chamber contains a siphon tube through which water, pumped from the
sludge suspension from the outside of the chambers to the inside,
can be removed. The anodes, shown removed from the model in the
figure, are situated alternately between each of the chambers, fac-
ing the cathode screens.
Initial sludge cake was formed by rotating the discs through the
sludge suspension while drawing vacuum through the siphon tubes.
Following the formation of the initial sludge cake, electroosmosis
was started. Electroosmosis was carried out at power densities of
0.25 watts/cm2 of immersed cathode area (about kj® em^). Continuous
rotation of the discs allowed sludge accumulating on the faces of
the discs to be carried out from beneath the sludge suspension for
final drying by warm air circulation. The dried cake could then be
scraped from the cathode surfaces.
Several serious problems were encountered in operation of the model
multiple disc electroosmotic pump. Although solids accumulated
efficiently on the cathode surfaces, they tended to adhere badly,
falling off and back into the sludge suspension before final drying
could be achieved. Warm air drying resulted in the evaporation of
water from the cathode surfaces with concurrent formation of carbon-
ate scale. The scale then inhibited water transport through the
cathode screens. In addition, the effluent water always contained
a high concentration of solids.
Very high energy was required to move water, in comparison to ex-
perience with the simple electroosmosis cells described in the pre-
vious section of this report. For example, in the simple cells,
electrical consumption ranged between 0.05 and 0.15 watt-hr/ml of
water pumped, in contrast to values ranging between 0.25 to 1.0
found for the disc device.
22
-------
a
5AJ23-8/2976
-------
Because of the difficulties encountered in operation of the model
multiple disc elec troosmotic pump, effort on it, was terminated and
redirected to the inclined plane device which showed considerably
more promise in preliminary testing.
Inclined Plane Electroosmotic Pump
A photograph ani\ drawing of the apparatus are shown in Figs. 11 and
12. The device was designed to dewater sludge from levels of about,
2fc solids to over JOfo. As was pointed out earlier, pumping is di-
rectly related to power consumption. Thus, the constant power supply
described before was used as the energy source for the unit. As can
be seen, the cathode screen is oriented as an inclined plane. The
screen is 50-mesh stainless steel. A platini/ed titanium anode and
scraper assembly are carried above the cathode screen by a carriage
which can move the assembly up the incline. As the anode assembly
moves, sludge is accumulated between the electrodes and thickened to
a level of about 15$ solids. The accumulated cake is finally carried
above the liquid level of the sludge suspension where it is dried to
a solids content of about JOfo and collected at the terminus of the
incline. Liquid is collected in the trays beneath the incline. In
a continuous operation, a group of anode carriages would follow one
another up the incline sequentially. After dropping the solid cake
at the terminus of the incline, each anode assembly would be returned
to the start by conveyor and dropped again into the sludge suspension.
Because the sludge cake shrinks as it dries, it was necessary that
the anode follow the decrease in cake dimensions to maintain elec-
trical contact. Therefore, the motor controller described earlier
was used to activate the anode assembly and lower the anode to main-
tain proper electrical contact. As will be seen, a great deal of
success was achieved with the Inclined Plane device.
DEWATERING IN THE INCLINED PLANE APPARATUS
Dewatering tests were carried out in the prototype apparatus with
several different sludges including digested and activated sludges
from the Los Angeles Hyperion Treatment Plant and digested sludge
from the Los Angeles County Treatment Plant. The digested sludges
ranged between 2 and 3$ in solids content and were virtually im-
possible to filter or settle. The activated sludge was roughly 0.6$
in solids content and settled and filtered significantly more easily
than the digested sludge. Successful dewatering tests were also
carried out with sludges pre-concentrated to high solids levels.
24
-------
Figure 11. Prototype Apparatus
-------
TO ELECTRICAL
CONTROL BOX
MOTOR
INTERNAL
THREADS
FIG.
SCRAPER AND ELECTRODE
DETAIL
SCRAPER^
MOTOR AND ANODE
ROD SUPPORT
SUPPORT ROD
ANODE
CATHODE SCREEN
TEFLON SCRAPER
FIG.
ANODE LEVEL CONTROLLER
AND CARRIAGE
ANODE SUPPORT ROD
CARRIAGE
SCRAPER
SUPPORT
TUBE
DRIVE
MOTOR
SLUDGE
RESERVOIR
FIG.
ASSEMBLED
APPARATUS
DRIVE
SCREW
SCREEN
CATHODE
EFFLUENT
RECEIVER
SCRAPER
Figure 12. Schematic of Prototype Electroosmotic Pump
26
-------
Wet Sludges
Typical results with wet sludges, reported in Table IV, indicate
higher total water flow than those achieved in the over-under cell
discussed earlier. The higher water flowrates appear to result from
a very large cathode screen area in this apparatus compared to the
actual active cathode area which lies directly under the anode (50
cm^). Thus, gravity filtration, which is normally a small percentage
of electroosmotic flow, became more significant.
Values for gravity flow were estimated by measurement of water flow
at zero voltage in the apparatus, and the corrected values for elec-
troosmotic flow are presented in column 6 of Table IV. In an actual
installation with this type of apparatus, many anodes would be used.
Thus, the ratio of active cathode area to inactive area would be
large, and gravity flow would be of minor significance.
As can be seen, electroosmotic pumping yielded sludge cakes of over
30^ solids with the digested sludges and cakes of about 15^ with the
activated sludge. No explanation is yet available for the difficulty
in dewatering the activated sludge, which was also observed in the
over-under cells discussed earlier. It may be that a considerably
greater concentration of bound water is present in activated sludge.
It should be noted that calculations of fuel values indicate that
the solids levels of the digested sludges are sufficiently high to
support auto-incineration, and that it is unlikely that activated
sludge itself would ever be subjected to incineration without blend-
ing with primary sludge.
Except for the activated sludge, the pumping efficiencies observed
for measurements in the prototype apparatus are comparable to those
obtained in the small over-under cells discussed before. Activated
sludges were found to be fairly easy to gravity filter and tended
to pass through the 50-mesh screen cathode of the prototype pump
very easily. Thus, the value for the activated sludge is thought
to be in error as a result of fast gravity filtration through the
inactive part of the cathode (note that the correction for gravity
filtration is nearly 50^ of the total flow).
It is thought that this consistency in pumping efficiency values for
electroosmotic pumping, in both this apparatus and in the smaller
over-under cells, is an indication that order of magnitude improve-
ments in energy costs for electroosmotic pumping will probably not
be attainable.
2?
-------
TABLE IV
RESULTS OF SLUDGE DEWATERING BY ELECTROOSMOSIS
IN PROTOTYPE SLUDGE DEWATERING APPARATUS
(Hyperion Sludges, Anode Area 50 cm , Starting Anode to Cathode Spacing, 2.0 cm)
Sludge Source
Digested (Hyperion)
Digested (L.A. County)
Digested^ ' (Hyperion)
Activated (Hyperion)
Initial
Suspended
Solids
Content
2.7
3.0
2.8
2.0^
Sludge
Cake Dry
Solids
Percent
32
33
29
16
Power
Level ,
watts
12
12
12
5
Total
Vater Flow-
Rate (total),
ml/hr
195
330
139
470
(c)
Corrected
Vater Flowrate
Electroosmosis,
ml/hr
175
185
90
250
Pumping
Energy
watt hrs
ml
0.0?
0.07
0.14
0.025
Anode
Assembly
Speed,
cm/hr
10.7
13.2
10.7
10.7
to
00
(a)
(b)
(c)
A 100 mesh screen was placed over the 50-mesh cathode screen normally used.
Settled from 0.6^ sludge.
See Text.
-------
Pre-concentrated Sludges
As will be discussed in a later section of this report, significant
cost advantages in operating an electroostnotic pump can be gained if
the process is used with sludges that are relatively high in solids
content. Because of these advantages, a series of experiments was
conducted to determine if any of the important operating parameters
of electroosmosis in the inclined plane apparatus were seriously
affected when the process was carried out on pre-concentrated sludges.
Tests were conducted with digested sludges, raw sludges, and mixtures
of raw and activated sludge.
Digested sludge was pre-concentrated in hatches hy electroosmosis in
a large, high-current-capacity cell. Raw and activated sludges were
pre-concentrated by gravity filtration.
The results of the series are presented in Table V. Comparison with
the data shown in Tables I, II, III, and IV shows that energy expendi-
tures similar to those required for wet sludges are required for de-
watering the pre-concentrated sludges. It is also noteworthy that
pre-concentrated sludges could be successfully dewatered without the
formation of an initial sludge film on the cathode screen surface.
As with the wet sludges, energy expenditure per unit volume of water
removed increased with increased power input to the pumping apparatus.
The dependence is illustrated clearly in Fig. 13. As before, the
increased energy used at higher power levels is attributed to in-
creased heating in the sludge cake and to reactions underway at the
electrode surfaces.
Based on the data of Table V, one can conclude that electroosmosis
is capable of dewatering pre-concentrated sludges with an electrical
efficiency similar to that obtained with very wet sludges.
ELECTRODE MATERIALS
It was pointed out in the foregoing discussion that the anodes used
in investigating electroosmotic pumping have been, in general, made
of either platinum or platinized titanium. The anodic reaction under-
way during electroosmosis is strongly oxidizing, and the anode must
be oxidation resistant. Platinum is, of course, essentially inert
to oxidation. Platinized titanium, developed for anodic service in
the electrochemical manufacture of chlorine, is similarly resistant
to anodic oxidation. However, both materials are too expensive for
service in electroosmotic dewatering applications. Platinized
titanium, for example, costs approximately $120/ft2. In order to
make electroosmosis economically more attractive as a unit process
in sewage treatment, more reasonably priced materials were required.
Thus, a testing program was carried out in which appropriate materials
were sought.
29
-------
TABLE V
ELECTROOSMOTIC DEWATERING OF PRECONCENTRATED SLUDGES
Initial
Solids
Percent
13.3
13.3
13.3
13-3
13.3
13.3
15.6
15.6
9.1
9.1
9.1
, 9.1
9.1
10.4
10. 4
10.4
10.4
10.3
10.3
10.3
10.3
10.3
10.3
10.3
10.3
10.3
8.7
12.7
12.7
12.7
10.5
10.5
10.5
9.9
12.5
13.2
13.3
13.4
13.3
13.3
11.8
Sludge Type
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Digested
Raw
Raw
Raw
Raw
50/50 Raw-Act.
50/50 Raw- Act.
50/50 Raw-Act.
50/50 Raw-Act.
50/50 Raw-Act.
50/50 Raw- Act.
50/50 Raw-Act.
50/50 Raw-Act.
50/50 Raw-Act.
50/50 Raw- Act.
50/50 Raw-Act.
Power
Level ,
Watts
11.0
11.0
10.6
10.9
10.6
10.6
10.1
21.4
10.1
10.3
10.5
21.8
23.0
23.0
19.9
42.0
85.0
9.6
31.2
87.0
70.0
43.0
19.1
10.2
26.0
19.0
9.8
9.9
9.8
19.9
9.7
20.3
43.7
9.8
21.8
9.9
9.6
43.6
9.6
9.7
5.3
Elapsed
Time,
Minutes
38
46
44
50
50
50
60
38
50
60
70
37
35
37
35
20
12
65
29
14
15
20
29
52
30
35
31
40
45
30
40
30
15
30
15.75
40
35
12.28
18.92
35
44
Final
Solids
Percent
24.3
25.8
26.6
28.4
26.3
24.6
27.7
29.1
26.3
29.0
29.5
30.1
31.1
29.7
28.0
28.1
30.3
27.5
28.9
27. 2
31.5
27.9
29.6
27.5
27.6
25.9
25.1
26.7
26.3
27.5
30.0
25.3
25.6
25.2
25.1
28.8
26.3
26.2
24.5
24.7
24.8
Pumping
Energy
Watt-hr/ml
0.11
0.12
0.10
0.12
0.10
0.12
0.15
0.17
0.10
0.10
0.12
0.10
0.10
0.13
0.10
0.13
0.16
0.11
0.15
0.20
0.18
0.16
0.086
0.12
0.11
0.11
0.09
0.09
0.09
0.12
0.06
0.13
0.13
0.05
0.09
0.10
0.11
0.15
0.08
0.10
0.07
30
-------
s
cc.
a
o.
0 10 20 30 40 50 60 70 80 90 100
POWER LEVEL, WATTS
Figure 13. Pumping Energy as a Function of Power Level
31
-------
Candidate anode materials were screened by subjecting them to electro-
chemical oxidation in digested sludge obtained from the Los Angeles
Hyperion Plant. Promising materials were then selected for continued
life tests.
To carry out the screening and life tests, a cell was constructed
which allowed six sample's to be tested in parallel and simultaneously
within a single sludge reservoir. Each of the candidate materials
was tested under conditions of constant current, and all the samples
were tested at the same current density.
Table VI shows the results of screening tests for a variety of candi-
date materials. Each of the samples was subjected to a current
density of 19 ma/cm2 of surface for seven hours, except where high
anodic polarization may have forced termination earlier.
As Table VI illustrates, the most promising materials for anode
application, compared to platinized titanium, were magnetite and
stannic oxide. The graphites and carbons all displayed relatively
high anodic corrosion, and the other metals, including the flame
sprayed refractories, were easily polarized, resulting in excessively
high voltage drops at the anode surface. The metals and refractories
apparently suffer relatively high anodic corrosion too, since the
weight losses shown in Table VI for those materials resulted from
short term (two to four hours) tests, compared to seven hours for
the other materials.
Because of the promise displayed by stannic oxide, continued investi-
gation of the material for anode application was carried out. Typical
of most ceramics, stannic oxide varies in its properties depending
upon the method selected for its preparation. Thus, the variation
of type and concentration of doping materials will result in samples
of variable density and electrical resistivity. Accordingly, a
series of stannic oxide recipes containing doping materials known to
be active sintering agents were prepared. Sample discs were pressed
at 14,000 psi and sintered at 1350-1400 C for five hours. The pro-
perties of the resulting ceramic discs were then evaluated for appli-
cation in anodic service. Characteristics sought included low elec-
trical resistivity, good structural integrity, i.e., high density,
and good oxidation resistance.
Some typical results for duplicate samples are illustrated in Table
VII. As can be seen, the formulations containing Sb2®j and ZnO dis-
play the lowest electrical resistivity and highest density combin-
ation. In continued testing, it was determined that sintering SnOp
in an atmosphere of Sb20-T consistently yielded samples with resisti-
vities comparable to that of Sample IV. However, the presence of an
additional agent like ZnO was required to obtain good structural in-
tegrity. Several other agents were found to serve as sintering aids,
as is illustrated in Table VIII.
32
-------
TABLE VI
ANODE TEST RESULTS
FOR HYPERION DIGESTED SLUDGES, CONSTANT CURRENT,
19 ma/cm OF ANODE SURFACE-SEVEN HOUR TESTS
Anode Loss
Sample Description
Grams/
Faraday
Moles/
Faraday
Remarks
Graphites
Speer 8826
Pocco AXF-5Q
Pocco AXF-5Q
Pure 4800
UCC AGLX-58
Rocketdyne Carbodyne
Carbodyne-Petroleum Coke
GE Pyrolytic
GE Pyrolytic
Graphite Cloth
UCC Karbate 22
Carbons
Vitreous
Speer 37
Pure 480
Pure 4799
UCC YBD
Flame Sprayed on Titanium Alloy
(5.1% Al, 2.5% Sn)
TiB2
TiC
TIN
WC
WSi2
Si
Other Metallic Samples
Platinized Titanium
Ti Alloy + Pt Foil
Oxide Coated Ti Alloy
Ti Alloy, Base'
Hastelloy C
Niobium
Tantalum, 10$ Tungsten
Oxides
(2$ Sb203, Q .1% ZnO)
Magnetite (Natural)
3.5
3.7
5.4
3.9
2.8
3.9
4.7
0.42
0.67
(5)
2.3
4.6
11.8
8.1
7.0
6.7
0.14
0.37
0.37
1.88
0.34
0.16
0.001
0.43
0.053
Not weighed
18
17.3
0.10
0.29
0.31
0.45
0.32
0.23
0.32
0.39
0.035
0.055
0.19
0.38
1.02
0.67
0.58
0.56
5x10
-6
0.26
0.19
6.5x10-^
Impregnated with
Halocarbon wax
Layer separation
Sealed edges
0.044 5.7X10-2*
High Anodic Polari-
zation
Constant current not
maintained
Early termination
Early termination
High anodic polari-
zation
Constant current not
maintained
(insulating film formed)
Pressed @ 14K psi
Sintered 5 hours 1350 C
33
-------
TABLE VII
STANNIC OXIDE ELECTRODES PRESSED AT
14,000 PSI, SINTERED AT 1400 C
Sampl e
I
II
III
IV
V
VI
Wt. % Sb203
0.1
0.1
0.1
2.0
2.0
-
Wt. $ ZnO
-
0.7
2.0
0.7
2.0
-
Vt. % V205
-
-
-
-
-
2.0
Density, g/cc
3.9
6.2
6.2
5.2
3.5
5.5
Resistance
ohms
3.7
200,250
150,300
0.6, 0.7
8, 15
2x1 07, 4x1 07
-------
TABLE VIII
STANNIC OXIDE COMPOSITIONS PEESSED 14,000 PSI,
SINTERED AT 1400 C, SbO- ATMOSPHERE
Sample
VII
VIII
IX
X
XI
XII
Wt. % ZnO
0.7
0.7
0.7
0.0
0.0
0.0
Wt. % CdO
0.0
0.0
0.0
1.1
1.1
0.0
Wt. % Bi203
0.0
0.0
0.0
0.0
0.0
3.2
wt. % Sb2o3
0.0
0.1
0.25
0.0
0.5
0.0
Density
5.9
5.3
4.9
5.2
3.9
4.9
35
-------
Several of the formulations shown in Tables VII and VIII were se-
lected for life testing in service. The results of the extended
tests are illustrated in Table IX along with some results for natural
magnetite and platinized titanium. As can be seen, the corrosion re-
sistance of the stannic oxide formulations was found to be lower than
magnetite and exceeded that of platinized titanium by a factor of only
five or less. Based on the results, stannic oxide would be expected
to be useful for electroosmotic pumping.
One can project an expected life for stannic oxide anodes in electro-
osmotic pump service from the data of Table IX. Vith a current
density of 0.125 ma/in2 (expected to be typical in electroosmotic
pumping), a corrosion loss of about 0.01 g/Faraday (conservative,
according to Table IX), and electrodes of density JO g/in3, about
0.01 in/in^/yr of electrode surface would corrode away. Thus elec-
trodes a half inch thick would be expected to have about a five-year
life.
36
-------
TABLE IX
ANODE LIFE TESTS
HYPERION DIGESTED SLUDGE, CONSTANT CURRENT OF 19 ma/cm"
Sample Description
Stannic Oxide IV
Stannic Oxide VII
Stannic Oxide VIII
Stannic Oxide X
Magnetite
Platinized Titanium
Cumulative Anode Vt loss for number of hours indicated
g'
162 hrs
0.0072
0.0056
0.0019
0.0059
-
0.0013
"ams/Farac
325 hrs
0.0096
0.0088
0.0028
0.0077
-
0.0023
lay
92 hrs
-
-
-
-
0.042*
-
162 hrs
4.8xlO~5
3.7xlO~5
1.3xlO~5
3.9xlO~5
-
0.7xlO~5
tnoles/Fara
325 hrs
6.4xlO~5
6.2xlO"5
2.2xlO"5
6.5xlO~5
-
1.2xlO"5
day
92 hrs
-
-
-
-
5.4xlO~4*
-
-------
ESTIMATES OF ELECTRICAL COSTS FOR ELECTROOSMOTIC PIMPING
From the results presented in the foregoing discussion, some prelim-
inary estimates of the electrical costs of electroosmotic pumping can
be made. Cost calculations can be based on the data presented in
Table III and on the assumption that the weight of sludge which can
be collected is equal to that contained in the volume of water re-
moved. Thus, with an average energy consumption for the early, more
efficient portion of the dewatering process being 0.11 watt-hr/cc,
and assuming power cost of $0.01/kwh, one obtains the curve shown in
Fig. 14. Plant costs are not included in the calculation.
It is to be noted that the estimated electrical costs for processing
fall dramatically with an increase in initial sludge solids content.
This cost reduction results because the cost of dewatering is deter-
mined by the amount of water that must be removed to achieve a de-
sired cake dryness level. To dewater sludge from 2fo to JOfo, one must
remove about twice the water than that necessary to achieve JOfo sludge
when starting with a k% suspension.
Electrical costs can be seen to be attractive if some preconcentration
of material to about 6-10$ solids is assumed. Thus, for mixed waste-
activated and primary sludges, which we have found to be gravity
filterable to solids contents as high as 20%, costs for dewatering
to burnable material should be dramatically lower than that of the
digested sludge, which does not settle as efficiently. Since it was
shown earlier that electroosmosis does, in fact, work well with pre-
concentrated sludges, one concludes that the process will be most
attractive if applied in tandem to a pre-treatment stage that thickens
the influent.
38
-------
o
to
o
i
i-
o
ec
ac
ui
O.
I/)
o
o
16
6 8 10 12 1
INITIAL SUSPENDED SOLIDS CONCENTRATION, PERCENT
16
Figure 14. Electrical Costs for Electroosraotic Dewatering
39
-------
ACKNOWLEDGEMENT
The work reported herein was carried out over the period 24 June 1968
through 23 July 1970 by the staff of the Rocketdyne Research division.
Dr. B. L. Tuffly, Manager, Environmental Sciences and Technology,
served as the Program Manager. Dr. Jerome Greyson was the Responsible
Scientist. Members of the Technical Staff contributing to the pro-
gram were Dr. H. H. Rogers, Dr. 0. Kalman, Mr. P. Faurote, and Dr.
H. Offner.
Dr. R. B. Dean, Chief, Ultimate Disposal Research Section of the
Advanced Waste Treatment Research Laboratory, Federal Water Quality
Administration, served as the Project Officer for the program. The
Rocketdyne staff wishes to express their appreciation for the interest
and cooperation provided by Dr. Dean throughout the course of this
research.
40
-------
REFERENCES
Adrian, D. D., et al (1968), "Source Control of Water Treatment
Waste Solids," Report No. EVE-14-69-2, FWPCA, March 1969.
Beaudoin, R. E., (1943), "Reduction of Moisture in Activated Sludge
Filter Cake by Electrooamosis," Sewage Works Journal, 15, 1153-63.
Cooling, F. F., et al, (1952), "Dewatering of Sevage Sludge by
Electroosraosis," The Water and Sanitary Engineer, London, November.
Glasstone, S., (1952), Textbook of Physical Chemistry, D. Van
Nostrand Co., New York, p. 1223-25.
Reuss, (1809), Mem, de La Soc. Imper. Des Naturalistes de Moscou,
P. 327.
McCabe, W. L. and Smith, J. C. (1956), "Unit Operations of Chemical
Engineering," McGraw-Hill, New York, p. 50.
41
-------
PUBLICATIONS AND PATENTS
Greyson, J., and H. Rogers, "Suspension Dewatering Apparatus,"
Patent Applied For, 1969.
Greyson, J., "Electroosmotic Sewage Sludge Dewatering," Yale
Scientific Magazine, XLIV (6), March 1970.
Greyson, J., and H. Rogers, "Dewatering Sewage Sludge by Electro-
osmosis," Proc. of the Fifth International Water Pollution Research
Conference, Pergamon Press Ltd., London to be published Spring 1971,
42
-------
APPENDIX I
SOURCE AND TYPES OF SLUDGES TESTED FOR EFFICACY
OF ELECTROOSMOTIC DENATURING
Source
Las Virgenes Water District
Tapia Plant, Calabassas, California
Los Angeles County
Joint Water Pollution Control Plant
24501 S. Figuroa St.
Harbor City, California
City of Los Angeles
Hyperion Treatment Plant
12000 Vista Del Mar
Playa Del Rey, California
Metropolitan Water District
Water Treatment Plant
LaVerne, California
Rinconada Water Treatment Plant
San Jose, California
Type
Activated sludge
Activated digested sludge
Aerobic filtercake-prepared
from gravity filtered with
filter and Rohm and Haas C-7
polyelectrolyte
Raw primary sludge
Anaerobically digested sludge
(about 10 days)
Raw primary sludge
Activated sludge
Anaerobically digested sludge
(about 21 days)
Alum water treatment sludge,
sand filter backwashings
Alum water treatment sludge,
reactor-clarifier underflow
43
-------
APPENDIX II
ACCELERATED DENATURING THROUGH TAILORING OF SURFACE PROPERTIES
ADJUSTMENT OF ZETA POTENTIAL TO INCREASE RATES
OF ELECTROOSMOSIS
Acceleration of electroostnosis through adjustment of the zeta poten-
tial of aludge particle surfaces was investigated by measuring the
electrophoretic mobility, i.e., the velocity with which a charged
particle drifts through a distance of 1 cm under a potential gradient
of 1 volt. Electrophoretic mobility is directly proportional to the
zeta potential and is, essentially, the inverse of electroosmotic
velocity. For particles of the order of 1 micron in diameter, elec-
trophoretic mobility can be measured by direct observation with a
travelling microscope, the suspension being placed in a tube of
rectangular or circular cross section. For these measurements, a
commercial micro-electrophoresis unit called the "Zeta Meter" was
procured. The Zeta Meter consists of a binocular microscope, a
micro-electrophoresis cell in which is drilled a drift tube of
circular cross section, and a precision power supply.
Initial experiments were conducted to assess the extent to which
variations in the pH and conductivity of the supernatant would
affect the electrophoretic mobility (EM) of digested sludge obtained
from the Los Angeles Hyperion Municipal Treatment Plant. Measure-
ments were carried out with suspensions of 1 cc aliquots of sludge
in 100 cc of distilled water, and pH and conductivity adjustments
were made to the mixture. Buffer tablets were used for pH adjust-
ment, and NaCl was used to adjust conductivity.
The results of the measurements are presented in Tables X and XI,
and Fig. 15. In Table X, the influence of pH on EM is shown over
the pH range 3 to 12. It is to be noted that at pH 3, the value of
EM is zero, resulting, apparently, from the partial neutralization
of the negative surface charge of the sludge by hydrogen ions.
Above pH 3, the value of EM is relatively independent of the acidity
of the solution. At high pH, EM increases. However, it is also to
be noted that the buffer tablets increase the specific conductivity
of the solutions with increases in pH. At pH 12, the conductivity
decreases. Thus, the increase in EM for the pH 12 solution may re-
sult from adsorption of the buffer salts imparting a higher negative
charge to the sludge particle surface.
-------
TABLE X
ELECTROPHORET1C MOBILITY
- PH
PH
3.1
4.1
6.8
8.0
12.0
Specific Conductivity x 10
(ohm-cm)
2,600
4,400
8,200
10,000
8,000
EM
Microns /Vol ts
sec cm
0
1.1
1.1
1.1
1.5
TABLE XI
ELECTROFHORETIC MOBILITY-CONDUCTIVITY
Concentration
Equivalent
NaCl/Liter
0.1
0.065
0.05
0.02
0.005
0.001
(ionic Strength) ' "~
0.30
0.25
0.22
0.14
0.0?
0.03
Specific s
Conductance x 10
(ohm-cm) ~1
12,000
9,000
4,950
2,925
750
260
EM
Microns /Volt
sec cm
0.9
1.3
1.7
1.6
2.0
2.4
45
-------
o-s
(IONIC STRENGTH)1/2 » I1/2
Figure 15* Electrophoretic Mobility Dependence on Ionic Strength
-------
Values of EM as a function of conductivity are shown in Table XI. It
is to be noted that EM is strongly dependent upon solution conductivity.
In Fig. 15, the ratio EMQ/EM, where EM0 is the mobility at infinite di-
lution and EM is its value at the measured concentration, is plotted
versus the square root of ionic strength. At values of ll/2 less than
about 0.15, the data follow a straight line, while at higher concentra-
tions, the data deviate from the line. Such behavior is typical of
suspensions which follow Debye-Huckel theory and is indicative of the
influences of the dissolved salt on the surface charge of the sludge
particles.
It was this influence of dissolved materials on surface charge which
was to be exploited to increase the electrophoretic mobility of sludge
and thus increase the electroosmotic velocity of water out of sludge.
It was expected that significant variations in mobility could be
effected by small additions of surface active agents or of salts with
multiple negative charges. These species could adsorb to the surface
of the sludge particles and impart to them an enhanced negative sur-
face charge, thus increasing the zeta potential and the electrophoretic
mobility. A series of experiments was carried out to determine whether
beneficial increases in electrophoretic mobility could be obtained by
additions of multivalent anionic salts and several commercial deter-
gents.
Solid sludge was obtained from the sedimented material resulting from
centrifuged (about 1 hour at 17,500 rpm) Hyperion digested sludge.
The supernatant itself was collected, filtered through a 34 Millipore
filter, and again centrifuged for an hour. Measurements were carried
out in supernatant, water, and ammonium bicarbonate solution (to simu-
late sludge supernatant). Weighed amounts of sludge solids and vary-
ing concentrations and types of surface active agents were added to
the liquids. The surface active agents which were tried are listed in
Table XII, and results are shown for the three suspending liquids in
Tables XIII, XIV, and XV.
As can be seen, regardless of the nature or concentration of the addi-
tive tried, significant increases in electrophoretic mobility for the
sludge solids were not obtained. In view of the results, it was con-
cluded that increasing the electroosmotic flowrates of water out of
sludge by adjustment of sludge zeta potential was not a promising
approach and the effort was terminated.
-------
TABLE XII
SURFACE ACTIVE AGENTS TESTED
Type, It
Organic
Surfactant
Chemical Formula
Alconox
Calgon
Sodium Acid Phosphate
Potassium Sulfate
Miracle White
Super Cleaner
Boric Acid
Sodium Silicate
Sodium Tetraborate
Sodium Citrate
Sulfanilic Acid
Igepon T-73*
Igepon AP-78*
Igepon AC-78*
Igepon Cn-42*
Igepal CO-630*
**Gafac Lo-529*
A long chain alkyl sulfonate
(NaPO,)
v 3 x
..,-_
3 6 -)
0 CH
ii i
7H C-N-CH2CH0SO Na
_H C-0-CHQCH0SO Na
, 13H23 ,
C H C-N-CH2CH2SO Na
CTJ c\(r*u r*u r\\ nu t
19Hi9~u\tH2 2 '82
Aryl 0(CH0CH00)n 0
/^ „ \ - - 'p '
(c8-c20)
HO ONa
Anionic
Anionic
Anionic
Anionic
Anionic
Non-ionic
Anionic
* Product of General Aniline and Film Corp., 140 W. 41st St., New York.
Manufactured by Monsanto, 800 N. Lindbergh Blvd., St. Louis, Mo.
*~*GAFAC Surfactants are mixtures of mono and diesters; increase in
product number indicates increase in n.
48
-------
TABLE XII
(Concluded)
Surfactant
Chemical Formula
Type, If
Organic
Gafac GB-528*
Gafac RE-610*
Gafac RS-710*
Santomerse ME-b
PVP ND K-30*
Alkyl 0(CH2CH2)nx ^0
P
HO XONa
Aryl 0(CH2CH20)n 0
HO
\
OH
Alkyl 0(CH0CH00)n 0
2 2 \
/ \
NO OH
Sodium salt of linear
alkyl-arl sulfonate
H2C
H2C
N
CH,
C=0
-CH
Anionic
Anionic
Anionic
Anionic
Viscosity
alternating
material
Product of General Aniline and Film Corp., 140 V. 41st St., New York.
Manufactured by Monsanto, 800 N. Lindbergh Blvd., St. Louis, Mo.
-------
TABLE XIII
ELECTROFHORETIO MOBILITIES IN SLUDGE SUPERNATANT
Sampl e
Supernatant + Solid
Supernatant + Solid H
Alconox
Supernatant + Solid
Supernatant + Solid 4
Na2HP04.7H20
Supernatant + Solid
Supernatant + Solid -t
Supernatant + Solid
Supernatant + Solid 4
Alconox
gm Solid
Sludge/
1000 ml
3.5
2.4
2.9
2.9
1.1
1.1
3.6
3.6
gm Added
Material/
1000 ml
0
1.23
0
1.37*
9
1.2
0
.127*
Conductivity
ohm cm x 10
6.1
7.2
6.8
6.8
6.7
8.2
6.8
6.7
EM
Hicrons /Volts
sec cm
1.32
1.89
1.51
1.44
1.52
1.30
1.39
1.52
*1 ml of concentrated solution in water added.
TABLE XIV
ELECTROPHORETIC MOBILITIES IN WATER
Sample
Water + Solid
Water + Solid +
Alconox
Water + Solid
Water + Solid +
NapHPO , • 7H 0
Water + Solid
Water + Solid +
gm Solid
Sludge
1000 ml
3.5
3.6
3.0
3.0
2.6
2.6
gm Added
Material/
1000 ml
0
1.76
0
1.36
0
1.43
Conductivity
ohm cm x 10
6.8 x 10~2
1.6
5.0 x 10~2
1.8
5.4 x 10~2
1.95
EM
Microns /Volts
sec cm
1.84
2.57
1.89
1.91
1.48
1.74
50
-------
TABLE XV
ELECTROFHORETIC MOBILITIES IN NH.HCO,
4 3
SOLUTION ( 2600 ppm)
Sampl e
Solid
Solid + Calgon*
Solid + Calgon*
Solid
Solid + Calgon
Solid
Solid + Alconox
Solid
Solid + Alconox
Solid
Solid + Alconox
Solid
Solid + Miracle White
Super Cleaner*
Solid + Miracle White
Super Cleaner*
Solid + Miracle White
Super Cleaner*
Solid + Miracle White
Super Cleaner*
Solid + Miracle White
Super Cleaner*
Solid
Solid + H_BO_
Solid
Solid + Na2B,0
Solid
Solid + Na2SiO,
Solid
Solid + NH2C6H4SO -H.H20
gm Solid
Sludge
1000 ml
2.6
2.6
2.6
2.4
2.4
3.2
3.2
1.4
1.4
2.9
2.9
2.3
2.3
2.3
2.3
2.3
2.3
1.0
1.0
.95
.95
.62
.62
1.3
1.3
gm Added
Material/
1000 ml
0
.083
.42
0
.27
0
.128
0
.380
0
.767
0
.107
.457
.800
1.77
5.30
0
1.51
0
1.34
0
1.7
0
1.6
Conductivity
ohm cm x 10
3.0
2.8
2.8
3.2
3.4
3.0
3.0
2.9
2.9
3.1
3.8
3.0
3.0
2.5
2.4
3.1
3.4
3.2
3.2
3.1
3.2
3.1
3.1
3.2
3.2
EM
Microns /Volts
sec ' cm
1.40
1.48
1.49
1.44
1.52
1.47
1.64
1.57
1.55
1.60
2.01
1.28
1.22
1.41
1.57
1.49
1.49
1.43
1.47
1.51
1.56
1.48
1.47
1.53
1.46
* Material dissolved in water prior to adding to sludge suspension
51
-------
TABLE XV
(Concluded)
Sampl e
Solid
Solid + Na Citrate
Solid
Solid + Na_ Citrate
Solid
Solid + IGEPON T-73
Solid
Solid + IGEPAL CO-630
Solid
Solid + GAFAC LO-529
Solid
Solid + GAFAC RE-610
Solid
Solid + GAFAC RS-710
Solid
Solid + IGEPON AP-78
Solid
Solid + IGEPON CN-42
Solid
Solid + GAFAC GB-520*
Solid
Solid + PVP Type ND K-30
Solid
Solid + IGEPON AC-78
Solid
Solid + SANTOMEESE ME-b
gm Solid
Sludge
1000 ml
2.8
2.8
1.1
1.1
1.5
1.5
1.3
1.3
1.3
1.3
.91
.91
2.1
2.1
1.0
1.0
3.9
3.9
3.7
3.7
2.1
2.1
3.7
3.7
2.8
2.8
gm Added
Material/
1000 ml
0
.123
0
2.4
0
1.6
0
2.5
0
2.8
0
4.5
0
2.0
0
2.5
0
3.9
0
1.5
0
2.8
1
2.6
0
2.4
Conductivity
ohm cm x 10
3.1
3.1
3.1
4.8
3.0
4.5
2.9
2.9
2.9
2.8
3.0
2.7
2.8
2.8
2.8
3.6
3.1
3.7
3.5
2.9
3.1
2.9
3.1
3.0
3.1
3.0
EM
Microns/Volts
sec cm
1.65
1.27
1.43
1.63
1.35
1.59
1.40
1.57
1.40
1.74
1.29
1.52
1.41
1.68
1.29
1.69
1.25
1.51
1.35
1.47
1.67
1.39
1.37
1.51
1.27
1.69
* Diluted to 40 ml
52
-------
ACCELERATING ELECTROOSMOSIS THROUGH SELECTH) FILTERING MEDIA
An alternative to adjusting the zeta potential of sludge is the se-
lection of a suitable filtering medium with an intrinsic zeta poten-
tial greater than that of the sludge itself. The potential gradient
giving rise to electroosraosis could then be applied across the filter
medium rather than the sludge suspension. Because of the elevated
zeta potential in the filter medium, electroosmotic flow velocity
vould be expected to be greater through it than could be achieved
with identical voltage gradients across the sludge suspension. Such
a configuration would be, in essence, an electro-filter, with the
driving force for filtration being an electrical gradient rather than
a hydraulic gradient as is used in ordinary filtration.
The apparatus used for investigating the efficacy of electroosmotic
pumping through supplementary filtering media is shown in Fig. 16.
It consists of a glass cell of approximately 200-cc capacity. The
cell is divided into an effluent and sludge chamber. At the facing
ends of the chambers, pyrex 0-ring flanges have been sealed (Sargent
Catalog No. S-40223). The cell chambers can thus be separated con-
viently by screen electrodes or sandwiches of membranes and screen
electrodes. The effluent chamber contains a liquid leveling tube to
preclude gravity filtration, and the sludge chamber is fitted with a
stirring shaft which passes through the end opposite the 0-ring flange.
The stirring shaft serves also as an anode and its effective position
relative to the screen cathode or sandwich can be adjusted.
Provision was made in the cell to pump sludge through the sludge
chamber while withdrawing effluent from the effluent chamber. Pump-
ing from a volume of sludge, large relative to the total effluent
withdrawn, allowed one to carry out measurements of electroosmotic
flow out of the sludge under conditions of constant solids content.
The rotating anode was located about 1 cm from the membrane surface
in the measurements discussed here. In most runs, a spiral stainless-
steel wire cathode was used and this was kept almost in contact with
the membrane (Fig. 17). Use of the wire electrode provided free cir-
culation of liquid near the cathode and ready removal of gas bubbles.
In control runs, stainless-steel screen (approximately 50 mesh) was
clamped between the 0-ring joints and served both as the cathode and
as a base for the accumulation of a sludge "membrane". The purpose
of the latter was to provide a comparison in flowrates between sludge
itself and the filtering media.
In Table XVI, the data obtained for a variety of membranes and filter-
ing media are presented. The column headings in the table, i.e., elec-
troosmotic permeabilities, kg, specific resistances, R, and other
parameters were derived from the equations
Q = keA(v/d) II-1
I = (|/R) (v/d) H-2
53
-------
0-RING SEALS
FOR CATHODE
AND MEMBRANES
—"•) [-•— 6 MM I .
D.
r~ " ^
1 9 MM ID • • • • —i
1 1. nn i • u • *
r"i n
V
Figure 16. Electroosmotfc Membrane Filter Cell
-------
CATHODE
(WIRE SPIRAL)
MEMBRANE —' (- £
SLIDGE CAKE DEPOSIT
(HC THICK)
ANODE
(PLATINUM SCREEN)
LIQUID SLUDGE
Figure 17. Membrane Location in Composite System
55
-------
where Q and I are the volume flowrate and current, respectively. A
is the cross-sectional area of the membrane, and v/d is the voltage
gradient. For a composite medium, i.e., membrane and liquid sludge
(and in some cases also an adhering sludge cake), the total resistance
R, is the sum of the individual layer resistances, and I and Q are
constant throughout the cell. By means of the relation
H = hR + (I - h )R II-3
m m x* m7 s J
the membrane resistance R can be obtained from the layer widths and
the liquid sludge resistance Rg . This latter quantity is 'given by
R = I — II-4
s AKS
where K$ is the specific conductance of the liquid sludge (5.7 x 10~3
(ohm cm)~l). If, during a test run, a visible cake forms on the
membrane surface, the effect of this layer may be taken into account
by including an additional term, hcRc, on the right side of Eq. II- 3
The permeability coefficient for the membrane, (ke)m, is obtained from
Eq. II- 1 and II-2 upon multiplying Q/I by Hn/A. Using this method of
computation, one obtains the parameters for the individual membranes
from the composite system data shown in Table XI.
The permeabilities, resistances, and power values ih^ReA) shown for
the individual membranes refer to a situation where the anode and
cathode are in direct contact with the membrane surface. This would
correspond to full utilization of the membrane properties for the
water transport. However, the volume flowrate, Q, must be continuous
throughout the cell. Because of this, if a sludge cake deposit forms
on the membrane surface (which was found for the glass frit, ceramic
Si02, and polyurethane foam), the amount of liquid reaching the mem-
brane may be affected. Thus, depending upon the properties of the
sludge cake formed, the true maximum flow through the membranes in
the composite systems may not be achievable.
The data in Table XVT do indicate higher volume rate per consumed
power for the sludge as a barrier material than for the membranes in
the composite systems. Also, despite the large range in pore size
(300 ,OOOA to 2^A pore diameters) and the diversity in chemical proper-
ties, the power requirements do not seem to vary greatly for the in-
dividual membranes in the composite systems. Thus, it appears that
the formation of sludge cake and penetration of sludge particles in-
to the pores of the membranes may have, in fact, limited the attainable
volume rate for a given power input.
56
-------
TABLE XVI
ELECTROOSMOTIC DEWATERING DATA IN COMPOSITE SYSTEMS
Membrane Material
Sludge Cake
Glass Frit (M)
Ceramic (SiO^)
Polyurethane Foam
Porous Teflon
(Extra Coarse)
(Chemware Teflon/
Halon)
Porous Teflon
(Extra Fine)
Dialysis Membrane
Van ¥ater & Rogers
#25225
Nalfilm D-40
^Jalco Chem. Co.,
Chicago, 111.
Composite Data
Volume
Rate,
Q x 10-3
ml/sec
2.8
12.0
6.4
2.1
2.4
7.5
1.0
1.2
ke x 104
cm sec
per v cm"-'
1.75
1.2
.48
1.2
.35
.17
.17
.88
A,
cm
12.6
5.1
8.2
11.4
8.2
8.2
8.2
11.4
ohm cm •"
18.6
94.9
73.6
29.4
37.2
248
32.7
16.7
Power,
watt-hrs
ml
.26
.28
.45
.33
.42
.82
.82
.58
Derived Membrane Data
(ke)ffl x 104
cm sec 1
per v cm
.22
.11
.16
.72
.0074
.001
.0012
.015
V
ohm cm"1
140
978
211
50
1,775
38,100
4,670
1,000
Qm,
cm
.1
.15
.8
1.3
8.9 x 10~5
11.3 x 10~5
4.6 x 10"3
8.4 x 10-3
Power
watt-hrs
ml
.07
.16
.37
.21
.09
.80
.28
.10
UI
-------
Control experiments were performed by using an aqueous NaCl solution
(3,500 ppm NaCl) to ascertain if the sludge limits the flow through
the membranes tested. The NaCl solution had approximately the same
specific conductance as the sludge, and the membrane parameters were
calculated in the way described for the sludge experiments. Table
XVII shows the data obtained. The results with the NaCl solutions
appear to confirm that the properties of the sludge determine the
transport of water through the membranes. For the porous teflon
extra coarse membrane, even the flow direction changes when NaCl
solution is used. For the dialysis membrane, no water transport
appeared to take place with NaCl solution while a small flowrate was
measured with the sludge. For the glass frit, a more favorable volume
rate/power was obtained with the salt solution than with the sludge.
In general, it is to be noted that in these experiments with NaCl
solution, one finds a greater variation of flowrates (and directions)
from one membrane to the next than in the runs with the sludge. This
is in accord with the variety of membrane materials tested which range
from hydrophobic to hydrophilic and from high to low porosity. The
lack of such changes through the series of membranes investigated with
the sludge points to the probability that the sludge surface properties
become controlling even in cases where sludge cake formation at the
membrane surface cannot be visually observed. We attribute such be-
havior to the adsorption of sludge solids onto the membrane pore sur-
faces, thus imparting the surface characteristics of the sludge solids
to the membrane. Based on these data, it was concluded that no ad-
vantage could be gained through use of supplementary filter media.
58
-------
TABLE XVII
ELECTROOSMOTIC DENATURING DATA FOR NaCl SOLUTIONS
(1)
Membrane Material
Glass Frit (M)
Porous Teflon
(Extra Coarse)
(chemware Teflon/
Halon)
Dialysis Membrane
(Van Waters &
Rogers #25225)
Nalfilm D-40
(Nalco Chem. Co.,
Chicago, 111.)
Composite Data
Volume
Rate,
Q x 10--
ml/sec
7.3
2.8
Reverse
Dir.
0
.52
ke x 104
cm sec"l
per v cm"-'-
1.3
.25
.12
A,
cnr
5.1
8.2
8.2
R,
ohm cra-1
50.5
64.0
23.8
Power ,
watt-hrs
ml
.26
(.53)
1.4
Derived Membrane Data
(ke)m x 104
cm sec"^
per v cm
.17
.019
.0023
Rm, _
ohm cm
379
8400
1240
V
cm
.15
8.9 x 10~3
8.4 x 10"^
Power,
watt-hrs
ml
.11
(.035)
.2?
VJl
§
3500 ppm
Reversed Polarity
------- |