WATER POLLUTION CONTROL RESEARCH SERIES • 11023 OAA 03/72
HYPOCHLORITE GENERATOR
FOR TREATMENT OF
COMBINED SEWER OVERFLOWS
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes 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, development and demonstration
activities in the Environmental Protection Agency, through
inhouse research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, D.C. 20^60.
-------
HYPOCHLORITE GENERATOR
FOR TREATMENT OF
COMBINED SEWER OVERFLOWS
by
IONICS, INCORPORATED
65 Grove Street
Watertown, Massachusetts 02172
for the
OFFICE OF RESEARCH AND MONITORING
ENVIRONMENTAL PROTECTION AGENCY
Program No. 11023 DAA
CONTRACT No. 14-12-490
and
GRANT No. 11023 DME
March, 1972
For sale by the Superintendent of Documents, U.S. Oovarnment Printing Office, Washington, D.C. 20402 - Price $1.00
-------
EPA Review Notice
This report has been reviewed by the Environmental Protection
Agency and approved for publication. Approval do&s not
signify that the contents necessarily reflect the views and
policies of the Environmental Protection Agency nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
ii
-------
ABSTRACT
An advanced electrolytic generator has been developed for on-site pro-
duction of sodium hypochlorite for disinfection of overflows from
combined sewer systems.
In this system an electrochemical cell electrolyzes sodium chloride
brine to chlorine gas and sodium hydroxide solution, which are reacted
immediately outside the cell to produce a 5 to 10% sodium hypochlorite
solution. Significant advances in safety and economy have been realized
by use of a hydraulically impermeable cation exchange membrane. The
most critical components, the dimensionally stable anode and the ion-
exchange membrane, have both operated for over 3000 hours with no
deterioration of performance.
System operation has been given a first order economic optimization.
At a current density of 240 ASF, the cell potential is 3.7 volts. The
generator requires 1.6 KWH of electricity and 2.1 pounds of salt per
pound of sodium hypochlorite. Salt utilization is over 80%. The
operating cost for systems larger than 500 pounds of hypochlorite per
day is projected to be 3 to 4 cents per pound of hypochlorite. This
cost is significantly below that of truck delivered solution. Such
economy of operation should make the generator useful for a wide
variety of water treatment applications.
The first field unit is scheduled for installation in the Northeast
Sanitary District in Somerville, Massachusetts, under Grant 11023 DME.
The estimated operating and maintenance cost is estimated to be 3.7C
per pound of chlorine or 0.13C per thousand gallons of overflow.
This report was submitted in fulfillment of Program No. 11023 DAA,
Contract No. 14-12-490 and Grant 11023 DME under partial sponsorship of
the Office Research and Monitoring, Environmental Protection Agency.
111
-------
COOTENTS
Section Pa?e
Abstract
Figures vi
Tables viii
Conclusions 1
Recommendations 3
I Introduction 5
II Design of Hypochlorinator 7
III Laboratory Development 23
IV System Optimization 65
Acknowledgments 75
V References 77
VI Publications and Patents 79
VII Glossary 81
VIII Appendices 85
v
-------
FIGURES
No. Title page
1 Electrochemical Hypochlorite Generator 8
2 Hypochlorite Generator Block Flow Diagram 9
3 Molded Cell Frame 10
4 Basic Stack Components 11
5 Cell Press and Skid in Process of Fabrication 13
6 Detailed Flow Diagram 15
7 Mass Balance for Full Scale Hypochlorite Generator 18
8 Estimated Capital Cost of Hypochlorite Generator 22
9 Small Scale Test Cell and Control Apparatus 25
10 Diaphragm Cell with Sheet Electrode 28
11 Diaphragm Cell with Expanded Cathode 28
12 Effect of Diaphragm Material on Internal Cell
Resistance 32
13 Typical Voltage-Current Curve for Sheet Electrode
Diaphragm Cells 33
14 Effect of Salt Utilization on Current Efficiency
for Sheet Electrode Diaphragm Cell 34
15 Effect of Salt Utilization on Current Efficiency
for Expanded Cathode Diaphragm Cell 37
16 Membrane Cell with Expanded Electrodes 40
17 Three Compartment Cell 40
18 Dependence of Caustic Concentration on Salt Utilization
for Two Compartment Membrane Cell Operated in Mode I 43
19 Dependence of Current Efficiency on Salt utilization
for Two Compartment Membrane Cell Operated in Mode I 44
20 Effect of Salt Concentration in Cathode Feed on Cell
Performance 46
-------
FIGURES (Continued)
Title
21 Effect of Salt Utilization on Cell Performance.
Cathode Feed Water 47
22 Effect of Salt Utilization on Cell Performance.
Cathode Feed 5% NaCl 48
23 Dependence of Current Efficiency on Salt Utilization
for Three Compartment Cell Operated in Mode I 50
24 Two Cell Stack and Associated Test Apparatus 57
25 Hypochlorinator Optimization Program 66
26 Treatment Cost at the Boston University Bridge Facility 78
27 Hypochlorinator Cell Frame 85
28 Cell Filter Press - Jack Support Frame 86
29 Cell Filter Press - Stationary Platen 87
30 Cell Filter Press - Movable Platen 88
31 Cell Filter Press - Assembly Drawing 89
Vll
-------
TABLES
No. Page
Yearly Operating Cost of the Hypochlorite
Generator at Somerville, Mass. 20
Yearly Operating Costs of 1000 Ib/day
Hypochlorite Generator 21
Evaluation of Diaphragm Materials 30
Effect of Separation on Expanded Cathode Cell 36
Comparison of Anode Operating Potential 54
Comparison of Cathode Operating Potential 55
Effect of Varying Water Feed to Cathode 59
Comparison of Brines from Three Grades of Salt
with Pretreated Brine 61
9 Amortized Capital Equipment Costs 70
10 Annual Operating Costs 71
11 Nomenclature Used in Economic Analyses 72
Vlll
-------
CONCLUSIONS
1. On-site electrochemical generation of hypochlorite is an efficient,
economical means of disinfecting combined sewer overflows. In
most cases hypochlorite solution can be generated for considerably
less than commercial prices for truck delivery.
2. A generator .ystem has been developed, which consists of a brine
saturator of conventional design, an electrochemical cell in which
chlorine gas and sodium hydroxide solution are produced and a re-
actor in which these materials are reacted to produce sodium
hypochlorite solution. The emphasis in the design has been on
safety, reliability and unattended operation.
3. The principal components in the electrochemical cell are a DSA
anode, a mild steel cathode and a fluorinated ion-exchange mem-
brane separator, all of which have been satisfactorily life tested
for over 3000 hours.
4. The most favorable configuration for an on-site cell is a two-
compartment cell having a membrane separator and expanded metal
electrodes adjacent to the membrane.
5. Hypochlorite cells have been run at high current densities, up to
500 ASF, without excessive loss in current efficiency or de-
gradation of materials. The optimum current density depends
upon various economic factors but generally is in the neighbor-
hood of 250 ASF.
6. The major operating variables are catholyte concentration, current
density and salt cut. A computer program, included in this report,
was developed which predicts the optimum operating conditions for
given values of salt and power cost and other economic parameters.
-------
RECOMMENDATIONS
1. The laboratory modules have met or exceeded the performance
specifications in the original design. Work should proceed on
construction, installation and testing of the field unitj as
scheduled.
2. Testing of the unit should include observation of component aging
and measurement of performance with time under real intermittent
operating conditions.
3. A study using modern methods of dynamic programming and incorpor-
ating hydrographic data from various regions in the country should
be performed to develop the optimum policy with regard to pro-
duction capacity of the unit, the fraction of time the unit should
be designed to run, size of brine saturator and size of product
storage tank.
4. Because of the variability of chlorine demand in outfalls during
a storm or between storms, a means of controlling the dosage to
the combined effluent should be developed.
5. A study should be made of the effects of calcium, magnesium and
sulfate in the salt and in the feed water to determine what con-
centrations of these species in the brine can be tolerated. Several
possible means of removal should be evaluated to determine which
is most economical.
-------
SECTION I
INTRODUCTION
Many American cities possess combined storm and sanitary sewers. In-
terceptor sewers are usually sized to handle approximately 3 to 5 times
the average dry weather flow. When a severe storm occurs, the com-
bined sewer flow to the interceptor may increase to twenty to one
hundred times the dry weather flow. Overflow weirs are generally
located throughout a combined sewer system so that raw sewage is
directed to treatment facilities during dry weather but, during wet
weather, is directed to the local water course and will not back up
into homes and streets under storm conditions. Installation of
separate sewer systems in older cities is considered prohibitively
expensive.
A minimum treatment of these overflows is brief retention and dis-
infection by addition of relatively small doses of chlorine gas or
hypochlorite solution. Chlorine gas is highly toxic as well as
corrosive. As a result of public concern regarding the dangerous nature
of chlorine and the widespread news coverage of chlorine accidents,
many cities are reluctant to use chlorine gas.
A safer alternative to use of chlorine gas is use of sodium hypochlorite
solution. This is chemically identical to the bleach solutions used in
most households. Advantages of the use of purchased hypochlorite are
low capital cost and low maintenance since little equipment more than a
storage tank and an automatic feed device is required. Disadvantages
are that purchased solution is expensive, it requires truck delivery and
the solution decomposes slowly but regularly which makes it undesirable
to keep large quantities on hand.
In a study funded by the Environmental Protection Agency, a system was
designed for on-site electrochemical generation of hypochlorite solution.
This study, which was highly successful, resulted in the development of
a system consisting of an electrochemical cell and a reactor, which pro-
duces a high quality solution of hypochlorite at prices which, for
almost any reasonable commodity costs, are much lower than that of
truck-delivered hypochlorite solution.
In this report the evolution and design of the system is described,
results of life tests of components are presented, and operating,
maintenance and capital costs axe estimated. The first field test unit
has been constructed for the Metropolitan District Commission of the
Commonwealth of Massachusetts. Details of the application are dis-
cussed.
-------
SECTION II
DESIGN AND OPERATION OF HYPOCHLORINATOR
The major components of the hypochlorite generator are a stack of
electrochemical cells and a caustic-chlorine reactor. In the cells
sodium chloride is decomposed to chlorine and sodium hydroxide (caustic).
Chlorine and caustic are combined in the reactor to produce sodium
hypochlorite solution (hypochlorite). Figure 1 shows the assembled
hypochlorite generator. A block flow diagram appears in Figure 2.
The electrochemical cell is the heart of the hypochlorite generator
system. It consists of two electrode compartments separated by a
membrane. Saturated brine is fed to the anode compartment. At the
anode surface, chlorine gas is generated. The effluent from the anode
compartment is sent to a gas-liquid separator where the chlorine gas
is separated from the depleted brine, which is discarded. The membrane
in the cell acts as a .hydraulic barrier which permits electrochemical
transport of sodium ions from the anode to the cathode compartment. A
small quantity of water is fed to the cathode compartment. At the
cathode hydrogen gas and hydroxyl ions are generated. The effluent
from this compartment goes to a gas-liquid separator in which the
hydrogen is removed from the caustic solution. The hydrogen is diluted
with air to an innocuous level and vented to the atmosphere.
Chlorine and caustic are fed to the reactor with caustic in slight
excess. The reactor is water cooled to, avoid decomposition of hypo-
chlorite and formation of oxygen. Insoluble gases, mainly oxygen, are
removed in a gas-liquid separator before the hypochlorite is sent to
the product storage tank.
Brine feed for the system comes from the salt storage tank, or lixator,
which is used for both salt storage and production of saturated brine.
The lixator is an item of commerce available in a variety of sizes, the
choice of which depends upon the expected rate of salt usage. Generally,
the lixator is obtained from the salt supplier. Because of the crystal
properties of sodium chloride, the bed of salt acts as a filter for the
saturated brine.
Since the hypochlorite is not used as it is produced, a storage tank
is provided for it.
The basic components of the cell stack are the molded cell frame and.
the membrane package. The cell frame is shown by itself in Figure 3
and with electrodes installed in Figure 4. The membrane package appears
on the right in Figure 4. The cell frame is injection molded from
asbestos filled polypropylene with overall dimensions of 24" x 24-3/4"
x 2". The cavity on one side of the frame forms the anode compartment
-------
00
Figure 1. Electrochemical hypochlorite Generator
-------
Brine
Water
CELL
STACK
Chlorine
SEPARATOR
Spent Brine
WASTE
VENT
Hydrogen
SEPARATOR
Caustic
COOLER
VENT
Inerts
REACTOR
Product
Hypochlorite
Figure 2. Hypochlorite Generator Block Flow Diagram
-------
Figure 3. Molded Cell Frame
-------
Figure 4. Basic Stack Components
Left: Cell Frame with anode attached
Right: Membrane package supported on cell frame
-------
of the adjacent cell. A 3/8" wall in the center of the frame provides
hydraulic isolation between cells. Electrical connection between the
anode and the adjacent cathode is achieved by using special through-
connectors which pass through two 1-3/4" diameter holes in the wall.
These through-connectors carry about 265 amperes apiece while main-
taining the hydraulic isolation between the compartments.
The compartments are fed from the 1" diameter manifolds at the bottom
of the cell. The channel that leads from the manifolds at the com-
partment is drilled and tapped to receive a specially designed flow
restrictor fabricated from Kel-F plastic (polytrifluorochloroethylene) .
The flow restrictors distribute the flow evenly to each compartment
throughout the cell stack. The restrictor appears in Figure 3. The
gases and electrolyte exiting from each compartment pass through four
1/2" diameter holes at the top of the compartment and into the product
manifolds located at the top of the frame.
The anode is an expanded titanium screen coated with a special metal
oxide coating. This is the so-called dimensionally stable anode or
DSA. It rests in a recessed ledge in the face of the molded cell frame.
The cathode is constructed from a sheet of expanded mild steel and is
fitted in the other side of the cell frame in a similar manner.
The membrane package consists of an XR cation-transfer membrane laminated
between 2 frames of 1/32" thick high temperature PVC, which have cutout
portions corresponding to the active area of the cell and the manifold
holes. A die-cut gasket of 1/16" ethylene propylene rubber is then
bonded onto each side of the membrane frame to produce a combination
membrane and gasket. The membrane package is fitted onto the face of
the assembled cell block by slipping it over two 1/4" diameter nylon
line-up pins that are fitted into holes drilled in the border of the
cell frame. These line-up pins insure a proper alignment of all the
components in the assembled stack.
The cell components are fitted into a skid mounted press similar to the
type used for a plate-and-frame filter apparatus. The skid is shown in
the process of construction in Figure 5. The hydraulic connections to
the cell manifolds are made through the stationary end plate. Besides
the cell the skid contains the pumps, filters, and product handling
equipment required by the system. All feed piping is corrosion resist-
ant FVC with no metallic components exposed to corrosion by the brine
feed.
The cell effluent system, including the reactor, is mounted at the end
of the skid directly adjacent to the fixed endplate. All of the com-
ponents are constructed of either glass, PVC, or titanium. The lower
portion of the chlorine seal leg is constructed of industrial glass
piping to allow visual determination of the reactor operating pressure.
12
-------
Figure 5. Cell Press and Skid in Process of Fabrication
-------
Two Hastelloy level probes are also inserted in the seal leg to shut
down the system-if there should ever be a loss of liquid seal.
Figure 6 shows a detailed flow diagram of the final generator system.
Saturated brine from the lixator is filtered and fed to the anolyte
manifold of the cell stack by a positive displacement metering pump.
The pressure of the feed stream is monitored to warn of a pump failure
or clogged filter system.
The water feed to the cathode compartment is first reduced from standard
city supply pressure to about 1 psi and then filtered and metered in a
manner identical to that of the brine. Pressure sensors are also in-
stalled after the pumps to detect pump or filter malfunctions.
The mixture of chlorine and spent brine from the anolyte compartments
of the cell system is fed to the chlorine seal leg assembly. The
chlorine gas flows overhead to the reactor while the disengaged spent
brine flows out of the seal leg by gravity.
Ths mixture of hydrogen and caustic from the cathode compartments is fed
into the hydrogen disengager. The hydrogen gas is removed by means of a
blower system which dilutes the hydrogen with air to well below its ex-
plosive limit and vents it to the atmosphere. The caustic solution is
cooled in a shell-and-tube heat exchanger to about 30° C before it enters
the reactor.
The reactor consists of a piece of 4" diameter PVC pipe approximately
40" in length. The caustic flows up from the bottom of the reactor and
reacts with the chlorine also fed from the bottom. The resulting sodium
hypochlorite solution overflows from the reactor outlet by gravity into
an accumulation tank.
The chlorine gas is sparged into the reactor through a piece of 1" PVC
pipe which projects into the solution from the top of the reactor. The
pipe, drilled with a number of 1/32" sparge holes, has a 15 foot long,
1/4" diameter titanium cooling coil wrapped around it to absorb the heat
of reaction. This reactor gives a reaction yield greater than 96%. For
a short time during the startup of the unit there is insufficient caustic
present to absorb all of the chlorine gas. To avoid undesirable operating
conditions, the pH of the reactor effluent is monitored and additional
caustic from an auxilliary supply is added whenever the pH drops below
8.0.
An electrical control panel provides for automatic start-up and shut-
down of the hypochlorinator and has also provisions for emergency shut-
down in case of equipment malfunction. In normal operation when there
is a demand for hypochlorite, the control system sequentially turns on
the blower system; turns on hydrogen detector, chlorine detector and
pH control system; opens appropriate feed valves; starts pumping system;
14
-------
LIXATOR
P-4
HYDROGEN
EXHAUST
<0> - > WASTE
SURGE
ACCUMULATOR
PRODUCT TRANSFER, PUMP
P-
BRINE
METERING
PUMP
Figure 6.
Detailed Flow Diagram
-------
and turns on the rectifier. There is a time delay of 20 seconds before
the safety interlocks are activated to al±ow the system to begin normal
operation. After the system has operated for sufficient time to satisfy
the hypochlorite demand, all of the equipment, except the hydrogen blower
and the transfer pump is automatically shut down. The blower and trans-
fer pump continue to operate for 10 minutes to insure that all of the
hydrogen is vented and the hypochlorite has been forwarded to the
storage tank.
The control panel also has interlocks to shut down the cell system if
any of the following conditions occur:
• blower failure;
• high cell temperature;
• high reactor temperature;
• chlorine or hydrogen release;
• loss of seal in chlorine seal leg;
• high or low current limit;
• high pressure drop through filters;
• loss of pump pressure.
If any one of these conditions occurs, the system immediately shuts down
and a malfunction light indicates the nature of the problem. An alarm
horn is also activated to summon an operator. After the malfunction is
corrected, the system can be restarted by means of a reset button.
The power supply was obtained from Enthone, Inc., Brooklyn, N.Y. The
rectifier is of the saturable core type with a capacity of 1500 amps at
36 volts D.C. The rectifier has a stepless control over the range from
100 amps to 1500 amps and has safety features such as high and low
current shutoffs and thermal overload protection.
Cell Operation
A number of operating conditions for the full scale cell were investigated
to optimize temperature, efficiency, salt utilization, and voltage. The
most favorable operating conditions are:
Cell Active Area 2.2 ft
Current 525 amps
Current Density 240 amps/ft
Cell Voltage 3.7-3.8 volts/cell
16
-------
Cell Temperature 87 - 89° C
Brine Feed Rate 58 ml/min cell
Salt Utilization 81%
Water Feed Rate 110 ml/min cell
Caustic concentration 2.0 - 2.3 N
Hypochlorite Production Rate 135-140 ml/min cell
Overall Electrical Efficiency 77%
An overall mass balance for the operation of the hypochlorite generator
is shown in Figure 7. This mass balance was derived from data taken
during test runs of the full scale two cell stack. These data were then
scaled up to represent the operation of the 12-cell module which is
presently under construction. The data reflect operation at a current
density of 240 ASF and a temperature of 86° C. The operating voltage of
the cells under these conditions was 3.8 volts per cell.
Component Life
All of the critical components in the system have been tested in actual
operation for over 3000 hours without any sign of deterioration or
dunctional impairment.
Membrane - The XR membrane has been tested for more than 3000 hours at
current densities between 200 and 400 ASF. In all of the tests that have
been run on the present membrane there has been no membrane failure or
any apparent attack on or deterioration of the material. Based on these
tests it is felt that the membrane should last for at least two years
in continuous service and probably considerably longer.
Electrodes - The DSA's have been run for the same period of time as the
membranes. Only one anode failure was experienced during this program
and that failure was attributable to obvious mistreatment during a test
run. According to experience and based on data supplied by Electrode
Corporation the anodes should have a minimum life of three years. After
the end of this period the electrodes can be recoated.
Cell Frames - The cell frames are constructed from high temperature
asbestos filled polypropylene. This material was tested extensively in
both hot wet chlorine and caustic. It undergoes an initial surface '
attack which roughens and whitens the surface and results in a slight
weight gain. No further attack is observed.
All of the auxiliary equipment in the system was chosen on the basis of
maximum economical resistance to the operating conditions. In the test-
17
-------
Spent Brine
"*
X = 56.8
HO = 51.3
2
Cl" = 3.4
Na+ = 2.1
Current = i300 amp
Cl- = 17.2
2
0_ = 1.0
1
Na+
OH~ <
H20
Ce]
>
Brine Feed
k H = D.56
"2
Caustic Stream
Z =261.
HO = 241.5
Na+ = 11.3
OH~ = 8.3
11.3
2.03
47.3
.1
Z = 131.5
H^O = 97.9
Cl- = 20.7
Na+ = 13.4
^
H20 = 205.
i
J-JICJ- 1,0
O 3 1 O
°2 1-°
Product Hypochlorite
Z = 278.
HO = 247.
NaOCl = 17.2
NaCl = 15.0
Reactor
•
Figure 7 , Mass Balance for Full Scale Hypochlorite Generator
(based on data of 6.17.71). All flows in pounds per hour
-------
ing and evaluation period all of the components have performed satis-
factorily.
System Economics
The overall economics are dependent on the size of the plant and the
local costs of electrical power and salt. However two detailed cases
will present an outline of the present economic picture. The first
case is the actual installation of the twelve cell module at Somerville
and the other will be a representative facility designed for continuous
one ton per day production.
The field test unit was designed for installation in the Somerville
Marginal Conduit Pretreatment Facility. This facility is being con-
structed to divert combined overflow, which was being dumped into the
Mystic River basin, into the tidal portion of the river. All overflow
is to be screened and disinfected with hypochlorite. The conduit will
act as the detention chamber with a 9.5 minute contact time for the 5
year storm and progressively longer times for less severe loads.
The Somerville interception station has been designed on the basis of
240 hours per year or storm water overflow. The average flow of this
overflow will be 90 MGD and the desired dosage level will be 4.3 ppm of
sodium hypochlorite. The consumption of hypochlorite will be 32,200
pounds of chlorine equivalent.
A balance was made to determine the fraction of time the unit should
operate. A high capacity unit, which runs a small fraction of the time
would refill the storage tank rapidly but the capital cost would be high.
A low capacity unit running almost constantly would decrease the unit
cost but the refill rate might not be 'sufficient to handle two storms
in close succession. A trade-off indicated a unit with a capacity of
15 Ibs/hr would be near optimal. This unit would operate approximately
25% of the time.
Based on the volume and frequency of storms the generator system was
designed for a capacity of 15 pounds of chlorine equivalent per hour.
Since the yearly consumption is 32,200 pounds per year, the system will
be running 2,400 hours per year or 27% of the time.
The yearly operating expenses of the hypochlorite generator at this
installation are shown in Table 1.
Based on a yearly capacity of 32,200 pounds this results in an operating
net of 9.1C/lb of chlorine equivalent. At that location the quoted price
for delivered hypochlorite is 15£ per pound of chlorine equivalent.
19
-------
Table 1
Yearly Operating Cost of the Hypochlorite Generator at Somervilie,Mass.
Electricty
Demand Charge - 21 KW @ $20.6/KW $1,298.00
Usage Charge - 4,270 KWH/mo. @1.8£/KW 863.00
Salt 32.2 tons/yr @ $40/ton $1,290.00
Water 13.50
Process Water 50,000 gal/yr @ $0.15/gal 7.50
Cooling Water 40,000 gal/yr @ $0.15/gal 6.00
Maintenance 325.00
Replacement membranes 45.00
Replacement Anodes 30.00
Labor @ $5.50/hr 250.00
Total Operating and Maintenance Costs $2,926.50
There are several factors in this application which make the production
cost per unit of hypochlorite high at this location. Between the demand
charge and the relatively low power usage in this application the cost
per KWH is over 2.5£. The salt used at the Somerville installation is
a purified grade delivered in 100 pound bags. The cost per ton of salt
is $40. Using a standard grade of bulk delivered salt the cost would
drop below $20 per ton.
A more representative cost of operation of a generator installation is
that estimated for a unit with a capacity of 1000 pounds per day operating
350 days per year. Economic assumptions in this estimate are: incremental
power charge of l.OC per KWH including demand charge, bulk delivery of
salt at $15 per ton, negligible cost of process water (as indicated by
the previous example), 175 hours of labor at a yearly cost of $1000, and
amortization over 20 years at 5%.
The estimated costs under these assumptions are shown in Table 2. At an
annual production rate of 3.5 x 105 pounds this results in an operating
cost of 4.06C and a capital amortization cost of 1.37C per pound of
available chlorine.
Capital costs will be higher with the electrochemical generator than
with purchased hypochlorite. While capital costs can be estimated with
20
-------
Table 2
Yearly Operating Costs of 1000 Ib/day Hypochlorite Generator
Electrical Costs $5900
Salt Cost $5250
Maintenance $3075
Membranes 1,245
Anodes 830
Labor 1,000
Total Operating & Maintenance Costs $14,225
Annual Amortization of Capital $ 4,800
some accuracy, there is a wide divergence in methods of capital amort-
ization. Estimated capital costs are shown in Figure 8.
It can be seen from the above illustrative cases that the hypochlorite
generator is far more economical than delivered sodium hypochlorite and,
in some cases, is almost competitive with liquid chlorine. Although
the comparative costs vary depending on factors of size and location,
the system offers significant economic advantages in a majority of
cases.
21
-------
to
rt
H
H
&
C
•H
-P
O
u
f-)
td
•P
•H
U
0)
4J
-P
W
W
10
10'
10'
10
Figure 8.
10 10 10
Capacity of Unit - Pounds of Chlorine Equivalent per Day
Estimated Capital Cost of Hypochlorite Generator
10
-------
SECTION III
LABORATORY DEVELOPMENT
Throughout the development of the hypochlorite generator, three design
criteria were considered to be of prime importance. These criteria
were safety, economy, and simplicity of operation. These three prin-
cipals were responsible for the path of development which the hypo-
chlorite generator took in the transition from a 2 by 2 inch laboratory
cell to the present system, which by modularization could be expanded
to produce tons of chlorine equivalent per day.
Operational safety of the system was the most important consideration
in the development of the hypochlorite generator. Any practical method,
based on present technology, of producing sodium hypochlorite from
sodium chloride results in the generation of chlorine gas, hydrogen
gas and sodium hydroxide. Each of these products presents a potential
hazard. To minimize these hazards the system was designed: to contain
a minimum inventory of hydrogen, chlorine or caustic; to minimize the
possibility of contact between hydrogen and chlorine; to provide a safe
means of disposal of the hydrogen gas; and to shut down automatically
if any malfunction should occur.
The operating expense of the system depends primarily on the efficient
use of two commodities, salt and electrical power. Real salt utilization
is defined as the fraction of salt fed to the cell, or cell stack which
is decomposed. Efficient use of salt requires that this be as high as
possible. The efficient use of power requires a high current efficiency
and a low cell voltage. Current efficiency as used here is the ratio of
current theoretically required to that actually required per unit of
product hypochlorite. The minimum values of current and voltage are
343 ampere hours per pound of chlorine and 2.15 volts per cell, the
reversible cell potential. Any practical cell will operate above these
values.
A factor which can affect the economics of the system is the cost of
maintenance and replacement parts. If the system were to undergo re-
placement of electrodes or other components frequently, this would have
an undesirable effect on the operating cost and possibly even decrease
the overall capacity of the unit. It is then highly desirable to have
long-life components, provided the lifetime is consonant with initial
cost.
With regard to simplicity of operation, the hypochlorite generator
diverges significantly from industrial practice. Most industrial
caustic-chlorine cells require highly trained personnel as operators.
Operation of these cells is based on delicate flow balances and close
23
-------
monitoring of the systems operation. Since the hypochlorite generator
is to be designed for unattended operation, it must be susceptible to
automatic control including start-up and shut-down.
Electrochemical Cell
In a caustic chlorine cell, brine solution is fed into the anode com-
partment where chloride ion is discharged at the anode forming chlorine
gas.
2 Cl" -» Cl t + 2e~ (1)
^
At the cathode, water is electrolyzed producing hydrogen gas and
hydroxyl ions.
2 HO + 2e~ - 2 OH~ + H t (2)
^ £*
Current is carried through the electrolyte between the electrodes
principally by sodium ion (Na+). A separator is usually placed be-
tween the electrodes. The purpose of the separator is twofold. First,
it provides a barrier that prevents the hydrogen and chlorine gases
from coming into contact with each other. Second, it provides a
hydraulic barrier to prevent hydroxyl ions from migrating into the
cathode compartment.
A number of small scale cells were built and tested utilizing a variety
of geometric configurations, electrode materials, and diaphragm materials.
The small scale test apparatus is shown in Figure 9. A series of tests
was run on each of these cells to determine current efficiency, oper-
ating voltage, and materials reliability.
The size of the experimental apparatus was chosen such that the pro-
duction rate of chlorine would not be hazardous in the laboratory and
yet the cell would be large enough to illustrate the effectiveness of
materials and modes of construction, and performance would be repre-
sentative of full scale operation.
The materials of construction for those parts other than electrodes in
contact with chlorine gas or hypochlorite were Lucite or chlorinated
polyvinyl chloride. The Lucite spacers allow visual inspection of the
interior of the cell to determine ease of gas-electrolyte separation and
electrolyte level. The experimental cell volume is 75 cm3, and the
active electrode area is 30 cm2 for most tests.
The flow rates of water and brine to the cell were provided by positive
displacement metering pump capable of + 1% accuracy.
24
-------
Figure 9. Small Scale Test Cell and Control Apparatus
-------
In many single cell tests reference electrodes were introduced into the
cells to allow separation of the voltage across the cell into anode
and cathode potentials and resistive potential drop. Saturated calomel
electrodes were used. These were connected ionically to the active
electrode surfaces by Luggin capillaries. The various potentials in the
cell were measured with a vacuum tube voltmeter.
Current efficiency was determined from the rate of production of free
chlorine in the anode compartment expressed in gram equivalents per
Faraday. The method of analysis of the anode compartment effluent is
the lodometric Method (1).
This experimental program showed that the most economical and practical
design was a two compartment membrane cell with independent water feed
to the cathode. The best anode was found to be the recently developed
DSA (dimensionally stable anode) and the most acceptable cathode was
found to be mild steel.
Based on the very favorable results from the small scale cell testing,
work was begun on developing and testing a full scale model. This was
followed by testing of a two-cell stack and finally the 12 cell module
was constructed. Experimental work on each phase of this development
is described in subsections below.
Cell Configuration
The internal geometry of the electrochemical cell is a compromise be-
tween the conflicting requirements that the resistance of the electro-
lyte path be low and that hydroxyl ion (OH~) generated at the cathode
not migrate to the anode. In a high current density caustic-chlorine
cell, a separator prevents bulk mixing of the solutions surrounding the
two electrodes.
Migration of hydroxyl ion into the anode compartment can reduce the
current efficiency of the cell in several ways. The pH of the anolyte
usually runs between 3 and 4. Significant migration of OH" into the
anode compartment causes this to increase and increases the occurrence
of oxygen evolution at the anode.
H20 -» 1/2 02 t + 2H+ + 2e~ (3)
Increased pH also increases the amount of hypochlorite in the anolyte
by affecting the equilibrium:
Cl2(aq) + H20 £ HOC1 + Cl" + H+ (4)
26
-------
which increases the probability of chlorate formation:
3HOC1 -» CIO ~ + 2C1~ + 3H+ (5)
The products O2 and 0103" are essentially useless for disinfection.
The electrochemistry of chlorate formation is discussed in detail in
References (2) and (3).
Basically, three cell configurations were studied, the diaphragm cell,
the membrane cell and the three compartment cell containing both
membrane and diaphragm.
Use of an open cell without diaphragm was briefly considered. However
current efficiencies observed ranged from 25% to a high of 55%. High
velocity operation which was not tested, would increase the efficiency,
but it would not be possible to produce hypochlorite of sufficient con-
centration for practical storage.
Diaphragm Cell
The distinctive feature of a diaphragm cell is a porous diaphragm,
usually asbestos, which separates the anode and cathode compartments.
This is shown in Figure 10. Brine is fed into the anode compartment
and flows through the diaphragm into the cathode compartment. The
hydraulic flow through the diaphragm is' set at a rate that prevents the
hydroxyl ions from migrating into the anode compartment. Current
efficiencies around 90% can usually be obtained.
The diaphragm cell does however have an important disadvantage. Since
chlorine and caustic are both extremely corrosive materials only a few
diaphragm materials will survive in the cell environment. Traditionally
asbestos has been used but it lacks mechanical strength. Other materials
which exhibit high corrosion resistance, i.e.. Teflon, if fabricated
into a diaphragm, usually exhibit high electrical resistance which is
undesirable.
Another drawback of diaphragm cells is the necessity for close hydraulic
control. There are two flows out of the anolyte compartment, chlorine
gas product and anolyte liquid flowing through the diaphragm. To pre-
vent chlorine gas flow across the diaphragm and to avoid flow of anolyte
liquid out with the chlorine gas, a hydraulic balance must be maintained
at all times. Since the pressure drop across the diaphragm increases
with time, skilled operators are needed to monitor and adjust the
typical caustic-chlorine cell.
27
-------
Chlorine Gas
Caustic + Hydrogen
Anode
Compartment
Anode
Cathode
Compartment
Cathode
Brine
Figure 10. Diaphragm Cell with Sheet Electrode
ne Gas "*
Anode
1
^ <-!-. J
Anode j *"
^ * + ?
f \^
Brine
— > Causti
bhode
apartment
athode
Caustic t Hydrogen
Figure 11. Diaphragm Cell with Expanded Cathode
28
-------
Probably the most critical component in a diaphragm cell is the diaphragm
itself. The diaphragm must be chemically inert, have uniform pore size,
prevent the caustic generated at the cathode from migrating into the
anode compartment, and have low electrical resistance. For many years
the caustic-chlorine industry has used slurry-deposited asbestos
diaphragms against a porous steel cathode. Because of the requirement
of minimum cell maintenance, the slurry deposited diaphragm, which
requires skilled manpower at the installation is undesirable. Con-
sequently, considerable effort was expended in the search for a suitable
throw-away diaphragm which could be easily replaced.
A program was undertaken to evaluate the ability of various diaphragm
materials to keep hydroxyl ions away from the anode and their short-
term resistance to degradation. The chlorine current efficiency of a
cell is directly proportioned to ability of the diaphragm to act as a
hydroxyl barrier. The current densities studied in these tests ranged
from 100 to 500 ASF which covers the probable range of interest. The
diaphragm materials under consideration were porous polyethylene (Porex
Corp.), porous Teflon (Chemplast), woven Teflon (Stern & Stern, Inc.),
fuel cell and electrolytic grades of asbestos (Johns-Manville Corp.),
polypropylene battery separators (National Lead Corp.) and polypropylene
matte (Pellon Corp.).
These diaphragms were tested in cells like those shown in Figure 10 with
sheet electrodes placed 2.5 cm apart. The brine feed rate in these
tests corresponded to 50% nominal salt utilisation. The anolyte was
maintained at a slight hydraulic head, 1 to 2 feet of water, above the
catholyte. The hydraulic head required to maintain proper flow through
the diaphragm depended on the porosity of the diaphragm.
The various diaphragm materials studied were supported with polypropylene
cloth on the catholyte side of the diaphragm and glass cloth on the
anolyte side. Only those diaphragm materials having considerable
strength were used without support. Table 4 presents the experimental
data obtained from a number of porous diaphragm materials. Figure 12
shows the internal voltage drop for various diaphragm materials over
the current density range 200 to 500 ASF, A minimum voltage curve is
included as a reference in this figure. The minimum voltage is that
which would exist across the electrode gap with concentrated brine at
the operating temperature and current density without any diaphragm.
The fuel cell grade asbestos contributes almost nothing to the overall
internal resistance. The desirable properties of this material were
confirmed in a variety of subsequent tests.
Electrochemical performance of this type of cell is characterised by the
relationship between cell voltage and current density and by the re-
lationship between current efficiency and salt utilisation. The effect
of current density on salt utilisation is small. These relationships
are shown in Figures 13 and 14.
-------
Table 3
Evaluation of Diaphragm Materials
Diaphragm Material
Teflon Cloth
ii ii
ii ii
ii H
ii ii
Porous Teflon (Zitex E-1002)
II tl
11 11
Porous PVQYuasa 0202G)
H ii ii
ii 11 ii
Porous PVC (Yuasa 620G)
H ii 11
H ii ti
H H ti
Polypropylene Battery Separator
(Polysep 40)
Electrolytic Asbestos 0.006"
ii 11 H
ti 11 11
Electrolytic Asbestos 0.010"
11 ii ii
ii n 11
ii n M
n it n
Electrolytic Asbestos 0.018"
ii n n
ti n ii
ii ii ii
n n n
it it it
11 ii n
n n n
ti n M
n n n
Electrolytic Asbestos 0.023"
n n M
n n n
11 n n
Current
Density
ASF
100
150
200
350
400
200
300
400
200
200
300
200
200
200
300
300
200
300
400
200
200
300
400
500
200
200
200
300
300
300
300
300
400
400
200
300
400
500
Total Cell
Voltage,
volts
5.0
-
5.4
7.6
7.2
5.2
5.9
6.5
4.6
4.8
6.0
5.3
5.1
5.1
5.7
5.9
5.3
6.2
-
5.1
5.7
6.7
7.2
8.0
4.35
-
-
4.85
-
-
4.3
4.3
5.9
6.9
5.1
6.0
6.9
7.7
Resistive
Voltage Drop
volts
_
-
-
-
-
1.75
2.28
2.80
1.4
1.35
2.4
1.7
1.25
1.65
2.1
2.65
1.5
2005
-
1.55
1.85
2.55
2.65
3.6
-
-
-
-
-
-
_
-
-
-
1.6
2.65
3.65
4.35
Current
Efficiency
(%)
55.5
59.
56.
52.1
46.4
25.7
66.7
66.7
64.2
73.7
17.5
47.7
51.6
76.4
36.0
26.1
84.3
74.9
71.5
7008
80.9
77.7
84.2
75.8
59.0
52.5
91.
52.3
76.5
67.5
57.5
73.0
45.1
66.0
85.3
82.2
77.5
74.4
•
ft
1
1-1
H O
Q) O
O
25
40
35
30
48
36
42
52
73
50
56
40
38
38
50
49
39
43
51
42
35
45
53
62
59
60
60
59
53
53
60
60
59
60
43
40
55
63
30
-------
Table 3 (Continued)
•Evaluation of Diaphragm Materials
Diaphragm Material
Fuel Cell Asbestos 0.010"
, Current
Dens ity
ASF
100
200
200
200
200
200
200
300
300
300
300
400
400
400
400
500
500
Total Cel]
Voltage,
volts
4.8
5.9
4.8
4.8
4.9
5.8
4.8
5.55
5.7
6.2
5.8
6.1
6.6
7.1
6.2
6.8
7.0
ft
o
Resistive
Voltage Di
volts
-
-
_
1.6
1.37
1.5
-
2.3
1.9
2.35
-
2.5
2.25
2.6
-
2.95
Current
Efficiency
(%)
86.
89.5
70.5
86.5
87.4
87.2
93.9
85.7
92.5
78.5
90.4
83.0
92.3
63.9
95.5
87.6
91.7
ft
EH
U
H 0
H
U
40
55
51
51
48
40
44
53
50
46
45
56
48
56
51
60
55
31
-------
in
4J
rH
Q)
U
A 0.006" Electrolytic Asbestos
O 0.010"
D 0.023" " "
O 0.010" Fuel Cell Asbestos
A Polysep 40
• Yuasa 0202
-------
w
-P
8.
I
Platinized Titanium Anode
Stainless Steel Cathode
Fuel Cell Asbestos Diaphragir
Operating Temperature
~ 50° C
I
100
200 300 400
Current Density - ASP
500
Figure 13. Typical Voltage-Current Curve for
Sheet Electrode Diaphragm Cells
33
-------
0.8
0.6
u
0)
•r4
U
•H
£ 0-4
3
U
0.2
0.0
0.2
0.4 0.6
Salt Utilization
0.8
1.0
Figure 14. Effect of Salt Utilization on Current Efficiency
for Sheet Electrode Diaphragm Cell
34
-------
In the region of interest the relationship between cell voltage in volts,
V, and current density in ASF, i, is linear and can be closely repre-
sented by:
V = 3.30 + 0.0077i (6)
Experimentally it is much more convenient to use nominal salt utilization
than real salt utilization. Nominal salt utilization is defined as the
fraction of salt fed to the cell which would be decomposed if the
current efficiency were 100%. It is then equal to real salt utilization
divided by current efficiency. "Salt utilization" without qualification
refers to nominal salt utilization in this report.
Expressing current efficiency, 77 , and nominal salt utilization, f ,
fractionally, the data in Figure 14 can be expressed by:
?7 = 1.01 + 0.33 log (1-f) (7)
These expressions are necessary for the computer optimization discussed
in Section IV.
Diaphragm Cell with Expanded Cathode
Use of an expanded cathode is a. refinement of the sheet electrode cell
construction. This is shown in Figure 11, page 28 . The drawing is
placed next to that of the sheet electrode cell to facilitate comparison.
This configuration allows the cathode to be placed next to the diaphragm
which reduces the inter-electrode spacing and the internal resistance
of the cell.
To permit leak-tight assembly in the cell, an ethylene-propylene rubber
gasket was formed around the expanded nickel cathode.
The current efficiencies with this cell configuration were generally
lower than the 85 to 90% observed with the sheet electrode (at 200 ASF
with a fuel cell asbestos diaphragm). A high concentration of hydroxyl
ion exists at the cathode surface. If the diaphragm sees this con-
centration, the current efficiency may be lowered. Several techniques
were used to provide a slight separation between cathode and diaphragm
in order to aid transport of hydroxyl ion away from the diaphragm.
Results of these tests are given in Table 4. In all tests the cell
voltage is lower than that of the sheet electrode cell.
The best current efficiency obtained was 85% using a polypropylene
screen spacer between the diaphragm and the cathode. The polypropylene
screen is constructed with vertical ribs 0.025 inches thick and with
35
-------
0.25 inch spacing between ribs; the horizontal crosslinks are 0.008
inches in diameter. In addition to the spacer, a row of 1/8 inch holes
was punched along the top and the bottom of the cathode to aid the
circulation of electrolyte through the space between the cathode and
diaphragm.
Table 4
Effect of Separation on Expanded Cathode Cell
Type of Separation
between cathode
and diaphragm
No separation
No separation
0.020" separation
0.025" Saran Screen
0.025" polypropylene
screen
Current Density
ASF
200
200
200
200
200
Operating
Cell Voltage Temperature
volts °C
3.7
4.2
4.2
4.65
3.60
55
65
63
Current
Efficiency
35
76
76
53
85
It was felt that the relatively small mesh of the expanded cathode
(Exmet #5/0-1/32" holes) might have caused low current efficiencies.
An expanded cathode having somewhat large holes (Exmet #3/0-1/16" holes)
was tested. The larger holes were expected to permit greater circulation
of electrolyte around the cathode and reduce the hydroxyl ion concen-
tration near the diaphragm. Current efficiencies with the larger mesh
ranged from 81 to 92%.
In this configuration, when the cell was not being operated, the nickel
cathode suffered severe attack. This attack was more pronounced with
the expanded cathode than with the sheet cathode, presumably because of
the great number of surface irregularities. Without an applied potential,
the cell becomes a galvanic cell with the nickel becoming the anode.
Salt deposits on the cathode were found to contain large amounts of nickel
by emission spectrum analysis. If this sort of cell were to be used, it
might be necessary to drain the cells, or maintain a potential between
the electrodes when not in use.
The relationship between salt cut and current efficiency was determined
using the best expanded cathode and diaphragm configuration. This curve
is shown in Figure 15. The relationship between salt utilization, f,
and current efficiency, 77 , is reasonably expressed by:
?7 = 1.10 + 0.805 log (1-f)
(8)
36
-------
1.0
0.8
0.6
u
c
Q)
-H
D
•H
H
0.2
0.0
I I
0.2
J I
0.4
0.6
_J I
0.8
1.0
Salt Utilization
Figure 15. Effect of Salt Utilization on Current Efficiency
for Expanded Cathode Diaphragm Cell
-------
The current efficiency at high salt cuts using the expanded cathode is
slightly higher at modest salt utilization and much lower at high salt
utilization than with the plate cathode.
The relationship between voltage, V, and current density, i, can be
approximated by:
V = 3.3 + 0.0036i (9)
The computer optimization indicated that this cell would operate at
higher current densities but at higher total costs than the cell with
two plate electrodes.
Difficulties with Diaphragm Package
To permit simple field replacement of the diaphragm, it was considered
desirable to have a gasket diaphragm package. A frame of ethylene-
propylene rubber (EPR) was compression molded around a fuel cell asbestos
sheet. However, during a life test, it was noticed that carbonaceous
deposits tended to collect on the diaphragm. The most likely source of
this material was from the EPR portion of the composite diaphragm. The
EPR rubber showed definite signs of chemical attack. Apparently the
carbon filler in the EPR was being released into the anolyte and was
being filtered on to the diaphragm.
Several possible solutions to the problem were tested. The EPR rubber
was coated with a 3% solution of vinyl chloride-aerylonitrile copolymer
(dynel) in acetone. This coating appeared effective for only two days
operation. Two other procedures were more successful. The EPR rubber
portions of the composite diaphragm were replaced entirely with acrylic
polymer. In a similar procedure acrylic polymer was substituted for
EPR only where wet chlorine would be likely to contact the gasket'
thereby retaining the resilient qualities of EPR necessary for effective
sealing. Fabricating satisfactory diaphragm packages using acrylic
resin was impeded by bubble formation in the plastic. This resulted
from the presence of a volatile component in the acrylic molding resin.
Bubble formation was virtually eliminated by reducing the molding
temperature below that suggested by the manufacturer.
These diaphragms failed after approximately 200 hours of life testing.
The mode of failure, however, was due to lack of wet strength on the
part of the asbestos rather than to failure of the acrylic gasket.
The most serious problem was the rather unpredictable variation of wet
strength from sample to sample. Early in the program a piece of fuel
cell grade asbestos had a life in excess of 1200 hours. Later testing
of the same type of diaphragm resulted in a significantly shorter life.
38
-------
The supplier, Johns-Manville Co., indicated that the samples were the
same type and quality during the test period. The mode of failure of
the diaphragm was sagging of the asbestos which left holes in the
diaphragm.
Before computer optimization was applied to the system, the hypochlorite
generator was expected to operate only 200 to 450 hours/year. With
annual replacement of the diaphragm on a routine basis, 1200 hours would
be a satisfactory diaphragm life but 200 hours would be insufficient.
The computer study indicated that production of hypochlorite solution
at lower rates would be cheaper overall. This lower rate of production
would require that the cells be running from 1000 to 4000 hours/year.
Under these conditions a much longer diaphragm life, in fact longer than
the best life observed,is required.
Several attempts were made to create a composite sandwich diaphragm in
which an integrally bonded Teflon cloth would support the asbestos paper
to overcome its lack of wet strength. Several problems were experienced
in making these composite diaphragms. One problem was excessive flash-
ing of rubber into the diaphragm portion of the composite. This ruined
the porosity of the resulting diaphragm. Flashing was overcome by care-
fully measuring the amount of uncured rubber in the center of the mold.
Movement of the Teflon cloth was overcome by cutting the Teflon cloth
larger than the outside dimensions of the mold and maintaining tension
on the cloth during molding. After the molding operation was completed,
the excess Teflon cloth was trimmed away. An intermittent problem was
the apprearance of pressed wrinkles in the asbestos cloth which resulted
in holes in the resulting diaphragm.
Another approach to the asbestos support problem was to sew the Teflon
cloth and diaphragm together with Teflon thread in a criss-cross fashion
to create a quilted sandwich of Teflon cloth and asbestos. This failed
because of the holes which the needle left in the asbestos sheet.
During the time when the diaphragm cell was under active consideration,
no totally satisfactory diaphragm was developed. The necessity of this
development was obviated when the benefits of the membrane cell, de-
scribed below, were discovered.
Membrane Cell
Iii a membrane cell, Figure 16, the anode and cathode compartments are
separated by a cation exchange membrane. This membrane must have good
chemical resistance since it is exposed to caustic on one side and
freshly generated chlorine on the other side. There is no direct
hydraulic flow from the anode to the cathode compartment. The only
water that crosses the membrane is endosmotic water which is associated
with the ions being transferred. The anolyte liquor and the chlorine
39
-------
Chlorine -f Spent Brine
T-1
Anode
Compart n
Anoc
rua •
1
lent '
1
Na+ 1
1
VJ
^1
I
cat:
Com;
Cathc
K
I
iode
partment
:>de
Wa-hRi
r
Caustic + Hydrogen
Water or Dilute Brine
Figure 16. Membrane Cell with Expanded Electrodes
Chlorine Gas
Spent Brine
Brine
° y
I
1
1
1
1
Na+— f-
Anode |
Membrane |
i " **
1
-*•
i i
i 1
i I
i |
1
|
1 1
~*l
i i > Cathode
i '^ ^
i | Diaphragm
\^-\^
Caustic + Hydrogen
Figure 17. Three Compartment Cell
40
-------
exit from a common port and are sent to a gas-liquid separator where
the product chlorine gas is separated from the depleted brine. Since
there is no delicate pressure differential to be maintained, this cell
does not have the hydraulic flow and control problems associated with
a diaphragm cell.
A cathode feed stream is needed to sweep out the sodium hydroxide that
is produced. This stream can consist of either fresh water or the dis-
engaged anolyte solution. In the experiments performed with membrane
cells, feed rates to the cathode compartment were used which resulted
in formation of caustic solution between 0.5 and 4.5 N in the cathode
compartment.
The major difficulty with a membrane cell is that fact that state-of-art
membranes allow more back-migration of hydroxyl ions into the anode
compartment than a good diaphragm, with a corresponding reduction of
current efficiency. With the membrane cells used in this study, the
optimum concentration of caustic in the cathode compartment has been
about 10%, or slightly over 2N, and the resulting current efficiency,
based on gaseous chlorine, was approximately 65% to 75%. The reason
that "gaseous" is emphasized is that in a cation membrane cell a large
fraction of the OH~ that migrates across the membrane reacts with the
chlorine product, producing hypochlorite. The hypochlorite ion is
reported as "Free Chlorine" in standard tests but it does not necessarily
appear in the product hypochlorite. Thus when the total anode product
(liquid and gas) is analyzed, the indicated current efficiency is higher
than when the gaseous product alone is measured. The conclusion that
may be drawn from these results is that although the membrane cell over-
comes the primary disadvantages inherent in diaphragm cells, these
advantages are obtained at the expense of current efficiency. Two modes
of operation were investigated:
• Mode I - spent anolyte fed to the cathode compartment
• Mode II - water or dilute brine fed to the cathode compartment
In mode I saturated brine is metered to the anode compartment and the
exiting stream is separated into liquid and gas phases. The liquid
phase, which is depleted brine, is fed by gravity to the cathode com-
partment. This system has the advantage of being simple since it re-
quires only one feed pump and does not require flow balancing. There
are the disadvantages that caustic concentration cannot be varied
independently of salt utilization and that the salt that is not decomposed
in the anode compartment cannot be recovered.
A full series of tests on the small scale membrane cells operating in
mode I was made. The various relationships that were found in this
series of tests are outlined below.
41
-------
In this mode of operation the normality of the caustic produced is
strictly a function of the caustic current efficiency and the salt
utilization. The experimental relationship between caustic normality,
c, and salt utilization, f, is shown in Figure 18. The caustic
normality is almost zero at zero salt utilization and increases linearly
to approximately 4.0 at 100% salt utilization. This can be expressed
as:
c = 0.0697 + 4.20 f (10)
Product hypochlorite concentration is consequently also a unique
function of salt utilization.
Figure 19 shows the relationship between current efficiency and salt
utilization. The current efficiency decreases with increasing salt
utilization from a high of about 85% current efficiency at very low
utilization to a low of less than 65% at high utilization. Expressing
current efficiency as 77 , the data closely correlate to:
T\ = 0.879 - 0.241 f (11)
The two dependent variables, caustic normality and current efficiency,
are then related by:
77 = 0.883 - 0.0574 c (12)
In mode II water or dilute brine is fed to the cathode compartment of
the two compartment membrane cell. In this type of cell, current
efficiency improves as the sodium hydroxide concentration in the cathode
compartment is lowered. Current efficiency also improves somewhat with
increasing chloride concentration in the catholyte. By maintaining a
separately metered feed to the cathode compartment, the current effi-
ciency can be controlled independently of the salt utilization.
Another advantage of this method is that the depleted brine can be
recirculated to the lixator for resaturation. Using this flow scheme
it is possible to obtain 100% salt utilization. Although this mode of
operation is more flexible than feeding the depleted brine to the
cathode, it is more complex and requires additional pumps and auxiliary
equipment.
A series of tests was carried out on the two-compartment cell with
membrane separator using various cathode feed compositions. The catho-
lyte caustic normality is determined largely by the volumetric feed rate
to the cathode compartment. The cell was operated over a range of
caustic normalities at three levels of sodium chloride concentration in
42
-------
§
•H
•P
10
M
-P
C
0)
o
C
O
U
U
-H
-P
W
u
o
Current Dansity 200 ASF
0.2
0.4 0.6
Salt Utilization
0.8
1.0
Figure 18. Dependence of Caustic Concentration on Salt Utilization
for Two Compartment Cell Operated in Mode I
43
-------
o
c
0)
•H
D
•H
W
+J
(D
U
1.0
0.8
0.6
0.4
0.2
0.2
0.4 0.6
Salt Utilization
o
o
o
0.8
1.0
Figure 19. Dependence of Current Efficiency on Salt Utilization for
Two Compartment Membrane Cell Operated in Mode I
44
-------
the catholyte feed (0, 3% and 5%) and at two levels of salt utilization
(50% and 90%). The purpose of these experiments was to determine the
performance relationships for a cell operated in mode II and to deter-
mine the best feed composition.
The effect of salt concentration in the cathode compartment feed is
shown in Figure 20. Over the range of caustic normalities from 1 to 4N,
use of 3% salt feed instead of water increases the efficiency by about
two percentage points and use of 5% salt feed increases it by about 6
percentage points. Since this salt is essentially unrecoverable,
electric power would have to be very expensive relative to the cost
of salt to justify use of salt water as an independent cathode feed.
In Figure 21 current efficiency is plotted against caustic normality
for two levels of salt utilization with water feed. The current
efficiency is higher for 50% salt utilization. However, this differ-
ence becomes small at 3N caustic concentration.
The same comparison is made with 5% salt feed in Figure 22. Although
both curves are higher than with water feed, salt utilization has no
distinguishable effect.
The relationship between cell voltage, V, and current density, i, was
determined for a membrane cell with expanded metal electrodes was
determined. At 75° C with water feed and caustic concentration of
2.3N this is expressed by:
V = 2.3 + 0.0061 i (13)
Three Compartment Cell
In an attempt to combine the advantages of a membrane cell and a
diaphragm cell into a single unit, the three compartment cell was
designed. This is shown in Figure 17, page 40. The anode compartment
is like that of the two compartment membrane cell. Brine is fed in
the normal fashion and both the depleted brine and chlorine gas exit
from a single port in the top of the anode compartment. Feed to the
cathode is introduced into the center compartment from which it flows
through a diaphragm into the cathode compartment.
This type of cell can be run in the same two modes as the two compart-
ment membrane cell.
In iRode I the anolyte liquor, separated from the chlorine gas, is fed
to the center compartment. The pressure in the center compartment
establishes its own level. This can be measured in a head tube which
also serves to release any trapped gases from the center compartment.
45
-------
1.0
0.9
0.8
u
0)
•H
D
•H
m
4-1
w
-P
c
(U
3
u
0.7
0.6
0.5
1-
o
o
+ 5% Nad solution
£ 3% NaCl solution
O - water
Two compartment membrane cell
Current Density - 250 ASF
Salt Utilization - 50%
0.4
0.3
1234
Caustic Concentration - N
Figure 20. Effect of Salt Concentration in
Cathode Feed on Cell Performance
46
-------
c
0)
•H
O
•r-l
W
4->
C
-------
0)
•H
U
•H
4-1
4-1
3
U
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
+ 50% Salt Utilization
A 90% Salt Utilization
Two compartment membrane cell
Current density - 250 ASF
Cathode Feed 5% NaCl
01 2 3 4 E
Caustic Concentration - N
Figure 22. Effect of Salt Utilization on Cell Performance
48
-------
This design combines all of the advantages of the diaphragm and
membrane cells while eliminating most of their difficulties.
There are several advantages to this design. There is no need to
maintain a delicate flow balance in the anode compartment. After
disengagement, the anolyte liquid flows by gravity into the head tube
where pressure balance is maintained by the level of the liquid. There
are several resistances to back-migration of hydroxyl ions. The
hydraulic flow from the center compartment to the cathode prevents
the najority of hydroxyl ions from migrating into the center compart-
ment, A large fraction of the ions that do back-migrate into the center
compartment react with the dissolved chlorine in the anolyte entering
the center compartment to produce hypochlorite. Consequently the
hydroxyl ion concentration at the membrane is much lower than in the
two-compartment cell.
The relationship between salt utilization , f, and current efficiency,
77, is shown in Figure 23. This can be expressed as:
7? = 0.34 log (1-f) + 1.03 (14)
Comparing these results to those obtained for the two compartment cell,
Figure 19, it can be seen that the current efficiency for low salt
utilization is much higher with the three-compartment cell. At higher
salt utilization (above ~ 75%) the current efficiencies become quite
similar for both cells. These results indicate that, although the
three-compartment cell exhibits a significant advantage over the two-
compartment cell at low salt utilization, this advantage decreases if
the economics require the system operate at high salt utilization.
The relationship between voltage, V, and current density, i, for this
cell is given by:
V = 2.3 + 0.0054 i (15)
For the same current density this cell, with a 1/2 inch interelectrode
spacing has a slightly higher operating voltage than the two-compart-
ment cell.
In mode II water or dilute brine is fed to the center compartment. The
advantage to this design over any other design involving a diaphragm
is that the diaphragm is not exposed to chlorine gas. Consequently the
diaphragm material can be selected because of the uniformity and
appropriateness of its permeability rather than because of its resist-
ance to chlorine.
The cell in this mode proved to be essentially inoperable. With water
feed the cell violtage was quite high and variable. This was due to the
49
-------
1.0
o
n
0)
•H
u
•H
IH
M-i
H
-P
C
o;
3
U
0.8
0.6
0.4
O
0.2
I 1
0.2
0.4
0.6
0.8
1.0
Salt Utilization
Figure 23. Dependence of Current Efficiency on Salt Utilization
for Three Compartment Cell Operated in Mode I
50
-------
high resistance of the center compartment which in operation con-
tained a very dilute solution of caustic. Use of 3 or 5% sodium
chloride feed reduced but did not eliminate these difficulties.
No reliable operating data were obtained for this mode of operation.
The difficulties experienced in the laboratory strongly suggested that
the three compartment cell in mode II would not have the required
reliability in the field.
Membrane Evaluation and Membrane Package Fabrication
Of the cation-exchange membranes presently available in commerce the
only type which would withstand exposure to both concentrated caustic
and nascent chlorine was a fluorinated membrane from DuPont. The
product is designated XR for a membrane with Teflon cloth backing and
R for a similar membrane without backing. (No membrane development
was performed on this contract.)
Early samples of this membrane had a life-in-service of about 100 hours.
These samples were of two-ply construction and the mode of failure was
spalling of the resin from the membrane. Subsequent samples have been
of one-ply construction. These have withstood up to 3000 hours of
cumulative testing without observable changes in properties. A life-
in-place of two to three years is anticipated.
This latter type of membrane was used in all membrane cells for which
data are reported.
Just as it was desirable to have a packaged diaphragm for field re-
placement in a diaphragm cell, it is desirable to have a packaged
membrane in a membrame cell. The package used at the present time
consists of a fairly rigid frame surrounding the membrane. Since the
membrane swells upon both heating and immersion in an aqueous solution,
the membrane must be stretched prior to lamination in the frame. A
special j ig was made to stretch and hold the membrane during the
fabrication process. Once the membrane is inserted in the jig and
stretched to the desired extent, it is laminated between PVDC sheets
using a special high temperature PVC resin cement.
A careful specification of membrane construction overcame a difficulty
encountered with the membranes for the cell. The membranes were
specified to be 24" x 24" with the active ion exchange film covering
the entire surface. When the membranes were received, they were
approximately one inch undersize and the edges of the film were buckled
and wrinkled. It was impossible to use this membrane since the buckled
edges would not allow the cell to seal properly. The membrane is a
special laminate of an active film on a Teflon cloth backing, and the
buckling difficulty arises from edge effects in producing large
laminated areas. To resolve this problem membranes were specified
51
-------
which had an active area, measured between borders of the resin fabric,
only slightly larger than the active area of the cell and also had
a wide border of unlaminated Teflon cloth. This design allows the
membrane to be stretched in place after which the excess cloth is cut
off.
Comparison of Cell Configurations
Economic analysis using the data accumulated from the small scale cell
testing indicated that most economical operation would be with high
saltutilization. In this light the two compartment membrane cell had
a combination of advantages not possessed by any other configuration.
This configuration was the simplest to fabricate, operate and maintain.
The caustic concentration could be varied independently of salt util-
ization. This allows the cell to be run at high salt utilization with
reasonable current efficiency. The membrane provides an impermeable
barrier between the hydrogen and chlorine. The membrane proved to be
a reliable separator that did not have to be cleaned or replaced as was
the case with diaphragms. Test membranes have been run for over 3000
hours without any sign of deterioration or functional impairment. The
system was simple to operate. After setting the flows to the anode
and cathode compartments no further adjustments were necessary.
The current efficiency of a diaphragm cell is highly dependent upon the
amount of salt that is decomposed in the cell. The relationship between
current efficiency and salt utilization is shown in Figure 14. It is
apparent that attempting to utilize a large portion of the salt feed
results in very low current efficiencies. In industrial cells only 50%
of the brine is decomposed since the efficiency drops off rapidly be-
yond this point.
The study of diaphragm established four significant drawbacks for th
present application: poor salt utilization, difficulty of control,
high maintenance and operational hazards. It is impossible to operate
a diaphragm cell at a high salt utilization without suffering severe
losses in current efficiency. The diaphragm cell required careful
adjustments of head levels and flow rates to maintain a proper pressure
balance across the diaphragm. The diaphragms tended to become fouled
and plugged with time, requiring the cell to be dismantled for diaphragm
replacement. Since the diaphragms were made of porous material which
tended to degrade with time, a significant possibility of mixing hydrogen
and chlorine existed.
The three compartment cell offered the safety and stability of a membrane
cell combined with the high current efficiency of a diaphragm cell, pro-
vided it could be operated at moderate salt utilization. At high salt
utilization the efficiency rapidly approached that of a two-compartment
membrane cell. The internal resistance however was higher because of
the greater interelectrode distance.
52
-------
The most economical mode of operating the two-compartment membrane cell
is with water feed to the cathode compartment. Although there is a
certain increase in current efficiency when there is sodium chloride
in the cathode feed, this increase in efficiency does not offset the
additional salt cost. The increase in current efficiency obtained by
running at a 50% versus 90% salt utilization is comparatively small.
If the spent anolyte is to be discarded, the economics indicate that
the high salt utilization is desirable. It might be possible to re-
cycle the spent brine to the saturator but it is unlikely that the
savings due to the increase in current efficiency would offset the
additional equipment cost and complexity.
Anode Material
A number of anode materials were tested. Among them were the Englehard
3000, 7000, and 9000 grades of platinized titanium, Asahi Kasei platin-
ized titanium, resin impregnated graphite, and Electrode Corp. DSA
electrodes.
The anode potentials determined in the laboratory are shown in Table 5
for these materials. A Luggin capillary was inserted in the cell and
the tip placed within one tip diameter of the electrode surface. The
potential between the working electrode and a saturated calomel electrode
was measured with a high impedance vacuum true voltmeter. The electrode
with the lowest operating voltage was the expanded DSA electrode. The
expanded electrode allows the Luggin capillary to be inserted from the
back so it does not interfere with the current flow within the cell.
The measurement of voltage of the other electrodes may be slightly in-
fluenced by the presence of the Luggin capillary but only by a few milli-
volts at most.
These comparative tests were not run for long enough to provide any in-
formation about aging characteristics except that the graphite was
observed to ablate or spall at current densities above 150 ASF. The
usable lifetime of the DSA electrodes can be assumed to be about 5 years
based on their use under similar conditions in the chlorine industry.
Cathode Material
Six materials were evaluated as cathodes. These were Hastelloy C,
Monel 400, 316 stainless steel, Inconel 625, nickel 200 and mild steel.
The cathode potentials of these materials at various current densities
are shown in Table 6. The voltages shown are referred to a saturated
calomel reference electrode and the method of test is similar to that
used for anodes described above.
53
-------
Table 5
Comparison of Anode Operating Potential (Volts)
Current
Density
(ASF)
200
300
400
500
Platinized Titanium
(Englehard Incl.)
Series 3000
1.85
(38°G)
1.85
(45°C)
1.90
(53°C)
1.95
(58°C)
Series 2000
1.85
(38°C)
1.95
(39°C)
2.00
(43°C)
-
Series 9000
1.75
(38°C)
1.85
(44°C)
1.95
(50°C)
-
Platinized
Titanium
(Asahi)
1.85
(38°C)
2.00
(44°C)
2.10
(51°C)
2.20
(58°C)
Resin
Impregnated
Graphite
2.25
(40°C)
2.50
(46°C)
-
-
DSA on Expanded
Titanium
(Electrode Corp.)
1.00
(48°C)
1.00
(53°C)
1.00
(65°C)
-
* Measured against saturated calomel reference electrode
-------
Table 6
Comparison of Cathode Operating Potential (Volts)
Ln
(Ji
Current
Density
ASP
200
300
400
500
Stainless Steel
316
1.80
(39°C)
1.95
(45°C)
2.10
(48°C)
2.20
(55°C)
Hastalloy C
1.55
(38°C)
1.50
(45°C)
1.75
(40°C)
-
Monel
Alloy 44
1.65
(38°C)
1.80
(45°C)
1.95
(50°C)
-
Nickel
200
1.50
(35°C)
-
1.60
(56°C)
-
Inconel
625
1.50
(39°C)
-
1.60
(44°C)
-
Mild
Steel
-
1.65
(45°C)
-
-
* Measured against saturated calomel reference electrode
-------
Unlike the selection of anodes, there was no obvious choice for cathode
material. Hastelloy, Inconel and nickel have the lowest potentials but
only by small fractions of a volt. Intermittent cell operation causas
the cell to be inactive for long periods of time during which time the
cathode is exposed to highly oxidative species. Hastelloy C has the
best corrosion resistance. However it is expensive and difficult to
fabricate. If the cell were to have sheet cathodes, Hastelloy C might
well be used. However, for an expanded cathode it is a less desirable
choice. The other materials would all require a trickle current for
cathode protection during the periods that the cell was not active.
After provision is made for a trickle current, then the advantage lies
with the cheapest material. The final decision to use mild steel
cathodes was probably strongly influenced by the fact that the chlorine
industry has found that mild steel provides the lowest cost cathodes.
Gasketing Material
A gasket, in the form of a flat gasket, is used as a seal between each
of the compartments in the cell. Several different gasketing materials
have been evaluated for use in the anolyte section of the cell. This
compartment provides the most corrosive atmosphere. Tygon, butyl rubber,
silicone rubber, and Viton all exhibited severe attack by the corrosive
environment. Teflon exhibits no attack or swelling, however it is fairly
rigid and is subject to cold flow. Neoprene appears to undergo a super-
ficial attack on exposed surfaces but this is not progressive. Neoprene
gaskets have been in intermittent service for at least 6 months.
Ethylene propylene rubber (EPR) is claimed to have superior chlorine
resistance. This material appears satisfactory based on testing over
several months. This material appears satisfactory based on testing
over several months. EPR was selected as the anode gasket material with
neoprene as an alternative.
In industrial practice butyl rubber has shown exception resistance to
caustic attack. This material has been successfully in the cathode
compartment. However, to prevent the possibility of gaskets being inter-
changed during cell assembly, the anode gasketing material will be used
in the cathode compartments.
Full Scale Unit Design and Performance
After determining the most favorable cell configuration, a single full
scale cell was built and tested. The full scale cell had an active
area of 2.2 square feet. The cell compartments were constructed from
high temperature PVC which had shown in tests to have satisfactory re-
sistance to wet chlorine. The full scale cell and test rig appear in
Figure 24.
56
-------
01
Figure 24. Two Cell Stack and Associated Test Apparatus
-------
For the full scale cell, end plates were designed which would distribute
the sealing pressure evenly over the gasketed areas and yet would be
light in weight. These were fabricated from 1/4" by 3" stainless steel
stock with the short dimension against the end block. The end plate
design can be seen in Figure 24.
End blocks were constructed from 1" thick lucite to allow for visual in-
spection of gas release and foaming inside the anode and cathode compart-
ments. Exposure to chlorine rapidly turned the anode end block opaque.
The cathode end block, however, remained transparent through the testing
of this cell.
A fairly conventional optimization was used to determine the operating
current density. A tradeoff was made between capital costs which de-
crease with increasing current density, and power costs, which increase
with increasing current density. With computer techniques, second order
effects like the effect of current density on current efficiency and
discontinuous functions like electric power rate structures can be con-
sidered in this optimization. For the design case the optimum current
density was approximately 240 amperes per square feet (ASF) and the cost-
current density curve was quite shallow.
In this cell heat build-up can affect operating current density. The
energy which corresponds to the excess of operating voltage over 2.15
volts is converted into heat. The energy which relates to the 2.15
volts is converted into chemical energy of reaction. Since the quantity
of brine fed to the anode compartment is constrained by the necessity of
obtaining high salt utilization and the quantity of water fed to the
cathode compartment is limited by the requirement of producing a con-
centrated hypochlorite solution, the operating temperature of the cell
shows a steady increase with current density. Since high temperature
produces high conductivity, this is desirable but not to the point that
the electrolyte boils. A desirable temperature is 85 to 90° C. With
85% salt utilization, a current density of 240 ASF produces an operating
temperature of 88° C. With other economic constraints it is possible
that the current density should be held below the economic optimum to
avoid too high a cell temperature.
The current efficiency of the large cell was found to be higher than that
of the small cells (80% vs 65%) . This was confirmed by a large amount
of operating data near 240 ASF. Since there is no external circulation,
essentially all convection past the electrode and membrane surfaces is
due to internal gas-lift recirculation. This is improved both by in-
creasing the height of the electrode and increasing the depth of the
compartment. Both the anode and the membrane can be affected by polar-
ization layers and it is not clear at this time whether one or both
are affected.
Analysis showed that the salt utilization has little effect on either
cell voltage or current efficiency. The cell voltage decreases slightly
58
-------
as the salt cut is increased because the temperature rises as the brine
flow role is lowered. There is simultaneously a slight increase in the
current efficiency since less dissolved chlorine is carried out by the
anolyte.
Varying the water feed to the cathode has a more significant effect than
varying the brine feed. As the amount of water fed to the cathode is
decreased the caustic concentration is increased. The higher hydroxyl
ion concentration causes a greater back migration across the membrane
which decreases the cell efficiency. This relationship is shown in the
table below.
Table 7
Effect of Varying Water Feed to Cathode
NaOH Naod % System Temp
Normality Normality Efficiency Cell Voltage °C
2.00 1.83 77 3.83 86
2.75 2.50 72 3.65 90
The cell voltage is lower in the case of the higher caustic concentration.
This decrease in voltage is due to the increased temperature and the more
conductive catholyte.
Since it is desirable to utilize at least 80% of the salt feed, the
present limits of caustic concentration are between 2.00N and 2.85N.
The lower limit is economic in natu-re since, as the concentration of
the hypochlorite solution decreases storage and handling costs increase
rapidly. The upper limit is presently set by the heating effects in
the cell stack. Several methods of cooling are presently being con-
sidered which may allow the production of more concentrated sodium
hypochlorite.
The next step in the development program was the design and testing of
a two cell stack. The two cell stack was used to test and evaluate
methods of electrical connection and hydraulic manifolding of the cells.
It was clearly desirable to have a series-connected cell stack. There
are several alternative means of internal connection. Leading the
current out of each cell is expensive, wasteful of power and conducive
to leaks. Bipolar sheet electrodes are ideal from the standpoint of
electrical connection, but the operating cell voltage is much higher
because of the increased inter-electrode spacing. The solution selected
was to weld circular lugs on the back of each electrodes. These are
connected through the cell walls. For the 16" by 20" electrode, two lugs
were used.
59
-------
Current efficiency of the two cell stack was found to be essentially the
same as that of the single large cell. The voltage per cell, V, at
60° C is related to current density, i, by:
V = 2.4 + 0.00595 i (16)
The slope of this curve decreases slightly with temperature.
Effect of Calcium and Magnesium
All commercially available sodium chloride contains some calcium (Ca)
and magnesium (Mg) salts as naturally occurring impurities. Further,
although usually smaller, quantities of these materials are introduced
into the brine through the feed water. In the electrochemical cell
these impurities can cause difficulties by formation of insoluble
hydroxides. A diaphragm cell is most sensitive because the precipitates
tend to plug the diaphragm. However, difficulties can occur because of
scale formation on membrane and electrode surfaces. In general cheaper
grades of salt contain more impurities. This problem was only given a
cursory examination during the present program. The decision was made
to use quite pure salt at the Somerville installation despite its higher
cost to avoid operating difficulties. Further research is needed to
determine what concentrations of calcium and magnesium, and of sulfate
can be tolerated in the feeds to the hypochlorite generator and what
means of pretreatment would be practical.
Brine prepared from rock salt contains between 150 and 300 ppm of Ca
and 0-50 ppm of Mg . Brine from food grade salt usually contains
5-25 ppm of Ca++ and 0.5 ppm of Mg++ . Experimentally some difficulty
was encountered with fouling of the diaphragms because of the presence
of calcium and magnesium. After several weeks of operation the three
compartment cells showed a gradual rise in the pressure drop across
the diaphragms. When the cells were disassembled, a white gelantinous
precipitate was found on the center compartment side of the diaphragm.
Analysis of this residue showed it to be primarily calcium and magnesium
hydroxides. Although food grade salt was used to prepare this brine,
the pH of the center compartment was sufficient to form precipitates.
One method of pretreatment is the injection of sodium carbonate and
sodium hydroxide into the brine saturator. These additives cause the
precipitation of calcium as calcium carbonate and magnesium as magnesium
hydroxide. The solubility products of these two components are:
[Mg^l x [OH~] = 1.2 x lo"11 (18° C) (17)
[Ca++] x [CO31 = 0.8 x 10~8 (25° C) (18)
60
-------
Thus if the solution in the lixator is made . 01N in each of hydroxide
(OH~) and carbonate (003") the equilibrium concentrations of calcium
and magnesium would both be below 1 ppm.
A study was performed to determine the efficacy of this brine treatment
procedure. Three samples of brine were prepared, the first using rock
salt, the second using rock salt treated with one gram per liter of both
sodium hydroxide and sodium carbonate, and the third using food grade
salt. The results are shown in Table 8.
Table 8
Comparison of Brine from Three Grades of Salt with Pretreated Brine
++ ++
ppm Ca ppm Mg
Untreated rock salt 195 3
Treated rock salt 16 0.2
Food grade salt 14 0
Pretreated brine < 5 < 1
It can be seen that the treatment method effects over a tenfold improve-
ment in hardness and produces a brine somewhat better than that obtained
with food grade salt. The practicability of this or similar treatment
on a large scale has not been investigated.
Reactor Development
In the reaction of caustic with chlorine:
2 NaOH + Cl_ -» NaCl + NaOCl + H0O (19)
to form hypochlorite, a slight excess of caustic must be maintained.
If the reaction is allowed to proceed in an acid environment, caused by
an excess of chlorine, the following reactions occur.(5)
Cl + HO -» HOC1 + Cl" + H+ (20)
2* £
HOCl ^ OC1~ + H+ (21)
2HOC1 + OC1~ -» C10~ + 2C1~ + 2H+ (22)
61
-------
The rate of reaction (22) is very sensitive to pH. Since this reaction
converts a weak acid (HOC1) to a strong acid (HC1), the sequence is
autocatalytic and can produce boiling with the release of oxygen, HCl
gas and chlorine. It also converts hypochlorite to chlorate. Sodium
chlorate (NaClO3) is very stable and is useless for disinfection. Thus,
it is essential to avoid over-chlorination and the inherent problems that
are associated with it.
The need to avoid over-chlorination requires use of a concurrent reactor
rather than a counter-current reactor. In a counter-current reactor the
nearly neutral hypochlorite at product concentration would encounter a
large stream of entering chlorine. The stream would rapidly become
acidic producing localized over-chlorination. In the co-current mode
as the hypochlorite becomes more concentrated and as the pH of the
stream decreases, the chlorine is depleted.
The product hypochlorite exits from the reactor through a side-arm via
gravity overflow. Any inerts in the gas stream are vented through the
top. The gravity overflow design offers a simple method of flow control
which avoids the need for level controllers or flow regulators. The
reactor constructed in this manner is capable of operating over a wide
range in flow rates without the necessity of changing any control
settings.
In the design of the reactor, consideration must be given to removal of
the heat of reaction. The heat of reaction of chlorine with caustic is
about 600 BTU per pound of chlorine. Since the rate of chlorate
formation increases with increasing temperature, it is necessary to cool
the reactor to suppress this reaction. Production of 10% hypochlorite
would be accompanied by a 60° F rise in temperature without cooling.
Production of 5% hypochlorite would cause a 30° F rise. The cooling
can be most easily accomplished by circulating cooling water through
a titanium tubing coil incorporated in the reactor.
Inefficiencies in the cell actually help avoid the possibility of
chlorate formation in the reactor. Generation of caustic at the cathode
is essentially stoichiometric. The anode under any conditions generates
some chlorine and some chlorine is lost by solution in the spent brine
stream. Consequently there is always a small but useful excess of
caustic over chlorine in the cell products. This will help ensure the
stability of a well-run reactor.
Initial tests were conducted with a tall, narrow packed bed reactor.
Co-current upflow design performed very well. The glass helices used
for packing gave very good liquid-vapor contact with essentially no un-
reacted chlorine gas escaping with the inerts. There is also no evidence
of channeling or localized over-reaction.
The column run without cooling confirmed the temperature rise calculations,
In the initial studies the temperature of the starting reactants was
62
-------
approximately 25° c and the product had a temperature of between 55 -
60° C. In the actual production unit where the reactants are fed to
the reactor at a temperature in the area of 70° c cooling will be
necessary. Above 80° C chlorate formation becomes quite significant.
Several additional runs with using the small scale lab reactor pro-
duced varying results. The reaction efficiency ranged between 50% to
90% with product temperatures ranging between 45° to 80° C of a feed
temperature of 23° C. The high temperatures were caused by over-
chlorination which also accounts for the poor reaction efficiency of
some of the runs.
In order to control the temperature an internal cooling coil for the
reactor was fabricated from 1/4" titanium tubing.
Since the reaction is quite rapidr good gas-liquid contact is not ex-
tremely important. The design consisting of a titanium coil in an
annular space described in Section II was selected in preference to
the packed bed because of its relative simplicity.
Hydrogen Disposal
For a chlorine production rate of 15 Ib/hr, 1.6 cubic feet per minute
of hydrogen is produced. Although this flow is quite small it is de-
sirable to provide a safe means of disposing of the hydrogen to elimin-
ate the possibility of fire or explosion. After considering various
methods of disposal it was decided that the easiest and most efficient
way to dispose of the hydrogen would be to vent it into a duct where it
would be diluted with a large volume of air and exhausted to the outside.
Hydrogen in air has an upper explosive limit of 74.2 volume percent and
a lower explosive limit of 4.0 volume percent. (4) Allowing a safety
factor of three, a hydrogen concentration of 1.3% in the duct would
require 150 CFM of air. In a four inch diameter duct, this flow pro-
duces a velocity of about 30 feet per second with an essentially
negligible pressure drop. The resulting Reynolds number is 64/000
ensuring turbulent flow and thorough mixing of the hydrogen with air.
This system requires a small blower installed in the duct vented to
the outside. A safety interlock automatically shuts down the cell if
the fan were to fail.
63
-------
SECTION IV
SYSTEM OPTIMIZATION
The basis for optimization of the hypochlorite generator is minimum
total cost to the user. A computer program was developed to permit
the optimization of operating variables, like salt utilization and
current density for a particular cell configuration and for specific
economic inputs. The program, as presented in Figure 25, applies to
a diaphragm cell with economic inputs suitable to the Boston University
Bridge site. Use of this program for other cell configurations re-
quires that the appropriate internal relationships, given in Section HI/
be used. Economic inputs which depend on location are indicated with
a dagger (t) at the end of the line, which is not part of the program.
The program is written in Fortran IV for GE400 timesharing. Other
systems should require at most modification of the input and output
statements.
In this program the chlorine production rate, J, is expressed in
theoretical amperes, that is, the current which would be required if
the generation process were 100% efficient. Values are assumed for the
independent variables, salt utilization and current density. Capital
equipment costs (cells, rectifier, brine storage tank, pumps and cell
cooling, if required) are calculated. These are multiplied by the
capital amortization factor, F:
F = ± (23 )
1 - (1 + I)'N
which depends upon the interest rate, I, and the number of amortization
periods, N. Provision is made for amotizing the cells, which are likely
to have a shorter lifetime than the other equipment, over a different
amortization period. The equations used are listed in Table 9. When
these quantities are summed and multiplied by an industrial mark up
factor (1.7 was used) the result is the cost per year of owning the
equipment. Annual operating costs (salt, energy, electrical power
demand charge and water) are calculated by the equations in Table 10.
The nomenclature for these equations and the computer symbols used are
listed in Table 11. Total annual cost is the sum of amortized capital
costs and operating costs.
The program stores the results of the first calculation, increments the
independent variables separately and recalculates. If a new set of
independent variables gives a lower annual cost than the stored value,
the old results are replaced by the new results.
65
-------
Figure 25. Hypochlorinator Optimization Program
REAL INT>NC*NR*M,K1tfM*K*KV/*KWCHS*KWHR
K=.00482
IN1=.07
NR=30.
NC=30.
ANHR=240. t
PUMS=3-
ENRS=1.
COOLSP=1.
DEKCH=1.
DOSE=4.3 t
OFLOto=90. t
SALS=.022
Q = 6.
Df = 1 • 2
SELFAC=1.7
CELS=120.
M=.003
V=2.4
FOD=.1
FOYSTR=0.
FM = 1 .
KUM=1.
CDM=1.
FT2M=1.
ANCOSM=10E10
C
C COMPUTATION OF SEVERAL VARIABLES INDEPENDENT OF
C THE DO-LOOP
C
ANHRS=ANHR/FOD
DESCAP=118.98*DOSE*OKLOV*FOD
RNC=-NC
AMC=INT/(1.-(1.+INT>**RNC>
RNR=-NR
AMR=INT/(1.-(1.+INT)**RNR)
DD=DF*DESCAP
C
C
C THIS IS THE BEGINNING OF THE DO-LOOP TO DETERMINE THE
C MINIMUM ANNUAL COST. ALL PARAMETERS DEFINED IN THE
C DO-LOOP ARE DEPENDENT ON THE VALUES OF CD AND/OR F.
C
C
DO 1* I=l>10
11=1*50
DO 2, J=5,9
RJ=J
RJ=RJ/10.
CD=II
66
-------
Figure 25. (Continued)
F=RJ
40 CONTINUE
CE=.146*AL06Cl.-F>+l.Ql
CEL=/( 1000.*CE)
REC=(KW/40.)**.67*1340.*AMR
REC=2.*REC
FT2=DD*100./(100-*CE*CD>
ROKSAL=ANHRS*K*DD/Q/F
IFCROKSAL.GT.20000.> GO TO 50
BRSTR=(ROKSAL/10000.>**.37*1900.*AMR
GO TO 51
50 BRSTR=(3800*CROKSAL/20000.>**.355+1500.>*AMR
51 PUM=DD/F*K*PUMS*AMR
HYT=C GO TO 55
PRINT 99
99 FORMAT( " KW>500 ">
GO TO 55
53 DEMCS=0.
GO TO 52
54 DEMCS=2.9*12.***.2*AMR*2500«
KWHR=KW*ANHR/(FOD*12.>
IFCKWHR.LE.800.) GO TO 60 '
IFCKWHR.GT.800..AND.KWHR.LE.1200.) GO TO 61 t
IFCKWHR.GT.1200..AND.KWHR.LE.4000.) GO TO 62 f
IF(KV/HR«GT.4000. .AND.KV/HR.LE.20000. )GO TO 63 f
IFCKV/HR.GT.20000..AND.KWHR.LE. 120000. )GO TO 863 f
IF(KVHR.GT.120000.) GO TO 865 t
60 KfcHCS=KUHR*.059
GO TO 64
61 KViHCS = 800.*.059+CKV/HR-800. >*.05
GO TO 64
62 KtoHCS = 800«*.059+. 05*400. •••• 0275*(KV/HR-1 200. )
GO TO 64
63 KWHCS=800.*.059+20.+2800.*.0275+(KUHR-4000.)*.016
67
-------
Figure 25. (Continued)
GO TO 64
863 K\vHCS=400.2-KKtoHR-20000>*.012
GO TO 64
865 KV/HCS=1600.2+(KVHR-120000. )*.0097
GO TO 64
64 ENEKG=DEMCS+KfoHCS*12.
IFCCD.LE.250. ) GO TO 70
UATER=< 1 . 1E-6*CD**2+CD*1 .4E-4-. 1 1 >*FT2*DF
IFCVATER.GT.22. > GO TO 71
COOLl=(FT2*COOLSP+78. >*AMR
COOL2=(FT2*COOLSP+270. )*AMR
GO TO 72
71 COOL1=(FT2*COOLSP+1 1 . 6*WATER** .495 >*AMR
COOL2=(FT2*COOLSP+40-2*WATER**.495)*AMR
72 WAT=VJATER*7.2E-05*ANHRS t
A 1 =COOL 1 *SELFAC+WAT
A2=COOL2*SELFAC
IFCA1.LT.A2) GO TO 73
COOL1=0.
GO TO 80
70 COOL1=0.
VAT=O.
73 COOL2=0.
IFCFOD.LT. 1 . ) GO TO 81
80 ANC01=(CEL+REC+BRSTR+PUM+COOL1+COOL2)*SELFAC+SALT+ENERG+
&WAT
ANCOS=ANCO 1 +LABOR*SELFAC
GO TO 82
81 ANC01=(CEL+REC+BRSTR+2.*PUM+COOL1+COOL2+HYT>*SELFAC+
&SALT+ENERG+\»(AT
C
C
C THE MINIMUM ANNUAL COST OF OPERATION IS
C DETERMINED BY RETAINING MINIMUM VALUES OF:
C F,CD,Kto,FT2 AND ANCOS AS THE PROGRAM GOES
C THROUGH THE ITERATIONS IN THE DO-LOOP.
C
C
ANCOS=ANCO 1 +LABOR*SELFAC
1202 FORMAT ( 5X, F4. 3> 5X* F5. 0* 5X* F8 • 2)
82 IFCANCOS.GE.ANCOSM) GO TO 90
ENERGM=ENERG
SALTM=SALT
CELM=CEL
RECM=REC
CEM=CE
FM=F
CDM=CD
68
-------
Figure 25. (Continued)
FT2M=FT2
ANCOSM=ANCOS
90 CONTINUE
2 CONTINUE
1 CONTINUE
PRINT 100
1 00 FORMAT (////, 6X* "F", 6X* "CD'S 8X* "MINANCOST'S 8X, "KWS 6X*
&"FT2",//>
PRINT 101*FM*CDM,ANCOSM,KV/M*FT2M
101 FORMAT(F9.2,F8.1,5X,F9.1*F12.l,F10-2*//>
CLCOST=ANCOSM/(DOSE*ANHR*OFLOfo*8.33/24. )
PRINT 102,CLCOST
TCEL=CELM/AMC
102 FORMAT<" THE COST OF GENERATED CL IS'SF6.3*" $/LB"*/)
103 FORMAT<3X,"THE ANNUAL ENERGY COST IS $",F8-2>
104 FORMAT(3X^"THE ANNUAL SALT COST IS S",F8.2)
105 FORMAT(3X>"THE TOTAL CELL COST IS £"*F8.2)
106 FORMAT(3X*"THE TOTAL RECTIFIER COST IS SM,F8.2)
107 FORMAT(3X>"CELL COST-$/YR="^F8.2*3X>"RECT COST-$/YR="*F8.2>
PRINT 103*ENERGM
PRINT 104,SALTM
PRINT 107*CELM*RECM
TREC=RECM/AMR
PRINT 105,TCEL
PRINT 106*TREC
CAPCOS=(ANCOSM-ENERGM-SALTM-V/ATM)/AMR
108 FORMAT(3X*"TOTAL CAPITAL EQUIPMENT COST IS $">F8.2>
PRINT 108*CAPCOS
109 FORMAT (3X,"PERCENT CURRENT EFFICIENCY=",F3.0)
CEM=CEM*100.
PRINT 109*CEM
95 CONTINUE
END
69
-------
Table 9
Amortized Capital Equipment Costs
J.D-C -F
c c
Cells: A = :
c 1-77
1340 p°-67.F
Rectifier: A = —
r 40
1900 V °'37oF
Brine Storage Tank: A^ = 1n nnn
O _LU jf UUU
* = ( n + 1500) F
v 20,000 r
J-D-k -C 'F
s p r
Pumps : A = —
** s
Cooling Spacers: A = ( — + 78)F
3 m 1-77 r
J-D-C
** s
SUPPORTING EQUATIONS
J-D (V0 + iR)
P =
1000 7?
J-D-t -k
r s
Vs = N,-f
a
77 = 1.01 + 0.146 In (1-f)
* If more than 2000 Ibs of salt is to be stored.
** The cost of using municipal water is compared to that of using
runoff. The lower total cost alternative is selected.
70
-------
Table 10
Annual Operating Costs
J-k «t -C.,
Salt: A^ = ^-£-
J-t -C (V + iR)
Energy: °
e 1000
Demand Charge: A = C -P
_3
Water: A = 7.2 x 10 W-t
w r
SUPPORTING EQUATION:
(1.18 x 10~6i2 + 1.4 x 1Q~4 i - 0.11)
w , _
71
-------
Table 11
Nomenclature Used in Economic Analysis
Symbol
AU
£c
Ce
C,
D
F,
1
J
Nc
p
R
V
vs
w
Computer
Designation
BRSTR
GEL
DEMCS
KWHCS
COOLl
PUM
REG
SALT
COOL2
WAT
GELS
DEMCH
ENRS
SALS
PUMS
COOLSP
DF
AMC
AMR
CD
DESCAP
K
Q
KW
M
ANHRS
V
ROKSAL
WATER
CE
Significance
Amortized Brine Storage Cost, $/year
Amortized Cell Cost, $/year
Kilowatt Demand Charge, $/year
Energy Cost, $/year
Amortized Cost for Cell Cooling Using
Municipal Water, $/year
Amortized Brine Feed Pump Cost, $/year
Amortized Rectifier Cost, $/year
Salt Cost, $/year
Amortized Cost for Cell Cooling Using
Runoff, $/year
Cost of Municipal Water for Cooling,
$/year
Factory Cost of Cells, $/ffJ
Demand Charge, $/kilowatt
Electrical Energy Price, $/kilowatt
hour
Salt Price, $/lb
Cost of Brine Feed Pump,
$/lb salt hr"1
Factory Cost of Cooling Spacers,
$/ft2
Design Contingency Factor
Amortization Factor for Cells,,year~-'-
Amortization Factor for Other
Equipment, year"-*-
Salt Utilization, expressed as
decimal
2
Current Density, amps/ft
Rate of Usable Chlorine Production,
theoretical amperes
Conversion Factor, 0.00482 Ib
NaCI/ampere hour
Number of Salt Deliveries per year
Power Rating of Rectifier,kilowatts
Effective Internal Resistance of Cell,
ohm-ft2
Fraction of Time in Operation,
hrs/year
Extrapolated Zero-Current Cell Voltage,
volts
Salt Storage Capacity, Ibs of salt
Cooling Water Required, gals/ft2
Current Efficiency, expressed as decimal
72
-------
The costs of operating labor and some capital equipment like the press
that holds the cells do not appear in this calculation. These are not
strong functions of the independent variables and do not affect the
optimization.
An example of the use of this program is shown in Figure 26. This was
the result of a paper study of the use of the hypochlorite generator
at the Boston University Bridge Facility. The assumption was first
made that the unit would operate only during storm overflow conditions
(Point 1). In this application, since the power required for pumping
during overflow conditions is very much more than the power demand of
the cells, the hypochlorite generator does not carry a demand charge
if it is not operating during the storm. This mode of operation
(Points 2 to 6) does require product storage tanks which on an amortized
basis cost much less than the demand charge. In this mode the ratio of
production rate to consumption rate becomes a third independent variable.
This is used as the abscissa in the figure. To give an indication of
how the operating variables are related, the conditions at optimum are
given for the various points calculated. For reference at this in-
stallation an average overflow of 100 MGD is expected for 300 hours per
year. The dosage rate is 3 ppm. The quoted treatment cost does not
include estimated fixed costs of $700 per year which amounts to
0.056C/Kgal. The treatment cost with electrochemically generated
hypochlorite can be compared to that using truck-delivered solution
which was estimated at 0.7£/Kgal-at this location
73
-------
1.2
0)
5
IH
o
H
n)
-P
w
o
u
-p
q
1.0
0.8
0.6
0.4
0.2
data i
point ASF
1
2
3
4
5
6
400
200
250
400
600
700
P
KW
255
10
25
55
163
310
Area
ft2_
125.
11.7
22.
27.6
46.1
71.1
0.8
0.85
0.85
0.85
0.85
0.80
operating during
storm
truck delivered
hypochlorite
not operating
during storm
0 0.2 0.4 0.6 0.8 1.0
Ratio of Production Rate to Consumption Rate
Figure 2Q. Treatment Cost at the Boston University Bridge Facility
74
-------
ACKNOWLEDGMENTS
Initial development of the hypochlorinator was performed under Pro-
gram 11023 DAA of the Environmental Protection Agency, which was
monitored by Allyn C. Richardson. The field unit was constructed
under Grant 11023 DME to the Metropolitan District Commission under
the supervision of Benjamin Fink, Director of Park Engineering and
Chief Park Engineer. The opportunity to locate the hypochlorinator
at the Boston University Bridge facility was kindly provided by Frank
Burgin, Chief Engineer, Construction Division, M.D.C. The support and
encouragement of William Rosenkranz of the Environmental Protection
Agency throughout the project is especially appreciated.
75
-------
SECTION V
REFERENCES
1. "Standard Methods", 12th Ed. American Public Health Assn.,
New York (1965).
2. Ibl, N. and D. Landolt, J. Electrochem. Soc. 115, 713 (1968).
3. Foerster, F., Trans. Am. Electrochem. Soc., 46^ 23 (1924).
4. Lewis, B.and G. von Elbe, "Combustion, Flame and Explosions of
Gases," Academic Press, New York (1951).
5. Sconce, J. S., "Chlorine, Its Manufacture, Properties and Uses,"
Reinhold, New York (1962), p. 81 et. seq.
6. "Report on B.U. Bridge Storm Detention and Chlorination Station,"
C. A. Maguire and Associates, Boston, Mass. (1967).
77
-------
SECTION VI
PUBLICATIONS AND PATENTS
Michalek, S. A. and Leitz, F. B., "Cloromat - On-Site Generator of
Hypochlorite," delivered at the 44th Annual Conference, Water Pollution
Control Federation, San Francisco, California(1971).
Leitz, F. B., Accomazzo, M.A. and Michalek, S. A., "Development of
Electrochemical Hypochlorite Generator," to be presented at the 141st
National Meeting Electrochemical Society, Houston, Texas (1972).
No patentable items were developed on this contract.
79
-------
SECTION VII
GLOSSARY
Brine - A concentrated solution of salt. Here brine refers to a
solution of sodium chloride at or near saturation.
Chlorine Equivalent - The amount of hypochlorite which would be re-
quired to perform the same oxidation as a given amount of chlorine.
Current Efficiency - The ratio of material produced to that which would
be theoretically produced for a given current flow. The current
efficiency for an overall process is the product of the efficiencies
of the subordinate processes.
Endosmosis - The passage of water through a membrane caused by flow
of current through that membrane.
Expanded Metal - A sheet of metal which has been slit and stretched
to several times its original length. The resulting holes are roughly
diamond shaped.
Internal Cell Resistance - The resistance to current flow presented
by the electrolytic path between the electrodes. In practice it is
convenient to use the slope of a plot of voltage versus current for
internal cell resistance.
Ion Exchange Membrane - A sheet of plastic which has been chemically
reacted to contain fixed sites of one charge. An electric current
passed through this kind of membrane will be carried almost exclusively
by species of the opposite charge.
Over-chlorination - The condition in a caustic-chlorine reactor caused
by either local or general excess of chlorine over caustic.
Salt Utilization (Nominal) - The fraction of chloride fed to the anode
compartment which would be converted to chlorine of the anode process
were 100% efficient.
Salt Utilization (Real) - The fraction of chloride fed to the anode
compartment which is converted to chlorine.
81
-------
SECTION VIII
APPENDICES
APPENDIX I. BOSTON UNIVERSITY BRIDGE FACILITY
Delays in construction of the Somerville Marginal Conduit have made
initial testing at an alternative site desirable. The temporary site
is to be the Boston University Bridge Storm Detention and Chlorination
Facility (B.U. Bridge) in Cambridge, Massachusetts. This facility was
designed to reverse the rapid degradation of the Charles River Basin (6) .
Its goal is to raise the basin to a water quality classification of "C",
that is, suitable for boating and fishing.
The portion of this project of interest to the present paper is the
provision for treatment of excess storm flow before disposal. During
storms, flows in excess of 146 MGD are routed to the B.U. Bridge facility.
There the mixed influent is screened and chlorinated with hypochlorite
solution. The dosage rate which is expected to be between 2.7 and 3.5
ppm will be controlled to yield 1 ppm residual at the outlet of the
tanks. Residence time in the tanks is 8 minutes at the design flow of
233 MGD. The treated effluent is discharged to the Charles River basin.
This facility was designed to use truck-delivered hypochlorite. It
contains storage facilities for 8000 gallons. The estimated annual de-
mand at this site is 50,000 Ibs of chlorine equivalent per year which
is 40% of the total design capacity of the field unit.
Results of the on-site tests will be reported by the E.P.A. upon com-
pletion of those tests.
83
-------
APPENDIX II. HYPOCHLORINATOR CELL FRAME AND FILTER PRESS DRAWINGS
Following are reductions of the drawings for the hypochlorinator cell
frame and various drawings for the filter press which was specially
designed to contain the electrochemical cell stack.
84
-------
00
Ul
Figure 27.
Hypochlorinator Cell Frame
-------
03
'6 r*,n
Figure 28. Cell Filter Press-Jack Support Frame (Drawing 5403-FP 203)
-------
00
Figure 29. Cell Filter Press-Stationary Platen (Drawing 5403-FP202)
-------
oo
CD
Figure 30. Cell Filter Press - Movable Plate (Drawing 5403-FP204)
-------
(V
co
"V _-n ,_ „.
Figure 31. Cell Filter Press - Assembly Drawing (Drawing 5403-FP201)
-------
1
Access/on Number
w
5
2
Subject Field &, Group
05D
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Title
Hypochlorite Generator for Treatment of Combined Sewer Overflows
1Q Author(s)
TFTV7 PRANTf R
MICHALEK, STEVEN A0
GREATOREX, JOHN L.
16
21
Project Designation
EPA Program 11023
DAA & Grant 11023 DME
Note
22
Citation
23
Descriptors (Starred First)
*Waste treatment, *water quality control, *water treatment, *Electrochemistry,
Process Design, Brine, Chlorine, Computer Programs, Combined sewers
25
Identifiers (Starred First)
*Electrolytic hypochlorite generator, *Hypochlorite, Electrolytic Cell,
Somerville, Massachusetts
27
Abstract
An advanced electrolytic generator has been developed for pn-site production of sodium
hypochlorite for disinfection of overflows from combined sewer systems. In this system
an electrochemical cell electrolyzes sodium chloride brine to chlorine gas and sodium
hydroxide solution, which are reacted immediately outside the cell to produce a 5 to 10%
sodium hypochlorite solution„ Significant advances in safety and economy have been
realized by use of a hydraulically impermeable cation exchange membrane. The most
critical components have operated for over 3000 hours with no deterioration of perform-
ance. The generator requires 1.6 KWH of electricity and 2.1 pounds of salt per pound of
sodium hypochlorite. The operating cost for systems larger than 500 pounds of hypo-
chlorite per day is projected to be 3 to 4 cents per pound of hypochlorite„ This cost is
significantly below that of truck-delivered hypochlorite solution,, Such economy of
operation should make the generator useful for a wide variety of water treatment
applications. The first field unit is scheduled for installation at a Metropolitan
District Commission facility, in Somerville, Massachusetts. (Leitz, Ionics)
Abstractor
FRANK
B.
LEITZ
Institution
Ionics,
Inc.
WR:102 (REV. JULY 1969)
WRSIC
SEND. WITH COPY OF DOCUMENT. TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 20240
* GPO: 1970-389-930
------- |