WATER POLLUTION CONTROL RESEARCH SERIES • 11010 EGO 01/71
PHOSPHORUS REMOVAL
BY
FERROUS IRON AND LIME
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. 20U60.
-------
PHOSPHORUS REMOVAL BY FERROUS IRON AND LIME
by
RAM) DEVELOPMENT CORPORATION
Cleveland, Ohio
and
THE COUNTY OF LAKE
Painesville, Ohio
for the
ENVIRONMENTAL PROTECTION AGENCY
Project No. 11010 EGO
January, 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price 75 cents
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does 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.
-------
ABSTRACT
When used in primary treatment, ferrous iron was effective in remov-
ing more than 80 percent of the phosphorus contained in sewage, with
spent pickle liquor a satisfactory source of the metal ion. Required
iron dosages ranged from the theoretical minimum of 2. 7 to a maximum
of approximately 3. 1 weights of ferrous iron per weight of phosphorus
contained in the sewage. When advanced means, such as filtration,
are available for more complete removal of the insolubilized phosphate
precipitate, phosphorus removals in excess of 90 percent are indicated.
To obtain optimum floe formation and to prevent carryover of soluble
iron in the effluent the pH of the sewage had to be adjusted to the range
of 7. 5 to 8. 0 by the addition of a strong base, such as lime. Required
lime dosage was 1.6 to 1.7 weights of lime per weight of iron. Total
iron in effluent was high, approximating 10 mg/1 Fe, and may require
filtration for removal.
Overall suspended solids removals of 61. 5 percent were attained over
a 23-month plant experimental program, with BOD and COD removals
of 61.6 and 55.3 percent, respectively.
The supplemental use of coagulant aids was generally not beneficial.
The combined chemical cost for ferrous chloride pickle liquor and
lime was $1. 09 per million gallons of sewage treated, for each mg/1
of contained phosphorus. The total chemical costs for treating a
sewage containing 13 mg/1 of phosphorus as typically received during
this work, would be $14. 17 per million gallons.
The weight increase in sludge solids due to the precipitation of phos-
phorus by ferrous iron and lime was approximately 100 pounds per mil-
lion gallons of raw sewage per mg/1 of contained phosphorus, with
enhanced removals of suspended solids representing an additional 420
pounds as sludge. The sludge volume increase varied to double that
obtained in normal treatment.
Filtration of the digested sludge proceeded normally and the precipitated
iron phosphate remained insolubilized through anaerobic digestion. Hy-
draulic overloading of the digesters adversely affected digestion and
subsequent settling, however, the resulting filter cake was adequately
stabilized.
This report was submitted in fulfillment of Grant Project No. 11010
EGO under the sponsorship of the Environmental Protection Agency,
Water Quality Office.
111
-------
TABLE OF CONTENTS
Section Page
I CONCLUSIONS 1
II RECOMMENDATIONS 3
III INTRODUCTION 5
IV MENTOR PLANT FACILITY ' 7
Description of Influent Sewage 7
Description of the Mentor Plant 12
Chemical Feeding and Handling Equipment 17
Sludge Handling 24
V DATA PRESENTATION AND EVALUATION 27
Ferrous Chloride Precipitation;
No Supplementary Additives 31
Lime Precipitation 34
Ferrous Chloride and Lime 36
The Use of Supplementary Additives 45
VI SEWAGE TREATMENT
WITH FERROUS IRON AND LIME 55
Chemical Costs 57
Equipment Costs 57
Treatment Costs 59
VII ACKNOWLEDGEMENTS 61
VIII REFERENCES 63
IX APPENDICES 67
-------
LIST OF FIGURES
Page
1 PLAN OF MENTOR PLANT FACILITY 8
2 TYPICAL DIURNAL FLOW
AND TOTAL PHOSPHATE VARIATION 10
3 WILLOUGHBY-MENTOR SEWAGE TREATMENT PLANT
FLOW SHEET 13
4 PHOTOGRAPH:
OVERALL VIEW OF THE MENTOR PLANT .... 15
5 PHOTOGRAPH:
VIEW OF THE WEST PRIMARY TANK 18
6 PHOTOGRAPH:
ENLARGED VIEW OF THE WEST PRIMARY TANK . 18
7 PHOTOGRAPH:
WASTE PICKLE LIQUOR RESERVOIR 20
8 PHOTOGRAPH:
VIEW FROM ABOVE IRON DELIVERY SPIGOT AT
RAW SEWAGE INLET CHANNEL 20
9 PHOTOGRAPH:
ORIGINAL PICKLE LIQUOR FEED PUMP 21
10 PHOTOGRAPH:
POSITIVE DISPLACEMENT PICKLE LIQUOR PUMP 22
11 PHOTOGRAPH:
VIEW OF IRON FEED PUMP REPLACEMENT ... 22
12 PHOTOGRAPH:
LIME FEEDER 23
13 PHOTOGRAPH:
VIEW OF SLUDGE CONDITIONING TANKS 25
VI
-------
LIST OF TABLES
No. Page
I Variations in Mentor Primary Plant Detention Time ... 16
II Condensed Summary of Overall Experimental Results . . 29
III Overall Summary of Contaminant Removals:
Ferrous Chloride Addition Without Other Additives . . 32
IV Overall Summary of Contaminant Removals:
Lime Addition Without Other Additives 34
V Comparison of Contaminant Removals:
Ferrous Chloride and Lime Addition 35
VI Overall Summary of Contaminant Removals:
Ferrous Chloride and Lime Addition 37
VII Comparison of Contaminant Removals:
Ferrous Chloride and Lime Addition,
Separately and in Combination 39
VIII Oxidation of Ferrous Iron-Supernatant
Laboratory Test Results: 41
DC Oxidation of Ferrous Iron-Sediment
Laboratory Test Results: 42
X Effect of Varying Lime to Iron Ratios 43
XI Overall Summary of Ferrous Iron and Lime Dosage Levels 44
XII Comparison of Overall Contaminant Removals:
Ferrous Chloride, Lime, and Supplementary Additives . 46
XIII Typical Iron and Phosphorus Analyses During Sludge
Handling 53
XIV Sludge Dewatering Tests:
Use of Coal for Sludge Conditioning 54
VII
-------
In Appendix
XV Summary of Experimental Conditions 68
XVI Summary of Experimental Results: Solids Removal . . 69
XVII Summary of Experimental Results:
Oxygen Demand Reduction 70
XVIII Summary of Experimental Results: Phosphorus Removal 71
Vlll
-------
Section I
CONCLUSIONS
1. The use of ferrous iron as an additive to sewage was effective in
removing more than 80 percent of the phosphorus contained in
the sewage.
2. Spent pickle liquor from the acid-processing of steel was a satis-
factory low-cost source of ferrous iron.
3. Required iron dosage ranged from the theoretical minimum of
2. 7 to a maximum of approximately 3. 1 weights of ferrous iron
per weight of phosphorus contained in the sewage.
4. A slight excess of iron in addition to the theoretical amount re-
quired by the phosphate demand of the sewage was necessary. The
excess iron hydrolyzed to form a floe of ferrous hydroxide in
which the finely-divided phosphate precipitate was entrapped and
caused to settle. Otherwise, the phosphorus, although largely
insolubilized as ferrous phosphate, was swept into the plant efflu-
ent.
5. It was also necessary, when ferrous iron was used in primary
treatment, that the pH of the sewage as measured subsequent to
the iron addition be adjusted to the range of 7. 5 to 8. 0 by the addi-
tion of a strong base. Unless this was done, excess soluble fer-
rous iron was not efficiently hydrolyzed and settled and appeared
in the effluent in high concentrations. Lime was satisfactory and
is the cheapest of available bases. A dosage of approximately
1. 6 to 1.9 weights of lime per weight of iron was required.
6. Even with the addition of lime, total iron residuals in plant efflu-
ent were approximately 10 mg/1 Fe. However, approximately
90 percent of the effluent iron was insoluble. Filtration may be
required where such iron levels are not acceptable. A subsequent
biological or other treatment could also reduce the effluent iron
to acceptable levels.
7. A valuable by-product of the use of ferrous iron and lime for phos-
phorus removal in primary treatment is the enhanced removal
of suspended solids and oxygen demand from the sewage. Since
overall suspended solids removals were 61.5 percent and BOD
and COD removals were 61. 6 and 55. 3 percent in an experimen-
-------
tal operation, it may be expected that the degree of treatment
will be the same or higher in future applications.
8. No additives other than the base (lime) were necessary with ferrous
iron. The supplemental use of polyelectrolyte, sodium silicate
and coal as further aids to sedimentation results in only incremen-
tal increases in the separation of the solids from the sewage, and
did not appear to be justified by the additional cost.
9. The combined chemical cost for ferrous chloride pickle liquor and
lime was $1. 09 per million gallons of sewage treated, for each
mg/1 of contained phosphorus. In the case of a sewage containing
13 mg/1, as at Mentor, the cost would be $14. 17 per million
gallons using the unit chemical costs experienced.
10. The weight increase in sludge solids due to the precipitation of
phosphates by ferrous iron and lime was approximately 100 pounds
per million gallons of raw sewage per mg/1 of contained phospho-
rus. The enhanced removals of suspended solids as sludge re-
presented an additional 420 pounds.
11. The sludge volume increase was variable depending, in part, upon
the efficiency of suspended solids removal prior to the use of iron,
but in no case appeared to be double that obtained in normal treat-
ment.
12. The operation of existing sludge digesters and dewatering facili-
ties may be adversely affected by the additional weight and volume
produced with ferrous chloride and lime additions due to over-
loading. However, any other procedure that has the fortunate re-
sult of increased suspended solids and phosphorus removals would
similarly affect sludge digesters.
13. Precipitated phosphorus remained insolubilized through anaerobic
digestion.
14. De-watering of the sludge produced from the use of ferrous iron
and lime proceeded routinely, with enhanced filterability.
15. All of the mechanical requirements of the ferrous iron-lime treat-
ment method can be met by conventional equipment properly used.
-------
Section II
RECOMMENDATIONS
The use of ferrous iron pickle liquor and lime, as investigated in this
program, is recommended as a means of removing 80 percent or more
of the total influent phosphorus in a primary sewage-treatment plant.
Specific recommendations include the following:
1. The pickle liquor (aqueous solution of impure ferrous chloride or
sulfate) should be introduced into the sewage at the inlet end of the
plant train. When adequate turbulence is not available for thorough
mixing, the use of a flash-mixing step is recommended.
2. An initial iron dosage of 3. 1 parts by weight of contained ferrous
iron per part by weight of contained total phosphorus by weight is
recommended, with subsequent empirical adjustment in each plant
situation.
3. Although the rate of iron addition may be controlled manually,
greater removal efficiency and economy are attained when the rate
is continuously varied in proportion to the phosphorus content of
the plant influent as it continually varies. The use of an automatic
phosphate analyzer as a means of controlling the iron feeder is
therefore recommended.
4. Hydrated lime, calcium hydroxide, is introduced subsequent to the
iron addition and prior to settling. Thorough mixing is required,
and the continuous preparation of the lime as a water slurry prior
to mixing is recommended.
5. An initial lime dosage of 1. 9 weights of hydrated lime per weight
of added iron is recommended. Required lime dosage will vary
according to the alkalinity of the water and the acid content of the
pickle liquor, and, as in the case of iron, should be further adjus-
ted in each situation. The introduction of only enough lime to in-
sure a blue color in the hydroxide floe is an accurate rule of thumb
for purposes of control adjustment.
6. Re-design of existing primary sedimentation units is not necessary.
However, requirements for thorough mixing of the added chemicals
may require the introduction of air or mechanical mixing in some
locations. This would generally be true in cases where lime is
added directly to the sewage in a dry form, rather than as a slurry.
-------
If air is required, an initial rate of approximately 20. 5 cubic feet
per minute per MGD of sewage is recommended, with further re-
finements during plant operation.
7. Although the use of ferrous iron and lime results in a substantial
increase in the quantity of sludge produced, it is not an inherent
requirement of the method that the capacity of such equipment in
existing plants be increased in every instance. Existing equip-
ment may often be adequate, and an investigation of this possibil-
ity is strongly recommended when the cost estimate for each in-
stallation is prepared.
8. The use of ferrous iron without lime or other chemical supple-
ments has been reported to be satisfactory for phosphorus re-
moval in activated sludge secondary treatment, but is presently
regarded with reservation for use in primary treatment ahead of
trickling filters or carbon adsorption columns because of the high
floe carryover from primary treatment. When lime is used with
ferrous iron the effluent floe carryover is greatly reduced, sug-
gesting that the combination might be an inexpensive and effective
means of phosphorus removal in secondary plants based on pro-
cesses other than activated sludge. Plant investigation of this
approach is strongly recommended.
9. It is evident that any evaluation of the effect exerted by chemical
additions on sewage treatment plant operation must take overall
solids balances into account. However, precise data during con-
ventional operation of the Mentor Plant are presently unavailable,
except on a theoretical basis. It is therefore recommended that
an empirical study of plant solids balance be undertaken to better
evaluate the effect of chemical additions on the production of par-
ticulate matter throughout the treatment system.
10. Although the recommendations listed here are based on the assum-
ption that anaerobic digestion will be used for sludge disposal, the
sludge produced following sewage treatment with ferrous iron and
lime may be more economically dewatered and disposed of by
other methods. The investigation of methods alternative to anaero-
bic digestion is recommended.
-------
Section III
INTRODUCTION
Chemical precipitation has received considerable attention as a method
of separating dissolved phosphates from sewage. The addition of
soluble compounds of aluminum and ferric iron has already been demon-
strated to be effective; but these materials, along with the supplemen-
tary additives which are sometimes required, are relatively expensive
and the investigation of cheaper additives has continued.
Ferrous iron is also reactive with soluble phosphates and *is perhaps
the cheapest of all potential chemical precipitants. Ferrous sulfate is
produced commercially as a low cost source of iron in fertilizers.
More important, ferrous iron--as either the chloride or sulfate--is
also currently available at low cost as a waste by-product in the acid
pickling of steel. However, the cost would increase with greater use.
This report covers an investigation of the use of ferrous iron as a phos-
phate precipitant in an operating primary sewage-treatment plant. The
work was performed at the 4-MGD Willoughby-Mentor Wastewater
Treatment Plant by the Ohio County of Lake and the Rand Development
Corporation, of Cleveland, Ohio, under FWQA Grant WPRD 172-01-68.
A first objective was to determine the level of total phosphorus remov-
al which could be attained by ferrous iron in primary treatment, and
to establish by experimentation the cheapest source of ferrous iron
which could be used to sustain this level. The use of supplementary
additives such as lime, polyelectrolyte and ground coal, was also in-
vestigated, as were the side-effects of both the iron and the supple-
ments.
Ferrous iron pickle liquor was found to be entirely satisfactory for the
removal of more than 80 percent of the influent phosphorus, with the
requirement for pH adjustment and flocculation of any excess iron met
by the use of lime. Significant increases in removal of suspended
solids and oxygen demand were also noted. The experimental program
then continued as an investigation of the practical application of the
method in an operating plant, with emphasis on matters of plant design
and control, and cost.
Sections I and II of the report comprise a summary of the conclusions
and recommendations. Section IV is a discussion of the Mentor Plant
facility, and all facilities and resources used during the program. A
brief discussion of the procedures and analyses devised to measure
process efficiency, as well as detailed presentation of experimental
results is included in Section V. An overall evaluation of the results
is included in Section VI, in which projected costs are also listed.
-------
Section IV
MENTOR PLANT FACILITY
The Willoughby-Mentor Wastewater Treatment Plant is located on
approximately 37.6 acres of land in Mentor City, Ohio, north of Lake
Shore Boulevard (S. R. 283) and approximately 1,800 feet east of
Hopkins Road.
Wastewater treatment at the Mentor Plant consists of comminution,
bar screening and primary sedimentation. The plant was designed for
an average flow of 4, 000, 000 gallons per day, but hydraulic loads of
up to 12, 000, 000 gallons per day, the design maximum, have been re-
corded. Throughout the body of this report the Mentor Plant is fre-
quently referred to, since most of the experimental work pertinent to
this program was carried out at that facility. Figure 1 is a layout of
the plant.
This section is a summary of pertinent data concerning Mentor's in-
fluent sewage and the facilities and resources available during this
program for its treatment. Mention of special conditions which in-
fluenced experimental results is made where appropriate. A brief
discussion of the procedures and analyses devised to measure process
efficiency, as well as the detailed evaluation of experimental results
is deferred until Section V.
Description of Influent Sewage
In summary, the influent at Mentor is a high-solids Wastewater with a
high proportion of colloidal suspended matter. Occasional concentra-
tions as high as 1000 mg/1 total solids and 350 mg/1 suspended solids
have been recorded. Total phosphorus concentrations fluctuate widely
throughout any given day, and typically vary between 7 and 25 mg/1,
with an overall average of 13.0 mg/1 as P.
Flow Rate
Flows to the Mentor Plant have increased steadily as connections to the
sewerage system are completed. For example, plant reports* indicate
the average daily flow rate has increased from 2. 08 million gallons per
day in 1967 to 4. 36 MGD, at a design population of 40, 000 persons, in
1970. In all, nearly three billion gallons of sewage were treated dur-
ing the two years of this program. From June 1, 1969, the average
influent flow at Mentor averaged somewhat above 3, 000, 000 gallons
-------
00
SEWER FIFE INFLUENT
SEWER PIPE INFLUENT
PVC CHLORINE LINE
GIF FORCE MAIN
SEWER PIPE OUTFALL
CIP PRIMARY INFLUENT LINE
SEWER PIPE OUTFALL
SEWER PIPE OUTFALL
CIP DIGESTED SLUDGE LINE
CIP SUPERNATANT LINE
CIP RAW SLUDGE LINE
WILLOUGHBT-MENTOR WASTEWATER
TREATMENT PLANT
PRIMARY TANK
BLDG.
WEST
PRIMARY
EAST
PRIMARY
DIGESTOR
CONTROL
BLDG.
WASTE PICKLE
LIQUOR STORAGE
IN PLANT
CONTROL BLDG
DIGESTER
NO. 1
(REACTIVE)
DIGESTER
NO. 2
(NON-
REACTIVE)
PUIP
. STA:K»
IRON ADDITION
CHEMICAL COAGULANTS APPLIED AT THIS POINT;
( LIME, POLYELECTROLYTES. AND CHLORINE )
FIGURE 1
Plan of Mentor Plant Facility
-------
per day overall. This flow was far from uniform over a twenty-four
hour period, sometimes receding to a rate of less than 1, 000, 000 gal-
lons per day between 6:00 a.m. and 9:00 a.m.* and gradually building
to over 4, 500, 000 gallons per day by 1:00 p.m. The flow would then
hold at this figure until approximately 6:00 p.m. , thereafter decreas-
ing slightly from 6:00 to 8:30 p.m. with a slight buildup in the late
evening hours. A typical diurnal flow pattern for August 20, 1969,
through August 26, 1969, is shown in Figure 2.
Solids Loading
Normally, the incoming solids are high at Mentor during the week and
low on the weekends. Further variations occur when a heavy rain
following a dry period flushes large quantities of solids into the plant.
Hydraulic overloading accompanies such rains, reducing the overall
degree of treatment that can be attained.
A specific influence on plant performance is the occasional introduc-
tion of solids on a gross basis that are not normally associated with
the sewage-treatment train at Mentor. Throughout the program, pri-
mary sludge produced in the Madison, Ohio, Imhoff tanks was trans-
ported by truck to the Mentor Plant and added to the incoming raw
sewage at a wet well upstream of the analytical sampling station. Ad-
verse scheduling requirements did not permit mixing this material
with Mentor's own digested sludge.
During a typical three-month period a total of 95, 000 gallons of
Madison sludge was thus added. Analysis of the sludge indicated an
average of 10. 3 percent total solids by weight (57. 8 percent volatile
and 42.2 percent non-volatile). Of the 82,000 pounds of solids, 90
percent were in the suspended form. There were days when as many
as 4, 000 gallons of sludge were added, representing 3, 100 pounds of
solids or 40 to 60 percent of the influent suspended solids. No delete-
rious effects on mechanical operation of the Mentor Plant were noted
as a result of these sludge additions. Madison sludge additions were
therefore regarded as a component of the sewage in this program, but
the precise effects on treatment were not susceptible to separate in-
vestigation. The subject is not given further notice, but it is to be
borne in mind when interpreting the influent suspended solids concen-
tration.
There is also a large recirculating load of solids in the digester super-
natant which imposes a periodic burden on the sedimentation process.
The cause is traced to overloading in the sludge digesters and the re-
-------
6AM
--26.4
A A A FLOW - WEEKLY AVERAGE
8/20/49 through 8/24/69
PHOSPHORUS- AVERAGE OF
70 GRAB SAMPLES
9AM
12Noon
3PM
6PM
Time of Day
FIGURE 2
9PM
12M
3AM
6AM
Typical Diurnal Flow and Total Phosphate Variation
-------
suiting decreased digestion time. Approximately 0. 8 percent of the
plant's overall hydraulic load is digester supernatant return, but inter-
mittent operation results in contributing nearly 35 percent of the
Mentor Plant's suspended solids load during short periods. Subsequent
chemical treatment with ferrous chloride and lime imposes an addition-
al burden due to doubled sludge production.
Supernatant liquid from the digester is returned at a point where the
influent flow is distributed between the two primary tanks. The pre-
cise point of return is off-center in the inlet pipe, favoring distribution
to the west primary. As a result of a hydraulic imbalance, noted later
is this section, most of the digester supernatant solids are subjected
to extended detention times in the west primary. Study of those per-
iods when no supernatant is added indicates that the two tanks provide
unequal treatment since the west tank can operate more efficiently.
The addition of supernatant solids therefore tends to equalize perfor-
mance, but to an undetermined extent.
Phosphorus Concentrations
Phosphorus concentrations in the Mentor raw sewage vary widely dur-
ing any single day. The total input quantity varies to an even greater
degree as a result of flow variations, since the highest concentrations
normally occur during the period of highest flow.
Seventy influent grab samples were averaged to investigate the fluctua-
tions in phosphorus content over a 24-hour period. The influent sampl-
ings started at 8:00 a.m. on a Monday and showed periods of high con-
centration that closely paralleled the flow and which were attributed to
the Monday wash day. Another 24-hour test was made on the following
Thursday and results did not indicate the extreme fluctuations obtained
on the previous Monday, but varied widely nonetheless. Results ob-
tained from both of these tests indicated that phosphorus concentrations
may fluctuate from 7. 35 to as much as 24. 70 mg/1 as P, but that aver-
age concentrations of 11. 60 to 15. 70 mg/1 during the week are more
nearly typical, with an average of 13. 0 mg/1 overall. A typical phos-
phorus concentration pattern, for August 20, 1969, through August 26,
1969, is shown in Figure 2.
The proportion of orthophosphate, PC>4, to total phosphate was also
observed to vary widely and numerous tests were conducted to deter-
mine if there was a pattern to the variations. During one 48-day test
in the autumn of 1968, the proportions were observed to range from
38 to 85 percent and could not be correlated with either influent total
phosphorus or flow rate. The average concentration of orthophosphate
11
-------
during this period was 66. 4 percent of 16. 6 mg/1 total phosphorus as
P. Since the proportion was found to be completely independent of
flow rate, impracticably frequent manual monitoring would have been
required to provide precise data for chemical dosing. As a conse-
quence, all work in this program was based on total phosphorus analy-
ses reported as "P".
Definite effluent improvement was obtained by changing the rate of
ferrous chloride addition to accommodate variations in influent phos-
phorus concentration. It is evident in Figure 2 that the influent phos-
phorus concentration generally begins to increase in the late morning
hours and to decline before midnight. However, more precise pre-
dictions are not feasible when based on statistical correlations. The
ability to make continuous phosphorus analyses would have permitted
more frequent changes in the iron feed settings and allowed feed rates
to conform more closely with minimum theoretical requirements. A
continuous phosphorus analyzer is therefore recommended for use in
future applications.
Description of the Mentor Plant
The overall plant consists of 3 main structures for plant operation:
a main plant control building, primary tank control building, and a
digester control building. The complete system contains a number of
steps and is represented schematically in Figure 3.
The main plant control building houses the chemical feed and sludge
filtering equipment, as well as a plant superintendent's office, analy-
tical laboratory, and facilities for'indoor loading of dewatered sludge.
The second floor is used for chemical storage and handling to the feed-
ers. The basement level houses equipment for pumping digested
sludge from the sludge holding tank to a vacuum filter on the main
floor.
The primary tank control building houses air compressors rated at
200 cubic feet per minute for primary sewage flocculation, and pri-
mary sludge withdrawal equipment on the main floor. Ventilation is
provided to purge raw sludge wells prior to and during sludge draw-
ing. The lower level houses the raw sludge pumps and associated
piping.
The digester building contains the heat exchanger and required pump-
ing and piping equipment.
12
-------
FIGURE 3
WILLOUGHBY-MENTOR SEWAGE TREATMENT PLANT
CHEMICAL STORAGE
8
a
FLOW SHEET
CHEMICAL COAGULANT
'FEEDING (LIME AND POLYMER)
« o i
n[
iW
JAGE
<
"** ^1 — 1 ^n T\
SEWAGE ^-"•n | PT| *"
PTTMP ., _ — " "7w
METERING
» r
^ RAW SLUDGE
GA
HEATING DIGESTER f
1 .r— L-!
J J SLUDGE
SLUDGE 1 Ki NO. 1
PUMP g| ,_ (REACTIVE)
r I, If
EXTERNALyn PRIMARY 1
EXCHANGER I | !>LUDGE
DIGESTER SUPERNATANT
PRIMARY SET!
SEDIMENTATION SEWA
^^J^
\
WASTE
3 BURNER
9
r -i
^ DTRRSTER ^
NO. 2 " >
(NON- DIGEST!
KTTYEJ^LUDGE
m j
POST CHLORINATION
FINAL EFFLUENT
SLUDGE CONDITIONING
CHEMICALS
/FILTRATE TO PRIMARY SETTLING
-------
Pump Station
All waste-waters enter the plant property through a 48-inch trunk sewer
and a 27-inch trunk sewer and are carried to a junction manhole ahead
of a pump station located on the plant property. Sewage is pumped to
a 22, 400-gallon wet well, providing 8 minutes detention time at an aver-
age flow of 4 MGD. Flows are manually by-passed to the outfall sewer
in the event of pump station malfunction or power failure. One com-
minutor with an ultimate capacity of 11 MGD is provided, and provi-
sion is made for a by-pass through manually cleaned bar screens.
The pump station itself consists of four vertical non-clog pumps with
motors located on an intermediate floor. Two of the units have a capa-
city of 2 million gallons per day and the others 4 million gallons per
day. Pump operation is sequenced by a wet well liquid level control
system, providing a total capacity of up to 12 million gallons per day.
Flow Metering
All influent sewage is metered by a flow tube located adjacent to the
plant control building in a 24-inch cast iron force main. Facilities
for by-pass of total plant flow are installed downstream from the meter-
ing device at the by-pass manhole. The plant flow is indicated, total-
ized, and recorded on a metering panel installed in the plant control
building. The plant flow totalizer was susceptible to occasional mal-
function throughout the program, requiring repairs which interrupted
the experimental program and interfered with continuous data collec-
tion.
Primary Settling
Raw sewage passes from the 24-inch inlet pipe into a control chamber
at the primary tank control building for distribution to the primary
tanks. The wastewater discharges from the control chamber through
an 18-inch cast iron pipe to the center flocculation compartment of each
primary tank. Two circular primary settling tanks each 70 feet in
diameter, of the combined air-flocculation and settling-compartment
type, are provided. The flocculation compartments are circular, with
air diffusers for mixing. Figure 4 is a view of the west primary.
The primary settling tanks provide an average detention time of 2 hours
and 20 minutes at 5 million gallons per day of wastewater flow. How-
ever, changes in pumping sequence often result in variations in rates
of flow of 100 GPM or more. Typical detention periods for observed
14
-------
Figure 4 Overall view of the Mentor Plant from the
north-west. The west primary tank is in the fore-
ground, with the primary tank control building along-
side it. The main plant control building is located at
the southern end of the plant property.
flows during 1970, the last year of the program, are listed in Table I.
Values are computed on the basis of the volume of the tanks between
the inlet wall and the overflow weirs; flow rates are not listed for days
of equipment malfunction.
Monthly average detention times, as indicated in Table I, range from
1. 86 to 2. 70 hours. On a daily basis the fluctuations in flow rate are
even wider, ranging from 2 to 8 MGD. Actual detention time can there-
fore range from 1. 16 to 4. 66 hours over a 24-hour period. This fluc-
tuation was observed to have a marked effect on settling efficiency dur-
ing ferrous chloride and lime additions, and exerted a corresponding
effect on solids removal.
Weir overflow rates are highly variable in the same approximate pro-
portion as detention time. The extreme turbulence resulting from
hydraulic overloads was observed to be the principal cause of chemical
floe overflow during experimental testing. While some turbulence is
15
-------
desirable for mixing purposes, the efficiency of the chemical addition
procedures was affected by very high flows and minimized somewhat
the degree of process optimization which could be attained.
TABLE I
Variations in Mentor Primary Plant Detention Time
Avg. Rate of Flow Detention Time
Month
January, 1970
February
March
April
May
June
July
August
September
October
November
December
MGD
4.285
5. 008
4.852
4.784
4. 547
4. 587
3.457
3. 534
4.287
4. 046
4. 371
4. 606
GPM
2970
3470
3360
3320
3150
3180
2400
2450
2980
2690
3040
3200
In Hours
2. 18
1. 86
1.92
1.95
2. 05
2. 03
2.70
2.64
2. 18
2. 30
2. 13
2. 02
A more specific influence on plant performance was exerted by an im-
balance in the distribution of raw sewage between the two primary tanks.
The east tank carries approximately 60 percent of the hydraulic load,
with a corresponding effect on detention time in each primary tank.
Subsequent tables, including the performance summaries, have been
weighted to give a more accurate indication of overall contaminant re-
movals by recalculating average effluent contaminant concentrations
(as mg/1) from the actual weights (in pounds) of contaminant materials
handled by each primary tank.
Application of Air for Mixing Purposes
Since the influent at Mentor is divided between two primary settling
tanks, the technique of introducing air into the mixing areas of one
tank while eliminating air in the other was studied during iron and
lime addition to determine the effect, if any, of air on flocculation ef-
ficiency. Eighty-five cubic feet of air per minute could be introduced
into one tank or divided equally between both tanks. The former prac-
tice had been to operate without air since greater removals of suspend-
ed solids could be attained without its use.
16
-------
Evaluations were made over two-week intervals, while the air was
changed from one tank to the other at thfe end of each week and by divi-
ding the air equally between the two tanks. This procedure was follow-
ed with and without chemical additions for study of the variations be-
tween the two tanks. The evaluation was hampered somewhat by the
flow imbalance discussed above, but it was determined both visually
and by the reductions of contaminants, that under all conditions except
extreme hydraulic overloading the use of some form of mixing to en-
hance that obtained from normal influent turbulence was desirable.
Poor mixing resulted in a cloudy supernatant high in phosphorus con-
tent. The use of approximately 20 cubic feet per minute for each MGD
of sewage flow appeared optimum for the Mentor sewage. However,
equipment requirements required adopting a fixed rate of 42. 5 CFM,
corresponding to a flow rate of 2, 125, 000 gallons of sewage per day,
as a standard. The application of air is pictured in Figures 5 and 6.
Plant Outfall
The primary tank effluents flow over peripheral "V" notch weirs and
discharge into a 48-inch diameter outfall sewer that extends into Lake
Erie for a distance of approximately 2, 000 feet. The design maximum
of the outfall sewer is approximately 30 million gallons per day.
Chemical Feeding and Handling Equipment
A total of six different chemical additive types was of significance to
the experimental program and included chlorine, coal, ferrous chloride,
lime, polyelectrolyte, and sodium silicate coagulant. The following
is a discussion of the equipment and procedures devised for their use.
Chlorine
Chlorination at Mentor is provided by two solution-type feeders with
facilities provided for weighing, handling, and feeding from one-ton
cylinders. The feeders for pre-and post-chlorination are sized to
deliver chemicals at a maximum rate of 800 Ibs. per day, or 24 mg/1
at the design rate of flow.
Pre-chlorination facilities are currently calibrated to deliver an esti-
mated 12 mg/1 of chlorine. It can be applied at two separate points:
a manhole ahead of the wet well, and a point a few feet upstream from
the lime addition station, the latter case being the usual practice.
Post-chlorination is applied at the effluent manhole. Chlorine is intro-
duced to the sewage at all times for disinfection purposes; a maximum.
17
-------
Figure 5 View of the west primary tank. Note the t\iT •
bulence at the center from the applied mixing air.
Figure 6 Enlarged view of the west primary tank. The
turbulence at the center results from 42. 5 CFM of air
forced into the tank for mixing.
18
-------
chlorine residual of apporximately 4 mg/1 is maintained in the plant
effluent, with the outfall sewer to Lake Erie providing the necessary
contact time.
During the program reported here, a study of chlorine additive levels
was made to gauge the effect of such parallel additions on treatment
with iron and lime, and to evaluate their effect on concentrations of
residual iron and phosphorus in the effluent. None of the tests showed
a noticeable effect on treatment characteristics in either case, even
at concentrations in excess of 40 milligrams per liter. The subject is
therefore not given further attention as a consequence, but it is to be
borne in mind that under certain conditions chlorine can oxidize ferrous
iron to the ferric state.
Coal Feeding
The temporary coal feeder used in this program was housed between
the two primary settling tanks to facilitate delivery of the coal to either
one or both of the primary clarifiers. A volumetric dry feeder of the
type normally used for lime was installed.
Ferrous Chloride
The ferrous chloride pickle liquor was fed into the raw sewage up-
stream from the comminutor and bar screens as shown in Figures 3,
7 and 8. The basic equipment consisted of a 10, 000-gallon fiberglass
tank for pickle liquor storage (Figure 7), a pump for mixing the mate-
rial during and after loading operations and a proportioning pump for
feeding the ferrous chloride solution according to the influent flow.
The quantity of iron solution added to the raw sewage influent was reg-
ulated by manual adjustment of the pump speed and percentage of
pumping time.
Thp original iron feed pump shown in Figure 9 proved inadequate,
since it was equipped with ball-check valves that became corroded by
---the ferrous chloride pickle liquor. The gravity-actuated valves were
located on each side of the pump diaphragm and residue accumulated
on the valve seats which affected the quantity of solution which could
be passed. Pump motor malfunction and piping failures also contribu-
ted to the problem of maintaining a reliable feed system for the iron
solution. Consistent dosing levels could not be maintained as a con-
sequence, and steps were undertaken to revise the system.
The iron feed equipment used after March, 1970, consisted of a Moyno
progressing cavity pump controlled by a variable speed drive. This
19
-------
Figure 7 Waste pickle liquor reservoir. Access to the
raw inlet channel is in the right foreground.
Figure 8 View from above iron delivery spigot at raw
sewage inlet channel. Ferrous chloride pickle liquor was
fed on a diurnal basis.
20
-------
Figure 9 Original pickle liquor feed pump. The
equipment was subsequently replaced by the positive
displacement pump depicted in Figure 10.
pump is shown in Figures 10 and 11, and was capable of continuously
supplying ferrous chloride solution from the storage tank at an average
rate of approximately ten gallons per minute. A removable nylon filter
was installed in the feed line between the storage tank and the pump,
from which accumulated solids could be withdrawn. A coarse filter in
the feed lines from the delivery truck was provided for additional pro-
tection. As with the previous chemical feed device, the pump was
cycled to co-ordinate with the intermittent operation of the wet well
pumps, with the cycling time dependent upon the combination of pumps
necessary to keep the wet well at a constant level. The change in
equipment proved adequate throughout the rest of the operation and
permitted reasonably predictable feed rates to be established.
Lime
Hydrated lime in bagged form was loaded into a hopper feeding a lime
batching column for wastewater coagulation (Figure 12). A collector
is provided for dust control. A side stream of raw wastewater was
diverted into the batching column. An adjustable solids feeder dropped
the lime into the side stream where flow turbulence mixed it into the
raw sewage.
21
-------
Figure 10 Positive displacement pickle liquor pump.
J
Figure 11 View of iron feed pump replacement. An al-
ternate view of the positive displacement pump used dur
ing the latter phase of this program.
22
-------
Figure 12 Lime feeder on first floor of main plant
control building. Dry lime is fed to the raw sewage.
The lime feeder had been installed for other work prior to this pro-
gram, and was originally intended for intermittent operation to feed
lime in proportion to the volume of wastewater flow based on an elec-
tronic signal from the magnetic flow meter. The original design
criteria for lime feeding equipment did not provide for feeding at the
rate required in this study when the equipment was operated as a func-
tion of changes in iron feed. The dry lime feeder was therefore man-
ually adjusted twice a day, at 12 noon and midnight. The average
daily requirements could be satisfied in this manner but optimum per-
formance could not be attained.
i
Poly electrolyte
An 800-gallon polyelectrolyte solution feeder in the main plant control
building was operated periodically throughout the program, with poly-
mer metered into the raw sewage influent a few feet downstream from
the lime station. Addition of polyelectrolyte was a fairly routine pro-
cedure throughout the experimental program and was suspended from
time to time to examine other additive combinations.
23
-------
Sodium Silicate
Additions of sodium silicate were manually charged upstream from
the iron addition station by regular plant personnel and representatives
of the chemical supplier.
Sludge Handling
Primary Sludge Removal
Sludge is withdrawn by gravity from each of the primary settling tanks
through telescoping valves to sludge wells located with the primary
tank building, and then pumped to the primary digester. The primary
sludge pumps operate automatically by liquid level control equipment.
Sludge may also be pumped directly from the settling tank hoppers with-
out use of the telescoping valves or sludge wells. The scum from the
primary tanks is also withdrawn into the sludge wells and pumped with
the primary sludge to the digester.
Sludge Digestion and Disposal
The sludge consists of the unstabilized sediment and scum from the
primary settling operation including the unseparated grit. Settled
solids concentrations ranged between 4 and 7 percent. The disposal
steps employed at the Mentor Plant are commonly those of digestion,
filtration, and ultimate disposal by land fill.
Normal digestion is accomplished by storing the raw primary sludge
for periods of up to a month or more in two 65-foot diameter tanks
with a maximum sidewater depth of 18 feet 6 inches. The digesters
are sometimes subjected to severe overloading with a corresponding
effect on efficiency.
Two sludge recirculation pumps and a heat exchanger are located in
the digester control building. The primary digester is reactive, with
continuous mixing of the tank contents. The system was designed for
the sludge to be heated to approximately 98° F. by passing it through
an external heat exchanger, but overloading and cold weather condi-
tions frequently result in temperatures ten to twenty degrees lower
with occasional temperatures thirty degrees lower.
The gas produced in the digestion process is either burned at a waste
gas burner or used as a fuel for the external heat exchanger. An
24
-------
auxiliary natural gas supply is provided as needed to augment the gas
produced in the digestion process. The sludge is then pumped to the
secondary, non-reactive, digester where separation of the solids
occurs by settling.
Each digester is provided with four supernatant outlets at different
elevations for manual selection of supernatant draw. The supernatant
from the secondary digester is returned to the raw sewage entering
the primary settling tank at a point right at the stationary flow splitter.
The combination of a plant hydraulic imbalance, decreased digestion
time, and uneven supernatant split between the two primary tanks,
results in a high suspended solids return to the west tank, as discussed
earlier in this section
Digested sludge is withdrawn to a 7, 500-gallon sludge holding tank at
the plant control building from which it is pumped through a positive
displacement pump to the sludge filtering process. In normal plant
practice the sludge is first conditioned for filtration by the addition of
ferric chloride solution and lime slurry stored in tanks, as pictured in
Figure 13. The solids are then separated by a rotating vacuum, filter,
with 132 square feet of drum area.
Figure 13 View of sludge conditioning tanks. The fer-
ric chloride tank for sludge conditioning is in the right
foreground, with the belt conveyer for filtered and dried
sludge beyond the doorway. The tank to the rear of the
ferric chloride reservoir is for polyelectrolyte.
25
-------
Dewatered sludge from the vacuum filter is transferred by belt con-
veyer to dump trucks or storage vehicles for disposal, either by use
as a soil conditioner or, as is most often the case, by land fill.
26
-------
Section V
DATA PRESENTATION AND EVALUATION
The experimental work was performed over a period of twenty-four
months and was conducted in parallel with routine plant operation. The
work is divided into three categories:
1. Ferrous chloride and lime addition
2. The use of supplementary additives
3. Coal as a settling aid
Plant operations were conducted 24 hours per day seven days per week.
Operating data were recorded hourly, summarized in 24-hour periods
and combined with the daily analytical results. The daily summaries
were then converted to monthly reports and submitted to the Ohio
Department of Health. ^
However, the gross data alone cannot be used for process optimization
or to draw final conclusions. The ferrous chloride dosage efficiency
cannot be typified by the tests using the prototype iron feed equipment
for example; and to generalize on the degree of treatment without re-
gard for plant hydraulic conditions, or the distribution of applied mix-
ing air, could result in misleading conclusions. Also, these evalua-
tions were done sequentially rather than simultaneously in parallel
equipment. Specific operating periods are therefore selected from the
gross data according to the following criteria:
1. Periods of extreme hydraulic overloading are excluded.
2. Periods of-gross mechanical, electrical or supply failure
a/re excluded.
3. When the type of additive was changed during a test the
period required for attainment of equilibrium with the new
additive is excluded.
4. Operating periods of less than five days are excluded when
data for other tests under the same conditions are available.
The selected data are tabulated first as summaries of each type of
test, twenty-five in all, with_a brief description of experimental condi-
tions which prevailed. Nine different chemical additives were evalua-
ted, including iron, lime and coal, representing thirteen separate
combinations, with the same basic equipment and procedures employed
for the study of each.
1. Ferrous chloride, FeClz (as pickle liquor in all cases)
2. Hydrated lime, Ca(OH)2
3. Ferrous chloride and lime
4. Polyelectrolyte, Dow A-23
27
-------
5. Ferrous chloride with Dow A-23
6. Lime with Dow A-23
7. Ferrous chloride and lime with Calgon 3000 polymer
8. Ferrous chloride and lime with Diamond 630 polymer
9. Ferrous chloride and lime with Diamond 640 polymer
10. Ferrous chloride and lime with Dow A-23
11. Ferrous chloride and lime with sodium silicate
12. Ferrous chloride and lime with coal
13. Dow A-23 with coal
Separate tables for solids removal, oxygen demand reduction, and in-
formation pertinent to phosphorus removal are given. All of the ma-
terial is taken directly from the operating records and is presented in
Tables XIV through XVII. in the Appendix: A condensed summary of
overall experimental results is included here in Table II. Further
condensations of data are tabulated throughout this section as appro-
priate.
Analytical results are based on manually collected 24-hour composite
samples, taken at the raw inlet and the primary effluent outfall for
each settling tank. Grab samples were taken at several locations as
required by the testing program. Routine analyses were performed in
the Mentor Plant laboratory, with more specialized tests and labora-
tory simulations performed by specially assigned personnel. All ana-
lytical procedures were taken from "Standard Methods for the Examin-
ation of Water and Wastewater. "3
It became apparent early in the experimental program that an analyti-
cal bias was introduced by the plant hydraulic imbalance discussed
earlier in Section IV. Effluent analyses obtained from the two primary
settlers ordinarily had been averaged to report an average value for
the entire plant. However, the effect of this procedure was to unduly
favor the results obtained from the west primary, which handled only
40 percent of the plant flow. Conversely the east primary was typi-
cally less efficient due to shortened detention times. The experimen-
tal results shown in Tables XIV through XVII have therefore been ad-
justed by recalculating average effluent concentrations from the actual
weight of contaminants handled in each primary tank to equalize the
effect of the hydraulic imbalance and provide a more accurate estimate
of process performance.
Laboratory simulations using a jar test technique were made through-
out the program to examine the mechanisms of phosphorus removal
and primary sedimentation as they pertained to the use of chemical
additives. Unless specified otherwise in the text, all such simulations
28
-------
TABLE II
Condensed Summary
Test
No.
3)
4)
5)
6)
9)
10)
11)
12)
13)
14)
17)
18)
19)
20)
21)
22)
23)
?4)
?S)
No.
Chemical Additions Days
by Type and Combination In
Ferrous Chloride.1 Fed?
Ferrous Chloride} FeCl2
Hydra ted Lime, Ca(OH)2
Hydrated Lime, Ca(OH)2
Ferrous Chloride & Lime
Ferrqus Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Polyelectrolyte, Dow A-232
Ferrous Chloride & Dow A-23
Lime & Dow A-23
Ferrous Chloride,
Lime & Calgon 3000
Ferrous Chloride,
Lime & Diamond 630
Ferrous Chloride,
Lime & Diamond 640
Ferrous Chloride,
Lime & Dow A-23
Ferrous Chloride, Lime &
Sodium Silicate3
Ferrous Chloride, Lime &
Coal4
Dow A-23 & Coal'
Teat
4
7
5
7
11
40
9
13
17
31
11
2
19
11
10
89
17
22
5
of Overall Experimental Results
Raw Influent Chemical Suspended
Sewage Acidity Additions Solids
Flow as CaCO^ me/Liter win
.Gallona ms/Llter Fe
10, 696, OOO7
26, 8 10, OOO6
11.025.0007
18,410,000°
40,480,000S
158,800,000s
38,250,000s
54,730,000s
76.500.000s
108,810, OOO9
50,710,000s
5,090,000s
57,000,000s
52,272,000s
40,500,0008
397,040,000s
68.510.000s
83,600,000s
15.165.000s
-
-
—
-
48
26
21
-
-
-
46
21
21
45
30
-
41.0
49.0
-
49.0
38.8
42.2
39.6
33.7
-
48.0
-
47.0
35.7
43.0
46.0
46.0
48.0
Lime
69.6
63.0
78.0
75.0
68.7
82.0
68.3
-
-
88.0
101.0
75.6
82.0
83.0
87.0
67.0.
Poly-»0.74z
Other
-
-
.
.
_
_
_
0.65
0.43
0.47
0.26
0.26
0.25
0.45
12.03
30.0
51.6
7 Red
28.5
14.0
61.0
53.3
73.8
52.0
58.8
67.9
65.2
56.1
5.2
38.0
70.6
64.4
71.4
62.3
70.9
59.3
44.4
7 Red
28.3
27.0
16.7
17.3
59.2
63.2
61.9
63.2
55.7
28.8
27.6
23.1
51.7
51.4
64.0
34.0
66.5
46.2
39.1
Total Effluent
Phosphorus Iron
COD (as P) pH me/Liter
%Rejd
32.8
14.7
27.4
34.8
51.7
56.3
68.4
56.2
47.8
32.7
30.7
32.6
64.6
50.3
50.0
61.0
52.8
51.1
63.1
%Red
23.4
26.0
41.8
44.7
81.6
74.8
83.2
81.4
79.8
0.0
42.0
47.4
85.4
80.4
81.5
82.5
80.9
82.3
8.3
IN
7.7
7.5
7.4
7.7
7.3
7.6
7.5
6.9
7.4
7.7
7.4
7.3
7.6
7.6
7.2
7.4
7.3
7.5
OUT
7.3
7.0
8.5
8.6
7.6
7.6
7.7
7.9
7.2
7.5
6.9
9.1
8.1
8.0
7.5
7.6
7.5
7.5
OUT
42.5
-
8.7
13.2
10.8
11.0
10.6
-
36.0
-
8.3
9.9
10.4
12.8
13.1
15.8
Pickle liquor; 7-107. Fe by weight
2 Dow A-23 polyelectrolyte; 0.3% solution
3 Grade 40 silicate; 40% Na2Si03
4 18 x 80 mesh coal; to east primary,
5 Minus 80 mesh coal; to east primary
6 Air applied for mixing; 85.0 CFM west tank only
7 Air applied for mixing; 85.0 CFM east tank only
8 Air applied for mixing; 42.5 CFM to each tank
' No air applied for mixing
-------
were performed with 1000 milliliter samples in the following manner:
1. Chemical addition with rapid mixing (100 RPM) for 2 min.
2. Slow mixing (20 RPM) for 10 minutes
3. Settling for 20 minutes after mixing
4. Supernatant sampling by pipette
A variable speed stirrer (1-100 RPM range) with six shafts was used,
permitting multiple sampling and comparisons.
Conclusions and comments specific to the various plant and laboratory
tests are included in the text for continuity. Discussion of the process
variables, with recommendations for future experimental work and for
application of the recommended treatment method, are included in
Section VI.
Selection of Waste Pickle Liquor as Ferrous Iron Source
Work was started in the summer of 1968 toward the design and pur-
chase of equipment necessary to treat the Mentor influent, and an
evaluation was made of alternate ferrous iron souces. Two forms of
ferrous iron are most commonly available:
1. Ferrous sulfate, FeSC>4, available in certain locations as
waste pickle liquor from sulfuric acid steel pickling, but
generally sold in the crystal form, copperas - FeSO^ Tt^O.
2. Ferrous chloride. FeC^i available mainly as waste pickle
liquor, containing 7 to 10 percent iron and normally 1 to 5
percent free hydrochloric acid.
Hydrated ferrous sulfate costs approximately $0. 05 per pound of con-
tained metal. On the basis of the minimum theoretical dosage require-
ments, and assuming the insolubilization of 13. 0 mg/1 phosphorus, as
P, the minimum chemical cost would be $14. 60 per million gallons of
sewage treated. Added to this would be the cost of feeding and of
storage and handling equipment, since the material tends to cake in
storage. If the capital costs were to be amortized over a long period,
the expense is negligible. However, the costs of materials and equip-
ment to efficiently handle large quantities of powdered ferrous sulfate
on a comparatively short term basis were prohibitive.
The use of waste pickle liquor was investigated. The pickle liquor was
obtained from a continuous hydrochloric acid steel-pickling operation,
and was made available to the Mentor Plant simply for the cost of haul-
ing. The cost per pound of contained iron was approximately $0. 02,
or less than half that for granular ferrous sulfate, with minimum capi-
tal investment for storage.
30
-------
Laboratory tests indicated that either material was suitable, the fer-
rous chloride form being equally effective in phosphorus insolubiliza-
tion as ferrous sulfate. A contract was then let for weekly shipments
at the following specifications:
Quantity 10, 000 gal /week (approximate)
Iron as Fe 11. 000% Maximum
Hydrochloric acid 1. 000% Maximum
Manganese 0. 100% Maximum
Copper 0. 003% Maximum
Chromium . 0. 005% Maximum
Zinc 0. 006% Maximum
Suspended solids 10 mg/1 Maximum
Appearance Green as opposed to brown,
signifying lack of hydrolysis
A typical analysis of the material as delivered to Mentor is as follows:
Iron as Fe 10.2%
Hydrochloric acid 0. 61%
Wt/gal. 10.20 pounds
Freezing point -11°F.
Manganese 0. 036%
Copper 0. 001%
Chromium 0.002%
Zinc 0.002%
The iron in the material delivered to Mentor generally averaged 9. 3
percent iron overall, with 0. 52 percent free acid. The iron and acid
ranged from 7. 1 to 11.0 percent and 0.4 to 0.6 percent, respectively.
The density averaged 1.20, or ten pounds per gallon.
Ferrous Chloride Precipitation; No Supplementary Additives
Bench scale studies demonstrated that thorough mixing of ferrous iron
solution with raw sewage for a period of one minute or less forms an
insoluble ferrous phosphate precipitate which settles only slowly over
periods as long as three hours. The principal unresolved question
prior to actual plant operations with ferrous chloride, then, was wheth-
er .the insoluble ferrous phosphates could be induced to settle on a
plant scale without the use of lime.
The ability of ferrous iron to precipitate phosphate when used alone
was also demonstrated during an associated program where sewage
was filtered through ground coal. When ferrous iron, as the sulfate,
was added to raw sewage the average phosphorus removal was 82 per-
31
-------
cent as compared to 30 percent by filtration through coal alone. *»
relatively low degree of pH dependence was exhibited by ferrous iron
when the filter was prepared from ground coal high in pyritic sulfur.
The oxidation of the pyrite released ferrous iron and acid to solution
when the coal was first wetted. The highest phosphorus removals
were observed during the first hours of operation when the water was
acidic, diminishing rapidly with a decrease in the production of both
ferrous iron and acid. The reaction of ferrous iron with soluble phos-
phorus was insensitive to pH at the levels tested, and near-perfect re-
moval of phosphorus was attained whenever means were available for
separation of the fine solid precipitate from the sewage stream.
Plant testing was undertaken first to provide a baseline for later eval-
uations. Two separate plant tests are included in Table III, character-
ized mainly by differing mixing air applications. The influence of air
on settling is a factor in these experiments, since the work discussed
in Section IV had shown 85. 0 CFM to be excessive. Air was applied
to a different primary during each test which permitted combining the
experimental results to obtain an overall summary of performance.
The analytical results listed in the combined data summary in Table
III are weighted according to the proportions of flow treated in each
separate test, a procedure also followed in similar tables throughout
the rest of this section.
TABLE III
Overall Summary of Contaminant Removals:
Ferrous Chloride Addition Without Other Additives
Test Sewage Additive Percent Contaminant Removal Effluent
No. Treated mg/1 Suspended BOD COD Phos. pH Iron
Gallons Iron Solids as"P" In put mg/1
3) 10,696,000 41.0 28.5 28.3 32.8 23.4 7-7 7.3
4)26,810,00049.0 14.0 27.0 14.7 26.0 7.5 7.0 42.5
Combined Results:
37,506,00046.5 18.1 27.4 19.9 25.2 7.6 7.1 42.5
32
-------
In general, the addition of ferrous chloride was found to be effective
in the insolubilization of phosphates on a plant scale. However, a
finely-divided floe was observed to form in the settling tank, with con-
siderable carryover into the effluent which resulted in low suspended
solids removals. Treatment operations were found to be sensitive to
even hourly changes in flow rate and influent sewage composition, re-
quiring several hours to regain equilibrium.
Influent suspended solids content was variable, ranging from 136 to
316 mg/1, with an average of 193 mg/1 during Test 4. Overall apparent
suspended solids removals were exceptionally low throughout both of
the iron-addition periods, ranging between 0. 0 and 39. 9 percent, re-
spectively. Overall suspended solids removals for the combined tests
was 18. 1 percent which is less than removals in primary treatment
without iron. Removals tended to fluctuate widely on a day-to day
basis and were observed to be mainly a function of hydraulic overload-
ing and the accompanying weir overflow turbulence. It was evident
from the turbidity and floe formation in the 24-hour composite samples
that the decrease was due to carryover of finely-divided precipitates
into the effluent.
Reductions in oxygen demand generally varied with suspended solids re-
movals as would be expected in primary treatment. Overall BOD and
COD removals were 27.4 and 19.9 percent respectively.
No apparent phosphorus removals higher than 42 percent were recorded
in either test period. A low of zero was observed during one day of
extreme hydraulic fluctuation, with a combined average of 25. 2 percent
attained.
At the start of iron addition the effluent was water-clear, with no visi-
ble turbidity. The two plant tests were operated consecutively and after
about two days of iron addition the effluent assumed a pale yellow color
with a visible, finely-divided iron floe. A portion of the dissolved iron
was not hydrolyzed in the settling tank, but continued to hydrolyze in
the effluent. Analyses of the effluent during Test 4 showed an average
total iron concentration of 42. 5 mg/1, over 80 percent of it insoluble.
Careful observation indicated that the primary settlers were inefficient
in collecting and removing the precipitated iron floe. It was evident
that much of the light material was merely being stirred up by turbu-
lent conditions to be discharged into the effluent.
During a parallel program at Texas City, Texas, waste ferrous sul-
fate pickle liquor was added directly to raw sewage prior to primary
sedimentation, with insolubilized phosphate precipitates carried over
33
-------
to the aeration tank of an activated sludge secondary. The secondary
process provided longer detention time for reaction and more complete
mixing. The precipitated phosphates and iron hydroxides were entrain-
ed in the biomass of the biological secondary and separated from the
waste stream during final clarification.
However, the simple addition of iron to a primary system, without a
biological secondary, necessitates the use of another chemical to in-
crease the pH of the sewage to aid in hydrolyzing the excess iron to an
insoluble floe, which can then be removed during the settling process.
The most readily available chemical at Mentor which fulfilled these
criteria was lime and plant testing was undertaken to define the re-
quirements for its use.
Lime Precipitation
Table IV illustrates that the addition of lime was an effective form of
treatment for phosphate removal. Treatment operations were not espe-
cially sensitive to hourly changes in flow rate and influent sewage com-
position. Lime treatment of raw sewage was observed to be less effec-
tive in reducing BOD than was the case with ferrous chloride addition
alone or simple sedimentation. Suspended solids removals of 61.0 and
53. 3 percent were also recorded which was higher than the combined
18. 1 percent removal observed with ferrous chloride addition when used
alone. The two separate experiments involving raw wastewater treat-
ment with lime resulted in 41. 8 and 44. 7 percent removal of phosphorus,
respectively, with a combined average of 43. 7 percent.
TABLE IV
Overall Summary of Contaminant Removals
Lime Addition Without Other Additives
Test Sewage Additive
No. Treated
Gallons
5) 11,025,000
6) 18,410,000
rng/1
Lime
69.6
63. 0
Percent Contaminant Removal
Suspended BOD COD Phos. pH
Solids as"P" In Out
61.0 16.7 Z7.4 41.8 7.4 8.
53.3 17.3 34.8 44.7 7.7 8.
5
6
Combined Results:
29,435,000 66.5 56.0 17.1 32.0 43.7 7.6 8.6
34
-------
Overall influent suspended solids concentrations during the two periods
of lime addition were comparable to those for ferrous chloride, aver-
aging 212 and 199 mg/1, respectively. Allowing for solids contribution
by the lime itself, actual removals of influent sewage solids are pre-
sumed to be even higher than those shown in the table.
Brenner^ reports that calcium alone of the common mineral additives
requires a high free alkalinity of the sewage to convert dissolved phos-
phorus to a solid form. Sufficient calcium is normally available tore-
act with phosphorus at high pH levels, so that the lime dosage require-
ment is independent of phosphorus concentration. The calcium ions en-
ter into a precipitation reaction with phosphates, approaching comple-
tion at the requisite pH of 11. Lime also reacts with the bicarbonates
and with magnesium in hard water. The precipitate mixture then con-
tains mainly calcium carbonate, calcium hydroxyapatite, and possibly
magnesium hydroxide, while the effluent water is considerably softened.
The resulting sludge volumes are reported to be tripled.
The addition of lime in plant tests 5 and 6 raised the pH to an overall
average of 8.6, less than optimum for complete reaction, but sufficient
to obtain enhanced phosphorus conversion. An examination of Table V
shows that in respect to all contaminant removals, save for BOD, lime
alone appeared to out-perform ferrous chloride alone for primary treat-
ment.
TABLE V
Comparison of Contaminant Removals:
Ferrous Chloride vs. Lime Addition
Sewage Additive, Percent Contaminant Removal Effluent
Treated mg/liter Suspended BOD COD Phos. pH Iron
Gallons Iron Lime Solids as"P" In Out mg/1
37,506,000 46.5 -- 18.1 27.4 19-9 25.2 7.6 7.1 42.5
29,435,000 -- 66.5 56.0 17.1 32.0 43.7 7.68.6
The influence of lime on settleability also affected the apparent phos-
phorus removals, which increased almost two-fold with the use of
lime. High removals of phosphorus had been observed with ferrous
chloride addition, when taking into account the tendency of the light
precipitates to carryover from the primary settlers to the effluent.
35
-------
Ferrous Chloride and Lime
The use of ferrous chloride without other additives had been demon-
strated to be beneficial in precipitating phosphorus, but was also found
to be detrimental to settling efficiency because of the tendency of the
iron-phosphate precipitate to remain in suspension. It was also evi-
dent in the tests with ferrous chloride alone that a small fraction of
the dissolved iron was not hydrolyzed in the primary settlers, but con-
tinued through the sedimentation process to appear eventually in the
effluent. Laboratory experiments had confirmed that the use of hydra-
ted lime along with ferrous iron produced a more dense and particulate
floe, and plant tests were therefore initiated.
The flocculation of ferrous iron is pH dependent, approaching comple-
tion only in a slightly alkaline medium. The pH of the Mentor raw
sewage typically varies between 7 and 8, or near-optimum, for ideal
flocculation as indicated in the laboratory. The reaction proceeds
most effectively when the iron is retained in the ferrous form, denoted
by a blue color in the resulting floe. The lime slurry contributed hy-
droxide ions which were shown in laboratory tests to promote the for-
mation of ferrous hydroxide. The flocculation of the excess iron by
lime suggested that it could be added in conjunction with ferrous chlo-
ride. Lime was therefore added to adjust the pH to a consistent level
of approximately 7. 5 to 8.0 and provide a medium for forming a dense
floe with efficient settling properties. The average ferrous chloride
dosing levels used during testing with ferrous chloride and lime varied
from 43. 0 to 50 mg/1 as Fe. A slurry of hydrated lime, Ca(OH)2, was
metered into the sewage downstream from the iron addition station at
an average rate that varied from 55. 0 mg/1 in Test 7 to 103. 0 mg/1
in Test 8. In the absence of continuous monitoring analyses, the meth-
od of adding only enough lime to maintain the blue color was found to
be.an extremely sensitive means of control.
The addition of iron and lime together during preliminary tests, while
demonstrably effective in reducing the phosphorus content and in im-
proving the removal of colloidal solids from raw sewage, was of course
dependent on the variable phosphorus content of the sewage. When the
content of added iron exceeded that necessary to react with the contain-
ed phosphorus, virtually all of the phosphorus was precipitated, with
the excess iron forming an effective floe in the presence of the lime.
When the phosphorus content of the sewage exceeded the amount of
available iron, most of the iron was thereby consumed in the precipi-
tation of the phosphorus with insufficient remaining for floe formation.
36
-------
The addition of lime with ferrous chloride increases the pH of the
sewage to a consistent level for precipitation and is valuable in aiding
the hydrolysis of the excess iron to an insoluble floe. The formation
of a floe is important as a means of converting the otherwise finely-
divided iron phosphates into solids of manageable size.
During the tests when iron was added on the basis of average phospho-
rus content it was observed that floe formation, as judged by grab
samples taken from the settling tank influent, proceeded satisfactorily
only in the early morning hours. Eight grab samples of raw sewage
were then taken at 3-hour intervals and analyzed for phosphorus con-
tent. Results ranged between 6 mg/1 at 6:00 a.m. and 38 mg/1 at
3:00 p.m. , confirming a deficiency of iron during the daylight hours.
The iron concentrations were subsequently doubled during the daytime
period of average high phosphorus content, and reduced at night. Floe
formation was observed to improve noticeably, with a marked increase
in the consistency of floe formation and daily phosphorus removals.
Extended plant testing was then undertaken to refine the treatment
method. The results of five separate plant tests are included in Table
VI, representing a total of 90 days of operation and nearly 370 million
gallons of sewage treated. The application of 85 CFM of mixing air
on only one primary settler in Tests 7 and 8 was shown to be excessive
in earlier work and those tests are therefore omitted. As in other
similar tables, the analytical results listed in the combined summaries
are weighted according to the proportions of actual sewage quantities
that were treated in the separate primary settling tanks.
TABLE VI
Qverall Summary of Contaminant Removals:
Ferrous Chloride and
Test
Sewage
No. Treated
Gallons
9) 40,
10) 158,
11) 38,
12) 54,
13) 76,
480,
800,
250,
730,
500,
000
000
000
000
000
Lime Addition
Additive, Percent Contaminant Removal
mg /liter Suspended BOD
Fe
49. 0
38.8
42.2
39.6
33.7
Lime
78.0
75,2
68.7
82.0
68.3
Solids
73.8
52.0
58.8
67.9
65.2
59.2
63.2
61.9
63. 2
55. 7
COD
51.7
56.3
68.4
56.2
47.8
Phos.
as"P"
81.6
74.8
83.2
81.4
79.8
PH
In
7.3
7. ,6
7.5
6.9
7.4
Out
7.6
7.6
7.7
7.9
7.2
Effluent
Iron
mg/1
8.7
13.2
10.8
11.0
10. 6
Combined Results:
368,760,000 39.3 74.3
61.5 6.1,1 55.3 78.4 7.4 7.6 11.6
37
-------
The addition of ferrous chloride was found to be entirely effective for
the insolubilization of phosphorus on a plant scale, with the further
addition of lime effective in settling the iron-phosphate precipitate.
Effluent iron concentrations were markedly reduced as a consequence.
The sensitivity of ferrous chloride addition to hourly changes in flow
rate and influent sewage composition was also thereby minimized.
Influent suspended solids concentrations varied during these tests in
the same approximate proportions as were observed when the sewage
was treated with ferrous chloride alone or lime alone,^ ranging from an
average low of 164 mg/1 in Test 9 to an average high of 193 mg/1 dur-
ing Test 12. Daily variations from 70 to 400 mg/1 were also recorded,
the latter figure attributed mainly to the addition of sludge from the
Madison Plant. Overall suspended solids removal for all test periods
was 61.5 percent, relatively high when hydraulic overloading and the
introduction of extraneous solids to the influent is considered. Re-
movals fluctuated less widely on a day-to-day basis than was noted
during additions of ferrous chloride or lime alone. It was evident
from the clarity of the effluent, even after samples were allowed to
stand for several hours, that more efficient settling had occurred.
Reductions in oxygen demand were not observed to accompany in-
creased solids removals to the extent that occurred during the lime
addition periods. Overall BOD and COD removals were 61. 5 and
61. 1 percent, respectively, with occasional removals in excess of 75
percent observed for both contaminants.
No phosphorus removals lower than 30 percent were recorded in any
test period with iron and lime, and the lower removals were attributed
in every case to equipment malfunction or dosage inefficiency. Ex-
treme hydraulic fluctuations were not as influential as even simple
equipment failure. An overall combined average of 78.4 percent phos-
phorus removal was attained. Removals in excess of 90 percent could
have been attained had means been available for removal of the iron
phosphate precipitate in the effluent.
An examination of Table VII shows that in respect to all contaminant
removals, the combination of ferrous chloride and lime together yield-
ed results superior to that obtained when used separately. The value
of lime additions to raw sewage treated with ferrous iron was there-
fore demonstrated. Contaminant removals were high, and reduced
effluent iron concentrations were regarded as a valid characteristic of
the method.
38
-------
TABLE VII
Comparison of Contaminant Removals:
Ferrous Chloride and Lime Addition,
Separately and in Combination
Sewage Additive, Percent Contaminant Removal Effluent
Treated mg/liter Suspended BOD COD Phos. pH Iron
Gallons Fe Lime Solids as"P" In .Out mg/1
37,506,000 46.5 -- 18.1 27.4 19.9 25.2 7.6 7.! 42.5
29,435,000 -- 66.5 56.0 17.1 32.0 43.7 7.68.6
368,760,00039.374.3 61.5 61.1 55.3 78.4 7.47.6 11.6
Effluent Iron and Phosphorus
Iron was not regarded as a normal contaminant in this work, but was
investigated in light of the possible contamination of the sewage effluent
by iron in excess of that required for precipitation. There was conclu-
sive evidence during testing with ferrous chloride alone that even when
less than stoichiometric quantities of iron were used a finite residue of
iron remained in the effluent.
Grab sample analysis indicated an average of 1. 0 to 3. 0 mg/1 of iron
in the plant effluent when no chemical treatment was applied. Asa
whole the period of ferrous chloride addition without lime was charac-
terized by the visible formation of a hydroxide floe in the settling
tanks. Analyses of the effluent showed an average iron concentration
of 42. 5 mg/1, ranging from 21 mg/1 to 57 mg/1 as an apparent function
of input iron feed rates and pH fluctuations. Analysis of the plant efflu-
ent showed that over 80 percent of the iron was in an insoluble form,
indicating that iron precipitation and hydrolysis reactions had been
largely completed, but that maximum settling efficiency had not been
attained.
Subsequent plant testing with ferrous chloride and lime together result-
ed in significant reductions of both total and soluble effluent iron. It
can be seen in Table VII, a comparative summary of contaminant re-
movals, that total effluent iron concentration decreased from the 42. 5
mg/1 recorded with ferrous chloride alone to an approximate average
39
-------
of 11.6 mg/1 for ferrous chloride with lime. During the tests with
ferrous chloride addition alone, 80 percent of the total effluent iron was
insoluble, as compared with 90 percent insoluble effluent iron during
tests with ferrous chloride and lime together.
A laboratory study was undertaken to determine the ratio of soluble to
insoluble forms of iron and phosphorus in the principal process streams.
A standard laboratory fiberglass filter was used for solids determina-
tion. Test results indicated that over 90 percent of the effluent iron
was insoluble and in suspended form and therefore amenable to separa-
tion by microstraining or filtration. Effluent phosphorus was found to
be 85 percent insoluble in the same tests, and it was noted that the pro-
portion of insoluble phosphorus compounds increased with time, even-
tually reaching levels in excess of 90 percent. It may therefore be pre-
sumed that retention of a chemically-treated effluent in a holding basin,
or secondary process, would result in higher plant removals.
It is evident that if high concentrations of insoluble iron exist in the
effluent and if the greater portion of the phosphorus in the effluent is
also in an insolubilized form, then the reported phosphorus "removals"
will not accurately reflect reaction efficiency. The concentrations of
insolubilized effluent iron were successfully reduced, along with in-
creased phosphorus removals, during treatment with both iron and lime
as a supplement; this subject is discussed later in this report.
Since over 90 percent of the 11.6 mg/1 effluent iron concentration re-
corded during testing with ferrous iron and lime was insoluble, it can
be assumed that the actual degree of phosphorus insolubilization was
higher than the 78.4 percent overall average recorded in Table VII
would indicate. The theoretical weight ratio requirement of Fe:P is
2. 70:1, requiring at least 7. 6 mg/1 of iron as Fe to insolubilize the
2. 8 mg/1 of effluent phosphorus noted for the combined tests. Since
10-. 4 mg/1 of iron were available for this purpose, it is reasonable to
assume that the actual degree of phosphorus precipitation was in excess
of 90 percent, but not reflected in the recorded removals due to inability
to permanently remove all of the precipitate from the plant stream.
Oxidation of Ferrous Iron
Numerous laboratory tests were conducted to determine the oxidation
state of the iron in the chemically-treated effluent and sludge at Mentor.
Since ferrous iron oxidizes rapidly to ferric at pH levels above neutral,
and since ferric iron hydrolyzes under more acidic conditions than does
ferrous iron', it was hypothesized that the high effluent iron concentra-
tions recorded during plant testing (42. 5 mg/1) with ferrous chloride
40
-------
alone could be lowered by rapid formation of ferric hydroxides above
the minimum iron dosage requirements for phosphate precipitation.
If so, the conditions for effective treatment (air, chemical dosage,
influent chlorination) could be modified in future applications to favor
oxidation.
In general, the tests showed a high degree of oxidation of ferrous iron
to Łhe ferric form. The oxygen demand of the sewage proved to be of
secondary importance, since the oxidation of the iron took place before
the oxygen demand of the sewage was satisfied. Laboratory findings
indicated that the greater portion of iron found in both the sample sedi-
ment and supernatant was in the ferric form.
During one representative test, summarized in Tables VIII and IX, a
constant weight ratio of Fe:P of 3. 1:1 was maintained in a 1000 rnl
sample of raw sewage. Weight ratios of lime to iron were varied and
the sample supernatant and sediment tested for presence of ferric iron
with diphenyl phenanthroline. Molar dosage equivalents are not given
to avoid mixing nomenclature as would be required in the case of lime.
Samples 1, 2, 3, and 4 were prepared by adding 80, 100, 100 and 120
mg/1 of lime, respectively. Tests were conducted using the laboratory
procedures outlined in Section V, but with certain variations. The
effect of air was determined by mixing Sample No. 4 for 30 minutes with
100 ml/min Of air dispersed through a ceramic filter disc 4 square
inches in area, followed by 30-minute settling. The sediment was sam-
pled by decanting and then diluting the sample with concentrated hydro-
chloric acid to 100 milliliters. Aliquots were taken for analysis.
Test Conditions: Raw Sewage 1000ml
Phosphorus 18. 4 mg/1 as P
Treatment: Iron 60 mg/1
Lime varied
Test Results:
TABLE VIII
Oxidation of Ferrous Iron - Supernatant
Laboratory Test Results:
Sample
No.
1)
2)
3)
4)
Weight
Ratio
Lime:Fe
1. 3:1
1.7:1
2. 0:1
2. 0:1
Weight
Ratio
Lime:P
4. 3:1
5.4:1
6.5:1
6. 5:1
Sample Supernatant, mg/liter
Total
Iron
14. 1
8.8
6.6
4.4
Phos.
as"P"
2. 1
1.5
1.3
2.2
Ferrous
Iron
2.8
1.9
1.4
1. 1
Ferric
Iron
11.3
6.9
5.2
3.3
Percent
Phosphorus
Removal
88.7
91. 7
93.2
88.3
41
-------
TABLE IX
Oxidation of Ferrous Iron - Sediment
Laboratory Test Results:
Sample
No.
1)
2)
3)
4)
Weight
Ratio
Lime:Fe
1. 3:1
1. 7:1
2. 0:1
2. 0:1
Weight
Ratio
Lime:P
4.3:1
5.4:1
6.5:1
6.5:1
Sample
Total
Iron
42. 0
46.0
53.0
53.0
Sediment, mg /liter
Ferrous
Iron
8.4
10. 0
10. 9
3.4
Ferric
Iron
33.6
36. 0
42. 1
49. 6
High weight ratios of ferric to ferrous iron were noted in the sample
eupernatants and sediments. Especially high ratios were noted when
air was used for mixing.
Ferrous Chloride and Lime Dosage Requirements
The concentration of phosphorus in the raw sewage at Mentor normally
varies by a factor of five or more throughout any day, with the total
input quantity varying to a greater degree as a result of flow variations
since the highest phosphorus concentrations normally occur during
the periods of highest flow. It was therefore important that precipi-
tants be added at a concentration sufficient to meet the phosphorus
demand. Addition of an excess adds to chemical cost and increases
the concentration of soluble iron in the effluent.
When ferrous chloride pickle liquor was added directly to raw sewage
before primary sedimentation, without the use of supplementary addi-
tives, a negligible fraction of the phosphorus was removed in the pri-
mary process and insolubilized phosphorus precipitates were carried
over into the effluent. At the theoretical weight ratio of 2. 7:1, as-
suming ideal mixing and reaction efficiency, 100 percent phosphorus
insolubilization could be attained. The actual Fe:P proportions used
during test No. 3 were on the order of 3. 46:1, but only 26. 0 apparent
percent removal was attained. However, a considerable portion of the
effluent iron was present in the form of an insoluble iron-phosphate
precipitate so that the degree of insolubilization could not be defini-
tively measured simply by the extent of apparent removal.
Laboratory tests were performed to study the effect of various lime-
to-iron ratios on phosphorus removal. The weight ratio of ferrous
42
-------
chloride to phosphorus used was 2. 3:1 as Fe to P. Lime addition
ratios were then varied as a function of iron addition. Samples 1, 2,
and 3 were prepared by adding 50, 75, and 100 mg/1 of lime, respec-
tively. Standard laboratory procedure was followed by an additional
20-minute settling period before analysis.
Test Conditions: Raw Sewage 1000 ml
Phosphorus 14. 0 mg/1 as P
pH 7.9
Treatment: Iron 45 mg/1
Lime varied
TABLE X
Effect of Varying Lime to Iron Ratios
Sample Weight
No. Ratio
1)
2)
3)
Lime:Fe
1. 1:1
1.7:1
2.2:1
Weight
Ratio
pH
Lime : P
3. 6:1
5.4:1
7.2:1
8. 1
8.7
9.1
Total
Iron
Soluble Phosphorus
Iron
mg/1 concentration
8.8
5.7
4.3
1. 2
0.7
0.6
3.3
2.6
2. 1
Percent
Phosphorus
Removal
76.4
81.4
85. 0
The high lime-to-iron ratios used in this test were observed to reduce
the total quantities of effluent iron and phosphorus, but exerted little
effect on the ratio of insoluble to soluble iron. The increased phospho-
rus removals were attributed to enhanced settling and precipitation by
the lime itself. The extent or relationship of these two variables could
not be quantified as a general rule, however, so that dosage levels
must ultimately be derived on an empirical basis in plant tests.
Table XI is a list of comparative dosage levels and removals attained
with ferrous chloride and lime as indicated by the plant tests tabulated
earlier. The presence in the effluent of the insolubilized iron phosphate
precipitate is ignored, as is the probability of increased removals in
future applications should the precipitate be permanently removed from
the plant waste stream, by means such as filtration.
Inspection of Table XI shows that Tests 9 and 12 were characterized by
high weight ratios in every category, yet attained phosphorus removals
slightly lower than those of Test 11, in which the ratios were uniformly
43
-------
the lowest. The weight ratio of iron to phosphorus used in Test 13 was
lower than that required by theory for completion of the phosphorus
precipitation reaction, suggesting that a finite portion of the precipita-
tion was carried out by the lime, at a correspondingly greater chemical
cost.
High lime ratios were also examined in Test 10, with phosphorus re-
movals lower than those in Test 11 using a comparable iron to phos-
phorus ratio. Test 10, of 40 days duration, comprised the longest
single period on the table and probably represented the most nearly
typical case with no attempt at process optimization. Test 11 was a
shorter period but was characterized by generally better performance
in contaminant removals of all types and uniformly low chemical weight
ratios that still satisfied the theoretical requirements.
TABLE XI
Overall Summary of Ferrous Iron and Lime Dosage Levels
Test Additive, Influent Weight Weight Weight Percent
No. mg/liter Phosphorus Ratio Ratio Ratio -Phosphorus
Fe Lime mg/1 Fe:P LimerFe Lime:P Removal
9)
10)
11)
12)
13)
49.0
38.8
42.2
39.6
33. 7
78. 0
75. 0
68.7
82.0
68.3
14.
12.
13.
12.
12.
1
3
7
9
9
3.
3.
3.
3.
2.
4:1
2:1
1:1
1:1
6:1
1.
1.
1.
2.
2.
6:1
9:1
6:1
1:1
0:1
5.
6.
5.
6.
5.
5:1
1:1
0:1
4:1
3:1
81. 6
74.8
83.2
81.4
79.8
Combined Results:
39.3 74.3
12.9
3.1:1 1.9:1
5.8:1
78.4
Data from actual plant testing therefore indicate that with a 3. 1:1
weight ratio of Fe:P and a weight ratio of lime to iron varying from
1. 6:1 to 1. 9:1, phosphorus removals of approximately 80 percent were
attained. A lime to iron weight ratio of 1. 9:1 is recommended for the
conditions as existed at Mentor in future applications with the expecta-
tion that refinements in procedure and dosing equipment will permit
reductions in the required chemical quantities. With filtration or bio-
flocculation of the residual iron phosphate precipitate in the effluent,
removals in excess of 90 percent may be expected.
44
-------
The Use of Supplementary Additives
The addition of ferrous chloride alone was demonstrated to be benefi-
cial in precipitating phosphorus, but was also found to be detrimental
to settling efficiency because of the tendency of the iron phosphate pre-
cipitate to remain in suspension and carry over into the effluent. The
use of hydrated lime produced a more dense and particulate floe and in
conjunction with ferrous chloride proved to be an efficient means of
separating the iron phosphate precipitate from the main plant stream.
However, some iron and phosphorus, averaging 11.6 and 2. 8 mg/1,
respectively, remained in the effluent mainly as solids.
Tests were therefore conducted to determine the value of various sup-
plemental additives, separately and in combination with ferrous chloride
and lime, The specific purpose of these tests was to find a means of
reducing effluent iron and phosphorus concentrations in a manner that
would not interfere with the degree of treatment. Laboratory tests
were conducted using various concentrations of ferrous iron and lime;
several polymers of the anionic, cationic, and nonionic types were
also tested. Order of addition, time of contact and type of mixing and
flocculation were also studied. Plant tests with Dow Purifloc A-23
and Diamond Shamrock 630 and 640 polymers, and with sodium silicate
are summarized in Table XII. Tests with ferrous chloride and lime,
separately and in combination, are included for purposes of comparison.
Laboratory results are discussed where pertinent throughout the text.
Polyelectrolyte Polymer
The principal form of chemical treatment employed by the Mentor
Plant prior to the start of this program involved the use of Dow A-23
polymer as a settling aid. Three separate plant tests were conducted
with this material for comparison with the results obtained with ferrous
chlo.ride and lime addition. The original tests were characterized main-
ly by differing air applications. The influence of air on settling was a
factor in these experiments, since parallel work had shown 85. 0 CFM
to be excessive and detrimental to settling efficiency. Air was applied"
to different primary settlers during Tests 15 and 16 and these tests
are therefore not included. Test 14, however, covered a 31-day oper-
ating period under stable conditions without any application of air and
is considered indicative of the overall degree of treatment typically
attained by the Mentor Plant. Overall results for Test 14 are listed
in Table XII.
It is evident that since polymer is not itself a phosphorus precipitant
the principal advantage commending its use is enhanced settling. In-
45
AWBERC LIBRARY U.S.
-------
TABLE XII
Comparison of Overall Contaminant Removals:
Test
No.
Chemical Additions
by Type and Combination
Ferrous
Chloride. Lime, and Supplementary Additives
Total Raw
No. Sewage
Days Flow
Gallons
14)
17)
18)
19)
20)
21)
22)
23)
Ferrous Chloride,2 FeCl2
Hydrated Lime, Ca(OH)2
Ferrous Chloride & Lime
Polyelectrolyte, Dow A-23^
Ferrous Chloride & Dow A-23
Lime ^ Dow A-23
Ferrous Chloride,
Lime 6. Calgon 3000
Ferrous Chloride,
Lime 6. Diamond 630
Ferrous Chloride,
Lime & Diamond 640
Ferrous Chloride,
Lime & Dow A-23
Ferrous Chloride, Lime &
Sodium Silicate'
11
12
90
31
11
2
19
11
10
89
17
37,506
29,435
368,760
108,810
50,710
3,090
57,000
52,272
40,500
397,040
68,510
,0005
,0005
,0006
,0007
,0006
,0006
,0006
,0006
,0006
,0006
,0006
Chemical Suspended
Additions Solids
mg/Liter
Fe Lime Other %Red
46.5 -- -- 18.1
66.5 -- 56.0
39.3 74.3 -- 61.5
0.65 56.1
48.0 -- 0.43 5.2
88.0 0.47 38.0
47.0 101.0 0.26 70.6
35.7 75.6 0.26 64.4
43.0 82.0 0.15 71.4
46.0 83.0 0.45 62.3
46.0 87.0 12. 04 70.9
BOD
7.Red
27.4
17.1
61.1
28.8
27.6
23.1
51.7
51.4
64.0
34.0
66.5
Total Effluent
Phosphorus Iron
COD (as P) pH mg/Liter
%Red
19.9
32.0
55.3
32.7
30.7
32.6
64.6
50.3
50.0
61.0
52.8
%Red
25.2
43.7
78.4
0.0
42.0
47.4
85.4
80.4
81.5
82.5
80.9
IN
7.6
7.6
7.4
7.7
7.4
7.3
7.6
7.6
7.2
7.4
7.3
OUT
7.1
8.6
7.6
7.5
6.9
9.1
8.1
8.0
7.5
7.6
7.5
OUT
42.5
--
11.6
--
36.0
--
8.3
9.9
10.4
12.8
13.1
1 Refers to results of single test listed only; all others combined results for type of chemical addition
2 Pickle liquor; 7-107. Fe by weight 5 Results from two different tests; 85.0 CFM air to different primary In each
Dow A-23 polyelectrolyte; 0.3% solution
4 Grade 40 silicate; 40% Na2S103
Air applied for mixing; 42.5 CFM to each tank
No air applied for mixing
-------
spection of Table XII shows that the suspended solids removal •with
A-23 was approximately 56 percent.
Reductions in chemical oxygen demand were generally a function of
suspended solids removal. BOD removals were uniformly low for all
three forms of treatment, with lime addition actually appearing to be
detrimental as discussed earlier in this section. Lime dosage was
however not optimized for its use as a sole additive.
It is apparent that the addition of polymer on a plant scale was more
effective than separate additions of ferrous chloride or lime in every
respect, save that of phosphorus removal. The fine polyelectrolyte
floe was observed to spread uniformly through the primary settler and
to aid in settling, with minimum effluent turbidity.
Ferrous iron or lime when used alone resulted in an effluent with high
contaminant concentrations. Iron without lime resulted in an unsightly
brown effluent with a suspended solids content only slightly reduced
from that of the influent. Numerous laboratory tests showed that poly-
mers were not capable of settling the insolubilized iron-phosphate pre-
cipitate to any significantly greater degree than normal primary sedi-
mentation alone.
Reference to Table XII shows that the average removal of suspended
solids during a plant test with ferrous chloride and polymer was 5. 2
percent. The removals of BOD and COD were 27- 6 and 30.7 percent,
respectively. Phosphorus removals averaged 42.0 percent. These
results do not differ greatly from those obtained from the uae of ferrous
chloride alone. The reduction in oxygen demand was not observed to
be significantly different from that attained when polyelectrolyte alone
was added. However, the contribution of solids by ferrous chloride
addition apparently counterbalanced the influence exerted by polymer
on settling, resulting in negligible net overall removals. Some settling
of the insolubilized iron-phosphate precipitate undoubtedly occurred,
but at the expense of overall performance. The use of polymer and
ferrous chloride together is therefore considered to be the most un-
favorable additive combination tested.
The use of 66. 5 mg/1 lime alone was shown in Tests 5 and 6 to yield
more acceptable results than ferrous iron alone; however, the overall
contaminant reductions were poor. The tabulated figures in Table XII
show somewhat improved contaminant removals by the use of 0.47
mg/1 polyelectrolyte in conjunction with 88. 0 mg/1 of lime. It was
observed that the disparity in performance between the primary settling
tanks was much less when polymer was used. Air was alternately
47
-------
applied to each primary tank during lime addition, with a correspond-
ing narrower spread in performance results. In any case, the data
indicate that settling efficiency was diminished, with suspended solids
removals declining from 56. 0 percent with lime addition to 38. 0 per-
cent when polymer and lime were used together. The use of a poly-
mer therefore did not improve the results obtained with lime addition
and, in respect to solids removal, was detrimental.
The final step in the sequence of testing was to use polymer as a sup-
plement to ferrous iron and lime additions. Numerous laboratory tests
were conducted to determine if polyelectrolyte compounds could be of
use in aiding the settling of the iron-phosphate precipitate and in so
doing improve overall phosphorus removals and reduce effluent iron
concentrations. As discussed earlier in this report, phosphorus re-
movals in excess of 80 percent were obtained using ferrous chloride
and lime.
Because settling conditions in full scale operation are more variable
than in the laboratory, tests were run on settling rates using various
concentrations of A-23 polyelectrolyte along with 60 mg/liter of iron
and 80 mg/liter of lime. Settling rate was definitely increased by in-
creasing amounts of polymer up to 1. 25 mg/1 with as little as 0. 1 mg/1
showing only slightly faster settling rates than tests with no polymer.
The function of settling vs time appeared to be straight line with about
ten minutes settling time required with no polymer. Greatest efficiency
was achieved only with the high polymer concentrations.
The effect on settling of the insolubilized iron-phosphate precipitate
was determined by taking grab samples of untreated raw sewage and
controlling the dosage of the chemical additives. The following test
was a typical example.
Test Conditions: Raw Sewage 1000 ml
Phosphorus 15. 1 mg/1
Acidity as CaCC>3 65 mg/1
Treatment: 55 mg/1 Iron
75 mg/1 Lime
Test Results:
Percent
Polymer Added Phosphorus Removal
T"5 No polymer added 82. 5
2) 0. 5 mg/1 Dow A-23 83.5
*3) 0. 5 mg/1 Dow A-23 83.0
* Polymer added 4 min. after lime addition
48
-------
The results indicated that no significant change in phosphorus removal
occurred as a result of polymer addition.
The polymer tested in the following experiment was "Atlasep 2A2" of
the weak-moderate anionic form. Supplementary additions of iron
were made to a sample of raw sewage to better gauge the effect of
polymer addition.
Test Conditions:
Treatment:
Test Results:
Raw Sewage
Total Iron
Lime
Polymer
1000 ml
120 mg/1
130 mg/1
varied
Polymer Added
Total Iron
mg/1
1) No polymer added
2) 0. 1 mg/1 "2A2" before lime
3) 0. 5 mg/1 "2A2" before lime
4) 0. 1 mg/1 "2A2" after lime
5) 0.2 mg/1 "2A2" after lime
6) 0. 5 mg/1 "2A2" after lime
23. 0
31.0
31.0
18.0
18. 0
18. 0
No significant change in residual iron content of the effluent was ob-
served when the polymer was added after the lime; however, the addi-
tion of polymer before the lime appeared to increase the effluent iron
content.
A great many polymer types--over thirty in all--were evaluated in the
laboratory and found to be equally ineffective in reducing effluent phos-
phorus and iron. Results with the six tested in the following experi-
ment typify the results obtained.
Test Conditions:
Treatment:
Raw Sewage
Phosphorus
Iron
Lime
Polymer
1000 ml
11.4 mg/1 as P
57 mg/1 as Fe
80 mg/1
0.3 mg/1
49
-------
Test Results:
Phosphorus Total
as "P" Fe
Polymer Added
1) No polymer added
2) Allstate #6
3) Atlasep 2&Z
4) Atlasep 5^5
5) Calgon 3000
6) Calgon WC 618
7) Nalco D 2339
mg/ liter
1.6
1.3
1.3
1.3
1.6
1.3
1.3
18.5
15.9
14.6
12.3
15. 0
14.1
14. 1
Pe rcent
Phosphorus
Removal
86.0
88.6
88.6
88.6
86.0
88.6
88.6
No significant reductions of effluent phosphorus beyond that obtained
with iron and lime alone were observed. The slightly lower removals
recorded in Tests 1 and 5 were accompanied by an equally slight rise
in effluent iron content.
In light of the work described above, plant testing was undertaken to
acquire data for comparison with ferrous chloride and lime treatment.
Four separate plant tests are included in Table XII, along with the com-
bined summary of results obtained with ferrous chloride and lime. All
data are derived from the tables introduced at the beginning of this
section.
In general, the addition of polyelectrolyte was found to result in an
overall degree of phosphate removal that was slightly superior to that
attained with ferrous chloride and lime alone although the significance
of the slight improvement is questionable. Suspended solids removals
for the two tests were ten percent higher. BOD removals were notably
lower for the Dow A-23 tests, averaging 34. 0 percent.
An examination of the results listed in Table XII shows that Tests 19
and 21 using Calgon 3000 and Diamond 640 polymers, yielded the most
favorable results. Contaminant removals were generally higher than
those obtained with the other polymers.
A comparison of results in Table XII shows that the addition of lime was
the most significant factor. The limeriron ratio without polymer was
1.9:1, as compared with ratios of 2. 1;1 and 2.2:1 used during supple -
mental additions with polymer.
50
-------
The uniform addition of large amounts of lime means that little oppor-
tunity for dosage deficiency exists on an hour-to-hour basis; and while
this results in increased chemical feed, enough lime is available at a
given time to meet the lime demand exerted by iron addition. Increased
efficiency of settling and phosphorus precipitation with increased re-
movals, and minimization of effluent iron all follow as a consequence.
The addition of ferrous chloride and lime together was sufficient to
attain high levels of performance, but removals were further enhanced
by supplements of polymer. Polymer may or may not prove beneficial
where iron and lime dosages are optimized.
Sodium Silicate
Sodium silicate was investigated as an additive with iron and lime in
both laboratory and plant tests. The silicate which was not activated
before use produced no significant improvement in treatment. Results
of the plant test are shown in Table II,
Settling With Coal
Ground coal sized to 18 x 80 mesh was used as an additive with iron
and lime while minus 80 mesh ground coal was used with Dow A-23
polymer. Laboratory tests indicated that a small dense floe which
settled faster than floe without coal was produced. Results of plant
tests, however, as shown in Table II, indicate that treatment was not
significantly improved by the coal.
Sludge Handling
The use of iron and lime adds to the weight and volume of sludge, most
of which is due to phosphate removal and higher suspended solids re-
movals. The remaining portion consists of calcium carbonate and
iron hydroxide. There was no significant influence of the use of ferrous
chloride and lime on routine plant operation throughout the program
other than this additional weight and volume of sludge, with a corres-
ponding increase on the load in the digestion process and sludge handl-
ing equipment.
Variations in plant load and in the mode of digester operation did not
permit a long-term study of sludge handling requirements during this
program. The recirculation of suspended solids to the primary settlers
from the digester supernatant also interfered with the determination of
solids balances, a condition which prevailed at Mentor even prior to the
start of ferrous iron and lime addition.
51
-------
Raw Sludge Production
The quantity of sludge removed at Mentor varied widely from day to
day, depending mainly on one or a combination of three factors:
1. Composition of the sludge, including the proportion
of chemical additives.
2. Characteristics of the digester supernatant and
frequency of return.
3. Hydraulic overloading, with consequent inconsistencies
in the quantity of sludge removed during the period of
overloading.
Short term tests indicated that in no case was the volume of sludge
produced by treatment with ferrous chloride and lime more than double
that produced by primary treatment without chemicals. Part of the
increase was due to increased solids return from the overloaded
digester.
Sludge Digestion
The No. 1 digester was designed for a minimum of 28 days digestion
at 98 F. Prior to the iron-lime phosphate removal test period the
digester residence time was a minimum of 36 days while during the
test period residence time ranged from 8 to 14 days. The decrease in
digestion time arose from three major factors:
1. Increase in plant wastewater flow rate from an average
of 2. 39 MGD in 1968 to 4. 36 MGD in 1970.
2. Increased suspended solids removal and the formation of
inorganic iron phosphate along with other iron, calcium,
and magnesium precipitates from the iron-lime treatment.
3. The creation of a primary settler-digester recycle load
of up to 100 percent due to poor No. 2 digester settling.
This was caused by the short digestion time and by a
low digestion temperature due to heat exchanger over-
loading.
Although digester residence time was shortened, stabilization did
occur and the superior dewatering characteristics of the iron-lime
sludge aided in its efficient disposal. Further, a decrease in both
total and soluble phosphorus concentrations in the digester supernatant
liquor was observed during and immediately following the periods of
chemical treatment. The volatile solids of the digested sludge aver-
aged 41. 9 percent during a typical test with iron and lime. Some de-
terioration in sludge digestion was observed during treatment with fer-
rous chloride and lime when accompanied by low digester temperatures
52
-------
as determined by a decrease in gas production. Sludge was examined
on several occasions and only trace amounts of soluble iron and phos-
phorus were found, although large amounts of insoluble iron and phos-
phorus were recirculating in the digester supernatant as shown in
Table XIII.
•TABLE XIII
-Typical Iron and Phosphorus Analyses During Sludge Handling
Note: all quantities
expressed in mg/1
Total Iron
Soluble Iron
Total Phosphorus (asP)
Soluble Phosphorus
Primary Sludge Analysis
Raw ^Digested
3750.0 1020.0
105.0 26.0
1140.0 393.0
81.5 32.7
Digester
Supernatant
3950. 0
trace
1240. 0
6.6
Gas Production
Treatment of sewage with ferrous chloride and lime was not observed
to inhibit the bacterial destruction of organic solids by the formation
of methane and carbon dioxide in the digesters. However, gas produc-
tion did decrease as temperature in the heated digester decreased, a
condition typically observed with high solids influents, hydraulic over-
loading and cold weather operation.
Supernatant Return
The return of digester supernatant liquor to the plant process adversely
affected the overall removal of phosphorus. The suspended solids in the
supernatant return averaged approximately 13, 900 pounds per day dur-
ing a typical test, ranging from approximately 2,200 to 14, 800 pounds
per day, at concentrations of between 0. 5 and 8. 5 percent. Less than
15 percent of the returns contained fewer than 1. 0 percent solids, and
the average was 4. 0 percent.
The settling properties of the high solids supernatants were extremely
poor, which can be attributed to the reduced residence time and tem-
perature in the heated digester. On many days the sludge that was with-
drawn for filtering contained less than five percent more solids than the
53
-------
supernatant itself. Laboratory cylinder tests indicated no continued
settling over a twenty-four hour period. However, adequate settling
was observed when the sludge was diluted with an equal volume of
water. Additions of polymer were ineffective.
All factors indicated that a simple digester system overload existed
that could be corrected by the addition of digester volume and heating
capacity. 9
Sludge Dewatering Tests and Conditioning with Coal
The Mentor Plant disposes of digested sludge solids b,y conditioning
with ferric chloride and lime, filtering on a rotary vacuum filter and
land filling of the resultant filter cake. Conditioning of the digested
sludge before filtration is accomplished by adding approximately 17
parts of lime and 5 parts of ferric chloride per 100 parts of sludge
(all by weight and on a dry basis). The chemicals are mixed with
water on a two pounds per gallon basis before adding to the sludge.
Laboratory tests were made to determine the dewatering properties of
sludge conditioned with pulverized coal in addition to the ferric chloride
and lime. The test method used a 0. 1 sq. ft. Buchner funnel in the
manner similar to that normally used with a test leaf filter. A vacuum
was applied to the Buchner, a coarse filter paper fitted to it, and the
funnel was immersed in conditioned sludge. The filtration cycle was
one minute with one half the time immersed and the rest drying. The
test variables and results are summarized in Table XIV. The results
indicated no significant improvement from the use of pulverized coal.
TABLE XIV
Sludge Dewatering Tests:
Use of Coal for Sludge Conditioning
Vacuum Percent Filtrate Solids, dry
Inches of Percent Sludge gal per sq Ibs per sq
Date
3-13-69
3-13-69
3-14-69
3-14-69
4-03-69
4-03-69
Coal Mercury
1/2 Ib/gal
-30+80 mesh
1/2 Ib/gal
-30+80 mesh
—
1/2 Ib/gal
-325 mesh
—
16.
12.
16.
16.
13.
0
5
0
0
5
Solids
5.
5.
10.
10.
8.
2
2
8
8
5
Solids
23.
16.
22.
23.
18.
7
8
0
0
5
ft per
28.
27.
28.
29.
36.
hr
6
3
6
5
2
ft per hr
20.
13.
38.
40.
56.
2
2
8
4
1
(clear)
9.
0
8.
5
16.
8
42.
3
All sludge pre-conditioned with ferric chloride and lime.
54
-------
Section VI
SEWAGE TREATMENT WITH FERROUS IRON AND LIME
This section is a discussion of phosphate precipitation in primary
sewage treatment by ferrous iron additions and is based on the experi-
mental results presented in Section V. Emphasis is placed on prefer-
red parameters of operation and suggestions for succeeding applica-
tions of the method. Mention of options available with the use of
ferrous iron is made where appropriate. The Section concludes with
an evaluation of expected treatment costs.
Work at the Mentor Plant demonstrated that additions of ferrous
chloride without supplements to an existing primary process were
effective in precipitating the contained phosphorus in the raw influent.
The precipitation reaction proceeded to rapid completion under turbu-
lent conditions and once the phosphorus was insolubilized, the overall
efficiency of removal depended on the particulate material being de-
livered to the primary settler in a settlcable form.
The use of a primary settler to separate the insolubilized precipitate
was not effective when ferrous iron was used alone, due to a tendency
of the finely-divided precipitates to carry over into the effluent. This
is not a process characteristic, but is the direct result of the physical
characteristics of primary treatment. With means available in a
secondary plant for separation of the iron-phosphate floe from the
waste stream, phosphorus removals in excess of 90 percent have been
reported. However, the use of a primary system alone, or in com-
bination with supplemental processes where carryover of the floe is
undesirable, requires the addition of a base such as lime or caustic
soda to attain adequate separation.
The minimum amount of iron additive required is based on the weight
of contained iron necessary to react stoichiometrically with the con-
tained phosphate as "P" with no allowance for excess, and is 2. 70:1.
Plant tests indicate that 3. 1 weights of ferrous iron per weight of con-
tained phosphorus, as P, are required for efficient phosphorus preci-
pitation, provided that this is taken as an overall average for a day's
operation. A diurnal division of chemical feed was used in this pro-
gram, with specific quantities needed derived from statistical observa-
tion of fluctuations in the influent phosphorus concentrations.
Criteria for Selection of a Base
When a strong base is added to the sewage subsequent to the iron addi-
tion, a bulky floe is formed which entrains the iron phosphate preci-
55
-------
pitate and flocculates iron present in excess of the stoichiometric
minimum, -- the latter appearing as hydroxides. Separation of the
phosphate from the sewage is then accomplished by settling of the
mixed floe. Control of pH is an important factor. Ferrous iron re-
acts with phosphates in sewage of neutral pH, but raising the pH to the
range of 7. 5 to 8. 0 insures satisfactory floe formation.
The addition of lime to raw sewage pre-treated with ferrous iron pro-
moted floe formation during the work at Mentor, which aided in set-
tling. The proportion by weight of lime to iron was most influential
on effective settling. Plant data indicate that the preferred proportion
of lime to iron, by weight, ranges from 1. 6:1 to 1. 9:1, the precise
proportion requiring experimental adjustment in each plant application.
There is limited evidence that contribution of calcium ions by the lime
results in the formation of a certain amount of calcium phosphate even
at relatively low pH levels; but calcium precipitation is not considered
to present a particular cost advantage in this case.
Data from actual plant testing therefore indicate that with a 3. 1:1
weight ratio of Fe:P and a weight ratio of lime to iron varying from
1. 6:1 to 1. 9:1, phosphorus removals of approximately 80 percent can
be attained. A lime-to-iron weight ratio of 1. 9:1 is recommended in
future applications with the expectation that refinements in procedure
and dosing equipment will permit reductions in the required quantities.
With filtration or bioflocculation of the residual iron-phosphate preci-
pitate in the effluent, removals in excess of 90 percent may be expec-
ted.
The addition of additives supplemental to ferrous chloride and lime is
recommended for further study. When considered in terms of overall
performance, the degree of treatment attained with their use was not
significantly different from that obtained with ferrous chloride and
lime alone. The principal advantage of using coagulants would be to
accomplish enhanced suspended solids removal. However, at the
Mentor Plant there was no provable need for their routine use. A
precise determination of the advantages obtained from coal use could
not be determined. The use of supplements might be valuable should
ferrous chloride and lime be used in a plant other than at Mentor. An
evaluation of these materials, with particular emphasis on cost com-
parison, could be of advantage; however, optimization of ferrous
chloride and lime additions should be undertaken at the same time to
insure a base line of control.
56
-------
Chemical Costs
The cost of phosphorus removal by the addition of ferrous iron pickle
liquor and lime is reflected mainly in the two categories of chemicals
and equipment. The labor required is minimal, and consists mainly
of monitoring, unloading and storage of chemicals, and the filling of
feeders. On the basis of plant experience these activities can be
accommodated by the normal plant staff, and provision for labor spe-
cific to the phosphorus removal operation is therefore not included as
a general requirement.
Chemical Sources
Ferrous iron reacts with dissolved orthophosphates and is the cheap-
est of the known chemical precipitants. A common commercial form
is the hydrated salt, produced as a low cost source of iron in fertili-
zers. However, ferrous iron--either as the chloride or sulfate--is
also available as spent pickling acid, "pickle liquor", from the acid
cleaning of steel.
Laboratory work prior to plant testing at Mentor demonstrated the
equal effectiveness of both common ferrous forms--chloride and sul
fate--and both are recommended for use as local supplies permit. The
specifications for ferrous chloride pickle liquor used at Mentor were
generally satisfied by the supplier; however, occasional delivery of
materials high in trace elements, or low in iron content; did occur.
No deleterious effects on treatment were noted by such deviations,
suggesting that wider latitude can be employed in future applications.
Testing of random samples of waste pickle liquor is therefore suggest-
ed to determine on a plant scale whether restrictive specifications will
be required in future applications. It should be understood that pickle
liquor or other additives must contain no impurities which would cause
the effluent to present any hazard or to interfere with subsequent use
of the discharged water.
At present ferrous iron pickle liquor is available at no cost except that
of freight and possibly nominal handling charges. The delivered cost
of iron as pickle liquor, based on truck haulage of thirty miles, has
been quoted at $0. 02 per pound of contained iron. This price is as-
sumed in the estimation of operating costs. The delivered cost of hy-
drated lime, in bags, ranges typically between $20. 00 and $23. 00 per
ton. The latter value is used in this estimate.
Equipment Costs
Actual equipment cost will vary in every case according to both the
plant capacity and the phosphorus concentration of the wastewater, and
57
-------
must therefore be discussed here only in general terms. Installation
of certain equipment and facilities discussed here but not used during
the experimental program may be expected to increase substantially
the effectiveness of ferrous chloride and lime additions.
Equipment requirements may be divided into two groups: Items re-
quired as a minimum in any plant for the handling and feeding of chem-
icals, and items which may or may not be required for the handling
of sludge.
Chemical Feed Equipment
Pickle liquor is received as an acidic solution containing 5 to 10 per-
cent by weight of iron; and a suitable non-corroding storage reservoir
is required. For optimum phosphorus removal and economy the pickle
liquor is metered into the raw sewage on a continuous basis by a
variable-rate positive displacement pump with the rate varied in pro-
portion to the total amount of phosphorus contained in the influent.
As in all methods of chemical phosphate removal, the rate of chemical
feed should be proportioned to the concentration of influent phosphorus.
Where the daily fluctuations in phosphorus content are reasonably well
known and predictable the pump may be adjusted manually, though this
method usually results in the addition of either an excess or deficiency
of chemical at any given time. A desirable alternative is the installa-
tion of an automatic phosphorus analyzer, with iron feed control con-
tinuously adjusted on the basis of the running analysis. Several experi-
mental analyzers are available with roughly comparable performance,
at prices ranging between twenty and thirty thousand dollars.
As in the case of any chemical reaction, effective mixing is required
and a zone of turbulence is recommended for effective mixing of
ferrous iron solution with sewage. Approximately 40 gallons of 7 per-
cent pickle liquor are required per million gallons of sewage for each
mg/1 of contained phosphorus, and optimum results are obtained only
when the liquid is well mixed into the sewage. On the basis of this
plant study special mixing equipment is not needed when the pickle
liquor can be introduced into a region of turbulence, such as at a
pump station. Air was used in this program for additional mixing due
to the hydraulic overloading at the Mentor Plant and subsequently
shortened primary detention times. In special cases flash mixing
might be necessary.
Lime must also be metered into the sewage, in proportion to the rate
of iron addition. The feeder may be adjusted manually, subject to the
58
-------
same conditons that apply to iron; but when automatic analysis and
proportioning control are available, the lime feeder can be controlled
by the same signal that controls the iron feeder plus adjustments to
attain a suitable pH.
Lime may be introduced in the dry form, but is not efficiently mixed
except in a zone of extreme turbulence. Continuous conversion of the
dry lime to a water slurry after metering is recommended, with the
slurry then mixed into the sewage as in the manner of the pickle liquor.
All of the chemical handling equipment can be added to an existing pri-
mary treatment plant and used, or not used, as desired without inter-
ference with the basic plant structure or operation. In the event the
plant is later expanded to secondary activated sludge treatment, where
lime is not required as a supplement to the iron, the installation can
be used directly with only the retirement of the lime handling facility.
Treatment Cost
At the recommended dosage of 3. 1 weights of iron per weight of con-
tained phosphorus, as P, a total of 25. 9 pounds of iron is required
per million gallons for each mg/1 of contained phosphorus, represent-
ing an iron cost of $0. 52.
At the recommended maximum dosage of 1. 9 weights of hydrated lime
per weight of iron, equivalent to 5. 9 weights of lime per weight of
phosphorus, 49.2 pounds of lime are required per million gallons for
each mg/1 of phosphorus, representing a lime cost of $0. 57.
The combined chemical cost, then, is $1. 09 per million gallons of
sewage treated, for each mg/1 of contained phosphorus. Total chemi-
cal cost is this value times the phosphorus concentration of the influent
wastewater in milligrams per liter. For example, in the case of a
sewage containing 13 mg/1 phosphorus, as was typically received at
the Mentor Plant, the total chemical cost would be $14. 17 per million
gallons.
Sludge Handling
The weight and volume of sludge produced in a conventional primary
plant—and thus the size and actual cost of digesters or other sewage
handling equipment--vary in each case with the plant flow and with the
concentration and nature of the solids removed from the sewage. The
requirement for special sludge handling conditions, if necessary at
59
-------
all, therefore can be measured only for each plant situation.
The weight increase in sludge solids due to the precipitation of phos-
phates by ferrous iron and lime was difficult to determine at the
Mentor Plant by direct observation due to the digester overload condi-
tions which existed. The use of the chemicals increased the suspended
solids removals from an average of 35. 8 percent when no chemicals
were used to an average of 61. 5 percent. At an average of 190 mg/1
suspended solids content in the raw sewage this is an increase of approx-
imately 420 pounds to a total of approximately 975 pounds removed per
million gallons of raw sewage.
The use of ferrous chloride and lime will produce approximately 100
pounds of various iron and calcium phosphate, hydroxide, carbonate,
etc. , compounds per mg/1 of phosphorus contained in the raw sewage.
At approximately 13 mg/1 of phosphorus in the raw sewage, this repre-
sents 1300 pounds of additional sludge per million gallons. The expect-
ed total sludge solids produced was approximately 2300 pounds. For an
extended period, an average of 6134 pounds per million gallons were
pumped to the digesters with more than one half of that contributed by
the supernatant return.
The important factor of sludge volume increase is also variable accor-
ding to influent sewage composition and plant efficiency, but in no case
appears to be more than twice that obtained in normal primary treat-
ment.
Waste sludge from ferrous chloride and lime treatment was thickened
and disposed of by the techniques normally employed at the Mentor
Plant. The increased volumes of sludge caused digester temperature
to drop and solids concentration in the digester supernatant to increase
to very high levels. The phosphorus contained in the sludge remained
insoluble throughout sludge treatment and was not resolubilized in the
digester supernatant.
The determination of sludge handling cost for iron-lime treatment
must be made for each plant. As a maximum the cost difference will
be that of doubling the capacity of the facility. As a minimum no in-
crease is necessary. In the Mentor Plant, the capacity of the digesters
and filters was taxed, to be sure, but not to the point of breakdown.
An increased capacity would have resulted in a greater improvement
in degree of treatment due to the recirculation of fewer solids in the
digester supernatant, but was not mandated by the use of ferrous iron.
60
-------
Section VII
A CKNO WLED GEMEN TS
Appreciation is expressed to Ralph G. Christensen, Project Officer,
and John J. Convery and Dr. Sidney A. Hannah, Technical Advisors,
of the Environmental Protection Agency for their assistance in con-
ducting this project. The local EPA Water Quality Office in Cleveland
furnished valuable information during the early phases of the program.
The experimental and demonstration work was conducted at the
Willoughby-Mentor Wastewater Treatment Plant of the County of Lake
in Mentor, Ohio. Assistance was received from Dudley B. Rose, the
Lake County Sanitary Engineer, in this work. Superintendant Fred
Vargo and his staff at the Mentor Plant were most cooperative in pro-
viding pertinent data on plant operations and sharing laboratory space.
The experimental work was performed by the late E. Thomas Alvord,
Chief Chemist of the Rand Development Corporation and Project
Engineer, with the technical assistance of Warren Blanchard, Ronald A.
Chiancone, Dennis M. Gaughan, Clyde M. Marr, John Nawalanic,
Bertram C. Raynes, and James Surman.
The report was written by E. Thomas Alvord, Dennis M. Gaughan, and
Clyde M. Marr. Stuart S. Carlton provided pertinent commentary and
manuscript review which are gratefully acknowledged, as are the use-
ful suggestions made by Bertram C. Raynes.
The manuscript was typed by Mrs. Lois King and Mrs. Carolyn Noland.
Work on this project was conducted under the supervision of Mr.
Thomas Colpetzer, P. E. , Project Director.
61
-------
Section VIII
REFERENCES
1. Ohio Department of Health Reports of Operation, Sewage Treat-
ment Works at Mentor, Ohio (January 1967-December 1970).
2. Ibid. , Reference No. 1.
3. Standard Methods for the Examination of Water and Wastewater,
Twelfth Edition, 1966. American Public Health Association, Inc.,
170 Broadway, New York 19, N. Y.
4. Carlton, S.S., et al. , "Development of a Coal-Based Sewage -
Treatment Process. " Office of Coal Research, Research and
Development Report No. 55; U.S. Dept. of the Interior, Washing-
ton, D. C. (1971).
5. Connell, C.H. , "Phosphorus Removal and Disposal from Municipal
Wastewater at Texas City, Texas". Federal Water Quality Adminis
tration Research Grant No. 17010 DYB to University of Texas,
Medical Branch (Summer, 1969).
6. Brenner, Richard C., "Combined Biological-Chemical Treatment
for Control of Phosphorus", Presented at the Advanced Waste
Treatment and Water Reuse Symposium, Dallas, Texas (January
12-14, 1971).
7. Stumm, Dr. Werner and Singer, Phillip C., "Oxygenation of
Ferrous Iron: The Rate-Determining Step in the Formation of
Acidic Mine Drainage. " Federal Water Quality Administration
Publication No. 14010 ---06/69 (December, 1968).
8. Encyclopedia of Chemical Technology, Volume 12, Ed, by Kirk,
Raymond E. , and Othmer, Donald F. The Inter science Encyclo-
pedia, Inc. New York, 1954. Under Silica and Silicates (Soluble),
p. 327.
9. The Effect of Short-Term Temperature Variations on Methane
Production, by R.E. Speece and Jan A. Kern, Journal Water
Pollution Control Federation, page 1993, November, 1970.
10. Connell, op. cit. , Reference No. 5.
63
-------
Additional Sources
11. Alvord, E. T. , and Gaughan, D. M., Interim Report: "Pollution
Abatement Program for Mentor, Lake County, Ohio". Federaj.
Water Quality Administration Research Grant No. 172-01-68 to
Lake County, Ohio (January, 1971).
12. Alvord, E.T., and Surman, J. "Pollution Study of Marsh Creek
and the Mentor Marsh. " Federal "Water Quality Administration
Publication No. 11010 EGO 09/70 (September, 1970).
13. Anon. , "Studies on Removal of Phosphates and Related Removal
of Suspended Matter and Biochemical Oxygen Demand at Lake
Odessa, Michigan", Wastewater Section, Division of Engineering,
Michigan Department of Public Health, and the Dow Chemical Co.
(October, 1967).
14. Earth, E. F. , "Mineral Controlled Phosphorus Removal in the
Activated Sludge Process". Journal Water Pollution Control
Federation, Vol. 39, No. 8, 1362 (August, 1967).
15. Earth, E. F. , et al. "Summary Report on the Effects of Heavy
Metals on the Biological Treatment Processes11, Journal Water
Pollution Control Federation, Vol. 37, No. 1, 86 (January, T9~65).
16. Babbitt, H. E. , and Baumann, E. R., "Sewerage and Sewage
Treatment". John Wiley and Sons, 1958.
17. Buzzell, J. C. and Sawyer, C. N. , "Removal of Algae Nutrients
from Raw Wastewater with Lime". Journal Water Pollution
Control Federation, Vol. 10, R 16 (October, 1967).
18. Carlton, S. S., et al., "Investigation of the Use of Coal for
Treatment of Sewage and Waste Waters". Office of Coal Research,
Research and Development Report No. 12, U. S. Department of
the Interior, Washington, D.C. (1967).
19. "Development of Phosphate Removal Process", Federal Water
Quality Administration Research Grant No. WPRD 51-01-67 to
Detroit Metro Water Department (July, 1970).
20. Eliassen, R. , and Tchobanoglous, G. , "Chemical Processing of
Wastewater for Nutrient Removal". Journal Water Pollution
Control Federation, Vol. 40, No. 5, R 171, 37 (May, 1968).
64
-------
21. Fair, G. M. , and Geyer, J. C. , "Water Supply and Waste-Water
Disposal". John Wiley and Sons, 1954.
22. Fair G. M. , Geyer, J C. , and Okun, D. A. , "Water and Waste -
water Engineering, " Vol. 1, Water Supply and Waste-water
Removal. John Wiley & Sons, Inc., New York, N. Y. (1966).
23. Gaughan, Dennis M. , and Carlton, Stuart S. , "Report and
Recommendations for Phosphorus Removal at the Willoughby-
Mentor Wastewater Treatment Plant, " Rand Development
Corporation Publication No. 422-71 (April, 1971).
24. Hannah, Sidney A., "Chemical Precipitation of Phosphorus",
Presented at the Advanced Waste Treatment and Water Reuse
Symposium, Dallas, Texas (January 12-14, 1971).
25. Johnson, E. L., Beeghly, J. H., and Wukasch, R. F. , "Phospho-
rus Removal at Benton Harbor - St. Joseph, Michigan", Dow
Chemical Company, Midland, Michigan.
26. Recht, H. L. , and Ghassemi, M. , "Kinetics and Mechanism of
Precipitation and Nature of the Precipitate Obtained in Phosphate
Removal from Wastewater Using Aluminum (III) and Iron (III)
Salts" Environmental Protection Agency, Water Quality Office,
Report No. 17010 (EKl) (64/70) (April, 1970).
27. Recht, N. L. , and Ghassemi, M. , "Phosphate Removal From
Wastewaters Using Lanthauum Precipitation. " Environmental
Protection Agency, Water Quality Office, Report No. 17010
(EFX) (04/70) (April, 1970)
28. Singer, P. C. , "Anaerobic Control of Phosphate by Ferrous Iron".
Presented at 43rd Annual Conference--Water Pollution Control
Federation, Boston, Massachusetts (October, 1970).
29. Smith, R. , "Cost of Conventional and Advanced Treatment of
Wastewater" Journal of the Water Pollution Control Federation,
Vol. 40, No. 9, 1546 (September, 1968).
30. Tenney, M. W. and Stumm, W., "Chemical Flocculation of
Micro-organisms in Biological Waste Treatment". Journal
Water Pollution Control Federation, Vol. 37, No. 1, 1370-1388
(January, 1965).
65
-------
31. Thomas, E. A., "Phosphate Elimination in the Activated-Sludge
Plant at Mannedorf and Phosphate Fixation in Lake and Sewage
Sludge". Chemical Abstracts, Vol. 65, 5206 (1966).
32. Thomas, E. A., "Phosphate Hypertrophy of Waters. Necessity
and Technical Possibility of Throttling Feed". Chemical
Abstracts, Vol. 68, 24418 (1968).
66
-------
Section IX
APPENDICES
TABLE XV: Summary of Experimental Conditions
TABLE XVI: Summary of Experimental Results:
Solids Removal
TABLE XVII: Summary of Experimental Results:
Oxygen Demand Reduction
TABLE XVIII: Summary of Experimental Results:
Phosphorus Removal
Page No.
68
69
70
71
67
-------
TABLE XV
Summary of
Test
No.
Chemical Additions
by Type and Combination
No.
Days
In
Test
Experimental Conditions
Flow
Rate
mgd
Influent
Acidity
as CaC03
mg/Liter
Application
Chemical
Additions
mg/Liter
t'e
1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
12)
13)
14)
15)
16)
17)
18)
19)
20)
21)
22)
23)
24)
25)
None
None
Ferrous Chloride} FeCl2
Ferrous Chloride, FeCl2
Hydrated Lime, Ca(OH)2
Hydrated Lime, Ca(OH)2
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Polyelectrolyte, Dow A-23
Polyelectrolyte, Dow A-23
Polyelectrolyte, Dow A-23
Ferrous Chloride & Dow A-23
Lime & Dow A-23
Ferrous Chloride,
Lime & Calgon 3000
Ferrous Chloride,
Lime 6. Diamond 630
Ferrous Chloride,
Lime & Diamond 640
Ferrous Chloride,
Lime & Dow A-23
Ferrous Chloride, Lime &
Sodium Silicate3
Ferrous Chloride, Lime &
Coal4
Dow A-23 & Coal5
9
7
4
7
5
7
13
11
11
40
9
13
17
31
7
5
11
2
19
11
10
89
17
22
5
3
2
2
3
2
2
3
2
3
3
4
4
4
3
2
2
4
2
.210
.670
.674
.830
.205
.630
.678
.710
.680
.970
.250
.210
.500
.510
.520
.338
.610
.545
3.000
4.
.752
4.050
4.460
4.030
3.
3.
800
033
- .
— —
41
49
-
- -
43
50
49
38
48 42
26 39
21 33
_
-
-
48
.0
.0
.0
.0
.0
.8
.2
.6
.7
.0
of Air
for Mixing
cu.
ft./min.
Lime Other
.
-
_
-
69
63
55
103
78
75
68
.
-
—
-
.6
.0
.0
.0
.0
.0
.7
82.0
68.3
.
-
-
-
0.65
0.42
0.45
0.43
88.0 0.47
46 47.
21 35.
21 43.
45 46.
30 46.
48.
,0
,7
0
0
0
0
Poly-»
101.0 0.26
75.
82.
83.
87.
67.
6 0.26
0 0.25
0 0.45
0 12. O3
0,, 30. 04
0.74^ 51. 65
85
85
85
85
85
85
85
.0
.0
.0
.0
.0
.0
.0
85.0
42.5
42.5
42.5
42.5
42.5
No
West Tank
East Tank
East Tank
West Tank
East Tank
West Tank
East Tank
East Tank
Each Tank
Each Tank
Each Tank
Each Tank
Each Tank
Atr
85.0
85.0
85.
42.
42.
42.
42.
42.
42.
0
5
5
5
5
5
5
East Tank
West Tank
East Tank
Each Tank
Each Tank
Each Tank
Each Tank
Each Tank
Each Tank
42.5 Each Tank
42.5 Each Tank
Pickle liquor; 7-107. Fe by weight
Dow A-23 polyelectrolyte; 0.37. solution
Grade 40 silicate; 407. Na2Si03
18 x 80 mesh coal;added to east primary only
Minus 80 mesh coal; added to east primary only
68
-------
TABLE XVI
Summary of Experimental Results: Solids Removal
Test Chemical Additions
No. by Type and Combination
1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
12)
13)
14)
15)
16)
17)
18)
19)
20)
21)
22)
23 >
24)
25)
None
None
Ferrous Chloride J- FeCl2
Ferrous Chloride, FeCl2
Hydra ted Lime, Ca(OH).
Hydra ted Lime, Ca(OH>2
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Polyelectrolyte, Dow A-232
Polyelectrolyte, Dow A-23.
Polyelectrolyte, Dow A-23
Ferrous Chloride & Dow A-23
Lime & Dow A-23
Ferrous Chloride,
Lime & Calgon 3000
Ferrous Chloride,
Lime & Diamond 630
Ferrous Chloride,
Lime & Diamond 640
Ferrous Chloride,
Lime & Dow A-23
Ferrous Chloride, Lime &
Sodium Silicate3
Ferrous .Chloride , - Lime &
Coal4
Dow A-23 & Coal5
No. Flow
Days Rate
In mgd
Test
9
7
4
7
5
7
13
11
11
40
9
13
17
31
7
5
11
2
19
11
10
89
17
22
5
3.210
2.670
2.674
3.830
2.205
2.630
3.678
2.710
3.680
3.970
4.250
4.210
4.500
3.510
2.520
2.338
4.610
2.545
3.000
4.752
4.050
4.460
4.030
3.800
3.033
Chemica 1
Additions
ing/Liter
Fe
_
"
41.0
49.0
_
-
43.0
50.0
49.0
38.8
42.2
39.6
33.7
.
-
-
48.0
-
47.0
35.7
43.0
46.0
46.0
48.0
Lime
—
™
—
-
69.6
63.0
55.0
103.0
78.0
75.0
68.7
82.0
68.3
—
-
-
-
88.0
101.0
75.6
82.0
83.0
87.0
67.0.
Poly-»0.74z
Other
.
-
.
-
..
-
_
-
-
-
-
'-
-
0.65
0.42
0.45
0.43
0.47
0.26
0.26
0.25
0.45
3
12.0
30.04
51. 65
Total
Solids
mg/Liter
IN
845
840
834
805
752
752
730
651
753
785
748
757
755
925
810
844
778
752
710
755
752
758
640
720
790
OUT
705
766
755
797
690
702
719
690
774
779
739
773
731
700
742
719
792
753
721
702
743
782
762
722
739
Suspended
Solids
mg/Liter
IN
132
198
226
193
254
188
119
138
164
181
187
193
178
198
233
187
152
171
146
174
171
183
216
157
167
OUT
91
118
164
166
99
88
64
67
43
87
77
62
62
87
139
100
144
106
43
62
49
69
63
64
93
%Red
31.0
40.5
28.5
14.0
61.0
53.3
46.2
51.5
73.8
52.0
58.8
67.9
65.2
56.1
40.4
46.5
5.2
38.0
70.6
64.4
71.4
62.3
70.9
59.3
44.4
1 Pickle liquor; 7-10% Fe by weight
2 Dow A-23 polyelectrolyte; 0.3% solution
3 Grade 40 silicate; 40% Na2Si03
* 18 x 80 mesh coal;added to east primary only
5 Minus 80 mesh coal; added to east primary only
69
-------
TABLE XVII
Sunmarv of Experimental Results:
Test Chemical Additions
No. by Type and Combination
1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
12)
13)
14)
15)
16)
17)
18)
19)
20)
21)
22)
23)
24)
25)
None
None
Ferrous Chloride} Fed.
Ferrous Chloride, FeC^
Hydrated Lime, Ca(OH),
Hydrated Lime, Ca(OH)2
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Polyelectrolyte, Dow A-232
Polyelectrolyte, Dow A-23
Polyelectrolyte, Dow A-23
Ferrous Chloride & Dow A-23
Lime & Dow A-23
Ferrous Chloride,
Lime & Calgon 3000
Ferrous Chloride,
Lime & Diamond 630
Ferrous Chloride,
Lime & Diamond 640
Ferrous Chloride,
Lime & Dow A-23
Ferrous Chloride, Lime &
Sodium Silicate3
Ferrous Chloride, Lime &
Coal4
Dow A-23 & Coal5
No. Flow
Days Rate
In mgd
Test
9
7
4
7
5
7
13
11
11
40
9
13
17
31
7
5
11
2
19
11
10
89
17
22
5
3.210
2.670
2.674
3.830
2.205
2.630
3.678
2.710
3.680
3.970
4.250
4.210
4.500
3.510
2.520
2.338
4.610
2.545
3.000
4.752
4.050
4.460
4.030
3.800
3.033
Oxygen Demand Reduction
Chemical
Additions
mR/Llter
Fe
-
41.0
49.0
.
-
43.0
50.0
49.0
38.8
42.2
39.6
33.7
.
.
-
48.0
-
47.0
35.7
43.0
46.0
46.0
Lime Other
-
-
69.6
63.0
55.0
103.0
78.0
75.0
68.7
82.0
68.3
0.65
0.42
0.45
0.43
88.0 0.47
101.0 0.26
75.6 0.26
82.0 0.25
83.0 0.45
87.0 12. O3
48.0 67.0 30. 04
Poly-^0.742 51. 6 5
BOD
ms/LlLar
IN
153
192
187
155
143
224
161
135
174
198
205
201
192
173
188
173
192
173
151
173
222
136
239
180
184
OUT
123
194
134
113
119
185
72
86
71
73
78
74
85
123
144
145
139
133
73
84
80
98
80
97
112
XRed
19.5
-
28.3
27.0
16.7
17.3
55.3
36.4
59.2
63.2
61.9
63.2
55.7
28.8
23.3
16.2
27.6
23.1
51.7
51.4
64.0
27.9
66.5
46.2
39.1
COD
mx/Liter
IN
232
251
505
385
482
450
380
320
542
410
515
433
395
263
326
319
352
260
460
445
412
405
490
505
468
OUT
182
203
339
328
350
293
148
166
262
179
163
190
202
177
171
228
244
175
163
221
206
158
231
247
273
ZRed
21.4
19.1
32.8
14.7
27.4
34.8
61.1
48.2
51.7
56.3
68.4
56.2
47.8
32.7
47.6
28.6
30.7
32.6
64.6
50.3
50.0
61.0
52.8
51.1
63.1
1 Pickle liquor; 7-lOZ Fe by weight
2 Dow
A-23 polyelectrolyte; 0.3%
solution
3 Grade 40 Silicate; 40Z Na2Si03
4 18 x 80 mesh coal; added to east primary only
5 Minus 80 mesh coal; added to east primary only
70
-------
TABLE XVIII
Summary of Experimental Results:
Test Chemical Additions
No. by Type and Combination
1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
12)
13)
14)
15)
16)
17)
18)
19)
20)
21)
22)
23)
24)
25)
None
None
Ferrous Chloride, FeCl2
Ferrous Chloride, FeCl2
Hydrated Lime, Ca(OH)2
Hydra ted Lime, Ca(OH)2
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Ferrous Chloride & Lime
Polyelectrolyte, Dow A-232
Polyelectrolyte, Dow A-23
Polyelectrolyte, Dow A-23
Ferrous Chloride & Dow A-23
Lime & Dow A-23
Ferrous Chloride,
Lime & Calgon 3000
Ferrous Chloride,
Lime & Diamond 630
Ferrous Chloride,
Lime & Diamond 640
Ferrous Chloride,
Lime & Dow A-23
Ferrous Chloride, Lime &
Sodium Silicate3
Ferrous Chloride, Lime &
Coal4
Dow A-23 & Coal5
No. Flow
Days Rate
In mgd
Test
9
7
4
7
5
7
13
11
11
40
9
13
17
31
7
5
11
2
19
11
10
89
17
22
5
3.210
2.670
2.674
3.830
2.205
2.630
3.678
2.710
3.680
3.970
4.250
4.210
4.500
3.510
2.520
2.338
4.610
2.545
3.000
4.752
4.050
4.460
4.030
3.800
3.033
Phosphorus Removal
Chemical
Additions
me/Liter
Fe
-
41.0
49.0
-
-
43. 0>
50.0
49.0
38.8
42.2
39.6
33.7
_
Lime
-
_
-
69.6
63.0
55.0
103.0
78.0
75.0
68.7
82.0
68.3
_
Other
-
_
-
-
-
_
-
-
-
-
-
-
0.65
Total
Phosphorus
(as P)
me/Liter
IN OUT
15.5 16.4
14.0 16.2
16.7 12.8
14.2 10.5
16.3 9.5
14.8 8.2
12.1 3.7
10.5 4.2
14.1 2.6
12.3 3.1
13.7 2.3
12.9 2.4
12.9 2.6
13.1 13.1
0.42 -»17.8 15.5
-
48.0
-
47.0
35.7
43.0
46.0
46.0
48.0
-
-
88.0
101.0
75.6
82.0
83.0
87.0
67.0.
Poly-*0.74Z
0.45
0.43
0.47
0.26
0.26
0.25
0.45
12.0
30. 04
51. 65
17.6 13.3
12.4 7.2
19.0 10.0
13.7 2.0
10.7 2.1
11.9 2.2
13.1 2.3
13.1 2.5
14.1 2.5
13.3 12.2
%Red
-
23.4
26.0
41.8
44.7
69.4
60.0
81.6
74.8
83.2
81.4
79.8
0.0
12.9
24.5
42.0
47.4
85.4
80.4
81.5
82.5
80.9
82.3
8.3
Effluent
pH Iron
mg /Liter
IN
7.5
7.8
7.7
7.5
7.4
7.7
7.1
7.7
7.3
7.6
7.5
6.9
7.4
7.7
7.6
7.5
7.4
7.3
7.6
7.6
7.2
7.4
7.3
7.5
7.7
OUT
7.5
7.6
7.3
7.0
8.5
8.6
6.9
7.4
7.6
7.6
7.7
7.9
7.2
7.5
7.5
7.5
6.9
9.1
8.1
8.0
7,5
7.6
7.5
7.5
7.1
OUT
-
.
42.5
_
-
16.9
9.9
8.7
13.2
10.8
11.0
10.6
_
-
-
36.0
-
8.3
9.9
10.4
12.8
13.1
15.8
4.5
1 Pickle liquor; 7-107. Fe by weight
2 Dow A-23 p'olyelectrolyte; 0.3% solution
3 Grade 40 silicate; 407. Na2Si03
4 18 x 80 mesh coal; added to east primary only
Minus 80 mesh coal; added to east primary only
71
-------
1
Accession Number
w
5
Organization
1
Subject Field & Group
05D
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
County of Lake, Ohio
Court House
Painesville, Ohio kk077
Title
Phosphorus Removal by Ferrous Iron and Lime
10
Authors) Alvord, E. Thomas
Gaughan, Dennis M.
Marr, Clyde M.
Colpetzer, Thomas
Rose, Dudley B.
16
Project Designation
21
EPA Project #11010 EGO
Note
22
Citation
23
Descriptors (Starred First)
*Phosphorus Removal, *Waste Treatment, *Pickle Liquor, *Chemical Precipitation,
Ferrous Chloride, Iron, Lime Addition, Lake County Ohio
25
Identifiers (Starred First)
27
Abstract
When used in primary treatment, ferrous iron was effective in removing more
than 80.percent of the phosphorus contained in sewage, with spent pickle liquor
a satisfactory source of the metal ion. Required iron dosages ranged from the
theoretical minimum of 2.7 to a maximum of approximately 3.1 weights of ferrous
iron per weight of phosphorus contained in the sewage. When advanced means,
such as filtration, are available for more complete removal of the insolubilized
phosphate precipitate, phosphorus removals in excess of 90 percent are indicated.
Overall suspended solids removals of 6l.5 percent were attained over a 23-month
plant experimental program, with BOD and COD removals of 6l.6 and 55.3 percent,
respectively.
The combined chemical cost for ferrous chloride pickle liquor and lime was $1.09
per million gallons of sewage treated, for each mg/1 of contained phosphorus.
The total chemical costs for treating a sewage containing 13 mg/1 of phosphorus
as typically received during this work, would be $lU.17 per million gallons.
Abstractor
Patrick M. Tobin
institution Environmental Protection Agency
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
* OPO: 1970-389-930
------- |