EPA-600/2-76-294
December 1976
Environmental Protection Technology Series
TREATMENT OF EFFLUENT WATERS FROM
VEGETABLE OIL REFINING
£>
55E
UJ
O
%
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA 600/2-76-294
December 1976
TREATMENT OF EFFLUENT WATERS
FROM VEGETABLE OIL REFINING
by
Donald F. Gill, Jr.
James C. lelase
Archer Daniels Midland Co.
Decatur, Illinois 62525
Grant 12060 FDK
Project Officer
Clifford Risley, Jr.
U.S. Environmental Protection Agency
Region V
Chicago, Illinois 60606
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publication-
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
n
-------
FOREWORD
When energy and material resources are extracted, processed, con-
verted, and used, the related pollutional impacts on our environment and
even on our health often require that new and increasingly more efficient
pollution control methods be used. The Industrial Environmental Research
Laboratory - Cincinnati (lERL-Ci) assists in developing and demonstrating
new and improved methodologies that will meet these needs both efficiently
and economically.
This report, entitled "Treatment of Effluent Waters from Vegetable
Oil Refining." presents the development of a chemical pretreatment method.
Waste management personnel can utilize the results to reduce the BOD load
up to 71%. The Industrial Pollution Control Division and the project officer
can be contacted for further information on the subject.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
iii
-------
ABSTRACT
With overloaded municipal facilities and increasingly restrictive
Federal water quality regulations, it has become necessary to restrict
industrial wasteloads. Vegetable oil refineries generate a high BOD5
loading along with a high fat loading, which makes treatment difficult.
Original loadings with only surface skimming as treatment generated 0.0135
kg BOD5 per kg oil processed. After the grant period, the load was 0.0019
kg per kg processed for a reduction of 86%.
Major wastewater sources and processes are defined and described.
Refinery washwater from the oil water wash centrifuges was found to be the
major contributor of BODs and fat, with barometric cooling tower blowdown
and water from acidulation contributing smaller amounts to the load.
A calcium chloride pretreatment for the washwater which reduced the
overall BODs load by 71% was developed. This was done by the addition of
0.25% by weight calcium chloride to the washwater and pH adjustment to
between 5 and 6. Further research was done to devise methods to further
reduce the wastewater loadings to within municipal limits. Dissolved air
flotation was tried without success. A consultant was retained to inves-
tigate further chemical preatment. These studies along with the recom-
mended process for ferric chloride and lime treatment are presented.
It was felt by the grantee that the wastewater loads could be reduced
to within the municipal limits by making some process modifications, in-
creasing production supervision, and installing a clarifier to control
process upsets. This program resulted in a further reduction of 50%
and brought the waste strength to within municipal limits.
Cost for the system installed is $75,000 per year. Revenues from
recovered by-products range from $20,000 per year to $180,000 per year,
depending on the price of recovered material.
This report was submitted in fulfillment of Grant Number 12060 FDK
by the Archer Daniels Midland Company, Decatur, Illinois, under the partial
sponsorship of the U.S. Environmental Protection Agency. This report
covers a period from March 1970 to December 1973.
iv
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables viii
Acknowledgment x
Introduction 1
Conclusions 2
Reconraendations 4
Process Description 5
Wastewater Sources, Analysis, and Quantities 11
Primary Chemical Treatment 16
Treatability Studies and Design Basis 20
Description and Evaluation of Installed Wastewater .... 48
Management Facilities
Appendices 68
Bibliography 77
-------
FIGURES
Number Page
1 Soybean preparation and oil extraction flow
diagram (typical Lincoln & Decatur) 6
2 Oil refining flow diagram 7
3 Deodorization flow diagram 10
4 BOD5 and COD correlation for total West plant
effluent 12
5 Schematic of majonwastewater sources 13
6 CaClp water treatment flow sheet 19
7 Treatment flow diagram during water survey 23
8 West plant refinery skimmer effluent variability of
24 hr composite COD and flows 33
9 West plant refinery treated skimmer effluent
settling test results 37
10 Proposed treatment facility 39
11 Decatur wastewater treatment facility 49
12 West plant treatment flow diagram 51
13 Pounds BODg per pound oil processed 52
14 Pounds fat per pound oil processed 53
15 Clarifier pH vs. day ADM Decatur 60
16 Temperature vs. day ADM clarifier, Decatur,
Illinois 61
17 Clarifier COD vs. day ADM Decatur 62
18 Fat vs. day Decatur ADM clarifier 63
vi
-------
Number Page
A-l Rex wastewater clarifier 68
A-2 Walker skimmer 69
C-l Lincoln wastewater treatment facility 71
C-2 ADM Lincoln, Nebraska average mg/1 fat vs. month
January 1971-March 1972 74
C-3 ADM Lincoln, Nebraska suspended solids vs. month
January 1971-March 1972 75
C-4 ADM Lincoln, Nebraska average Ib COD/day vs. month
January 1971-March 1972 76
-------
TABLES
Number Page
1 Characteristics of Refinery Wastewater
Component Streams Grab Samples 14
2 Extraction Plant Loading .... 15
3 Laboratory Washwater COD Reduction by Calcium
Chloride Treatment and pH Adjustment 16
4 Wastewater BODs Before and After CaCl2 Treatment
Installation: East Plant Discharge to Municipal
Sewer 17
5 Laboratory COD Reduction with Gas Flotation 21
6 Chemical Flocculation Screening Test Results for
West Refinery Skimmer Effluent 24
7 Chemical Flocculation Results of Samples of West
Plant Refinery Skimmer Effluent 30
8 Laboratory Settling Test Results of Refinery
Skimmer Effluent 34
9 Summary of Investigative Results 35
10 Summary of Design Basis and Major Unit Sizes 40
11 Summary of Major Equipment Sizes 41
12 Summary of Construction Costs 45
13 Summary of Annual Estimated Costs 46
14 West Plant Effluent Analysis 54
15 Decatur West Plant BODs and Fat Per Pound
Oil Processed 56
16 Clarifier COD Reduction 58
viii
-------
Number Page
17 Decatur Clarifier Operation 64
18 Lagoon Operation COD (mg/1) 65
19 Installation and Operation Costs for West . . .
Refinery Waste Treatment System 66
20 Summary of Treatment Costs 66
21 Net Costs . . 67
C-l Analysis of Discharge to Municipal Sewer 72
IX
-------
ACKNOWLEDGMENTS
This study was conducted by the Archer Daniels Midland Company, Decatur,
Illinois. The initial treatment studies were conducted by James lelase under
the direction of R. S. White. In 1969, James lelase took over direction of
the study. The Roy F. Weston Company of Chicago, Illinois, was retained to
investigate secondary chemical treatment and to design a ferric chloride and
lime treatment system. Arnold Johnson, refinery project engineer, supervised
construction of the clarifier and treatment system. Donald Gill developed
the operation procedures and process modifications after the system was in-
stalled and prepared the final report. James Curry assisted with the analyt-
ical work throughout the grant period.
This report was submitted in fulfillment of Grant No. 12060 FDK. Clifford
Risley, Jr. was the project officer during the grant. David Rickles was the
assistant project officer during the research and construction phase. Ronald
Eng was the assistant project officer during the preparation of the final
report. He and James Scaief, research representative, Corvallis, Oregon,
provided assistance in the final report preparation. Special recognition
is made of the efforts of Jack L. Witherow of the Food and Wood Products
Branch, Corvallis Field Station, lERL-Cincinnati. Mr. George Rey, EPA staff
engineer, Washington Headquarters, provided technical assistance in the early
stages of the project.
-------
SECTION I
INTRODUCTION
The Archer Daniels Midland Co., Inc., processes oilseeds, primarily
soybeans. Oil is solvent-extracted from the seeds and refined at three
locations—two at Decatur, Illinois, and one at Lincoln, Nebraska. The
refining of vegetable oils generates several wastewater streams that have
caused the plants to exceed acceptable waste strength limits set by their
respective municipal authorities. The object of this project was to
analyze these waste streams and devise treatment methods by which their
and fat strength could be reduced to acceptable levels.
The initial problem was to identify the waste streams and to charac-
terize the wastewater. The second problem was to determine optimum treat-
ment for each source and finally to implement the necessary equipment and
procedures.
The initial investigation and design was done by ADM in conjunction
with a consulting firm. These investigations are described in Sections
V, VI, and VII.
Because of significant process modifications described in Section
VIII, it was determined that some of the process recommendations would not
be necessary. Provisions were made for their future installation if they
proved to be necessary.
A description of equipment installed and results obtained are con-
tained in Section VIII.
-------
SECTION II
CONCLUSIONS
The wastewater survey done after the calcium chloride system was
installed determined that for a soybean processing facility, the refinery
produces approximately 90% of the wasteload and the extraction plant con-
tributes only about 10%. Of the refinery wastewaters, the washwater is
the largest source of BODs and fat. For a facility processing 385,500 kg
(850,000 Ib) per day of oil, this stream contributes 450 to 22,700 kg
(1000 to 50,000 Ib) per day of BODs and fat in the form of water soluble
soaps. Greasy cooling water can contribute 340 kg (750 Ib) of fat and
BODs. Acid water contributes approximately 30 kg (70 Ib) of fat and
1800 kg (4000 Ib) of BODs per day. Floor washings and car washings can
contribute significant and highly variable loadings. Any of these sources
can contribute many times the normal loading under conditions where the
process is upset or uncontrolled.
Treatment of the washwater with 0.25% by weight calcium chloride
resulted in an overall reduction of 71% of the 8005, from 0.0135 to 0.0039
kg BODs per kg oil processed. Process modifications, increased process
control, and a clarifier resulted in a further reduction in total load of
50% of both the fat and COD. A greater reduction would have been obtained
had the plant not started to refine corn oil at approximately the same time
the clarifier was started. Installation of the clarifier in Lincoln, where
corn oil is not refined, resulted in a reduction of 65% of the COD and fat.
The overall reduction in Decatur for both the calcium chloride and clarifier
was 86%, from 0.0135 kg BODs per kg oil processed to 0.0019 kg BODs per kg
oil processed.
Laboratory studies indicated that secondary chemical treatment with
ferric chloride and lime would result in a reduction of 85% of the fat and
65% of the COD over the calcium chloride treatment. This would be an overall
reduction of 95% of the fat and 90% of the COD.
The calcium chloride treatment system was installed in 1970 at a cost
of $25,000. Treatment cost based on 10 years straight line depreciation is
$32,308 per year. The system removes approximately 1206 kg (2660 Ib)
per day fat and 3736 kg (8236 Ib) per day BODc. This is a cost of 2.34$
per kg (1.06
-------
mately 778 kg {1716 lb) per day BODs and 190 kg {420 Ib) per day fat.
This is a cost of 61.954 per kg {28.U per lb) fat and 15.154 per kg
(6.874 per lb) BOD5 removed.
The net cost for the treatment system is 14.774 per kg (6.7$ per lb)
fat removed, 4.564 per kg (2.074 per Ib) BODs removed, and 364 per 1000
liter ($1.36 per 1000 gal) water treated. This cost is offset by revenue
from recovered oil that can be sold. The price of this oil has varied over
the grant period from 8.84 to 39.7* per kg (44 to 184 per lb). This results
in a net treatment cost that has varied from 254 per 1000 liter (954 per
1000 gal) to a gain of-514 per 1000 liter ($1.93 per 1000 gal). On a
basis of oil processed, this is .0324 per kg (.0154 per lb) oil processed
to a gain of .0634 per kg (.0284 per lb) oil processed.
Installation of the ferric chloride and lime treatment system is
estimated to cost $100,300 per year compared to $43,000 for the system
installed, The estimated fat recovered would be approximately 324 kg
(714 lb) per day at a cost of $275 per day, or 84.94 per kg (38.54 per lb)
fat, which is 31% higher than the installed treatment.
-------
SECTION III
RECOMMENDATIONS
Further research should be done to develop uses for and modifications
of corn lecithin to reduce the wasteload from refining corn oil.
Additional study is also needed to determine the effect the ferric
chloride and lime treatment system would have on the reduced waste load
compared to the load at the time the secondary chemical treatment was done.
A further major reduction in 6005 would result in the development of a
good treatment system for the acid liquor. This was not an important consi-
deration during this grant period because after the development of the
calcium chloride treatment for wash water, the plant was well within the
BOD5 limit specified by the municipality.
-------
SECTION IV
PROCESS DESCRIPTION
This section describes a soybean extraction plant and refinery common
to Decatur and Lincoln. Additional corn germ extraction and refining in
Decatur is entirely analagous to the soybean process and will not be
described separately. Only those processes that generate significant
waste products will be described in detail. All processes operate 24 hours
per day and 7 days per week. The plant in Decatur has the capacity for
120,000 bushels per day soybean extraction, and 200 tons per day of corn
germ, and the capacity to refine 1.4 million Ib per day of vegetable oil.
Figures 1 and 2 describe extraction and oil refining respectively.
SOYBEAN PREPARATION AND EXTRACTION
The beans are initially cleaned by mechanical screening. They are then
cracked and cleaned by aspiration. This is followed by heat conditioning
in a steam tube drier. The cracks are flaked by flaking rolls and conveyed
to an extractor. The oil is removed by continuous countercurrent hexane
extraction. The flakes are then conveyed to the desolventizer toaster
where the volatile solvent is driven off by jacket and sparge steam.
The vapors are condensed and the liquor pumped to a gravity separator.
The water insoluble hexane is sent to solvent storage and the water dis-
charged to the sewer. The desolventized and toaster flakes are cooled,
blended and loaded. The oil-hexane mixture is pumped to distillation
where the solvent is evaporated off and condensed. The oil is then pumped
to storage. The condensate water from this process contains some entrained
oil and soluble protein but is not a major source of BODc or fat (See
Section V).
SOYBEAN OIL REFINING
The refining of vegetable oil may employ a combination of several
processes, depending on the product desired. These processes include
degumming, caustic refining, subsequent washing, vacuum drying, bleaching,
hardening, post bleaching, filtration and deodorization.
Degumming
Approximately 2% water is added to the oil to hydrate the gums (phos-
phatides). On hydration these gums become oil insoluble and are separated
from the oil in a centrifuge. The gums are then dried to produce lecithin.
This step can be bypassed; however, it decreases the amount of soap-
stock produced in caustic refining and in the case of soybean oil, produces
a valuable by-product.
5
-------
SOYBEANS
i
STEAM
CLEANER
TRASH
CRACKING
ROLLS
SEPARATOR
1
MEATS
\
STEAMER
HULLS
FLAKES ,
\_
FLAKING
ROLLS
SOLVENT
STEAM
EXTRACTOR
SPENT
FLAKES
—MISCELLA
DESOLVENTIZER
TOASTER
DUST
COOLER
GRINDING &
BLENDING
AIR.
—VAPORS
WATER
SOLVENT VAPOR:
SOLVENT
EVAPORATORS
SOLVENT
CONDENSER
SOLVENT
SOLVENT
STORAGE
ENTRAPMENT
SEPARATOR
EXTRACTOR
WASTEWATER
SOLVENT 8 WATER
EXTRACTED CRUDE OIL %
(TO STORAGE OR REFINING)
Figure!. Soybean preparation and oil extraction flow diagram (typical Lincoln and Decatur).
-------
CRUDE fc
OIL
VYMI C.T\
DEGUMMING
LECITHIN
CAUSTIC SOLUTION '
(Na
DEGUMMED fc
OIL 1
TO
LOADINI
OH) 1
CONTINUOUS
REFINING
r%
ACID TO OIL
STORAGE
SOAPSTOCK ^
ACIDULATION
ACID-
kl ^ WATER
NaOH
NEUTRA
^* ATION
WATER
1 ,
SOAPSTC
WASHIN
irK
/VI\
G
r
LIZ- TO WATER
TREATMENT
UDADING
CLAY
^
^ VACUUM
DRIER UL
.EACHING
SPENT CLAY
(TO LANDFILL)
CLAY
1
1 _ POST
BLEACHING
CATALYST AND
HYDROGEN J
l
H
""•(HYC
AR
)R(
DENING
X3ENATION)
UNHARDENED OIL
(TO LOAI
FILTRATION
DING)
k HARC
STEAM
*
DEODORIZER
FILTRATION
SOAPSTOCK
WASH WATER
FILTER CAKE
(TO LANDFILL) *
)ENED OIL 1
FILTRATION
FINISHED
OILS
(TO STORAGE)
SPENT CLAY
(TO LANDFILL)
FILTER CAKE
DEODORIZED
DISTILLATE
FILTER CAKE
(TO LANDFILL)
Figure 2. Oil refining flow diagram.
-------
Caustic Refining
Soybean oil is primarily triglyceride with approximately 0.5% to 1.0%
free fatty acids. The object of caustic refining is to convert the free
fatty acids to water soluble soaps and then to remove them by centrifugation.
A 12% sodium hydroxide water solution is mixed with the oil. The water soap
solution and oil is then separated in a centrifuge. The chemical equation
for this reaction is:
0 0
R-C^ + NaOH R-Qf + H.O
N)H VNa+
The water soap solution (called soapstock) is pumped to acidulation
where sulfuric acid is added on a batch basis to convert the soaps to fatty
acids. The fatty acids are insoluble in water and float to the surface.
The mineral salts and any water soluble material remains in the water layer.
The water layer is neutralized and discharged to the waste treatment
facilities. The chemical equation for this reaction is:
2 R-C( + H-SO, 2 R-( -i- NaSO.
VNa+
Water Wash
The oil from caustic refining contains trace amounts of soaps and
sodium hydroxide. Condensate water is added to the oil and agitated to
remove this residual material. The oil and water from this process is a
weak (\% to 3%) soap solution with a relatively high pH.
Vacuum Drying
The oil from water washing contains trace amounts of water that can
blind the filters in subsequent filtration operations. In vacuum drying
the oil is heated under a vacuum of 28" Hg to remove this trace moisture.
There is no significant waste stream from this process.
Bleaching
Trace amounts of color bodies are removed by absorption onto activated
bleaching clay. The clay and oil are slurried and heated under vacuum.
The clay is then removed in pressure filters and disposed of in the muni-
cipal landfill.
Hardening
Hardening is employed to raise the melting point of the oil for use in
shortening or margarine manufacture. This is done by the addition of hydrogen
to unsaturated oil molecules over a nickel catalyst. The catalyst is
8
-------
blended with the oil and heated. The hydrogen gas is then bubbled through
the mixture. The degree of hardening is determined by the temperature,
pressure and quantity of hydrogen gas added to the reaction. The catalyst
is filtered out of the oil after the reaction is complete. No significant
wastewater is generated by hydrogenation.
Deodorization
The purpose of deodorization is to remove volatile impurities that might
cause flavors and odors. This is done by steam distillation at temperatures
of 425°F to 475°F and 3 mm Hg absolute pressure. The high temperatures
are obtained with a dowtherm boiler heating system. The vapors are pulled
off through the vacuum system. (See Figure 3) A recycle deodorizer dis-
tillate scrubber removes the bulk of the organic material from the vapor
(99%+). Excess distillate is pumped to storage and sold as a valuable
by-product. Deodorizer distillate is a valuable source of a-tocopherol
(Vitamin E). The remaining vapors, primarily steam, are condensed in the
water scrubber along with any remaining organic material and drain to the
greasy cooling tower hot well.
-------
OIL IN
-Lr_J
I- I
COOLER
SCRUBBEF
JWATER
TO DEODORIZER
DISTILLATE TANK
OIL OUT
SPARGE STEAM
-£
GREASY
COOLING
TOWER
BAROMETRIC
CONDENSER
V
TO HOT1
WELL
Figure 3. Deodorization flow diagram.
-------
SECTION V
WASTEWATER SOURCES, ANALYSIS, AND QUANTITIES
All analyses except COD were conducted according to the test procedures
given in Standard Methods for Analysis o;F Water and Wastewater, 12th Edition.
The test procedure for COD is given in Appendix C. This modified procedure
was used because of the shorter reflux time and because it is more closely
related to BOD5 than the standard method. This modified COD will be denoted
by quotation marks. Figure 4 compares the COD analysis to the BODs analysis.
Flows and wastewater analysis for each wastewater source are presented
in Figure 5.
The condensate water is generated in heating coils and jacket steam
throughout the refinery. It is generally hot, clean water and most of it
is being recycled as washwater in the refining step.
Washwater generated in the oil washing step is high in BOD and fat
primarily because of dissolved soaps. The stream will contribute 1000
to 5000 Ib per day of fat or BOD5 to the sewer load. If the oil is not de-
gummed before caustic refining, some water soluble phosphatides will be
removed in this step and will cause higher loads. This is generally the
case when corn oil is being processed.
Floor washings are highly variable for obvious reasons. It is charac-
terized by high temperature and high pH because of the washing equipment
and soaps.
Greasy cooling water can be a major source of BODs depending on the
operation of the deodorizer scrubber and atmospheric conditions. (See
deodorization description Section IV) Any organic material that passes
through the scrubber becomes emulsified in the greasy water in the vacuum
system. The greasy cooling tower dissipates the heat of condensation of
the steam primarily by evaporation of the greasy water. In cold weather
the water loses more heat by convection which decreases the amount of
evaporation necessary. When the amount of water entering the tower via
steam condensation is greater than the wind drift plus evaporation, it
becomes necessary to blow down the tower to the sewer. The organic material
in the water is primarily short chain free fatty acids. Typical COD values
range up to 3000 mg per liter. For an overflow of 30,000 gal per day,
750 Ib of COD would be discharged to the sewer based on 3000 mg per liter
of fat.
11
-------
""
BODg » 354- 1 .01 COD"
95% CONFIDENCE LIMITS
5000 —
1000 2000 3000
4000
"COD" mg/l
Figure 4. BOD and COD correlation for total West plant effluent.
12
-------
REFINERY
CO
134,000-183,000 GPD
SKIMMER
CLARIFIER
0x5: rn
r~ CD
H
OD m
O2
~O rn
3 O
*s
CO
m
CO 6
/: 20,000 -30,000 GPD,
FEWATER CHARACTERS
i
CO
X
m
HO
%5
r x
is
DD H
O rn
P s
o
X
ro
o
0
o
TJ
^
CO
-n
k
L 0-100,000 GPD, -CO
'EWATER CHARACTERIS
CONDENSATE WATER
oH
CO
1
PLANT
CI2 TREATMENT
PROCESS SEWER
_l
EXISTING SAMPLING AND
RECORDING STATION
Figure 5. Schematic of major wastewater sources,
-------
Tank car washings vary from one to four cars per day. The effluent is
characterized by high temperature, 160°F, and a variable amount of emul-
sified oil. The BODs can be anywhere from 2000 to 60,000 mg per liter.
(Table 1)
The neutralized acidulation water flow and COD is highly variable de-
pending upon the amount of flush water being used on the centrifuges and
whether or not the oil is being degummed prior to the refining. The
phosphatides which would be removed in degumming greatly increase the
strength of the acidulation water due to their water solubility. Fresh
water can increase the moisture content of the soap from 50% to 95%.
While this does not affect the total BOD5 load, it does reduce the residence
time in treatment facilities and reduces the processing time which is ne-
cessary to get good oil separation. While the BODs is very high (25,000
mg per liter) for this stream, the fat is relatively low (300 - 500 mg
per liter). This is the only waste stream from the refinery that is not
composed primarily of hexane soluble material. A typical day would yield
loadings of 4000 Ib per day BOD and 70 Ib of fat based on 20,000 gal at
a 25,000 mg per liter BODs and 400 mg per liter fat.
The extraction plant wastewater consists of condensate described in
the process section. At 150,000 gal per day and 300 mg per liter BOD,
this stream contributes only 375 Ib per day of BOD. The temperature is
high (180°F) and the pH is near neutral (6-8). Flows were estimated
from process design requirements. (Table 2)
TABLE 1. CHARACTERISTICS OF REFINERY WASTEWATER COMPONENT STREAMS GRAB
SAMPLES
Source
Acidulation
Tank
Acidulation
Tank
Acidulation
Tank (Settled
Sample)
Tank Car
Washing (Crude
Oil)
Tank Car
Washing
(Hydrogenated)
Sampling
Date Time
9-09-69
9-11-69
4 p.m.
9-11-69 9:30 a.m.
Wastewater Characteristics
COD (mg/1) pJi Temp. (°F)
26650 5.6 1150-160
29220 4.6 150-160
24760 5.0 150-160
2-6 p.m.
9-11-69 (Composite) 60400 6.0 150-160
9-11-69 (Composite)
4655 6.1 150-160
-------
TABLE 1. CHARACTERISTICS OF REFINERY WASTEWATER COMPONENT STREAMS GRAB
SAMPLES (Continued)
Source
Sampling
Date Time
Wastewater Characteristics
COD (mg/1) pj]_ Temp. (°F)
Tank Car
Washing
(Hydrogenated) 9-16-69 (Composite)
2100
6.7 150-160
TABLE 2.
EXTRACTION
PLANT LOADING
Sample
Date
Extraction Plant
9-24-69
9-26-69
9-30-69
10-01-69
10-02-69
Average
BOD,
(mg/T)
325
496
370
360
370
384
Estimated Flow
(GPD)
150,000
150,000
150,000
150,000
150,000
150,000
15
-------
SECTION VI
PRIMARY CHEMICAL TREATMENT
The initial objective was to determine by laboratory experimentation
whether a chemical pretreatment of any waste water stream was feasible.
Several chemicals, including polymers, polyelectrolytes, sodium chloride and
calcium chloride were used. Calcium chloride was found to be the most
economical and efficient coagulant at dosages of approximately 0.25% by
weight. The equation for this reaction is:
2 R-Qf + Ca++ 2 Na+ + (R-G( ) Ca++
N0-Na+ 0-2
The normal pH of the washwater was in the range of 10.5 to 12.5 and
while some separation was accomplished at this pH, a more rapid and effective
coagulation and separation occurred at a lower pH. After the addition of
the calcium chloride to the washwater, an adjustment to the 5 - 6 pH range
with sulfuric acid greatly enhanced the speed of flotation and made the
floe more dense. Table 3 shows COD reductions using this chemical treatment
on the washwater. The average reduction in COD was 91.7%.
TABLE 3. LABORATORY WASHWATER COD REDUCTION BY CALCIUM CHLORIDE TREATMENT*
AND pH ADJUSTMENT
COD Before Treatment COD After Treatment
Test (mg/1) (mg/1)
1 25142 2280
2 26560 3750
3 39000 3176
4 51000 3276
5 29366 2061
6 44662 4008
7 17766 934
Average 33356 2783
*Calcium chloride added was 0.25%.
+Average reduction was 91.7%.
16
-------
Because of the success of the laboratory investigations, primary treat-
ment units were installed at all ADM refineries. Figure 6 shows the plant
layout used for this system. The washwater from the water wash centrifuges
is discharged into the 5500 gal tank on a continuous basis. Calcium chloride
is added by a chemical feed pump and a roto-meter was installed to insure
proper proportioning. The residence time in the 5500 gal tank is dependent
upon the rate of refining and the amount of floor washings being discharged
to the system. Because of this, some of the floe is usually recovered in
sedimentation tanks later in the system. Acid is added to the second tank
to increase the density of the floe and to increase the rate of the rising
flocculated soaps and oils. The effluent from this tank is pumped to an
oil skimmer located outside the refinery. The recovered soaps and oils
are pumped to soapstock storage and acidulated.
Table 4 shows the average Decatur East BODs figures for water discharged
to the municipal sewer for May 1968 before the calcium chloride treatment
was installed and May 1970 after the calcium chloride treatment was installed.
No acidulation is done at Decatur East so the 93.3% overall reduction is
indicative that the process is as effective as expected from laboratory
studies on washwater only. Overall BODr was reduced at the West Plant by
71%, the difference being due to acidulation. (See Table 14, Section VIII)
Part of the original grant agreement was to investigate a washwater
recycle process with financing from the United States Department of Agricul-
ture. The washwater would have been recycled through an ion-exchange unit
to convert the soaps to free fatty acids. The grantee submitted a proposal
for $45,000 for a pilot unit which was rejected. Development of the calcium
chloride treatment had minimized the importance of the ion-exchange process
so the project was dropped.
TABLE 4. WASTEWATER BOD5 BEFORE AND AFTER CaCl2 TREATMENT INSTALLATION:*
EAST PLANT DISCHARGE TO MUNICIPAL SEWER
BOD5, May 1968, BODs, May 1970,
Before CaClz After CaClz
Test (mg/1) (mg/1)
1 1340 335
2 1620 70
3 1270 30
4 1730 170
5 2710 380
6 2650 100
7 2720 245
8 2350 125
9 1595 70
10 2250 50
17
-------
TABLE 4. WASTEWATER BODs BEFORE AND AFTER CaCl2 TREATMENT INSTALLATION:*
EAST PLANT DISCHARGE TO MUNICIPAL SEWER (Continued)
Test
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Average
BOD5, May 1968,
Before CaCl2
(ma/1 )
2055
2710
5200
3620
2100
2100
2100
2100
2100
2100
2100
2100
2100
2070
1900
2100
3610
3300
2590
2360
2460
2358
BOD5, May 1970,
After CaCl2
(mg/1 )
155
680
545
85
155
90
80
100
25
55
25
100
85
105
55
40
25
85
90
50
620
156
*Reduction in 6005 was 93.3%.
18
-------
, CaCI2
.1600
GAL.
LARGE LINE 6
OIL- CaCI2
5500 GAL.
•WASHWATER
WATER
OIL- CaCI,
H2S04
1 ...
s.
1
> f
2000 GAL.
TO SOAPSTOCK _
TO OUTSIDE
TREATMENT
TO SOAPSTOCK
Figure 6. CaCl2 water treatment flow sheet.
-------
SECTION VII
TREATABILITY STUDIES AND DESIGN BASIS
While chemical treatment of the washwater from the refineries resulted
in an overall reduction in the total waste stream from approximately
50,000 P.E. to 20,000 P.E., there was still a problem resulting from other
wastewaters in the refinery complex which did not respond as readily to
the calcium chloride treatment. Further studies were needed to reduce the
organic load to within acceptable limits.
A consulting firm was retained to assist in the investigation of
secondary chemical treatment and to design a treatment process based on these
investigations. At the same time, the grantee was to investigate the pos-
sibility of treatment with air flotation and process modifications to reduce
the waste load.
GAS FLOTATION STUDIES
Laboratory studies indicated that finely dispersed gas bubbles produced
by electrolysis will rapidly remove most suspended solids from a waste stream
and materially reduce the BOD.
Two grids consisting of an anode and cathode were immersed in the bottom
of a chamber and low voltage current was passed between them. Electrical
current was supplied to the grids with a 20 ampere rectifier. A direct
current of approximately 2-5 volts and 4-8 amperes was introduced to the
grids which were one half square foot in area. Hydrogen was generated at
one grid and oxygen at the other in the form of very fine bubbles evenly
distributed over the area. As the bubbles rose, they attached themselves
to the suspended solids and entrained oils and floated them to the top.
With this procedure, settleable solids as well as suspended solids
were floated to the surface. Satisfactory results were usually obtained within
fifteen minutes. (Table 5) The data obtained from these experiments showed
a reduction of organic loading of approximately 50%, from 7021 to 3231
mg per liter, beyond the results obtained with the primary treatment which
consists of the already outlined calcium chloride system.
20
-------
TABLE 5. LABORATORY COD REDUCTION WITH GAS FLOTATION
Material
Greasy Water
Wash water
Skimmer Effluent
Water to Sanitary
District
Test
1
1
1
2
3
1
2
Before COD
(rag/1)
3005
29,019
1,606
2,565
2,850
1,140
1,238
After COD
(mg/1 )
248
111
331
1,250
1,650
331
228
An electrolysis unit was then installed in the refinery skimmer. One
hundred sixty square feet of 1/2 inch steel floor grate was used for the
anode and cathode. The grids were placed in the bottom of the skimmer
on bricks approximately 3 inches apart. Power was supplied by a 1500
ampere Clinton rectifier model number R1512HO-S. Results seemed to be
promising; however, after a few weeks of operation, the production of gas
bubbles all but stopped. Upon emplying the skimmer and inspecting the grids,
the cathode was in excellent condition and showed no corrosion; however, the
anode was almost completely eroded.
Laboratory studies were resumed, and it was found that on prolonged
usage, the anode was indeed eroded while the cathode remained in good condi-
tion. Studies were then undertaken to find a suitable anode material to
withstand this eroding effect.
Anodes of iron, aluminum, titanium, 316 stainless steel, Inconel,
cast iron, monel, nickel and high silica iron were tried without success.
A combination of 316 stainless steel mesh with graphite supports for the
anode gave the best laboratory results.
A plant scale installation was made with grids consisting of 18 mesh
stainless steel with graphite cross bars. After a period of two weeks,
the skimmer was drained and the grids inspected. The anode consisting of
stainless steal and graphite was badly deteriorated. The stainless steel
mesh was almost completely gone and the graphite had decreased in size
considerably. The efficiency had deteriorated over the two week period
to a point where few bubbles were observed coming to the surface of the water,
During the first few days of operation, the process worked very well
producing a diffuse bubble blanket.
21
-------
Another consultant was then employed to develop a workable anode. This
firm produces material used specifically for this application and they felt
their material would stand up under the conditions of the process. The material
consisted of graphite bars 48 inches long, 6 inches wide and 1 inch thick
coated with lead dioxide. Eight bars were purchased and installed in the
Imhoff tank. Copper buss bars coated with glyptol, an insulating material,
to prevent corrosion were used to connect the graphite bars. The coating
did not withstand the corrosion and in two days was almost completely destroyed.
Graphite connecting bars were then tried and proved to be more successful.
After 1 week of operation, the graphite anodes were badly eroded and began
breaking under their own weight.
Advances made in other methods of waste control had minimized the impor-
tance of the development of this system. At this point, the electrolysis system
was abandoned as being impractical.
SECONDARY CHEMICAL TREATMENT
Studies were undertaken by the consultant in September 1969 to do the
following:
1. Establish the flows and characteristics of the West Plant Refinery,
Extraction and Subsurface drainage streams.
2. Do chemical/physical treatability studies on the refinery skimmer effluent.
3. Develop process design parameters from the above treatability studies.
4. Propose a treatment facility capable of treating these wastewaters.
The treatment requirements were set by the Decatur Sanitary District and
consist of the following:
a) For 95% of the time the total Biochemical Oxygen Demand of waste from
all industry facilities, in the aggregate, shall not exceed 5780
Ib per day (34,000 P.E.) as any 10-day average nor exceed 7,200 Ib
per day (42,500 P.E.) as any 3-day average.
b) The total fats-oils-grease of the waste from the three plants shall
not exceed 1000 Ib per day as a monthly average, shall not exceed 1,00
Ib per day for more than 10% of the samples, and shall be less than
2,000 Ib per day for 97% of the samples. The sanitary district reserves
the right to re-examine these criteria if, after they are met, phy-
sical and treatment problems resulting from the grease content of the
waste continue.
Figure 7 is a location plan of the sampling points used in the study.
Grab samples, representative of skimmer effluent, were collected at the lagoon
inlet (sampling point No. 1) at 2-hour intervals over a period of 3 days
from September 9 to September 12, 1969.
22
-------
SKIM TANK
O
PUMP
STATION
\
TO DSD
\ FROM REFINERY
\
DULATION
FACILITIES
®
LAGOONS (TVIO IN SERIES)
..J
UNDERDRAIN-
OL SEPARATOR
TO
HOMEWOOD LAKE
Figure 1. Treatment flow diagram during water survey.
23
-------
TABLE 6. CHEMICAL FLOCCULATION SCREENING TEST RESULTS FOR WEST REFINERY SKIMMER EFFLUENT
Raw Sample,
Collection
Date and
Time
Type
of
Sample
Characteristics
Treatment
COD
mg/1
PH
Temp
OF
Coagulants Coagulant Aid
Type Dosage, rnq/1 Type Dosage, rng/1
Qualitative Observations and Conclusions from Jar
Test Results
9-9-69
11:30 a.m.
Grab
6,400 6.2
155
FeCl3 100 to 1000
ro
Ca(OH),
CaCl,
200 to 1000
200 to 2000
Fe2(S04)3 100 to 1000
FeCU dosages below 1000 mg/1 did not s^eein to have
any significant effect on breaking the emulsion.
The supernatent with Fed, dosages at less than
1000 mg/1 after centrifugation still was very
turbid. 1000 mg/1 dosage gave food floes and
relatively clear supernatent. The Fed, treat-
ment at a dosage of 1000 mg/1 was judgea as
effective.
Floe formation was poor at 200 mg/1 dosage.
Supernatent was turbid after centrifugation.
However, 400 mg/1 to 1000 mg/1 dosage gave good
floes and clear supernatent after centrifugation
floe formation was almost instantaneous. The
pH at these dosages valued from 7 to 8. The
Ca(OH)2 treatment above 400 mg/1 was judged as
effective.
There was no apparent flocculation. No change in
clarity after centrifugation, no sludge formed at
the bottom. CaClp treatment at all dosages was
judged as ineffective.
No flocculation was observed up to 5000 mg/1
dosage. Supernatent was very turbid with very
slight sludge at the bottom after centrifugation.
pH was 5.9 at 100 rag/1 dosage and 5.4 at 500
mg/1 dosage. 1000 mg/1 dosage gave good
flocculation with clear supernatent after
centrifugation. pH was 5.0 Fe-fSOJ, treatment
at a dosage of 1000 mg/1 was jadgea as effective.
-------
TABLE 6. CHEMICAL FLOCCULATION SCREENING TEST RESULTS FOR WEST REFINERY SKIMMER EFFLUENT (Continued)
Raw Sample
Collection Type
Date and of .
Time Sample
—
Characteristics
COD
mg/1
Treatment
Temp Coagulants
PH
°F Type
A12(S04}3
Dosage, mg/1
100 to 1000
Coagulant Aid
Type
Dosaqe, mq/1
--
Qualitative Observations and Conclusions from Jar
Test Results
Flocculation was very poor up to 500 mg/1
dosage.
ro
9-9-69
3:30 p.m.
Grab
19,000 7.5
148
Fed-
200 to 1500
Ca(OH),
100 to 2000
supernatent clarity. However, the results were
still poor. The pH was 5.8 at 100 mg/1 and 5.3
at 1000 mg/1 dosage. AlJSO.k treatment was
judged as ineffective.
At 200 mg/1 dosage there was no noticeable effect.
At 300 mg/1 dosage the floes formed after 10 mins.
Some floated at the top, some settled down. The
sample was still very turbid. Centrifugation did
not improve the clarity. 500 to 1500 mg/1 dosage
gave large amount of floes through out the depth.
Centrifugation gave some sludge at bottom, some
scum at the top and still large amounts of floes
throughout the depth, indicating that the mass
density of the floes was same as that of the
wastewater. pH at a dosage of 500 mg/1 was 6.6
and that at 1500 mg/1 was 40. The FeCK treat-
ment on this sample was judged as ineffective.
Up to 500 mg/1 dosage there was no significant
change in the sample appearance. At dosages of
1000 mg/1 and above there was almost instant
floe formation. However, the supernatent was
still very turbid even after Centrifugation.
The pH at 500 mg/1 dosage was 8 and those at 1000,
1500, and 2000 mg/1 were 10.3, 11.1, and 11.5
respectively. The Ca(OH),., treatment was judged
as ineffective.
-------
TABLE 6. CHEMICAL FLOCCULATION SCREENING TEST RESULTS FOR WEST REFINERY SKIMMER EFFLUENT (Continued).
Raw Sample
Collection
Date and
Time
Type
of
Sampl e
Characteristics
COO
mq/1
PH
Temp.
°F
Treatment
Coagulants
type
Dosaqe,
mq/1
Coagulant Aid
Type
Dosage,
Qualitative Observations and Conclusions from Jar
mq/1
Test Results .
CaCl,
CT>
Fed.
100 to 2000
Fe2(S04)3 200 to 1500
y\i2(so4)3 200 to 1000
500
500
500
500
DOW
N-ll
DOW
N-11
DOW
N-ll
DOM
N-ll
At 200 mg/1 dosage there was no apparent change.
At 500 mg/1 and above there was floe formation.
However, the sample was still very turbid after
centrifugation. In this particular the floes
floated on the top. The pH at 200 mg/1 dosage was
6.9 and that at 1500 mg/1 dosage was 6.2. The
CaClp treatment was judged as ineffective.
Up to 300 mg/1 dosage there were no visible floes.
The mg/1 dosage formed very small floes after a few
minutes. 1000 mg/1 and layer dosages gave instant
floes. However, the sample was still very turbid.
Some floes floated to the top, some settled, and
some still remained in suspension throughout the
depth after centrifugation. The pH at 200 mg/1
dosage was 7.2 and gradually decreased to 5.7 at
dosage of 1500 mg/1. The Fe,(SO/!)0 treatment was
judged ineffective.
J4'3
Similar results as above except that the pH at 1000
mg/1 dosage was 6.5. A12(SO«), treatment was
also judged ineffective.
Using 4 of these polyelectrolytes along with 500
mg/1 Fed, gave large floes. However, the results
were judged as no better than Fed, treatment
alone. This treatment was also judged as in-
effective.
-------
TABLE 6. CHEMICAL FLOCCULATION SCREENING TEST RESULTS FOR WEST REFINERY SKIMMER EFFLUENT (Continued)
Raw Sample
Collection
Date and
Time
Type
of
Sample
Characteristics
COD
mg/1
Temp.
pH °F
Treatment
Coagulants
Type Dosage, mg/1
Coagulant Aid
Type Dosage, mg/
Qualitative Observations and Conclusions
1 Test Results
from Jar
9-10-69
Composite 3,900
of 2 hr.
grab sample
over 16 hr.
period.
6.2
ro
Cad,
Ca(OHf5
Ai2(so4}3
Fed.
Fed.
500 to 1500
500 to 1000
500 to 1500
500 to 1000
500
500
500
500
DOM
C-31
DOW
A-23
DOW
N-ll
DOW
There was apparent flocculation. The floes
settled; however, the supernatant was still
turbid after centrifugation indicating that
there was no destabilization of the emulsions.
The treatment with these chemicals was judged
ineffective.
At these dosages floes were observed; however,
the supernatent was still turbid after centri-
fugation. pH at 500 mg/1 dosage was 5.2 and that
at 1000 mg/1 was 4.6. This treatment was judged
ineffective.
Some floes formed. No clear supernatent after
settling.
Some floes formed. No clear supernatent after
settling.
Relatively clear supernatent. However, still lot
of turbidity after settling.
Considered as less effective than DOW N-ll; however,
better than C-31 and A-23. The combination of 500
mg/1 Fed, and 5 mg/1 of DOW N-ll marked the best
in this series. COD was reduced from 3900 mg/1 to
1300 mg/1. This treatment was judged as effective
on this composite sample.
-------
Flows were obtained by timing the total duration of pumping over a given
time interval. Due to some difficulties in recording the flows, it was not
always possible to determine the flow rates and the total flows corresponding
to the time or period of sampling. Since the effectiveness of the Imhoff tank
located in between the skimmer outlet and the lagoon inlet was considered
negligible, no sampling was done to define its performance.
Other grab samples from the acidulation facilities, lagoon effluent and
subsurface drainage were collected from sample points 3, 4, and 5, respectively.
Flows for these discharges were estimated from process design requirements.
Results of this survey in conjunction with data generated by ADM is presented
in Section V and will not be.presented here.
The wastewater survey revealed that BOD (as approximated by COD) and oils
was the major problem in the effluent from the West Plant. Further, comparisons
of the three major sources of wastewaters revealed that the refinery wastewaters
contribute over 80% of the organic load. Consequently, emphasis was placed
on the chemical/physical treatment of the refinery wastewaters for BOD5 and
oil reduction.
Since results of the analysis of lagoon effluent samples showed little,
if any, equalization of other significant or consistent improvement in quality
over the skimmer effluent samples, major emphasis was placed on the treatment
of the refinery skimmer effluent wastewaters. Grab and 24-hour composite
skimmer effluent samples obtained during the problem definition survey were
evaluated to determine the most effective and economical process for reduction
of BODs and oils. Grab versus composite samples were tested to determine the
necessity and advantages of equalization prior to further treatment. Laboratory
tests included pH adjustment, activated carbon absorption, inorganic and or-
ganic chemical coagulation for emulsion destabilization, and gravity settling
and air flotation for separation of chemical floes and oils.
Coagulation screening tests were performed on grab as well as composite
samples of skimmer effluent for refinery wastewater. These studies were
qualitative in nature and were conducted to determine the types of coagulants
best suited for the wastewater and to investigate the need for an equalization
basin.
Grab Samples
Fifty-mi Hi liter samples were used for these summary tests. Different
coagulants with and without different polyelectrolytes were screened. After
coagulant(s) addition, the samples were slowly stirred and observed for floe
formation and supernatant clarity. The coagulant which did not form floes in
30 minutes and/or did not produce a clear supernatant were not evaluated further.
The results of these studies are presented in Table 6. Although not shown
in Table 6, the results of varying wastewater pH with acid and lime did not
produce a significant reduction in COD or improvement in wastewater clarity.
Also tried and found not effective was the use of activated carbon—tried
along, at various adjusted wastewater pH's, and in combination with several
coagulants. No benefit was observed to justify the increased cost of this
chemical absorbent.
28
-------
It may be noted that the coagulants that worked on one grab sample
(9/9/69 - 11:30 a.m.) did not work at all on the other (9/9/69 - 3:30 p.m.).
Also, it may be noted that in some cases the floes settled, and in the others
they floated. These results indicate that for unequalized samples one single
approach or solution may not consistently work on the skimmer effluent waste-
water .
Composite Samples
At this point the same coagulants were tried on the composite samples of
the skimmer effluent. The results (presented in Table 6) were very encouraging.
The combination of ferric chloride (FeCIs) and calcium hydroxide (Ca(OH)2)
seemed to produce the best and most consistent results. This treatment combina-
tion was subsequently quantitatively investigated for COD and oil removal.
Quantitative studies on coagulation were performed on seven 24-hour compo-
site samples. These samples were obtained by continuous collection of the
wastewater from the discharge side of the skimmer inlet pump. The composite
samples were approximately at room temperature due to a loss of heat on standing,
whereas the temperature of the wastewater going to the skimmer varied over a
range of 110° to 155°F. In order to simulate composite skimmer effluent,
the samples were mixed well and allowed to stand for 30 minutes and the top
oil layer was skimmed off. The subnatant was used for the studies.
The results of the composite sample coagulation studies are presented in
Table 7. Quantitative performance was judged on the basis of COD alone with
simultaneous qualitative visual inspection of floe settleability and super-
natant clarity. The results indicated that the combination of 100 mg per
liter FeCl3 plus 300 to 500 mg per liter Ca(OH)2 generally gave the best
results. Increasing the FeCls dosage to 200 mg per liter did not give any
better results. In fact, in many cases, the floe settling rate was observed
to be slower.
The variability of the 24-hour composite treated skimmer effluent COD's
are presented in Figure 8. These results show that for the present refinery
skimmer effluent wastewater characteristics, FeCl3 and Ca(OH)2 coagulation
followed by sedimentation would reduce COD's to equal to or less than 850
mg per liter for 50% of the samples, and equal to or less than 1,500 mg per
liter for 90% of the samples. This represents approximately 65% COD reduction
on the average.
At the time of these investigations no specific limits for fats, oils
and greases (fat) were imposed by the municipality on the effluent. Hence,
in order to avoid an excessive analytical work load on the personnel, only
four samples of raw and treated skimmer effluent were analyzed for fat.
The results indicated that the skimmer effluemt FOG can be substantially reduced
by Fed3 and Ca(OH)2 coagulation followed by gravity sedimentation. The
maximum and minimum fat concentrations of the treated skimmer effluent were
266 mg per liter and 10, respectively.
29
-------
TABLE 7. CHEMICAL FLOCCULATION RESULTS OF SAMPLES OF WEST PLANT REFINERY SKIMMER EFFLUENT
CO
o
Raw Wastewater ^
Characteristics Chemical Dosage
Date of Hours
Sampling Composited "COD"
9-12-69 24 3048
9-14-69 24 1682
Fed,
Oil pH mg/r
2155 6.3 100
100
100
200
200
200
635 6.5 100
100
100
Ca(OH)
mq/1 2
200
300
400
200
300
400
200
400
600
Treated Wastewater
Characteristics
mg/1
—
—
__
—
794
572
568
620
"COD"
%
Reduction
„
--
—
__
68.5
73
66
66
63
pH
6.3
—
—
6.2
6.4
--
7.2
7.4
7.5
Qualitative Observations and Conclusions
from Jar Test Results
Very poor flocculation and very poor
settling. Good flocculation. Very slow
settling floes. Treatment considered
effective.
Results no better than above. However,
larger sludge volume. Treatment con-
sidered effective.
Good flocculation, floe settling slow.
Many floes stayed in suspension. Clear
supernatant.
Almost instant floe formation. Good
flocculation. Floe settling very good.
200
200
200
200
400
600
673
594
60
64.5
Very clear supernatant. Treatment con-
sidered effective.
6.7 Good flocculation. Floe settling slow.
Many floes still in suspension, clear
supernatant.
6.8 Almost instant floe formation. Good
7.0 flocculation. Floe settling very good.
Very clear supernatant. Floes judged
to be less dense than with 100 mg/1
Fed, + comparable CafOH)^ dosage.
Treatment considered effective.
-------
TABLE 7. CHEMICAL FLOCCULATION RESULTS OF SAMPLES OF WEST PLANT REFINERY SKIMMER EFFLUENT (Continued)
Raw Wastewater
Characteristics Chemical Dosage
Date of 'Hours
Sampling Composited "COD" Oil
9-15-69 16 2523 842
9-16-69 24 1629
9-17-69 24 2155
Fed,
pH mg/r
6.2 100
100
100
6.2 100
100
100
200
200
200
8.2 100
100
100
100
Ca(OH)2
mg/1
200
300
400
200
300
400
200
300
400
200
300
400
500
Treated Wastewater
Characteristics
mg/1
846
678
762
1051
835
851
_ _
794
867
_ _
--
-_
1787
"COD"
I
Reduction
66.5
73
69.5
35.6
48.7
47.8
__
51.3
47
„_
—
__
17
PH
—
6.7
7.0
7.5
6.6
6.7
6.8
—
..
7.1
Qualitative Observations and Conclusions
from Jar Test Results
Good flocculation. Floe settling slow.
Clear Supernatant.
Good flocculation. Floe settling very
good. Clear supernatant.
Floes not well defined. Poor f'loc-
culation. Supernatant turbid.
Very good flocculation. Very good
floes. Floe settling very good.
Very good supernatant.
Floes not well defined. Poor floc-
culation. Supernatant turbid.
Good flocculation. Loose floes. Floe
settling slow. Clear supernatant.
This sample was not considered normal
by ADM. It rained heavily on this day.
pH of raw wastewater was adjusted to 6.2
with addition of H9SO,. The oil liberate!
was skimmed off before conducting test.
CaC(OH)2 dosage of 400 mg/1 and 500 mg/1
along wfth 100 mg/1 Fed, gave good
flocculation and clear supernatant.
Floes settled fast. However, COD
reduction was poor.
-------
TABLE 7. CHEMICAL FLOCCULATION RESULTS OF SAMPLES OF WEST PLANT REFINERY SKIMMER EFFLUENT (Continued)
CO
ro
Raw Wastewater
Characteristics
Date of Hours
Sampling Composited "COD" Oil pH
9-18-69 24 4993 ~ 5.6
9-19-69 24 2050 266 5.8
Chemical Dosaqe
Fed,
mg/13
100
100
100
100
100
100
100
100
Ca(OH)?
mg/1 *•
200
400
600
800
300
400
500
600
Treated Wastewater
Characteristics
mg/1
..
—
4047
..
783
804
840
--
"COD"
%
Reduction
— —
—
19
--
61.5
61
59
*•""
PH
«•_
—
7.1
—
6.2
7.0
8.2
8.3
Qualitative Observations and Conclusions
from Jar Test Results
Flocculation test gave negative results.
Floes were almost throughout the depth.
No settling, no flating of floes.
Dow N-ll at 1 to 2 mg/1 dosage along with
those coagulants also did not help.
Small floes. Slightly turbid super-
natant. Floe settling good.
Small floes. Clear supernatant. Floe
settl ing very good.
Flocculation Procedure: 500-ml sample in 1 liter beaker, chemical added; 2 minutes rapid mix at 100 rpm; 15 minutes flocculation at 40 rpm;
supernatant analyzed after 15 minutes settling.
-------
7000
5000
£ 4000
Q
O
O
1000
200 —
6000 gl80
CD
Q
Z
w
o
3000 -120
2000 -noo
o
80
A FLOW
O COD
I I I I I III II
10 20 3040506070 80 90 95
PERCENT OF TIME
FLOW AND COD ARE EQUAL TO OR LESS THAN
GRAPH VALUES
Figure 8. West plant refinery skimmer effluent variability of 24 hr
composite COD and flows.
33
-------
TABLE 8. LABORATORY SETTLING TEST RESULTS OF REFINERY SKIMMER EFFLUENT
Suspended Solids and
Sludge Production
24-hr Composite Sample
CO
•£»
Collection
Date
(Ending)
9-13-69
9-14-69
9-15-69
9-17-69
9-19-69
COD, mg/1
3,048
1,682
2,523
2,155
2,050
OH, mg/1
2,155
635
848
--
266
Coagulant
Fed 3, mg/1
100
100
100
100
100
Treatment
Dosage
Ca(OH),, mg/1
400
400
400
500
400
Settled Supernatant Duality
Laboratory*
Overflow Rate
gas/sq ft/day
1,200
600
400
200
400
400
200
1,200s
600
400
300
125
mg/1
1,314
946
998
1,050
557
778
1,734
840
820
825
794
778
COD
% reduction
57
69
67
65.6
67
69
19.5
59
60
59.5
61
62
mg/1
__
--
266
"
97
131
•-
__
_-
10
._
--
on
% reduction
--
87
--
85
85
--
—
96
--
--
Sludge
Volume
mg/1
of Waste
70
65
50
50
43
60
105
105
80
65
57
43
Suspended
Sol Ids
after
Chemical
Treatment, mg/1
904
622
884
922
..
--
--
--
Remarks and Observations
Clear supernatant. Turbidity
judged as 10 Jackson Turbidity
Units (JTU).
Visual inspection of the sus-
pension during the test indi-
cated that similar removal could be
obtained at overflow rates of 600 gpd/sq
Clear supernatant.
Clear supernatant. Turbidity judged as
JTU.
ft.
10
Samples were pretreated 1n the settling cylinder Itself, using the Indicated coagulant dosages and flocculated for 15 minutes prior to settling. These
coagulant dosages were judged as optimum by jar test analysis. The samples were at room temperature.
*One-L1ter graduated cylinder settling test, (14" water depth).
'six-Liter cylinder settling test, (14" water depth),
tin all the tests a thin scum layer was observed floating on the top. This was excluded from the supernatant sample.
-------
A series of laboratory-scale settling tests were performed at room
temperature using skimmer effluent wastewater treated with the optimum coagulant
dose of ferric chloride and calcium hydroxide (determined by a modification of
the previously described screening procedure). A summary of results is presented
in Table 8. A small percentage of solids floated to the liquid surface in all
the coagulation and sedimentation studies, indicating the need for some type
of skimming equipment in the clarifier.
The relationship between clarification overflow rate and the treated
effluent COD is presented in Figure 9. Generally, suspended solids removal
is used as a criteria in determining the design overflow rate for a sedimenta-
tion tank. However, in this particular case, with the removal of flocculated
wastewater BODs being the prime consideration, supernatant COD was used as a
criteria to determine the overflow rate. The data shown in Figure 9 is typical
of the several settling tests run and is summarized in Table 8. Interpretation
of the laboratory overflow rates indicated that in most cases a clarifier over-
flow rate of approximately 750 gal per square foot per day will produce an
effluent with a minimum COD. A scale up adjustment factor of 1.5 was applied
to the laboratory observed overflow rate to account for hydraulic turbulence,
temperature induced turbulence, inlet and outlet losses, etc. In one case
the floe settling was very rapid and a laboratory overflow rate as high as 1200
gal per square foot per day also produced very good supernatant. For design
purposes, however, an adjusted overflow rate of 500 gal per square foot per
day was used.
The effectiveness of chemical coagulation of grab samples taken from the
skimmer outlet at different times of day was not consistent, indicating that
effective equalization is necessary for reliable treatment. The chemical
coagulation of 24-hour composite samples from the skimmer effluent, in most
cases, gave good results, indicating that 24-hour equalization should be sufficient.
Chemical coagulation of equalized skimmer effluent (using a combination
of ferric chloride and calcium hydroxide) followed by flocculation and sedi-
mentation reduced the COD to 850 mg per liter or less in 50% of the samples and to
about 1,500 mg per liter or less in 90% of the samples. The behavior of two
samples taken on September 17 and 18, 1969 stresses the importance and the need
for dependable in-plant treatment with calcium chloride, and effective oil
separation and removal in the new in-plant and the existing refinery wastewater
effluent skimmers. A summary of the investigative results is given in Table 9.
TABLE 9. SUMMARY OF INVESTIGATIVE RESULTS
Item . Amount
Equalization:
Detention Time, Hours 24
35
-------
TABLE 9. SUMMARY OF INVESTIGATIVE RESULTS (Continued)
Item Amount
Coagulation:
Ferric Chloride (FeCl^) Dose, mg/1 100
Calcium Hydroxide (Ca(OH)2) Dose, mg/1
Minimum, mg/1 300
Average, mg/1 400
Maximum, mg/1 500
Clarification:
Overflow Rate (lab), gal/sq ft/day 500
Average COD Reduction, % 65
Average Fat, Oils, and Greases Reduction, % 85
Treated Effluent:
COD, mg/1, 50% Probability Value 850 (1,300)*
90% Probability Value 1,500 (2,200)
Fats, Oils, and Greases, mg/1 10-270 (10-400)
Clarifier Sludge
Minimum Solids Concentration, % 1
Maximum Solids Concentration, % 2
Average Sludge Production, % of
Wastewater Flow 6
*Figures in parenthesis indicate future production rate.
General Basis of Design and Recommended Process
The design of the treatment facilities is based on the estimated 90%
probability-of-occurrence value of future West Plant refinery skimmer effluent
flows and loadings. Chemical dosages are based on the results of laboratory
investigations using compsited skimmer effluent samples. A scale-up factor
of 1.5 has been applied to the results of laboratory overflow rates to obtain
design overflow rates for the clarifiers. Flocculator-clarifier units are
provided in duplicate for maintenance and operating flexibility.
The design of chemical feed systems is based on the requirement that
minimum supervision and operating manpower will be available. Hydrated lime,
instead of pebble lime, will be used. However, since the average lime requirement
is only 725 Ib per day as Ca(OH)2, the additional cost of hydrated lime over
pebble lime is negligible. The capacities of the lime feeder, lime mixing
tank, and transfer pumps are designed such that a whole day's lime slurry re-
quirement can be prepared in one to two hours during the day shift. The lime
slurry storage tank is designed to store three days' average supply of lime
-------
24 HR. COMPOSITE SAMPLE
A 9-12-69 TO 9-13-69
O 9-18-69 TO 9-19-69
TREATMENT CONSISTED OF OPTIMUM
DOSAGE OF FeCI3 AND Ca(OH)2
COAGULANTS
CO
CO
"U
m
5
m
-I
s
O
o
1400
1200
1000
800
600
400
200
0
I 1 I I I 1 [ I I I I I
8
10 II
12
OVERFLOW RATE, HUNDREDS OF GAL/DAY/SQ. FT.
Figure 9. West plant refinery treated skimmer effluent settling test results
-------
at a 10% slurry concentration. This will also allow the storage of weekend
lime slurry feed requirements, and thereby minimize operator attention over
the weekend. The capacity of the lime dry storage hooper is designed for 1.5
times the capacity of a delivery truck. This provides approximately 75 days'
storage at the average rate of use.
Ferric chloride will be purchased, stored, and fed in liquid form. Acid
resistant materials will be used through this system.
Equalization of skimmer effluent is necessary to level out the variability
of flows and loadings prior to chemical treatment, as was demonstrated during
the treatability survey. A portion of the existing lagoon will be modified to
serve as the equalization basin. The remainder of the existing lagoon will
provide a second basin to serve as emergency storage for excessive surface
runoff, production leaks, and spills and other miscellaneous flows. These
accumulated wastewaters will be bled into the equalization basins for further
treatment prior to discharge to the municipal sewer system.
The design assumes that the necessary changes in the refinery effluent
pumping and skimming facilities, as previously discussed, will be completed,
and that these units will remain as part of the proposed treatment scheme.
The present Imhoff tank and lagoon will be abandoned or modified for use in
proposed new treatment facilities. Since the refinery wastewaters are pre-
sently receiving CaCl2 treatment, no further adjustments were made to account
for the changes when this treatment is moved in-plant.
Description of Proposed Facilities
The schematic flow diagram for the proposed treatment facilities and
the preliminary plot plan are shown in Figure 10.
The wastewater from various refinery sources after being collected in the
existing sump, will be pumped to the existing skimmer, where free floating
oil present in the wastewater will be removed. The skimmer effluent will then
flow by gravity to the equalization basin (T-|), which will provide an equaliza-
tion period of one day at design flows. Manually-diverted influent or any
overflow from the equalization basin will be stored in an emergency storage
basin (TIQ) with a capacity of 180,000 gal and will be pumped back (via
Pump PIS) to the equalization basin at a manually controlled rate during
the periods of low in-flows, for retreatment. Under normal conditions, the
emergency basin will remain empty.
The equalized skimmer effluent will be pumped (using two centrifugal
pumps, P-J, P2) to the flash mix tank.
A constant flow rate, adjusted manually according to the level in the
equalization basin, insures more uniform operation and performance of the
downstream facilities, and simplifies control of the coagulants feeding systems.
Lime and ferric chloride will be added automatically in proportion to the waste-
water flow. Lime will be fed from a recirculated storage tank system with
a flow-activated control valve. Ferric chloride solution will be fed directly
38
-------
CO
HYDRATED LIME
(LI) SILO 12x20'
(LV> _ (LLCO)
*X—| {—*»TO DSD
STORAGE 11
LIME MIX
TANK
LIME CONVEYOR
SPLITTER
FLASH
MIXER
(LLCO) (FIG!
SKIMMINGS
-HOLD
TANK
EQUILIZATION
EMERGENCY DISCHARGE
FLOCCULATOR
CLARIFIERS
HOLD
TANK
T-5
AcoN
STORAGE . ACIDULATION
iBASIC T-IO EMERGENCY OVER
FLOW TO DSD
HLA
LLCO
FR
FT
F 1C
TS
HIGH LEVEL ALARM
LOWLEVEL CUT OFF
FLOW RECORDER
FLOW TOTALIZER
LI
NC
NO
FLOW INDICATOR CONTROLLER
TORQUE SWITCH
LEVEL INDICATOR
LOW LEVEL INDICATOR
NORMALLY CLOSED
NORMALLY OPEN
ACIDULATION
ADM
FACILITIES
Figure 10. Proposed treatment facility.
-------
from the storage tank by a variable-speed metering pump controlled by the
wastewater flow. The lime and ferric chloride feed systems will be housed
in a small control building. The lime storage bin will be located outside
the building.
From the flash mixer, the wastewater will then flow by gravity through a
flow splitter box to two 25-foot diameter flocculator-clarifiers, where the
coagulated wastewater will be flocculated by gentle stirring in the floccula-
tion zone before being settled in the outer clarification zone. The clarifier
effluent will then flow by gravity to the existing sewer leading to the muni-
cipal sewer system. The clarifier sludge will be raked to the center wells
of the flocculator-clarifier units and pumped to sludge handling facilities to
be provided by ADM.
The skimmings collected from the equalization basin and the flocculator-
clarifiers will be pumped separately using two positive displacement pumps
(P3, PS) to the acidulation facilities. Separate handling and use of progres-
sive-cavity pumps are proposed to minimize plugging during winter months due
to the hardening of the skimmer oils.
TABLE 10. SUMMARY OF DESIGN BASIS AND MAJOR UNIT SIZES
General Design Basis Probability of Occurrence
50%5UI
Refinery Skimmer-Effluent:
Flow, gpd 195,000 293,000
COD, mg/1 4,000 6,400
COD, Ib/day 6,500 12,000
Population Equivalent 38,000 70,000
Fats, Oils, and Greases, mg/1 1,550 *
Fats, Oils, and Greases, Ib/day 2,500 *
Expected Treated Refinery Effluent:
COD, mg/1 1,300 2,200
COD, Ib/day 2,100 4,200
Population Equivalent 12,500 25,000
Fats, Oils, and Greases, mg/1 300 *
Fats, Oils, and Greases, Ib/day 500 *
Estimated Total West Plant Effluent:
COD, mg/1 700 1,200
COD, Ib/day 2,500 4,700
Population Equivalent 15,000 28,000
Fats, Oils, and Greases, mg/1 220 *
Fats, Oils, and Greases, Ib/day 740 *
40
-------
TABLE 10. SUMMARY OF DESIGN BASIS AND MAJOR UNIT SIZES
(Continued)
General Design Basis
Probability of Occurrence
50%90%
Estimated Sludge Production:
Ib/day
Percent Solids
gal/day
Chemical Consumption:
Lime, Ib/day
gal/day at 10%
Fed3, Ib/day
gal/day at 40%
1,350 2,000
1.5 1.5
11,000* 16,000*
725 1,100
875 1,300
160 240
500 725
insufficient data available to calculate.
TABLE 11. SUMMARY OF MAJOR EQUIPMENT SIZES
Item
Equalization Basin, T-j:
Number
Dimensions
Capacity
Detention Time
Mixer, M-l
Materials of Construction
Effluent Pumps (Pi, P2)
Number
Type
Capacity
Control
Skimmings Holding Tank
Number
Capacity
Amount
One
50' X 50' X 18' deep (161 SWD)
300,000 gal (40,000 cu ft)
24 hours at 90% probable flow
One 15-HP platform-mounted
(Existing lagoon will be modified
sheet pile, side walls, and
concrete lined bottom
Two (one alternating standby)
Non-clog, centrifugal
250 gpm at required TDK
Manual, liquid-level override
One
1000 gal
41
-------
TABLE 11. SUMMARY OF MAJOR EQUIPMENT SIZES (Continued)
Item
Skimmings Pump (Pa):
Number
Capacity
Type
Control
(T2,
Flocculator-Clarifiers
Number
Required Flocculation Zone
Detention Time
Dimensions
Design Overflow Rate for
Settling Zone
Detention Time in
Clarifier Zone
Overall Dimensions
Volume of Each Unit
Material of Construction
Sludge and Scum Collection
Sludge Pumps (P4,
Number
Type
Capacity
Control
Skimmings Pump
Number
Type
Capacity
Control
Lime Feed Facilities
Lime Storage Bin:
Number
Capacity
(T6):
Amount
One
50 gpm
Positive-displacement,
progressive cavity
Liquid-level in skimmings tank
Two
20 minutes at 90% probable
flow
8' diameter X 6' height
500 gal per sq ft per day
5 1/2 hours at 90% probable
flow
25' diameter X 14' deep
(includes 2' freeboard plus
2' sludge zone)
6,900 cu ft or 51,500 gal
Steel
Units provided with surface skimmer.
sludge raked to center well
Two (one alternating standby)
Positive-displacement,
progressive-cavity
50 gpm at required TDH
Time controlled, constant
speed
One
Positive-displacement,
progressive-cavity
50 gpm at required TDH
Liquid-level in skimmings tank
One
30 tons of hydrated lime, or 75
days at normal feed rate
42
-------
TABLE 11. SUMMARY OF MAJOR EQUIPMENT SIZES (Continued)
Item
Volume
Material of Construction
Material Stored
Lime feeder:
Number
Capacity
Type
Mixing Tank (Ty):
Number
Volume
Dimensions
Volume of Each Unit
Material of Construction
Sludge and Scum Collection
Sludge Pumps (P4, PS):
Number
Type
Capacity
Control
Skimmings Pump (PS):
Number
Type
Capacity
Control
Lime Feed Facilities
Lime Storage Bin:
Number
Capacity
(T6):
Volume
Material
Material
Lime Feeder:
Number
Capacity
Type
of Construction
Stored
Amount
1,500 cu ft
Steel
Hydrated lime
One
800 Ib per hour
Volumetric
One
1450 gal or 190 cu ft
7' diameter X 7'depth (5' SWD)
6,900 cu ft or 51,500 gal
Steel
Units provided with surface
skimmer, sludge raked to center
well
Two (one alternating standby)
Positive-displacement,
progressive-cavity
50 gpm at required TDH
Time controlled, constant speed
One
Positive-displacement
progressive-cavity
50 gpm at required TDH
Liquid-level in skimmings tank
One
30 tons of hydrated lime, or 75
days at normal feed rate
1,500 cu ft
Steel
Hydrated lime
One
800 Ib per hour
Volumetric
43
-------
TABLE 11. SUMMARY OF MAJOR EQUIPMENT SIZES (Continued)
Item
Mixing Tank (T;):
Number
Volume
Dimensions
Lime Slurry Concentration
Mixer, M-2
Material of Construction
Lime Slurry Transfer Pumps (Py, PS):
Number
Type
Capacity
Control
Lime Slurry Storage Tank (Tg):
Volume
Dimensions
Storage Capacity
Mixer, M-3
Material of Construction
Lime Slurry Feed Pumps (Pg, PIQ):
Number
Type
Capacity
Control
Ferric Chloride Feed System:
Storage Tank (Tg):
Number
Capacity
Dimensions
Material of Construction
Material stored
Ferric Chloride Feed Pumps:
(hi. PIZ)
Number
Type
Amount
One
1450 gal or 190 cu ft
7' diameter X 7' depth (5' SWD)
10%
One at 3 HP
Steel
Two (one alternating standby)
Non-clog, centrifugal
50 gpm at required TDH
Manual
3,500 gal or 467 cu ft
9' diameter X 10' deep (81 SWD)
2,400 Ib of lime or 3 days at
normal feed rate
One at 2 HP
Steel
Two (one alternating standby)
Non-clog centrifugal
10 gpm at required TDH
Flow rate from equalization
basin effluent
One
10,000 gal or 1,430 cu ft or
24 days at normal feed rate
14' diameter X 10' deep
Steel tank with acid-resistant
lining
40% ferric chloride solution
Two (one alternating standby)
Positive-displacement pump
44
-------
TABLE 11. SUMMARY OF MAJOR EQUIPMENT SIZES (Continued)
Item
Capacity
Control
Material of Construction
Emergency Storage Basin (T-jg)
Number
Dimensions
Capacity
Materials of Construction
Emergency Pump (PIS):
Number
Type
Capacity
Control
Amount
0-40 gph
Flow rate of equalization basin
effluent
Acid-resistant
One
50' X 30' X 18' deep (SWD)
180,000 gal
(Existing lagoon will be modified)
sheet pile side walls, and concrete
lined bottom
One
Non-clog, centrifugal
250 gpm
Manual, liquid-level override
Cost Estimate for Ferric Chloride and Lime Treatment with Clarification
Preliminary estimates of the construction and annual (fixed and operating)
costs for the proposed wastewater treatment facilities are shown in the
following tabulations. The total construction cost included a construction
contingency, but excluding engineering design costs, is estimated to be
$393,000. The total annual cost, including operating costs (labor, maintenance,
utilities, and chemicals) and fixed cost {depreciation, interest, insurance,
and taxes) is estimated to be $100,300.
TABLE 12. SUMMARY OF CONSTRUCTION COSTS*
Item
Equalization Basin
Flocculator-Clarifiers
Lime Feed System
Emergency Storage Basin
Skimming Tanks
Control and Chemical Building
Amount
$54,000
95,000
19,000
32,000
6,000
40,000
45
-------
TABLE 12. SUMMARY OF CONSTRUCTION COSTS* (Continued)
Item Amount
Subtotal 258,000
Piping (at 15$) 37,000
Electrical (at 10%) 26,000
Instrumentation (at 6%) 16,000
Site Work 5,000
Subtotal 342,000
Construction Contingency (15%) 51,000
Total Construction Costs 393,000
*Based on 1969 construction prices.
TABLE 13. SUMMARY OF ANNUAL ESTIMATED COSTS1*
Item Amount
Operating Costs:
Labor Cost at $5.00/hr, 8 hrs/day, 260 days/hr $10,400
Maintenance
Mechanical at 7% of $ 65,000 4,500
Structures at 2% of $168,000 3,400
Electrical and Instruments at 7% of $ 42,000 2,900
Chemicals:
Lime, 135 tons/yr at $26.00/ton 3,500
Ferric Chloride, 30 tons/yr at $80.00/ton 2,400
Utilities:
Electrical at 0.015/KWH, 700 KW-hr/day 4,000
Total Operating Costs 31,100
46
-------
TABLE 13. SUMMARY OF ANNUAL ESTIMATED COSTS*
(Continued)
Item Amount
Fixed Costs:
Depreciation (18 years straight line) 22,000
Interest (10%) 39,300
Insurance and Taxes (2%) 7,900
Total Fixed Costs 69,200
Total Annual Costs 100,300
*Costs are rounded to nearest $100, 1969 prices.
47
-------
SECTION VIII
DESCRIPTION AND EVALUATION OF INSTALLED
WASTEWATER MANAGEMENT FACILITIES
Process evaluation was done for the equipment in Decatur. Figure 11
is a flow diagram for the treatment facilities. Flows and concentration
are approximate and are based upon operating records.
FACILITIES AT DECATUR WEST PLANT
It was decided by ADM to install one 186,000-gal clarifier at the
Decatur West Plant. It was felt that this clarifier along with process
modifications described below would bring the waste strength to within
acceptable limits. If this was unsuccessful, the process recommended by
the consultant could be installed incorporating this clarifier. This was
successful, so the ferric chloride and lime flocculation system was not
installed.
1. Condensate Recycle: The condensate generated in the refinery, while
not having any BOD, contributes a significant hydraulic load to the
treatment facility. Therefore, this water was recycled for use in
refining and water wash. Because of this, there was no net increase
in hydraulic load as production increased resulting in longer residence
time in the treatment facility. Water recycled is approximately 50,000
gallons per day.
2. Soapstock Concentration: It was found that far too much flush water was
being used on the refining centrifuges. This resulted in soapstock with
moisture in the 80% to 90% range. This caused the acidulation facili-
ties to be overloaded and resulted in poor splitting due to inadequate
reaction time. By simple external piping modifications, the soapstock
moisture was decreased to 30% to 50%. This gave a residence time suf-
ficient to treat the soapstock and resulting acid liquor even after
expansion. The water eliminated from the waste stream was at least
50,000 gallons per day.
3. Oil Spills and Malfunctions: Projects were undertaken to eliminate
possibilities of oil losses due to malfunctions. All areas where the
possibility of an oil spill existed were piped to the calcium chloride
treatment system. Each vacuum vessel which could overflow to the baro-
metric cooling tower via the ejector system was equipped with high level
alarms.
48
-------
-£»
ID
CaCI2 FLOOR
J 1
SOAPSTOCK 5500
WA9HWATFFT TAMI/ ^u»u*
WHwrivYMi trt TANK
H2S04
ACIDULATION
*
ACIDULATION
25,000 GAL/DX
^ CONCE
(SALE
42,000 PPD
DRY BASIS
RAW SOAPSTOCK
STORAGE
RAW SOAPSTOCK
HHAIN- SLOWDOWN GREASY WATER
uimiNv, 0-30,000 GPD COOLING TOWER
0-1000
R
DO GPD i
NaOH
i ,
NEUTRALIZATION
1 ' \ I, i . C
PROCESS
*Y WASTEWATER
:NTRATED SOAPSTOCK
:ABLE BY-PRODUCT)
I40O PPD
FLOATING MATERIAL]
(SKIMMINGS)
TO SOAPSTOCK
STORAGE
PPD * *
*T DEODORIZER
MUNICIPAL WATER
ZEOLITE SOFTENING
(MAKEUP WATER)
•JBOILER
^ CLEAR WATER
COOLING TOWER
v (HOLDING LAGOON L
^ ^^n V/I»L/I t»\j • L.Mwwi'i p
L MECHANICAL
^* SKIMMING TANK
1 \
(SANITARY WASTE
TO
DECATUR MUNICIPAL -—
SEWERAGE SYSTEM
SOLVENT
EXTRACTION
WASTEWATER
150,000 GPD
450 PPD BOD
ZEOLITE
REGENERATION
SLOWDOWN
SLOWDOWN to
7000 COD
-In ARIF1FR M.
-|v/L.Mr{ir lt.rl ^
J 150.000 GPD
_ 1 2O6 -500 FAT
500-1000 COD
AUTOMATIC SAMPLER
"AND FLOW METER
FROM CAUSTIC REFINING
Figure 11. Decatur wastewater treatment facility.
-------
It has been difficult to obtain meaningful operational data on the
clarifier for several reasons. The refinery started refining corn oil
weekly at approximately the same time the clarifier was put into operation.
It was found that the wastewater generated by this process was more difficult
to treat than that generated by soybean oil.
The clarifier offset the high sewer load generated by the refining of
corn oil. Therefore, it is difficult to make a meaningful before and after
comparison in terms of overall load. Reduction in COD will be given for
a particular treatment step in terms of before and after treatment, as well
as a net per pound processed reduction in strength after the clarifier was
installed.
Waste treatment facilities consist of chemical flocculation equipment,
a 30,000 gal skimmer, a 150,000 gal holding lagoon and a clarifier. Figure
12 is a plot plan of the existing treatment area at Decatur.
The calcium chloride treatment system is described in Section VI.
All refinery wastewater and floor washings are treated in this system. As
can be seen from Table 3 of Section VI, this system reduces the COD by
approximately 92%.
Table 14 presents Ib BODs and Ib fat per pound oil processed data from
April 1969 to June 1974. The calcium chloride treatment was started in
October 1969. This data is presented in Figures 13 and 14 with significant
elements of the treatment research and installation noted. Fat analysis
was not done by the city before 1970. BOD5 and fat analysis for October,
November, and December 1970 were not available.
The refinery was down due to a fire November 1969, and expanded pro-
duction resumed April 1970.
Table 15 presents population equivalents, oil processed and Ib 8005
and fat per pound of oil processed for the six month period before the cal-
cium chloride treatment was installed and the six month period after the
expanded refinery was started. There was a net reduction of 71 % of the
6005. No fat tests were run before the calcium chloride treatment was started
because the City of Decatur did not require them at that time. The last
part of Table 13 is for 1973 after the clarifier was installed and all pro-
cess modifications were made. BODs and fat reductions for each process
are summarized as follows:
Before Treatment:
.0135 Ib BOD per Ib OIL processed
.0041 Ib FAT per Ib OIL processed, estimated
After Calcium Chloride Treatment:
.0039 Ib BOD per Ib OIL processed
.00099 Ib FAT per Ib OIL Processed
71% reduction in BOD
50
-------
TO DSD
TRUCK
LOAD-OUT
SKIMMER
TANK'
TO SOAPSTOCK
r~
i
%-£'
PUMP
\
\
\
HOUSE
JPUMP HOUSE
[OVERFLOW
\
\
SKNMME
CLARIFIER SUMP
I SOAPSTOCK
2 RECOVERED OIL
3 COOKER
4 NEUTRALIZING
5 SULFURIC ACID
6 CAUSTIC
SKIMMER
BYPASS VALVE
GREASY WATER
OVERFLOW
OVERFLOW
SKIMMER BYPASS
TO CLARIFIER
SETTLING BASIN
TANK
10 SANITARY
LINE TO CITY
Figure 12, West plant treatment flow diagram.
-------
Ol
ro
z.o
CD
O
O
03
O
x 1.0
o
CaCI2 TREATMENT INSTALLED
STUDY COMPLETED
CLARIFIER STARTED
CORN OIL REFINING STARTED
ELECTROYSIS SYSTEM ABANDONED
TREATMENT COMPLETED
1969 I 1970 I 1971 I 1972 I 1973 I 1974
Figure 13. Pounds BOD per pound oil processed.
-------
OJ
CD
3
•>.
r
CD
O
TJ
TO
O
O
m
CO
0)
m
o
O
01
2.0
1.0
IL
n
11
T
l
JlJ1!
1969 I 1970 ' J97l ' 1972 ' 1973 ' 1974
MONTH
Figure 14. Pounds fat per pound oil processed.
-------
After Total Treatment System:
.0019 Ib BOD per Ib OIL processed
.00050 Ib FAT per Ib OIL processed
5155 reduction in BOD over primary
chemical treatment
50% reduction in FAT over primary
chemical treatment
86% overall BOD reduction
86% probable FAT reduction
= .0036 Ib per Ib processed
The clarifiers and system modifications resulted in a further reduc-
tion of 50% of the BODs and fat per pound processed. The overall reduction
in BODs was 86%. This is probably close to the overall fat reduction
for the system based on the overall reduction in the second step.
From the primary chemical treatment, the calcium chloride treatment
effluent gravity flows to a sump where it is mixed with barometric con-
denser cooling tower overflow and surface water from the acidulation area
and pumped to the skimmer. The skimmer has a volume of 30,000 gal and is
diagrammed in Figure A-2 of Appendix A. The skimmer retains any oil spilled
in acidulation and some additional floe from the calcium chloride treatment
system. All skimmings go to acidulation.
TABLE 14. WEST PLANT EFFLUENT ANALYSIS*
Oil Ib BOD/ Ib FAT/
Production, BOD, Fat, Ib Oil Ib Oil
Month 1b/day Ib/day Ib/day Processed Processed
Apr. 1969 407,000 6,331 ~ .0155
May 407,000 8,186 ~ .0201
Jun. 407,000 3,281 — .0081
Jul. 409,000 8,874 « .0217
Aug. 348,000 4,845 — .0139
Sep. 490,000 4,420 — .0090
Oct.. 339,000 2,159 ~ .0063
NOV.:;
Dec.
Jan. 1970
Feb.
Mar.
Apr. 603,000 2,040 893 .0034 .00150
May 726,000 3,417 448 .0047 .00062
-------
TABLE 14. WEST PLANT EFFLUENT ANALYSIS* (Continued)
Month
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
Jan. 1971
Feb.
Mar.
Apr.
May
»»-^
Jun.
Jul.
Aug.
Sep.
r
Oct.
Nov.
Dec.
Jan. 1972
Feb.
Mar.
Apr.
May
•w^
Jun.
Jul.
Aug.
Sep.
w ff m
Oct.
Nov.
Dec.
Jan. 1973
Feb.
Mar.
Apr.
May
i iv* jr
Jun.
W M 1 1 •
Jul .
Aug.
Oil
Production,
Ib/day
803,000
852,000
829,000
660,000
1,043,000
712,000
788,000
895,000
963,000
845,000
930,000
1,130,000
915,000
1,101,000
1,208,000
792,000
801,000
696,000
868,000
1,020,000
958,000
1,118,000
920,000
1,107,000
872,000
753,000
635,000
854,000
705,000
1,262,000
1,046,000
742,000
1,021,000
987,000
1,047,000
861,000
980,000
835,000
729,000
BOD,
Ib/day
2,703
3,026
3,536
2,567
—
—
--
1,480
1,534
1,640
4,450
2,470
1,480
1,596
1,994
5,142
3,900
2,926
3,593
3,450
2,188
2,688
3,117
2,156
2,338
2,034
1,414
1,921
1,089
892
764
1,836
1,156
1,887
3,757
1,683
952
1,156
1,1224
Fat,
Ib/day
587
599
1,066
722
-
—
--
460
650
699
2,479
633
427
600
741
2,157
785
590
1,124
1,149
375
557
1,523
402
818
652
807
864
303
330
226
241
139
472
517
568
183
430
228
Ib BOD/
Ib Oil
Processed
.0034
.0035
.0043
.0039
--
—
"•—
.00165
.00159
.00194
.00478
.00219
.00162
.00145
.00165
. 00649
.00487
.00420
.00414
.00338
.00228
.00240
.00339
.00195
.00268
.00270
.00223
.00225
.00154
.00071
.00073
.00250
.00110
.00190
.00360
.00190
.00100
.00140
.00170
Ib FAT/
Ib Oil
Processed
.00073
.00070
.00128
.00109
— —
--
•• ~
.000514
.000675
. 000827
.002660
.000560
. 000466
.000545
.000613
.002723
.000980
.00084
.001295
.001130
.000390
.000498
.001650
.000363
.000938
.000866
.001270
.001011
.000430
.000261
.000216
.000320
.000140
.000480
.000490
.000660
.000190
.000520
.000310
55
-------
TABLE 14. WEST PLANT EFFLUENT ANALYSIS* (Continued)
Month
Sept.
Oct.
Nov.
Dec.
Jan. 1974
Feb.
Mar.
Apr.
May
Jim.
Oil
Production,
Ib/day
733,000
942,000
900,000
1,048,000
1,030,000
1,294,000
1,089,000
1,074,000
1,167,000
1,363,000
BOD,
Ib/day
935
1,700
1,632
2,856
1,525
2,535
1,904
2,740
2,671
3,420
Fat,
Ib/day
424
559
496
1,217
437
820
529
730
645
786
Ib BOD/
Ib Oil
Processed
.00130
.00180
.00180
.00270
.00148
.00196
.00175
.00255
.00229
.00251
Ib FAT/
Ib Oil
Processed
.000580
. 000590
.000550
.001160
.00420
.000630
.000490
.000680
.000550
.000580
*Total effluent approximately 300,000 to 400,000 6PD through grant period.
+Down for refinery fire.
TABLE 15. DECATUR WEST PLANT BOD5 AND FAT PER POUND OF OIL PROCESSED
Month
Apr. 1969
May
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
Avg. Popula-
tion Equiv.
37,200
48,100
19,300
52,200
28,500
26,000
12,700
3,100
8,400
FAT, BOD,
Ib/day Ib/day
6,331
8,186
3,281
8,874
4,845
4,420
2,159
Oil Produc-
tion/day,
Ib
407,000
407,000
407,000
409,000
348,000
490,000
339,000
Ib BOD/
Ib Oil
.0155
.0201
.0081
.0217
.0139
.0090
.0063
Ib FAT/
Ib Oil
Calcium chloride treatment started
Jan. 1970
Feb.
Mar.
5,800
6,600
13,900
56
-------
TABLE 15. DECATUR WEST PLANT BOD5 AND FAT PER POUND OF OIL PROCESSED
(Continued)
Calcium chloride treatment started
Month
Apr.
May
Jim.
Jul.
Aug.
Sep.
Avg. Popula-
tion Equiv.
12,000
20,100
15,900
17,800
20,800
15,100
FAT,
Ib/day
893
448
587
599
1,066
722
BOD,
Ib/day
2,040
3,417
2,703
3,026
3,536
2,567
Oil Produc-
tion/day,
Ib
603,000
726,000
803,000
852,000
829,000
660,000
Ib BOD/
Ib Oil
.0034
.0047
.0034
.0035
.0043
.0039
Ib FAT/
Ib Oil
.00150
.00062
.00073
.00070
.00128
.00109
Total system in operation
Jan. 1973
Feb.
Mar.
Apr.
May
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
10,800
6,800
11,100
22,100
9,900
5,600
6,800
7,200
5,500
10,000
9,600
16,800
241
139
472
517
568
183
430
228
424
559
496
1,217
1,836
1,156
1,887
3,757
1,683
952
1,156
1,224
935
1,700
1,632
2,856
742,000
1,021,000
987,000
1,047,000
861,000
980,000
835,000
729,000
733,000
942,000
900,000
1,048,000
.0025
.0011
.0019
.0036
.0019
.0010
.0014
.0017
.0013
.0018
.0018
.0027
.00032
.00014
.00048
. 00049
.00066
.00019
.00052
. 00031
.00058
.00059
.00055
.00116
The overflow from the skimmer gravity flows to a sump where it is
pumped to the clarifier. The clarifier is a 183,500 gal Rex Wastewater
Clarifier (Figure C-l of Appendix C). The clarifier is 50 ft in diameter
with a surface area of 1950 sq ft. This provides an overflow rate of
75 gal per sq ft per day based on a flow of 150,000 gal per day. A 60 day
average of clarifier influent versus clarifier effluent COD's resulted in
an average reduction of 45% through the clarifier. Samples were based on
24 hour composites but were not flow proportioned. This data is presented
in Table 16.
57
-------
TABLE 16.
CLARIFIER
COD REDUCTION
Date
11/08/71
11/09/71
11/10/71
11/11/71
11/12/71
11/15/71
11/16/71
11/17/71
11/18/71
11/19/71
11/22/71
11/23/71
11/24/71
11/26/71
11/29/71
11/30/71
12/01/71
12/02/71
12/03/71
12/06/71
12/07/71
12/08/71
12/09/71
12/13/71
12/14/71
12/15/71
12/16/71
12/17/71
12/20/71
12/21/71
12/22/71
12/27/71
12/28/71
12/29/71
12/30/71
1/03/72
1/04/72
1/05/72
1/06/72
1/10/72
1/12/72
1/13/72
1/14/72
Influent
COD (mg/1)
1369
790
3579
1790
2900
1970
3080
2100
1840
2250
1840
2260
9731
3550
1210
5030
10000+
4460
3760
2780
2160
8760
6000
2580
1540
4740
1800
5665
1950
825
3500
2407
2064
3326
8750
1080
2455
10000+
2170
2275
3375
1725
3560
Ef f 1 uent
COD (mg/1)
1316+
790
1555
1420
1050
579
1100
526
684
420
920
893
2314
1210
1210.,
2680?
5360
2310
1900
1700
1440
4950§
1560
2580
2830
2260
1800
3141#
1900
2000
1250
1700
1746
1267^
3168#
585
1820
2300
1171
1455
2250
1725
3000
58
-------
TABLE 16. CLARIFIER COD REDUCTION (Continued}
Influent Effluent
Date COD (mg/1) COD (mg/1)
1/18/72 1595 1740
1/19/72 2465 2247
1/20/72 1885 725
1/21/72 1667 943
1/24/72 1392 956#
1/26/72 2375 1241
1/31/72 1692 1692
2/01/72 1690 1400
2/02/72 2000 693
2/03/72 1181 1155
2/07/72 970 908
2/08/72 1196 700
2/09/72 1847 960
2/10/72 2000 763
2/11/72 1456 640
2/14/72 2408 2254
MEAN 2997 1642
*Reduction in COD is 45%.
+Acid liquor to lagoon.
^Overflow soap-stock tank.
iAcid liquor to clarifier-black
#Corn oi1.
Clarifier influent and effluent was monitored for 20 consecutive days
from March 24 to April 14, 1972, and analyzed for fat, COD, pH, and tempera-
ture. This data is presented in Table 17 and Figures 15, 16, 17, and 18.
Samples were taken on a 24 hour composite, but were not flow proportioned,
A reduction in fat of 53% and a reduction in COD of 41% was noted.
The temperature of the effluent was stabilized at 34°C. The pH was
slightly decreased throughout the period. The pH of the influent was high
April 9 through April 11 due to a leaking seal on a caustic pump. The clari-
fier kept the pH below 9.5 throughout this period and returned to normal quickly
after the seal was repaired.
High waste strength when the plant was on corn oil was primarily because
corn oil was not degummed before refining. This was because no appreciable
market for corn lecithin existed. Water soluble phosphatides are separated
with the soapstock and some come out in the wash-water. Their gums are not
readily precipitated by calcium chloride and a high wasteload results.
-------
12
II
10
9
8
7
6
5
4
3
BROKEN CAUSTIC LINE
--INFLUENT
— EFFLUENT
I I I I I I I I I I 1 I I I I I I I I I I
24 25 26 27 28 29 30 31
I 2 3 4 5 6 7 8 9 10 II 12 13 14
DAY MARCH-APRIL 1972
Figure 15. Clarifier pH vs day ADM Decatur.
-------
60
50
1 40
m
m 30
0
20
10
n
- A A /
,-' • ,'--.y \
• s
i
—
\
\
" — J 1 x — — • — -^
l^x'
.-.--- IMPI IIPMT
" *~ ~ inrL.ucni
•*•.«»•
F FFLU E NT
1 I 1 1 1 1 1 1 1 1 1 i 1 1 1 1 1 1 1 1 1
24 25 26 27 28 29 30 31 I 23 4 5 6 7 8 9 10 II 12 13 14
DAY MARCH-APRIL 1972
Figure 16. Temperature vs day ADM clarifier, Decatur, Illinois.
-------
ov
5000
o
o
4000
3000
2000
1000
INFLUENT
EFFLUENT
j CORN |
\
1 I L I L I
1 J_ J_ I I I I I I I I I I I
24 25 26 27 28 29 30 31 I 2 345 6 78 9 10 II 12 13 14
DAY MARCH-APRIL 1972
Figure 17. Clarifier COD vs day ADM Decatur.
-------
INFLUENT
EFFLUENT
CO-CORN OIL PROCESSING
SO * SOYBEAN OIL PROCESSING
CO
1800
1600
1400
1200
1000
800
600
400
200
I I I I I I I I I I I I
I
I
I
I
I
I
I
1
I I
24 25 26 27 28 29 30 31 I 2 3 4 5 6 7 8 9 10 II 12 13 14
DAY MARCH-APRIL 1972
Figure 18. Fat vs day Decatur ADM clarifier.
-------
Gums in the soapstock are not converted to fatty acids and remain in the acid
water phase. This was and has remained a problem. However, since 1972
a small market for corn lecithin has been cultivated by sales, which has
partially resolved the problem.
The COD of the acid liquor is highly variable depending on the nature
of the soaps and the amount of floor washings in a particular batch. Therefore,
the lagoon influent COD's give only a range of values that can be expected
at a given time. The volume of acid liquor is approximately 25,000 gal
per day produced in three or four equal batches. The steady increase in
strength of the effluent indicates-a need to clean the lagoon periodically,
every two or three months. (Table 18)
Economic Analysis of Decatur West Refinery Waste Treatment System
From the process evaluation, it will be assumed fat losses were reduced
.0036 Ib per pound processed by the calcium chloride and clarification sys-
tem. This is recovered as acid oil, the price of which is highly variable.
This is an average of 3089 Ib per day based on a daily production of 858,000
Ib of oil. Table 19 is the total installation and operation cost of the cal-
cium chloride treatment system and the clarifier. Tables 20 and 21 are
a summary of waste treatment costs on a per Ib 8005, per Ib fat removed, per
gal treated, and treatment cost per kg oil processed. The oil recovered must
be processed by acidulation at a cost of 2
-------
TABLE 17. DECATUR CLARIFIER OPERATION* (Continued)
Date pH Fat COD Temp. pH Fat COD Temp.
(mg/1) (mg/1) (C°) (mg/1) (mg/1) (C°)
4/08/72
4/09/72
4/10/72
4/11/72
4/12/72
4/13/72
4/14/72
Average
6,
10.
10,
10.
6.
7,
6,
1216
244
253
47
2136
613
682
544
3416
627
454
247
2757
1421
2375
1388
44
44
53
6.1
5.8
7.6
9.4
6.8
6.4
5.9
270
133
66
18
87
118
191
223
960
808
250
418
467
433
773
818
34
33
34
*Reduction is 59% for fat and 41% for COD.
TABLE 18. LAGOON OPERATION COD (mg/1)1
Date
1/05/72
1/15/72
2/07/72
2/22/72
2/28/72
3/06/72
Average
Influent
14,600
8,650
11,475
7,618
19,260
14,250
12,642
Effluent
3,580
5,600
6,235
7,194
9,250
10,000
6,976
*COD removal is 45%.
65
-------
TABLE 19. INSTALLATION AND OPERATION COSTS FOR WEST REFINERY WASTE TREAT-
MENT SYSTEM
Item Cost per year
Calcium Chloride (Costs for 1969):
Depreciation @ 10% per year based on total
capitalized cost of $25,000 $ 2,500
Calcium Chloride @ 7i per Ib dosage of 0.25% 21,308
Maintenance 1,000
Labor 4 hours per day 7,500
$ 32,308
Clarifier (Costs for 1971):
Depreciation @ 5.5% per year based on total
capitalized cost of $96,621 $ 5,367
Utilities 5,000
Interest 9,662
Insurance and Taxes 1,932
Maintenance
Mechanical 5% 5,000
Structures 1,000
Labor 1 man 15,000
$ 42,961
Total Treatment Cost $ 75,269
TABLE 20. SUMMARY OF TREATMENT COSTS
Item Treatment Clarifier Total
Cost per day $ 88 $118 $206
BOD removal (Ib per day) 8236 1716 9952
Cost £ per Ib BOD removed 1.06 6.87 2.07
Fat removal (Ib per day) 2660 420 3080
Cost i per Ib fat removed 3.3 28.1 6.7
Cost i per gal .059 .077 .136
(based on 150,000 gal per day)
Cost t per kg oil produced .019 .026 .045
66
-------
TABLE 21. NET COSTS (COST MINUS REVENUES*)
Acid Oil
Price
4* per Ib
18* per Ib
Cost per
1000 liters
$.251
-$.510
Cost per
1000 gal
$0.95
-$1.93
Cost per
kg Oil Processed
.032*
-.063*
Revenue: 4* per Ib Acid Oil Revenue = (.04-.02) (3080) = $61.76 per day;
18* per Ib Acid Oil Revenue = (.18-.02) (3080) = $494.08 per day.
67
-------
HA
-INLET
MD RAIL
50' *
DRIVE UNIT,
RRIDGE
T3
T3
n>
ex
-J.
X
3>
SKIMMER
^SCUM BAFFLE
u
WEIR
WELL
EFFLUENT
TROUGH
SWEEP ARMS
TORQUE
TUBE
-a rn
-o o
n i—i
oo x
STEEL TANK'
Figure A-1. Rex wastewater clarifier (186,500 gallons).
-------
VO
7'-6"
/
OIL DISCHARGE
O
INLET
^57-6"
Figure A-2. Walker skimmer.
-------
APPENDIX B. ANALYTICAL PROCEDURES
Fats. Oils, and Greases - Analytical Procedure Used During R.F.W. Survey
1. The samples were thoroughly shaken (and heated, if necessary) and
were acidified with 5 ml of concentrated h^SCty and heated on a steam
bath for 15 to 30 minutes.
2. The samples were cooled and 50 ml of hexane added to the sample bottle
(100 mg of hexane was added if an overabundance of grease was present).
3. The sample bottle was shaken manually for 30 seconds at about 240
shakes per minute rate.
4. The entire contents of the bottle were poured into a one-liter separatory
funnel and allowed to separate. The water layer was decanted into
a one-liter graduated cylinder and the volume recorded. The hexane was
then allowed to filter through a filter-eel precoated, whatman No. 1
covered, buchner funnel. The precoat consisted of about 30 ml of a
10 gm/1 slurry of filter-eel.
5. The sample bottle, transfer funnel, and separatory funnel were rinsed
with about 10 - 15 ml of hexane and this was added to the 50 ml and also
filtered.
6. The filtered hexane was collected in test tubes in the vacuum flasks
and transferred to 30-ml capacity, pre-tared aluminum weighing dishes
and evaporated on a steam hot plate. Any water-hexane emulsion was
also filtered and evaporated.
7. The evaporated dishes were placed in a 103°C oven for at least one
hour, transferred to a desicator and re-weighed.
8. Fats, oils, and greases, mg/1 =
Final Weight - Initial Weight of Dish
in mg in mg x 1000
ml sample
70
-------
A-D-M COD Test Procedure
1. Prepare sample of aliquot and place in 250 ml reflux flask. (Size
of sample depends on strength of aliquot - 10 ml if COD is 1000
below or proportionately less if COD is higher.)
2. Add 10 ml potassium dichromate (0.125 n), 15 ml cone, sulfuric acid
and pumice to each flask and reflux for 15 minutes.
3. Set control and titrate with ferrous ammonium sulfate (0.067 n) to
determine normality .125 (10 ml potassium dichromate, 15 ml
ml used ,
cone, sulfuric acid, cool to 60°C, add 3 drops ferroin , 50 ml
distilled water and titrate).
4. After 15 minutes cool flask to below 60°C and titrate with ferrous
ammonium sulfate (0.067 n) recording ml used.
5. Calculate COD = A-B x normality x 8000
ml of sample
A = ml of titrant for control
B = ml of titrant for sample
Normality = 1.125
ml of control
Ferroin Indicator = 1.485 grams 1.10 Phenanthroline and 0.695 grams
Ferrous Sulfate per 100 ml distilled water.
71
-------
APPENDIX C. WASTE TREATMENT FACILITIES AT ADM, LINCOLN, NEBRASKA
The calcium chloride treatment is essentially the same as that in
Decatur and will not be discussed here.
All refinery effluent is pumped to a 30,000 gal split box where it is
skimmed and overflows to a series of sumps where it is mixed with blowdown
from the barometric condenser cooling tower and pumped to the skimmer.
The skimmer effluent is pumped to the clarifier. The water from the clari-
fier flows through the sampling station to the sanitary sewer. (Figure C-l)
The water from acidulation is also pumped to the split box. The clari-
fier water is always dark in color due to the acid liquor; however, the clari-
fier does an effective job of BOD§ removal. Acidulation, a major contributor
to the sewer load, was shut down in May 1971 and started again in September
when the clarifier was put into operation. A drop in the sewer load was noted
when acidulation was shut down and no increase was noted when it was started
again with the clarifier in operation. Refer to Figures C-2, C-3, C-4, and
Table C-l. Therefore, a before and after comparison was made for the period
after the clarifier was started and before acidulation was shut down.
TABLE C-l. ANALYSIS OF DISCHARGE TO MUNICIPAL SEWER
Month
Jan. 1971
Feb.
Mar.
Apr.
May
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
Jan.- 1972
Feb.
Mar.
COD
Ob/day)
3191
4128
4467
4116
1589
1182
1590
1907
1497
1589
1510
1472
1765
1118
662
Suspended
Solids (mg/1)
756
982
780
735
272
256
248
309
209
225
183
155
145
199
97
Fat
(mg/1)
310
571
508
484
233
206
229
246
213
229
181
160
5150
99
73
72
-------
CO
CdCI2 FLOOR DRAINS
BLOWDOWN
0- 30.000 GPD
0-1000 PPM
FAT
GREASY'WATER
COOLING TOWER
DEODORIZER
JL
J
SOLVENT
EXTRACTION
WASTEWATER
PERIODIC SK
SOAPSTOCK <
TO
H2S04
~L
ACIDULATION
25,000
GPD
NaOH
NEUTRALIZATION
ACIDULATtON PROCESS
7000 PPD '
DRY
BASIS
WASTEWATER
^CONCENTRATED SOAPSTOCK
(SALEABLE BY-PRODUCT)
MUNICIPAL WATER
ZEOLITE SOFTENING
MAKEUP WATER
ZEOLITE
-HBOIL.ERI-
REGENERATION
BLOWDOWN
CLEAR WATER
COOLING TOWER
BLOWDOWN
MECHANICAL
SKIMMING TANK
-JCLARIFIER
^
30,000 PPD
DRY BASIS
1500 PPD
RAW SOAPSTOCK
STORAGE
RAW SOAPSTOCK
[^FLOATING MATERIAL i
(SKIMMINGS)
TO SOAPSTOCK
STORAGE
SANITARY WASTES!
TO
LINCOLN MUNICIPAL
SEWERAGE SYSTEM""
AUTOMATIC SAMPLER
AND FLOW METER
FROM CAUSTIC REFINING
Figure C-l. Lincoln wastewater treatment facility.
-------
These data are presented in Figures C-2, C-3, and C-4. A comparison
of parameters for January - April, 1971 with October 1971 - March 1972
gives the following average reductions:
Parameter
Fat
SS
COD
% Reduction
66.3
78.7
65.6
The higher reduction in suspended solids was expected, since this
type of material is treated especially well by the clarifier.
-------
Ul
500
400
300
200
100
ACIDULATION CLARIFIER STARTED
I I I I I I I i I I I I I I I
J FMAMJJASONDJ FM
MONTH JANUARY-1971 THRU MARCH" 1972
Figure C-2. ADM Lincoln, Nebraska average mg/1 fat vs month January 1971-March 1972.
-------
CO
c
m
o
m
o
CO
O
o
CO
1000
900
800
700
600
500
400
300
200
100
0
ACIDULATION ICLARIFIER STARTED
I i I I I I I 1
1 I I 1 1
MAMJJASONDJ
MONTH JANUARY-1971 THRU MARCH-1972
M
Figure C-3. ADM Lincoln, Nebraska suspended solids vs month January 1971-March 1972.
-------
o
o
a
o
5
o
z
m
zo
m
5000
4000
3000
2000
1000
^ACIDULATION
DOWN
CLARIFIER
rSTARTED
1 1 1
J F M
| |
A N
1 1 1 1 1 1 1 I 1 1
AJJASONDJFM
MONTH JANUARY-1971 THRU MARCH-1972
Figure C-4. ADM Lincoln, Nebraska average Ib OOD/day vs month January 1971-March 1972.
-------
SECTION X
BIBLIOGRAPHY
Bailey, A. E. Industrial Oil and Fat Products, New York, Interscience
Publishers, Inc. 1945.
Standard Methods for the Examination of Water and Wastewater, 12th Edition.
New York, American Public Health Association.1965.
Weston, R. F. "Conceptual Design Report, Upgrading of West Plant Wastewater
Treatment Facilities. West Chester, Pennsylvania. January 1970.
73
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-294
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Treatment of Effluent Waters from Vegetable Oil
Refining
5. REPORT DATE
December 1976 (Issuing date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Gill, D. F. and lelase, J. C.
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Archer Daniels Midland Co.
4666 Paries Parkway
Decatur, IL 62525
10. PROGRAM ELEMENT NO.
1BB037 ROAP 21 BAA Task 20
11. CONTRACT/GRANT NO.
12060 FDK
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Laboratory - Gin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A detailed investigation was done to characterize the wastewater from a vegetable
oil refinery. A calcium chloride chemical treatment was installed which resulted
in a net decrease in waste load of 71 percent from .0135 Ib BOD5 per pound oil
processed to .0039 Ib BOD5 per pound oil processed. A further reduction of 50
percent to .0019 Ib BOD per pound oil processed was made by the addition of a
clarifier to the treatment system. The net cost of treatment depends on the
market value of the oil recovered and will range from $.032 per pound processed
to a gain of $.063 per pound oil processed
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Clarification
Precipitation (Chemistry)
Refining
Vegetable Oils
Waste Treatment
Chemical Removal (Water Treatment)
Vegetable Wastes,
Economics, Vegetable
Industry, Treatment
Costs
13B
7A
6H
13H
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
89
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
79
U.S. GOVERNMENT PRINTING OFFICE: 1977-757-056/5577 Region No. 5'H
------- |