EPA-600/2-76-304
December 1976
Environmental Protection Technology Series
PROCEEDINGS SEVENTH NATIONAL SYMPOSIUM ON
FOOD PROCESSING WASTES
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-600/2-76-304
December 1976
PROCEEDINGS SEVENTH NATIONAL SYMPOSIUM
ON FOOD PROCESSING WASTES
April 7-9, 1976
Atlanta, Georgia
Co-sponsored by
NATIONAL CANNERS ASSOCIATION
AMERICAN MEAT INSTITUTE
SOUTHEASTERN POULTRY & EGG ASSOCIATION
PACIFIC EGG & POULTRY ASSOCIATION
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory-Cincinnati, 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.
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly more efficient pollution con-
trol 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.
The Seventh National Symposium on Food Processing Wastes was co-sponsored
with the National Canners Association, American Meat Institute, Southeastern
Poultry & Egg Association, and Pacific Egg & Poultry Association. The pri-
mary purpose of these symposia is the dissemination of the latest research,
developmentxand demonstration information on process modifications waste
treatment, by-product recovery and water reuse to industry, consultants and
government personnel. Twenty-five papers are included in this Proceedings
as well as the final registration list.
These symposia will be continued; if you are interested in participating
or wish to receive additional information about the Eighth, contact:
Industrial Pollution Control Division
Industrial Environmental Research Laboratory--Ci
Environmental Protection Agency
Cincinnati, Ohio 45268
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
iii
-------
CONTENTS
Foreword
111
ERA'S INDUSTRIAL ENVIRONMENTAL RESEARCH PROGRAM 1
Dr. Wilson K. Talley
THE EPA REGION IV PROGRAMS 7
John A. Little
WHY ARE WE HERE? AN INDUSTRY OVERVIEW 10
Walter A. Mercer
A WASTEWATER TREATMENT STUDY FOR SKOKOMISH PROCESSING PLANT 21
Dr. S. S. Lin, Dr. Paul B. Liao, and Max W. Cochrane
RECLAMATION AND TREATMENT OF CLAM WASH WATER 42
R. R. Zall, L. F. Hood, W. 0. Jewell, R. L. Conway,
and M. S. Switzenbaum
PILOT PLANT PRODUCTION OF A FUNCTIONAL PROTEIN FROM FISH 67
WASTE BY ENZYMATIC DIGESTION
G. 0. Bucove and G. M. Pigott
AN IMMOBILIZED-ENZYME PILOT PLANT FOR THE TREATMENT OF 83
ACID WHEY
Dr. M. Charles, Dr. R. W. Coughlin, and
Dipl.-Ing. K. Julkowski
IN-PLANT CONTROL TECHNOLOGY FOR THE FRUITS AND VEGETABLES 100
PROCESSING INDUSTRY
Kenneth V. LaConde and Curtis J. Schmidt
LAND DISPOSAL OF WINERY WASTEWATER 110
Dr. Larry L. Russell, Dr. John N. DeBoice, and
Dr. Walter W. Carey
LOW WASTEWATER POTATO STARCH/PROTEIN PRODUCTION PROCESS- 118
CONCEPT, STATUS, AND OUTLOOK
J. R. Rosenau, L. F. Whitney, and R. A. Elizondo
-------
CHANGES IN ORGANIC AND INORGANIC CONSTITUENTS OF WASH WATER 129
UPON RECYCLE IN A PROTOTYPE LEAFY-GREENS WASHER
R. C. Hoehn, P. B. Gearing, M. E. Wright, and
W. H. Robinson, Jr.
TOMATO FLUME WATER RECYCLE WITH OFF-LINE MUD REMOVAL 157
George E. Wilson, Wally R. Rose, and Jerry Y. C. Huang
PROTOTYPE APPLICATION OF SINGLE PARAMETER SLUDGE AGE CONTROL 189
TECHNOLOGY—A CASE HISTORY—PERFORMANCE OPTIMIZATION BY
APPLICATION OF SLUDGE AGE CONTROL TO EXTENDED AERATION
TYPE TREATMENT PLANT FOR FOOD PROCESS WASTES
Calvin G. Brown
CONTROL OF COLOR PROBLEMS DURING RECYCLING OF FOOD PROCESS 237
WATERS
A. Hydamaka, P. Stephen, R. A. Gallop, and L. Carvalho
THE TOXICITY OF FOOD PROCESSING EFFLUENTS TO FISH 257
D. W. Bissett
CHARACTERIZATION AND POTENTIAL METHODS FOR REDUCING WASTE- 273
WATER
D. L. Grothman, L. J. P. Scully, P. M. Berthouex, and
D. 0. Dencker
POULTRY PROCESSING WASTEWATER—ADVANCED TREATMENT AND REUSE 298
Daniel T. McGrail
ALTERNATIVES FOR TREATING POULTRY PROCESSING WASTEWATER 308
Dr. Franklin E. Woodard
PROCESSING EGG BREAKING PLANT WASTE 331
J. M. Vandepopuliere, H. V. Walton, W. Jaynes, and
0. J. Cotterill
WATER USAGE IN POULTRY PROCESSING—AN EFFECTIVE MECHANISM 338
FOR BACTERIAL REDUCTION
F. A. Gardner and F. A. Golan
TREATMENT OF PACKINGHOUSE WASTEWATER BY SAND FILTRATION 355
Dr. M. L. Rowe
WASTE TREATMENT FOR SMALL MEAT AND POULTRY PLANTS 357
Jack L. Witherow
EVALUATING AND TREATING POULTRY PROCESSING WASTEWATER 410
W. K. Whitehead
A MEAT PACKER'S SOLUTION TO MEETING 1983 EFFLUENT REQUIREMENTS 432
Joseph A. Home!, Jr-. and Jack McVaugh
-------
EVALUATION OF VIBRATORY BLANCHER-COOLER FOR SNAP BEANS AND 450
LIMA BEANS
J. L. Bomben, W. C. Dietrich, J. S. Hudson, E. L. Durkee
R. Rand, J. W. Farquhar, and D. F. Farkas
TREATMENT OF MEATPACKING PLANT WASTEWATER BY LAND APPLICATION 470
Anthony J. Tarquin
REGISTRATION LIST 485
vii
-------
ERA'S INDUSTRIAL
ENVIRONMENTAL RESEARCH PROGRAM
by
Dr. Wilson K. Talley1
As most of you in attendance today know, the U.S. Environmental Protection
Agency, EPA, was created by Presidential order in December of 1970. This
was, at least in part, a result of an increased public awareness of and concern
with the environment. Both the President and the Congress initiated a coor-
dinated effort through this new agency that would, as a matter of explicit
national policy, make environmental concerns an integral and important part
of our economic and social life.
The creation of EPA brought together 15 different, but related, programs
from several Federal Government agencies concerned with such environmental
problem areas as: air and water pollution, solid waste management, pesticides,
water supply, radiation, noise, and toxic substances. The overall mission
of EPA is the enhancement and maintenance of environmental quality in a
way that is consistent with other national goals. Specific functions performed
by EPA include:
- setting and enforcing environmental standards;
- researching the causes, effects, and control of environmental problems;
- assisting States and local governments through a variety of planning
and waste treatment facility construction grants;
- providing technical assistance and disseminating information on
environmental problems and their solutions; and
- demonstrating how to protect and enhance the environment.
Considerable effort has been expended upon balancing our environmental
goals and priorities with our national economic and energy objectives.
1Assistant Administrator, Office of Research and Development, U.S. Environ-
mental Protection Agency, Washington, DC
-------
Thus far, it appears that we have been successful. A recent public opinion
poll states in summary: "Even during a time of recession, high unemployment
and rising fuel costs, the public does not voice a readiness to cut back^
on environmental control programs to solve economic and energy problems."
Indeed, many of our energy and materials problems can be solved or,
at the very least, reduced by a return to the old conservation ethic -
"Use less, use it better, use it longer, and thus waste less."
A portion of the success of EPA's environmental programs can be directly
attributed to a strong research and development program. The Office of
Research and Development (ORD) provides EPA with the data and information
it needs to fulfill its responsibility as a regulatory agency.
Principal outputs of ORD include:
- Information which contributes to the scientific and technical bases
for regulatory standards.
- Standardized methods to measure and assure quality control in programs
to assess environmental quality, implement regulations and enforce
standards.
- Cost-effective pollution control technology and incentives for accep-
tance of environmentally sound options.
- Scientific, technical, socio-economic, and institutional methodologies
that are needed to judge environmental management options and to
balance these options against competing national needs.
By necessity, our research program is both multidisciplinary and multi-
media in nature. It is multidisciplinary because it covers virtually every
EPA area of technical responsibility and it must provide support to each of
EPA's programs and regional offices. The research program is multimedia
because pollution occurs in all media, especially that produced as a by-
product of industry. Finally, and sadly, correction of one pollution problem
(e.g., water) can result in other pollution problems, if care is not taken.
ORD's resources consist of about 1,800 technical and support personnel
located in 15 laboratories and in Washington, D.C. For this fiscal year
(1976), our budget is approximately $250 million. It is readily apparent
that our program is not entirely in house. In addition to the research
conducted by ORD scientists and engineers, we manage a large extramural grant
and contract program in cooperation with colleges and universities, industrial
organizations, research institutes, and State and local governments. Finally,
to avoid unnecessary duplication of federal skills and capital investment,
we enter into interagency agreements with other federal agencies.
\
We reached these operating modes over a long period of time. To
accomodate to the present state, we have recently completed a reorganization
of the Office of Research and Development. Among other objectives, the
-------
reorganization establishes clear lines of responsibility, authority, and accoun-
tability. We are striving to reduce the amount of paperwork, substituting
closer working relationships between field research personnel and the head-
quarters staff. To accomplish this, I have reduced the number of administrative
levels and given more responsibility to the research laboratories.
Unhappily, certain well intentioned restrictions on our operations
have created a "Catch-22" situation. Chief among these is a restriction
on the fraction of our budget that can be allocated to travel. The ceiling
is well meant, certainly, but it fails to recognize that each of our labora-
tories conducts a research program national in application and largely extra-
mural in execution. Well, the solution appears easy—simply shift from
extramural to a more in-house R&D effort by hiring more ORD staff. Unfor-
tunately, this conflicts with attempts to restrain the growth of the federal
bureaucracy, so we must seek other solutions.
Organizationally, ORD is now divided into four functional offices:
The Office of Monitoring and Technical Support; the Office of Health and
Ecological Effects; the Office of Air, Land, and Water Use; and the Office
of Energy, Minerals, and Industry.
The Office of Energy, Minerals, and Industry responsibilities include
assessing, developing, and demonstrating technology to abate pollution
from industrial point sources. This office also plans and administers a
large comprehensive energy/environmental research, development, and demon-
stration program involving EPAa and 15 other federal agencies.
Within the Office of Energy, Minerals, and Industry, the headquarters
staff is responsible for long range program planning and for staff support
of the Office. The responsibility for implementation of the research, deve-
lopment, and demonstration programs of-OEMI rests with the Industrial Environ-
mental Research Laboratory in Research Triangle Park (RTP), North Carolina
and the Industrial Environmental Research Laboratory, Cincinnati, Ohio.
In the industrial area, the RTP laboratory has the lead role for iron,
steel and ferroalloys, petroleum refining, petrochemicals, agricultural
chemicals, and textiles. The Cincinnati laboratory has the lead role for non-
ferrous metals, metal finishing, inorganic chemicals, specialty chemicals,
rubber® plastics and organics, pulp, paper and wood products, food products,
and miscellaneous industries. Note that the divisions of responsibility are
by industry, rather than by medium - air, water, land - or by pollutant.
This is a step toward finding complete solutions, rather than partial, sub-
optimum palliatives.
Historically, these industrial research programs have been concerned
with the development and demonstration of new or improved, cost-effective,
waste management technology with industry-wide applicability. As a result,
EPA has been able to establish economically and technically feasible effluent
guidelines and treatment parameters for liquid-waste discharge permits.
These programs have also assisted in the implementation of ambient air quality
standards and the development of New Source Performance Standards. Research
of this type will continue in order to respond to the technology requirements
-------
of legislation including both the FWPCA and the Clean Air Act.
A major portion of EPA's research and development program for the food
processing industry is implemented through the Corvallis, Oregon, Field
Station of the Cincinnati Industrial Environmental Research Laboratory's
Food and Wood Products Branch. Because of the fundamental importance of the
food industry, EPA has had an active R&D program for this industrial segment
for over nine years. During this time, our research staff has developed
a close working relationship with industry, universities, and other Federal
agencies. This cooperation is absolutely necessary, because without it,
our research program would be operating without critical interfaces with
other segments of the research and development community.
During the past nine years our research and development program has
participated in over 100 extramural/contract and grant projects in the
food processing area. Our financial contribution has amounted to over 15
million dollars. These projects have covered the whole spectrum including:
secondary treatment, joint municipal and industrial treatment, process
modifications, product and by-product recovery, as well as process water
reuse and recycle. Some of the more significant accomplishments to date
of this R&D program would include:
- Commercial demonstration of dry caustic peeling of peaches and white
potatoes.
- Commercial recovery of protein and lactose from cheese whey by ultra-
filtration and reverse osmosis.
- Full-scale fungal fermentation of cheese whey with recovery of fungal
mass for incorporation into animal feed.
- Full-scale biological treatment systems for several segments of the
food industry which demonstrated that "exemplary" effluents can be
economically achieved.
- A demonstration of process water reuse (zero discharge) for the beet
sugar industry.
=- Commercial utilization of sugar cane bagasse and leafy trash for
power generation.
- Cost-effective methods of reducing water use in poultry processing.
In addition to these, there have been over 55 other projects completed
to date, with approximately 70 reports published and available to the public.
Nearly all of these projects have been reported on at one of the previous
six National Symposia on Food Processing Wastes. Proceedings from each of
these Symposia are also available.
I would like to digress for a minute to say that symposia such as this
one serve a very useful purpose. The development of new or improved proces-
-------
sing methods, treatment systems, etc., is useless unless the information is
made available to and used by industry. Therefore, I would like to personally
acknowledge this year's co-sponsors: The National Canners Association,
American Meat Institute, Southeastern Poultry and Egg Association, and the
Pacific Egg and Poultry Association.
We have been (and still are) aware of some of the unique problems with
which this industry is faced. Taken together, the various segments of the food
industry have a modest R&D budget and most of it is used to insure the produc-
tion of safe, nutritious foods. This requires keeping abreast of and
implementing the requirements of several other Federal agencies and depart-
ments in addition to EPA such as FDA, USDA, OSHA, NIOSH, etc. Some of the
food processing industry is also faced with: short processing seasons
(2 months or less); significant variability in processing (hour to hour,
day to day, and year to year) which compounds end-of-pipe treatment problems;
very low profit margins; inability to acquire money for significant capital
expenditures—whether for treatment or for plant renovation; as well as high-
strength wastes (raw waste BOD of several thousand mg/1). An additional
complicating factor is that a large number of the processing plants are very
small; in some cases they are a family operation.
Currently, there are several significant projects underway on process
modifications, by-product recovery and process water reuse and recycle.
During the next two and one-half days, you will be hearing presentations on
many of these. I would like to briefly mention two of these, since I believe
they will be looked upon as "major milestones." Both projects have as their
objective the treatment (including polishing) of the processing plant effluent
water and then evaluating the quality of the treated water for potential reuse
back in the processing plant. One project is at a poultry processing plant
in Maryland, and the second is at a fruit processing plant in the State of
Washington. Because of the complexity of these efforts, a very active and
involved committee has been established for each study. Besides our industrial
waste R&D personnel, the committees include participants from industry,
EPA's Health Effects Research Laboratory in Cincinnati, FDA, and USDA.
The next point I would like to discuss briefly is the future. Where
do we go from here? In fiscal year 1977, ORD will have approximately the
same budget as it has this year. The planned budget for the industrial pro-
gram will also be about the same, but there will be some changes within the
industrial area. We will be looking at those industries having toxic dis-
charges in more detail. There is also a need for additional work in assessing
the magnitude of air pollution problems and the state-of-the-art for control
of noncriteria and hazardous pollutants.
In ORD's recent "Report to Congress", which represents our first attempt
at presenting a five-year overview, I pointed out that there will need to be
a significant increase in expenditures for industrial pollution control
research, development, and demonstration efforts in the next few fiscal
years. This is required if we are truly serious about approaching the water
quality goals of the 1980's. If this increase does not happen, the much
needed control technology will not be developed and there will very likely
be a significant time delay in achieving our goals.
5
-------
As I mentioned earlier, EPA is aware of some of the unique problems
facing the food processing industry. That is why we have had a very active
R&D program in this area. There is a definite need to develop cost effective
waste management systems to comply with 1983 requirements and beyond. This
cannot be done exclusively through the development of more sophisticated
kinds of add-on treatment technology; we need and want to encourage additional
efforts on modifications in processing. In fact this very point was recently
stressed by our Administrator, Russell E. Train, in his "Year-End Report to
the Members of EPA." And, as an engineer, I must confess to a personal
distrust of add-on and retrofit solutions as optimal.
We are making progress but we still have the greatest challenge ahead of
us. Our country's commitment to environmental improvement and integrity
has been, and continues to be very real. With your continued support and
cooperation, I think we will be able to achieve our goal of a "quality environ-
ment" for everyone, and do it in consonance with the necessary production of
food and fiber.
Thank you.
-------
THE EPA REGION IV PROGRAMS
John A. Little, Deputy Regional Administrator
EPA, Region IV
Good morning and welcome to Atlanta—headquarters for the dogwood tree,
azaleas, and EPA, Region IV. I don't want to sound like a Chamber of
Commerce spokesman but, it's a fact, you couldn't have chosen a better time
to visit the city.
Two weeks ago, I and others from the Regional Office, accompanied Russell
Train to Tampa where we met the public at an environmental town hall meeting.
And I'm happy to report that people are still very much concerned about
environmental quality. Attendance estimates ran from 350 to 500. Many of
them stood throughout the entire three hour question and answer session.
Questions ran the gamut from Mirex to why one of our engineers allegedly
scheduled a meeting with a Baptist deacon on a Sunday afternoon.
There were a number of inquiries into the Agency's position on returnable
beverage containers. We were charged with wanting to "ban-the-can."
Several workers in the container manufacturing industry felt their jobs
were in jeopardy. Russ assured them that was not the case. We are not trying
to ban-the-can, or put anybody out of work. He stressed, and I would emphasize
here this morning, each of our regulatory actions is subject to the most
thorough, hardnosed economic analysis performed anywhere in government.
We do this to make sure our regulations are as sound as possible and to
demonstrate our good faith to the industries we regulate. It is our way
of avoiding as much disruption and conflict as possible. In short, it is
our way of trying to make sure our regulations are truly in the public interest.
Several years ago it was commonly believed that environmental goals
and economic progress were incompatible. A number of industry spokesmen
feared that compliance with environmental requirements would result in the
loss of millions of jobs and the forced closing of numerous plants. Recent
studies conducted by EPA and outside consultants have demonstrated that these
fears were greatly exaggerated. According to the most recent quarterly report
on economic dislocation resulting from environmental controls, only 75
plants have been closed during the past five years (January 1971 through
September 1975) as a result of environmental regulation. These closings,
many of which were only partially due to environmental factors—resulted
-------
in the loss of only 15,700 jobs—or sixteen thousandths of one percent of
the total U.S. labor force.
What is more important, these losses have been more than offset by the
creation of an entirely new industry—an industry devoted to the production,
installation and operation of anti-pollution equipment. A recent study by
a firm of Wall Street analysts for the council on environmental quality found
that environmental legislation has generated an industry employing over 1.1
million new workers. What this means is that, on balance, environmental
controls created seventy times more jobs than they destroyed. The message
is quite clear. Environmental protection is not only good for America.
It is also good for our National economy.
More than that, environmental protection is also the key to continued
industrial growth. Several years ago pollution levels had reached the point
where the public was unwilling to tolerate any increase until the backlog
of problems had been cleaned up. Not only the environment but also the public
had reached the saturation point. Public protest against growth at the expense
of the environment became a political force capable of blocking growth.
This makes clear a basic point: the greatest obstacle to growth is not pollu-
tion control, but the pollution itself. Thus, pollution control has become
the key to growth.
As you know, a key provision of the Federal Water Pollution Control
Act Amendments of 1972 was a stipulation that "It is the National goal that
the discharge of pollutants into the navigable waters be eliminated by
1985." As the key mechanism in attaining this goal Congress created a
nationwide federal water pollution abatement system for point sources
called the National Pollutant Discharge Elimination System, or NPDES for short.
To date, Region IV has issued 519 permits to canners and food processors
in the eight southeastern states. The categories include meat products
(red meats and poultry), dairy products, canned and processed fruits and
vegetables, seafoods, grain mill products, bakery products, sugar and
confectionery products, fats and oils, beverages, and miscellaneous preparations.
We have received 11,512 applications from all types of industry in the
Region and 8,095 permits have been issued. This represents an issuance
total of 70 percent. We feel Region IV is doing quite well in this regard.
And if we're not doing as well as we think we are, please let us know.
In closing, let me appeal to you to see that your own self interest
is best served by successfully solving environmental problems. Certainly
you can solve them better than anyone else. It is by now quite clear that
first-class pollution control systems can be installed without destroying
healthy, profitable business operations. And it also is increasingly clear
that sound environmental management is the key to future economic growth
now
The need for a strong commitment by American business leaders is greater
than ever before. The fact is that the environmental problems of this
8
-------
country are a long way from being solved and behind us. We have made a good
start toward cleaning up the most offensive conditions of pollution control
facilities and top-flight operation of those facilities will demand intensive
attention by top management throughout the next decade.
In addition to these major problems of air and water pollution, several
major environmental problems still remain to be solved. We are just beginning
to understand the subtle, chronic, adverse effects which chemical compounds
released into the environment may have on human health. We are now plunging
ahead with mammoth efforts to expand energy production and we will have to
grapple with all the problems of preventing environmental damage from those
efforts. Finally, all the complex questions of siting future industrial
plants and planning intelligently for economic growth are just now coming
into focus.
As America faces all these challenging problems. We have a real need
for constructive business leadership. The entire country has an enormous
stake in the successful resolution of these problems, but no one has a
greater stake in that success than you do. We in government need your help.
-------
WHY ARE WE HERE?
AN INDUSTRY OVERVIEW
By
Walter A. Mercer
Association Senior Scientist
National Canners Association
Berkeley, California
INTRODUCTION
In each life, I believe, there are moments of time in special situations,
which crystalize and become memory capsules of "happenings," - events
whose associated warmth and brightness become underlined reference points
in the book of memories which, in part, we compile each day of our lives.
For me, this is a "happening" to be remembered - if the words chosen convey
the intended message of this presentation, and the manner of their presenta-
tion arouses responsive thoughts - then, perhaps it can be noted as worth-
while.
My history of association with the birth and growth of these symposia
gave me the opportunity to speak for the 70-year-old Association of the food
canning industry. Sixty-three of those years measures the history of the
Industry's research efforts to improve its water and waste management prac-
tices while responding to its public responsibility to expand production
of safe and wholesome foods.
We are hopeful that this presentation will be in accord with the feelings
and hopes of all segments of the food processing industry. We meet for this
symposium, in the center of an area so important to all of us for its prod-
uction and processing of poultry, as well as other food commodities. We
must recognize that our future food supplies for the consumers of this
Nation - and to a great extent those of the World - depend on a prosperous
and productive agriculture.
At some time in the future - far, far in the future, we all hope -
each of us must return to Nature those elements which we borrowed, to form
the human package which is each of us. That one package - because we are
human - has two recognitions - that which we think we are, and another,
which those about us recognize as carrying our name. Regardless of personal,
or society's evaluation of our worth, in Nature's materials analysis scheme,
we are not much:
Water enough to fill a 10 gallon keg,
Fat for seven bars of soap,
Carbon for 9,000 writing pencils,
Phosphorus for 220 matches,
Magnesium for one laxative dose of salts,
Iron enough for one nail,
Enough lime to whitewash a chicken coop, and
Sulfur enough to rid one dog of fleas.
10
-------
To these and other materials essential to the structure and functions
of the human body, must now be added traces of chemicals not a part of our
original birthright - unavoidable because we did not in the past - nor
is it possible now, or in the forseeable future, to live in an environment
completely free of man-made organic and chemical substances.
Industry - Assigned Significance of This Symposium
On a day in November, 1968, in Berkeley, California, a discussion
group consisted of representatives from the food processing industry, various
universities engaged in food related research, and state and federal agencies
concerned with agricultural production and the environmental problems re-
sulting from the growing and processing of foods.
Other participants in this first, full-scale, two-day, meeting of the
National Canners Association's Technical Committee on Environmental Research,
were four staff members of the, then, Federal Water Pollution Control
Administration:
Mr. James Boydston and Mr. Kenneth Dostal from the Pacific Northwest
Regional Laboratory at Corvallis, Oregon.
Mr. George Keeler, Research and Development Grants Program, Washington,
D.C., and
Mr. William Pierce of the California - Nevada Basins Project Team.
In the years, which have followed, these and other EPA staff people
have participated in other food industry meetings. At that first meeting it
was my privilege, to outline a proposal previously endorsed by a number of
industry and university persons. Essentially, the proposal called for the
organization of an annual workshop, or similar-type forum, for presentation,
discussion, and distribution in printed form, of environmental research
results obtained by industry, government and universities.
The first fruit of the proposal, and its discussion, was the concept
of national symposia to be co-sponsored by government and industry. The-
First National Symposium on Food Processing Wastes came in April, 1970,
in Portland, Oregon.
James Boydston, Chief, Waste Treatment Research Program, in the Federal
Water Quality Administration's Pacific Northwest Water Laboratory, presented
the objectives of that first symposium. I quote, in part, from his remarks:
"This meeting beginning today marks a milestone, I think, in the
National program of research and development of water pollution
control methods for the food processing industry. The program
will, I hope, be the first of a continuing series of symposia -
to make current information available in open forum meetings of
11
-------
this type, and to give you an opportunity to question those experts
who are actually developing and demonstrating new and improved
treatment methods."
I continue to quote:
"This meeting marks a further milestone in that it truly represents
a cooperative, coordinated program between industry and govern-
ment, to mutually solve troublesome water pollution problems.
This cooperation is demonstrated by the (government-industry)
co-sponsorship of this meeting."
"Our research goals should include the refinement of conventional
methods of treatment; the development of processes capable of higher
degrees of treatment or completely closed-loop systems where the
treated effluent is reused within the processing plant; efforts
should be expanded on processing methods to reduce the quantity
of water required per ton of product; finally, to develop profitable
by-products from the wastes resulting from current processing
methods."
Mr. Boydston's closing remarks were prophetic and full of promise, I
quote:
"In closing let me say that we are faced with serious challenges:
we have a deteriorating environment and an aroused public. We know
a few answers and a lot of problems. But, I think that cooperation
such as shown here today between industry and government can make
the food processing industry a leader in the fight for clean water."
That First Symposium received an enthusiastic response from industry,
suppliers to the industry, universities, and government leaders. The
obvious result was the scheduling of the symposia on an annual and regional
basis.
Mr. Carey, speaking as President of the National Canners Association,
wishes me to express our appreciation and gratitude for the opportunity,
over these seven years, to be recognized as a co-sponsor of this National
Symposium on Food Processing Wastes. The intensive planning'which has
always gone into each Symposium has resulted in comprehensive coverage of
environmental research dealing with the many facets of water use and waste
generation in the production and preservation of foods of all kinds. In
looking backward, from this time, near the end of a career of research
efforts in a broad area of food industry problems, I recall with pleasure
and gratitude the many government-industry committee assignments and the
responsibilities for cooperatively-financed, government-industry research
efforts.
Associated with these bench-mark events of the past and present, are
the personal relationships and channels of communication formed with the
technical and scientific staff of the government activity now known as
the Environmental Protection Agency. My experiences reach back to the days
12
-------
when the U.S. Public Health Service was still responsible for water quality
and pollution abatement research. Members of the Association staff and
the food canning industry, which we represent, wish to express our sincere
appreciation for the unique and fruitful relationship out of which came
opportunities to assist in the development of water quality criteria and
the technical and economic data needed for establishment of environmental
control regulations expected to be reasonable and attainable, and to allow
economic survivial.
We are particularly grateful to that EPA team located in Corvallis,
Oregon and the Research and Development Grants Branch in Washington, D.C.
Many names could be recited in recognition of their unstinting efforts to
assist us in characterizing and accurately describing the unique complexity
and diversity of the food processing industry. Over the years, working
together on demonstration projects - or separately, but mutually interested
and concerned - we have developed technological and economic data for inplant
process modifications and innovations, which are now in use or will be used
if time is permitted for the necessary engineering and economic adjustments.
Of immeasurable value have been the contributions made by these and
other EPA scientists to our knowledge and research capabilities devoted
to the treatment and management of residual food materials. Government
printing and distribution of the symposia proceedings have provided a virtual
library of reference materials applicable in many ways to the technical
aspects of environmental problems.
We interpret a portion of the report of the National Commission on
Water Quality as recognition of the value of this EPA scientific input into
solving the environmental problems of low-profit margin industries in regional
areas. I quote from the Commission's report:
"Finally, the results of many (of the) studies point to a need
for technical advisory services to the localities and to small
industries. The compliance requirements are technically, legally,
and administratively complex. The planning requirements and their
long-term implications reach beyond the capacity of many small or
even moderate-sized communities to cope."
I continue to quote:
"The Commission is convinced that such a field advisory service,
operating out of regional offices of EPA would help localities,
states, and the Federal government save money. It would help to
expedite grant applications. It would help to assure that commu-
nities select the best long-range treatment option, in terms of
cost effectiveness and environmental quality. Properly staffed,
it could also be an essential information clearing-house to keep
localities and industries abreast of innovative technologies and
economically and environmentally preferable solutions to their
individual treatment problems."
13
-------
This recommendation and others, in the report, are completely in line
with my industry's experiences with EPA scientists - a rewarding and
profitable association in the best meaning of industry-government cooperation.
Mr. Carey wishes me to add this special message which I now quote:
"The lofty intent of the American public to protect our environment
was imperfectly translated into laws by the Congress. Those laws
are being imperfectly translated into regulations and the regula-
tions themselves are being imperfectly enforced. In too many cases
these successive distortions have resulted in actions which are
unrelated to the basic intent of environmental protection."
"You who have scientific and technical competence are the ones
who we must look to to keep all of the machinery properly directed.
While you may not control the legal machinery itself, you are the
advisers upon whom all must lean."
That is why we are here today.
What Will Happen to Food Production?
On Sunday, March 28, 1976, at some minute during that day, the World's
population reached four billion human individuals. More than one-third
of that population is under the age of 15 years.
Although the World's population is growing at a slightly slower rate
than our experts believed a year ago, many of those just under 15 will soon
reach their child-bearing years.
During the last 16 years, the World's population grew from three billion
to four billion. Only 13 more years will be required to reach a count of
five billion humans on Earth.
Each day, somewhere, babies are born to approximate 328,000 in numbers.
The estimated 133,000 deaths each day leave us with a net increase of 195,000
new humans - 135 new humans each minute of each day - 2.25 food-dependent
humans each second of each minute.
Who will provide the food, clothing, medical care, educational oppor-
tunities, the hope that anesthetizes the misery of hunger and illness?
In our United States, we have five percent of the World's population.
By mid-July we will have 215 million Americans - 86 times more than we had
200 years ago.
Today, this opening of the Seventh National Symposium on Food Processing
Wastes, marks the passing of an interval of time within which technical groups
of the various segments of the food industry have worked, increasingly, to
obtain, organize, and explain, for government teams, the complexities of
food processing operations: non-continuous seasonal operations, variabilities
14
-------
between processing of different products, and the differences in water use
and waste discharges for the same plant processing the same product. It
was our hope that detailing such complexities would enable the promulgators
of effluent control regulations to utilize the individual industry economic
evaluations and cost/benefit equations which were the intent of Congress.
In the meantime, and for many years in the past, the industry has moved
as rapidly as possible - considering its operational burden of more than
2000 federal food regulations - to establish the controls required to prevent
excessive and unneeded discharge of organic food residuals into the streams
of the Nation.
For mid-December, 1975, the Arthur D. Little Opinion Research Corporation,
summarized interviews with more than 1,000 persons over 15 years of age.
Responses from those interviewed were the basis for ranking public concern
about 30 national issues. Seventy percent of the persons interviewed ranked
the "high cost of food" as the number one national problem of most personal
concern. Inflation was placed third in the listing of personal concerns
for 62 percent of those interviewed.
Placed 10th in the listing of 30 national issues of concern was air
and water pollution. This matter of personal public concern was ranked
equally with "problems of the elderly" by 45 percent of the respondents.
Today, consciously or not, we, the people - the food consumers -
have thrown a challenge to the World's agriculture - double your production
in 25 years!
As previously indicated, the World's population is expected to double
before the end of this century - 24 years from now. Today, at the beginning
of each new Earth Day - almost 200,000 new mouths to feed.
Agriculture and food processing must grow, process and distribute almost
as much additional food as is now produced. If we meet the additional
need of improving the diet of the World's people, at the same time, World
agriculture must more than double its production - in one generation.
The last doubling of food production by the World's farmers1 required
35 years - now it must achieve a doubling in 25 years.
In one form or another, government takes one out of every three dollars
we earn as a nation of people. Even though, the bite is already that deep,
we as a people continue - or remain silent but concerned - when self-
appointed consumer protectors demand more government programs - leading
to federal, state and city budget deficits - and meddling in our private
business and personal affairs.
The public needs to recognize that a double-edged regulatory system
that outlaws insect and mold fragments in food, on the one hand and severely
restricts the development of pesticides on the other, is bound to reduce
agricultural production efficiency, increase food costs, and reduce the
quantity and quality of the food produced.
15
-------
Other Conflicts in Regulations
Unfortunately, other Federal agencies have brought on a proliferation
of regulations which are both conflicting and confusing. The result is
that the environmental problems of food production and processing are becoming
increasingly difficult and costly.
Examples include the following:
Aerators, to meet wastewater treatment standards proposed by EPA, which
generate noise exceeding OSHA standards;
Power-driven equipment required for EPA waste treatment compliance that uses
excessive amounts of energy (FEA).
Porous noise insulating surfaces (OSHA) that do not meet FDA and USDA
sanitation requirements, and
EPA proposed recycling of food processing waters to minimize liquid dis-
charges or reuse of treated wastewaters as opposed to FDA responsibility
for the safe use of such waters as they affect the wholesomeness of processed
foods.
Recently, Mr. S. Donald Wiley, Chairman of the Board of Directors,
National Canners Association made this eloquent plea, which I quote as
follows:
"To all who speak for the consumer - for that body of diverse
individuals we call the public - we extend an invitation to work
toward goals of environmental improvements - scientifically
determined to be needed - and demonstrably achievable goals -
which will allow the grower and processor of foods to survive
and to continue the increasing production of foods - not only
for ourselves but, whenever possible for those who need and must
have our grains and other foods to survive."
"The food processing industry - all segments - growers and pro-
cessors - that constitute the whole of Agriculture - ask only
that regulatory agencies, including EPA, require of us only that
which is truly in the best interest of, and desired by, the silent
majority of the people for who you speak and act - that you weigh
as prudently as possible, each action which you feel is a step
in the stairway of regulation on regulation - intended to reach
the pronounced goal of an eco-system restored to its virginal
purity."
Mr. Wiley concluded these remarks with the following:
"Any legislation, any regulatory procedure, which affects directly
or indirectly, the growth and processing of foods should be carefully
16
-------
weighed against the possibility that it may impede the production
of food or push its costs beyond the financial reach of any
significant segment of our people."
Because of the continuing, concerted effort to impose "humanistic"
science over objective science, books, and articles, and television dialogues
have dealt, emotionally, with chemicals in the environment. Without authentic
evidence they stand guilty, until proven innocent, of many of the pollution
ills and health disorders of current concern to the general public.
Food additives, animal feed additives, cosmetics; air, water and chemical
emissions from factories, the burning of fossil fuels - each and all have
been selected, at times for attack by consumer!sts and ecologists.
Extraordinary attention has been given to pesticides, plant growth
regulators, and chemical fertilizers. Groups and associations have been
formed to mount opposition to their use - opposition so furious and tenacious
as to be astonishing in its degree.
An now, to add the ultimate in confusion to that already created in the
public mind, we have the February, 1976, EPA presentation to the National
Press Club, included was the statement which I quote.
"Most Americans had no idea, until relatively recently that they
were living so dangerously. They had no idea that when they went
to work in the morning, or when they ate their breakfast - that
when they did the things they had to do to earn a living and keep
themselves alive and well - that when they did things as ordinary,
as innocent and as essential to life as eat, drink, breathe, or
touch, they could, in fact, be laying their lives on the line.
They had no idea that without their knowledge or consent, they were
often engaging in a grim game of chemical roulette whose results
they would not know until many years later."
End of quote.
Apparently, we are to expect that Mother Nature will give us abundance
and yet live up to some degree of virginal purity and allure which she never
possessed at any time since or before Earth became man's home.
The original Georgia Colony in the 13th Century almost perished from
epidemics of disease and poisonings caused by the foulness of the streams
polluted with the up-stream carcasses of dead animals and decaying plants.
There were no trout in the Missouri River that Lewis and Clark explored
all the way to today's Montana. The water was too muddy.
There were no game fish in the Colorado River when Coranado gave it its
name - because of the heavy load of red silt borne seaward by the river.
17
-------
Our forefathers took Nature as they found it - and were content with it.
They endured storms, droughts, and disease - caught fish when they would bite,
dug wells for drinking water, tossed their slops out through the back door -
and prayed for rain for their crops and protection from the hordes of pests
that came, as imagined punishment, for their transgressions.
What About Cancer and the Environment?
In May, 1974, the National Center for Health Statistics made the report
that death rates from heart disease, cancer, strokes, and six other diseases
had declined over the past 20 years - but nobody can say why.
Data were gathered from 33.6 million death certificates recorded in
the United States in the 1950's and 1960's. These data showed declines in
9 of the 15 leading causes of death among Americans and increases in 6
others - accidental deaths, suicides, homocides, and deaths from cirrhosis
of the liver were among those on the increase.
The American Cancer Society's Environmental Cancer Research Project began
in 1974 to attempt to answer questions about how environmental changes affect
people. Prominent in the search for data is confirmation of the statements
all edging that some 85 percent of cancer is derived from environmental causes -
causative agents in air, food and water.
In selected situations, there can be no serious argument that prolonged
exposure to lung irritants such as asbestos, coal dust, irritants in smoke
and certain volatile chemicals may cause lung cancer, there appears to be a
connection between long exposure to unusual concentrations of certain chemicals
and liver cancer for workers in the manufacturing process. Industrial workers
should and can be protected against such hazards.
A recent publication of the American Cancer Society revealed that in a
survey of cancer deaths around the world - with more than 20 countries
included - Scotland has the highest cancer death rate for men while Denmark and
Chile rank equally as number one in the highest rates for women.
On the other hand Nicaragua has the lowest cancer death rate for both men
and women. From the top of the list of about 22 countries, the United States
is in the 18th place, from the top, for men and 19th for women. As compared
to the number of deaths after diagnosis and treatment for cancer, since 1937,
there has been a 65 percent increase in the number of those who live beyond
the crucial 5-year period. But lung cancer is on the increase. I think we
know the causes. I am told that 75 percent of these cancers are preventable.
There is not one shred of evidence, or even a basis of reasonable suspicion
that any human damaging effects have ever been caused by any traces of
pesticides in foods consumed in the United States.
Certainly some defects have been observed in test animals fed or Injected
with exceedingly large amounts of some agricultural chemicals. But it is a
long, long unscientific step from the observation of the effects of such abnormal
18
-------
dosages in animal experiments to the promulgation of regulations banning the
use of any agricultural chemical which increases the quantity or the quality
of man's daily diet.
I am most encouraged when I read that as compared to the 37-year life span
of people in the United States in 1775, that today the average life expectency
is almost 75 years. As a Nation we are better fed and we have countless
opportunities our ancestors never dreamed possible.
Better sanitation, insecticides, and new drugs are credited with this
remarkable increase in man's expected average life span.
Only a few, apparently, want to go back to those days of hazardous living
in the absence of agricultural chemicals and drugs.
Cleaning the Water - At Last
Finally, I wish to summarize an editorial, entitled, "Cleaning the Water."
It appeared in the February, 1976 issue of Environmental Science and
Technology. The author is Russell F. Christman, Editor of the journal.
Recently, EPA has reported that the annual benefits to be gained from an
$11 billion water quality improvement cost will result in approximately one-half
that amount for recreational benefits. Around $2 billion of the total cost
will accrue to actual improvements in human health and aesthetic benefits.
An estimated cost of $375 billion will meet the requirements of P.L. 92-500
for capital construction costs for publicly-owned treatment plants. An additional
$134 billion will be borne by industry, for the construction of privately-
owned facilities, not including annual operating and maintenance costs.
Mr. Christman emphasizes that a water supply or waste treatment for an
industry or municipality can be unique to the individual situation. Optimal
use of financial resources, he states, are not likely to result from mandatory
and uniform national requirements.
Mr. Christman believes that these estimated social expenses required -
without a sharp definition of the kind and worth of the improvements to be
expected - constitutes a strong argument for a research approach to an objective
understanding of the program for which public spending to the point of national
and public sacrifice is claimed to be necessary. Mr. Christman makes this
important declaration, which I quote:
"It is clear that organized programs must further identify the nature,
sources, and mechanisms of the most serious threats to human health
and the technical alternatives to control them."
This clearly-voiced statement is an indictment of the apparent disregard of
government decision-makers for thorough evaluation of the public benefit/
public cost equation in arriving at environmental control requirements. The
outstanding objective appears to have been administrative simplicity.
19
-------
This does not mean to say that we have no environmental problems to solve.
The task before us will require the industrialist, government official, scientist,
and citizen - each must play an important role - but it must be a symphony
of effort in which all of the risks, the costs, and the benefits are known.
Our World, as we know it today, depends on this.
That is why we are here.
20
-------
A WASTEWAT1R TREATMENT STUDY
FOR SKOKOMISH SALMON PROCESSING PLANT***
by
Dr. S. S. Lin*, Dr. Paul B. Liao* and Max W. Cochrane**
INTRODUCTION
The Skokomish River supports a large Indian commercial fishery which
enables tribal fishermen to supplement their income and food supply
during the fishing season. With no means for preserving the fish other
than smoking, much of the harvest is marketed as fresh, which often
results in a weak market position for the fishermen and quality loss
in the fish.
A salmon processing plant was built to provide a more efficient system
for handling the fishery resource. The plant, designed by Bosworth
and Carroll, consists of a fish-preparation area, smokehouses, refrigeration
and freezing capacity, and a retail outlet. The plant markets fresh,
fresh-frozen, and smoked-fish products. Freezing capacity allows the
plant to smoke fish continuously throughout the year. The location and
layout of the salmon processing plant are depicted in Figures 1 and 2,
respectively.
Pre-bled salmon were used for processing in the past. However, fish
now are hand butchered at the plant. This increases blood and solids
in the waste flow and imposes a higher loading in the wastewater
treatment facility.
In 1973 EPA and the Washington State Department of Social and Health
Services informed the Skokomish Tribal Council that wastewater generated
at the salmon processing plant must be treated before discharge into
receiving waters in order to meet effluent-limitation standards. The
Council retained Kramer, Chin & Mayo, Inc. (KCM) to provide technical
*Kramer, Chin & Mayo, Inc., Consulting Engineers, Architects and Applied
Scientists, 1917 First Avenue, Seattle, Washington.
**Environmental Protection Agency, Corvallis, Oregon.
***This project was sponsored by the Environmental Protection Agency,
under Grant No. 803911, and the Skokomish Indian Tribal Council,
Shelton, Washington.
21
-------
LOCATION MAP
VICINITY MAP
FIGURE 1 LOCATION AND VICINITY MAPS
SKOKOMISH PROCESSING PLANT
SHELTON, WASHINGTON
22
-------
WASHWATER
1
FRESH FISH.
DRESSED
SORTED
BY WEIGHT
FREEZING
BRINE
ro
co
I
SMOKING
PACKING
SHIPPING
11 FRESH SALMON
21 FROZEN SALMON
31 SMOKED SALMON
FIGURE 2 PROCESS LAYOUT OF SKOKOMISH
SALMON PROCESSING PLANT
SHELTON, WASHINGTON
-------
services and to design a waste-treatment facility for the processing
plant. Construction of the treatment facility was completed, and has
been in operation since July 1975. The treatment facility consists of
an extended-aeration system and two identical aerobic polishing ponds,
as shown in Figure 3. Table 1 summarizes the criteria used in the
design of the wastewater treatment facility. Treated effluent is
disposed of in the drainfield through a 6-inch perforated plastic pipe.
TABLE 1. SUMMARY OF DESIGN, WASTEWATER TREATMENT FACILITIES
Waste Characteristics
Source = Process water, no sanitary waste
Minimum Flow = 0 gpd
Maximum Flow = 20,000 gpd
Peak Flow =18 gpm
Daily Flow Variation = Continuous over approximately 18 hours
Average BOD Concentrations - 500 mg/1
Average Suspended Solids Concentration =100 mg/1
Maximum BOD Loading = 80 Ib/day
Extended Aeration System
Aeration Tank: Volume = 20,000 gallons
Detention Time at Maximum Flow = 24 hours
Maximum Oxygenation Available « 240 Ib 02/day
Maximum Loading » 0.03 Ib BOD/cu ft/day
Clarifier: Volume - 3,000 gallons
Detention Time at Maximum Flow =2.7 hours
Surface Area - 98 sq ft
Overflow Rate at Maximum Flow = 202 gpd/sq ft
Polishing Ponds
Surface Area « 0.37 acre
Loading at Maximum Flow = 21.5 Ib BOD/acre/day
Volume - 364,000 gallons
Detention Time at Maximum Flow » 18 days
Design of the treatment facility was based upon literature review and
daily grab samples for the determination of waste characteristics. The
literature review was limited since very little has been published about
the characteristics of salmon"processing waste. Thus, the reliability
of biological systems for salmon-processing waste-control is subject to
verification and substantiation. A water-quality monitoring program
was begun in September 1975 to evaluate the performance of the newly
installed treatment facility. This study is scheduled for completion
in March 1977. Results of this study will be analyzed 1) to determine
24
-------
ro
in
SALMON
PROCESSING
PLANT
PUMP STATION
DIVERSION
BOX
6" PLASTIC I PERFORATED I
DRAIN PIPE
WASTEWATER
PACKAGE PLANT
FIGURE 3. FLOW SCHEME OF THE WASTE
TREATMENT SYSTEM
SKOKOMISH PROCESSING PLANT
SHELTON. WASHINGTON
-------
if the control efficiencies of the system are adequate to meet waste
effluent limitations and 2) to determine if the EPA effluent guidelines
are realistically achievable. Also, the results will be used to develop
engineering criteria for future design. This paper presents an evalua-
tion of the operation data obtained from September 1975 to February
1976. An evaluation of the entire study will be presented in an EPA
publication following the completion of the project in March 1977.
CHARACTERISTICS AND TREATABILITY OF SALMON PROCESSING WASTE
Only the wastewater generated at a fresh-frozen salmon processing plant
will be discussed because the Skokomish salmon processing plant does
not include the canning process. The fresh-frozen salmon process is
essentially the same throughout the industry. The only major factors
affecting the waste characteristics are the geographic location and
the size of the plant. Generally, the processing of Pacific salmon
as a fresh or frozen commodity is considered to have smaller waste loads
and waste flows than the canning segment of the salmon industry.(1)
Very little has been published about the characteristics of waste
generated by salmon processing. Possibly the most reliable data for
salmon-processing waste was provided by a seafood-waste survey(1) of
six plants in three areas of Alaska and one area of the Northwest.
Table 2 lists summary statistics of the waste loads from all hand-
butchered salmon processes studied during the survey. These results
can be used to determine the typical raw-waste loadings resulting from
fresh-frozen or hand-butchered salmon-canning processes in both Alaska
and the West Coast when characteristics of the wastewater are not known.
Biological treatment of seafood-processing wastes has not been fully
practiced in U.S. seafood industries except at a small, pilot, blue
crab processing plant in Maryland and at full-scale systems at two
shrimp processing plants in Florida.'"' However, sufficient nutrients
are available in most seafood wastewaters, indicating that such waste-
waters are amenable to aerobic biological treatment.
The Federal Water Quality Control Administration believes that seafood
wastewater can be treated by domestic sewage-treatment methods with
some waste strength adjustment.(2) in the Current Practices Report(3)
the need for testing and developing optimum operational characteristics
is outlined.
In a report issued in 1972, Riddle^ ' studied the efficiency of
biological systems for smelt and perch wastewater. He found a 90%
removal of unfiltered BOD5 after 10 days of aeration, and 90% removal
of filtered BODs after 2 days aeration in a batch reactor. Tests
in a continuous reactor showed that maximum BODs removal (80% soluble
and 45% unfiltered) could be achieved with a 7.5-hour detention time,
sludge recycling and a 3-day sludge age, or a 5-day detention time
with no sludge recycling.
26
-------
TABLE 2. HAND-BUTCHERED SALMON PROCESS SUMMARY*
Parameter
Production, ton/hr
Process Time, hr/day
Flow, I/sec
(gal/min)
Flow Ratio, 1/kkg
(gal /ton)
Sett. Solids, ml/1
Ratio, 1/kkg
Sol. Solids, mg/1
Ratio, kg/kkg
Sus. Solids, mg/1
Ratio, kg/kkg
5-Day BOD, mg/1
Ratio, kg/kkg
COD, mg/1
Ratio, kg/kkg
Grease/Oil, mg/1
Ratio, kg/kkg
Organic-N, mg/1
Ragio, kg/kkg
Ammonia-N, mg/1
Ratio, kg/kkg
pH
Temp. , Deg. C.
Mean
2.13
6.34
2.15
34.1
5,040
1,210
1.02
5.15
193
0.971
236
1.19
493
2.48
1,070
5.36
341
1.72
80.9
0.407
2.12
0.011
6.73
13.2
Std. Dev.
1.09
1.80
1.09
17.2
3,100
743
1.19
5.99
155
0.782
185
0.933
179
0.900
601
3.03
628
3.16
40.0
0.202
0.794
0.004
0.318
2.51
5% Min.
0.733
3.67
0.754
11.9
1,410
338
0.109
0.547
37.5
0.189
47.6
0.240
233
1.17
332
1.67
15.0
0.076
29.0
0.146
0.979
0.005
6.25
9.19
95% Max.
4.38
8.98
4.39
77.5
13,000
3,120
4.18
20.5
600
3.12
722
3.54
923
4.35
2,600
13.1
1,770
8.39
181
0.314
4.14
0.120
7.13
15.7
*From Environmental Associates, Inc., *'Draft - Canned and Preserved Fish
and Seafoods Processing Industry," February 1974.
In a later study, Robbins reported that an activated-sludge plant
in Japan has been designed especially for fish-waste treatment. The
wastewater flow is approximately 0.27 mgd, and the 5-day BOD concentra-
tion ranges from 1,000 to 1,900 mg/1. The results of pilot-plant
studies using a 10-hour aeration time, and the organic and hydraulic
loadings, are listed in Table 3. Bulking occurred when the organic
loading rate exceeded 0.31 Ib/cu ft/day.
27
-------
TABLE 3. ACTIVATED SLUDGE PLANT RESULTS*
FISH PROCESSING WASTES
Parameter
^H^^HBAVHHBBHBA|HHIIHVIBIIHvHIIIBIIIBIIBBBBailviBVillwl^iHin
BOD5 (mg/1)
%
Removal
Raw
Waste 0.075
1 ,000 5
99.5
BOD Loading
0.
mmm^Hm •^•••M. !!••!•
10
99.
14
^^(•••••••HHIBIII—
0
(Ib/cu
0.
HPHpHHHiBaHHmpHgBqHHHaiggt,,,,,
13
98.
ft /day)
21
.»^«M»«miM~«<
7
0.
27
97.
26
3
^Obtained from Environmental Associates, Inc., ''Draft - Canned and
Preserved Fish and Seafoods Processing Industry,*' February 1974.
The EPA effluent limitations guidelines for salmon processing plants
are presented in Table 4. Based upon the weight of fish processed,
the recommended effluent limitations for each processing plant can then
be determined.
TABLE 4. RECOMMENDED EFFLUENT LIMITATIONS GUIDELINES FOR SALMON
PROCESSING PLANTS(1)
Maximum 30-Day Average
Parameter (kg/kkg) (Ib/ton)
5-Day BOD 3.2 6.4
Total Suspended Solids 2.0 4.0
Grease and Oil 4.9 9.8
*0nly for West Coast hand-butchered salmon.
METHODS
Operation
Various operational conditions at the Skokomish salmon-processing waste-
treatment facility are examined in this report and are summarized in
the following paragraphs.
28
-------
1. Test Schemes
Two test schemes for the study are depicted in Figure 4. Scheme A
indicates that extended aeration and aerobic ponds are in series, but
the aerobic ponds are in parallel. Similarly, Scheme B shows that
extended aeration and two aerobic ponds are in series.
2. Extended Aeration System
Two major parameters, dissolved oxygen (DO) content and food-to-
microorganism (F/M) ratio in the aeration tank, can be controlled during
the study period. The DO content can be varied by adjusting impeller
Immersion. The F/M ratio (BOD/MLVSS) can be varied by adjusting MLVSS
(mixed liquor volatile suspended solids) concentration, which is controlled
by regulating the sludge recycling rate and the quantity of sludge wasted.
3. Aerobic Ponds
Two aerobic ponds are used to polish the effluent out of the extended
aeration system. Surface area and retention time are the key factors
affecting treatment efficiencies of aerobic ponds. Both factors can be
varied by adjusting water depth.
Sampling
Eight sampling stations for both schemes are shown in Figure 4. The
locations of each sampling station for both schemes are identical and
are described as follows:
Sampling Stations Location
a Incoming raw wastewater
b Aeration tank
c Return sludge and excess sludge pipe
d Effluent of the extended aeration process
e Aerobic Fond No. 1
f Aerobic Pond No. 2
g Effluent of Aerobic Pond No. 1
h Effluent of Aerobic Pond No. 2
Both grab and composite samples collected in the field were preserved,qv
In accordance with Methods for Chemical Analysis of Water and Wastes.
The daily and weekly sampling schedules are shown in Tables 5 and 6,
respectively. The operational schedule for the entire study is presented
in the Detailed Project Plan, which was submitted to the U.S. Environmental
Protection Agency in July 1975.
Testing
All chemical and biological tests were analyzed in accordance with
Standard Methods.^ Field tests included flow measurement, pH, DO,
and temperature. Laboratory tests conducted in the KCM Laboratory were
BOD, COD, SS, VSS, TS, grease and oil, TKN, alkalinity, turbidity,
ortho-P, total-P, MLSS and MLVSS.
29
-------
AJ
RAW
WASTE
EXTENDED AERATION
"AEROBIC
POND NO. 1
AEROBIC
POND NO. 2
SCHEME A
AEROBIC
POND NO
?u©r
irl
EXTENDED AERATION
SCHEME B
AEROBIC
POND NO. 2
FIGURE A PROCESS LAYOUTS.
30
-------
TABLE 5. DAILY SAMPLING SCHEDULE
Sampling
Stations Sampling Frequency Tests
a Grab Samples/2 hour Flow, Temp, pH, DO
Eight-hour Composite BOD, COD, SS, VSS, TS, Grease/Oil*, TKN,
Sample Ortho-P, Total-P, Alaklinity, Turbidity
b One Grab Sample Daily MLSS, MLVSS, DO, Alk., pH, Temp.
c One Grab Sample Daily SS, VSS, Flow
d Same as (a) Same as (a) (Flow measurement is not
necessary)
e One Grab Sample Daily DO, pH, Temp.
£ Same as (e) Same as (e)
g Same as (d) Same as (d)
h Same as (d) Same as (d)
*0ne grab sample for sampling day.
TABLE 6. WEEKLY SAMPLING SCHEDULE
Week Monday Tuesday Wednesday Thursday Friday
1st S S
2nd S
3rd S S
4th S
Note: S indicates the day for sampling.
31
-------
It should be noted that this paper serves as an interim report since
the entire testing schedule has not been completed. Due to the low
flow entering the treatment facility, only one aerobic pond was in
operation during this phase of the study. Because of the small
quantity of extended aeration effluent reaching the pond and pond
seepage, sampling at Stations E through H has not commenced. This
paper emphasizes the evaluation of the operation of an extended
aeration system. However, the operation of aerobic ponds also will
be evaluated as the project progresses.
RESULTS AND DISCUSSION
The aeration tank of the extended aeration system was seeded before
the monitoring program was started in the beginning of September 1975.
One thousand gallons of returned activated sludge with a suspended
solids concentration of approximately 4,500 mg/1 was obtained from
the Renton Treatment Plant, Renton, Washington, and fed into the
aeration tank for seeding.
Water meters were installed for continuous recording of water
consumption during fish processing. The used water was then pumped
to the extended aeration system for treatment. Water utilized for
sanitary purposes was transported into a septic tank for treatment.
The duration of each pumping period also was recorded. Using pumping
duration and rate, the flow rate entering the treatment system was
then computed. In order to record flow rate between the extended
aeration system and the aerobic pond, a weir box was installed in a
corner of the aerobic pond. Flow rate was computed from a continuous
record of the depth of wastewater over the V-notch in the weir box.
Wastewater generated from fish processing was screened by a fine
screen installed in the drain entering the pump station.
As indicated previously, this paper presents the evaluation of
operational data obtained only from September 1975 to February 1976.
The intent of this portion of the study period was to monitor the
performance of the extended aeration system with a minimum control.
The daily operational conditions during this five-month period are
discussed in the following.
The monitoring program was started on September 12, 1975, but composite
sampling was not taken until September 17, 1975. Large salmon were
processed from September 12 to December 5, 1975. Large salmon herein
refers to an average weight of 10 pounds per salmon. Due to the
unexpected difficulty in obtaining fish for processing, no fresh fish
were processed from December 6, 1975 to January 30, 1976. Repacking
frozen fish, which produced no wastewater, was the only activity
occurring during that interval.
In order to maintain the performance of waste-treatment systems during
the time fish-processing wastes were unavailable, Purina Trout Chow was
fed into the treatment system. This fish food was added from January 16,
1975 to January 30, 1976. Approximately 9.5 pounds of fish food were
32
-------
dissolved into 560 gallons of water, and the solution was then fed into
the pump station.
Processing of fresh fish was resumed on February 2, 1976, when small
salmon became available. The processing of small salmon is expected
to continue for several more months. Small salmon herein refers to an
average weight of 11 ounces per fish. Variations which affected waste-
water production during this period are summarized as follows:
Intervals Processing Activities
9/12/75 to 12/5/75 Large salmon
12/6/75 to 1/15/76 No activity except repacking
1/16/76 to 1/30/76 Fish food addition
2/2/76 to 2/29/76 Small salmon
The following presents the results of different processing activities
during the five-month study period. iX
Wastewater Characteristics
The characteristics of wastewater generated from the salmon processing
plant varied daily. An attempt was made to correlate the waste flow
and pollutant strength with total weight of fish processed. In order
to generalize and reflect the fluctuation of operational data, average,
standard deviation and ranges were utilized. Wastewater characteristics
for large salmon processing, small salmon processing and fish food
addition are summarized in Tables 7, 8 and 9, respectively.
It can be seen that wastewater characteristics during fish food addition
were comparable to those during salmon processing, and the performance
of the treatment system was maintained when receiving no fresh-fish
processing wastewater. Approximately 20% of total weight of fish
processed was wasted. The wastage has been hauled away to sanitary
landfill for disposal. Comparison of Tables 7 and 8 indicates that
flow and pollutants generated per ton of fish processed for small salmon
is greater than that for large salmon. In other words, less flow and
fewer pollutants were generated during large-salmon processing.
High variations of wastewater characteristics were reflected in high
ratios of standard deviation to average. It is believed that the
fluctuation of wastewater characteristics may be reduced as more data
become available.
Treatment Efficiencies
Similarly, average and standard deviation were used to present various
treatment efficiencies for each parameter. Treatment efficiencies for
wastewater produced by large salmon processing, small salmon processing
and fish food addition are summarized in Tables 10, 11 and 12, respectively.
33
-------
TABLE 7. WASTEWATER CHARACTERISTICS FOR LARGE SALMON PROCESSING
Parameter
Flow, gpd
Process Time, hr/day
Fish Processed, Ib/day
Processing Wastage, Ib/day
Gallon of Wastewater per
Ton of Fish Processed
Turbidity, JTU
PH
DO, mg/1
Temp. , Deg. F.
Alk. , mg/1 as CaC03
BOD (total), mg/1
BOD (total), Ib/ton
GOD (total), mg/1
COD (total), Ib/ton
BOD/COD (total), mg/1
BOD/COD (total), Ib/ton
SS, mg/1
SS, Ib/ton
VSS, mg/1
VSS, Ib/ton
VSS/SS, mg/1
VSS/SS, Ib/ton
TS, mg/1
TS, Ib/ton
Grease and Oil, mg/1
Grease and Oil, Ib/ton
TKN, mg/1
TKN, Ib/ton
NH4+-N, mg/1
NH4+-N, Ib/ton
Total -P, mg/1
Total -P, Ib/ton
Ortho-P, mg/1
Ortho-P, Ib/ton
Average
691
6.5
3,153
747
723
100
7.21
7.88
51.5
128.5
687
3.26
2,057
10.51
0.328
0.328
502
2.5
265
1.3
0.541
0.541
1,638
7.47
283
1.46
207
0.86
6.04
0.0367
0.8
0.0061
0.34
0.00212
Standard
Deviation
237
1.0
2,658
614
599
49
0.13
1.04
3.6
88
445
2.5
1,120
8.6
0.105
0.105
224
1.38
147
0.86
0.175
0.175
861
2.76
142
0.93
197
0.49
7.25
0.0434
0.48
0.0083
0.16
0.0022
Range
405-1,215
5.5-8.0
324-9,970
188-1,733
240-2,500
24-205
6.9-7.7
6.8-11.9
38-58
10-388
79-1,708
0.92-8.4
595-4,864
2.64-34.14
0.133-0.546
0.133-0.546
250-950
0.83-6.25
100-700
0.46-4.17
0.278-0.778
0.278-0.778
575-2,525
3.28-12.38
24-541
0.078-3.53
2-850
0.006-2.16
1.2-26
0.0039-0.161
0.3-1.75
0.0011-0.0271
0.1-0.6
0.00024-0.00834
34
-------
TABLE 8. WASTEWATER CHARACTERISTICS FOR
Parameter
Flow, gpd
Process Time, hr/day
Fish Processed, Ib/day
Processing Wastage, Ib/day
Gallon of Wastewater per
Ton of Fish Processed
Turbidity, JTU
pH
DO, mg/1
Temp. , Deg. F.
Alk. , mg/1 as CaC(>3
BOD (soluble), mg/1
BOD (soluble), Ib/ton
COD (soluble), mg/1
COD (soluble), Ib/ton
BOD/COD (soluble), mg/1
BOD/COD (soluble), Ib/ton
BOD (total), mg/1
BOD (total), Ib/ton
COD (total), mg/1
COD (total), Ib/ton
BOD/COD (total), mg/1
BOD/COD (total), Ib/ton
SS, mg/1
SS, Ib/ton
VSS, mg/1
VSS, Ib/ton
VSS/SS, mg/1
VSS/SS, Ib/ton
TS, mg/1
TS, Ib/ton
Grease and Oil, mg/1
Grease and Oil, Ib/ton
TKN, mg/1
TKN, Ib/ton
NH4+-N, mg/1
NH4+-N, Ib/ton
Total-P, mg/1
Total-P, Ib/ton
Ortho-P, mg/1
Ortho-P, Ib/ton
Average
3,631
5.4
2,557
498
3,000
74
7.23
6.93
45.5
58.5
452
9.5
626
12.49
0.732
0.732
484
11.58
814
21.18
0.581
0.581
234
6.53
139
3.92
0.582
0.582
888
23.08
177
4.34
65
2.81
2.42
0.077
1.84
0.038
0.3
0.0078
SMALL SALMON PROCESSING
Standard
Deviation
959
1.3
472
49
1,420
54
0.12
2.20
1.65
9.35
103
2.8
146
1.14
0.175
0.175
194
3.5
201
9.49
0.153
0.153
90
4.51
67
2.91
0.151
0.582
230
10.2
55
1.19
64
4.94
0.94
0.064
0.74
0.016
0.2
0.005
Range
2,044-5,154
4.0-8.0
1,520-3,500
400-500
2,040-8,080
30-180
7.1-7.4
4.0-11.3
42-48
44-70
335-528
7.24-12.57
521-793
11.26-13.52
0.62-0.934
0.62-0.934
211-726
7.76-16.74
650-1,104
13.57-39.50
0.36-0.81
0.36-0.81
134-370
2.90-14.83
48-244
1.04-8.76
0.358-0.803
0.358-0.803
600-1,155
12.97-41.25
96-260
3.11-6.47
23-191
0.54-12.87
0.9-3.7
0.015-0.202
1.2-2.9
0.02-0.06
0.2-0.7
0.0034-0.0145
35
-------
TABLE 9. WASTEWATER CHARACTERISTICS FOR FISH FOOD ADDITION
Parameter
Flow, gpd
Fish Food, Ib/day
Turbidity, JTU
PH
Temp., Deg. F.
Alk., mg/1 as CaC03
BOD (total), mg/1
COD (total), mg/1
BOD/COD (total), mg/1
SS, mg/1
VSS, mg/1
VSS/SS
TS, mg/1
Grease and Oil, mg/1
TKN. mg/1
NH+-N, mg/1
Total-P, mg/1
Ortho-P, mg/1
Average
1,022
9.5
91.5
6.9
45
59
400
688
0.587
354
200
0.652
525
199
99
2.48
1.46
0.33
Standard
Deviation
1,017
0
32.4
0.217
1.58
13.8
56.8
131
0.064
134
60.9
0.355
219
203
10.6
0.80
0.39
0.12
Range
550-2,841
9.5-9.5
51.5-137
6.7-7.2
43-47
44-74
319-458
551-833
0.540-0.698
160-485
133-279
0.300-1.25
250-775
75-560
82-110
1.7-3.8
0.8-1.8
0.2-0.5
TABLE 10. EXTENDED AERATION TREATMENT
PROCESSING
Parameter
DO, mg/1
Retention Time, days
PH
SVI
BOD/MLVSS
MLVSS/MLSS
Overflow Rate, gpd /ft
BOD, %
COD, %
SS, %
VSS, %
TS, %
Grease and Oil, %
TKN, %
NH4+-N, %
Total-P, %
Ortho-P, %
WASTEWATER
Average
8.3
31
7.3
103
0.06
0 516/978
2 7.2
91
88
62
66
45
69
83
72
28
37
EFFICIENCIES FOR
Standard
Deviation
1.4
8
0.97
25
0.06
174/218
2.3
8
5
18
20
19
19
16
24
3
25
LARGE SALMON
Range
6.2-10.2
16.4-49
5.8-8.4
67-155
0.01-0.09
225/525-850/1,300
4.1-12.4
67-98
75-96
24-94
33-93
10-74
25-91
48-99
8-95
25-33
0-75
36
-------
TABLE 11. EXTENDED AERATION TREATMENT EFFICIENCIES FOR SMALL SALMON
PROCESSING WASTEWATER
Parameter
DO, mg/1
Retention Time, days
PH
SVI
BOD/MLVSS
MLVSS/MLSS 2
Overflow Rate, gpd/ft
BOD, %
COD, %
SS, %
VSS, %
TS, %
Grease and Oil, %
TKN, %
NH4+-N, %
Total-P, %
Ortho-P, %
Average
9.2
6.2
7.4
118
0.084
1029/1693
36.5
86 (90)*
83 (90)*
63
54
60
60
48
52
N/At
N/At
Standard
Deviation
1.6
2.1
0.22
41
0.036
313/468
13.8
8 (7)*
8 (5)*
30
29
7
22
35
35
N/At
N/At
Range
7.9-10.9
3-9.8
7.1-7.7
67-192
0.046-0.1414
550/850-1400/2200
20.9-62.7
79-98 (82-97)*
67-91 (84-94)*
5-84
6-80
52-67
17-74
0-94
4-62
N/At
N/At
*Soluble portion
tN/A indicates that effluent concentrations were either greater than or
equal to influent
TABLE 12. EXTENDED AERATION TREATMENT EFFICIENCIES DURING FISH FOOD
ADDITION
Parameter
DO, mg/1
Retention Time, days
PH
SVI
BOD/MLVSS
MLVSS/MLSS 2
Overflow Rate, gpd/ft
BOD, %
COD, %
SS, %
VSS, %
TS, %
Grease and Oil, %
TKN, %
NH4+-N, %
Total-P, %
Ortho-P, %
Average
9.7
29.6
7.4
97
0.039
540/1306
10.6
88
85
78
75
54
61
92
21
N/At
7
Standard
Deviation
1.4
12.7
0.26
47
0.039
114/93
10.8
7
6
12
6
10
16
0.5
17
N/At
N/At
Range
7.6-10.2
7.0-36.4
7.1-7.7
111-133
0.015-0.108
400/1200-700/1400
5.6-30.0
76-97
77-91
64-93
67-81
43-67
47-88
92-93
0-38
N/At
0-20
tN/A indicates that effluent
equal to influent
concentrations were either greater than or
37
-------
Lower ratios of standard deviation to average indicated that treatment
efficiencies were comparably consistent. Low wastewater generated
from the processing plants resulted in longer retention time, high
dissolved oxygen content and lower overflow rate. It is anticipated
that these operation conditions will improve when more fish are being
processed. The treatment efficiencies during fish food addition were
comparable to those during salmon processing. This finding demonstrated
that the addition of fish food could maintain the performance of the
treatment system when no salmon processing wastewater was available.
Average BOD removal efficiencies were more than 85%, which is within
the range for a biological secondary-treatment system. However, removal
efficiencies for SS were less than 70%, indicating the poor settleability
of solids in the clarifier. Poor settleability of solids (SS, VSS, and
TS) resulted from longer retention time and overaeration. The removal
efficiency of SS could be improved with the increase in wastewater flow.
Both ortho and total phosphorus contents in the effluent were either
greater than or equal to those for the influent during most of the
five-month study period. This could be due to overaeration resulting
from longer retention time. It was reported(10) that extended aeration
resulted in the depletion of carbon source, which is very likely to
cause the release of phosphorus in the aeration tank. Shorter retention
time would certainly minimize the problem.
Oxygen Uptake Rate v
Oxygen uptake for mixed liquor suspended solids was tested in the field
by using a YSI DO probe. The oxygen uptake rates for different dates
were computed and tabulated. Only the summary of the test results is
presented below.
Average 1.56 mg/l/hr
Standard Deviation 0.92 mg/l/hr
Range 0.5-4.2 mg/l/hr
No conclusion could be drawn based upon the available data. More tests
should be conducted as the project progresses in order to correlate
oxygen uptake rate with retention time, BOD/MLVSS, etc.
Sludge Filterability
Sludge disposal is an important factor in the operation of wastewater
treatment systems. So far, no sludge has been wasted. To facilitate
analytical work, a sludge filterability test was conducted to indicate
sludge characteristics. One hundred milliliters (100 ml) of sludge was
filtered under a pressure of 21 psi for 80 minutes, and solid concentra-
tions were measured at 0, 20, 40, 60 and 80 minutes. For this study,
an attempt was made to compare the percent of SS increase to that
obtained from conventional activated-sludge process treating of domestic
wastewater.('1) The comparison, as shown below, demonstrated that sludge
filterability for the Skokomish study was similar to that for conventional
38
-------
activated sludge. More data will be compiled as the project progresses
to substantiate this finding.
Skokomish Study
Activated Sludge
EPA Effluent Limitations
Filtering
Pressure Time Percentage of
(psi) (min) SS Increase
21 20 495
5 20 115
Effluent quality in terms of BOD, SS and grease and oil was checked
against the recommended EPA effluent limitations.(1) Comparison of
extended aeration effluent quality with EPA effluent limitations is
shown in Table 13. It can be seen that effluent quality of extended
aeration met all the recommended limitations. Based upon available
results, it can be concluded that the system is capable of meeting
EPA effluent limitations.
TABLE 13. COMPARISON OF EXTENDED AERATION EFFLUENT QUALITY WITH EPA
EFFLUENT LIMITATIONS
Average Weight of
Fish Processed
(Ib/day)
EPA Effluent
Limitations
(Ib/day)
Extended Aeration
Effluent Quality
(Ib/day)
Average Range
Large Salmon
BOD, Ib/day
SS, Ib/day
Grease and Oil,
Ib/day
3,153
3,153
3,153
10.1
6.3
14.7
0.4
1.5
0.7
0.03-1.3
0.3-3.0
0.1-1.2
Small Salmon
BOD, Ib/day 2,557
SS, Ib/day 2,557
Grease and Oil, Ib/day 2,557
8.2
5.1
15.4
2.1
3.1
2.2
0.2-3.5
0.6-10.7
1.1-4.1
39
-------
CONCLUSIONS
From the results obtained during the five-month study period, the
following conclusions may be made:
1. Small salmon processing generates more wastewater flow and higher
pollutant contents per unit of fish processed than does large salmon
processing.
2. Fish food addition was found capable of maintaining performance
of the wastewater treatment system when no wastewater was received from
the salmon processing plant.
3. Both longer retention time and overaeration result in poor
settleability in the clarifier and release of phosphorus in the aeration
tank. The latter consequently causes a high phosphorus content in the
effluent.
4. Sludge filterability for the extended aeration system was found
similar to that for conventional activated sludge.
5. The treatment system is capable of producing an effluent which
meets the EPA effluent limitations in terms of BOD, SS, and grease and
oil.
40
-------
1. ENVIRONMENTAL ASSOCIATES, INC. FOR U.S.E.P.A. Draft development
document for effluent limitations guidelines and standards of
performance - canned and preserved fish and seafoods processing
industry. February 1974.
2. U.S. DEPARTMENT OF INTERIOR, FEDERAL WATER CONTROL ADMINISTRATION,
ROBERT S. KERR WATER RESEARCH CENTER. Seafood waste water, Westwego,
Louisiana. April 1968.
3. SODERQUIST, M.R., et al. Current practice in seafood processing
waste treatment. Project 12060 ECF of Environmental Protection
Agency, Water Quality Office, 1970.
4. JOHNSON, E. L. and PENIS^TON, Q. P. Pollution abatement and by-
product recovery in the shellfish industry. National Symposium
on Food Processing Wastes • 1971, EPA and National Canners
Association, 1971.
5. JENSON, C. L. Industrial wastes from seafood plants in the state
of Alaska. Proceedings, 20th Industrial Waste Conference, Purdue
University, Engineering Extension Series No. 118, 1965.
6. MATUSKY, F. E., et al. Preliminary process design and treatability
studies of fish processing wastes. Proceedings, 20th Industrial
Waste Conference, Purdue University, Engineering Extension Series
No. 118, 1965.
7. RIDDLE, M. J., et al. An effluent study of a fresh water fish
processing plant. Water Pollution Control Directorate Reprint
EPT G-WP-721, Canada, 1972.
8. AMERICAN PUBLIC HEALTH ASSOCIATION, Standard methods, water and
wastewater. 13th Ed., 1971.
9. U.S. ENVIRONMENTAL PROTECTION AGENCY, OFFICE OF TECHNOLOGY TRANSFER,
Manual of methods of chemical analysis of water and wastes.
Washington, D.C., 1974.
10. SEKIKAWA, Y., et al. Release of soluble ortho-phosphate in the
activated sludge process. Kurita Central Laboratories, Yokohama,
Japan.
11. UN, S.S. Phosphate removal by the addition of aluminum (III) to
the activated sludge process. Ph.D. thesis, Department of Civil
Engineering, University of Washington, Seattle, 1972.
41
-------
RECLAMATION AND TREATMENT OF CLAM WASH WATER
by
R. R. Zall*, L. F. Hood*, W. J. Jewell**,
R. L. Conway*, and M. S. Swltzenbaum**
INTRODUCTION
The surf clam is a relatively recent addition to the Atlantic Coast shellfish
processing industry. The major industry growth occurred in the mid 1940's and
in 1958, when new offshore beds near Long Island were discovered. Animal
presence, coupled with improved harvesting methods, brought about substantial
increases to seafood production in New York State marine fisheries. Today
annual amounts of surf clam meats from New York approximate 2,000 metric tons
of finished product or about 7% of the total landings in the United States.
When growth occurs and plants enlarge to handle increased volumes, waste
generated from food processing activities becomes troublesome and difficult to
handle. The organic material wasted from clam processing sites place a burden
on sewage systems. If discharged without treatment, it probably pollutes
waters near the Coast where these types of plants operate.
Research Objectives and Plan
A team of professors and graduate students from the Department of Food
Science and Agricultural Engineering at Cornell University have been studying
the reclamation of protein and flavor material from clam wash water. The
purpose of the project was to assist the clam processing industry in develop-
ing methods to create salable products from sewage where the economics of
waste handling might be turned into a credit rather than debit to the industry.
After studying the problems at a single site, the Cornell group drew up a
research plan with the following objectives:
(1) Develop process flows to reduce water consumption and BOD discharge loads.
(2) Develop edible products from proteins and flavor materials in clam
mincing wash water.
(3) Develop methods that a seafood processing plant might use to reduce the
costs of sewage disposal.
(4) Define composition of the solids and flavor components in wash water
together with chemical reactions responsible for flavor deterioration in
wash waters.
*Department of Food Science, Cornell University, Ithaca, New York
**Department of Agricultural Engineering, Cornell University, Ithaca, New York
42
-------
This paper summarizes one phase of the study and partially reports the
progress of the ongoing project.
A Single Site Study
The research group formed an affiliation with a well-known Long Island surf
clam processing company. During 1975, five multi-day trips were made to the
plant in January, February, May, July, and November for the purpose of
obtaining samples and observing plant operations. In between periods were
utilized for evaluation of site sampling materials in laboratories and pilot
plants at Cornell's Ithaca campus.
The thrust of our work was to:
(1) Fully understand plant operations in order to recommend sound production
changes to reduce water consumption.
(2) Determine if seasonal variations occur in the chemical and micro-
biological properties of wash water and clam products.
(3) Conduct a bench top waste treatment study of waste flows generated from
plant operations.
(4) Use different processing methods to treat rinse waters with existing
plant equipment and with experimental equipment moved from Ithaca to
Long Island.
Mode of Operations ;
The plant is one of the twelve clam processing plants in the New York-New
Jersey-Maryland area. It is located on eastern Long Island and faces a small
harbor on two sides. The site occupies a land area of approximately31.5
acres. The processing building, refrigerated storage, ice making sheds,'and
garages are located on the site. Loading and unloading areas are strategical-
ly placed to handle raw receipts and shipments of finished goods.
A flow chart of plant operations is shown in Figure 1. Surf clams are
delivered to the plant by truck and transferred to refrigerated storage in
32-bushel capacity tote bins until utilized.
At processing time, tote bins are moved with forklift trucks to the dumping
table where the clams are dumped into a 190°F water tank and held for about
one minute. This operation removes surface silt and a black membrane above
the opening of the clam shell. In addition, the hot wash makes the clam
easier to handle by the shuckers. Incoming clams are transferred by belt
conveyor to a shucking station where up to 30 people open shells and harvest
clam meats. At this point shells and bellies are separated from edible clam
meats and sent to waste on an out-of-plant discharge belt conveying system.
Waste portions are separated outside the building into different fractions
like bellies and shells which are then disposed of via dump truck, garbage
43
-------
SilRF CLAM
RECEIVING
w
IHOT WATER BATH
- - 'SHUCKING LINE!
2:1
UJ
>-i
CO
-------
boat, or direct to the harbor.
The clam meat is taken by the shucker via gallon buckets to a dump table in a
wash area where the material is washed by a mechanical spray-cleaning rotating
chamber (Station 1).
At first glance, the shucking process may seem somewhat primitive, but it is
a workable method by which a shucker is graded and compensated for individual
productivity.
Rinsed meat is manually sorted on a slow moving belt by several inspectors at
a second table and washed again through a second stainless steel spray-
tumbling device similar to the Station 1 washer.
At this point, washed clams are transferred via plastic pails to a mincer
device that resembles a large meat grinder but without cutting knives in the
discharge port. Mincing is supposed to take out the last residues of sand and
tenderize the meat (Station 3)-
From this point, meat is either packed for fresh sales, frozen in trays for
frozen sales, or cooked in broth for canning purposes as shown in Figure 1.
Broth is made in steam kettles and then canned. Filled cans are further
processed in retorts, cooled, marked, and cased for truck load shipments to
sales outlets.
Water Usage and Waste
The plant uses potable water obtained from a town water system and returns all
liquid wastes to a municipal waste treatment plant system. Water cost is
40c/l,000 gallons plus 48
-------
TABLE 1. FLOW RATES AT WASHING STATIONS
Station
1
2
3
Rate
(gal/min)
60
10
5
Where measured
Outside settling basin
Copper drain pipe from bath
Outside settling tank
TABLE 2. DAILY WATER USAGE
Time
7-9 a.m.
9-10 a.m.
10-11 a.m.
11-12 a.m.
12- 1 p.m.
1- 2 p.m.
2- 3 p.m.
3- 4 p.m.
Total /day
1/8/75
12,000
5,900
5,000
4,300
2,000
5,300
4,900
2,800
38,200
Water usage
2/10/75
~Q 1
gaj.
11,850
3.200
4,000
6,000
2,000
5,000
5,100
900
38,050
2/11/75
12,000
2.900
4,900
3,800
2,400
4,200
4,800
4,400
39,400
Wastewater Characteristics
Wastewater was sampled at hourly intervals during operating days. Composite
samples of each station were prepared from individual hourly samples based
upon flow rates so as to form a representative sampling picture of the day's
actual operations.
Tables 3, 4, 5, and 6 show pertinent biological-chemical data needed by our
group to help evaluate the relative food values of these waste flows.
46
-------
TABLE 3. BOD AND COD ANALYSIS OF COMPOSITES FROM WASHING STATIONS
FEBRUARY 1975
Station
1
2
3
BOD
COD
mg/1
234 680
558 937
2340
TABLE 4. PROTEIN AND SOLIDS CONTENT OF HOURLY SAMPLES OF WASH WATER FROM
STATION 3 - FEBRUARY AND MAY 1975
Protein
Total solids
Protein as %
of total solids
Sample time
9:00 a.m.
9:30 a.m.
10:00 a.m.
10:30 a.m.
11:00 a.m.
11:30 a.m.
1:00 p.m.
1:30 p.m.
2:00 p.m.
2:30 p.m.
3:00 p.m.
3:30 p.m.
February
_a
0.17
0.11
0.24
0.36
0.14
0.08
0.06
0.10
0.10
0.06
0.09
May
0.69
-
0.46
-
0.51
0.42
-
0.08
0.36
-
0.30
0.36
February
%__
0.73
0.60
0.65
0.86
0.97
0.22
0.23
0.41
0.35
0.31
0.58
May
1.4
-
1.0
-
1.2
0.9
-
0.3
0.7
-
0.4
0.6
February
_
23
18
37
42
14
36
26
24
29
19
16
May
49
-
46
-
42
46
-
28
51
-
76
60
- = no sample drawn at this time
47
-------
TABLE 5. PROTEIN AND TOTAL SOLIDS IN COMPOSITE SAMPLES OF WASH WATER
Station
Station
Month
Protein
Protein as %
Total solids of total solids
1
2
3
January
February
January
February
January
February
0.15
0.06
0.16
0.09
1.07
0.33
%___
0.08
0.08
0.74
0.35
1.70
0.33
77
22
41
63
94
TABLE 6. CALCIUM, PHOSPHORUS, AND CHLORIDE CONCENTRATIONS IN COMPOSITE
SAMPLES OF CLAM WASH WATER
Month
Calcium
Phosphorus
Chloride
1
2
3
January
February
January
February
January
February
60
65
57
61
72
44
6
40
26
62
230
109
207
6
280
7
540
7
Microbial Quality of Clam Meat and Wash Water
The possibility of seasonal variations in wash water quality seemed probable
with changes in water temperature. In order to learn what these effects might
be, the following data were collected.
Tables 7, 8, and 9 show typical bacteriological data of the wash water and
clam meat during different seasons at the final station before packaging or
cooking.
48
-------
TABLE 7. STANDARD PLATE
COUNT OF CLAM AND WASH WATER AT STATION #
Month of sampling
Sample type
Wash water
Clam surface
Meat ground
in sterile
water
Clam solids,
washed with
sterile water
Time drawn
8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
3.
- = no sample drawn at
February
_a
14,000
32,000
60,000
-
-
-
28,000
130,000
50,000
-
-
10,000
32,000
18,000
-
-
6,300
18,000
13,000
-
-
this time
May
Jt/mT —
480,000
162,000
472,000
640,000
1,420,000
930,000
,
10,000,000
7,000,000
7,000,000
4,000,000
9,000,000
5,000,000
„.
8,000,000
3,000,000
6,000,000
4,000,000
9,000,000
5,000,000
...
4,000,000
1,000,000
300,000
220,000
370,000
101,000
July
107,000
86,000
83,000
94,000
109,000
98,000
204,000
188,000
153,000
192,000
191,000
167,000
101,000
128,000
76,000
131,000
127,000
72,000
42,000
42,000
29,000
33,300
12,000
14,700
November
242,000
167,000
200,000
125,000
-
-
395,000
1,300,000
228,000
152,000
-
-
340,000
550,000
76,000
176,000
-
-
18,000
42,000
43,000
61,000
-
-
49
-------
TABLE 8. COL I FORM COUNT
OF CLAM AND WASH WATER AT STATION 3
Month of sampling
Sample type
Wash water
Clam surface
Meat ground
in sterile
water
Clam solids,
washed with
sterile water
a - = no sample
Time drawn
8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
drawn at
February
_a
10
70
80
-
-
-
20
50
20
-
-
-
10
10
10
-
-
-
10
10
20
-
-
this time
May
_ ill ml
— — ____ ——-jf / mj.—
269
23
24
12
47
32
._ it I0
ff/g
248
70
309
42
13
43
.________* /0—_
yj g
237
62
167
20
20
38
& im-\
33
17
43
15
12
8
July
90
80
70
40
60
70
160
20
100
90
30
10
20
40
30
10
10
10
30
0
130
50
30
30
November
153
67
33
97
-
-
100
90
70
50
-
-
10
0
0
20
-
-
0
10
10
20
-
-
50
-------
TABLE 9. YEAST AND MOLD
Sample type Time drawn
Wash water 8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
Clam surface 8:30 a.m.
9:30 a.m.
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
Meat ground 8:30 a.m.
in sterile
water 9:3° a'm-
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
Clam solids, 8:30 a.m.
washed with .. ,n
sterile water 9:3° a'm'
10:30 a.m.
11:30 a.m.
2:00 p.m.
3:00 p.m.
a - = no sample drawn at
COUNT OF CLAM
February
" " a
20
20
30
-
-
_
10
10
10
-
-
10
10
10
-
-
20
10
10
-
-
this time
AND WASH
WATER AT STATION 3
.
Month of sampling
May
8
2
6
192
183
108
25
30
7
13
16
16
23
28
14
12
26
14
16
7
5
10
14
11
July
460
420
290
570
490
390
* i0
ff/g
540
400
590
630
520
480
A /_
— if/g
370
290
170
320
120
200
jt i—i
420
210
200
370
470
320
November
170
225
251
44
-
-
350
620
210
200
-
-
20
, 20-
30-
115
-
-
_
90
i 30
40
-
-
51
-------
Besides looking at the wash water and the clam surface, the tables also give
microbiological counts of meat washed with sterile water and then meat ground
in sterile water. The purpose of the analysis was to see if microorganisms
penetrated the meat interiors.
Table 10 shows that potable water temperature varies from 40° - 68°F.
TABLE 10. SEASONAL CHANGES IN WATER TEMPERATURE AT PLANT
Month Temperature (°F)
January 48
February 46
May 62
July 68
November 56
CONVERSION OF WASH WATER TO A CANNED JUICE PRODUCT
The research team trained itself to evaluate the flavor components in Station
3 wash water. Preliminary tasting sessions were used to identify flavor
characteristics peculiar to clam wash water. Comments such as fishy, bitter,
astringent, unclean, and musty were noted. (See Figures 2 and 3.)
The clam juice product being marketed by the firm was used as the reference
sample. This commodity is made by cooking clams in water, adding salt,
canning, and retorting. The product is used in institutions to make soups and
broths. It is a sweet, slight greenish-yellow, mild clam-flavored liquid.
Analysis of different commercial batches of juice showed that its composition
varied. Average composition was 3.5% total solids, 2.0% salt, and 1.5%
protein.
Because wash water at Station 3 was a dilute protein solution, it was neces-
sary to concentrate the liquid two- to three-fold to bring its solids
composition up to levels approximating those present in the marketed clam
juice.
Different concentration techniques were used to increase the total solids
content: (a) ultrafiltration, (b) open boiling, and (c) vacuum evaporation.
Materials were processed, canned, and retorted under different test conditions
and evaluated for flavor characteristics after different storage times.
Table 11 shows the composition of products concentrated by the different
methods.
52
-------
Directions: You h
and a
Taste
again
then
Tasting Session I
FLAVOR EVALUATION OF CLAM WASH WATER
ave before you 10 samples of processed clam wash water
sample of canned clam juice "R".
the reference sample "R" first and evaluate samples 1-10
st it. Samples 1-6 are unsalted so taste these first and
proceed to 7-10, the salted samples.
Flavor attribute Intensity ratine
Clam flavor
1
2
3
4
5
6
7
8
9
10
Much
less
-5
-4
Fish flavor
1
2
3
4
5
6
7
8
9
10
Mod.
less
-3
-2
Slightly
less
_i
Refer-
ence
0
Slightly
more
1
Mod.
more
2
4
Much
more
5
Figure 2. Taste Panel Work Form Comparing Selected Wash Water Flavors with a
Commercial Clam Juice,
53
-------
Tasting Session II
FLAVOR EVALUATION OF CLAM WASHWATER
Directions; As you taste the following 10 samples, please jot down a
few descriptive words which characterize the flavor of
these samples.
Please base your comments on the sample with respect to "R".
Sample
Description
,9
10
* Feel free to make comparisons such as, "#3 is more clammy than #4".
Figure 3. A Taste Panel Work Form to Describe a Flavor Profile of Clam Wash
Water
54
-------
TABLE 11. SELECTED CHARACTERISTICS OF CONCENTRATED SAMPLES OF WASH WATER FROM
STATION 3
Sample type
Clam juice
Juice from fresh clams
Ultrafiltration - p.m. shift
No salt, retorted
No salt, unretorted
Ultrafiltration - a.m. shift
No salt, retorted
No salt, unretorted
Salt, retorted
Salt, unretorted
Open boiling - a.m. shift
No salt, retorted
No salt, unretorted
Salt, retorted
Salt, unretorted
Vacuum evaporation - a.m. shift
No salt, retorted, 80°C
Salt, retorted, 80 °C
No salt, retorted, 50°C
Salt, retorted, 50° C
Protein
1.36
2.47
2.31
2.50
0.49
0.53
1.45
1.57
1.35
1.47
2.34
2.25
-
-
-
-
Total solids
%_
3.6
6.4
3.1
3.15
0.7
0.8
2.6
2.8
3.1
3.15
5.4
5.0
0.9
3.4
0.8
3.0
Protein as %
of total solids
37.78
38.59
74.19
79.36
70.00
66.25
55.62
56.14
43.55
46.67
43.33
45.00
-
-
-
-
The following process was developed to convert clam wash water into clam juice
which is now marketed by the company:
(1) Wash water from mincing pumped directly into a steam-jacketed kettle.
(2) Boiled for 30-40 minutes to concentrate and remove objectionable
volatiles.
(3) Concentrated juice strained to remove coagulated clam proteins.
55
-------
(4) Salt added.
(5) Product canned and retorted.
The finished product is equivalent to conventional clam juice made by boiling
clams in water in a kettle cooking-type operation.
Opportunities to Decrease Waste Volumes
Station 3 wash water contributed about half of the total organic load wasted
to the sewer in this plant operation. The removal of nutrients from the
waste flows would appreciably diminish waste treatment costs from this type
of an operation.
Station 1, as indicated in Tables 1, 2, and 3, uses 80% of the total water
and contributed about half of the total BOD load.
Water and waste flow continuously even though the clam meat which needs wash-
ing is batch produced by buckets of clams from a manual shucking operation.
Existing washing equipment can clean more than three times the amount of
clams now being sent through the system. If the operator at Station 1 would
operate the wash water only when clams traverse the tumbling spray system, it
would be possible to cut water consumption about 60%. In addition, wash water
spent in Station 2 could be used to pre-rinse clams at a dump table prior to
wash Station 1. Fresh water is now used to prewash clams before the Station 1
washing equipment.
By combining water economies from Station 2 with intermittent flow rather than
continual water use in Station 1, the total water used for both stations could
be reduced from about 70 gpm to 30 gpm. The drop in water consumption for six
operating hours would be 6 hours x 60 minutes/hr x 40 gpm, or 14,400 gallons
per day- This is half the daily volume now utilized.
Station 3 would generate little to no waste but would increase clam juice
volume by using the 5 gpm as product rather than waste as shown in Table 1.
WASTE TREATABILITY STUDY
Concurrent to the emphasis placed upon in-house methods to reduce waste, some
of the group carried out a waste treatability study. The objectives of this
phase of the project were as follows:
(1) Characterize the clam processing waste water as concentrated waste at
Station 3 and as composite waste of the three stations mixed in the
12-2-1 ratio.
(2) Determine if these kinds of waste are amenable to simple aerobic biolog-
ical treatment.
(3) Learn if usable food products would be captured from clam processing
56
-------
waste water through physical-chemical treatments such as a coagulation-
sedimentation system.
Project Approach
Waste samples were collected separately from the three stations in the
processing line; the first two being the washing operation with the third a
combination wash and mince operation. Samples were immediately frozen and
transported to Cornell in 20 gallon pails and stored in a -30°F walk-in
freezer. Composite mixtures were prepared by mixing appropriate amounts of
waste in a 12-2-1 ratio. Station 3 was further treated separately with
special flocculents to learn if organics therein might be economically
separated into more concentrated fractions.
Coagulation-Sedimentation Tests
Jar tests were used to determine optimal coagulation doses in conjunction
with appropriate turbidity measuring instruments.
Once optimal dosages were determined, the effluents were characterized by
parameters such as: TKN, COD, and total solids. Along with coagulation-
sedimentation tests, various combinations of centrifugation and filtration
were employed. The coagulants used in a series of tests included: tri-valent
solids, alum and ferric chloride, a synthetic polyelectrolyte (Dow A 23), a
cellulose gum (Herculus Type 7HF), and a natural polymer chitosan. The
results of these tests are presented in Tables 12, 13, and 14.
TABLE 12. SUMMARY OF OPTIMAL DOSAGES
Coagulant Concentration Turbidity reduction
mg/1 %
Composite waste
FeCl3 25 78
Alum 50 75
Chitosan 5 81
Dow A-23 1 62
Concentration waste (Station 3)
FeCl3 80 75
Alum 140 77
Chitosan 10 65
Dow A-23 7 66
Hercules cellulose gum 10 76
57
-------
TABLE 13. COMPOSITE WASTE RESULTS
Test
TS
Reduction
TKN
COD Turbidity
Centrifugation alone
Coagulation alone
Alum
Chitosan
Blank
Coagulation-Centrifugation
8.8
2.8
2.8
2.8
1.3
13.3
6.6
6.6
14.6
0.0
0.0
2.8
2.8
15.7
0.0
58.3
50.0
76.6
80.0
33.3
Alum
FeCl3
Chitosan
Blank
9.8
9.6
11.4
7-5
14.6
14.6
16.0
2.6
2.8
25.7
37.1
0.0
63.3
80.0
86.6
58.3
These tests indicate that clam processing waste does not appear to be
especially amenable to coagulation-sedimentation treatment. What did come to
light in determing the optimal coagulant dosages for the five coagulants used
in the study was that substantial turbidity reduction occurred with little
reduction in TKN, TS, or COD in dilute or concentrated wastes. We can only
conclude from these data that opportunities for chemical-physical treatment
do not appear promising.
Table 12, 13, and 14 show optimal dosage data from jar tests. Note the
percentage reductions in turbidity with little improvement in organic
composition. Chitosan, a shell waste product itself, exhibits interesting
turbidity reduction characteristics.
58
-------
TABLE 14. CONCENTRATED WASTE RESULTS
Test
Reduction
TS
TKN
COD Turbidity
Filtration alone
Centrifugation alone
Coagulation alone
Alum
FeCl3
Chitosan
Blank
Coagulation-Centrifugation
6.9
7.8
2.7
3.6
3.9
0.6
5-7
1.7
5.3
3.3
5.7
0.3
10.9
17.7
15.0
9.6
16.4
1.3
73.3
53.3
64.0
46.0
88.6
13.3
Alum
FeCl3
Chitosan
Blank
8.5
8.4
11.2
7.9
7.1
5.9
7.6
1.7
19.1
24.6
46.6
17.7
75.3
71.3
94.0
53.3
BIOLOGICAL TREATMENT
A bench scale study was carried out with liquid waste collected from the
single site clam processing plant to look at its biotreatability character-
istics. Fmir laboratory size aerobic lagoons were designed and operated at
volume retention times of 2.5, 5, 10, and 15 days. These lagoons were
operated on a draw and fill basis, daily, at a constant temperature of 20°C.
The tanks were sized as follows: (a) 10 liters to simulate a 15-day lagoon,
(b) the other three containers were 5 liters each. Aeration and mixing of the
liquids were achieved by using an air supply through a diffusion stone.
The units were seeded with a mixed liquid obtained from a nearby sewage
treatment plant. Feed for the units, clam processing wastes, was obtained
from composite samples obtained at the factory side and frozen in one-week
portion amounts for use during the testing period. No additional nutrients
were added to the feed since analysis of clam processing wastes showed the
material was not deficient in nitrogen or phosphorous. However, sodium
carbonate was added daily at the rate of 250 mg/1 food to maintain adequate
buffering capacity.
The units were fed and wasted daily for a period of three solid retention
times (SRT) before steady state data were taken. At steady state the pH
values of all four units remained relatively constant and the following
59
-------
parameters were monitored for mixed liquors: total chemical oxygen demand
(COD), total solids (TS), total volatile solids (TVS), suspended solids (SS),
total Kjeldahl nitrogen (NH -N), ammonia as nitrogen (NH,-N), nitrite (N02~N),
nitrate (NOg-N), pH, and temperature. The effluent analysis included SS,
total COD, and biochemical oxygen demand (BOD).
In these kinds of treatment, the principal objectives of an aerobic biological
process is the removal of colloidal and dissolved organic compounds by micro-
bial metabolism. Bacterial cultures in a continuously mixed and aerated
reactor utilize the soluble organic substrate for energy and growth.
Theoretically the end products of this aerobic process are carbon dioxide,
water, and microbial mass which is then removed by some kind of physical
process. An equation is shown to illustrate this process:
CNHa°b + N °Y + °2 bacteria C5H7°2N + C02 * H2°
Small amounts of solubilized phosphorous and nitrogen compounds are needed as
nutrients to keep the growth steady. Table 15 shows pertinent waste
characteristics of material used to feed the four systems.
TABLE 15. WASTE CHARACTERISTICS
_
Concentrated waste Composite waste
Analysis (Station 3) (Stations 1, 2, 3 mixed in 12:2:1 ratio)
~~~mg/l——~————~———————
Total COD 3590 837
Soluble COD 3672 614
BOD 2452 463
TS 5380 2528
TVS 4420 1204
SS 97 161
VSS 90 138
TKN 282 113
NH3-N 39.2 9.3
Alk 10 as CaC03 54.2 as CaCO
P04 150 40
TDS 2830 1695
* values are average of the nine food portions
60
-------
Chemical oxygen demand (COD) removal efficiencies are listed in Table 16.
TABLE 16. COD REMOVAL EFFICIENCIES
Date
8/1
8/4
8/6
8/7
8/13
8/14
8/15
8/18
8/20
8/21
Average
Solids retention
2.5 days 5 days
94.5
94.5
95.8
95.8
96.6
91.9
89.7 93.4
91.3
93.2
94.4
93.0 94.1
time (SRT)
10 days
96.4
97.8
97.9
96.9
91.2
91.2
89.9
94.4
15 days
98.9
98.9
98.9
98.9
95.1
94.9
94.0
97.0
Table 17 through Table 20 summarize the mixed liquor and effluent
characteristics of the reactors in the 2.5-, 5-, 10-, and 15-day
experiments.
It was found that greater than 90% COD reductions were achieved at a
hydraulic retention time of 2.5 days or more, and total nitrification occurred
in a retention period of five days.
The results of the study indicate that clam processing waste appears to be
readily amenable to aerobic biological treatment. Like operations might
consider variations of this kind of treatment attractive for on-site
pretreatment action where necessary.
61
-------
IV)
TABLE 17. 5 DAYS
RETENTION WITH AEROBIC TREATMENT - AUGUST 1975
Mixed liquor
Date
8/1
8/4
8/6
8/7
8/14
8/15
8/20
TS
1920
2155
2455
2270
2235
2210
SS
712
480
376
470
480
479
618
COD
538
456
440
432
534
518
543
TKN NH--N NO.-N NO.-N
J £* J
__/1 _
nig /i —
62 0 0 72
55 0 0 65
53 0 0 54
50 0 0 61
40 0 0 63
pH
6.
6.
6.
6.
6.
6.
6.
7
7
1
9
8
8
5
VSS
612
412
312
412
420
413
482
Effluent
SS
28
28
26
26
69
40
37
COD
40
40
32
32
75
61
82
BOD
14
7
26
19
30
10
-------
OJ
TABLE 18. 10 DAYS RETENTION WITH AEROBIC TREATMENT - AUGUST 1975
Mixed liquor
Date
8/1
8/4
8/6
8/7
8/14
8/15
8/20
TS SS COD
2015 618 489
2085 434 408
2365 416 416
2525 320 350
2306 494
422 445
2080 446 461
TKN NH -N NO -N NO -N pH VSS
J ** J
mg/
45 0 0 83 7.0 506
50 0 0 85 7.0 346
50 0 0 58 6.0 370
38 0 0 6.9 252
43 0 0 70 6.7
6.5 346
6.5 384
Effluent
SS
32
30
32
48
56
57
82
COD
g/ J-
24
16
16
24
82
82
124
BOD
10
25
19
26
30
-------
01
TABLE 19
Date
8/1
8/4
8/6
8/7
8/14
8/15
8/20
. 15 DAYS RETENTION WITH AEROBIC TREATMENT - AUGUST 1975
TS SS
1975 382
2015 422
2570 398-
2140 358
2165 440
386
2105 482
COD
408
399
432
367
432
424
465
Mixed liquor
TKN NH3-N N02-N NO^N pH
_ mo /I
—lug/ JL— — — — — — — —
49 0 0 80 6.9
49 0 0 88 6.9
54 0 0 83 6.0
45 0 0 6.7
42 0 0 83
6.5
VSS
332
346
334
308
370
336
396
SS
8
24
8
10
34
18
Effluent
COD
— mg/1 —
8
8
8
8
45
47
40 73
BOD
5
16
11
20
20
-------
cr>
en
TABLE 20.
Date
8/13
8/14
8/15
8/18
8/21
2.5 DAYS RETENTION WITH AEROBIC TREATMENT - AUGUST
Mixed liquor
TS SS COD TKN NI^-N NC^-N NC^-N pH
_ _ _ inn /I _
1952 515 216 66 5.6 32 20
1995 524 588 57 0 37 18
492 600 7.2
2207 470 530 56 5.6 31 14 7.1
2540 848 750 71 7.7 37 11 7.0
1975
vss
476
424
398
408
523
/
Effluent
SS COD BOD
96 61
46 39
28 96
64 81 38
70 68 62
-------
ACKNOWLEDGMENTS
This study describes one phase of an ongoing research project being sponsored
by the New York Sea Grant Institute under a grant from the Office of Sea
Grant, National Oceanic and Atmospheric Administration (NOAA), U. S. Department
of Commerce.
The authors are appreciative of the support provided by the following
individuals: John Plock, Sr., John Flock, Jr., II Joo Cho, Yongkeum Joh,
David P. Brown, William Walters, and J. H. Martin.
Finally the authors appreciate the support provided by Dr. Robert C. Baker
and other administration officers in the New York State College of Agriculture
and Life Sciences of Cornell University.
66
-------
PILOT PLANT PRODUCTION OF A FUNCTIONAL PROTEIN
FROM FISH WASTE BY ENZYMATIC DIGESTION*
G. 0. Bucove** and G. M. Pigott
**
INTRODUCTION
For several years, researchers of the Institute for Food Science and Technol-
ogy at the University of Washington have been investigating methods of uti-
lizing both marine food processing waste and fish stocks not normally used
for human consumption. The primary goal of this research has been to produce,
from waste and underutilized species, a product of higher nutritional and
economic value than ordinary fish meal, while minimizing pollution. The goal
has not been to solve an existing pollution problem, but to upgrade an exist-
ing use of raw material without creating a problem.
Of the approximately 70 million metric tons of fish caught annually worldwide,
between 30 and 40% is used directly for animal feed.'-'-) Much of this fish is
from species too small, too boney, or too oily for direct human use, and is
reduced to fish meal. Fish meal is a good animal food supplement, but loses
nutritional value during direct flame drying and has serious odor and liquid
waste problems.
Of the remainder of the world catch, which is processed in one way or another
for human food, up to 65% becomes a by-product or waste. A salmon fillet line,
for example, uses only about 35% of the whole fish. Canning lines are more
efficient, utilizing up to 65% of the whole f ish. ™) At present, much of this
waste becomes serious local pollution or, at best, is reduced to a low grade
animal feed. This does not have to be the case, since the non-edible or waste
portion (viscera, head, filleted frame, etc.) is very similar in terms of
proximate analysis to the edible portion, as seen in Table 1.
*This investigation is supported by funds from Sea Grant Contract No.
04-5-158-48.
**Respectively Graduate Research Assistant and Professor, Institute for Food
Science and Technology, College of Fisheries, University of Washington,
Seattle, Washington 98195.
67
-------
TABLE 1. PROXIMATE COMPOSITION OF A TYPICAL TRAWL-CAUGHT
Constituent
Water
Lipid
Protein
Ash
Whole
Fish
89.9%
3.5%
12.7%
2.7%
Edible
Portion
83.6%
0.8%
15.2%
1.1%
Non-edible
portion
81.2%
4.4%
11.7%
3.5%
Over the last ten to twenty years, there have been numerous proposed methods
to utilize fish processing wastes and underutilized species to produce a
high grade animal or human food product - usually called Fish Protein Con-
centrate or F.P.C. Few have been successful. The problem is great: to
produce an inexpensive, low lipid, highly nutritious product, free of fish
odor and taste, with long-term storage stability not requiring refrigeration
or extensive packaging. Most of the conventional fish protein concentrate
methods use an organic solvent extraction to remove lipid. These methods
entail complex, expensive engineering and have difficult problems with sol-
vent recovery and solvent-contaminated product and waste streams. The
F.P.C. from solvent extraction is furthermore a non-functional protein. It
cannot be rehydrated with water and therefore has a limited potential as a
food additive.(4) Many of the other methods have never successfully left the
bench laboratory stage due to economic and engineering scale-up problems.^)
This paper presents our ongoing research on a biological F.P.C. method devel-
oped to produce a functional, soluble, high-quality protein product, which
could have immediate marketability as a milk solids substitute in animal feed
formulations and a good potential as a human food additive.(6) we expect to
present a complete engineering report on the process this August at the First
International Congress on Engineering and Food at Boston, Massachusetts.
PILOT PLANT PROCESS
Figure 1 presents a flow chart of the enzyme digest F.P.C. process. A brief
overview here will be followed by detailed descriptions of the various steps.
First, the raw material is mechanically deboned. The skin and bones can be
dumped overboard if processed at sea, or reduced to meal or fertilizer ashore.
The fish flesh is homogenized with water and placed into the enzyme reactor.
68
-------
WHOLE FISH or FISH WASTE
water
acid
enzyme
NaOH
ANION
CONVERSION
Salt
H^O
v
DEBONER
HOMOGENIZER
ENZYME REACTOR
CENTRIFUGE
ANION EXCHANGE
REACTOR
CENTRIFUGE
CONCENTRATER
SPRAY DRYER
skin
bones
sludge
oil
H20
v
PACKAGE
Figure 1. Flow Chart - Enzyme Digest Fish Protein Concentrate.
69
-------
Pepsin is the enzyme chosen for this process, since it allows good yields of
soluble protein and has a low pH optimum,(pH 2), resulting in minimal bac-
teriological problems during processing. '' The enzyme breaks up the
natural long-chain, high molecular weight fish proteins into smaller poly-
peptides, peptides, and amino acids with an average length of three or four
amino acid residues. These shorter molecules are very soluble in water and
can be mechanically separated from most of the lipid present, and the insolu-
ble sludge fraction can be separated by centrifugation. The soluble protein
fraction, after centrifugation, is still at a low pH and must be neutralized
before being dried. This is accomplished in an anion exchange reactor, where
the H* ions are removed from the solution. The ion exchange resin is regen-
erated and recycled. The neutralized liquid is then concentrated and spray
dried.
SAW MATERIAL
The raw material for this process can come from various sources, mainly
process wastes, as from fillet and canning lines, and underutilized industri-
al (or trash) fish. The original development work on this process used
English sole fillet waste.'*,10) The process has also been shown to work
well on salmon waste, assorted trash fish, and shrimp from "dirty" shrimp
trawls, and on hake. If the in-plant type waste is treated in a sanitary
fashion, the flesh can be used directly for human food by various processes
of deboning and patty or artificial cutlet formation.(2»H) However, even
poorly treated scraps off the process room floor make acceptable input for
the enzyme process. Yields will be somewhat lower for very poorly treated
fish, since the lipid tends to become more tightly bound in lipo-protein com-
plexes, and emulsions result which lower the soluble protein fraction during
centrifugation. '^)
The data presented in this paper were developed using hake caught in bottom
trawls or as incidental catch in shrimp hauls. Hake is a good example of a
plentiful pelagic species which is difficult to market as a direct human food
fish. It has encountered fresh market resistance due to its texture, is
expensive to fillet, has too much water in the flesh to make good frozen
blocks, doesn't bind water well enough to make good minced products, and has
a limited frozen storage life, due to active enzyme systems in the flesh.
(13,14)
DEBONER
In this process, the first step could be done on shipboard. Bones and skin
could be dumped at sea, and the deboned flesh held in tanks below deck, kept
acidified at pH 3 to 4 for stability. When the vessel is in port offloading
the primary product, a quick disconnect hose could be used to pump the
slurried fish ashore for further processing. In this fashion, incidental
catches such as the up to six pounds of fish caught for each pound of shrimp
could be easily handled. Presently a great amount of extra equipment and
manpower is required to utilize such fish and, as a result, millions of
70
-------
pounds are dumped overboard annually. Figure 2 shows a Japanese type debon-
er In use on a Pacific Northwest shrimp trawler preparing samples from a
trashy shrimp haul. The bone and skin separation is quite good, the unused
portion being quite dry and low in flesh. Recovery can be up to 80% with
whole fresh fish, less with small fish, and as low as 50% from process
waste, depending on the amount of flesh left on the frame.
The original work on this process did not include a deboning step before the
enzyme reactor. The resulting slurry in the reactor required nearly constant
addition of acid to retain a pH of 2, due to calcium and phosphate buffering
activity. Deboning solved this problem, greatly reduced the precipitation of
protein during neutralization, increased the protein yield, and also decreas-
ed the ash content of the final product.
ENZYME REACTOR
The remaining process steps were carried out in the Food Processing Engineer-
ing laboratory at the University. Figure 3 shows the pilot plant developed
in the engineering lab for this process. First the fish is homogenized with
water. The homogenization thoroughly integrates the fish and water and allows
a faster initial reaction rate. Too much homogenization, however, has been
shown to increase emulsification of lipids and proteins, slowing the enzyme
activity and lowering yields.
Following the homogenizer, the fish slurry drops into the enzyme reactor
shown in Figure 4. The reactor was designed with a conical bottom, so that
the slurry can be mixed in two fashions: by pumping out the bottom and re-
circulating, and by mixing with a paddle. At the beginning of a run, when
HC1 is added, the slurry thickens to a paste-like consistency and the paddle
alone is insufficient. Thorough mixing is needed to distribute the enzyme
throughout the slurry. After the reaction proceeds one-half to one hour, the
mixture becomes very soupy, and the paddle works well for continued agitation.
The amount of water added can make a difference in viscosity and ease the
mixing, but each pound added must be removed later. Successive runs show
small increases in protein yield with greater amounts of water. The optimum
chosen in this process is a 2:1 ratio, with 2 parts by weight of deboned fish
to 1 part water. The original work used a reaction temperature of 37°C, the
temperature thought to be optimum. Work on the pilot plant has shown, how-
ever, that by using higher temperatures of 45-50°C the protein recovery is
faster and nearly the same yields can be accomplished with half the amount of
pepsin originally prescribed.
Figure 5 shows typical recovery curves for the reaction at 48°C. The percent
of the original protein that is recovered in the soluble fraction after
centrifugation is plotted against time. Yields are affected by the amount of
water added, amount of pepsin, mixing regime, temperature, quality of fish,
and separational force in centrifugation. The top line represents nearly
optimal conditions, giving a 73% recovery of protein; the second curve, with
half the amount of costly enzyme, gives only a few percent lesser yield and
71
-------
Figure 2. Deboner in Use
/1
-------
Figure 3. Pilot Plant
-------
Figure 4. Enzyme Reactor
7 A
-------
pH 2 U8°C
in
PERCENT
PROTEIN
RECOVERED
123
TIME (HOURS)
0.3$ ENZYME
ENZYME
ENZIME
Figure 5. Percent Protein Recovered in the Liquid Fraction,
-------
is clearly more economical.
The line at the bottom of Figure 5 shows the yield with no enzyme added and
is shown for reference. In the production of fermented fish sauces and other
types of autolysates (without added enzymes), the pH is not lowered to 2. and
protein yields approach 30 to 45%, but only after 48 or more hours. v°»")
ANION EXCHANGE
During the preliminary work on this process, the low pH liquid leaving the
centrifuge was neutralized by addition of base, in this case NaOH. The re-
sulting final product had a very high salt level of 25-30% (89% of which was
Nad). High salt is a drawback to any product to be used as a milk substitute
in feed formulations, and an alternative neutralizing process was neededA**)
Several ion exchange resins were investigated and continue to be investigated
to overcome the high content of the original product. Dowex 2-X8, a commer-
cial grade strong base anion exchange resin, has given good results. This
resin will neutralize the soluble protein fraction to pH 7 with 1/3 vol./vol.
wet bed resin. The resin is reconverted to the OH~ form with two volumes of
1 N NaOH. Both these volumes of resin and NaOH could be reduced in a commer-
cial operation, if 2-X8 were chosen, because the protein need only be neutral-
ized to a pH near 5, and because the reconversion is an equilibrium reaction
and approaches adequate completion with one volume of NaOH.
With a strong base exchange resin, a tank reactor appears to allow the best
control. Resin beads are added to the tank as the protein liquid is added
and adjusted to the proper effluent pH. The resin could then be separated
in a specially designed basket centrifuge which would retain the beads in the
basket for subsequent washing and reconversion. Also under investigation are
different weak base exchange resins which could be handled entirely in
columns, minimizing the possible mechanical breakdown of the beads.
DRYING
After neutralizing, the material is dried. The method of drying can have a
large effect on the final product.^-*-'' Spray drying is the best choice for
this type of product, since it minimizes loss of nutrients, enhances rehydrat-
ability, and tends to whiten the product. Spray drying, however, is not
economic with an incoming liquid less than 30% solids, so the liquid protein
fraction must be concentrated first. The concentration step should also
minimize nutritional losses, so either a low temperature, long time process
(as in vacuum drying) or some efficient high temperature, short time process
is required. The trouble with a proteinaceous liquid, such as in this case,
is that it tends to bake onto heat transfer surfaces quickly, which lowers
the dryer efficiency. We have used a vacuum rotary dryer at 40°C and 25 in.
Hg vacuum with good results, and are currently developing a submerged com-
bustion concentrating unit. This unit burns propane in a specially designed
chamber under the surface of the liquid, releasing hot gas bubbles which rise
through the liquid, acting as very efficient heat transfer surfaces.
76
-------
PRODUCT
The product leaving the spray dryer is a white, fluffy powder. It has the
characteristic bitter taste of small polypeptides and amino acids, but this
bitter taste is considerably less with the ion-exchanging and with shorter
enzyme reaction times than it was in the original process. The bitterness
should not be objectionable in animal feed formulations at the 5 or 10%
level. We also hope to present in August further information on the product's
potential use as a human food additive. A class this term in the School of
Home Economics at the University is trying the material in different food
formulations at varying levels and will produce taste panel data.
A proximate analysis of the product is presented in Table 2. It compares well
to fish meal at 30 to 50% protein, milk solids at ~35% protein, and many of
the organic solvent extracted FPC products at ~70% protein.
TABLE 2. PROXIMATE ANALYSIS - ENZYME DIGESTED HAKE FPC
Total Amino Acids 84.6%
Kjeldahl Protein 85.0%
Moisture 8.2%
Ash 8.9%
Lipid 0.8%
Lead <0.01 ppm
Mercury <0.5 ppm
Fluorides Neg. (AOAC)
Table 3 presents an amino acid analysis of the product compared to casein for
reference. The FPC has a very good essential amino acid profile. Lysine is
quite high; the aromatic amino acids are lower and limiting. Tryptophane was
lost to a great extent in the original work and is generally considered limit-
ing in acid hydrolyzed fish protein. By removing the bones and ions such as
Ca and PO^ which help to complex and precipitate tryptophane, the recovery is
much improved.
77
-------
TABLE 3. AMINO ACID ANALYSIS - ENZYME DIGESTED HAKE FPC
FPC ANRC Reference Casein
Isoleuclne
Leucine
Lysine
Phenylalanine
Tyro sine
Cystine
Methionine
Threonine
Tryptophane
Valine
Total Amino Acids
3.9
7.3
8.5
3.3
2.7
1.0
2.8
3.8
1.1
4.5
84.7
5.0
7.5
6.7
4.0
5.2
0.3
2.3
4.0
1.0
5.9
89.8
Table 4 gives the results of a Protein Efficiency Ratio (PER) assay performed
with the hake FPC.
TABLE 4.
PER
PROTEIN EFFICIENCY RATIO
FPC-Enzyme
Digest
FPC-Enzyme
Digest plus
Tryptophane
Casein
Control
Weight gain/Protein
consumed
Percent of Casein
Control
3.44
1.15
3.45
1.15
3.00
1.00
To confirm the retention of tryptophane, 1% (of the total amino acids) of
tryptophane was added to one diet with no significant change from the un-
changed FPC. The PER results indicate that this FPC product has a good
quality dietary protein.
78
-------
ECONOMIC SUMMARY AND CONCLUSION
Listed in Figure 6 are numbers for both an underutilized species (hake) and
fish waste (cod fillet trimmings). Hake gets up to 80% recovery deboned,
less if small or if a lot of shrimp waste is included. The fillet waste
gives a yield of about 50%. Hake is also slightly higher in protein, 13%
compared to 11 1/2% for the fillet waste. Hake is currently being landed
north of Seattle for between one and two cents per pound. One cent per pound
is a convenient figure for scrap, as it can vary from a few cents to a nega-
tive cent or two to haul it away.
Pepsin is the larges process expense at the current price of $28 per pound,
but we have confidence this cost can be substantially reduced. The raw
material being used is hog stomach mucosa at about 50 cents per pound, giving
about 4% yield of 1:10,000 N.F. pepsin, for a raw product cost of about
$12.50 per pound pepsin. We have ongoing research involving pepsin prepara-
tions and, again, hope to have more information later.
The sludge is about 9% protein, has a good amino acid array, and is proposed
as an animal feed, especially in formulations such as aquaculture fish feed,
where insolubility is a positive factor. Inside the box in Figure 6, 2,700
gallons of liquid at about 7.5% protein goes to the anion exchange. Nine
hundred gallons of mixed bed resin will bring the pH to 7, and require 1,800
gallons of 1 N NaOH for reconversion. It is anticipated, as mentioned
earlier, that both these quantities can be reduced. On the basis presented
here, the per final product pound costs for pepsin are $0.35; for raw materi-
al, $0.23; and estimating $0.20-0.25 for processing, we feel this product
could be put out the door at a cost on the order of $0.80 per pound. This
compares well with non-fat milk solids selling now at about $0.60 per pound
at 35% protein, and with casein selling at $1.65 per pound.
In conclusion, this is a process capable of producing an excellent product
with very few effluent or waste stream problems, promising marketability and
economics, and relatively simple engineering. This may well be a process
that will be able to leave the laboratory and make its way in the real world.
79
-------
HAKE
23400 i
18700 (80%)
2430 (13%)
$468 (.02)
\
28 Ib. Pepsin
$700 5
•aw product (Ib.)
deboned (Ib . )
protein (Ib.)
Dollars
~ "X
FPC
PLANT
70% protein
recovery
FISH WASTE
42000
21000 (50%)
2430 (11.6%)
$420 (.01)
^ Sludge 8400 Ib.
730 Ib. protein
Salt H20 & NaOH
*• 1800 gal.
T
One Ton FPC
1700 Ib. protein
Figure 6. Protein Balance - One Ton Final Product
(85% Protein),
80
-------
REFERENCES
1. FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS. Yearbook of
Fisheries Statistics, Vol. 37, 1974, Rome Italy.
2. SEA RESOURCES ENGINEERING, INC. Upgrading Seafood Processing Facilities
to Reduce Pollution. Paper submitted at the EPA Technology Transfer
Program in Plant Control, Industry Seminars for Pollution Control, New
Orleans and Seattle, 1974.
3. STANSBY, M. E., and OLCOTT, H. S. Composition of Fish. Industrial
Fishery Technology, Reinhold Publ. Corp., New York (1963).
4. SPINELLI, J., KOURY, B., and MILLER, R. J. Food Sci. 37: 599 (1972).
5. SCRIMSHAW, N. S. "The Economics, Marketing and Technology of Fish
Protein Concentrate." Tannenbaum, S. R., Stillings, B. R., and
Scrimshaw, N. S., Eds. MIT Press, Cambridge, Mass. (1974).
6. HALLGREN, B., SJORBERG, L. B., and STELLEMAN, J. "New Uses for Fish
Proteins." In Protein in Human Nutrition. Porter, J. W. G., and Rolls,
B. A., Eds. Academic Press, New York (1973).
7. HALE, M. B. Relative Activies of Commercially Available Enzymes in
Hydrolysis of Fish Protein. Food Tech. 23: 107 (1969).
Using Enzymes to Make Fish Protein Concentrates. Marine
Fisheries Review 35: 15 (1974).
8. McBRIDE, J. R., IDLER, D. R., and MacLEOD, R. A. The Liquefaction of
British Columbia Herring by Ensilage, Proteolytic Enzymes and Acid
Hydrolysis. J. of Fish. Res. Bd. of Can. 18: 93 (1961).
9. TARKY, W., AGARWALA, 0. P., and PIGOTT, G. M. Protein Hydrolysate from
Fish Waste. J. Food Sci. 38: 917-918 (1973), and
TARKY, W. Functional Proteins from Fish Waste by Enzymatic Digestion,
M.S. Thesis, University of Washington, Seattle, Washington, 1971.
10. HEGGELUND, P. 0. Studies to Upgrade the Pepsin Digestion of Fish Waste
for High Quality Protein Recovery. M.S. Thesis, University of Washing-
ton, Seattle, Washington, 1975.
11. MIYAUCHI, D., and STEINBERG, M. Machine Separation of Edible Flesh
from Fish. Fisheries Industrial Res. 6: 165 (1970).
12. SHENOUDA, S. Y. K., and PIGOTT, G. M. Lipid-Protein Interaction During
Aqueous Extraction of Fish Protein. I. Myosin-Lipid Interaction.
J. Food Sci. 39: 726-734 (1974).
81
-------
Lipid-Protein Interaction During Aqueous Extraction
of Fish Protein. II. Actin-Lipid Interaction. J. of Food Sci. 40:
520-523 (1975).
13. PERKINS, C., New England Fish Co., Personal Communication, 1975.
14. BENSUSSEN, M. College of Fisheries, University of Washington, Personal
Communication, 1976.
15. TATTERSON, I., and WIDSOR, M. L. Torry Tests Practical Value of Fish
Silage. Memoir No. 443, Torry Research Station, Aberdeen, Scotland,
1973.
16. DREOSTI, G. M. The Future of Powdered Products. Paper presented to the
Technical Conference on Fishery Products, Tokyo, Japan, Dec. 4, 1973.
17. LABUZA, T. P. Nutrient Losses During Drying and Storage of Dehydrated
Foods. CRC Critical Reviews in Food Technology, Sept. 1972.
82
-------
AN IMMOBILIZED-ENZYME PILOT PLANT FOR THE
TREATMENT OF ACID WHEY**
by
Dr. M. Charles*, Dr. R. W. Coughlin* and Dipl.-Ing. K. Julkowski
INTRODUCTION
We have previously developed an economical bench-scale process for
hydrolyzing the lactose in acid whey based on the use of a simple-
to-prepare inexpensive immobilized-lactase catalyst incorporated
in a fluidized-bed reactor(1). This process has now been scaled-
up and has been incorporated in a pilot plant having a maximum
daily capacity of 400 gallons of whey. The pilot plant also has
ultrafiltration and ion-exchange demineralization units and is
constructed in such a fashion that any sequence of the three unit
operations can be synthesized. Thus, the plant may be used to
obtain data for scale up and economic evaluation of a variety of
whey treatment strategies and at the same time to provide to
potential users adequate supplies of various products (e.g.,
hydrolyzed raw whey, hydrolyzed whey permeate) for their own
testing programs.
BENCH-SCALE WORK
Overview
A fluidized-bed hydrolysis reactor (1 inch diameter) employing a
catalyst comprising lactase immobilized on small alumina particles
was operated successfully at the Lehigh Valley Dairy (Allentown,
PA) to hydrolyze the lactose in raw (unfiltered) cottage cheese
whey. This work, which will be reviewed briefly, demonstrated
the technical feasibility of the process, the superiority of
fluidized-bed reactors, and the suitability of the catalyst, and
provided data which indicated a favorable economic prognosis and
which was used in the design of the pilot plant.
The Catalyst
The catalyst used in all bench scale experiments and in all pilot
plant work to date was prepared as follows(2):
(1) Lactase was adsorbed from solution onto 15Ou diameter
alumina particles having a mean pore diameter of 4000 A
*Department of Chemical Engineering, Lehigh University, Whitaker
Laboratory, Bethlehem, PA 18015
**The authors gratefully acknowledge support for this research
under the National Science Foundation Grant No. GI35997.
83
-------
2
and a surface area of approximately 4 cm /gm. These
particles are available commercially (Carborundum
SAEHA-32) and are presently employed in industrial
catalytic reactors.
(2) The adsorbed enzyme was then crosslinked with glutar-
aldehyde and washed with citrate-phosphate buffer.
This simple inexpensive preparation which used only FDA-approved
materials and which can be performed at room temperature with
very simple equipment provides a catalyst having the following
attributes:
(1) Excellent activity and long-term stability under actual
operating conditions.
(2) Good fluidization characteristics.
(3) Freedom from microbial attack.
(4) Good mechanical integrity; less than 1/2% catalyst loss
by attrition and elutriation for two weeks of on stream
operation.
(5) A pH profile(2) (Figure 1) which is relatively flat
over the normal pH range of acid whey.
(6) Temperature-reactivity characteristics (2) (Figure 2)
permitting stable active behavior at temperatures
between 55°C and 60°C which are close to the temperature
of whey leaving the cheese vats and which tend to dis-
courage microbial growth in the reactor.
More complete information concerning catalyst preparation and
characteristics is available in previous publications (2,3,4).
The Fluidized-Bed Reactor
The rate of enzymatic hydrolysis of lactose is dependent on both
lactose and galactose (reaction product) concentrations and hence
a plug-flow continuous reactor will provide greater productivity
than will other reactors(5). In theory, both fixed and fluidized-
beds can be operated as plug-flow reactors and will give the same
conversion of substrate for a given quantity of catalyst and re-
actor residence time assuming equal mass transfer resistances for
both reactors which is essentially true for the present case.
Therefore, the choice between the two will be dictated by mech-
anical considerations. In order to minimize reactor size and to
increase productivity it is necessary to use a catalyst particle
having a large active surface area per unit volume. In the case
of immobilized-enzyme catalysts this dictates the use of the
smallest practical particle (50u-200y). When such small particles
are employed in fixed-bed reactors, two major difficulties are
84
-------
100 -
Temperature of assay 37°C
-A Soluble lactase Max. 210,000 LU/gm
Lactase-Alumina Max. 12,250 LU/gm
Lactase-TiCl. treated
Stainless Steel Max.
1,690 LU/gm
PH
Figure 1. pH Profile.
85
-------
1 1 1
O Lactase-Alumina Max. 26,000 LU/gm
^ Lactase-TiCl4 treated
Stainless Steel Max. 2,930 LU/gm
100
80
o\<> -1
20
30
±
40 50 60
Temperature °C
Figure 2. Temperature Profiles.
70
80
86
-------
invariably encountered:
(1) Plugging results from particulates in the process
stream and/or microbial growth on the catalyst in
the column.
(2) High pressure drops.
Neither of these difficulties is encountered with fluidized-beds.
Furthermore, liquid fluidized-beds are quite stable and do not
exhibit the violent behavior of gas-fluidized beds; there is a
sharp well defined interface between the top of the catalyst bed
and the liquid and attrition of catalyst is negligible. Details
of the theory and operation of fluidized-bed immobilized-enzyme
reactors have been discussed by Charles et al(3).
Bench-Scale Experiments
Fluidized-bed reactors having a diameter of 1 inch, a height of
3 feet, and charged with lactase-on-alumina catalyst were used
to hydrolyze raw unfiltered whey at the Lehigh Valley Dairy
(Allentown, PA) in a number of long-term experiments conducted at
various temperatures, flow rates, etc. Typical results of these
experiments are given in Figure 3. In this case two columns each
charged with 40 gms of catalyst were operated at 60°C continuously
for almost 28 days with scheduled shut-downs for daily in situ
sanitization with losan solution. Other details of operation are
given in Figure 3. The important results of this experiment,
which are analyzed and discussed elsewhere in greater detail (1,2),
were:
(1) The reactors functioned without mechanical failure or
plugging despite the high particulate content of the
feed (the whey was not filtered); even large cheese
curd particles passed unimpeded through the reactors.
(2) After an initial rapid decay in apparent activity the
catalyst remained stable throughout the experiment;
the catalyst half-life (based on activity following the
initial drop) was greater than 60 days.
(3) The initial decay was caused by only partially reversible
adsorption of whey protein which apparently blinds only
some of the catalytic sites. After the initial adsorp-
tion of whey protein there appeared to be no further
significant substrate-related decay. Therefore, the
problem can be easily overcome by designing on the
basis of reactor residence time based on catalyst
activity following the rapid decay. (The problem will
be greatly diminished when whey permeate is processed.)
87
-------
90
70
10
60
120
Time (hrs)
180 240
300
360
420
Run 5
Two fluidized bed reactors in series
Column diameter - 2.54 cms
Unexpanded bed height - 82 cms
Initial activity - 7,000 LU/gm
Substrate - Acid whey
Expansion - 50%
Flow rate - 50 - 60 ml/min
Sanitized once daily
Column 2
12 16
Time (days)
Figure 3. Long Term Dairy Experiment-
24
8.0
6.0
4.0
x
4J
•H
>
•H
4J
U
28
-------
(4) Catalytic activity is not affected significantly by
daily sanitization with losan.
PILOT PLANT
General Description
The results of the Dairy experiments were used in the design of
pilot plant reactors having a maximum daily throughput of 400
gallons of whey. While the primary purpose of our work is to
evaluate the reactors and catalyst at the pilot scale, it is
clear that enzymic hydrolysis probably will be used in concert
with ultrafiltration and/or ion exchange (or some other demin-
eralization process) to produce commercially acceptable products
but it is not clear which sequence(s) of the three operations
will yield such products. Therefore we have included ultra-
filtration and ion exchange units and have designed the system
to facilitate synthesis of any sequence. Thus the system, which
is illustrated in Figure 4, permits us to:
(1) provide various products to potential users for use
in their own testing programs
(2) obtain data for scale up and economic evaluation of
the various processing alternatives.
It should be noted that in addition to enhancing product accept-
ability the use of ultrafiltration and/or demineralization might
lead to longer immobilized-enzyme catalyst half-life(2,6,7).
This warrants investigation at the pilot scale because a signifi-
cant increase in catalyst half-life will markedly decrease the
cost of hydrolysis(2).
Operating and Design Details
Raw whey may be hydrolyzed directly or pretreated by ultrafiltra-
tion and/or ion exchange and subsequently hydrolyzed. In the
first case, which is easily accomplished when fluidized-bed hy-
drolysis reactors are used but quite difficult with fixed-bed
reactors, raw whey is heated to the desired hydrolysis tempera-
ture (50-60°C) by passing it through a stainless steel shell and
tube heat exchanger (American Standard SSCF-03014) from which it
flows to the immobilized-lactase fluidized-bed hydrolyzers where
the whey lactose is converted to glucose and galactose. The pilot
plant has two hydrolyzers each being 3 inches in diameter by 6
feet high and containing approximately 4 kilograms of catalyst.
Good liquid flow distribution is provided by a conical inlet
section at the base of each column. Support screens which are
known to clog quickly, particularly when raw whey is processed,
are not used; flow to the column is started and stopped by a
quick-acting ball valve which prevents catalyst drainage.
89
-------
Product Tank
to
o
Particle Trap
S—-—SM
—
Intermediate
Tank
Filter Pump
Figure 4. Pilot Plant Flow Diagram.
Reactor Pump
Hydrolysis
Reactors
Rotometear
Heat Exchanger
-------
Hydrolyzed whey then flows through a particle trap to remove any
catalyst which might elutriate and subsequently to the ultra-
filtration and/or ion exchange units if post-hydrolysis treat-
ment is desired. It should be noted that very little of the
catalyst is elutriated; a liquid fluidized bed is characterized
by very gentle agitation of the catalyst particles and a very
sharp liquid-catalyst interface at the top of the bed. In con-
trolled experiments(8) less than 1/2% of the catalyst in a 1
inch diameter reactor (which is not as well-behaved as the 3
inch diameter pilot plant reactors) was elutriated during two
weeks of continuous operation.
If prehydrolysis treatment is desired, raw whey can be ultra-
filtered and/or demineralized. Ultrafiltration is achieved by
a Romicon (H15-45-XM50) filter which is of the type(9) already
proven for separation of protein from acid whey. It has a
molecular weight cutoff of 50,000 and an average flux of 15 gsfd
and permits recovery of valuable whey protein. Ion-exchange is
used for demineralization. The pilot plant demineralizer con-
sists of an anion exchange column containing 0.26 cubic feet of
Duolite ES-340 resin and a cation column containing 0.26 cubic
feet of Duolite C-20 resin. The columns are both 4 inches- in
diameter by 6 feet high and are operated in series.
PILOT PLANT RESULTS
Pure Lactose Solutions
Pure lactose solutions at various concentrations and temperatures
were processed through the pilot-plant reactors to provide base-
line hydrolysis data not subject to extraneous effects such as
whey-protein adsorption for comparison with similar data obtained
during bench-scale experiments. Results of a typical experiment
are given in Figure 5. Operating conditions are specified in the
figure. It should be noted that deviation from steady-state
conversion in this experiment was unquestionably associated with
reactor start up. These results are quite encouraging in that
for the same catalyst (lactase adsorbed on alumina and cross-
linked with glutaraldehyde) and the same normalized reactor resi-
dence of 5.07 x 10^ LU-min/ml the conversions of 5% lactose solu-
tions were 65% in the bench-scale reactors (1" diameter)(1,2) and
84% in the pilot plant reactor: The enhanced conversion at the
pilot scale is believed to be a result of better flow distribution
in the larger reactors. Such better flow distribution may be
partially inherent in the scale-up and may also result in part
from improved distributor design. The bench-scale reactor dis-
tributor was a modified bubble cap(2) which produced visible
channeling while the reactor distributor is simply a conical inlet
section as previously discussed. A comparison of results and
operating conditions for bench and pilot scale is given in
Table 1.
91
-------
100
90
80
70
/
3
O
H
Cfl
g
IS
o
u
60 -
50.
40-
30-
20_
10-
Reactor Diameter: 7.56 cm.
Unexpanded Bed Height: 96 cm.
Expansion: 40%
Catalyst Weight: 3.9 kg.
Catalyst Activity: 6500 LU/gm
Flow Rate: 0.5 Liters/min
pH: 4.7
© 2.5% Lactose; 43°C
A 2.5% Lactose; 23°C
0 5.0% Lactose; 52°C
I
I
I
_L
I
I
I
r
30 60 90 120 150 180 210 240 270 300 330 360
OPERATING TIME (MIN)
Figure 5. Hydrolysis of Lactose Solutions in
Pilot Plant Reactor,
92
-------
TABLE 1. COMPARISON OF PILOT PLANT AND BENCH SCALE RESULTS FOR
HYDROLYSIS OF LACTOSE SOLUTIONS
Reactor Diameter (cm)
Unexpanded Bed Height (cm)
Expansion (%)
Catalyst Weight (gm)
Catalyst Activity (LU/gm) (1)
(2)
Residence Time (LU-min/ml)
Lactose Concentration (wt.%)
Flow Rate (liters/min)
Conversion (%)
Bench Scale
2.54
10.00
60%
50
4400
507 x 102
5.0%
0.06
65%
Pilot Plant
7.56
96.00
40%
3900
6500
507 x 102
5.0%
0.50
84%
(1) Catalyst Activity Assayed At 37°C
(2) Based On Catalyst Activity At 37°C
Whey Permeate & Raw Whey
Initial experiments were conducted in which whey was first ultra-
filtered and subsequently hydrolyzed. Results for conversion of
lactose in whey permeate as a function of time are given in
Figure 6. The initial unsteady-state is associated with reactor
start-up.
The reactor has not as yet been operated with this feed-stock for
a sufficient period of time to observe the effects of catalyst
deactivation.
The results of experiments in which raw, unfiltered whey was fed
directly to the hydrolyzer are presented in Figure 7. (The num-
bers attached to the data point are average reactor temperatures;
some difficulty was experienced in maintaining constant tempera-
ture during these experiments.) It will be noted that in addi-
tion to the unsteady-state associated with start up there is a
slow but discernable decrease in conversion which can not be
completely explained by the temperature variations experienced.
Furthermore, an increase in activity (and hence enhanced conver-
sion) was observed directly after the catalyst bed was sanitized
with losan. (The sanitization operation involves washing the bed
in the expanded state for approximately 1 hour.) However, this
increase was lost soon after the feed of raw whey was started
93
-------
c
s
o
IOO
90 h
80
70
60
50
40
20
10
o 10
Temperotore: s\*
Substrate pH : 4.9
Lactose Con c. : ^.?
O-S L/rwn
Wt:
l?cacior Diameter : ?-5fe
_L
_L
30
4O SO
Time C
60
60 90
Figure 6. Hydrolysis of Whey Permeate in Pilot
Plant Reactor.
94
-------
too
4B Hour t>owrvTt<
loo ISO ZOQ 250 300 350 40O 4SO Soo SSo «PO
Operatina
Figure 1. Hydrolysis of Raw Whey in Pilot Plant Reactor.
95
-------
again and the slow decrease in conversion continued. This is
the same behavior exhibited at the bench scale and is consistent
with our theory that rapid adsorption of whey protein causes
step changes in activity to be superimposed on the natural decay
of the catalyst(2).
While direct comparisons between these results and those obtained
under comparable conditions at the bench scale are not possible
at present and must await the accumulation of long term on-stream
data it now appears that the improved reactor performance ob-
served during the pilot plant experiments with pure lactose solu-
tions promises better long-term results at the pilot scale than
were obtained at the bench scale with whey.
PROCESS ECONOMICS
At the present time the only meaningful determination of return
on investment must be based on the value of the components pre-
sent in the final hydrolyzed product and the waste treatment
savings realized. Such a determination has been done for the
conversion of raw whey to a final product containing 50% solids
and all of the original whey protein and in which 70% of the
original lactose has been hydrolyzed. The main features of the
calculation(2) which are based on bench-scale hydrolysis data are
given in Tables 2, 3, and 4.
TABLE 2. BASIS OF COST CALCULATIONS
Throughput
Volumetric Flow Rate
Lactose Conversion
Catalyst Activity (Initial)
Catalyst Half-Life
Total Catalyst Required
Reactor Diameter
Reactor Height
100,000 Whey Ibs/day
34.4 Vhr.
70%
3,500 LU/gm
60 days
710 kg
20 in.(50.8 cm)
22.5 ft. (6.86 m)
96
-------
TABLE 3. CAPITAL AND OPERATING COSTS
Equipment
Catalyst
Operating
Fixed
Total For Hydrolysis
Demineralization
Concentration (50% Solids)
Total
$101,000.00
$15,110.00
$371.00/day (Incl. Catalyst)
$70.00/day
9$/lb. Whey Solids
2$/lb. Whey Solids
4
-------
on the values of individual components (although this will not
effect waste treatment savings which constitute almost 20% of
the calculated product value). A determination of this market
value must await the results of product testing by prospective
users.
-------
1. Coughlin,,R. W., Charles, M., Allen, B. R., Paruchuri, E. K.,
and Hasselberger, F. X. Increasing Economic Value of Whey
Waste Waters Using Immobilized Lactase. Presented at AIChE
Philadelphia Meeting, Nov. 1973.
2. Paruchuri, E. K. Ph.D. Dissertation, Lehigh University,
Bethlehem, PA (1976).
3. Charles, M., Coughlin, R. W., Allen, B. R., Paruchuri, E. K.,
Hasselberger, F. X. Lactase Immobilized on Stainless Steel
and Other Dense Metal Supports. In "Immobilized Biochemicals
and Affinity Chromatography". R. B. Dunlay, ed., Plenum
Press, NY (1974).
4. Charles, M., Coughlin, R. W., Paruchuri, E. K., Allen, B. R.,
Hasselberger, F. X. Enzymes Immobilized on Alumina and
Stainless Steel Supports. Biotechnology and Bioengineering
17:203 (1975).
5. R. W. Coughlin, M. Charles. Comparisons of Potential Reac-
tors for Immobilized Enzymes. Presented at ACS Symposium
on Antibody and Enzyme Engineering, Purdue Univ., Jan. 1974.
Enzyme Technol. Digest 3(2), p.69 (Nov.1974).
6. Quinn, M. R., Beuchat, J., Miller, J., Young, C. T., and
Worthington, L. E. J. Food Sci. 40:467 (1975).
7. Ford, J. R., and Pitcher, W. H. Paper presented at Whey
Products Conference, Rosemont, IL, Sept. 18, 1974.
8. Unpublished results.
9. Personal communication, B. R. Breslau, Romicon, Inc.
99
-------
IN-PLANT CONTROL TECHNOLOGY FOR THE
FRUITS AND VEGETABLES PROCESSING INDUSTRY
by
Kenneth V. LaConde* and Curtis J. Schmidt*
During 1974, SCS Engineers, under contract with the En-
vironmental Protection Agency, conducted a nationwide
study of the preserved fruits and vegetables processing
industry to develop effluent limitations guidelines.
Throughout 1975, additional work to investigate, identify,
and describe current in-plant technologies was performed
for this same agency.
Initially, in 1974, approximately 500 processing plants
representing 58 subcategories were visited to obtain the
followi ng:
1. Basic manufacturing processes and unit operations.
2. Wastewater volume and characteristics. Wet sampling
was performed at selected plants to both verify and
expand the data base.
3. Current in-plant practices utilized to reduce waste-
water volumes and pounds of pollutants (BOD and SS).
4. Current treatment technologies, performance effici-
encies, alternatives, and costs.
All data received, after careful screening, was placed in a
computer bank, the output of which resulted in plant and
subcategory averages. These data were in turn recycled
through the individual companies to be verified for accu-
racy. Comments were subsequently received and the data,
wherever necessary, modified. Additionally, new input was
also received from the current processing season (1974) to
expand the existing data base. All of these information
points were then, once again, entered into the computer.
This resulted in the final data base which EPA subsequently
used to develop the effluent limitations guidelines as
promulgated on October 21, 1975. The work effort of SCS
Engineers relative to in-plant control technologies util-
ized this final data base as a starting point.
*SCS Engineers, Long Beach, California and Reston, Virginia
100
-------
In-plant controls are usually the most effective way to
reduce waste load generation in fruit and vegetable pro-
cessing plants. These alone cannot accomplish the required
efflu-nt discharge limitations, but they can reduce the
waste load by a factor well over 50 percent, reducing the
cost of city discharge or waste treatment accordingly.
In-plant controls not only reduce the waste load, but are
also employed to improve processing efficiency and product
quality. Nearly all of the 500 plants visited could have
improved their in-plant control and water conservation
measures.
The first task at hand was to define those subcategories to
be investigated. Each of these were reviewed in terms of
wastewater volume, pounds of BOD and SS generated, and size
of pack. Concurrent with this, further development of the
economic study necessitated review of certain specific sub-
categories.
From a combination of the above factors, the following sub-
categorias were finally chosen for investigation: canned
apricots, brined cherries, sour cherries, sweet cherries,
caneberries, dried fruit, canned peaches, canned pears,
canned plums, strawberries, baby food, canned asparagus,
canned dried beans, canned beets, canned corn, frozen corn,
canned and frozen carrots, mushrooms, canned peas, frozen
peas, fresh pack pickles, processed pickles, sauerkraut
canning and cutting, frozen spinach/greens, canned spinach/
greens, canned snap beans, frozen snap beans, sweet potatoes
tomato products, peeled tomatoes, and potato chips.
Each of the individual data points which comprised the com-
puted averages were then carefully examined. Previous com-
puter runs had expressed each of the reported data points
in terms of flow, BOD, and SS. All results, except as
otherwise noted, were expressed in terms of raw product
received. Further, within each subcategory, arithmetic
and log averages were computed using both individual data
points and individual plant averages. This defined individ-
ual plant performances and afforded an opportunity to com-
pare the waste parameters from various plants within each
subcategory.
Those plants which exhibited significantly lower raw waste
loads than the average were identified, and the data care-
fully re-examined. The emphasis of this examination was
placed on those in-plant controls and/or technologies which
were positively identified as a contributing factor to the
reduced raw waste loads.
101
-------
This resulted in a list of plants to be investigated i-n more
detai 1.
As previously mentioned, the majority of information relative
to individual plant operations had been gathered throughout
the 1974 contract period. It was, therefore, necessary at
this time for a third review of all information on hand.
The scope of this examination involved the following:
Examination of the various in-plant unit processes
and their potential effect on pollutants generated.
Comparison of these individual operations to other
plants within the same category.
An evaluation of the observed unit process differences
and their significance to the performance averages.
Relationship of product styles, size of container
(industrial vs. retail pack) and other similar vari-
ables to waste load differences.
The formation of various questions to clarify what
particular steps or operations these plants possessed
that enabled them to produce a significantly lower
waste load.
This list of questions was further and more completely
developed to lend input to existing in-plant qualitative
information. This checklist, referred to as the In-Plant
Control Check Sheet, encompassed the basic unit processes
for a majority of fruit and vegetable preservers. These
were raw product cleaning, product conveyance, peeling,
blanching, product cooling, canning, waste handling, and
cleanup.
The plants were also asked to detail other conservation
systems not covered by the check sheet. Additionally,
they were requested to lend input to waste management
philosophy within their respective companies. For example,
who is responsible for waste control within the respective
processing plants? In what terms are equipment modifica-
tions considered? In terms of water use reductions? Im-
proved case yields? Labor savings? Energy/chemical
savings? All of these? Other considerations? An addi-
tional question concerned the use of by-products. Are any
useable by-products generated from any specific equipment
that was installed for pollution abatement purposes?
Further, are any revenues received from these by-products
to help off-set the cost of this equipment?
102
-------
Upon receipt of the completed forms, the individual plant
comments were compared to the original plant visit informa-
tion,files. Additions or corrections, where appropriate,
were made. As a further step to better delineate the unit
processes, follow-up phone calls were made for more detailed
clarification. This point in time, then, represented the
most current and complete data bank gathered throughout the
contract period.
By and large, the results of these examinations were not
surprising. Each of those plants responding had taken
positive action in one way or another to reduce their waste
loads. Taken as a whole, certain unit processes and/or
technologies were readily identified as having an impact on
raw waste load reductions. Among these were:
Separation of low and high strength waste streams
Installation of low-volume high-pressure cleanup
systems
Dry in-plant transport of products
Countercurrent reuse of wash/flume/cooling waters
Dry handling of solid wastes
Dry caustic peeling
Changeover from water to steam blanching (where
possible)
Air cooling after blanching
An active and progressive waste management program
As a result of this current work, we would like to present
one case study to illustrate what can be achieved by pro-
gressive and conservation-minded management. Technologies
that will be discussed are: separation of waste streams,
blanching, air cooling, solid waste handling, and waste
management.
103
-------
CASE STUDY
Plant "A" is a processor of frozen carrots. The plant
operates on a typical schedule of 5 days per week, 10 to 11
months each year. Normal production is set for two shifts
per day with additional cleanup time as needed. Average
throughput is approximately 8 raw tons per hour.
Figure 1 is a flow diagram of this plant's operation prior to
1974. As can be seen, the manufacturing processes were
quite typical. Carrots were delivered in bins either di-
rectly from the field or from a fresh sorting shed. They
were washed, trimmed, conventionally lye peeled and diced
or sliced. Blanching was accomplished by means of a water
blancher; post blanch cooling was by water. Water usage
was high with a daily average approaching 1.5 mgd. BOD
levels were similarly high averaging about 2400 mg/1.
This plant, landlocked by city growth, was already paying
a surcharge for its effluent disposal, but the city,
recognizing that expansion was needed for industrial users,
initiated discussions for a revised surcharge system.
Tentative proposed surcharge formulae indicated that if
this system was adopted, monthly discharge rates could
approach $4,00£).
At this point, corporate management decided to institute
changes in the form of capital investments. The engineers,
when designing the modifications considered each waste
contributing point within the processing sequence. Figure
2 shows the results of these efforts; Table 1, the benefits.
Carrot Washing
Water used to wash the incoming carrots had previously
been combined with the main plant effluent stream.
However, observation and testing of this stream indicated
that its chief components were sand and dirt, some field
debris, and occasional carrot particles. Becuase of
the plant's location, they decided to isolate this stream
from the higher concentration waste streams and discharge
it separately under NPDES permit. A double screening
system was installed (not shown on diagram) to effectively
remove the various solid particles from the waste stream.
This not only lowered the suspended and settleable
solids but also in part, contributed to reduced BOD
discharge. Freezer defrost water was also diverted to
merge with this stream.
104
-------
BINS
WASH
LYE PEEL
TRIM
DICE/SLICE
WATER BLANCH
WATER COOL
FREEZE
LIQUID
WASTE
DIRT, SOLIDS
SOLIDS, LYE
SOLIDS
ORGAN
--!
ICS, SOLIDS ^ I
1
ORGANICS
-J
I
CITY
SEWER
FIGURE 1. ORIGINAL FLOW, CARROT PROCESSOR "A1!
105
-------
SOL
WAS
SOL
DISP
.ID
5TE
LYE-PEEL
SLURRY
IDS
OSAL
BINS
WASH
DRY
CAUSTIC
PEE'
TRIM
DICE/
SLICE
STEAM
BLANCH
AIR
COOL
FREEZE
LIQ
WAS
DIRT, SOLIDS
f
SEPARATE
DISCHARGE
JUICES
~~
CONDENSATE
!
CI
SEI
JID
5TE
TY
tiER
FIGURE 2. MODIFIED FLOW, CARROT PROCESSOR "A"
106
-------
The effect of this separation should not be underestimated
These two components are responsible for approximately
60% of the total plant flow - about 350,000 gallons per
day. The alternative discharge of this large volume of
water to a city sewer system would have been extremely
expensi ve.
Peel ing
The peeling process was identified as the main source of
BOD and SS generation. The water required for peel
removal was a substantial percentage of the plant's total
effluent. A new lye peeler (ferris wheel type) was '•
purchased and installed.
The conventional high pressure water lye peel removal
system was replaced with a Magnuscrubber. The resultant
peel waste slurry was pumped directly to a holding tank.
Ultimate disposal of this waste fraction is to a sanitary
landfill.
Table 1. Carrot Process "A." Parameter Differences
Parameter
Flow (MGD)
Flow Ratio (Gal/Ton)
BOD (mg/1)
Sliced Recovery (%)
Diced Recovery (%}
Monthly Surcharge ($)
Before
1.5
9375
2400
50
65
4000*
After
0.61
3750
1600
62
72
500
* Estimated
1 0.35 MGD discharged under NPDES permit, 0.25 MGD to city
107
-------
Note that in addition to the expected water usage and
BOD reductions, finished product recoveries increased for
both sliced and diced. These increased yields are a
direct result of the new peel removal system. Although
numbers were not available, the plant personnel were also
aware of a reduction in caustic soda consumption.
Dry Solids Handling
The plant engineers recognized that every attempt should
be made to stop the introduction of solid wastes into the
effluent stream. The original plant flow allowed for
trimmings and solid waste to be flushed into gutters with
final screeing before discharge to the city sewer system.
This allowed for additional leaching of soluble solids
and was, in part, responsible for the plant's high BOD.
To overcome this problem, a separate series of dry solids
conveyors were installed to transport trimming table
and other wastes directly to outside waste bins.
Further, a concerted effort was made by management to
instruct plant employees on proper methods of solid waste
handling.
Blanchi ng
As shown in Figure 1, a hot water blancher had been used
over the years.
Enzyme deactivation and product quality met the necessary
criteria for continued use. Water blanchers, however,
have been shown to be a major source of water pollution.
The action of solids leaching, continuous spillage and
overflow contribute heavily to a plant's BOD and SS
loads. The waste stream is not only concentrated, but
is of substantial volume. The old water blancher was
removed and a stream blancher was installed. The only
effluent from this unit process is a highly concentrated
1-2 gpm condensate. The effect on total BOD and SS
load has been considerably lessened, yet the desirable
quality characteristics retained.
Cooling
Similarly, post blanch hydrocooling necessary to pre-
pare carrots for freezing and to optimize freezer
efficiencies was determined to be a contributor to the
plant's effluent load - to eliminate the flow altogether
was the objective. The conventional hydrocooling
system was replaced with an ambient air cooler. Working
on a fluidized bed principal, the product exits the
108
-------
falancher and is blown and vibrated over perforated screens
by the action of air jets blowfng through and upward to
contact the product. The action of the air and the
mechanical vibrations of the screen move the product
across the screen during which the required cooling
gradient is achieved. The use of air eliminates almost
all water contact after blanching except for one or two
fine water sprays between the blancher and cooler.
Water usage at this stage was estimated at no more than
2 gals/minute.
Capital Investment
The plant estimates their total expenditure for all the
improvements and modifications to be approximately
$90,000. Although no attempt was made to calculate
paybacks on investment capital, it can be seen in Table 1
that savings accrued due to reduced discharged volume
and pounds of pollutants to the city accounts for
approximately $42,000 per year. If one were to add a
monetary value to the increased yields, the investment
looks even more attractive.
Waste Management
We have briefly reviewed the components of a revised,
modified and updated carrot processor. We have alluded
to an active management role and corporate confidence
in terms of capital invested. One personal observation
should be added at this point. All of the employees
seemed to be cognizant of management's attempt to
reduce both solid and liquid wastes. The plant was a
model of cleanliness. They all appeared involved,
concerned and willing to do their share to help manage-
ment achieve its goals.
109
-------
LAND DISPOSAL OF WINERY WASTEWATER
by
Dr. Larry L. Russell*, Dr. John N. DeBoice*, and
Dr. Walter W. Carey**
INTRODUCTION
A wastewater treatment and disposal system has been constructed recently
for the Beringer Winery in St. Helena, California. The treatment system
consists of screening and neutralization prior to disposal of the wastewater
to seven percolation beds. The wastewater flow is usually less than
50,000 gpd and the Chemical Oxygen Demand (COD) of the wastewater can be
10,000 mg/1 or more.
The land disposal system was chosen because it provided the highest
level of treatment for the least capital and operating costs. The nuisance
problems, such as odors and gnats frequently associated with winery waste-
water treatment, are completely controlled by applying the wastewater to the
soil at a rate sufficiently low to ensure that aerobic conditions are main-
tained in the soil. Additionally, a very high level of treatment would be
required prior to discharge to a surface water, and the costs would be
excessive.
A study of the effects of the disposal system on the ground water
quality was conducted prior to initiating discharge. The study consisted
of a computer simulation of the chemical and biological reactions that occur
in the soil in conjunction with a dispersion analysis. The study showed
that land disposal of the wastewater was feasible and that the effects on
the ground water quality would be minimal.
An ongoing monitoring program consists of analyzing samples taken
from observation wells located upstream and downstream of the percolation
beds. In this manner, the quality of the ground water is monitored, and
changes would be observed prior to any significant impact on the ground
floater.
BACKGROUND
The construction of a large new winery at the Beringer Winery site
required the design of a new wastewater treatment and disposal system.
In a previous study at another Beringer Winery, the wastewater generation
*James M. Montgomery, Consulting Engineers, Inc., Walnut Creek, California
**The Nestle Company, White Plains, New York
110
-------
varied from 200-300 gallons per ton of grapes crushed. Using this informa-
tion, the system was designed to treat and dispose of about 50,000 gallons
per day (gpd) during the crushing season, which normally extends for 45
days in the fall. The wastewater flow rate decreases to about 20,000 to
30,000 gpd during the noncrushing season.
Several alternative treatment systems were considered, such as aerated
ponds, high-rate biological oxidation and percolation beds. A review of
the literature conducted by Stokes") indicated that the most reliable means
of treatment was the use of percolation beds, and that nuisances such as
odors would be completely controlled by proper operation and wastewater
application. A recent survey conducted for the Environmental Protection
Agency(2) indicated that approximately 80 percent of the wineries in the
United States depend on land treatment by either aerated ponds or percolation
beds. The remainder of the wineries utilize municipal treatment systems.
After considering the alternatives, the use of percolation beds was
evaluated in detail because a high level of treatment could be obtained
for low cost. The soils in the area of the winery are coarse alluvial
materials underlain at about 25 feet by a clayey sand. Percolation tests
were conducted at the chosen location; the soil exhibited percolation rates
of 10-20 feet/day which would easily accommodate a loading rate of 100,000
gal/acre/day proposed by Coast Laboratories in 1948\3). This loading rate
results in a depth of application of about 4 inches. The soil should be
allowed to dry thoroughly (approximately 6 days according to Coast Labora-
tories) prior to the application of more wastewater. The intermittent
application of the wastewater ensures that aerobic conditions are maintained
and that the biological growths, which act to "blind" the soil and reduce
its percolative capacity, will be controlled by the drying out periods.
Permeability studies conducted on the coarse alluvial soils showed
that the soil was quite permeable, having permeabilities in excess of
2,000 gal/day/sq. ft. The hydraulic gradient is such that the Darcy velocity
is about 3 feet/day.
During the planning stages of this study, a computer simulation was
conducted of the biological and chemical reactions that occur in the soil.
The model, described in detail by Russell (4,5), is capable of accurately -*«
predicting the quality of a percolating water applied to a soil for treatment
and disposal. The simulation indicated that a hardness, alkalinity, and
TDS pickup of up to 400 mg/1 (hardness and alkalinity expressed as CaCO-J
would be experienced due to the dissolution of the calcite in the soil by
the C0? produced as a result of the biological oxidations of the organic
material.
The results of the computer simulation were used as inputs to a dispersion
analysis; and the effects of the discharge on the receiving water, the
Napa River located 2,000 feet from the percolation beds, were evaluated.
Figure 1 shows a cross section of the aquifer and the projected area of
influence denoted as the plume. The steady state operation of the treat-
ment and disposal system would result in an estimated increase in the TDS
of the ground water of about 200 mg/1.
Ill
-------
WASTE
PONDS
240
UJ
Ul
O 220
I
UJ
UJ
200H
^
u ^
1
1
1
1
1
iv%*®***^
AGROUND SURFACE
1
1
1
1
1
1
•
1
NAPA
RIVER
FIGURE 1: CROSS SECTION OF THE AQUIFER
-------
The ground water in this aquifer is not used for irrigation or domestic
purposes. The entire ground water flow joins the Napa River; and as long
as the TDS change is within the range indicated by our studies, the effects
of the discharge are minimal.
Figure 2 shows the layout of the percolation beds and their orientation
relative to the movement of the ground water. There are seven percolation
beds, as shown in Figure 2, and the wastewater is applied to a different
bed each day to allow the beds to dry and to ensure that aerobic conditions
are maintained. The new winery is located to the southwest of the percola-
tion beds, and the Napa River is located about 2,000 feet from the percolation
beds in the direction of the ground water movement.
The wells are denoted by the letters "U", "N", "M", and "S." The
well marked "U" is the upstream well, and the other wells denoted "N",
"M", and "S" are the north, middle, and south wells. These wells are used
to monitor the quality of the ground water on a monthly basis during the
crushing season and a quarterly basis during the rest of the year.
WASTEWATER CHARACTERISTICS
The wastewaters generated during the production of wine are primarily
from the washing and cleanup of the fermenters and storage tanks. Signifi-
cant quantities of inorganic solids such as bentonite and diatomaceous earth
are used in the "racking" or fining of wines, and these inorganic solids
combine with the organic solids removed from the wine to form a sludge at the
bottom of the tanks. Most of the organic material in the wastewater is
biodegradable.
Most of the data in the literature on winery wastes concerns wastes
that contain stillage. Very little data has been reported on non-stillage
winery wastewaters. The Beringer Winery has no stillage wastes, and the
treatment system was designed on the basis of a small amount of data collected
at another Beringer Winery and the limited data in the literature. A com-
prehensive sampling program was initiated to provide the data to characterize
the wastewater and to ascertain that the treatment system was suitably
designed for the wastewater.
Samples were taken by an automatic sampler every one and one-half
hours for a period of two months. Four samples were deposited in the same
bottle to form a six-hour composite sample. Half of the six-hour samples
were preserved with sulfuric acid and the other half were preserved with
mercuric chloride. The samples preserved with sulfuric acid were analyzed
for hardness, calcium, TDS, COD, and total Kjehdahl nitrogen (TKN). The
remaining samples were analyzed for alkalinity, pH, conductivity, and odor.
The results of the sampling program are shown in Table 1. Note the
large amount of variation in the concentration of these constituents. The
values reported are average daily concentrations. The variation in the
results is due to the batch nature of the cleaning operations. The minimum
113
-------
FIGURE 2: WASTEWATER TREATMENT
AND DISPOSAL SYSTEM LAYOUT
114
-------
values reported for the hardness, IDS, and COD reflect the concentrations
found in the cooling water, which amounts to 10,000 pgd.
TABLE 1. WINERY WASTEWATER QUALITY (Non-crushing Season)
Hardness*
mg/1
Mean 185
Max 580
Min 80
*as CaC03
Alkalinity*
mg/1
284
1,070
0
TDS
mg/1
870
2,000
200
COD
mg/1
2,950
25,700
100
EtL
6.5
9.7
3.9
TKN
mg/1
12.5
74.5
1.2
The samples rarely had any foul odor; most of the odor was due to wine,
and the odor of the wastewater was not unpleasant.
GROUND WATER QUALITY
One stipulation of the discharge permit was that the ground water quality
would be monitored by ovservation wells located up and downstream of the
percolation beds. This phase of monitoring has been conducted since the
fall of 1974. The ground water monitoring program was supplemented during
the wastewater monitoring period by taking additional samples weekly.
The samples were preserved by refrigeration and the analyses were conducted
as soon as possible.
The upstream water quality data is presented in Table 2. Note the rather
high COD of 45 mg/1, which reflects the use of the aquifer as a receiving
water for wastewaters from sanitary drain fields and from previous waste-
water disposal practices.
TABLE 2. UPSTREAM GROUND WATER QUALITY
Hardness*
mg/1
Alkalinity*
mg/1
TDS
mg/1
COD
mg/1
pH
185 180 300 45 6.9
*as
115
-------
The downstream water quality data is shown in Table 3. The wells are
identified as noted in Figure 2. The bulk of the effect of the discharge
is seen in the middle well ("M"), as would be expected from the dispersion
pattern that results from a line source discharge, such as the percolation
beds. Note the small effect of the discharge on the COD of the ground water;
a significant fact in light of the large COD concentrations found in the
wastewater. The hardness, alkalinity, and TDS pickups are less than the
maximum concentrations predicted by the computer simulation and the dis-
persion analysis, which most likely reflects a more rapid and thorough out-
gassing of the C02 at the soil surface during the drying period than was
anticipated. As more C02 is outgassed, the ability of the wastewater to
dissolve calcite is reduced.
TABLE 3. DOWNSTREAM GROUND WATER QUALITY
Well
N
M
S
*as
Hardness*
mg/1
300
425
240
CaC03
Alkalinity*
mg/1
260
470
315
TDS
mg/1
340
370
315
COD
35
65
40
RH
6.3
6.7
6.8
SUMMARY AND CONCLUSIONS
The installation of the treatment and disposal system at the Beringer
Winery has been an unqualified success. The percolation beds have been
in use for nearly one full season, and the effects on the quality of the
ground water have been minimal. No odor or nuisance problems have occurred
and the percolation beds have been operated with a minimal amount of energy
and manpower. The intermittent loading practice has resulted in a manageable
and efficient wastewater treatment system.
The soil type and ground water gradient have combined to make an ideal
situation for a wastewater disposal system of this nature. The high rate
of percolation has been maintained throughout the period that the percolation
beds have been in operation. The quantities of inorganic and organic sus-
pended solids applied to the percolation beds have not significantly affected
the percolation bed operation.
The impact of this treatment system on the environment has been minimal.
This system is a good example of the integrated approach to wastewater
treatment wherein an effective treatment system has been developed that
combines energy conservation with the maintenance of high standards of
environmental quality.
116
-------
REFERENCES
1. Stokes, R. D., "An Evaluation of Current Practice in the Treatment
of Winery Wastes," Masters Thesis, University of New South Wales,
Australia, 1967.
2. Associated Water and Air Resources, Inc. and Environmental Research and
Applications, Inc., Status of the Art of the Wine Industry SIC. 2084.
Prepared for the Environmental Protection Agency, December 1971.
3. Coast Laboratories. Report on 1947 Grape Still age Disposal Report
to the Wine Institute, July 1948.
4. Russell, L. L., "Chemical Aspects of Ground Water Recharge with Waste-
waters," Ph.D. Thesis, University of California, Berkeley, 1976.
5. Russell, L. L. and Thomas, J. F., "Increase of TDS by Ground Water
Recharge." Presented at the Annual Conference of the Water Pollution
Control Federation, Miami, October 1975.
117
-------
LOW WASTEWATER POTATO STARCH/PROTEIN PRODUCTION PROCESS -
CONCEPT, STATUS, AND OUTLOOK**
by
* * *
J. R. Rosenau , L. F. Whitney , and R. A. Elizondo
ABSTRACT
A process for producing starch, feed grade protein meal, and pulp from cull
potatoes with minimal water use has been developed and is being "tuned".
Potatoes are ground and wet sieved to remove pulp. Recycled juice is used
in the sieving process to flush starch from the pulp. The pulp, after
pressing or centrifuging to remove residual juice, is dried. The starch-
juice mixture is passed through an elutriation type liquid cyclone yielding
a starch-rich underflow and a starch-free juice overflow. The use of low-
volume elutriation in the cyclone minimizes protein and fiber in the starch
underflow which is dewatered and washed in a vacuum filter (or basket
centrifuge), dried and ground. Air classification of the dried starch
further reduces impurities and improves whiteness.
A bleed stream from the recycled juice removes the protein introduced by the
potatoes. This juice, which contains 4% solids (w.b.) and 50% protein (d.b.),
is concentrated and dried. The use of ultrafiltration, or heat and acid
induced precipitation, to concentrate and fractionate juice solids is
currently being investigated.
INTRODUCTION
Research relating to food processing wastes can, in most cases, be divided
into three areas: treatment of food processing wastes, process modification
to reduce waste loadings, and by-product recovery. Of these, the last areas
are often the most exciting from a plant operation standpoint since they
usually can be implemented at a profit.
Cull potatoes have been processed for starch and pulp for many
years.(1»2»3^4,5) Traditionally, starch processing has included grinding
*
Department of Food and Agricultural Engineering, University of
Massachusetts, Amherst, MA 01002.
**
This investigation is being supported by funds from the Environmental Pro-
tection Agency (Grant No. R-803712-01-0), Agway, Inc., the Maine Potato
Commission, and the University of Massachusetts.
118
-------
(in the presence of S02 which retards browning) and wet sieving with an
excess of water to separate starch from pulp. The pulp is dewatered by
pressing or centrifuging and dried in a rotary drier for cattle feed. The
starch is refined, originally by settling or tabling, more recently, by
centrifugal or multistage liquid cyclone systems. Vacuum filtration or
basket centrifugation is used for final starch dewatering prior to drying,
grinding and bagging.
Non-starch constituents washed from starch granules give rise to large BOD
loadings in the effluent streams, and many starch processors have been
forced to close because waste processing systems capable of handling the
discharges are prohibitively expensive. Pollution problems have played an
important part in the closing of over twenty plants in the New England area
alone .
The closing of these plants has lead to additional problems. Culls have
been dumped into landfills or back onto the growers' fields (which leads to
potential disease propagation). Meanwhile, the price of potato starch has
climbed to about 13
-------
ro
o
WHOLE POTATOES
I
WASH
I
SULFITE-
GRIND
i
PRESS OR FILTER
I
PULP
FLASH
DRY
SCREEN
AIR CLASSIFY
T
STARCH
JUICE
-•-PEEL
I
EVAPORATE
I
SPRAY DRY
T
PROTEIN MEAL
Figure 1. Initial Starch/Protein Meal Process Flow Chart
-------
particles of pulp that were small enough to pass through the 200 mesh screen
with the starch but were not removed by air classification - even when pre-
ceded by grinding at the most intensive setting on an Alpine American cross
flow mill.
A critical review of the above results - especially in view of the ability
of the traditional process to produce starch of extreme purity - led to a
refinement that drastically improved starch quality, lowered wash water
demand still further, and reduced capital equipment costs for the process.
MODIFIED STARCH PRODUCTION PROCESS
Figure 2 depicts the modified starch production process. Potatoes are
ground as before and the slurry wet-sieved by flooding with recycled juice.
The pulp remains above the 140-mesh screen while juice and starch pass
through. The slurry is pumped to a liquid cyclone wherein the starch is
spun to the underflow while the overflow is largely recycled to the screen.
Excess juice at 4% total solids (w.b.) passes to the juice processing
system.
The underflow starch is dewatered and washed on a belt vacuum filter (or
some other device such as a basket centrifuge, decanting centrifuge, or
rotary vacuum filter). In our trials to date, we have used a pilot (3
horizontal belt filter produced by Straight Line Filter, Inc. The starch
cake produced tested about 65% total solids (w.b.). The cake was dried in a
tray drier at 140°F.
Liquid from the starch dewatering and washing step is recycled back to the
liquid cyclone and is pumped into the tangential elutriation inlet. This
particular design is a patented feature of the Bird Machine Co. This recycle
stream is necessary to aid in removal of colloidal materials from the starch
which otherwise interfere with the filtration step.
Rough tuning of this scheme to date using a wash stream of 25 pounds of
water per 100 pounds of cull potatoes processed has produced starch of 0.3%
protein (d.b.) and a reflectance reading of 87% even without air
classification. Even better results should be achieved as additional experi-
ence is gained with the method and it is revised and tuned accordingly.
Upon air classification of the above starch, the reflectance increased to
89% while the protein level decreased to 0.16%. Small samples of this
starch may be obtained on request.
PRELIMINARY ECONOMICS
Currently, cull potatoes are available at $0.75 per cwt. Each hundred
weight should yield ten pounds of starch which, at 13<: per pound, is worth
$1.30. In addition, each hundred weight should yield four pounds of dried
pulp, at $80.00 per ton, worth $0.16. Finally, a yield of 4 pounds of 50%
protein meal, at $200. per ton, should be worth $0.40. Value of all
121
-------
Washed Cull Potatoes
Sulfite
\r
Grinder
Recycle Juice
V V
14 O mesh Sieve
Starch
Slurry
Pulp
Press
or
Centrifuge
Juice
Balance Tank
Overflow
Liquid Cyclone
Sulfited
Water
Under-
flow
Vacuum Filter
c
o
Pulp
Pulp Drier |
T
Dried Pulp
Defoamer
Juice (4 % solids)
Drier
Air Classifier
1
Fines for
Animal Feed
Stare h
Figure 2. Revised Starch Production Flow Chart
122
-------
products is estimated at $1.86.
Energy costs for drying the starch and pulp and concentrating (with a four-
effect evaporator) and drying the juice should be $0.13 (assuming $3.00 per
million Btu). This leaves $0.98 per cwt. to cover profit, capital costs,
and operational expenses. While it is premature to attempt to evaluate
closely all economies of the process, communications with industry have
indicated that these figures suggest a potentially profitable enterprise -
especially since there are plants in existence that already have some of the
appropriate equipment in hand.
It is interesting to note that test plots in the Connecticut River Valley
have produced twenty ton of pototoes per acre. In the above terms, this
corresponds to 4,000 pounds of starch ($520), 1,600 pounds of pulp ($64),
and 1,600 pounds of 50% protein meal ($160). In comparison, soybeans pro-
duce about 700 pounds of protein per acre and alfalfa, about 2,400 pounds.
Since most of the true protein (about 1/3 of the crude protein) of potato
juice is heat coagulable, it has potential as an egg white substitute in
some food systems. In addition, researchers at the University of
Minnesotav^) have succeeded in breeding potatoes to twice their normal
protein content.
JUICE PROCESSING
There are many possible processing schemes that could be applied to the 4%
solids (w.b.) juice which contains about 50% crude protein (d.b.). Many of
these schemes have been investigated particularly in the low concentration
range by researchers at the Eastern Regional Research Center of the U.S.
Department of Agriculture^* 9). The rapidly changing animal feed •market
(which includes the possibility of liquid feed concentrates), the relatively
high solids concentration of the juice produced by the starch process
described in this paper, and the rapidly changing state of the art of ultra-
filtration and the reverse osmosis technology C11*12) suggests that con-
tinued work is appropriate.
Figures 3, 4, and 5 outline the three broad categories believed most
promising, each of which carries a number of sub options to further
maximize profitability: 1) the use of ultrafiltration to concentrate
protein; 2) the use of heat and acid induced coagulation to concentrate or
fractionate protein; and 3) concentration of juice without fractionation.
The profitabilities of these options and sub options depend upon permeation
rates within the ultrafiltration and reverse osmosis unit operations. These
in turn depend on the previous processing history, concentration, flow rate,
membrane type and configuration, system design, temperature, and membrane
condition. The main thrust of the research planned for this coming year
will focus on these flux rates so equipment and energy requirements of the
various options may be more accurately determined for comparison with the
more obvious methodology of vacuum evaporation followed by spray drying. A
final goal of the project will be to design a complete plant and perform its
detailed cost analysis. An additional goal is to review progress on a
regular basis with industrial leaders so those portions of the work which
123
-------
Juice (4 % solids)
Wet Pulp
(Opt.) Liquid Cyclone
or
(Opt.) Clarifier
Slud g e
Pulp Drier
Ultraf iltrat ion
Retentate
Spray Drier
Drie d Pulp
High Protein
Juice Solids
Reve rse Osmosis
an d / or
Vacuum Eva p •
Water
A
I
(Opt.) Liquid Concentrate
for Animal Feed
Spray Drier
Low Protein
Juice Solids
Figure 3. Juice Processing Flow Chart with Ultrafiltration
-------
IX)
en
Juice ( 4 % solids )
Acid
Heater
Wet Pulp
Li quid Cyclone
or
Clarifier
Sludge
Pu Ip Drier
Oi CM
( Opt.) Liquid 8| S[_
Fertilizer
Spray Drier
Dried Pulp
High P r otein
Juice Solids
Reverse Osmosis
a n d / o r
Vacuum Evap.
Water
(Opt.) Liquid Concentrate
for Animal Feed
(Opt.) Liquid Concentrate
for Animal Feed
Spray Drier
Low Protein
Juice Solid s
Figure 4. Juice Processing Flow Chart with Heat Precipitation
-------
r\>
01
Juice (4 % solids)
Wet Pulp
Liquid Cyclone
or
Clarifier
Sludge
Pulp Drier
Dried Pulp
Reverse Osmosis
and/or
Vacuu m Eva p.
Water o|
(Opt.) Liquid Concentrate
for Anim a I Feed
Spray Drier
Ju i ce Solids
Figure 5. Juice Processing Flowsheet without Protein Fractionation
-------
have application to existing hardware can be implemented without delay.
FOAMING PROBLEMS
One of the properties of potato juice is that it foams readily. Samples of
potato juice were transported to the Cornell Machine Co. plant in Spring-
field, NJ to evaluate their "Versator" defearning machine. The juice was
agitated to produce foam and processed in their eight and sixteen inch dia-
meter machines which operate on the principle of sheeting a thin film of
juice and foam together over a concave disk spinning in a vacuum chamber.
The larger machine proved much more effective and has been ordered for
addition to the pilot equipment as shown in Figure 2.
Recent tests on unfractionated juice have shown that foaming within the
vacuum evaporator can be controlled with Down Corning FG-10 antifoamer used
at 300 ppm.
SUMMARY
••* «..
A low water process with no process water discharge has been developed for
the production of starch, pulp, and protein meal from cull potatoes. The
process currently produces starch of nearly equal quality to that produced
by traditional methods. Ongoing research will continue to improve starch
whiteness and characterize membrane flux rates in juice processing ultra-
filtration and reverse osmosis operations. Results of this work will lead
to a complete plant design and cost analysis.
127
-------
1. TREADWAY, R. H. Potato Starch. Potato Processing ed. by Talburt, W.
F. and Smith, 0., AVI Publishing Co. (1959).
2. TREADWAY, R. H. Manufacture of Potato Starch ed. by Whisler, R.L. and
Paschall, E. F., Starch: Chemistry and Technology Vol. II, Academic
Press (1967).
3. HICKS, C. P. Starch Refining 2 - Quality, Yields, and Equipment.
Process Biochemistry 5(7):30 (1970).
4. HOWERTON, W. W. and TREADWAY, R. H. Manufacture of White Potato Starch.
Industrial and Engineering Chemistry 40(8):1402 (1948).
5. HEMFORT, H., HUSTER, H. and Heiineier. Low Water Consumption in Preparing
Potato Starch. Reviewed by Peterson, N. B. Edible Starches and Starch-
Derived Syrups. Noyes Data Corp., Park Ridge, NJ (1975).
6. STROLLE, E. 0., CORDING, J., JR., and ACETO, N.C. Recovering Potato
Proteins by Steam Injection Heating. J. Agr. Food Chem. 21(6):974
(1973).
7. ANON. Protein Recovery from Potato Starch Process Biochemistry, p. 51
of May 1968.
8. KAMPHUS, G. G. Verbal Communican. Stork-Friesland, Stationsweg 84,
Gorredijk, Holland (1975).
9. STABILE, R. L., TURKOT, V. A., AND ACETO, N. C. Economic Analysis of
Alternative Methods for Processing Potato Starch Plant Effluents.
Proceedings of the Second National Symposium of Food Processing Wastes,
Denver, CO (1971).
10. DISBOROUGHS, S. and WEISER, C. J. Protein Comparisons in Selected
Phureja-Haploid Tuberosum Families. Department of Horticultural
Science, University of Minnesota, St. Paul, MN 55108.
11. PORTER, M. C. Concentration Polarization with Membrane Ultrafiltration.
I & EC Product Research and Development 11:234 (1972).
12. GOLDSMITH, R. L. and HORTON, B. S. Membrane Processing of Cottage
Cheese Whey for Pollution Abatement. Final Report of EPA Project
12060 DXF (1971).
128
-------
CHANGES IN ORGANIC AND INORGANIC CONSTITUENTS OF WASH WATER
UPON RECYCLE IN A PROTOTYPE LEAFY-GREENS WASHER*
by
_ „ , ** ** *** ***
R. C. Hoehn, P. B. Geering, M. E. Wright, and W. H. Robinson, Jr.
INTRODUCTION
Food-processing industries reportedly are one of the major consumers of
fresh water in the United States, using about 99 billion gallons annually.
Fresh water not only is costly but also the supply sometimes becomes
inadequate, most often when wells are the source, and especially in rural
locations where industries provide their own water. Restrictions on
industrial discharges under the provisions of the National Water Quality
Act Amendments of 1972 (Public Law 92-500), coupled with increasing costs
of obtaining fresh water, have caused most industries to study their
operations and find ways to minimize water consumption.
Industries that process leafy vegetables use large volumes of water,
mainly during four operations: 1) washing; 2) blanching; 3) cooling, and
4) transporting product. Three types of washers are most commonly used in
commercial operations: immersion, spray-belt, and rotary-spray. Estimates
of wash-water requirements per unit weight of product range from 50 per
cent of a total 3 to 5 gallons per pound (gal/lb) [ 25.0 to 41.6 liters per
kilogram (jt/kg)] necessary for complete processing (2) to 73 per cent of a
total 1.1 to 1.5 gal/lb ( 9.1-12.5 £/kg) cited by Bought3)
One obvious way to conserve water in a greens-washing operation is
to modify the system in some manner to permit recycling of the wash water
itself. Of course, the chief concern is that the recycled water be of
This investigation was supported in part by funds from the Environmental
Protection Agency, Food and Woods Product Branch, under Grant No.
S802958-01-0.
Respectively, Associate Professor and Graduate Research Assistant,
Civil Engineering Department, Virginia Polytechnic Institute and
State University, Blacksburg, Virginia 24061.
Respectively, Associate Professor and Graduate Research Assistant,
Agricultural Engineering Department, Virginia Polytechnic Institute and
State University, Blacksburg, Virginia 24061.
129
-------
sufficiently good quality throughout the period of use to insure that the
product quality will not be degraded beyond acceptable limits. However,
relatively little is known about the effects of recycling on product
quality, though water conservation is practiced in other ways in processing
operations, so in 1973 a proposal was made to the Environmental Protection
Agency that a project be funded to design and evaluate a full-scale,
immersion-type washer that would permit the recycling of wash water during
greens-washing operations. The proposed system was to be a further modifi-
cation of one constructed earlier that had been shown to be more effective
than conventional systems for removing insects and trash from leafy greens.^'
The proposed project was approved to begin in May, 1974, and the washer was
tested during the period from October through December, 1975.
The objective of this paper is to describe the modified system and to
present a summary of data showing its performance during five trials, four
when collard greens were being processed and one when spinach was being
processed. The evaluation was conducted at Exmore Foods, Incorporated, a
frozen-food company on the Eastern Shore of Virginia. Detailed information
concerning water and product quality, water usage rates, and wastes generated
per ton of vegetable processed during the five trials will be presented.
THE WASHING SYSTEM
A diagram showing water- and product-flow routes is presented in
Figure 1. The two washers, designed to wash 4,000 Ib (1773 kg) of product
per hour (hr), were each 4 feet (ft) 1.1.22 meters (m)] wide and 16 ft (4.88 m)
long. The bottoms were comprised of three sections, each V-shaped to
facilitate the collection and removal of grit when the washers were drained.
The maximum depth in each tank was 3 ft (0.92 m), and the nominal capacity
of each was 688 gal [2.60 cubic meters (m3)]. Water in each washer was
recycled through a separate settling tank 8 ft (2.44 m) long by 4 ft (1.22 m)
wide, with a maximum depth of 4 ft (1.22 m) and a nominal capacity of
718 gal (2.71 m3).
Both washers were identical and were operated in series. Each was
equipped at the input end with three banks of sprayers (4 nozzles each),
one bank placed at water level and two above the incoming product. One of
the overhead banks of sprayers was later taken out of service in order to
maintain high pressure on the nozzles while reducing the overall recircu-
lation rate from 200 gpm(757 Jl/min) to approximately 125 gpm (473 &/min).
This was necessary because it was discovered that the washer drain system
and the return sump pump (described later) could not handle the larger flow.
As product entered, it was spread and vigorously agitated by the
sprayers and was pushed toward the first of three,center-mounted, paddle
wheels covered with flattened, expanded metal. The three, revolving
paddle-wheel drums propelled the product through the washer, alternately
submerging and releasing it. Insects and leaf fragments floated to the
surface inside each drum while product was submerged, and water, forcefully
dispensed through a stationary bank of three spray nozzles (positioned
inside each drum) prevented leaves from becoming entangled in the expanded-
130
-------
i
i
'
i
WASH
2
<8)
j
WSHI
B» ^BB
«t-
B&-
•t-
ER
i
i
«*v
B>-
«*-
ER
1
i
4
SJ
— «•»-
_ — -*-
\
^^^^^ i
— «^-
— — -^-
MBMH^
i
mf^mmm
FRESH
WATER
INPUT
i
SETTL
T*
— »••
1
1
1
1
i
(S3
.ING
iNK 2
r t i
1
t
OVERFLOW FROM
SETTLING TANK 2
Q TO SETTLING TANK 1
®
SETTLING
TANK 1
i
r
n PR<
-
-PRODO
-WATER
3DUCT S
DRAIN TO
WASTE
CT FLOW
FLOW
AMPLING SITES
0 WATER SAMPLING SITES
O FLOW METERS
Figure 1. Diagram of experimental recirculating washer system.
131
-------
metal covering and provided additional cleaning. The sprayer system in
each washer was designed to operate at a pressure of 35 pounds per square
inch (psi) [2.5 kg per square cm (kg/cm2)], 57 per cent of the flow being
dispensed through the sprayers at the head of each washer. An exit
conveyor— made of an open-mesh, plastic belting with flights every 2 ft
(0.6 m)—carried product out of each washer.
Water and floating trash that collected inside the rotating drums flowed
out of the washers through surface side-drains into a trough leading to a
box containing a submersible sump-pump. This pump forced water into the
settling tanks through a moving-belt filter made of polyester screen that
trapped trash and carried it to a collection box. Streams of compressed
air at the end of the conveyor were directed upward through the belt
across its width to force trash into the collection box. Each settling
tank was 8 ft (2.44 m) long, 4 ft (1.22 m) wide and a maximum of 4 ft (1.22 m)
deep and was baffled to prevent floating debris from getting into the pump
that forced water back to its respective washer. The nominal capacity of
each tank was 688 gal (2.60 m3) and provided a detention of approximately
7 min for grit to settle when the recirculation rate was 100 gpm (378.5 £pm).
Fresh, chlorinated water was added to the system at only one point:
the second settling tank. The overflow was carried through an HS-flume into
the first settling tank. The volume in excess of that required to make up
for losses in each washer, caused by carry-over on the product.»was discharged
as waste from the first settling tank through a second HS-flume. Flow
meters were installed to permit monitoring of fresh-water input and recircu-
lation rates, which could be controlled by opening or closing gate valves,
through each washer-settling tank system. The HS-flumes were equipped with
continuous stage-recorders and were calibrated so that depth of flow through
the flumes could be converted to volume rate of flow-
An overhead photograph of the system, along with the field trailer and
rotary tumbler used to remove some of the grit and insects from the unwashed
product, is presented in Figure 2. Not shown are the components of the
conventional washing system that interfaced with the experimental system.
Prior to entering the first washer of the prototype system, dry product was
carried by conveyor belts through inspection lines where yellowed and
damaged leaves, weeds, and any other foreign materials were removed by hand.
Product feed rates into the washing system varied throughout the day,
mainly as a function of the condition of raw product as it came in from the
fields. The more foreign material and unacceptable product present, the
slower the feed rate had to be to insure thorough sorting.
METHODS
The prototype of the experimental system interfaced with the commercial
operation so that greens from the dry inspection belts could be conveyed
to either the experimental or conventional washers. When the experimental
system was in operation, greens passed through it and then were rewashed
and further processed by the conventional system. Because the two systems
were interfaced, the experimental operation was subjected to the same
variables one would encounter in a commercial operation (delays due to
132
-------
Figure 2. Components of experimental washing system: 1. Field trailer. 2. Ro-
tary tumbler. 3. Conveyor from tumbler to dry inspection belts. 4. Washer No.l.
5. Side-drain trough to sump. 6. Conveyor for product from Washer No. 1.
7. Washer No. 2. 8. Settling tank No. 2. 9. Fresh-water influent. 10. HS-
Flume for monitoring flow between settling tanks. 11. Discharge of sump pumps
from Washer No. 1. 12. Conveyor belt filter for removing larger particulates.
13. Trash collector. 14. Settling tank No. 1. 15. HS-Flume for monitoring
waste discharge from Settling tank No. 1.
133
-------
equipment failure, work breaks, etc.)- One disadvantage was that the type
greens available at the times chosen to operate the experimental system
was dependent upon harvesting schedules, and, because schedules could not
be rigidly followed, the original plan of washing a variety of greens in
the experimental system was not possible. However, replicate washings of
one type green did serve to demonstrate the variations one might encounter
in the quality of both wash water and final product resulting from differences
in several cuttings of the same green.
Description of a Typical Trial
Immediately before processing was to begin, the operation of the
experimental system was checked, and water-flow rates were adjusted and
recorded. The time when product feed to the system actually began was
recorded, and soon thereafter, usually 15 min later, samples of water and
product were collected for analysis at sites shown in Figure 1. Samples
then were collected at hourly intervals throughout the entire period of
actual plant operation. Many tests of water quality were performed on site
between sampling periods, but samples to be used for some analyses were
preserved and returned to the university laboratory.
The total number of packages of processed product were recorded
hourly, and periodically meters were read to determine recirculation rates
and fresh-water input rates. At the end of the trial, trash collected
from the washing system was collected and weighed, the tanks were drained,
and the grit was collected. Samples of grit, trash, and frozen product
were returned to the university laboratory for analysis.
Water and Product Analyses
Water quality analyses included biochemical- and chemical oxygen
demand (BOD and COD, respectively), total chlorine residual, total and
volatile suspended solids, and total bacterial population densities as
determined by plate counts. Standard Methods for the Examination of
Water and Wastewater(5) Was followed for determinations of BOD, COD,
chlorine residual (by the amperometric method), solids, and total plate
count.
A portion of each product sample was frozen for later analyses of
residual insects and bacteria present on the greens. Insect counts were
determined by a gasoline-flotation test used routinely at Exmore Foods,
Inc. and described by Frey(6/. Product bacterial densities were determined
also by methods used at Exmore Foods, Inc., namely by plating aliquots of
serial dilutions (on Total Plate Count Agar) made after vigorously shaking
11 grams (g) of product in 90 milliliters (ml) of sterile, buffered, dilution
water and counting the colonies that developed after 24-48 hr incubation.
Another product portion was weighed immediately after collection,
blotted with paper towels to remove excess water, and reweighed. The
samples then were frozen and later were dried at 105 degrees Celsius (°C)
for 24 hr and reweighed. The moisture content of each sample was calculated
134
-------
from the relationship:
w
W + ¥.
w d
where: M = moisture content on a wet basis
W = weight of water
W
W, = weight of dry matter.
Moisture contents of the packaged product also were determined, and
reasonable estimates of the total weight of product processed per hour
could be calculated by multiplying the total package weight by a fraction
obtained by dividing the moisture contents of raw and finished products.
Product feed rates through the washing system could be determined, and the
observed changes in product- and water quality could be related to the
total, fresh weight of product. This method of expressing data permitted
more valid comparisons from trial-to-trial than if product- and water
quality changes had been expressed as functions of time because, as has
been mentioned previously, product feed rates varied considerably.
Grit Analysis
Grit— including sand, silt, and clay— was collected at the end of
each trial, dried at 105°C for 24 hr, and weighed. A particle-size
analysis then was performed: the standard hydrometer method(7) was USed
for particles smaller than 50 nanometers (nm) and a sieve analysis(8)
was used for larger particles.
RESULTS AND DISCUSSION
Operational and waste-characterization data describing the washing
system and wash-water quality, respectively, will be presented in this
section, along with representative data to illustrate the following:
1) the effectiveness of the first washer-settling tank subsystem in
the total washing operation,
2) the effects of different cuttings of collards in altering wash-
water quality,
3) the variations in water quality caused by the action of a washer
and settling basin,
4) the relative variations in wash water quality observed when spinach
and collards were processed, and
5) product quality as reflected by bacterial densities and insect
counts.
In addition, the wastes generated per unit weight of fresh product processed
will be presented and compared to data presented by Bough.
135
-------
Operational Data
Table 1 shows specific details of the five evaluations of the
experimental washing system. The water usage by the experimental system,
expressed as volume per unit product weight, was from 2 to 3 times less
than that estimated by Bought) for a commercial cannery and from 3 to 5
times less than an estimate given by Whittington et^ al^' ', the latter
being for the Exmore Foods operation.
TABLE 1. PERFORMANCE DATA FOR THE PROTOTYPE WASHING
SYSTEM DURING FIVE TRIALS
Processing
Trial
#1**
#2
#3
#4
#5
Time
4.
5.
6.
6.
4.
,hr.
0
8
8
8
2
Processed
Product
Weight*
i
Tons
4.9
11.0
9.5
8.7
4.4
103 kg
4.4
10.0
8.6
7.9
4.0
Average Product
Feed Rate
Tons/hr
1.2
1.9
1.4
1.3
1.0
103
1
1
1
1
0
kg/hr
.1
.7
.3
.2
.9
Water Use
per Unit
Product
gal/lb
0.62
0.42
0.48
0.55
0.75
A/kg
5.16
3.50
3.99
4.58
6.25
Flow Rate
of
Makeup Water
gal/min
12.7
18.3
15.2
16.6
15.2
£/min
48.1
69.3
57.5
62.8
57.5
(Trials #1-4: Collards; #5: Spinach)
*
**
Fresh weight
Estimated data involving product weight based on the average, fresh
weight of other collard cuttings washed during the project
It should be pointed out, however, that the flow data in Table 1 are measured
values, not estimated, and the quantity of water conserved when operating the
prototype actually is more apparent when the volume rates of flow are compared.
The two conventional washers at Exmore Foods are operated at flow rates of
35 gal/min (132.5 fc/min) each for a total of 70 gal/min (265 £/min). Bough^3)
estimated 74-84 gal/min (280-318 Jl/min) and 180-200 gal/min (681-757 S./min)
used in dunker washers and reel washers, respectively, at a commercial
cannery. The prototype recirculating system, on the other hand, was operated
at an average rate of only 15.6 gal/min (59 £/min), a considerable savings
indeed (77 per cent less than at Exmore Foods).
136
-------
Chemical Characterization of Waste Water
Table 2 presents selected chemical characteristics of waste water in
settling tank No. 1 at the end of a processing day. The low and high
TABLE 2. CHEMICAL CHARACTERISTICS OF WASH WATER WASTED FROM
SETTLING TANK NO. 1 AT THE END OF THE PROCESSING DAY
Waste Component,
mg/Z
TSS
VSS
COD
BOD5
Average
171
110
356
80
Collards
Low
109
61
262
66
High
220
155
439
97
Spinach
689
57
211
46
Processed Fresh Weight
Tons 8.5 4.8 11.0 4.4
103 kg 7.7 4.4 10.0 4.0
concentrations were observed, respectively, on the days when the least and
greatest mass of collards were processed. Note the relative magnitude of
total suspended solids (TSS) concentrations present when spinach and collards
were processed. The much higher concentration obtained by washing spinach
reflects the difference in leaf shapes of the two greens, spinach having
a convoluted surface that can hold more dirt than the relatively smooth-
surfaced collard greens. Figure 3 clearly illustrates the differences.
Concentrations of three of the four chemical characteristics in
Table 2 reported by Bough(•>) in a study at a commercial cannery are presented
in Table 3. In general, the concentrations of the various waste components
were much greater at the cannery where Bough's work was conducted. The
reason is not clear, and the differences are surprising because the experi-
mental washing system in this study was a recirculating type, which permits
waste components to accumulate and become highly concentrated.
The differences in Bough's data and that derived from this study are
just as great when compared on the basis of the product weight processed.
Table 4 shows a comparison of the waste loads per ton calculated from data
collected during the two studies.
137
-------
CO
o
800
700
600
500
400
300
200
100
SITE
SPINACH
COLLARDS
•« SITE 5
er /
___--/ ___
>»SITE I
- —o^ SITE 5
•^
1234567 8 9
ACCUMULATED PRODUCT INPUT, kgXIO3
Figure 3. Variations in total suspended solids showing differences when
collards and spinach were washed. Water collected at Sites 1 and 5,
respectively, were settled water from the first settling tank entering
washer No. 1 and unsettled water entering the second settling tank from
washer No. 2.
138
-------
TABLE 3. CHARACTERISTICS OF WASTE EFFLUENTS FROM THE WASHING OF
COLLARDS AND SPINACH AT A COMMERCIAL CANNERY
REPORTED BY BOUGH (3)
Waste
Component
TSS
Dunker Washer
Reel Washer
COE)
Dunker Washer
Reel Washer
BOD5
Dunker Washer
Reel Washer
Waste Concentration, mg/£
Col lards
Mean
261
62
; 633
325
223
122
s*
116
22
268
125
115
51
Spinach
Mean
973
159
1322
528
263
185
s"
578
,28
366
155
37
,64
Standard deviation
TABLE 4. A COMPARISON OF WASTE LOADS" GENERATED IN THE PROTOTYPE
WASHING SYSTEM WITH THOSE CALCULATED BY BOUGH (1973) IN A
COMMERCIAL CANNERY. DATA GIVEN ARE POUNDS WASTE
: PER TON OF FRESH PRODUCT*
Component
TSS
VSS
COD
BOD,.
Col lards, Ib/ton
This Study5"* Bough's Value"1"
0.72 1.79
0.43
1.72 9.37
0.42 2.51
Spinach, Ib/ton
This Study Bough's Value"1"
4.89 5.63
0.42
1.81 12.37
0.35 3.41
•kit
J
or kg per 2 x 10 kg of product
Average of three trials
Dunker Washers plus Reel Washers
139
-------
The waste loads in this study were the sums of three components: 1)
waste in wash water carried out of the system on the product, 2) waste
discharged continuously from the first settling tank, and 3) waste
remaining in the washers and settling tanks at the end of the processing
day.
(3\
If, in fact, Bough's estimates of water usage were reasonably
accurate (he stated that they were based on intermittent estimates of flow),
the best explanation for the observed differences in wash-water characteris-
tics and waste loads (Tables 3 and 4) is that there were vast differences in
the cuttings of greens processed during the two studies. Differences in
organic constituents and loads (COD and 8005) could be explained if the
commercial washers damaged the product more than the prototype recirculating
system, but there is no evidence that such damage actually occurred. More
data of the type presented in Tables 3 and 4 will have to be collected
before the observed differences between Bough's results and those of this
study can be explained.
One additional comparison among data in Table 4 should be made; namely,
that for waste loads of spinach with those of collards. Note that the total
loads for three of the four waste water components (VSS, COD, and 6005) per
ton of product processed, were similar in magnitude regardless of whether
spinach or collards were being processed. As was discussed previously,
however, the suspended solids load contributed by spinach was much greater
than that contributed by collards.
Relative Effectiveness of the Two Washer-Settling Tank Subsystems
Figures 4 and 5 illustrate what was observed in every trial, namely
that water in the first washer-settling tank subsystem became progressively
"dirtier" than that in the second subsystem as more and more product was
processed. There are two reasons. First, most of the cleaning actually
did occur in the first subsystem; second, fresh make-up water was added to
settling tank No. 2, providing a continuous dilution of the wash water in
the second subsystem. A mass accounting within the total system, based on
concentrations in the units at the end of a processing day, showed that an
average 71 per cent of the suspended solids and organic matter was contained
in the first washer and settling tank. Figures 4 and 5 clearly reflect
this fact and, in addition, illustrate that the rate of increase in waste
water constituents was much greater in the first subsystem. Note also that
the water quality changes are expressed as functions of accumulated product
input rather than operating time, because feed rates in the several trials
varied considerably making invalid any comparisons based on time. Figure 6,
showing the product-feed rate for one trial, is typical of the observed
variations in other trials.
Variations in Water Quality with Different Cuttings of Product
In retrospect, it was fortunate that 80 per cent of the effort in this
project was with one type green, because the data showed that wash-water
quality can vary considerably even when the same product is being processed.
t) also observed considerable variation, as is evidenced by the high
140
-------
1000
OSITE I
A SITE 2
ASITE3
• SITE 4
S SITE 5
SITES
ACCUMULATED PRODUCT INPUT, kg x 10'
Figure 4. Variations in total suspended solids concentrations of water from
all six sampling sites during the washing of spinach. Sites 1, 2, and 6 were
within the first washer-settling tank subsystem. Sites 3, 4, and 5 were
within the other subsystem.
141
-------
250
OSITE I
A SITE 2
A SITE 3
SUE 4
SITES
D SITE 6
01234
ACCUMULATED PRODUCT INPUT, kgxIO3
Figure 5. Variations in chemical oxygen demand concentration of water from
all six sampling sites during the washing of spinach. Sites 1, 2, and 6 were
within the first washer-settling tank subsystem. Sites 3, 4, and 5 were
within the other subsystem.
142
-------
2500
2000
o»
1500
4s,
CO
1
u.
g
a.
1000
500
BREAK
2 3 4
' '•- "--- . . ^^ 'j
OPERATJNG TIME, hr
6
Figure 6. Variations in product flow rates during a typical processing day.
-------
standard deviations of the mean concentrations he reported (See Table 3).
Figures 7 through 9 illustrate differences in only three of the many
waste water characteristics monitored during this project when collards
were being processed. Data collected during the first trial do not appear
on any of these figures because no record was made of the cumulative product
input.
The reductions in TSS, evident in Figure 7, is typical of other
observations of solids and reflect the overall results of interactions
within the system between different settling rates of variable-sized
particles, dilution, and variations in cleanliness of different loads
of vegetables brought in from the fields during the day. Chemical oxygen
demand (Figure 8) and 8005 (Figure 9), indicative of the concentrations of
small particulate or dissolved organic matter, usually increased contin-
uously throughout the washing period. Differences in the absolute magni-
tudes of COD and 6005 concentrations were observed and probably reflect
differences in the ages of the various cuttings, a factor that affects the
thickness of the protective leaf cuticle and the general tenderness of the
leaves themselves.
Effects of Washing Action and Sedimentation of Water Quality
Figures 10 and 11 illustrate the effects that vigorous product
agitation had on water quality. The changes in COD brought about by
propelling the greens through the washer are reflected by the area between
the two curves. Water entering the washer (site 1) originated from the
settling tank; water leaving the washer at the side drains entered the
sump (site 2). Hence, differences in water quality at those two locations
represent differences induced by vigorous agitation within the washer.
During the fourth trial with collards and during the only spinach-washing
trial, the increases of COD caused by the washing action were on the order
of 10 to 20 per cent. The highest increases, about 40 per cent, were
observed early in the day when the water was relatively clean.
Figures 12 and 13 show the effectiveness of one of the settling tanks
in removing suspended solids. The area between the curves shown in each
figure represents the apparent changes in suspended solids concentrations
as the water moved through the settling tank, but these data do not
accurately reflect the total removal of suspended solids that actually
occurred. In the spinach-washing trial, approximately 132 Ib (60 kg) dry
weight of grit was collected, 63 per cent being found in settling tank No. 1
and 12 per cent in washer No. 1, indicating once again that most cleaning of
product occurred in the first washing system. The material in the settling
tank was mostly silt (55 per cent) and fine sand (34 per cent) while that
in the washer was mostly fine sand (70 per cent) and medium sand (20 per
cent). Therefore, while the tanks were effective traps for the larger grit
particles, the smaller ones, like clay, were present in water collected at
both the head (site 6) and back of the tank, causing the apparent difference
in total suspended solids to be quite small. The actual removal that must
have occurred, based on the weight of sediment collected from the settling
144
-------
20
O TRIAL 2
0 TRIAL 3
A TRIAL 4
8
10
ACCUMULATED PRODUCT INPUT , kg X 10
Figure 7. Variations in total suspended solids concentrations in settled wash
water from the first washer-settling tank subsystem on three occasions when
different cuttings of collards were being processed.
145
-------
500
o»
E
f»
§
<
400
300
o 200
100
O TRIAL 2
DTRIAL 3
A TRIAL 4
8 9
K>
ACCUMULATED PRODUCT INPUT, kgXIO"
Figure 8. Variations in. chemical oxygen demand concentrations in settled
wash water from the first washer-settling tank subsystem on three occasions
when, different cuttings of collards were being processed.
146
-------
300
200
O TRIAL 2
A TRIAL 4
0 I 2 3 4 567
ACCUMULATED PRODUCT INPUT , kg X
8
3
9 10
Figure 9. Variations in the five-day biochemical oxygen demand of settled
wash water from th6 first washer-settling tank subsystem on two occasions
when different cuttings of collards were being processed.
147
-------
E
o
I
a
§
UJ
o
O SITE !
SITE 2
TIME OF OPERATION,hr
Figure 10. Variations in concentrations of chemical oxygen demand (COD)
in water in the first washer. The area between the two curves represents
the COD added to the wash water as collards passed through during one day's
operation.
148
-------
250
012345
TIME OF OPERATION, hr
Figure 11. Variations in concentrations of chemical oxygen demand (COD) in
water in the first water. The area between the two curves represents the
COD added to the wash water as spinach, passed through during one day's
operation.
149
-------
6
TIM EOF OPERATION.hr
Figure 12. Variations in concentrations of total suspended solids in wash
water.,hefore and after passage through the" first settling basin.. The area
between the two curves represents the apparent removal of suspended solids
by the settling basin during one day's operation when collards were being
processed.
150
-------
O SITE I
A SITE 6
01 2345
TIME OF OPERATION.hr
Figure 13. Variations in concentrations of total suspended solids in wash
water before and after passage through the first settling basin. The area
between the two curves represents the apparent removal of suspended solids
by the settling basin during one day's operation when spinach was being
processed.
151
-------
tank, was greater than. 2000 mg/SL per hour. From Figures 12 and 13, it can
be seen that the greatest removal observed at any time appeared to have
been less than 50 mg/£.
The distribution of grit in the system showed that most of the larger
sediment particles (sand) settled out in the washers before the water ever
reached the settling basin. The average detention time in the settling
tanks was approximately 7 min, but the intermittent flow of water from the
sumps to the settling tanks probably produced sufficient turbulence to
reduce the effectiveness of the settling tanks for trapping the smaller
particles, that is, the silt and clay(2 to 50 nm and less than 2 nm,
respectively).
The data showing changes in concentrations of volatile suspended
solids through the settling tank indicated that their removal was poor.
These solids are the tiny, leaf constituents, produced by the violent
washing action, which pass through the moving filter belts into the
settling tanks. Their specific gravity is much less than the inorganic
sediment and, while the volatile fraction of the grit recovered from the
bottom of the tanks was not determined, it was not unexpected, in light
of the particle-size distribution of the grit, that the organic particles
were recirculated and not settled. It is possible that chemical coagulant
aids, such as organic polymers, would be effective in increasing the
settling efficiency of these tiny particles; settling of the fine inorganic
particulates also would be aided.
Bacteriological Quality of Wash Water and Product
Table 5 is a summary of the observed changes in bacterial quality of
product and wash water through the washing system during four or the five
trials. No bacterial densities on product were determined during the first
trial washing of collards. Note from the table that the bacterial densities
in the washwater on some occasions increased even though chlorine was
maintained at a reasonably high level. The greatest increase was observed
when the chlorine concentration decreased to zero. In spite of the high
total plate counts in water during all trials (average 2.75 x 105 colonies
per ml, range 1.7 x 103 to 1.3 x 106) the product itself always was cleaned
well.
In addition to the presence of chlorine in the wash water, a second
factor may account for the decrease in product total plate counts. A foam
was observed to accumulate in the washing system during each washing trial,
the accumulation beginning very soon in the operating day and continuing to
increase throughout the day. It is suspected that the foaming action was
evidence that organic compounds leaching from the greens behaved as surface-
active agents and caused additional cleaning by detergent action. Thus,
even though the wash water generally became more heavily contaminated with
bacteria as more and more product was washed, the product itself was always
cleaner after washing than when it was first introduced into the system.
152
-------
TABLE 5. COMPARISONS* OF BACTERIAL POPULATION DENSITIES (TOTAL
PLATE COUNTS) IN WASH WATER AND ON PRODUCT BETWEEN SAMPLES
TAKEN FROM THE HEAD OF THE FIRST WASHER AND THE BACK OF
THE SECOND WASHER AT THE BEGINNING AND END OF A
GREENS-WASHING TRIAL
Trial Water ProductTotal Chlorine, mg/fc""
Beginning*** End*** Beginning*** End*** Beginning End
2 +243% +380% -11% >-99% 2.6 0
3 -64% -68% -96% -68% 0.4 2.3
4 -50% -61% >-99% -67% 1.2 2.3
5 -58% +76% >-99% -68% 1.3 1.6
*
A plus sign before a number Indicates an increase in bacterial density
from the head to the back of the washing system. A minus sign indicates
a decrease.
**
Residuals measured amperometrically and are the concentrations measured
at the back of the second washer at the beginning and end of the
processing period.
"Beginning" signifies a sample taken within the first hour of operation.
"End" signifies a sample taken immediately before termination of the
trial.
The differences in the reductions on product at the beginning and end
of a washing operation, shown in Table 5, are somewhat misleading because,
for some unexplained reason, bacteria populations on the last loads of
product being fed to the system were 25 times higher than those on the first
loads, averaging 3.3 x 10^ colonies per gram and 1.3 x 10^ colonies per
gram, respectively.
While the total counts of bacteria on product leaving the second washer
ranged from 4.6 x 103 to 1.1 x 106 per gram, it should be remembered that
the product had not been blanched. Blanching does reduce the population
densities to much lower levels.
Insect Removal Efficiency
The insect densities on the fresh cuttings of collards and spinach that
were processed during this project were too low for there to be any real
test of the insect-removal efficiency of the experimental system. Fall
crops, compared to spring, are always less heavily infested with insects.
153
-------
The maximum number of insects observed during four collard-processing
operations was 2 per 100 g of product. During the spinach-processing
trial, the maximum observed was 12 per 100 g. Frey et jgJU (4) reported
an average of 65 aphids and 63 insect fragments per 100 g of spinach
harvested in spring on the Eastern Shore of Virginia. Washings of
spring crops with the experimental system used in this project are
planned for" the future.
SUMMARY AND CONCLUSIONS
It may be concluded that the experimental washing system, as a
whole, performed well during the four trials with collards and the one
with spinach. While the water quality did deteriorate as more and more
greens were processed, cleaning of the greens while they passed through
the system was evident at all times. Reduction of bacterial population
densities on the greens were observed in all trials, probably because
chlorine residuals were maintained at high levels and because of a
detergent action in the washwater induced by dissolved organic compounds
leached from the vegetables.
Conclusions derived from this study are:
1. Most of the actual cleaning of product occurred in the first
subsystem; an average 71 per cent of organic and inorganic waste water
constituentswere present in the first washer and settling basin.
2. The settling tanks were effective in removing small particles
such as silt and fine sand. In one trial when sufficient grit for analysis
was recovered, 63 per cent of the total was found in settling tank No. 1.
Washer No. 1 collected the most of the fine and medium sand (12 per cent
of the total weight of grit collected). Differences in total suspended
solids concentrations in the wash water from the head to the back of a
settling tank did not accurately reflect the actual solids removal that
occurred.
3. There were distinct differences in final wash water quality when
different cutting of collards were processed, the differences most probably
being a function of product maturity.
4. The water-usage rate in the experimental system, was, on the
average, 15.6 gal/min (59.0 Jl/min), which is 77 per cent less than that of
the commercial washers at Exmore Foods, Incorporated, where this study was
conducted* Water use per unit product was an average of only 0.56 gal/lb
(4.66 Jt/kg) compared to an estimated usage of 1.5 - 2.5 gal/lb ( 12.5-20.8
SL/kg) at Exmore Foods.
5. The average chemical concentrations of the wash water being dis-
charged from the system at the end of the trials were high, both in organic
and inorganic suspended solids (110 and 61 mg/fc, respectively, when collards
were washed and 632 and 57 mg/&, respectively when spinach was washed).
The magnitudes of the solids concentrations probably reflect the fact that
154
-------
there was considerable turbulence in the settling tanks because of the high
recirculation rates within the system and because the lighter particulates
(clay and small leaf fragments, e.g.) did not settle well.
6. The average concentrations of COD and BODs in waste water being
discharged from the system at the end of the trials were similar to those
of weak municipal sewage, averaging 356 and 80 rng/A, respectively, when
collards were processed, and 211 and 46 mg/&, respectively, when spinach
was washed.
7. Concentrations of both organic and inorganic compounds in wash
water were much lower than those reported by BoughA^) t as were the waste
loads generated per unit weight of product processed. The best agreement
of data between the two studies was for total suspended solids. Bough's
values for organic waste loads (COD and BOD, Ib/ton) were from 5 to 6
times higher than those calculated for this study.
8. Recirculation of wash water in immersion-type, leafy-green
washing systems is a promising modification of existing processing
methods for reducing water consumption and concentrating waste loads
so that they can be more easily treated.
ACKNOWLEDGEMENTS
The authors would like to thank the Environaeatal Protection Agency
for financial support through Grant Number S802958-01-0 that made this
project possible. They also are indebted to the management and staff of
Exmore Foods, Incorporated (Exmore, Virginia) for permitting this study
to be conducted at their plant and for their help in the actual execution
of the project.
155
-------
REFERENCES CITED
1. NATIONAL CAHNERS ASSOCIATION. Liquid wastes from canning and freezing
fruits and vegetables. Water Pollution Control Series No. 12060 EDK-08/71.
U. S. Government Printing Office. Washington, D. C. (1971).
2. WHITTINGTON, J., PEARSON, S., NOTTINGHAM, R., and SCHOENEMANN, D.
Managerial Staff at Dulany Foods, Inc., Personal communication. Exmore,
Virginia (1973).
3. BOUGH, W. A. Composition and waste loads of unit effluents from a
commercial leafy greens canning operation. J. Milk Technol. 36: 544
(1973).
4. FREY, B. C., WRIGHT, M. E., and HOEHN, R. C. Modification of a leafy
vegetable immersion washer. Trans. Amer. Soc. Ag. Engr. 17: 1057 (1974).
5. AMERICAN PUBLIC HEALTH ASSOCIATION. Standard methods for the examination
of water and waste water, 13th edition, New York (1971).
6. FREY, B. C. Modification of a leafy vegetable immersion washer.
Unpublished Master's Thesis. Virginia Polytechnic Institute and State
University. Blacksburg, Virginia (1973).
7. DAY, P. R. Hydrometer method of particle-size analysis. In "Methods of
Soil Analysis, Part I", Amer. Soc. Agron., Madison, Wisconsin (1965).
8. BUCKMAN, H. 0. and BRADY, N. C. "The Nature and Properties of Soils",
7th ed. The MacMillan Company, New York (1969).
156
-------
TOMATO FLUME WATER RECYCLE WITH
OFF-LINE MUD REMOVAL
George E. Wilson1, Wally R. Rose2, Jerry Y. C. Huang1
SUMMARY AND CONCLUSIONS
The wastewater management objective for a tomato processor is to imple-
ment sound and performance guaranteed systems which minimize wastewater
related costs. Using performance parameter values found in this project,
it was demonstrated that installation and operation of an in-plant water
recycle system with off-line mud removal would result in approximately
50% savings in the total annual wastewater related costs. For the 35 ton/hr
plant evaluated, annual savings would amount to approximately $47,000.
The '75 season performance parameter values were obtained from investi-
gation of four modes of operation; conventional cleaning; conventional cleaning
with water recycle; disc cleaner with water recycle; and, disc cleaner with
recycle and chemical coagulation-flocculation. Water consumption and total
solids balances were made on each mode.
Not surprisingly the daily average tonnage of tomatoes processed in-
creased substantially with disc cleaning and water recycle as compared to
the conventional system. An increase of 26% in the tonnage of tomatoes
processed was realized with the disc cleaner with water recycle and chemical
flocculation. These increases in the daily tonnage of tomatoes processed
may be primarily due to the virtual elimination of solids accumulation in
the dump tank with consequent impaired product flow. No incident of
temporary shutdown of shift operation for dump tank clean-up was encountered
during the modes of operation with water recycle.
With respect to the water consumption, the following findings were
established in this study:
1. The majority of daily water usage was operational (48-61%) followed
by clean-up (31-44%) and filling (6-8%). There were no significant
variations in percentage usage in the various modes of operations.
In all modes of operation approximately 7% was filling; approximately
55% operational; and approximately 39% clean-up.
2. A 26% decrease in the average total daily water usage was realized
when disc cleaner with water recycle and chemical flocculation
relative to the conventional system was applied.
1. EUTEK, INC., Process Development and Engineering, Sacramento, CA.
2. National Canners Association, Berkeley, CA.
157
-------
3. A decrease in the average unit water consumption rate relative to
counter-current flow conventional cleaning occurred when water
recycle measures were applied. A 41% decrease in average unit
water consumption rate was realized when the disc cleaner with
water recycle and chemical flocculation (164 gal/ton) was applied
relative to the conventional system (278 gal/ton).
With respect to the total solids balances, the following conclusions
were drawn:
1. With the water recycle measures implemented, the soil solids removed
from the dump tank per tonnage of tomatoes processed were signifi-
cantly reduced; the soil solids lost to the sewer per tonnage of
tomatoes processed were reduced substantially.
2. The estimated soil solids incoming to the plant per unit weight
of tomatoes processed ranged from 10 to 20 Ibs/ton, having an average
of 13 Ibs soil solids per ton of tomatoes processed. The total
soil loaded was estimated from the sum of soil solids which were
collected from the dump tank, lost to the sewer, and removed from
the sludge thickener. It appeared that the amount of soil to the
plant varied considerably, depending on the type of soil from which
the tomatoes were grown, the moisture content of the soil when
the tomatoes were harvested, and the method of tomato harvesting.
3. Efficient clarification of the thickeneK overflow requires surface
loading rates of less than 1,000 gpd/ft . Approximately one-
half of the gravity settleable soil solids overflowed from the
thickener at surface loading rates of 2,000 gpd/ft^.
INTRODUCTION
.With the advent of mechanical harvesting of tomatoes, tomato processors
noted an increase in the soil accumulations within the flume system. Most
of the soil accumulated in the initital flumes or bin dumps. Velocities
within the bin dump were insufficient to scour settled soil solids from the
base. The resultant accumulation of soil in the bin dump eventually
impaired product flow and required processing downtime to remove.
There are two widely practiced procedures for mitigating the accumula-
tion of soil in the bin dump. The first is to employ a high overflow rate
from the bin dump, this overflow discharging to the plant's sewer system.
The second procedure involves processing downtime to drain off excess
liquids and hand shovel the accumulated soil into an adjacent receptacle.
Several adverse impacts result from the current procedure for handling
bin dump mud. In the case of the high overflow rates, the excess water used
adds to the hydraulic surcharge to the sewer system from this seasonal industry.
The high soil loadings discharged to municipal treatment systems result in
operation and maintenance problems. As a consequence of the necessity for
158
-------
periodically shutting down the bin dump to remove accumulated soil, the time
requirements for processing a given tonnage of tomatoes are extended resulting
in reduction of the plant's productivity. This, in turn, results in further
excess water use through the down-time period and for the additional required
cleanup and washdown.
Project Objectives
During the 1974 processing season, a jointly sponsored EPA-NCA project
was undertaken to evaluate alternative water recycle system configurations.
Bin dump model studies were then undertaken in the spring of 1975 to develop
design data on an efficient system for intercepting soil solids in the bin
dump and transporting them to the solids removal system without interfering
with product flow.
The objective of the '75 season tomato water recycle project was to
demonstrate a water recycle system which when used in conjunction with a
normal bin dump operation would significantly reduce the adverse impacts
associated with current practices. An essentially closed loop recycle system
was employed. The recycled water was used to maintain adequate scouring
velocities within the bin dump without detrimentally affecting product flow.
A solids removal system was constructed within the closed loop to remove the
settleable solids and thereby prevent their accumulation within the bin
dump. This recycle system was expected to eliminate excessive water usage
related to high bin dump overflow rates as well as those related to clean-up
down-time and extended operational period for processing a given tonnage of
tomatoes.
Process Design, Operation, and Data Acquisition
Facility Design and Installation
The key process elements were a solids trapping false bottom; an
ejector for solids transport; a screen with screenings discharge hopper;
a soil solids separating swirl concentrator; a sludge thickener; and a
chemical coagulation-flocculation system. As shown on Figure 1, the soil
solids passed through the false bottom and were transported by an ejector
to the gravity screen. Here vines, rocks, and debris were separated and
the settleable soil solids in the water were routed to the swirl concentrator.
Solids within the swirl concentrator under-flow were further concentrated
within the sludge thickener. A mechanical device was provided to enhance
the gravity thickening process. Thickened sludge was removed from the
bottom of the thickener for final disposal.
The swirl concentrator overflow was recycled to the bin dump through
standpipes to provide a motive force for both the bin dump scour jets
and the solids transporting ejector. The sludge thickener supernatant
was returned as make-up water to the bin dump or was discharged to the
sewer. The general configuration of the required modifications to existing
bin dump tanks is illustrated on Figure 2. This configuration was developed
159
-------
SCREEN
SWIRL
CONCENTRATOR
RETURN WATER FROM
SORTING 8ELJS
LEVEL CONTROL
MAKE-UP WAT
UNDERFLOW
GRIT TUBE
THICKENER a
SOLIDS STORAGE
FJG.I FLOW DIAGRAM OF WATER RECYCLE SYSTEM
-------
STANDPIPE FOR fi . .
SCOUR MANIFOLD
.^-
HOPPER
ynptrE
VT
.-
EJECTOR
IE
. -BM DUMP LEVEL CONTROL
STILLING WELL
VNDPIPE FOR
XJR MANIFOLD
'N
1
k
"in
X'SCOUR MANIFOLD
,^ T-
- BIN DUMPE
BOTTOM}
>
:R AXIS
BIN DUMP-N
1 >
)
SUPPLY FROM
SCOUR
SUF=PLYFROM
EJECTOR
END VIEW SECTION
SIDE VIEW SECTION
FIG 2 FULL SCALE GRIT COLLECTOR a TRANSPORT SYSTEM DESIGN
-------
from model studies/ ' It is capable of intercepting and removing 80%+
of the settleable soil entering with the product without altering product
flow. The false bottom area was 4' x 4' with 1/2" tubes spaced at 1-1/4"
center-to-center.
The installed system is shown on Plates 1A and IB. The ejector-
transported soil-laden water was pumped to the gravity screen mounted
above the swirl concentrator. The screenings were collected in a gondola.
(2)
The screened water flowed by gravity to the swirl concentrator. '
Grit and sand in the incoming wastewater were separated and discharged with
the underflow from this unit. An orifice opening controlled the underflow
rate. As designed, the underflow rate was 100 gpm. With the maximum
design incoming flow rate of 500 gpm the swirl concentrator overflow rate
was approximately 400 gpm. Overflow was returned to the bin dump through
the scour jets and transporting ejector.
Underflow loaded the ten foot diameter gravity thickener at approxi-
mately 2,000 gpd/ft2. The thickener overflow rate was approximately equi-
valent to the swirl concentrator underflow rate. Level control was provided
at the dump tank to regulate the amount of thickener overflow returned
to the system. This prevented overflow from the bind dump. When the bin
dump was topped-off, thickener overflow was bypassed to the sewer. This
assured a single overflow point from the water recycle system to the sewer.
Only clarified water was discharged to the sewer.
In order to increase the thickener underflow solids concentration
a vibrating thickening mechanism was installed to break down bridging of
solids within the thickening zone. Vibration was activated by pneumatic ,
impactors. In addition to the internal mechanism an impactor was mounted
externally on the cone of the thickener to assist the flow of solids to
the cone apex. Pneumatic impactors provided impacts at 60-psi pressure
at controlled time intervals.
Thickener underflow was withdrawn periodically through a 6" diaphragm
valve. Underflow sludge was collected in gondolas for subsequent disposal.
Operational Modes and Monitoring Arrangement
During the study period, the following four modes of operation were
investigated:
1. Conventional cleaning without water recycle;
2. Conventional cleaning with water recycle;
3. Disc cleaning with water recycle; and,
4. Disc cleaning with water recycle with chemical coagulation-
flocculation.
162
-------
PLATE 1A VIEW OF IN-TAKE PUMP, SCREEN, SWIRL CONCENTRATOR,
SLUDGE THICKENER AND TUBE FLOCCULATOR
PLATE IB GENERAL VIEW OF WATER RECYCLING FACILITIES
163
-------
The monitoring program objective was to evaluate each of the four modes
of operation in terms of total water consumption and total soil solids removed.
Conventional Cleaning—The conventional system consisted of four
stages: dump tank, inside flume, distribution flume and, final rinse step.
Tomatoes were mechanically transported by elevators between stages.
As shown on Figure 3, counter-current use of water was employed in
the conventional system. The distribution flume water was recycled between
the distribution flume and tank-3. A portion of water equal to the final
rinse was pumped to the inside flume. Excess inside flume water flowed
by gravity to tank-1. There was an internal recycle loop between tank-1
and the dump tank. Excess water overflowed from tank-1 to the sewer.
A significant savings in water consumption had already been realized in
the counter-current mode of operation relative to systems in which each
stage operates independently. Composite water samples were collected from
the tanks on each loop such that water quality represented the water in each
of the flumes as designated 5W from the dump tank, 2W from the inside flume
and 6W from distribution flume. There were a total of five flow meters
monitoring the various flow rates.
The water samples were used to determine the solid concentrations
for mass balances. During the conventional cleaning mode a steal plate
was placed on top of the false bottom. Soil solids accumulated within
the dump tank were hand shoveled into gondolas for subsequent disposal.
Soil samples from the gondolas were collected for the solid concentration
determinations.
Conventional Cleaning with Hater Recycle—In this mode of operation,
the water recycle system was added to the conventional mode of operation.
Soil solids were intercepted by the false bottom and were transported by
the ejector and pump to the water recycle and off-line mud removal system.
On Figure 4 are shown the flow metering the sample collection points
for this mode of operation. There were a total of 8 flow meters and 10
sample points from which composite samples were collected for water quality
analyses.
Irregular deliveries of field harvested product occurred during this
mode of operation. As a consequence, the plant frequently operated with
partial shifts. Oftentimes the processed tomato tonnage was significantly
less than the plant's capacity. Thus, while effective interception, transport,
and removal of settleable soil was effected by the water recycle system,
its affect on the productivity of the plant and the full ramifications on
water use was not accurately assessed.
jDisc Cleaner With Mater Recycle —In this mode of operation, tomatoes
were transferred from the dump tank to the disc cleaner and then to the
distribution flume and the final inspection stage as shown on Figure 5.
164
-------
-{M} *-
-H-
LEGEND
'TOMATO PRODUCT
I 'PROCESS WAfER
'*"' MAKE UP WATER
'WATER METER
'WTER QUALITY SAMPLE
STATION
DUMP
TANK
INSIDE
FLUME
r
\
ORDINARY
DISTRIBUTION
FLUME
SUBMERSIBLE
FINAL RINSE
V
FINAL
.INSPECTION
SEWER
(FURTHER PROCESSING)
FIG.3-FLCW DIAGRAM OF CONVENTIONAL CLEANING SYSTEM
165
-------
SCREEN & SWIRL
CONCENTRATOR
SOLIDS
SOLIDS SLURRY
H • WATER METER
® • WATER QUALITY SEWER
SAMPLE STATION
*3> 'BINDUMP LEVEL CONTROL
(FURTHER PROCESSING)
FIG.4- FLOW DIAGRAM OF CONVENTIONAL CLEANING
WITH WATER RECYCLE SYSTEM
166
-------
HD-t-
SCREEN &
SWIRL
CONCENTRATOR
—MH
SLUDGE
THICKENER
—Bh
—B-
SOLIDS
DUMP
TANK
DISC
CLE'ANER
r
H
<,
_N
\
V.
7 ,
MSC RINSE
*
DISTRIBUTION
OFONARY
PUMP
ri)
SUBMERSIH
PUMP
t\\
FLUME
-HTANK-3
\JF\tWL RINSE
LEGEND
FINAL
INSPECTION
TOMATO PRODUCT
PROCESS WftTER
MAKEUP WATER
SOLIDS SLURRY
WATER METER
SEWER
BIN DUMP LEVEL CONTROL
WATER QUALITY
SAMPLE STATION
(FURTHER PROCESSING)
FIG.5-FLCW DIAGRAM OF DISC CLEANER WITH
WATER RECYCLE SYSTEM
167
-------
The disc cleaner replaced the inside flume conventional cleaning process
to remove tightly adherent smear soil. In this unit the long exposure of
tomatoes to the turbulent action of water was replaced by short exposure
to the vigorous mechanical shipping action of soft rubber discs/ '
The action of the discs moved the tomatoes through the unit, loosened and
shipped off the soil and threw the heavy soil water into the tray beneath
the device. The rubber disc unit was equipped with small whirl jet spray
nozzles, located to provide full coverage of the last 2 feet of the unit
with no overlap. Water was collected and pumped to the gravity screen and
swirl concentrator.
Disc Cleaner With Water Recycle and Chemical Flocculation—The flow
diagram in this mode of operation was similar to the previous flow scheme
except for the addition of the internal chemical coagulation and flocculation
at the thickening stage as shown on Figure 6. The primary study objective
in this mode of operation was to evaluate the effectiveness of chemical
coagulation-flocculation with respect to soil solids removal efficiency and
wastewater quality to the sewer.
The chemical coagulation-flocculation mechanism consisted of four
components: recirculation pump, tube flocculator, slip-stream turbidimeter,
and chemical feeding mechanism. The recirculating pump withdrew from the
grit tube, a 10" diameter tube surrounding the swirl concentrator underflow.
Grit settled through the tube while turbid underflow waters were withdrawn.
A portion of the flow was directed into the slip-stream turbidimeter which
continuously monitored the turbidity. The chemical feed was automatically
controlled to be proportional to the level of turbidity of the water.
The coagulant mixing and flocculation processes were accomplished in
a novel tube flocculator(4) comprised of a series of discrete pipe sections.
These pipes were coiled around the outer wall of the thickener, each down-
stream section having a progressively larger diameter with diverging
transition members connecting adjacent sections. The coil curvature and
pipe diameter were carefully pre-determined to effect a flow condition velo-
city gradient yielding optimum coagulation-flocculation results in minimum
time.
Plant Operation and Data Acquisition
The plant operation and data acquisition commenced with the tomato
season start during the latter part of August, 1975. The schedule of opera-
tions are listed chronologically in Table 1. The study period covered the
entire season of tomato processing which lasted about 2 months.
The plant operated three 8-hour shifts per day, 6 days per week.
There were .approximately 4 operators involved in each shift. One operator
was controlling the rate of tomatoes to the bin dump and the other three
operators were involved in the following tasks:
168
-------
BIN
WASHER
. 5
SCREEN 8
SWIRL
CONCENTRATOR
J
©
«,
«
x-s.
DUMP
TANK
TANK-I | 1
•TOMATO PRODUCT
PROCESS WATER
MAKE UP WATER
SOLIDS _ " 'SOLIDS SLURRY SEWER
i -WATER METER
3> 'WATER QUALITY
SAMPLE STATION
& 'BIN DUMP LEVEL CONTROL
(FURTHER PROCESSING)
FIG.6- FLOW DIAGRAM OF DISC CLEANER WITH WATER
RECYCLE SYSTEM WITH FLOCCULATION
169
-------
-£UTEK-
TABLE 1
CHRONOLOGICAL LISTING OF THE MODES OF OPERATION
Period of Operation
Modes of Operation
3 Sept - 9 Sept, 1975
Conventional Cleaning
10 Sept - 14 Sept, 1975
Conventional with Water
Recycle
16 Sept - 20 Sept, 1975 &
30 Sept - 8 Oct. 1975
Disc Cleaner with Water
Recycle
22 Sept - 29 Sept, 1975 &
9 Oct - 11 Oct, 1975
Disc Cleaner with Water
Recycle and Chemical Coagula-
t1on-Flocculat1on
170
-------
1. Organization of tomato bins for processing;
2. Removal of vines and debris from vine collectors;
3. Maintaining and clean-up of the processing facilities.
Three full-time NCA operators, one in each shift, were at the plant
throughout the study period. The duties of these operators were:
1. Taking water meter readings at the beginning and end of each
shift;
2. Taking sludge samples, one at each shift;
3. Taking tomato product samples in each shift at each washing
stage;
4. Recording events and general observations;
5. Recording total number of bins processed;
6. Tabulating results.
ISCO automatic water samplers were used at all sampling points except
two, one at the influent water to the swirl concentrator and the other at
the sludge thickener overflow. At these sampling points, an automatic
Roto Vee sampler developed by EUTEK was used. The Roto Vee sampler con-
sisted of sample collector, motor and remote controlling timer. The Roto
Vee, because of its simple design and operation, was effective 'in collecting
samples having relatively high solids concentrations such as the wastewater
incoming to the swirl concentrator. Sampling for sludge solids concentra-
tions was done with a core sampler which collected representative mud samples
from deposited soil solids.
Interpretation of Results
Analysis of Water Consumption
Water usage in a typical tomato processing plant can be categorized
as:
1. Filling
2. Operation or process
3. Clean-up
Filling water represented that used to fill the dump tank and flumes.
It was normally required at the end of each washing period and prior to
the beginning of the following operation shift. Process water used for
171
-------
operation was that which was continuously utilized during the shift opera-
tion, such as bin wash, continuous spray onto trash belts for cleaning pur-
poses and final rinse waters. Clean-up water was used both during the
operation period for cleaning floors and for clean-up tanks, flumes and
other equipment during the normal clean-up period. Total water consumption
in a typical tomato processing plant was the sum of all the water used for
each above-mentioned purpose.
The Average Water Consumption Rates—The total and unit water consump-
tion rates have been averaged and summarized in Table 2.
In the conventional cleaning mode, which represented an average of
six days operations, the average amount of tomatoes processed was 481
ton/day; total water usage was 133,500 gal/day; yielding an average unit
water consumption rate of 278 gal/ton of tomatoes processed.
As indicated on Figure 3, the conventional cleaning system utilized
an effective counter-current flow scheme, thereby reducing total plant water
usage significantly below that of other tomato plants. Conventional tomato
plants without counter-current use of process waters have been reported to
utilize approximately 1200 gal/ton of tomatoes processed.(5) Significant
savings in water volume used can be achieved through utilizing counter-
current flow measures. In this study, the comparison of the water recycle
system with off-line mud removal was made relative to a system in which maximum
water conservation through counter-current measures had already been effected.
The second mode of conventional operation with water recycle averaged
over four consecutive days of operation, the average amount of tomatoes
processed was 445 ton/day; average water consumption was 109,100 gal/day.
The average unit water consumption rate was 245 gal/ton or tomatoes processed.
Due to the intermittent and irregular delivery of field harvested product
during this period, the resulting unit water usage rate should be viewed
as suspect. Shift operations were frequently limited by quantity of tomatoes
to process. It would be expected under normal operating conditions in which
quantity of field harvested products did not limit plant operation that
the unit water usage in this mode of operation would be similar to that found
in the other water recycle modes of operation.
Under the mode of operation of disc cleaning with water recycle, over
twelve days of operation, the average amount of tomatoes processed was
594 ton/day and the average water consumption was 92,800 gal/day. The
average unit water consumption rate was 156 gal/ton of tomatoes processed.
In the final mode of operation of disc cleaning with water recycle with
chemical flocculation over eight days of operation, the average amount of
tomatoes processed was 605 ton/day; with an average water consumption of
99,000 gal/day. The average unit water consumption rate was 164 gal/ton
of tomatoes processed.
Several observations and conclusions can be drawn:
172
-------
TABLE 2
SUMMARY OF AVERAGE WATER CONSUMPTION BY MODE OF OPERATION
CO
Mode of Operation
Conventional
Conventional w/
Recycle - w/o Floe.
Disc Cleaner w/
Recycle - w/o Floe.
Disc Cleaner w/
Recycle - w/ Floe.
Average
Tomato Ton-
nage Pro-
cessed
(ton/day)
481
445
594
605
Average Water Consumption (gal/day)
Filling
10,700
7,800
7,000
6,400
Operational
81,300
59,800
44,300
51,100
Clean-up *
41,500
41 ,500
41,500
41,500
Total
133,500
109,100
92,800
99,000
Average Unit
Water Consump
tion Rate
(gal/ton)
278
245
156
164
* Season Average Value
-------
1. The tonnage of tomatoes processed daily increased substantially
during the disc cleaner modes of operation. A 26% increase in
the tonnage of tomatoes processed was realized with the disc
cleaner with water recycling and chemical flocculation relative
to conventional cleaning without recycle. This increase may be
due to the increase in the soil solids removal efficiency from
the dump tank such that there was no significant accumulation of
solids in the dump tank which hindered the tomato processing operation.
No incident of temporary shutdown of the operation was encountered
during the modes of operation with water recycling measures.
2. A decrease of 26% in average total daily water usage is realized
with disc cleaning with water recycle and chemical flocculation
relative to conventional cleaning without recycle. Decreases in
water usage were noted in both clean-up and operation.
3. A significant decrease in the average unit water consumption rate
occurred when disc cleaning with water recycle was applied. A
41% decrease in average unit water consumption rate was realized
when disc cleaner with water recycling and chemical flocculation
was applied relative to conventional cleaning without recycle.
Table 3 presents the percentage of average water consumption by operation
mode. The majority of daily water usage was operational (48%-61%) followed
by clean-up (31%-44%) and filling (6%-8%) purposes. There were no signifi-
cant variations in percentage of daily total under various modes of operation.
Approximately 7% of daily water usage was filling; 55% for operational;
and, approximately 39% for clean-up purposes.
Flow Balance in Water Recycle Modes—The average flows to the swirl
concentrator during each of the three shifts for each operational mode
investigated are shown in Table 4. Flows to the swirl concentrator were
about 100,000 gal/shift for all three operational modes. Flows were
consistently lower during the second shift relative to the first and third
shift. There is no apparent explanation.
For the operational modes with water recycle the only source of water
discharged to the sewer was the thickener overflow. The average thickener
overflow to the sewer (equal to the total make-up water) is also shown in
Table 4. The volume of thickener overflow to the sewer was approximately
the same for all three water recycle operational modes.
Mass Balance Analysis—On Table 5 are presented the total solid concen-
trations under the various operational modes. For the dump tank and inside
the distribution flumes, there was no significant difference in solid
concentrations. However, there was a significant reduction in solid con-
centration to the sewer with water recycle. Eight thousand mg/1 of solid
overflowed to the sewer in the conventional cleaning system. Between
300 to 3,000 mg/1 of solid overflowed when recycle and reuse measures
were used.
174
-------
TABLE 3
PERCENTAGE OF AVERAGE WATER CONSUMPTION BY OPERATION MODE
—i
01
Mode of Operation
Conventional
Conventional
w/Recycle - w/o Floe.
Disc Cleaner
w/Recycle - w/o Floe.
Disc Cleaner
w/Recycle - w/Floc.
Fi 11 i ng
Purpose
8
7
8
6
Operational
Purpose
61
55
48
52
Clean-up
Purpose
31
38
44
42
Daily
Total
100
100
100
100
-------
TABLE 4
AVERAGE TOTAL FLOW TO SWIRL CONCENTRATOR AND OVERFLOW TO SEWER DURING EACH SHIFT
Mode
of
Operation
Conventional
w/Recycle -
w/o Floe.
Disc Cleaner
w/Recycle -
w/o Floe.
Disc Cleaner
w/Recycle -
w/Floc.
Average Total Flow to Swirl Concentrator
Shift 1
Quant.
(gai)
118,600
124,600
104,800
% of
Shift
Total
Flow
385
373
284
Shift 2
Quant.
(gal)
90,200
105,500
88,300
% of
Shift
Total
Flow
260
286
246
Shift 3
Quant.
(gai)
100,700
102,400
106,800
X of
shift
Total
Flow
325
308
287
Average Total Overflow to Sewer
Shift 1
Quant.
(gal)
21,500
15,600
17,800
% of
Shift
Total
Flow
65
42
47
Shift 2
Quant.
(gal)
16,900
17,200
16,500
% of
Shift
Total
Flow
49
46
45
Shift 3
Quant.
(qal)
21 ,400
18,000
17,600
% of
Shift
Total
Flow
65
49
47
-------
TABLE 5
REPRESENTATIVE SOLID'S CONCENTRATIONS BY OPERATIONAL MODE
Mode of Operation
Conventional
Conventional with
Recycle
Disc Cleaner
with Recycle
Disc Cleaner with
Recycle and
Chemical Coag-Floc.
Dump Tank
(mg/1)
8,000
4,000
—
—
Inside
Flume
(mg/1)
750
1 ,000
~
—
Distribu-
tion
Flume
(mg/1 )
350
300
400
500
From
Disc
Cleaner
(mg/1)
—
—
2,000
2,000
Overflow
to Sewer
(mg/1 )
8,000
3,000
1,000
800
Into
Swirl
Concentra-
fe)
~
8,000
6,000
7,000
Overflow
From
Swirl Con-
centrator
Tma/1 )
—
7,000
5,000
6,000
Slurry
From
Dump Tank
m
60
64
64
60
Slurry
From
Thickener
m
—
30
32
30
-------
The suspended solids concentration in the overflow from the sludge
thickener during the '75 season study was two-fold higher than observed
during the '74 season study. During the '74 season study, the thickener
had been loaded at approximately 1,000 gal/day/ft2. Turbulence in the
thickener was low enough to insure clarification of the swirl concentrator
underflow waters. Thickener overflow solids concentrations were approxi-
mately 300-500 mg/1. In the '75 season study, the sludge thickener overflow
rate was increased to 2,000 gal/day/ft2. The increased turbulence presumably
resulted in an increase in the overflow solids concentrations.
It would appear that to achieve acceptable solids concentrations in the
thickener overflow the loading should be maintained at approximately
1,000 gpd/ft2 or less.
The solids concentration in the influent to the swirl concentrator
ranged from 6,000 to 8,000 mg/1 for all water recycle modes of operation.
The solids to the swirl concentrator were those intercepted by the false
bottom and transported by the ejector excluding those separated by the
gravity screen. The false bottom was quite effective in intercepting
settleable soil solids. In the conventional cleaning mode with recycle
the bin dump solids concentration was 4,000 mg/1 as compared to 8,000 mg/1
in the influent to the swirl concentrator.
Solids separated by the screen were primarily tomato seeds. Approximately
30 ft3 of solids were collected from the screen daily. Since these solids
were primarily generated from broken tomatoes they were not taken into account
as soil solids loaded to the plant with incoming raw tomatoes.
Comparing the swirl concentrator overflow solid concentrations with
that of influent, it was noted that a reduction of only 1,000 mg/1
in solid concentration was realized. This was true for all modes of opera-
tion. This result indicated that the majority of the solids were not settle-
able. The solid removal efficiency by the swirl concentrator was apparently
affected by the particle size distribution of the soil, in addition to the
other design parameters. The particle size distribution was, in turn, dependent
on the type of soil from which the tomatoes were grown. The type of soil,
the moisture content of the soil when the tomatoes were harvested and the
method of tomato harvesting all affected the amount of soil solids carried
to the processing plant. These "unaccountable" factors might explain the
variations in some of the results obtained during the period of study.
During the mode of operation of the disc cleaner with water recycle
and chemical coagulation-flocculation, a series of jar tests were conducted
to determine the optimum coagulant (Calgon Cat-Floe) concentration. In
view of the daily variations in soil characteristics, jar tests were conducted
for several days. Results indicated that the desired coagulant concentra-
tion ranged from 7 to 15 mg/1.
Difficulties were encountered in sampling during the latter part of
the study period. The composite samplers were out of order, and, conse-
quently, the grab sampling technique was used. Due to the sharp variations
in solids concentrations it was difficult to collect representative grab
samples. This made it difficult to evaluate the quality of water overflowed
to the sewer.
178
-------
Under conventional cleaning, the total soil solids which were loaded
to the plant were estimated from the sum of soil solids accumulated in the
dump tank and lost with overflow from the flumes to the sewer. The former
was determined by the total volume and concentration of soil solids shoveled
into the gondolas. The latter was determined by the total overflow volume
and soil concentration to the sewer. For example, on September 4th there
were a total of 11,779 Ibs of soil solids to the plant. Of this, 7,384
Ibs were estimated to be lost to the sewer while 4,395 Ibs were shoveled
out of the dump tank into gondolas. There were a total of 489 tons of tomatoes
processed; the soil loading per unit of tomatoes was, for this particular
day, 24 Ibs/ton of tomatoes processed, or, slightly in excess of 1% of raw
product weight.
Under the conventional mode with water recycle the total soil solids
incoming to the plant were from three sources: soil solids removed from the
dump tank, lost to the sewer, and removed from the thickener. For example,
on September 10th there was a total of 5,451 Ibs of soil to the plant.
Of this, 1318Ibs were removed from the dump tank, 2,103 Ibs were lost to
the sewer, and 2,030 Ibs were removed from the thickener. Since there
was a total of 433 tons of tomatoes processed on this day, the unit soil
loading per ton of tomatoes processed was 13 Ibs/ton. Similar daily analyses
were made on all data collected.
Average Soil Loadings—Table 6 presents the average total soil loadings
and the unit loading rates under each mode of operation. As shown by the
second column of Table 6, approximately 3,800 Ibs/day of soil solids were
removed from the dump tank in the conventional mode without recycle.
This represents a unit soil loading of 7.9 Ibs/ton of tomatoes processed.
A significant decrease in soil solids removed from the dump tank was noted
as the water recycle was implemented. For example, there were only 1,447,
1293 and 1244 1bs/day of soil solids removed from the dump tank for conven-
tional with water recycle, disc cleaner with water recycle, and disc cleaner
with water recycle and chemical coagulation-flocculation, respectively.
The unit soil solids removed from the dump tank were 3.3, 2.2, and 2.1
Ibs/ton respectively for these three modes of operation.
Data is presented in the fourth and fifth columns of Table 6 that
indicates that there were 12.5 Ibs of soil per ton of tomatoes processed
lost to the sewer under conventional cleaning whereas there were 4.1, 1.3
and 0.9 Ibs/ton lost to the sewer in modes with water recycle. This indicates
that soil solids lost to the sewer decreased as water recycling was implemented.
Three point seven (3.7), 6.4, and 8.3 Ibs of soil/ton of tomatoes processed
were removed from the thickener in the conventional cleaning, disc cleaning,
and disc cleaning with chemical flocculation modes, respectively. This
indicates that the unit weight of soil solids removed from the thickener
increases as more sophisticated removal measures such as chemical floccu-
lation were implemented.
In the last column are presented the average soil solids loading per
unit of tomatoes processed. Twenty, 11, 10, and 11 Ibs of soil/ton of
tomatoes processed were accounted for in the conventional, conventional with
179
-------
CO
o
TABLE 6
AVERAGE SOIL LOADINGS BY OPERATIONAL MODE
Mode of Operation
Conventional
Conventional
w/Recycle - w/o Floe.
Disc Cleaner
w/Recycle - w/o Floe.
Disc Cleaner
w/Recycle - w/Floc.
Total
Tomato
Processed
(tons/
day)
481
445
594
605
Soil Solids Removal
From Dump Tank
(Ibs/day)
3796
1447
1293
1244
(Ibs/ton)
7.9
3.3
2.2
2.1
Soil Solids Lost
to Sewer
(Ibs/day)
6029
1807
747
547
{Ibs/ton)
12.5
4.1
1.3
0.9
Soil Solids Removed
From the Thickener
(Ibs/day)
~
1628
3827
5038
(Ibs/ton)
—
3.7
6.4
8.3
Total Sof
Solids to
the Plant
(Ibs/day)
9825
4882
5867
6829
Soil Per
Unit of
Tomato
(Ibs/ton)
20
n
10
11
-------
recycling, disc cleaner with recycling, and disc cleaner with recycling
and chemical flocculation modes, respectively.
Several important conclusions can be drawn from the observations above:
1. The apparent plant capacity for tomato processing increased with
disc cleaning and water recycle.
2. The soil solids removed from the dump tank per unit weight of
tomatoes processed decreased significantly as water recycling
was applied.
3. Soil loads to sewers per tonnage of tomatoes processed decreased
as water recycling was implemented.
4. Soil solids removed from the thickener per unit weight of tomatoes
processed increased as water recycling was applied.
5. Incoming soil solids per ton of tomatoes processed ranged from
10 to 20 Ibs/ton, having an average of 13 Ibs of soil per ton of
tomatoes processed.
Distribution of Soil Loadings—An analysis of soil solids distribution
under each operational mode is presented in Table 7. With conventional
washing, 39% of the soil was accumulated in and removed from the dump tank.
The balance of 61% was discharged to the sewer.
The percentage of soil solids discharged to the sewer sharply decreased
as water recycling was implemented. In the conventional mode with water
recycle, 37% of the soil solids were discharged to the sewer. For both
modes of operations with the disc cleaner, the percentage of soil solids
discharged to the sewer decreased substantially.
Soil solids accumulating in and removed from the dump tank were reduced
significantly in modes of operation with recycle. The majority of the settleable
soil solids were effectively intercepted and transported to the swirl con-
centrator. Only a small percentage of soil solids accumulated in the dump
tank.
Economic Significance of Water Recycle With Soil Removal
An economic evaluation was made of a typical tomato processor dischar-
ging to a municipal sewer. Charges for this service are usually established
by the municipal agency so that the operating cost of its facility are
varied and distributed according to usage. However, this practice often
may be unfavorable to the average tomato processor who operates on a seasonal
basis.
181
-------
TABLE 7
DISTRIBUTION OF SOIL LOADINGS BY OPERATIONAL MODE
00
ro
Mode of Operation
Conventional
Conventional
w/Recycle - w/o Floe.
Disc Cleaner
w/Recycle - w/o Floe.
Disc Cleaner
w/Recycle - w/Floc.
Soil Solids From
Dump Tank
(*)
39
30
22
18
Soils Discharged
Into Sewer
(30
61
37
13
8
Soil Removed
by the Thickener
(*)
—
33
65
74
-------
Cost-Effectiveness Evaluation
On Table 8 are listed the industrial wastewater charges for Sacramento
County, California as proposed for 1978.(6) Total charge is the sum of five
items: demand charges, loading charges, service charges, monitoring charges,
and potable water use charges. To illustrate the economic significance of
in-plant water recycle with soil removal for the tomato processor, an evalua-
tion was made based on the parameter values obtained in this study as listed
on Table 9.
The following assumptions are made:
1. Period of operation - 60 days.
2. Solid (sg 1.5) concentration of the mud from the dump tank - 60%.
3. Solid (sg 1.2) concentration of the sludge from thickener - 30%.
4. Total unit soil loading rate - 15 Ibs/ton tomato processed.
5. Unit soil loadings (based on the average solid distribution as
indicated in Table 6).
Soil
Solids Removal
Soil Solids Removed From Thickener
from Dump Tank (lbs/ton| (Ibs/ton)
Conventional 5.9
In-Pi ant Treatment 2.7 11.1
6. Solid hauling cost - $3.00/yd
7. In-plant facilities cost - $40,000
Useable life - 5 years, straight line depreciation
Operational and maintenance cost - $20,000 annually for all systems.
8. Service - 2 miles - 8" diameter or 16 diameter inch miles.
The total estimated annual cost consists of the charges for dischar-
ging into the municipal sewer system, sludge hauling cost for hauling
the collected solids to the disposal field, the annual capital cost for in-
plant pre-treatment facilities and annual operational and maintenance cost.
The industrial wastewater charges in Sacramento County by this typical tomato
processing plant with and without in-plant treatment are compared in Table
10. The evaluation indicated that approximately $1.80/ton of tomatoes pro-
cessed would be charged without water recycle and soil removal. Approximately
$.30/ton of tomatoes processed would be charged if in-plant treatment were
implemented.
183
-------
TABLE 8
MONTHLY INDUSTRIAL WASTEWATER UNIT
CHARGES SACRAMENTO COUNTY, 1978
1.
2.
3.
4.
5.
Demand Charges
Peak flow rate, per MGD
Peak BODc loading, per Ib/rlay
Peak SS loading, per Ib/day
Loading Charges
Volume, per MG
BODc per 1,000 Ib
SS, per 1,000 Ib
Service charge for handling the flow,
per diameter inch mile
Monitoring charge, per month
Potable water use charge, per 1,000 ft3
$ 1,069.00
0.39
0.10
85.25
43.00
42.80
17.00
126.75
0.85*
*$123 for the first 1,000 ft3/month
184
-------
TABLE 9
AVERAGE TOMATO PROCESSING PLANT WASTEWATER PARAMETERS
(BASED ON '75 SEASON STUDY)
Parameter
1. Average Unit Water Use Rate
(gal /ton)
2. Average Dally Tomatoes Processed
During Processing Season
(ton/day)
3. Average Daily (Peak Monthly)
Wastewater Flow Rate During
Processing Season, (gal/day)
4. Peak Daily Tomatoes Processed
During Processing Season.
(ton/day)
5. Estimated Peak Daily Wastewater
Flow Rate During Processing
Season, (gal /day)
6. Estimated Average Suspended
Solids (SS) Concentration of
Wastewater. (mg/1)
7. Average Chemical Oxygen Demand
(COD) of Wastewater (mg/l)
8. Estimated Average 5-Day
Biochemical Oxygen Demand of
Wastewater (mg/1)
Conventional
System
280
480
140,000
520
155,000
8,000
2,750
1,100
In-Plant
Treatment and
Conservation
Systems
160
600
90.000
670
100,000
800
2,900
1,000
185
-------
TABLE 10
ESTIMATED ANNUAL MUNICIPAL CHARGES
TOMATO PROCESSING WITH MO WITHOUT IN-PLANT TREATMENT
Charge Basis
1. Demand Charges Based On:
Peak Flow Rate
Peak BOD's Loading
Peak SS Loading
2. Loading Charges Based On:
Total Flow
Total BOD's' Loading
Total SS Loading
3. Services Charges
4. Monitoring Charges
5. Potable Water Use Charge;
Estimated Total Annual
Municipal Charges
Estimated (Average
Unit Costs for a
Conventional
System
($/ton)
$ 0.07
0.73
0.43
0.02
0.12
0.83
0.02
0.01
0.03
$ 1.76
Estimated Average
Unit Costs for an
In-Plant Treatment
System
($/ton)
$ 0.03
o.n
0.02
0.01
0.05
0.04
0.02
0.01
0.02
$0.31
186
-------
On Table 11 are compared the total annual costs with and without in-
plant pre-treatment. While both costs in sludge hauling and in-plant
O&M increase, due to the saving in the municipal charges, the total annual
unit cost savings is estimated to be approximately $1.31/ton of tomatoes
processed with in-plant treatment. An estimated 53% sayings would be realized
through use of the in-plant treatment system demonstrating that in-plant
treatment is a cost-effective approach.
REFERENCES
1. Bin Dump Tank Grit Collection and Transport System Model Study,
EUTEK, Report prepared for NCA, June 1975.
2. The Swirl Concentrator. EPA R2-72-008, September 19/2.
3. Cleaning and Lye Peeling of Tomatoes Using Rotating Rubber Discs.
Western Regional Research Center, USDA and National Canners Asso-
ciation, April 1974.
4. Flocculation Apparatus. U.S. Patent No. 3,933,642, January 20, 1976.
5. Liquid Hastes from Canning and Freezing Fruits annd Vegetables, EPA
12060 EDK, August, 1971.
6. Industrial Unit Charge Rates in Sacramento County, CA, Division of
Water Quality, Department of Public Works, County of Sacramento.
187
-------
TABLE 11
COMPARISON OF TOTAL ANNUAL WATER RELATED COSTS
WITH AND WITHOUT IN-PLANT WATER CONSERVATION SYSTEMS
Conventional
System
With In-Plant
Water Conser-
vation System
Municipal
Charges
($/ton)
1.76
0.31
Sludge Hauling
Cost
($/ton)
0.01
0.06
Annual Capital
and
Operation and
Maintenance
Cost
($/ton)
0.19
0.78
Total Annual
Cost
($/ton)
2.46
1.15
Potential Net Annual Savings = (2.46 - 1.15)(60}(600) = 47,160
-------
PROTOTYPE APPLICATION OF SINGLE PARAMETER SLUDGE AGE CONTROL
TECHNOLOGY - A CASE HISTORY - PERFORMANCE OPTIMIZATION BY
APPLICATION OF SLUDGE AGE CONTROL TO EXTENDED AERATION TYPE
TREATMENT PLANT FOR FOOD PROCESS WASTES**
by
Calvin G. Brown, P.E.*
INTRODUCTION
"The Lord by wisdom hath founded the earth; by understanding hath he
established the heavens."(1)
"Happy is the man that findeth wisdom, and the man that getteth understanding.
For the merchandise of it is better than the merchandise of silver, and the
gain thereof than fine gold.
She is more precious than rubies and all the things you canst desire are
not to be compared unto her."(2)
Because we have entered a new era of biological treatment kinetics,
or bio-kinetics understanding, let me briefly review well known recent
history pertinent to this paper's topic.
For decades, especially since 1914, when the activated sludge process
was first applied to wastewater treatment, engineers and scientists have
been seeking greater understanding of nature's process of stabilizing sewage
and other man-made wastes.
Lacking this vital understanding, operational data from many operating
plants was reviewed and used to develop empirical rationale, which were
then incorporated in the Ten States Standards (3). This procedure has
resulted in the perpetuation of very conservative design criteria use,
modified by trial and error, which today is still basically the State-of-
the-Art. The net result, however, has been very little increase in the
ability of designers to control or accurately predict effluent quality of
activated sludge systems.
Credit for the first major breakthrough in focusing attention on a
measurable parameter affecting effluent quality belongs to Garret in 1958 (4).
This was in regard to the hydraulic control of the sludge soluble rate.
In the mid 1960's, the U.S. Congress and N.Y.S. Legislature passed
historic water quality control legislation with grant-in-aid funding, and
the flood of research on treatment processes began.
In 1968 Jenkins and Garrison published their treatise on "Control of
Activated Sludge by Mean Cell Residence Time" (5).
189
-------
At the same time Professor A. W. Lawrence of Cornell University, in
collaboration with Professor McCarty, was developing and proving his unique
theory that optimal design and control of activated sludge was practical
utilizing only the single parameter, sludge age. Their paper, published
at Cornell University in 1969, was based on extensive laboratory experi-
ments (6).
A paper on control of solids retention time (SRT) was presented by
Walker in 1971 (7).
In late 1971 the author's firm was engaged to design a 24 MGD ter-
tiary treatment plant (8) requiring nitrification. Prof. Lawrence was
retained to determine whether one stage or two stage nitrification was
required for the domestic waste, utilizing sludge age control, in the vigo-
rous Western N.Y. climate. The results were presented in January 1973
(9) and excerpts appear in EPA's new Nitrogen Control Manual (10).
As explained later in more detail, the work of R. I. Dick published
in 1970 (11) on final settling tanks has proven invaluable ,in providing
the final element required to fully implement sludge age control on a prac-
tical and economical basis. The entire activated sludge process facilities
design and operation can now be properly implemented with confidence.
In 1973, the NYSDEC review staff advised our firm that our proposed
sludge age control technology, although showing promise, could not be approved,
since operational data of the "process" was not available in the United
States. At the same time, EPA's Region II R&D Screening Committee labeled
our findings (9) as the major breakthrough in advancement of bacteriological
treatment in 30 years. Metcalf & Eddy's new Wastewater Treatment Textbook
(12) also contained extensive detail on sludge age kinetics.
PROJECT SCOPE
The opportunity to apply the single control parameter of sludge age
to an existing conventional activated sludge type plant came in September
1974. Our firm was retained then for studies of and preparation of a report
on required modifications to Company X's Waste Treatment Plant in Western
New York, built in 1968. Company X produces over 100 tons of frozen
dinners a day, with over 500 employees. See Figures No. 2 and 3, which are:
2. Photo of Waste Treatment Plant; 3. Schematic of Process Flow Existing.
(Figure 1 not used).
The company's NPDES permit, issued in July 1974, dictated the scope
of studies and reports described in this paper by establishing final
effluent parameter limits.
190
-------
:-
Figure 2. EXISTING WASTE TREATMENT PLANT
(BUILT 1968)
191
-------
r\>
Figure 3. EXISTING WASTE TREATMENT FACILITY
-------
The pertinent NPDES final effluent limitations (effective 2/1/76) are:
Parameter Discharge Limitation #/D
Ave. Day Max. Day •
BOD5 23 46
TSS 32 64
CrT 0.2 0.4
Zn —0.4
pH 6.5 min 8.5 max
Fecal Coliform 200 (MPN/100ml/ 400(MPN/100ml/
30 day) 30 day)
Temp Not > 90°F
This paper describes the successful prototype application of sludge
age control to operation modifications of a conventional plant and to
design considerations for an extremely restrictive effluent quality
requirement.
In order to understand the data and changes in operation, the study
sequence of events was as follows:
DATE
Permit Issued 7/31/74
Engineers Retained 9/4/74
Report on 7 Day Monitoring Report - Schedule I 9/30/74
Start Routine Operations Sampling - Schedule II 10/1/74
Complete Report on Evaluation of Existing Facility 1/31/75
and Modifications Design - Based on 7 Day Monitoring
Data
Start 34 Day Test Period per Design Report 4/14/75
Recommendations
Complete 34 Day Tests 7/3/75
Complete 7 Day Air Flotation Unit Tests 9/9/75
193
-------
Complete Addendum A: Revised Design Report, 1/14/76
Based on 34 Day Test Results
Begin Preparation of Final Plans & Specs 2/16/76
Based on Addendum A
FROZEN FOODS PROCESSED
In order to properly evaluate the company's waste, it was necessary
to review the products processed and the probable variation in waste
characteristics. A typical week's production categories are given in
the following tabulation. These consisted generally of 30% sliced,
60% fried and 10% fruit and other products.
TABULATION - FROZEN FOODS PROCESSED
Day of Week
Mon Tues Wed Thur Fri Sat*
Product 9/23 9/24 9/25 9/26 9/27 9/28
Char Broil Beef Patty X X X X
Gravy & Sliced Beef XXX X
Gravy & Sliced Turkey X XXX
Gravy & Sliced Chicken X
Meat!oaf X XXX
Salisbury Steak X X X X X
Veal Parmagian X
Chicken Croquette
Man Size Beef Patty X
Turkey Croquette X
Fruit Cocktail X X
Orange Jello X XXX
Diced Peaches X
Strawberry Fruit XX XX
Applesauce X
Sliced Turkey X
Fried Chicken X
Fish X
Spaghetti X
Chicken Ala King X
Gravy & Beef X x
Gravy & Turkey X
*Saturday not a regular scheduled working day.
194
-------
The principal process sequences are mixing and slicing (flour &
spices), meat preparation & packing, oil cooking, breading, gravy and
noodles cooking, school lunch mini-pak and lunch room.
Since there is a single final discharge point waste stream, it was
not necessary to evaluate subprocess streams. A complete typical des-
cription is available in EPA Wastewater Characterization for the Specialty
Food Industry, Category 1. (13)
WASTEWATER CHARACTERISTICS AND ANALYTICAL PROCEDURES
The most important rules a designer must observe in establishing
waste characteristics for a process waste are
(1) Secure representative data over a statistically valid period
of time.
(2) Calibrate all flow measurement devices and check sampling
and analytical techniques.
(3) Determine correlation between processing operations and waste
generated. «
(4) Utilize an independent laboratory certified by EPA for intensive
testing, preferably having the same technician and approved
technique involved in all analysis during the tests.
The effects of departing from these rules are quite well illustrated
in this project.
The NPDES permit required an immediate monitoring of waste characteri-
zation over a 7 day period. Company officials requested that this test
data be used to provide the basis for a complete report on modifications
to the W.T.P. in order to comply with the NPDES compliance schedule.
We were to implement sludge age control as much as possible for this
test period.
This was accomplished, with our accompanying recommendation, however,
that sludge age control be fully applied to W.T.P. operations and that
a minimum of 30 days test data be acquired. This would provide a statis-
tically accurate basis for designing a lower first cost and operating
cost treatment system, we stated.
The following Tables illustrate the comparison between the original
design loading criteria, the 7 Day monitoring data and the actual 34 s
Day test data waste loadings to the W.T.P. All tests were made in
accordance with Standard Methods for the Examination of Water and Waste-
waters, 13th Edition, 1971, APHA NYNY.
195
-------
All sampling, but only solids analyses, were performed by company
personnel. All other sampling analyses were performed by Ecology and
Environment Inc. laboratory, Cheektowaga, N.Y. under the general direction
of Gerald L. Stobel, P.E. and direct supervision of Rocco Termini.
It is to be noted that the 7 Day data indicates that the loading
to the W.T.P. aeration tank is greater than the raw waste load of BODj-
to the W.T.P., whereas for the 34 Day average aeration tank loading is
about 2/3 of the raw waste BODr content.
In the discussion on plant operation, the reasons for this anomaly
will become evident.
The processing plant's production vs water use is listed in Table
No. 1 and 1A.
The characteristics of the raw waste and the aerator applied loadings
are in Table No. 2 and 3, while aerator-clarifier operating data is
listed in Table Mo. 4.
TABLE 1 . PRODUCTION VS WATER USE AND WASTEWATER FLOW
•
- 7 DAY TEST
Day
Mon
9/23
Tues
9/24
Wed
9/25
Thur
9/26
Fri
9/27
Sat ,
Min
Ave
Pro-
duction
Tons/Day
153.0*
123.6
128.3
112.7
110.1
89.2
89.2
119.5
Water Used
Gals.
101,690
105,360
102,330
102,810
104,990
86,100
6 Da.y
86,100
100,550
Water Use
Gals/Ton
665
850
800
910
950
965
Data Averages
665
840
Process
Wastewater
Flow
Gals.
84,800*
64,280
62,730
57,990
51,900
52,260
51,900
62,000
Process
Wastewater
Flow
Gals/Ton
555
520
489
515
470
585
470
522
196
-------
TABLE 1. (CONT.)
Day
Max
Min
Ave
Max
Pro-
duction
Tons/Day
153.0*
74.6
100.0
128.0*
*Maximum Data
Water Used
Gals.
6 Day
105,360
Data as
85.6
100.0
104.8
Water Use
Gals/Ton
Data Averages
965
Process
Wastewater
Flow
Gals.
(Cont.)
84,800*
Process
Wastewater
Flow
Gals/Ton
585
Percentage of Averages
79.2
100.0
114.9
83.7
100.0
136.7*
90.0
100.0
112.1
Note: Data reported on 9:00 AM to 9:00 AM; while each day's shift
start-up was 6:00 AM and production started at 7:00 AM. Thus
there is a lag between water use and waste generated recorded
for each day of the test period. Previous flows were reported
from midnight to midnight.
197
-------
Table 1A. BASE DATA - 34 DAY TEST PERIOD
APRIL § JUNE 1975
PRODUCTION AND FLOW
(including DoC.G,. Filtrate'Inflow)
WASTE FLOW
Date
Produc-
tion
(Tons/Day)
Convers
Apr
Apr
Apr
May
14
15
16
17
18
21
22
23
24
25
28
29
30
1
2
126.
127.
128.
135,
118.
118.
129.
123.
136.
117.
138.
126.
144.
135.
135.
n= 15
x=129
s= 8
n= 34
x=133
s= 8
W.T.P.
Aeration
Tank
Inflow Produc-
(incl sewage) tion
(GalsxlOOO) Date (Tons/Day)
ion Period
4
2
7
0
7
5
9
0
1
4
6
6
9
8
4
.5
.1
36
50
63
47
45
44
74
72
70
37
72
57
36
33.
47
n=
x=
s =
. 5 Jun
,9
.2
.2
.0
.7 Jun
.2
,7
.6
.7
. 3 Jun
.8
.3
.5
.8
Jun
Jul
15
52.7
14.7
Combined
Sludge
9
10
11
12
13
16
17
18
19
20
23
24
25
26
27
30
1
2
3
136
130
138
140
121
150
126
139
146
118
129
145
140
145
134
134
135
139
141
n=
Control
ol
08
.9
.5
.8
.5
.6
.0
.5
.8
.5
.0
.1
.7
,1
.9
.0
.5
.5
19
x=136.6
Data
s=
8,3
WASTE
FLOW
W . 1 . F .
Aeration
Tank
Inflow
(incl sewage)
(GalsxlOOO)
Period
65.
70.
56.
68.
47.
42.
69.
63.
56.
45.
54.
48.
60.
54.
44.
44.
44.
53.
39.
6
9
6
9
1
6
2
3
0
6
7
6
4
3
9
1
9
6
8
n=19
x=54.3
s =
9.9
n=34
.4
.9
x=53o6
s =
12.0
198
-------
Table 2
Unscreened Raw Waste Characteristics
VO
VO
Parameter
Flow Q (gpd)
BOD5 Ave
Max
BOD5/COD, Ave
TSS Ave
Max
Oil § Grease Ave
(Ether Soluble) Max
CrT Ave
ZnT Ave
pH Ave
Min-Max
Temp
Ave
Min
Max
7 Day Test
Ave 62,
Max 8 5 ,
.55
4.7-8.7
4.7-12.4
11.7
5.2
16.3
(1974)
000
000
#/D
1400
1580
1440
2000
320
690
.04
.53
.
34 Day
mg/1
2480
4010
.
3470
7240
1710
2800
6.5-
-
21
16
24
Test (1975)
49,000
69,000
#/D
1000
2300
55
1400
4150
690
1600
8.6
-
.5
.7
.0
-------
Table 3. Aerator Influent Waste Characteristics
Parameter 1968 Design
Flow
BOD5
COD
BOD5
TSS
Oil
, Q(gpd) Ave 100,000
Max
mg/1 £/D ,
Ave 825 687
Max
Ave 920 766
/COD Ave .90
Max
Ave 294 245
Max
§ Grease Ave 0 0
Max
7 Day Test
62,000
85,000
mg/1
3010
--
.53
.47
2450
--
370
--
(1974)
•f/D
1500
2420
2800
1220
2240
185
410
34 Day
mg/1
1490
1670
2800
1690
2440
400
600
53
77
ft
9
Test
,500
,500
53
49
!••«••«
(*calculated
pH
Temp
Ave 6 . 7
°C Ave 21°
Min
Max
4.1-6.7
12.3
4.2
19.8
4.
8-
20
10
27
8.5
.0
.1
.8
(1975)
f/D
660
1020
1250
ODult/
BOD5=1.33
750
1820
180*
360*
from COD)
-------
Table 4. Aerator - Clarifier Operating Data
Parameter
1968 Design
IVi
o
BOD5#/KCF Ave
Max
MLSS mg/1 Ave
Max
MLVSS/MLSS
Detention Time, hrs.
(Hydr)
52
__
--
10,000
20.4
F/M #BOD5/#MLVSS
Clarifier Loading
TSS#/sf/d
Sludge Age-days (SRT)
(Calculated)
RAS recycle % Q
Metals Cone. (Aerator)
Cr mg/1
Zn mg/1
Fe mg/1
Cu mg/1
25
7 Day Test (19 74)
113
181
10,700
13,800
.80
33
.26 (.43max)
28.6 (41max)
57
(4 min - 208 max)
100
(50 min - 170 max)
0.12
2.6
12.2
0.4
34 Day Test
50
77
7,100
10,800
.66
38.2
.13
25.7
8.3
(3.0 min -
250
(130 min -
Not
measured
tt
it
(1975)
(.29max)
(46. 5max)
9 . 6 max)
380 max)
-------
EXISTING TREATMENT PLANT FACILITIES
The existing treatment facilities were sized and operated up through
1974 in accordance with conventional design (Ten States Standards).
Operators were trained in state sponsored sewage treatment plant operator
schools. They found very little practical application to this process
waste treatment operation was possible.
It should be noted that the area in which Food Company X is located,
is not served by city sewers nor interceptors.
A brief explanation of the facilities and operation in September
1974 is as follows. These include many improvements in facilities and
operation initiated by company officials since 1968, and some at the
suggestion of county officials.
The daily waste characterizations for all test periods are illustrated
in Figure 4. Typical daily flow variation data is shown in Figure 5.
The unscreened process waste flows under pressure into two surge
tanks which hold about two thirds of an average day's flow. Bottom
air diffusers aid in breaking up emulsifying grease and solids and mini-
mize odors. pH can be adjusted by chemicals in the surge tank. Solids
and emulsified grease settled in the bottom two feet and were pumped by
scavenger several times a week.
Oversized trash pumps lifted the waste to an air flotation system
at 20-30 psia. This system was inadequate to produce or maintain a
float and about 10% oil and grease removals were effected during both
test periods. In October 1975, use of a large air compressor and rear-
rangement of the trash pumps in series provided a 60 psia discharge head
capability and resulted in increased oil and grease and solids removals.
Alum and polymer are normally added.
The Air Flotation Tank weir discharge was mixed with unscreened
domestic sewage in an intermediate wet well, together with the dewatered
solids filtrate from the D.C.6. unit, and lifted by oversized centrifugal
pumps to the aeration tank. Excess flow to the flotation unit was
recycled via gravity pipe to the surge tanks up to 100% recycle.
The influent to the aeration tank was discharged onto the aeration
tank surface alongside the clarifier RAS recycled flow at a point oppo-
site the aerator overflow to the clarifier.
The aerator is a Lightnin Mixer 30 HP rated at 3.2# 0,,/HP/hr., with
surface aerator and submerged blade, providing complete mixing in a
Suburbia aerator-clarifier system.
The aerator mixed liquor overflows to the clarifier, which had a
nominal 250 gpsfd rating at design flow of 100,000 gpd, and 8'-4"SWD.
A sludge draw off box with telescopic valve and oversized submerged sludge
pump provided RAS recycle and WAS by manual valving.
202
-------
A
\^
::TC/
/\
!u
= Kl
;;: -H
V
i..-.
-hhH-
ffR
_ JUH. 7> -.
<3 E
_l ~"
5
*>—
_ a>
Z £
S"
tf)
3
O <9
O . fe
— <0
-------
140
130
MAXIMUM DAY
65,750 GPD
AVERAGE DAY
49,600 GPD
. i! 1 [MINIMUM DAY
1 II I I 38,570 GPD
7am 8 9 1O 1 1 12 1pm 2 3 4 5 6 7 8 9' 1O 11 12 1om2 3 4 S 6 7
TIME (hours)
Figure 5.
TYPICAL DAILY FLOW FLUCTUATIONS
(34 Test Day Period- 1975)
204
-------
The clarified effluent flowed through a chamber with less than
5 minute detention, where Cl? is diffused into the flow. A 60°V notch
weir at the end of the chamber provided flow level for a BIF float operated
recorded, transmitter and recorder measurement system.
The effluent flows to Lake Erie above the lake high water level
via a 12" diameter outfall.
The plant is located at the east end of Lake Erie and is subjected
to 50-75 mph winter gales and is subject to infrequent surface water
flooding.
Because of high water table and poor subsoils, the plant is wholly
at or above ground level.
The initial plant was as described but without any surge tanks,
polymer mixing tanks, submerged aerator blade, DC6 (dual cell gravity
solids thickener) for WAS processing, laboratory, or flotatipn recycle
pipe to surge tanks.
Over the years, company officials found that equalization was
required; that polymers aided in securing better grease float; that a
submerged aerator blade prevented anaerobic sludge build-up on the tank
bottom; and the DCG unit resulted in less volume of sludge for scavenger
trucking to landfill. The flotation recycle pipe reduced flow variation
somewhat and smoothed loading to the flotation unit.
EXISTING TREATMENT PLANT OPERATION
In this paper, only those operating features will be discussed which
are related and significant to our discussion of sludge age control vs
conventional control. There are of course certain procedures which are
equally applicable to improving any plant operation. In order to compare
the two control methods the "before" and "after" operating conditions
will be summarized in this section and detailed operation will be de-
scribed in later sections.
The plant operator has a New York State Sewage Treatment Operator's
license and has taken the prescribed two week course. One shift operator
had taken the two week course in 1975. The other shift operators are
scheduled to take the course in the near future.
In 1974 one operator was on duty each 8 hour shift on plant produc-
tion days.
Daily settled 24 hour composites were analyzed for BOD- and solids,
pH, temperature and Cl? residual, of raw, effluent and aerator mixed
liquor. Monthly reports of BOD^, TSS and S.S. were based on Thursday's
averages for 1974. Prior to September 1974, scavenger pumping of surge
tanks and WAS'storage tanks, at about 50% of weir flow, kept the loading
205
-------
to the aeration tank at reasonable levels. Thus the plant appeared to
provide about 90% removals of BODg and T.S.S.
Cl« residual of 0.5 mg/1 was carefully maintained at the weir effluent
(not at the outfall discharge at Lake Erie).
Scheduled tests were supposed to be run every shift (3/day) on the
following parameters at six locations 1) surge tank, 2) flotation unit,
3) intermediate wet well, 4) aeration tank contents, 5) clarifier blanket,
6) weir effluent
a) TS, TVS, TFS
b) TSS, VSS
c) S.S.
d) DO
e) BOD,
f) Oil & Grease
(occasionally)
g) pH
h) SVI
i) Cli residual
The basic control of operation up through 1974 has been to use
scavenger service to pump out surge tanks, oil & grease holding tank,
waste sludge tank, and even clarifier if solids discharge to Lake Erie
became excessive so as to be visually detectable by adjoining property
owners and regulatory agencies.
At $45 per 1000 gallons in 1974 and 1975, for scavenger service
collection and disposal at remote sanitary landfill, the cost was ex-
cessive even to maintain normal secondary removal efficiencies. This
equalled weekly costs of $1400 to $1800. In 1976, the cost increased
to $70 per 1000 gallons for this service.
The chief operator has been most aggressive in seeking out informa-
tion from trades people, other operators and industry sources, and self
study on hdw to improve operation and effluent efficiencies.
Our evaluation indicatedd, however, that interaction of existing
facilities and their non-complimentary capabilities provided major
hindrances to more effective operation.
For instance, in order to decrease solids overflow in plant effluent
by excessively wasting activated sludge from the clarifier, the BOD,
removal efficiency decreased, as there were insufficient bacteria
left in the aerator for effective metabolism.
The following comparison between the major operational conditions
is given here as the discussion of the test period results are related
to these facts. Parameter data is in Tables No. 1 through 5.
The most significant cost savings resulting from sludge age control
was due to the large decrease in waste activated sludge volume and the
reduction of raw waste scavenging.
206
-------
OPERATING CONDITIONS
(with no capital cost revisions to facilities)
"Before"
(w/Conventional Operation)
1. WAS cone to DCG=l-2%
2. WAS cone from DCG=3-5%
3. Total Volume raw waste,
grease & oil, & WAS TO
Scavenger=36,000-38,000 g/p/wk
4. @$70/1000 gal annual
scavenger cost equals
$138,000
5. RAS recycle rate 10-25%
(actual rate due to in-
correct flow data was
20%-50%)
6. Grease emulsion formed
daily in surge tank bottoms
requiring bi-weekly scavenger
pumping
7. Excessive flow variations
resulted in frequent spill-
overs of wastewaters to the
parking lot adjacent to W.T.P.
8. Air diffusers in surge tank
plug up-level indicators
fail and pumps clog up from
oil, grease, debris solids
plugging small valves and
lines.
9. Trash pumps in parallel:
air bound pump empty at
times. Pumping at 32-35 psi
close to shut off head-
cavitation pronounced.
Hardly any grease float.
10. Wet well lift pumps to
aerator clog from sewage
debris-float controls
jammed frequently. Pumps
on-off 5-15 seconds due to
"After"
(w/Sludge Age Control Operation)
WAS Cone to CDG 2-3%
WAS Cone from DCG 6-15%
Total Volume raw waste, grease & oil
& WAS to Scavenger=27,000 g/p/wk
@$70/1000 gal annual scavenger
cost equals $98,000;
scavenger = $40,000/yr. savings
RAS recycle rate ranged from
100%-400%
Grease emulsion pumped out infre-
quently once or twice a month-
balance handled by aeration thru
longer sludge age
Infrequent spillovers; during
test periods - none occurred.
Regular maintenance on more frequent
basis greatly reduced down time
and reduced frequency of downtimes.
Pumps in series pumping at 53-60
psi total: better performance
of pumps and air flotation system
(This was operation improvement)
Pumps throttled to permit 5-15
minute cycle on-off. Sewage
debris cleanout more often.
(This was operation improvement)
207
-------
oversizing and small wet well
level variation available.
11. Very little floe forms on
water surface of flotation
tank.
12. Chemical pumps clogged.
13. No effluent discharge from
flotation tank at times.
14. No D.O. residual in
aeration tank-30 HP
operated 24 hr/day during
production days.
15. Reading on influent pH
(adjustment in surge tanks)
S.S., and aerator D.O.
required hourly to provide
some control on processes
to prevent bulking sludge and
poor effluent.
16. Aeration unit foaming frequent.
17. Aeration unit short circuiting
18. MLSS variation unmanageable
at times, if too high over-
loaded clarifier, pumping by
scavenger from clarifier
only control available.
19. Either too little air apparently;
or too much air if low BOD
loading to W.T.P.
20. Sheared Floe aeration unit
affected settling characteris-
tics of MLSS.
21. Septic sewage in existing wet
well.
22. Difficulty in returning sludge
at low recycle rates or at very
high rates.
3-6" float forms on water surface
of flotation tank. Larger air
compressor used.
Same
More uniform discharge.
5 mg/1 to saturation D.O. measured
in aerator tank at all times, 30 HP
shut off at times.
No readings for control purposes
needed of influent pH and S.S.
and aerator D.O.
Infrequent
Lower average flow rates to aeration
unit reduced short circuiting.
Increasing recycle rate controlled
most MLSS variations
No problem with minimum or maximum
air supply.
No change - when reduced speed of
aerator is installed, improvement
is anticipated.
None
By throttling valves, and recycling
in 20-30 minute periods, instead of
very short or continuous recycling,
permitted good control.
208
-------
It is of greatest interest to note that most of the reported operating
problems for conventional activated sludge systems occurred in this plant
operation. Nearly all of these "ills" were reduced, brought under control.
or eliminated merely by applying sludge control technology principles,
without modifying existing facilities (except for a larger air source
for the flotation unit).
FIRST TEST PERIOD - SEPTEMBER 1974 (7 DAY)
In order to comply with the NPDES Permit Compliance Schedule, Company
officials directed our firm to develop the Seven (7) Day monitoring data
and submit it by September 30, 1974 and use that data as the basis of
designing necessary modifications. The engineering report was due January
31, 1975.
Our firm was commissioned to start on September 4, 1974. In our
proposal to perform the work, a minimum 30 day test was recommended,
however, before we could start testing all the unit operating facilities,
especially the flow measuring systems, were routinely checked or cali-
brated.
A 100% error in reported Weir Effluent Flow was discovered, as was
a similar variance in metered manufacturing plant waster use. After
determining this, we set up a pumping recording charge. The 7 day moni-
toring was commenced on September 23, 1974. Weir flow head was measured
hourly and calculated flow rate compared to recording equipment.
Several operating procedures were revised prior to the 7 day period.
The aim was to begin instituting sludge control if possible for this
test, even though in order to attain a true 8 day sludge age, for example,
a 16 day operation is required before data is valid.
a) The recycling of flotation tank effluent to surge tanks
was increased to reduce existing severe fluctuations of flow
to the aeration tank.
b) Lift pumps to aerator were throttled to increase 5-15 sec.
on-off cycles to -- 15 minute cycles minimum.
c) Recycle of RAS was increased fromm 25% to a minimum of 100%
to minimize clarifier solids overflow. Attempts were made to
utilize the equalization capacity of the surge tanks, but the
plant was so grossly overloaded that this was not possible for
the one week test.
d) Operators were requested to keep maximum forward flow from surge
tanks under 70 gpm (we found average flow was 35-50 gpm, not
70 gpm, as the recorder was incorrect as noted above).
209
-------
e) Wasting of solids was reduced as much as possible in order to
increase sludge age. When clarifier evidenced failure, wasting
was allowed until effluent solids were reduced to acceptable
levels.
f) Split composite samples over 24 hours were taken from six loca-
tions (previously described) at one hour intervals Monday
through Saturday. Solids were analyzed by the plant operators,
and all other parameters were analyzed by Ecology & Environment,
Inc., laboratory.
g) Previous reported data was suspect; the kinetics of the waste
were unknown; the process units maximum actual capabilities
were unknown; the Weir at the chlorine chamber was the only flow
measuring point; and operation control of the plant by the
Operators was primarily based on visual observation and opera-
tional knowledge acquired over years of experience.
h) The plant was severely overloaded during this first test week
and no raw waste was removed from the raw process stream, as
was done otherwise. All acquired flow data was by improvised
methods instituted by our firm.
i) In comparison to 1968 design criteria, the plant was overloaded
by two (2) times the BOD5 design loading, and six to eight
times the suspended solias design loading, on a weight basis.
j) The 7 Day monitored plant operation resulted in a hybrid combina-
tion of partial sludge control technique and previous hydraulic
control practices, which were resorted to by operators whenever
the effluent solids content indicated clarifier failure had
occurred.
Although additional manpower was added for the test period (one man
per shift), the lack of measuring and recording information, and lack
of hourly information on influent loadings and flows, made it impossible
to exercise adequate control by implementing frequency for required changes
in flow rates, recycle rates, flotation tank pressures, pump throttling,
sludge wasting, and the like.
The 7 Day results were analyzed and evaluated insofar as possible
(statistical evaluation was not possible due to lack of sufficient
occurrences, and correlation of data was nonexistent).
Tables 1 through 5 and Figures 4, 5, and 13 contain pertinent data
from the September 1974 monitoring period.
Some of the more pertinent findings of this test were as follows:
210
-------
a) The surge tanks were not used as equalization, but as storage,
as the surface aerator being oversized 100% (determined later)
appeared to overaerate. Over mixing of the waste on weekends
resulted in a low MLVSS on Monday morning and thus too small a
bacteria population to handle the organic loading beginning
Monday morning. Thus the raw waste had to be fed slowly to the
aerator to build up MLVSS.
b) The air flotation system operated at about 1/2 the desired
air pressure due to the pump shutoff head restriction, and was
furnished about 10% of the air required. Therefore its effi-
ciency was about 5-10% and the aerator and clarifier contained
what is reported to be excessive oil and grease for efficient
metabolism. At short sludge ages, grease and oils were not
effectively degraded, but at longer sludge ages much improved
degradation was noted.
c) The aeration tank measured 0 mg/1 dissolved oxygen (D.O.) when
checked several times each day at 1/2 depth. A profile showed
that 2-3 mg/1 D.O. was maintained at 1-2 feet depths, but at
6 foot or greater, the D.O. was 0 to 1.0 mg/1. This was believed
to be due to the high oxygen uptake of the waste. Later tests
showed that the short sludge ages, high sludge yield (synthesis)
and recorded MLVSS content caused this condition to exist.
The aerator influent pH was not effectively neutralized to an
acceptable range by the oxygen available in the aerator.
d) Highly variable rates of forward flow which also exceeded the
design hydraulic capacity appeared to result in excessive clari-
fier solids overflow to weir discharge, with proportional total
BODg in the effluent. Effluent soluble BOD5> however, remained
witnin acceptable limits, indicating that efficient capture of
suspended and colloidal solids would be the primary method of
meeting NPDES effluent quality requirements, after longer sludge
ages were made possible.
e) Due to the gross overloading, temperature did not appear to
have an identifiable affect on treatment or sludge settling.
The limited data seemed to indicate, however, a much larger
biological temperature coefficient than reported by other
authorities (12).
f) The oil and grease appeared to be readily biodegradable but an
excessive quantity was applied to the aerator. With short sludge
ages, degradation was not possible, and an excessive amount
entered the clarifier and was in the effluent.
\
g) The high wasting volume required in order to maintain control
over effluent solids resulted in excessive depletion of bacteria.
This in turn required larger MLVSS in the aerator, but this over-
211
-------
loaded the clarifier. Thus the two requirements resulted In
opposing operating requirements. The lack of facility capabi-
lity thus prevented the operation from being optimized beyond
the limit of the least capable operating unit, in this case the
clarifier.
h) The analysis of the 7 Day monitoring data indicated that the
number of occurrences of samples were insufficient to provide
a valid basis for predicting effluent quality or designing
units based on operating data secured.
i) Overloading any unit of the total system affects data generation
and analysis. This is why empirical formulae are suspect,
especially if derived from plants operating with one or more
units out of balance. For instance, if a final clarifier is
too small, providing excessive amount of equilization at the
head end of a plant will not result in improved effluent quality.
On the other hand, providing adequate clarification for maximum
solids loading will eliminate the need for additional equalization.
A preliminary design was prepared based on conventional aerator
loading data, F/M ratios, and conservative evaluation of kinetics as
derived from the 7 Day data.
Since the kinetic data was suspect, it is not presented in this paper.
The valid data is derived from the 34 Day Test is discussed in the next
section, together with the appropriate changes in operation and control.
However, the same kinetic evaluation procedure was used to derive opera-
tional charts for the operators to use for the 34 day test period.
The most important limitation in applying sludge age control is that
the kinetics cannot be estimated prior to development of the actual
kinetics of a given waste. Once the kinetics are determined, the appli-
cation to operation is easily accomplished and results are predictable
and accurate, once all parts of the process are in balance for the applied
loading variations.
The 7 Day test included sampling for heavy metals, Cr, Zn, Fe and
Cn. The permit final effluent limitations for Cr was 0.2 #/D daily average,
0.4 #/D maximum day, and for Zn 0.4 #/D maximum day. Based on future design
flows, this results in 0.4 mg/1 Cr daily average, and maximum day of
0.6 mg/1 Cr and Zn.
The September '74 data is of interest in that there was a significant
luxury uptake of all heavy metals as well as phosphorus and nitrogen.
Considerable amounts were removed in the sludge wastes.
The following tabulation on a concentration basis illustrates this
quite well. Although not statistically valid, the results for effluent
were well below the required limits and therefore no further sampling or
testing was performed during the 34 Day test and due to the high cost for
analyses.
212
-------
Average Dally Concentration In mg/1
Parameters Influent Aeration Tank Effluent
MLSS
Org N
P(T)
--
10.4
30.4
0.07
1.2
1.3
0.1
10,000
28.3
61.9
0.13
2.6
12.9
0.4
—
2.1
4.6
0.09
0.3
N.M.*
N.M.*
Zn/Tx
Fe,Tx
Cu,T)
*N.M. - Not measureable
SECOND TEST PERIOD - 1975
The company officials reviewed the preliminary design and estimated
construction costs of over $600,000 submitted in January 1975, based on
conservative design criteria.
On about April 1, 1975 the NYSDEC technical review staff informally
indicated to Food Company X officials that the design concept was approve-
able, but in their opinion provided more than adequate modifications to
meet final effluent limitations.
The company therefore authorized our firm on April 4, 1975 to proceed
with our previously recommended minimum 30 Day test utilizing sludge age
control to optimize operation and secure kinetic data for determining
design criteria.
The testing began on April 14, 1975.
Initial attempts to control the operation were still made very diffi-
cult due to several factors.
a) In order to secure the kinetics, gross overloading of the plant
had to be reduced for the test period of managable levels.
Operators lacked the means to determine incoming loadings and
flow variations frequently enough to decide how much raw waste
to remove by scavenger in order to limit overloading of aeration
unit. At times Saturday processing was found to be necessary.
b) Air flotation unit could not be operated at optimum pressure,
and trash pumps clogged up constantly. Flow rate was erratic,
213
-------
as recycle line valve was manually adjusted after estimates
were made of throttling required.
c) The existing B.I.F. float measuring unit had been taken out of
service, and a Honeywell measuring unit with a water level sensor
was installed late in 1974. This sensor when coated with grease
gave high readings and required constant calibration. Weir heads
had to be read hourly.
d) Initial attempts to maintain 10-15 days sludge age were only
partially effective. Complete control was not maintained, primarily
because the sampling and analysis for operation was made every
three hours, and wasting once per eight hour shift. This proved
to be inadequate since the flow and load changed drastically
during a shift and during each hour. This was evidenced by rapid
rise and fall of aeration MLSS concentrations.
e) Proper Monday startup equalization fully was difficult since, weekend
composite data was not available.
f) With added manpower, the hourly sampling, operational sampling
and analysis and routine down-time problems, left insufficient
time for calculations and review prior to making adjustments.
The charts, gages, and controls, were scattered throughout the
plant and adjustments to all equipment were manually made.
Samples and special measurements were manually secured, and water
use meters were,read at some distance away from the plant control
locations.
g) The need to have better control, more frequent data collection
and operation changes, was most evident as when the aerator
was turned off. Dissolved oxygen levels dropped from 5-6 mg/1
to 0 mg/1 in less than 15 minutes.
h) The initial lack of control resulted in several severe over-
loads on the clarifier, and solids carryover to the effluent
for up to several days.
i) Operators reverted to conventional practices whenever problems
arose and the engineer was not present.
After three weeks of these problems the testing was suspended on
May 2, 1975. Our firm then prepared two operating charts, based on the
7 Day data, with the aim of securing an average sludge age of 8-10 days.
Those are Figures 6 and 7. Operators were trained in the use of these
two charts, and wasting and flow changes was scheduled every three hours
as a maximum. More frequent analysis and changes were made if the plant
appeared to be nearing overload conditions, or a rapid change in aera-
tion MLSS was noted, or the sludge blanket in the clarifier began to rise.
The sludge control charts No. 1 and No. 2 were developed from kinetics
derived from the 7 Day test in September 1974. Every three hours, or
214
-------
ro
en
3
M
a
o>
E
•a
o
O
O
O
o>
-------
ro
OPERATING CHART N°2
SLUDGE (W.A.B.)
GALLONS TO WASTE
5 6 7 8 9 10 11 12 13 14 15 16 17
8 HOUR AVERAGE of W.A.S. MLSS (mg/l.X1000)
19 20
Figure 7. OPERATOR'S CHART NO. 2: WAS WASTING
-------
more often, the operator measured the aeration tank MLSS, the RAS concen-
tration, the weir flow rate and recycle rate. With these measurements,
he entered the chart No. 1 and determined the correct recycle rate and
adjusted the pump recycle rate accordingly by manually adjusting the valve
setting. He entered chart No. 2 and determined the 8 hour wasting volume
of TAS to waste and proportioned this amount for the actual hours involved.
All other data was taken for our use in developing kinetics and waste
characterization.
Testing was resumed on June 9, 1975 and completed on July 3, 1975,
with very satisfactory results and a noticeable reduction in operating
troubles and downtime.
During this June testing, the D.C.6. unit broke down after two weeks,
so that data was secured without D.C.6. filtrate returning to the aerator
influent. The effluent BODg levels did not change noticeably from operation
with the D.C.G. in operation.
The Honeywell sensor was removed during June; therefore, a drum float
recorder was installed for weir flow measurement.
Again, because the raw waste load was not able to be accurately
determined, there were several clarifier overloads. For our purposes
this was not serious, for we were able to thus determine the maximum
capability of the clarifier and other units.
Special data secured during the 34 Day Test were: and B values,
zone settling data, oxygen uptake rates, dissolved oxygen levels, extended
BOD value, hourly BOD5 vs flow for a typical day, and temperature and pH
at all stations.
Table 3 lists the aerator influent waste characteristics, and Table
4 lists the aerator clarifier operational data. Figures 4 and 5 show
the raw waste characterization and overall flow variations data for the
complete test periods.
The regular operator's daily report is typified by the June 9, 1975
report reproduced as Figure 8. The additional report sheet showing sludge
concentrations and volumes of WAS are reproduced as Figure 9.
The weir flow chart is reproduced as Figure 10 for July 2, 1975.
This was a typical flow pattern for the June test period.
FINDINGS
The most noticeable results of applying sludge age control are as
follows. Biological unit efficiencies and overall plant efficiencies
are given in Table 5. Table 6 lists the official operator's monthly
report based on Wednesday's tests for March through September 1975.
217
-------
SEWAGE TREATMENT PLANT
DAILY LOG
TUES,
DATE:
, /
VISUAL OBSERVATION:
hour
hour(IMHOFF)
TEST
Total SoTicTs"
Total Solids Fixed
Total Solids Volatile
Suspended Solids
Suspended Solids Volatile
Suspended Solids Fixed
Dissolved Solids
Settleable Solids
Settleable Solids
Sludge Vol. Index
Dissolved Oxygen
B.O.D. - 5 Day
Putrescibility or Methylene Blue
Alkalinity
Acidity
Hydrogen Ion Concentration pH
Nitrate - (N)
Nitrite - (N)
Ammonia - (N)
Phosphate
Color - Turbidity
Oil/Grease
Residual Chlorine
Chemical Oxygen Demand
Bacteria of the Coliform Group
Zinc
Chromium
Temperature
Daily Flow
Outfall Gallons per Day:
Usage Gallons per Day:
7^5*
so
so
"7.0
tD
C)
Raw Waste Characteristics
Color:
Odor:
Figure 8. OPERATOR'S TYPICAL DAILY REPORT
218
-------
DATE v
JO,
OPERATOR'S LOG
TIME
FLOW
MLSS
AERATION
SETTLEABLE
SOLIDS
HR,
1 HR
R.A.S.
RETURN
FLOW
MLSS
RAS
/o.
*/ '$00
60
*£&«*-<£,
7 /OG
970
//.
WASTE SLUDGE
TIME
GALS
MLSS
/
'.- e «-J>
SfrJff
G>
Figure 9» OPERATOR'S DAILY SLUDGE § FLOW REPOR
219
-------
.•. •
Figure 10 . FLOW RECORDER CHART
220
-------
Table 5. Plant Effluent Characteristics and Removal Efficiencies
Biological System
Parameter 1968 Design 7 Day Test
#/D % R #/D
BODcfT-, Ave 82.5 90 140
L J (87.7
,F-, Ave ... 27
B> TSS Ave 24.5 90 85
(110) (87.9
(max) (5
0§G Ave 0 60
Overall Plant Systems
(1974) 34 Day Test
% R f/D
91.3 55
min-97.1 max)
98.2 27
91.9 120 '
min-95.2 max)
Day Data)
84.0 20
(50 max)
Removals
(1975)
% R
90.7
95.8
81.9
87.0
5 Days Data
BOD5,T-, Ave -- 140
(F) Ave
TSS Ave -- 85
0§G Ave
90.0
--
94.1
__
93.0
95.8
90.0
96.5
-------
Table 6. Operator's Monthly Monitoring Report Results
1975 Summer Months - Wednesdays
Parameter
BOD,
* Effective until 1/31/76
See text for ultimate limits.
T.S.S.
Permit Limit*
1975
March
April
May
June
July
August
Sept,
(#/Day)
Av.e Max
23.0
38.7
50.4
49.6
41,4
37.4
31.9
13,6
64.0
65.3
55.6
74.9
56.9
48.4
42.0
25.6
(#/Day)
Ave Max
32.0
25.8
67.4
90.1
17,2
243,4
56.9
23,1
72.0
29.7
100.4
152.9
29.4
599.7
134.3
39.6
222
-------
(1) The dissolved oxygen in the aeration tank rose to 5-6 mg/1
minimum to nearly saturation values. Under comparable loadings
prior to September 1974 no D.O. residual was observed.
Pure oxygen systems, it is noted, generally operate in this
range of D.O., which heretofore has been regarded as a major
advantage of these systems over air systems.
(2) The clarifier sludge blanket was nearly always controllable
even when clarifier neared maximum load. The quick adjustment
of recycled RAS from 50% to 40% of forward flow made this
possible.
(3) The aeration tank pH was nearly neutral (6.7 to 7.7) at all
times due to the excess D.O. available, even though the raw
waste pH varied from 4.7 to 8.7.
(4) The plant had fewer downtimes and equipment malfunctions.
(5) The aerator-clarifier BODg loading was from 50% to 150% of
1968 design criteria loading. The solids loadings varied from
three to seven times 1968 design criteria loading.
In spite of this, the system worked well until the clarifier
load exceeded its capability. This occurred about five times.
the overloading effect lasted for several days before stable
operation resulted (i.e. normal solids in effluent).
(6) Oil and grease were well degraded and effluent contained satis-
factory minimal quantities.
(7) Although the range of flow rate was still from 0 to 120,000
g.p.d. during a 24 hour period, the larger flow rates had no
affect on clarifier performance, until the maximum permissible
loading for a given recycle rate was reached. Only then did the
clarifier fail.
(8) The temperature ranged from 5°C to 28°C for the aeration tank
influent. There did not seem to be a noticeable effect on over-
all operation. This appears to be because of tradeoffs between
lower sludge yield at lower temperatures, lower settling rates
at low temperatures, and higher level of mixed liquor concen-
tration corresponding to lower F/M ratios.
The biological reaction temperature coefficient derived from the
34 Day test data agrees well with referenced authorities (10).
The 7 day value however, did not.
(9) The oxygen uptake rate of the MLVSS was found to be comparable
to domestic sewage MLVSS at the longer sludge ages. The aerator
thus has nearly twice the Horse Power required for oxygen transfer
as needed, utilizing sludge age control.
223
-------
(10) The aeration tank MLSS concentrations ranged up to over 16,000
mg/1 during a 3 hour period, and up to about 12,000 gm/1 for a
24 hour period, or about 9,500 #/D, considerably in excess of
system capacity as determined later.
(11) The effluent BOD- substrate was very consistently related to
sludge age. See Figure 12. Thus, appropriate capture of clari-
fier solids (suspended and colloidal) would have resulted in
94+ to 96+% removals of BODC over the 34 Day test period.
— — o
(12) The sludge age varied between 3 days and 10 days, a very satis-
factory range for our purposes. This was achieved utilizing
only the two operator charts prepared by our firm.
(13) The sludge settleability has historically been checked by cal-
culating the SVI. Recently it has been noted that industrial
process sludges have SVI's which do not correlate with settle-
ability as does sewage sludge.
The SVI of this process waste MLSS during the test period was
of limited value to use for control purposes. Most of the time
the applied solids concentrations to the clarifier was so high
(> 6000 mg/1) that settling was hindered. The standard zone
settling test for sewage resulted in very low settling velocities
in undiluted settlings tests. Even so, the clarifier solids
did not overflow the clarifier weir until the limiting solids
flux was reached (as determined later from the test data).
^Additional tests demonstrated that the sludge which appeared
to be non-settling in a standard SVI liter cylinder, when diluted,
settled rapidly and formed a good sludge blanket.
Thus the standard SVI one liter test for sewage MLSS settleability
is not a valid or reliable test for settleability of a high concentration
process waste MLSS.
CONCLUSIONS
The 34 Day Test Data was analyzed and evaluated. Figures 11, 12
& 13 illustrate the critical parameters needed to establish sludge age
control. These are, sludge yield QC-!» (the reciprocal of sludge age
0 ), specific utilization U (the true F/M ratio), sludge age 6 , effluent
shbstrate BOD5 and zone settling velocity of the clarifier influent solids.
Since this paper is concerned primarily with application to operation,
a second paper is under preparation which will present the development of
the kinetics from the operating data at non-steady state conditions
Literature research indicates that this can not be successfully done
for conventionally designed and operated plants. The application of sludge
224
-------
U2o(BOD) U2o (COD) QQ ( 8C ) tcmp.°C
AVG. .213 .390 .105 (9.6)
.2 .3 .4 .5
U(20oc) Spec. Utilization:
*BODR /*MLVSS/day & *CODR /*MLVSS/dqy
Figure 11. SLUDGE YIELD (0
vs SPECIFIC UTILIZATION (U) CHART
225
-------
PERMIT.,
REQLMREMSNT
PERKIITj
REQUIREMENT
UN.75 APPLO. BC
r.57 b3«.0
il W.T.R QPERAT, R
APR.-JUN.
20 22
SLUDGE AGE 6.
Figure 12. SLUDGE AGE 0C vs WEIR EFFLUENT
BOD5 (Filtrate) for Fully Loaded W.T.R
APR.- JUN., 1975 Data
226
-------
0.08 •
0.07-
0.06"
0.05+
o
O 0.04-
_l
IU
0.02
O
E
CO
UJ
z
o
N °-0'T
OX) 2- -
0.00-
34 DAY DATA, 1975
ZSV»0.42e<-4-2x|°
ZONE SETTLING VELOCITY
5 DAY DATA,1974
4—I—I—I—I—I—I—I—I—I—I—I—I
6 8
LO 12 14 16 18 20 22 24. 26 28 30
AERATION TANK MLSS x |Q3 mg/l
Figure 13. ZONE SETTLING VELOCITY CHART
227
-------
age control to an existing plant has made it possible to develop the non-
steady state kinetics, which are of course more valid than laboratory
steady state data.
The final effluent quality and removal efficiencies are found in
Table No. 5, together with the Chief Operator's official summer month
reporting for 1975 based on Wednesday operations, as shown in Table 6.
The NPDES effluent limits are given in #/D; therefore the efficiency
percentages are no longer meaningful, except to compare with comparable
loadings.
The final design chart, Figure 12, establishes from the waste's
kinetics the proper sludge age to maintain for a required effluent quality
of BOD5. The suspended solids capture requires a multi media or single
media filter with appropriate capability of capturing the necessary amount
of clarifier effluent suspended solids and colloidal solids to meet the
permit limitations.
The clarifier design is based on R. I. Dick's work (11), our modi-
fications, and data secured on the waste settling characteristics. The
existing clarifier's maximum capability was at aeration tank MLSS of
7800 mg/1. Even a slight excess resulted in failure. The new clarifier
was designed to be optimal for an aeration MLSS of 9400 mg/1, when utilizing
the existing aerator and converted clarifier for aeration.
It is extremely important that the aerator-clarifier be sized to
complement each other. If not, one unit will immediately restrict the
ability of the other unit to function at maximum design conditions and
prevent optimum treatment.
In designing the clarifier, the concentration of the applied solids
T.S.S. has a direct affect on the actual settleability of the solids,
or settling velocity.
Therefore, proper sizing of the clarifier for high MLSS concentrations
for a particular waste should be based on MLSS concentrations, and not
on empirical hydraulic overflow rates. This will result in the most
efficient capture of all settleable solids applied to the clarifier.
It appears that more investigations of actual clarifier operation
under high MLSS loadings should be conducted so that a rational basis for
determining the optimum clarifier depth can be developed.
FINAL DESIGN
Based on the 34 Days data, a final design providing for raw waste
rough screening, a revised air flotation system, conversion of existing
final settling tank as an aeration tank, a new larger and deeper clarifier
sized to balance economically the total aeration volume provided, and a
multi media filter following the clarifier, with appropriate flow measure-
ment and parameter indicators for optimum control.
228
-------
The estimated construction cost of this design is about $425,000,
or about 30% less than the more conservative design.
Figure 14 is the schematic of this proposed final design.
In April 1976 the NYSDEC verbally indicated approval of this revised
design to Food Processing Co. X officials. Plans and specs have been
authorized by the Company for a June 1976 completion date to comply with
a new and revised EPA compliance schedule. The modified plant is to be
on line by June 1, 1977. The company plans to have the air flotation
unit and multi media filter in service later this fall.
GENERAL OBSERVATIONS
a) One of the most important practical advantages of using the sludge
age technology is that all existing plant operators can be easily retrained.
It is also not necessary for an operator to have college training or to
be a licensed engineer in order to achieve and maintain very high quality
effluents.
Special sampling and analysis as required by regulatory permits can
be contracted to local certified laboratories, when beyond the analytical
equipment capability of the plant.
b) The main disadvantage of conventional systems is their lack of re-
liability of operation due to lack of control to produce consistently
a quality effluent under varying loadings, flow fluctuations and seasonal
changes.
Because the effluent quality can now be predicted and effectively
controlled by utilization of sludge age control, very effective as well
as reliable treatment systems are now available.
c) The author believes that the sludge age control technology is not
only applicable at any stage of pollution project development, but is has
universal application to all biologically treatable wastes to produce the
least costly treatment system for a given effluent quality.
d) Based on our observations so far of this technology in its application
to date, and years of searching the literature and reviewing the reported
interpretations of the cause of upsets, failures, etc., of conventionally
designed and operated treatment plants, the following statement is offered
for consideration.
It appears that due to inadequate understanding of the true inter-
relationship of all components of a treatment plant biokinetic and solids
process, all attempts to isolate one component, or treatment unit, and
determine the factors most affecting treatment have met with little real
success. The profession is still today attempting to determine total
kinetics piecemeal and by trial and error, and/or by very complicated
methodology and formulae.
229
-------
ro
CO
o
M.H.
/•(•'A' t&f 9/m^n —i
"" ~fSoM n*UJ \ \MuroA
__nrj
L~ — —
BiffS. 400
Figure 14. PROPOSED WASTE TREATMENT FACILITY
-------
These are beyond most designers' and operators' ken, and wind up
relegated to the bookshelves for ostentatious display of proficiency by
association.
The sludge age single control parameter use is practical, is simple
to adopt, easy to implement in operation, and to secure the objective within
the reliability and accuracy desired. It is less difficult to implement
for establishing design criteria as it eliminates considerable guess work
where conventional design is concerned.
e) One of the most perplexing problems involved in waste treatment is in
regard to those wastes containing so called "toxic" levels of components
such as heavy metals, or of intermittent shock loadings of these heavy
metals, oil and grease, ammonia, etc.
Based on our findings, there appears to be a luxury uptake of salts
and metals in high aerator MLSS in proportion to the concentration and food
input. The literature reports that certain concentrations of "toxic"
materials are inhibitory to treatment at certain low levels of MLSS concen-
tration. However, the concentration of "toxic" metals may be higher than
reported before being considered inhibitory to metabolism of certain
stronger wastes or high aerator MLSS.
This may be equally true of nitrogen, phosphorus, etc. Our findings
are that the MLSS under aeration will absorb and hold the appropriate
amount of trace metals and catalytic material such as phosphorus, needed
for optimum metabolism.
Although the total concentrations of metal ions were in excess of
recommended levels, no inhibition of treatment was noted.
The high levels of "toxic" metals maintained also provide an acclimated
waste and the ability to abort shock loadings.
This therefore can be considered a very important advantage of sludge
age control technology and resulting ability of a system to maintain
higher MLSS concentrations.
POSTSCRIPT
The Food Processing Company X Chief Operator reports that in April
1976 the Waste Treatment Plant is averaging 98+% removals of BOD5 with
loadings varying from 2000-8000 mg/1 per day. S.S. removals are averaging
96% removals. Both effluents are still slightly in contravention of final
effluent limitations prescribed by EPA.
Scavenger removal of waste is averaging
8000 gals/wk surge tank bottom solids
8000 gals/wk oil & grease float
8000 gals/wk 8%-15% WAS DCG discharge (from 1% WAS)
231
-------
24,000 gals/wk Total removed
This compares to about 27,000 gals/wk removed during 1975 test periods.
SUMMARY
In conclusion it may be well to look once more at the concept of the
single parameter sludge age control. What is it? What can it accomplish?
Where can it be used? And what are its limitations?
Sludge age control is the optimization of any biological waste treat-
ment system, where aerated sludge is recycled, by controlling only the
active sludge mass age to result in a predictable substrate BOD concen-
tration in the effluent discharge to public waters or a municipal sewer
system, resulting in the least costly capital investment and least annual
operating costs.
Its application to operation requires no operating tests for BOD,
COD, pH, N, P, oil and grease and so forth. It requires sludgge values
and flow data only be determined on a frequent basis.
It is applicable at any stage of a pollution abatement project:
a) To existing activated sludge type plant or aerobic digesters
b) During design of a new plant or modifications
c) During construction of a new plant or modifications
d) After completion of a new plant or modifications
e) After a new plant or modified plant is in operation
Modifications are progressively more expensive to implement as the
project stage advances from (a) to (e).
No special pilot plant is required for analysis of an existing
activated sludge type plant. Once kinetics are determined for an existing
waste, it is not necessary to run further pilot studies unless an extreme
change in characteristics occurs.
The only apparent limitation to its use is for application to
standard trickling filters, oxidation ponds and similar single flow
through systems.
It is most effective in cost reduction when applied to sewage waste
nutrient removals requirements, in that it permits the use of one stage
carbonaceous oxidation nitrification systems and lower chemical require-
ments for phosphorus removals. This latter is possible because of lowered
alkalinity produced by the nitrifiers.
The other most significant technical breakthrough results from
applying sludge age control to an existing treatment plant under non-
232
-------
steady state conditions. It is not possible to measure and determine
the kinetics of such a biologically treatable waste without the necessity
of setting up and operating a pilot plant under laboratory steady state
conditions. By properly setting up the testing program and evaluating the
data, non-steady state data can be utilized to determine the waste
kinetics and be used with confidence for basis of design. The era of
empirical basis of design is past - now comes the era of rational basis
of design for biologically treatable wastes.
Standard Operation Form
Due to the continuing plant overloading, the company decided to con-
tinue intensive sampling and data acquisition for operator use in maintaining
optimum control until the plant modifications were built. To facili-
tate this, Chief Operator Chester prepared a comprehensive operation
form which was placed in use in September 1975. See Figure 15-1 and 15-2.
Abbreviations Used
NPDES National Pollutant Discharge Elimination System
NYSDEC New York State Department Environmental Conservation
EPA Environmental Protection Agency
RAS Return Activated Sludge
WAS Waste Activated Sludge
DCG Dual cell gravity sludge thickener unit
SVI Sludge Volume Index
gpsfd gallons per square foot per day
gpd gallons per day
gpwk gallons per week
gpm gallons per minute
SWD side water depth
HP Horse power
(L Oxygen
hr hour
D.O. Dissolved Oxygen
ACKNOWLEDGMENTS
We are very grateful to the various Company X officials: Vice President,
Chief Engineer, and Chief Operator and shift operators for their support,
cooperation and patience in making the successful transition from con-
ventional operation to sludge age control operation under severe operating
conditions. Their personal interest was most appreciated.
Mr. Gerald L. Strobe!, P.E., Chief Engineer of Ecology & Environment,
Inc., Cheektowaga, New York and his Laboratory Chief, Rocco Termini,
provided excellent and high quality technical assistance in analyzing
and reporting a multitude of data.
233
-------
SEWAGE TREATMENT PLANT LOG
ro
CO
A.M
7
I
9
10
11
12
PJM
1
2
3
4
S
6
7
g
9
10
11
12
AJM
1
2
3
4
S
6
RAW PUMP PRESS
#1 SURGE TANK LEVEL
n SURGE TANK LEVEL
LIFT STATION LEVEL
1
IE
Q.
(C
m
DC
AIR COMPRESSOR PRESSURE I
CHLORINE PRESSURE I
CHLORiNE FtOW - LB/DAY I
„
CHART FLOW 1
SCALE INCHES 1
SLUDGE BED DEPTH I
D.C. G. UNIT
FLOW G.P.H. RATE
PMP STROKE % ANJONIC
PMP STROKE
RATE%CATONIC
% SOLIDS IN
I
S
SS
METER READING
METER
PRESENT
PREVI'OUS
FLOW IFT3)
FLOW GALS
FIREMAIN
TERMINAL
MN PLT S. T. PLT
'
OUTFALL
CONTROL TEST
MLSS
AERATIOM
'A HR SETTABLE
SOLIDS
TIME
1 HR SETTABLE
SOLIDS
R.AS. RETURN
FLOW GPM
i i
FIELD MEASUREMENTS
RAW
a
1
£
*.
1
SLUDGE WASTED
GALS
MLSS
SURGE TANK!
1 HR SETTABLE
SOLIDS
.
I
y
1
LIFT STA.
3
1
I
1st SHIFT OPERATOR
2nd SHIFT OPERATOR
3rd SHIFT OPERATOR
PLANT SUPERVISOR
AERATION
« HR. SETTABLE
SOLIDS
1 HR SETT ABLE
SOLIDS
1
,-
CLARIFIER
1 HR SETTABLE
SOLIDS
X
TEMP °C
WEIR
1 HR SETTABLE
SOLIDS
£
TEMP °C
RESIDUAL
CHLORINE
OUTSIDE TEMP. °C
Figure 15-1. STANDARD OPERATOR'S DAILY REPORT FORM (Adopted 9/75)
-------
Date.
AVERAGE DAILY PARAMETERS
PARAMETERS
TOTAL SOLIDS
TOTAL SOLIDS FIXED
TOTAL SOLIDS VOLATILE
SUSPENDED SOLIDS
SUSPENDED SOLIDS FIXED
SUSPENDED SOLIDS VOLATILE
DISSOLVED SOLIDS
SETTABLE SOLIDS % HR
SETTABLE SOLIDS 1 HR
SLUDGE VOLUME INDEX
DISSOLVED OXYGEN
B.O.D. - S DAY
PUTRECIBILITY OR
METHYLENE BLUE
ALKALINITY
ACIDITY
PH
NITRATE (N)
NITRITE (N)
AMMONIA (N)
PHOSPHATE
COLOR - TURBIDITY
OIL / GREASE
OIL
GREASE
RESIDUAL CHLORINE
C.O.D,
FECAL COLIFORM
ZINC
CHROMIUM
TEMPERATURE
RAW
PPM
IBS
LIFT STA.
PPM
LBS
AERATION
PPM
LBS
CLARIFIER
PPM
LBS
WEIR
PPM
LBS
Figure 15-2. STANDARD OPERATOR'S DAILY REPORT FORM
CAdopted 9/75)
235
-------
A special note of thanks to Gerald L. Devlin, P.E., my partner and
Vice President of Brown-Devlin Associates, P.C. who provided invaluable
technical expertise, counsel and advice throughout the project and pre-
paration of this paper.
BIBLIOGRAPHY
1. BIBLE, Proverbs 3:19
2. BIBLE, Proverbs 3:13-15
3. Recommended Standards for Sewage Works "Great Lakes - Upper
Mississippi River Board of State Sanitary Engineers" (1968)
4. GARRET, M. T., Jr., "Hydraulic Controls of Activated Sludge Soluble
Rate", JSIW, 30 253 (1958)
5. JENKINS, D. & GARRISON, U. E., "Control of Activated Sludge by Mean
Cell Residence Time", JWPCF 40 1905 (1968)
6. LAWRENCE, A. W. & McCARTY, P. L., "A Kinetic Approach to Biological
Wastewater Design & Operation", Cornell University 1969, ASCE
JSED Vol. 96 No. SA3 June 1970 p. 757
7. WALKER, L. F., "Hydraulically Controlling Solids Retention Time
in the Activated Sludge Process", JWPCF 43, 30 (1971)
8. BROWN-DEVLIN ASSOCIATES, West Seneca, N.Y., Nov. 1972, "Wastewater
Facilities Reports" (24 MGD tertiary Waste Treatment Plant utilizing
single stage nitrification, sludge age control and tertiary Phosphorus
Removal using Alum)
9. LAWRENCE, A. W. & BROWN, C. G., "Biokinetic Approach To Optimal
Design and Control of Nitrifying Activated Sludge Systems", presented
at the Annual Meeting of the New York Water Pollution Control Asso-
ciation, New York City, N.Y., Jan. 23, 1973 (Accepted for publication
late 1976 in JWPCF)
10. EPA Process Design Manual for "Nitrogen Control", Oct. 1975
11. DICK, R. I., "Role of Activated Sludge Final Settling Tanks",
JSED Proc. ASCE 96 No. SA2 p. 423 (1970)
12. METCALF and EDDY, "Wastewater Treatment" Textbook, McGraw Hill
1972
13. EPA Technology Series, "Wastewater Characterization for the Specialty
Food Industry", Dec. 1974
236
-------
CONTROL OF COLOR PROBLEMS DURING RECYCLING OF FOOD PROCESS WATERS
A. Hydamaka, P. Stephen, R. A. Gallop and L. Carvalho
Food Science Department, University of Manitoba
Winnipeg, Manitoba R3T 2N2 Canada
INTRODUCTION
Of immediate concern to the food processing industry are the legislative
guidelines with deadlines set for 1977 and 1983, for reduction of waste
effluents and the use of best available technology, with the aim of zero
discharge by 1985.
To deal with the above guidelines, the EPA and the U.S. food processing
industry began in 1970, a national program of research, development and demon-
stration of pollution control methods. The First National Symposium on
Food Processing Wastes (1) suggested that goals include the development of
processes capable of high degrees of treatment, the design of processing
methods to reduce the quantity of water required, the possible recycling of
water, the development of profitable by-products from the recovered processing
"wastes", and the ultimate disposal of sludges and other residues.
There are only limited reports in the literature, where water recycling
in the food industry has become a reality (2, 3, 4, 5, 6, 7). As an example
of what can be widely achieved, this paper will discuss the potential of
a closed-loop recycle system at the rinse stage of potato processing as part
of the "total systems" concept approach to the food industry. Water used
at this processing step will leach from the surface of the potato slices
color forming compounds such as enzyme substrates and enzymes. The colored
process water may be undesirable from an aesthetic viewpoint as it is foamy
as well, with appreciable microbial populations, and also may promote further-
development of browning if recirculated upon freshly exposed food tissue,
while also discoloring the latter. Browning reactions are important for
quality control in terms of the alteration of appearance, flavor, nutritive
value and possible toxicology.
Conventionally, potato processing plants dump their wastes to municipal
sewage disposal systems. This practice is no longer usually acceptable as
sewage surcharge costs have increased dramatically in recent years. Primary
and secondary treatment facilities on-site, only duplicate the municipal
waste treatment system. Industry is faced with increased costs to meet the
proposed guidelines for 1977 and 1985. For frozen potato products, the maximum
daily discharge in Ib/ton is set at BOD, 9.5, TSS, 17.5, in 1977, and
BOD, 1.6 and TSS, 2.7 in 1983 (8).
237
-------
Most of the pollution problems faced by the food industry have arisen
because for too long the industry has followed the "linear-process" once-use
system (Fig. 1) rather than cyclic systems (Fig. 2), with waste treatment
thought of as an added investment yielding no return as a result. EPA
studies conclude that at least 100 food processing plants will soon go out
of business because of pollution controls and at least another 300 may be
similarly affected (9).
The potato processing industry has made significant advances in recent
years in pollution control at the slice-rinse stage, by recovery of products
for commercial sale, and in reduced water use. The C-E Bauer Company (5)
reported that by use of the hydrasieve screen and cyclone, a potato processing
plant received $439 per day in solid waste products, and also saved $18,000
per year in sewage charges. Net pay-off of the initial investment of
$135,000 was about one year's operation. The J. D. Ferry washer system (6)
resulted in an annual savings of $11,000 to a potato plant. The Perfect
Potato Chip plant in Decatur, Illinois has installed the Sweco recovery
system, and expects to earn an annual gross revenue of some $10,000 through
starch reclamation (7). The above systems demonstrate to the food processor
that in some areas pollution control at the sources can be a profitable
venture.
The slice-rinse process proposed by the authors would include the above
systems for removal of solids such as starch. However, there would be complete
recycling of the starch-free rinse water with only a minimal amount of
make-up requirement. For recycling of the rinse water, quality control
factors such as physical, physical-chemical, biochemical and biological
factors, must be regulated. Included are the factors of "color, odor, foam,
and bacterial build-up." The use of activated carbon may be the essential
key to controlling all these factors. The present practice of chlorination
for bacterial control (now a controversial topic in wastewater management
due to potential health problems), the costs, supply problems, and the need
for massive dilution, requiring much fresh make-up water, would be eliminated.
The use of activated carbon and water recycling is part of our total
systems concept to the food industry. The concluding remarks of the Fourth
Symposium (1973) (10) suggested that this approach is the most ideal solution
to water and waste management in the food processing industry. Using this
approach former elements of waste become as important as the intended commer-
cial products.
The agricultural-food processing industries are largely producers of
solid wastes and wastewater. In the case of agronomic solid wastes, the
industry is faced with the disposal of massive amounts of wet carbohydrate
material, often of low economic value yet presenting a major potential
for costly treatment and disposal, so as to avoid causing pollution. Such
"wastes" are important on-site, to the total system concept proposed. The
authors have found that conversion of such wastes into activated carbons,
has produced adsorbents which are equal to or superior to commercially
available carbons (11). These carbons are then highly used in cyclic loops,
to rapidly reclaim process water, to permit continuous recycling of such
"aids" at each step, to a very high degree, without serious problems.
238
-------
n
DEEP
RECOVERY
SYSTEM
ENVIRONMENT
I
Figure i. Efficiency vs Inefficiency
Cyclic System vs Linear System + Pollution Control
$ dx vs $ (x + x2"*n)
239
-------
Figure 2. AUTOMATIC, HIGH EFFICIENCY QUALITY CONTROL SYSTEMS
-CYCLIC PROCESS
TISSUE
IN
FT1
CLE
IDEAL LEVEL
APPROX.BALANCE
Fl + F3 =F2 + F4
Fl _F2 = F4-F3
U)
o
z>
H
CL
CO
CO
LL)
O
o
o:
CL
RED ZONE
STOP AND
CORRECT
COURSE
•ORANGE ZONE
SLOWDOWN
AND DRIVE
CAREFULLY
GREEN ZONE
HIGH-SPEED
SAFE TRAVEL
Symbols:
P.M.— Purifier modules, one on load, one on standby
W — Waste solids , to use or destruction
R —Reject for re-purification or discharge
G—Accept for re—use
V —Cyclic base volume
dV—Make —up volume/cycle
n— No. of re —use cycles
-------
The potential use for carbon in such systems has been enhanced by a
fast process, the "Atomized Suspension Technique", which regenerates and
recovers powdered carbon in about 30 seconds (12) at low costs and high
yields (90% recovery and 95% of reactivation of adsorbing capacity).
The Proper Sequencing of the Sciences in Engineering Systems
Nature has obviously been created in a hierarchical, rational pattern
beyond basic change by man, which we must adequately understand and respect
if our practical engineering systems are to succeed, plus be efficient and
economical. Figure 3 schematically shows the "power-structure" of the
sciences, and of their interdependency in practical situations on each other,
and of course, on the great constant, time, to which all of them but mathe-
matics, which preceded, and which rules all of them in quantitative terms,
must submit.
All processes, in natural and man-made systems, must follow the sequence
of the sciences shown in Figure 3 qualitatively and quantitatively, in time.
Thus if we gather such data on a particular process, as we do, step by step
in a process line, then we can easily see how to improve the process itself,
its control, efficiency, and cost. Then predictive ways to greatly improve
processing equipment, product quality control, and its mirror-image, reduced
water and waste-management problems, along with smaller, simpler systems
for handling them, plus the costs and environmental constraints involved,
are all opened up widely to our advantage. In Sweden, only the most efficient
processors have survived the environmental constraints, and those who improve
their processes as we suggest should do well competitively in every way.
The Q. C. Criteria in Hater-Food Relationships
The quality characteristics of each grade of acceptable water must be
defined by the intended role/s for it. The human judgements expressed as
quality control criteria about the relative significance of the physical,
physico-chemical, biochemical and biological factors (see Figure 3) likely
to be in the water, or required to be in it, to enable the intended function
of it to be achieved efficiently express the maximum limits of undesirables
and the minimum limits of desirables (Figure 4), that should be in each grade
of water for each function.
The concept of "background levels" is very important in the management
of resources, especially water, air, land, and food composition. It is
foolish to try to work below natural background levels of exposure in most
situations. Over-purification of water before and after use, in huge volumes,
is a major cause of inefficiency in water-management. In virtually "closed-
loop" cyclic systems, purification needs and volumes/cycle are always minimal.
With repetitive use, all factors build up to equilibrium background levels
at appropriate rates (see Figure 5). The only purification needed between
uses is to keep each factor below a level where they can cause significant
defects in the system (e.g., by failing to remove enough residual soil each
cycle from foods being washed). The defects (pollutants) added to the
water each pass can be determined and adjusted; systems for their automatic
water for them can be determined and adjusted; systems for their automatic
deduction from such a flow (at appropriate rates equal to or greater than
241
-------
f\3
-fc.
no
Figure 3. PRACTICAL SEQUENCE OF THE SCIENCES, WITH AUTOMATIC Q.C.
PHYSICAL
MATHEMATICAL 11
•II. _
PHYSICO- CHEMICAL
BIOCHEMICAL
o:
o
o
LL
O
BIOLOGICAL
TIME
-------
Figure 4. QUALITY CONTROL CRITERIA
PO
-p.
CO
in
UJ
_1
ca
<
gc ...._
CO
UJ
O
CO
UJ
_l
CD
CO
UJ
O
2
13
D
B
NATURAL ,0 , .,
MAXIMUM (Background)
A
GRADES
NATURAL
MINIMUM
Q.C. GRADES ARE BENEFIT/COST JUDGEMENTS.
based on Numbers, derived from scientifically
sound, Sampling, Analytical, and Statistical
Procedures , for the determination of Physical,
Physico — Chemical, Chemical—Biochemical,
Biological, Aesthetic, and Economic Factors.
JUDGEMENTS ABOUT Q.C. GRADES.
decide the relative significance in quantitative
terms, of these Factors, in reasonable
priorities , in a Process step, or sequence of
same, for given functions.
(Background!
-------
Figure 5. REDUCING SUGAR CONTENT OF RECYCLED
FRENCH FRY RINSE WATER
lOOOr
rinse 1-8
centrifuge (I) rinses 9-13
centrifuge (2) rinses 14-18
centrifuge (3) rinses 19-23
surface wash
0
345
Trial Re-use
-g
6
~6
244
-------
their likely rate of addition with recyclic water use) should then be
installed in-line (just like a thermostatic cooler for heating systems, or
a cartridge filter on our car crankcase oil cyclic system) to maintain the
quality of the water within the green or orange-light zone (see Figures 2 & 3)
almost infinitely. If we set-up such systems at each function where water
is used to deduct excessive levels of undesirables and make-up deficiencies
of desirables (e.g., as is done for condensers and boiler-feed waters), and
is readily possible now, by on-line cartridge devices, in most situations,
very high degrees of re-cycling on-site become readily possible in a wide
range of situations, at slight costs, in every way.
As the recycling number "n" is increased, thevinput volume V required,
and the supply equipment capacity drops from V to —, the discharge "blowdown"
for excess solute removal (as in a boiler) becomesnL pumping costs will
rise somewhat, the collection and treatment system becomes ^ in volume, and
very much less than this in operating costs. And finally, significant
environmental pollution can be avoided when returning the spent materials
to nature.
If our cars were built the way most of our industry, university,
governmental and domestic water-using systems are designed (e.g. on the
linear flow-sheet) they would be unusuable. We can only use and afford them
because their water and oil systems were rationally designed on the cyclic
model almost a century ago. A moment's reflection on this paradox gives
us an idea of how much productivity and freedom we can readily regain by
using sensible engineering systems for water, air, and other processing
fluids everywhere at costs very much lower than we are now carrying,
with great gains, in many ways, while conserving and protecting our resources
and environment.
EXPERIMENTAL SECTION
Enzymic browning in the food processing industry has always posed
problems for food processors. Such enzyme systems also must be considered
when dealing with a closed loop cyclic system in the rinsing of french
fry slices.
Browning of potato tissue on exposure to air (oxygen) is the result of
action of the phenolase complex on organic constituents present in the potato.
This enzyme or complex can be divided into two types of reaction: the phenol
hydroxylase or cresolase activity, and the polyphenol oxidase or catecholase
activity. In the potato tuber, Schwimmer and Burr (13) have found that the
main substrate for the phenolase enzyme is L-tyrosine. It has been shown
that tyrosine is first converted to a diphenol compound by cresolase activity
and that this intermediate compound (3, 4-Dihydroxyphenylalanine) is oxidized
by catecholase activity into a corresponding quinone (o-quinone phenylalanine).
Removal of hydrogen results in the formation of a red compound, dopachrome
(5, 6-quinone indole-2-carboxylic acid) which subsequently undergoes poly-
merization to form brown melanin compounds (14).
It is obvious, therefore, that by removing substrate and/or enzyme from
solution, the browning reaction can be controlled. Activated carbon has long
been used to remove color and odor from various solutions, and was therefore
245
-------
used in this study, to attempt to control the browning problem found in
recycled rinse water.
The role that carbon would play in the reduction of "browning" was
observed by measuring the activity of the polyphenol oxidase enzyme and the
level of hydroxylated aromatic compounds present in solution, before and after
treatment.
Method for Determination of Enzyme Activity
The activity of the polyphenolase enzyme was measured by calculating
the initial velocity of the enzyme when a standard volume (0.20 ml) of potato
rinse water was combined with a standard volume (3.0 ml) of buffered catechol
solution (.01M, pH 6.0).
The equipment used for measurement of the initial velocity of the
enzyme included a Coleman UV-111 spectrophotometer, a Unicam SP45 Concen-
tration Readout Unit (used" to set the range of the recorder from 0.0-0.20
absorbance) and a Unicam AR25 linear recorder.
PROCEDURE
A 0.01M catechol solution (pH 6.0) was prepared by pipetting a 15 ml
aliquot of a 0.2M disodium hydrogen phosphate solution and a 110 ml aliquot
of a 0.2M sodium dihydrogen phosphate solution into a 250 ml volumetric
flask, adding 0.2753 g of categhol and making the flask up to volume with
deionized water. The solution was aerated at room temperature before analysis.
The wavelength used for the analysis was 410 nm. A "control" solution
was used to set the absorbancy of the spectrophotometer to 0. Three ml of
the buffered catechol solution was placed in a 10 mm cuvette. To this was
added 0.20 ml of deionized water. The cell was inverted several times and
the instrument adjusted to 0 and the scale expansion unit and recorder set
so that full scale deflection would be achieved at an absorbancy of 0.20.
The above procedure was repeated for an actual sample, but on addition
of the 0.20 ml of potato rinse water the recorder was started and the cell
immediately inserted into the spectrophotometer to enable a "progress curve"
to be obtained for the enzyme.
The initial velocity was obtained by drawing a tangential line to the
curve approximately 48 seconds after the reaction had begun. The enzyme
activity was expressed as the change in absorbancy units/minute.
Method for "Tannin" Determination (hydroxylated aromatics)
The method followed was that found in "Standard Methods for the Examina-
tion of Water and Wastewater" (15). Although the results are reported as
"tannin" they more properly should be referred to as "hydroxylated aromatic"
compounds.
The standard used for the preparation of the standard curve was tannic
acid and the wavelength used for analysis was 650 nm.
246
-------
Carbon Treatment of Rinses 1-5 of Potato Rinse Water
French-fry slices were rinsed with tap water (solid:liquid ratio of
3:10). The rinse water was then centrifuged, filtered through glass-fibre
filter paper (Whatman GL/C) and treated with either 0.10%, 0.30% or 0.50%
dosages of "Aqua Nuchar" powdered carbon. The samples were placed on a
rotary shaker at 300 rpm for 15 minutes to ensure maximum adsorption.
The adsorption isotherm for tannin and enzyme (Figure 6) shows that carbon
is a good adsorbent for these factors. Both factors can be reduced to low
concentration levels. Samples were taken before and after treatment for
analysis of "tannin" and enzyme activity. Carbon was removed from solution
by filtration through glass-fibre filter paper. This "rinse and treat"
combination was repeated for 5 rinses.
The object of this procedure was to obtain the minimum dose of powdered
carbon that would effectively control enzyme activity and would remove
"tannin-like" materials from solution. Incomplete removal of the enzyme would
only lead to a gradual buildup of the enzyme in the water loop, which would
soon bring about further browning as soon as fresh substrate contacted the
enzyme.
From Figures 7(a) and 8(a) it can be seen that a dose of 0.10% is
very effective in lowering both tannin and enzyme activity, but is not sufficient
to prevent accumulation of both enzyme and substrate at; each rinse stage.
Although the carbon does produce "aesthetically" acceptable water at this
level, i.e., there is little evidence of browning, this would be reversed
rapidly at higher rinse stages when the recycled water came in contact with
more substrate.
A level of 0.30% (Figures 7(b) and 8(b)) is much more effective in
controlling both enzyme and substrate and the rate of increase of both
parameters with increasing rinses is decreased.
At a level of 0.50% (Figures 7(c) and 8(c)) both parameters are almost
static after the first rinse. The residual enzyme activity was not measurable
up to and including the fifth rinse.
From the preceding work it can be seen that despite the fact that relatively
small dosages of carbon are required to produce aesthetically acceptable water,
relatively large dosages of carbon are required to limit the increase in
enzyme activity with successive rinses.
i fr
Therefore, attention was given to other possible treatments that the rinse
water might receive to halt the progress of the enzyme. Many substances
have been used to inhibit the enzymatic browning of food products. Such
substances include ascorbic acid, sodium insulfite, sodium chloride, malic
and citric acids.
There are advantages and disadvantages in the use of any of these compounds
Some compounds such as ascorbic acid are not suited because of cost while
others such as bisulfite can produce objectionable flavors and odors in the
food product itself and additional wastewater problems. Citric acid was
247
-------
IN)
£
2.40
2.25
2.10
1.95
I.8O
1.65
1.50
1.35
1.20
1.05
0.90.
TANNIN
CF =mg/l (TANNIN)
= AABS/MIN (ENZYME)
-0.30 -0.15 -0.00 0.15 0.30 0.45 0.60 0.75 0.90 1.05
Log CF
Figure 6. ADSORPTION ISOTHERM FOR TANNIN AND ENZYME REMOVAL
1.20
-------
O.I5j
0.12
0.09
Q06
0.03
* ENZYME ACTIVITY UNTREATED WATER
o ENZYME ACTIVITY CARBON TREATED A
WATER /1
(a)
Ql%(w/v)
ao
0.08
- 0.06
H
p 0.04
8
N
UJ
0.02
0
0.04
OJ03
0.02
0.01
0
- (b)
0.5% (w/v)
i -
12345
SLICE-RINSE RECYCLE
Fi gure 7. CONTROL OF ENZYME ACTIVITY BY ACTIVATED CARBON
249
-------
25
20
15
10
5
0
16
2 \2
DC.
Ul
O
* TANNIN CONC. UNTREATED WATER
o TANNIN CONC. CARBON TREATED
WATER
(a)
8
8 4
z
z 0
2345
0.3%(w/v)
j i i
0.5% (w/y)
12345
SLICE-RINSE RECYCLE
Figure 8. CONTROL OF TANNIN BY ACTIVATED CARBON
250
-------
chosen mainly because of cost and its successful use as a constituent and
additive in many food products.
Effect of pH on Enzyme Activity
French-fry slices were rinsed with water (water:sol ids ratio of 3:10)
to produce a total volume of 1500 ml of rinse water. The water was recycled
over fresh potato slices five times to achieve a fairly high level of enzyme
activity. Samples of 100 ml of the final rinse water were adjusted in pH
from 3-7. Following this adjustment the samples were dosed with 0.10 g
of carbon and the enzyme activity determined on each sample. Residual
activity of the enzyme was expressed as a percentage of the activity of the
untreated rinse water at the fifth rinse. Results of this work can be seen
in Figure 9. The enzyme was almost completely inactivated at a pH of 3,
and at a pH of 5.5 only 16% of the original activity remained.
Thus by altering the pH it appears that most of the enzyme is inactivated,
and at the same time the active portion of the enzyme which remains is severly
inhibited by the lowered pH. _,
From the point of view of cost, it would appear that the pH of the rinse
water should be maintained at a pH of between 5.0-5.5. An experiment where
rinse water is "rinsed and treated" (pH adjusted to 5.5) is shown graphically
in Figure 10. It is to be noted that the pH has little if any effect on the
tannin concentration but that the enzyme activity is held constant when
the pH is kept at 5.5.
The greatest dosage of acid required is in adjusting the pH of the original
feed water. To adjust the pH of 500 ml of tap water to 5.0 required 0.25 ml
of 1M citric acid. After each rinse only 1.1, 1.0, 0.35, 0.70 and 0.25 ml
of 0.1M citric acid were required to maintain the pH at 5. Thus of the total
acid added 67% was required to lower the pH of the original feed water.
Therefore the major cost of the use of citric acid would be in the initial
dosing of the "start up" and "make up" water used in the process. The amount
required at this stage would be dependent upon the buffer capacity of the water
at any plant location.
As citric acid is an organic acid it may be argued that addition of such
an acid would add considerably to the oxygen demand of the recycled water.
Up to the 5th rinse stage pH adjustments from 4.5-5.5 in our series of experi-
ments would add in theory between 100-340 mg/1 to the chemical oxygen demand.
However, in studies where potato rinse water (5th rinse) was adjusted to
pH levels ranging from 4.5-6.75 and then dosed with up to 0.2% of powdered
carbon, it can be seen from Figure 11 that, in fact, at the maximum dosage
of carbon, samples lowered to a pH of either 5 or 5.5 had a slightly lower
COD than the sample which was unadjusted. This result may be explained by
the fact that the additional COD of the citric acid is compensated for by
removal of slight amounts of protein which are denatured upon acid addition,
and then removed in the filtration step.
251
-------
100
90
o
UJ
cr
o
o
.-* 80
>
B 70
c
I 60
! 50
O)
§ 30
I 20
C
& 10
w
(O
O
CARBON CONC.(O.I%w/v)
% 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
PH
Figure 9. THE EFFECT OF ACTIVATED CARBON AND pH ON ENZYME
ACTIVITY OF RECYCLED POTATO RINSE WATER.
252
-------
50
r\>
01
CO
9
X
40
cn^
E \
yffiso
8«
20
LJ
S >o
LJ
0
+ TANNIN
o TANNIN
* ENZYME
• ENZYME
0
I
pH6.9
pH5.5
pH6.9
pH5.5
2345
SLICE- RINSE RECYCLE
Figure io. CONTROL OF COLOR FORMATION BY pH CONTROL
-------
O pH 6.75 NON ADJUSTED
* pH 4.5'
X p H 5.0
• pH 5.5 J
ADJUSTED WITH
CITRIC ACID
1850
0.00 0.05 0.10 0.15
GRAMS CARBON
(100ml)
0.20
Figure 11. EFFECTS OF CITRIC ACID AND ACTIVATED CARBON ON
THE CHEMICAL OXYGEN DEMAND OF RECYCLED RINSE
WATER (5X).
254
-------
CONCLUSIONS
The work to date at the Food Science Department, based on a complete
closed-loop recycling system, has been performed only on a laboratory scale.
To evaluate the economics of the process it would be necessary to simulate
actual plant conditions by scaling the design to at least a pilot plant stage.
It has been demonstrated in other studies that starch recovery systems can be
a profitable venture to the potato industry. The cost of a carbon design
system on a plant scale, needs to be evaluated. As part of the total system
concept, activated carbon could be produced on site from "waste organic solids"
of the industry, or from farms (e.g. straw, cobs), or from urban areas (e.g.
newspapers). We have used these materials for about a decade for these pur-
poses. Commercially powdered activated carbon sells for as little as 9
cents per pound (1973 data) (16) and regeneration shows a cost reduction of
up to 70% from the cost of continuously replenishing powdered carbon. On
the laboratory scale, carbon costing cannot be accurately evaluated because of
scale-up variables such as clean water make-up.
As indicated in the report a combination treatment of pH and carbon may
also prove to be beneficial. Control of enzyme activity by chemical means
alone is not recommended for the closed loop recycling system. Activated
carbon is felt to be an essential ingredient in the control of color. Carbon
at the same time removes other factors such as COD, odor, and foam. Studies
in our laboratory have shown that carbon is also effective in controlling the
possible bacterial build-up that could occur with continuous reuse, as predicted
by Figure 3.
References
1. EPA. Proceedings First National Symposium on Food Processing Wastes,
Portland, Oregon. Water Pollution Control Research Series 12060-40/70.
1970.
2. EPA. In-Plant Control of Food Processing Wastewater. Pollution Abatement
in the Fruit and Vegetable Industry, Volume 2. Washington, D.C. 1975.
3. Gallop, R. A. and Hydamaka, A. W. Water Management Systems for Potato
Processing Plants. Canada Department of Environment, Ottawa. August 1973.
4. Maxwell, W. A., Rogers, C. J. and Jackson, Gilbert. WateReuse,
Recycling and Energy Savings in the Poultry Processing Plant. National
Conference on Complete WateReuse, p. 291. April 1973.
5. Taylor, J. A. The C. E. Bauer Hydrasieve and High Efficiency Liquid
Cycline Equipment. Canadian Potato Chip Association Environmental
Seminar. Toronto, 1973.
6. Krolopp, 0. C. Raw Potato Slice Washing and Waste Water Cleanup.
Canadian Potato Chip Association Environmental Seminar. Toronto, 1973.
255
-------
7. Pettay, B. You Can Make Money in Pollutional Control. Potato Chipper,
p. 50. January 1975.
8. Air/Water Pollution Report, Vol. 11, No. 45, p. 447. November 1973.
9. Agri Division, Dunlap and Associates Inc. Economic Impact of Environmental
Controls on the Fruit and Vegetable Canning and Freezing Industries,
for the Council on Environmental Quality, p. 585. 1971.
10. Proceedings of the 1973 EPA-Cornell Agricultural Waste Management Con-
ference. Food Processing Waste Management, p. 318. New York, March.
11. Gallop, R. A. Activated Carbon from Wastes. Proceedings of the First
National Conference on Water Pollution Research. University of Toronto.
pp. 1-12. February 1971.
12. Chemical & Engineering News-. Process Regenerates Powdered Carbon.
p. 7. December 15, 1975.
13. Schwimmer, S. and Burr, H. K. Potato Processing. Structure and
chemical composition of the potato tuber, p. 12. AVI Publishing Company,
Westport, Connecticut. 1967.
14. Eskin, N. M., Henderson, H. M., and Townsend, R. J. Biochemistry of Foods.
Academic Press, Inc., New York, NY. pp. 70-71. 1971.
15. American Public Health Association, American Water Works Association,
and Water Pollution Control Federation. Standard Methods for the
Examination of Water and Wastewater. pp. 346-347. 1971.
16. Loven, A. W. Perspectives on Carbon Regeneration. Chemical Engineering
Progress, Vol. 69, No. 11, p. 56. November 1973.
256
-------
THE TOXICITY OF FOOD PROCESSING EFFLUENTS TO FISH
by
D.W. Bissett*
Whenever the word "toxicity"is used to describe the effects of
a certain material on some area of the environment, an alarm mechanism
sound in most people's minds. The word "toxic" is synonymous with
"poisonous" or at minimum, "harmful". In this context then, the Environ-
mental Protection Service of Environment Canada has incorporated toxicity
testing of effluents into its regulation development program to control
the harmful effects of these effluents to the aquatic environment and
in particular, fish.
One might well ask the question, "Why emphasize fish toxicity
testing in developing water pollution control regulations?" Indeed this
is a valid question as most effluent regulations usually specify relatively
common parameters such as BOD, suspended solids, ammonia nitrogen, phosphate
or pH.
In the United States, the enabling legislation with regard to
water pollution is the Federal Water Pollution Control Act passed in 1972.
This legislation was specifically designed to control water pollution by
the setting of standards respecting levels of point discharges to navigable
waters.
In Canada, there are a number of legislative documents relating
to water and water pollution in general. However only one Act, the
FISHERIES ACT, authorizes the Governor in Council to set regulations
respecting the deposit of deleterious substances into waters frequented
by fish. This act was originally passed in 1868 for the protection of
the Canadian fishery resource and indeed, a number of sections deal with
provincial rights to licence, distance of nets from one another, possession
or sale of fish, etc. Only in 1971 when Environment Canada was organized
from a number of other federal departments, was the act amended to permit
the establishment of regulations limiting the discharge of substances
which are harmful to fish or to the use of fish by man.
In order to establish regulations for specific industrial
discharges under the Fisheries Act, it is absolutely essential to evaluate
the effects of industrial wastewater effluents on fish. For each industrial
sector, a rather extensive programme of toxicity testing is carried out to
determine how acutely lethal effluents are and how effective existing
treatment methods are in removing toxicity. This paper describes in general
the toxicity testing methodology employed by EPS laboratories located across
Canada and in particular results which have been obtained for food processing
effluents.
*D.W. Bissett, Water Pollution Control Directorate, Environmental Protection
Service, Environment Canada, Ottawa, Canada
257
-------
What Does Toxicity Testing Involve?
In its crudest form, a fish toxicity test simply involves taking a
bucket of effluent, throwing in some fish and observing the results. No
doubt, the most extreme effect would be death, but one would not know if
the mortality was due only to a lack of oxygen available to the fish or
to some other toxic component in the effluent.
Before discussing the actual fish bioassay procedures in detail, a
general introduction of the terminology and techniques would be relevant. The
four most common terms used to describe toxic effects (1) are as follows:
acute - response occurs rapidly, usually within four days
chronic - involves a response which is lingering or occurs
over a long period of time
lethal - causes death or sufficient to cause it
sublethal - below the level which directly causes death
As a first assessment of industrial effluents, EPS has been looking
at acute lethal toxicity: that which causes mortality in a short period of
time. Lethal toxicity is usually expressed in terms of an LC50 (similar
terms are TL50 or TLM), which is the concentration lethal to 50 percent of the
fish over a specific time interval. A more detailed account of toxicity
measurements can be found in Sprague's "The ABC's of Pollutant Bioassay
Using Fish"(3).
While a variety of test methodologies exist for evaluation of toxic
effects,^ ' the bioassay procedures for determination of an LC50 have been
well defined ( ' 'and involve the following:
1) a series of test vessels, each with a different but constant
concentration
2) a group of similar fish in each container
3) observations of fish mortality at regular exposure times
for the duration of the test period
4) calculation of the LC50 results expressed as a concentration
which causes mortality to 50% of the fish
Toxicity tests can be conducted in a static, semi-static or
continuous flow (flow-through) mode. In the static test, the fish are
placed in a container of standing test water for the duration of the test
whereas in the semi-static mode, at regular intervals (eg. 12 or 24 hours)
the same group of fish are exposed to a fresh test solution. In the continuous
flow test, the fish are subjected to a test solution which is continually re-
newed. Each method has its several advantages. In the main, the static test
is less expensive and time consuming whereas the continuous flow test is
generally more sensitive if carried out properly.
258
-------
Specific test conditions may vary according to the laboratory or
type of effluent being examined. Some of the more important items which
have become standardized in EPS testing programs are as follows:
1) The standard test fish species is rainbow trout
(Salmo gairdneri Richardson)
2) Variation in fish size is controlled within the test and
limited to less than 10 grams.
3) The fish are acclimatized and maintained in holding facilities
at water temperature of 15°C±2°C. Tests are conducted at
15°C±1°C.
4) Dilution water for holding and testing must be dependable,
clean and obviously non-toxic.
5) The volume of test solution must be sufficient so that the
fish will not deplete any toxic constituents by uptake
(usually 1.5-2 litres per gram of fish per day).
6) Aeration of the test solution should provide a minimum
dissolved oxygen of 8 mg/1.
7) The pH of the test solution should not be directly adjusted.
Selection of the concentrations of effluent to be used in the toxicity
test is another important part of the testing. For preliminary or exploratory
testing, the span of concentrations should not be very restrictive. Once
the general range is determined the span can perhaps be narrowed but it
should always be sufficiently wide to cause from zero mortality to complete
mortality at either end of the range. With food processing effluents,
unless the effluent is consistently non-lethal, test solutions should cover
the full range from control or no effluent to undiluted effluent. Most
commonly, a geometric or logarithmic series using concentrations such as
100, 56, 32, 18, 10 percent effluent by volume, is employed.
Analysis of Test Results
Prior to a discussion of the results from various food processing
sectors, a brief outline of the method of calculation of LC50 values is
useful. The laboratory results of a typical toxicity test are presented in
Table 1. Fish mortalities for the 5 test concentrations plus control were
observed over a 96 hour period. Two of the major parameters which should
be monitored at regular intervals throughout the test are dissolved oxygen
and pH. Note the low initial dissolved oxygen concentration in the undiluted
effluent, a typical occurrence with untreated or partially treated food
processing effluents.
Determination of the time period where all observable responses
occur may aid in the identification of possible causes of mortality and
perhaps in future work allow the duration of the-test to be shortened.
259
-------
TABLE 1
TYPICAL TOXICITY TEST RESULTS
CONCENTRATION
% BY VOLUME
100
56
32
18
ro
o 10
CONTROL
DISSOLVED
OXYGEN
mg/1
3.8-9.5
6.3-9.4
7.0-9.2
7.6-10.0
8.4-10.6
8.4-10.8
INITIAL
pH
8.6
8.45
8.1
7.85
7.7
7.7
% MORTALITY
at (HOURS)
1
0
0
0
0
0
0
I
0
0
0
0
0
0
1
0
0
0
0
0
0
2
0
0
0
0
0
0
4
0
0
0
0
0
0
8
10
0
0
0
0
0
24
80
0
0
0
0
0
48
100
0
0
0
0
0
72
100
80
20
0
0
0
96
100
90
40
10
0
0
-------
The results of Table 1 indicate that mortalities occurred over the entire
96 hour test period for various concentrations of effluent. Frequently,
with food processing effluents, all mortality will occur within the first
24 hour period, although with some meat packing effluents, mortality has only
begun to occur after 72 hours exposure to undiluted effluent. Unfortunately,
most EPS laboratories presently lack facilities and staff necessary for the
exact determination of the toxic constituents responsible for the death of
the fish.
A number of methods of calculation of the LC50 value exist,
probably the most common being the graphical determination after Litchfield
and WilcoxonW . This method involves plotting the dose (or concentration
by volume) against the percent effect (mortality) on logarithmic - probability
paper, as shown in Figure 1 for the previous results. Note that 0 and 100
percent effects are not accommodated on the graph paper and are thus indicated
with an arrow. A straight line is fitted to the points by eye, placing most
emphasis on the results where partial mortality was observed and minimizing
the total vertical distance between the points and the line. The LC50 value
is the concentration where a 50 percent mortality effect occurs, in this case
34 percent effluent by volume.
The method of Litchfield and Wilcoxon allows the determination of
the slope of the line and confidence limits and provides a means for approximating
the confidence limits of effects other than the median or 50 percent effect,
the details of which will not be presented here.
Toxicity Tests Of Food Processing Effluents
The results of toxicity testing of food processing effluents have
been summarized for the following industrial sectors :
a) Potato Processing
b) Meat and Poultry Products
c) Dairy Products
d) Fruit and Vegetable Processing
e) Associated Industries
EPS surveys have mainly been directed towards plants utilizing some
form of biological treatment as these plants would, if designed and operated
properly, produce effluents with acceptable water quality parameters such
as BOD or suspended solids.
Rather than discuss the results of each plant in detail, the major
emphasis will be directed to those end-of-pipe treatment processes that produce
non-acutely lethal effluents and to possible causes of toxicity at other plants.
261
-------
981—
D
O
I
(0
0)
<
>
a:
o
l
o
o:
HI
Q.
9O
8O
70
6O
50
40
30
2O
1O
LC 5O=34% by volume
I
I
J
1O 2O 3O 5O 7O 1OO
EFFLUENT CONCENTRATION (%) BY VOLUME
FIGURE 1 PLOT OF TOXICITY TEST RESULTS ON
LO6-PROBABILITY
262
-------
a) Potato Processing
Three potato processing plants in Canada presently employ secondary
biological treatment systems. The results of surveys at these plants as well
as several other operations are presented in Table 2.
Plant B employs two aerated lagoons in series, with approximately 10
and 25 day retention times respectively. Due perhaps to some non-biodegradable
residue, toxicity tests have usually resulted in complete mortality with full-
strength effluent while producing no mortality in an 80 percent by volume
dilution.
The french fry operation (plant A) employing two plastic-media
trickling filters in series has been the subject of considerable study in the
last two years. Although BOD and suspended solids removals are acceptable, the
final effluent has always exhibited a lower lethal concentration than screened
effluent. To date, no specific cause for this phenomenon has been found, except
that suppression of certain components such as ammonia and suspended solids has
not resulted in reduced toxicity. A study into possible toxic chemical residuals
is continuing.
Potato chip operations which employ pretreatment such as screening
discharge effluents which are extremely toxic having LC50 values ranging from
12-44 percent.
b) Meat and Poultry Products
The data presented in Table 3 summarizes the results of an extensive
survey of meat and poultry packing plants and rendering operations across Canada.
Perhaps the most important observation to be made is that aerobic
biological treatment systems such as extended aeration activated sludge (Plants
E and H) and oxidation ditch (Plant D) in all cases produced effluents exhibiting
no mortality to fish.
Where anaerobic/aerobic lagoons or stabilization ponds were evaluated,
on only one occasion (plant E) did no mortality occur in the treated effluents.
Indeed, in two cases (Plant B and F) the effluent following treatment was at
least as toxic as the raw waste thought to be due in most part to the high
ammonia concentrations generated in the anaerobic degradation of protein
material.
Although ammonia has not been identified as being the sole component
responsible for acute lethal toxicity, there is an apparent relationship between
the ammonia concentration in an effluent and the resulting LC50 value as shown
by the data presented in Table 3. The systems which produce non-lethal effluents
are those which have the capability to nitrify ammonia to the nitrate form.
c) Dairy Products
Evaluation of the effluents from dairy products plants has been carried
out under both summer and winter climatic conditions as shown in Table 4. Values
263
-------
TABLE 2
POTATO PROCESSING PLANTS
PLANT §
PRODUCT
A
FRENCH FRIES
B
FRENCH FRIES
ro
o>
c
FRENCH FRY AND
DEHYDRATED
PRODUCTS
TREATMENT
TRICKLING
FILTER
TWO
AERATED
LAGOONS
EXT. AERATION
6 3 LAGOONS
EFFLUENT
SAMPLE
SCREENED
FINAL
SCREENED
FINAL
SCREENED
SEC. CLAR.
FINAL
BOD
mg/1
2050
115
1735
100
1385
9
<10
TSS
mg/1
1748
89
.
-
560
10
5
NH3-N
mg/1
4.7
12.3
5.4
5. 5
44
2.9
0.8
LC 50
%
32-56
18-32
50-100
85-100
25-50
ND
ND
FRENCH FRY
NO TREATMENT
RAW
1700
2480
(pH-11.8)
13.5
POTATO CHIPS
(SEVERAL PLANTS)
SCREENING
SCREENED
590-2100
600-1760
9.3-70.3
12-44
ND - NO*MORTALITY IN FULL-STRENGTH EFFLUENT
-------
TABLE 3
MEAT AND POULTRY PRODUCTS PLANTS
ro
cr>
01
PLANT g
PRODUCT
A
RENDERING
PLANT
B
RENDERING
PLANT
C
MEAT-PACKING
PLANT
D
MEAT-PACKING
PLANT
E
MEAT-PACKING
PLANT (SMALL)
F
POULTRY PACKING
PLANT
G
POULTRY PACKING
PLANT
H
POULTRY PACKING
PLANT
TREATMENT
STABILIZATION
LAGOONS (2)
ANAEROBIC/
AEROBIC LAGOONS
AIR FLOTATION
(CHEMICAL ADDITION)
OXIDATION
DITCH
ANAEROBIC
AEROBIC LAGOONS
STABILIZATION
POND
EXT. AERATION
PHOS. REMOVAL
EXTENDED
AERATION
EFFLUENT
SAMPLE
RAW
FINAL
RAW
FINAL
RAW
FINAL
RAW
FINAL
RAW
FINAL
RAW
FINAL
RAW
FINAL
RAW
FINAL
BOD
mg/1
4419
12
975
85
968
348
1065
<20
1900
350-650
1550
1200
1675
400
485-900
< 10- 100
TSS
mg/1
-
272
164
825
186
693
15
1300
155-90
200
130
325
150
130-400
20-100
NH,-N
mg/1
215
19.5
77.7
85.0
17.8
17.7
23
<0.2S
195-145
1-10
35
75
65
<4
34-44
1-14
LC 50
%
6.25
50-100
20-30
10-20
30-40
90-100
37-45
ND
16
ND-71
32-41
32
25-71
ND
12-50
ND
-------
TABLE 4
DAIRY PRODUCTS PLANTS
LC 50
PLANT 6
PRODUCT
A
Cheese §
Milk Powder
B
Process
Cheese
ro c
CT> Ice-Cream
ov
D
Butter §
Milk Power
E
Butter $
Milk Powder
TREATMENT
AERATED
LAGOONS (3)
AERATED
LAGOONS (3)
EXTENDED
AERATION
EXTENDED
AERATION
OXIDATION
DITCH
EFFLUENT
SAMPLE
RAW
FINAL
RAW
FINAL
RAW
FINAL
RAW
FINAL
RAW
FINAL
BOD
mg/1
1600
550(67)
1700
400(30)
1000(6400)
325(<15)
1800(600)
900(15-218)
1340(700)
TSS
mg/1
200(500)*
130
400-1900
150
2000(5000) j.
1800**(100)
375
2500**(25-1100) .
980(650)
30
Winter
14-40
32-50
14-24
32
60
ND
32-71
i 50-100
6-71
ND
Sunun
9-12
29-53
<6
ND
<3
95-ND
35
ND-35
14-19
ND
•Values in brackets indicate results obtained under summer conditions
**Bulking sludge at these plants
-------
for BOD and total suspended solids obtained in the summer sampling survey
are indicated in brackets where there was an appreciable difference from the
results of winter survey. In general there was some difference in effluent
quality parameters throughout the year although no definite relationships could
be developed between toxicity and BOD or suspended solids.
Problems with the lagoon systems freezing over in the winter,
resulting in partially anaerobic conditions and generation of hydrogen supphide,
caused mortality to occur in these systems. In fact, even in summer conditions
the effluent from Plant A exhibited toxicity due in part to a lack of
sufficient aeration capacity in the third lagoon.
The extended aeration systems were capable of producing a non-lethal
effluent provided that the "bulking" condition at these plants could be controlled.
Overloading of the systems and a high C:N ratio have been cited as possible causes
of the bulking problem.
The oxidation ditch at Plant E was the most stable system tested .
At no time have any fish died when exposed to undiluted treated effluent. Even
during the winter season, there was never any fish mortality and other effluent
quality parameters remained consistent at less than 10 mg/1 BOD and 30 mg/1
total suspended solids.
Results of the dairy sampling program indicate that properly designed
and operated systems are capable of producing effluents exhibiting no acute
lethal toxicity.
d) Fruit and Vegetable Processing
The data presented in Table 5 were obtained in a survey carried out
during the 1975 processing season.
For all commodities with the exception of Plant E producing fruit
juices and concentrates, the screened wastewater was acutely lethal due to a
number of factors in combination such as high oxygen demand and extreme pH. For
example, effluents from plants B and F had a pH ranging from 4.6 to 5.3 whereas
plant G had a pH of 9.6 to 9.8 when tested.
Most of the treatment systems examined were capable of producing
effluents which exhibited no acute lethal toxicity to fish. However, problems
were encountered at Plant C which employs an aerated lagoon prior to discharge
to a municipal sewer. This aerated lagoon system was not well operated, being
overloaded organically and lacking in sufficient nutrients, thus resulting in
a "bulking" sludge.
e) Associated Industries
Preliminary surveys were carried out at several other plants
associated with the food industry during the winter and spring of 1975, the
results of which are summarized in Table 6.
Although the treatment efficiency of these systems was not exceptional
(i.e. BOD removal ranging from 50 to 58 percent), the treated effluents with
one exception did not exibit acute lethal toxicity.
267
-------
TABLE 5
FRUIT AND VEGETABLE PROCESSING INDUSTRY
PLANT AND
PRODUCT
A
Canned
Fruit
B
Pie Fillings,
Jams, Jellies
en Canned
00 Vegetables (corn)
D
Canned § Frozen
(corn, Broccoli, plums)
E
Fruit Juices §
Concentrates
F
Canned
Tomatoes
G
Tomatoes
Products
TREATMENT
SOIL
BIOFILTRATION
AERATED
LAGOON
AERATED
LAGOON
OXIDATION
DITCH
EXTENDED
AERATION
SOIL
BIOFILTRATION
EXTENDED
AERATION
EFFLUENT
SAMPLE
FINAL
SCREENED
FINAL
FINAL
SCREENED
FINAL
SCREENED
FINAL
SCREENED
FINAL
SCREENED
FINAL
BOD
mg/1
980-1500
1008-2880
158-296
415-864
1182-5100
1145-1685
800-1300
15-95
500-5000
80-375
500-600
S-19
TSS
mg/1
90-157
128-178
31-42
1740-3250
310-580
240-370
26-350
15-63
160-675
20-55
290-365
6-19
LC 50
ND
24-27
ND
71-100
13-17
ND
ND
ND
12.5-18
SO-ND
13.5-36
ND
-------
TABLE 6
ASSOCIATED INDUSTRIES
PLANT AND
PRODUCT
TREATMENT
EFFLUENT
SAMPLE
BOD
mg/1
TSS
mg/1
LC 50
Distillery EXTENDED
(whisky, gin, Alcohol) AERATION
RAW
FINAL
800
340
58
12
50-100
ND
PO
CT>
10
B
Brewery
EXTENDED
AERATION
RAW
FINAL
1640
812
1505
260
28-71
= 100
Starch (corn)
Products
TWO-STAGE
ACT. SLUDGE
RAW
FINAL
1475
630
2000
280
50-77
4-22
Coffee, hot choc.,
instant tea, etc.
ACT. SLUDGE
§ POLISHING POND
RAW
FINAL
1250
550
410
36
>100
ND
-------
Perhaps of most concern in this preliminary survey were the results
from Plant C, the starch products plant. The effluent, following treatment
in a two-stage activated sludge system, is acutely lethal to fish having a
96 hour LC50 value between 4 and 22 percent by volume. The exact cause for the
rather poor waste treatment efficiency and the extreme toxicity, has not yet
been determined but may relate to the by-products generated in the degradation
of corn starch process effluents.
Application of Toxicity Test Results
As noted earlier, environmental regulations for the control of
discharges of deleterious substances are developed under the Fisheries Act.
Generally, the regulations allow the discharge of quantities of certain sub-
stances based on the levels achievable by best practicable process and treat-
ment technology (BPT) for a particular industry sector. In addition, where BPT
demonstrates that an effluent quality which exhibits no acute lethal toxicity can
be achieved, then a toxicity requirement can be established as a "catch-all",
thus reducing the number of regulated parameters.
As shown in Table 3 for the Meat and Poultry Products industry
several plants employed treatment systems which discharged effluents having no
fish mortality. Since these systems were identified as exemplifying BPT for the
industry, an objective is proposed wherein a minimum of 50 percent of the test
fish must survive for 96 hours in the undiluted effluent. A similar objective
has been proposed for the Potato Processing industry, based mainly on the
results of plant C.
Given that the possibility exists for some fish to die of natural
causes in the testing procedures, a requirement of 50 percent survival is
considered consistent with effluent quality obtained from current BPT systems.
Also, environmental consultants can not guarantee that biological treatment
systems will always discharge an effluent in which no fish mortality will occur.
The regulatory test procedure used to determine whether a plant
meets the objective will be the 96 hour continuous flow test incorporating the
conditions outlined in a previous section. A more simplified static bioassay
procedure may be used by the plants to monitor the toxicity of their effluents
at regular intervals.
Regulations incorporating these toxicity objectives as guidelines
have now been drafted for the Potato Processing Industry and Meat and Poultry
Products Plants. Effectively, these regulations and guidelines will come into
force by mid 1976. To follow are regulations for the Dairy Products industry
in 1977 and the Fruit and Vegetable Processing industry in early 1978.
SUMMARY
The testing of food processing effluents with respect to their toxicity
to fish is relatively new and not really understood in the industry. Consequently
many plants managers have shown some concern when approached on this subject.
270
-------
The objective of this paper has been to provide some insight into the aspects
of fish toxicity as it relates to the food processing effluents.
In summary then, the following points should be noted:
1) Government agencies will continue to monitor the multiple
effects of industrial discharges on the aquatic environment.
2) The major causes of fish toxicity from food processing effluents
are organic material exerting a high oxygen demand, cleaning
and disinfecting agents, and by-products such as ammonia and
hydrogen sulphide generated in the degradation of certain
waste constituents.
3) Properly designed and well-operated biological treatment systems
are capable of producing an effluent quality where acute lethal
toxicity is not observed.
Acknowledgements
The author wishes to acknowledge the staff in the five Regional offices
of the Environmental Protection Service who have conducted the studies from which
much of the data was obtained for this presentation. Special thanks are also
due to Ed Pessah, EPS Atlantic Region, not only for reviewing this manuscript
but for his advice in the past on EPS toxicity assessment programs.
271
-------
References
1. Pessah, E. "The Assessment of Industrial Wastes Using Fish
Bioassays", Proceedings of Annual Conference, (April 1974)
Pollution Control Association of Ontario, Toronto, Ontario,
pp. 65-81.
2. "Standard Methods for the Examination of Water and Wastewater",
13th edition, APHA, AWWA, WPCF, 1971.
3. Sprague, J.B., "The ABC's of pollutant Bioassay using Fish",
Biological Methods for the Assessment of Water Quality, ASTM
STP 528, American Society for Testing and Materials, 1973,
pp. 6-30.
4. Litchfield, J.T. and Wilcoxon, F., "A Simplified Method of
Evaluating Dose-effect Experiments", J. Pharmacology and
Experimental Therapeutics, Vol. 96, 1949, pp. 99-113.
272
-------
CHARACTERIZATION AND POTENTIAL METHODS FOR REDUCING WASTEWATER
D. L. Grothman*, L. J. P. Scully**, P. M. Berthouex***, D. 0. Dencker****
INTRODUCTION
The purpose of this project was to determine the wasteload from different hog
slaughtering processes, to characterize the waste streams, and to make changes
in processes and equipment in order to achieve"one or more of the following
objectives without reducing hog kill capacity or the quality of the carcasses
produced:
1. Reduce water use
2. Reduce and/or prevent product loss to sewers, and
3. Reuse wastewater where possible
It has long been apparent to those in the meat packing industry that there is
great potential for reduction of wasteload to the sewers through equipment,
process, and procedural changes. However, most emphasis has been placed
towards efficient end-of-pipe wastewater treatment.
Many in-plant waste reduction measures can be designed into new plants, but
in implementation older facilities present numberous problems. Also, new
equipment should be designed with water conservation and waste generation in
mind.
DESCRIPTION OF PLANTS
The majority of this study was done at the Madison plant of Oscar Mayer & Co.
with a few processes studied separately at the Davenport, Iowa, or Beardstown,
Illinois, plants.
The Madison plant includes a modern meat processing plant, spice processing,
and plastic film production areas as well as a hog kill rated at 1,000 head/
hour and an 80 head/hour beef kill. The facility includes edible and inedible
rendering. The plant's power house supplies all of its steam and about 80%
of its electrical power. Seventy-five percent of the water used is from the
*Project Engineer; Oscar Mayer & Co., Madison, Wisconsin.
**Peat, Warwick & Mitchell; Washington, D.C.; formerly Research Assistant;
Water Resources Center; University of Wisconsin, Madison, Wisconsin.
***Associate Professor; Department of Civil and Environmental Engineering;
University of Wisconsin, Madison, Wisconsin.
****General Sanitary Engineer; Oscar Mayer & Co., Madison, Wisconsin.
273
-------
Company's own wells. Wastewater at this plant is segregated. Clear water
drainage from roofs and parking lots, water used for cooling plastic extruders,
etc., is discharged to the city storm sewers. Sanitary sewage discharges
directly to the Madison Sewerage System. Manure water (wastewater from the
stockyards, stomach dumper, dehairing machine, and scald tank) is screened
to remove the large solids, settled to remove grit, and pumped to an Oscar
Mayer & Co. operated wastewater treatment site for biological treatment. The
plant greasewater system collects the wastewater from all of the floor drains
and processes throughout the entire plant. Greasewater is separately treated
by settling and dissolved air flotation, and is then combined with the manure
water for biological treatment (two stage trickling filter).
The effluent from biological treatment is discharged to a Madison Metropolitan
Sewerage District interceptor sewer. Sludge is dewatered by vacuum filter.
Madison also has a blood collection system on the kill floor which feeds a
blood recovery system. This is a part of a dual drainage system. During
production the blood drains are open for recovery of blood. When clean-up
starts, the blood drains are plugged and the greasewater drains are opened.
The Davenport plant includes a meat processing plant, complete edible and
inedible rendering, and a hog kill capable of processing 750 head/hour.
Davenport has its own power plant to supply steam and electricity. Wells on
the property supply 80% of the water used in the plant. Wastewater is
segregated in Davenport as in Madison except for a few roof drains which flow
into the greasewater drainage system. The manure water drainage system
collects water from the stockyards, scald tank, dehairing machine and stomach
dumper and sends it across the rotostrainer which removes the large solids.
From the rotostrainers the wastewater discharges to a city sewer. The grease-
water drainage from floor drains throughout the plant flows across two roto-
strainers to remove the solids and then through a dissolved air flotation tank
before discharging to a city sewer.
Beardstown was built in 1967 as a hog slaughtering and butchering facility with
a capacity of 750 hogs/hour. Since then, the plant has been increased in size
by 40% to provide additional room for ham canning and other operations. The
plant also has edible and inedible rendering. All water comes from wells located
on the property. All plant wastewater except clear water from roof and yard
drains is treated by screening and flotation before being pumped into anaerobic,
intermediate aerobic and final aerobic lagoons in series.
SLAUGHTERING PROCESSES
The process described is used in the Madison plant.
The hogs are anesthetized in a conveyorized carbon dioxide chamber. Unconscious
hogs are discharged onto an inclined steel-slat conveyor which has a stainless
steel collection trough along one side. The hogs are arranged on the conveyor
so their necks are over this trough and their carotid arteries and jugular veins
are severed with a knife. The blood is collected and pumped to a blood process-
ing area where it is coagulated and dried for use as an animal feed supplement.
274
-------
The carcasses drop from the bleed conveyor into a long shallow scald tank which
is filled with 140 F water to which lime has been added. The carcasses are
moved through the tank by mechanical "dunkers" which submerge the carcass and
push them through the tank. Total scald time is about five minutes. The
carcasses move out of the scald tank into a "Boss" dehairing machine. Hot
water showers lubricate the carcasses to prevent the dehairing paddles from
tearing the hide and flush away removed hair and toenails. The dehaired carcass
is gambreled a worker slits the hind legs and inserts a steel bar, the
gambrel, behind the achilles tendons. The gambrel is hung on the steel trolley
which is pulled by a chain along an elevated rail to transport the hanging hog
through the rest of the process.
The carcasses are dipped into a tank containing 290 F rosin which covers all of
the carcass except the last eight inches of the hind legs. The carcass is
showered lightly to harden the rosin- and the rosin is then stripped off by
machine and by hand. This removes hair not removed in the dehairing machine.
The carcass is conveyed through a singer where very fine hairs are burned off.
A set of rubber paddles strip off burned residues while the carcass is lubricated
with a water spray in the rail polisher. Tissue is cut from between the toes
and the carcasses are manually scraped with knives to remove or loosen remain-
ing soil. A final carcass shower rinses the carcass with large quantities of
water. Once the carcass has passed this point, any remaining soil or hair must
be removed by excision.
Disassembly of the carcass starts with removal of eyelids along with the surround-
ing skin and fat. The neck is almost completely severed so the head is left
hanging by a strip of skin along the jaw. USDA Insepctors slash open the salivary
glands to check for abscesses and they examine the outside of the carcass for dirt,
bruises, or signs of disease. The sternum (brisket) is split, the aitch bone is
split, and the bladder and uterus (pizzles and piggy bags) are removed. The anus
(bung) is cut around and freed from the surrounding tissue. Viscera (heart,
lungs, liver, stomach, pancreas and intestines) are removed and placed in an
examination pan which travels by conveyor alongside the carcass. The carcass
is split into halves by sawing through the backbone. Pieces of the dorsal
abdominal aorta are trimmed out and the stick wound in the neck is slashed open.
The urethra (male pigs) is removed. The membrane over the kidneys is slashed
open and the kidneys are slid out for examination. A final examination intern-
ally and externally for bruises, cuts, dirt, etc. is made and any unsatisfactory
parts are trimmed off. The kidneys are removed either to be washed, boxed and
frozen for animal feed or to be sent to inedible rendering. The head is washed,
removed and butchered (muscles from the cheeks, scalp and tongue are saved; the
head is split and the pituitary gland and hypothalmus are removed for pharma-
ceutical purposes; and the skull and jaw are ground up and sent to inedible
rendering). Leaf fat which lines the interior dorsal area of the carcass is
pulled out and sent to lard rendering. Loose fat and tissue is scraped from
the interior of the carcass. The stick wound on the neck is washed. Interior
carcass fat is trimmed a little more, then the carcass is weighed and sent to
be chilled overnight before being butchered.
The entire kill process is graphically shown on Figure 1.
275
-------
Figure 1. Madison Hog Kill Process Flow Sheet.
-------
The removed viscera is checked by USDA Inspectors for signs of disease. Diseased
viscera is sent to inedible rendering. Acceptable viscera is separated for
further processing:
1. The heart is slashed open, to remove the blood clots, chilled, and
held for use in processed meats.
2. The lungs are frozen for animal food or rendered in inedible
rendering.
3. The stomach is slit open so contents can be flushed out, and
tumble washed. The mucosa is removed for pharmaceutical process-
ing and the remainder of the stomach is washed, scalded and
frozen for animal food.
4. The caul fat (omentum) is removed, washed and sent to lard
rendering.
5. The liver is trimmed and chilled for use in processed meats.
6. The gall bladder goes to inedible rendering.
7. The large and small intestines, condemned viscera and pizzle
and piggy bags are sluiced to the hasher-washer where they
are slashed open and washed by the sluice water. The washed
solids go to inedible rendering.
SAMPLE COLLECTION AND ANALYSIS
Quantifying the water use and discharge of pollutants from the kill floor is
difficult in any plant because the water supply and the drainage pattern is
never simple and it is usually impossible to isolate each process from its
neighbors. The methods used to overcome this problem in characterizing the
Madison hog kill operation will be summarized here briefly. A variety of
methods including water meters, a bucket and stop watch, and chemical tracer
dilutions were used to determine the flows into and/or out of the various
slaughtering processes and production areas. Figure 2 shows the drainage
system from the kill floor. The sampling locations were chosen so far as
possible to isolate areas which were expected to have different characteristics.
Several different methods were also needed to collect representative samples
because the sampling point each had different physical characteristics. In
some of the sewers, the flow was intermittent or periodic. In most of the
sewers there was no way to continuously monitor flow rate so as to properly
make a flow proportioned composite sample. Because of this, a great reliance
was placed on tracking the pattern of water use over time and making arith-
metic composite based on numerous grab samples which were collected. Table 1
shows the sampling method and the flow measurement method used at each of
the sampling locations during the production shift. Similar but more
complicated problems were encountered during the clean-up shift when there
is a great deal of intermittent operation of valves and hoses. An example
of how this was handled is given later.
Samples were collected from all the sampling locations on several different
days and collected in order to characterize both the production and the clean-
up shift. Laboratory measurements included BOD, COD, organic carbon, grease,
several forms of solids, and total Kjeldahl nitrogen. The procedures used
were those specified in the EPA laboratory manual.
277
-------
Figure 2. Madison Kill Floor Layout,and Drainage Plan,
-------
Sampling Location
1. Bleed Floor Drain
2. Bleed Area Floor Drain
3. Bleed Conveyor Wash
4. Scald Tank
5. Dehair Floor Drain
6. Rail Polisher
7. Carcass Shower
8. Hasher-Washer Drain
9a. Stomach Washer
9b. Neck Washer
9c. Head Washer
10. "660" Grease Drain
11. Center Grease Drain
12. "330" Grease Drain
Table 1. SAMPLING AND FLOW MEASUREMENT
Sampling Method
Manual Short-Term Composite
Manual Short-Term Composite
Automated Graba
Manual Short-Term Composite
Q
Automated Grab
Automated Grab
Automated Grab
Manual Short-Term Composite
Manual Short-Term Composite
Manual Short-Term Composite
Manual Short-Term Composite
Automated Grab
Automated Grab
Automated Grab
METHODS
Flow Measurement
Time Known Volume
Time Known Volume
Time Known Volume
No Flowd
Totalizing Flow Meters in
Water Supply Lines
Totalizing Flow Meter
Totalizing Flow Meter
Lithium Dilution
Lithium Dilution
Time Known Volume
Time Known Volume
Lithium Dilution
Lithium Dilution
Lithium Dilution
Isco Automated Sampler.
Five to six grabs composited over 5 to 10 minutes.
•*
"One-inch diameter rubber bladder valve with compressed air supply controlled by
a solenoid.
Samples were from the scald tank which is dumped during clean-up.
279
-------
An enormous number of samples had to be collected, processed through the lab-
oratory, and the data from all these analyses had to be converted to inform-
ation. A system of checks and cross checks on the inventory of samples was
designed so that the project manager, the data key punch operators, and the
laboratory technicians could all see clearly what schedule was intended and
how the data production was progressing in accordance with that schedule.
This data management scheme is discussed in the final report of the project.
Table 2 shows the initial characterization of the production shift for the
Madison plant with flow and quantities of all pollutants given in terms of
thousand pounds of live weight killed. The summarizing tables were prepared
from tables for individual pollutants and for flow similar to Table 3. A
typical shift was defined from long term production records. The pounds per
shift and flow per shift are important numbers for the individual slaughtering
plant. In this paper we will express results more commonly in terms of mass
per thousand mass units live weight killed because this, hopefully, extra-
polates more reasonably to other slaughtering operations.
WASTEWATER CHARACTERIZATION
Table 2 and Table 3 are based on the reduction of a great number of samples.
Having great mass of raw data summarized in this form enables one to identify
quickly those areas in the kill floor which are the greatest contributors of
flow and of various pollutants. For example, Table 2 quickly shows that the
dehairing operation contributes roughly 1/4 of the flow from the kill floor;
the 660 grease drain is a large contributor of flow but is really not a very
substantial portion of the pollutant load, the stomach washer uses large
amounts of water and has a very heavy pollutant load, and the hasher-washer
is the other most important contributor. This kind of intiial characterization
quickly draws one's attention to the areas which hold the greatest potential
for conservation. For example, a ten percent reduction in the solids in the
hasher-washer drain is the equivalent in pollutional load reduction to elim-
inating the other solids load in any of the grease drains. It shows that,
while small reductions may be possible in certain areas such as the rosin
stripper or the rail polisher, these reductions cannot be important in terms
of the total load. Small changes should be made whenever possible if the
cost and effort is not too great. Clearly, a major investment of time and
money to reduce the contribution from the hasher-washer will bring substan-
tial reward.
The same kind of characterization and interpretation was done for the clean-
up shift. Because the flow rates during clean-up are so highly variable, the
pattern in time wds determined by a continuous tracer dilution technique and
flow profile such as shown in Figure 3 were developed for each sampling point.
At the same time grab samples were collected on which measurements of pollutant
concentration were made. Figure 4 shows the profile of BOD during the clean-
up shift which corresponds to the flow in Figure 3. This flow and concentra-
tion data are used then, to estimate the mass of BOD which was discharged past
that patticular sampling point during the clean-up shift. At every sample
point there was always an initial peak in concentration just as clean-up began.
In the examples shown, we see the concentration of BOD dropping from initial
values of one to two thousand to concentrations of one to two hundred within
the first hour. At some sampling points the concentration dropped even more
drastically than this because the area being cleaned was rather well defined
280 .
-------
Table 2. INITIAL WASTEWATER FLOW AND POLLUTANT LOAD CHARACTERIZATION OF THE PRODUCTION SHIFT.
EXPRESSED PER THOUSAND POUNDS OF LIVE WEIGHT KILL.
ro
00
Sample point
Bleed area floor drain
Bleed conveyor blood drain
Bleed conveyor wash drain
Dehair floor drain
+ scald tank
Rosin stripper
Rail polisher drain
"660" grease drain
Carcass shower
Center grease drain
"330" grease drain
Head washer
Stomach washer drain
Hasher washer drain
Neck washer
Flow
_
-
1.9
70.3
8.6
8.9
21.9
(20.1)
14.3
16.4
(4.5)
48.5
53.2
(8.6)
Total
solids
_
-
0.036
1.225
0.054
0.069
0.136
(0.092)
0.088
0.430
0.981
4.340
(1.020)
Total
volatile
solids
0
0
0
0
0
(0
0
0
0
3
(0
_
-
.030
.875
.028
.040
.070
.039)
.077
.175
.817
.730
.980)
Suspended
solids
_
-
0.004
0.649
0.005
0.009
0.021
(0.004)
0.033
0.086
0.671
2.346
(0.472)
Total
suspended
solids
_
-
0.003
0.580
0.004
0.008
0.018
(0.004)
0.029
0.080
0.606
1.880
(0.440)
Grease
_
-
0.002
0.180
0
0.005
0.012
0.008
0.074
1.454
3.590
(1.580)
Total
Kjeldahl
nitrogen
_
-
0.004
0.131
0.002
0.005
0.019
(0.008)
0.045
0.056
0.041
0.230
(0.020)
BOD
_
-
0.014
.536
0.009
0.020
0.037
(0.010)
0.025
0.169
0.537
2.700
(0.360)
COD
-
-
0.036
1.308
0.023
0.045
0.080
(0.030)
0.337
0.308
1.300
6.800
(0.900)
Total gallons/1000 Ibs LWK
Total pounds/1000 Ibs LWKa
244.0
7.36
5.84
3.828
3.208
5.325
0.533
4.047
Live weight kill
^'Carcass shower is included in 660 grease drain totals.
(^Head washer is included in 330 grease drain totals.
^3'Neck washer is included in hasher-washer drain totals.
10.237
-------
Table 3. COD SUMMARY OF THE INITIAL CHARACTERIZATION OF THE PRODUCTION SHIFT
ro
CO
ro
Sample point
Bleed area floor drain
Bleed conveyor blood drain
Bleed conveyor wash drain
Dehair floor drain
Scale tank
Rosin stripper
Rail polisher drain
"660" grease drain
Carcass shower
Center grease drain
"330" grease drain
(2)
Head washer
Stomach washer
Hasher washer drain
Neck washer^
Total
mg/1
_
-
2360
2230
320
610
440
(200)
2830
2260
-
3220
15340
(13580)
Pounds/
shift
_
.
52.5
1882.
34.0
64.9
114.5
(48.2)
485.0
444.3
-
1871.0
9780.
(1402.)
14728.2
Kilograms/
shift
_
-
23.8
854.
15.3
29.4
51.9
(21.8)
219.8
201.5
-
849.
4436.
(635.)
6680.7
Pounds
1000 Ibs LWKa
_
-
0.036
1.308
0.023
0.045
0.080
(0.03)
0.337
0.308
-
1.300
6.800
(0.90)
10.237
Percent
of load
.
-
0.4
12.8
0.2
0.4
0.8
(0.3)
3.3
3.0
-
12.7
66.4
(9.5)
100.
aLtve weight kill
*• 'Carcass shower is included in 660 grease drain totals.
'Z'Head washer is included in 330 grease drain totals.
^ 'Neck washer is included in hasher-washer drain totals.
-------
2
Q.
70
z
— 60
O 50
U.
~ 4O
UJ
30
UJ 20
H
tn
< ,0
SAMPLE POINT! 330 GREASE DRAIN
I
I
6O 120 180 240 300 36O
TIME OF CLEAN-UP SHIFTS, minutes
Figure 3. Clean-up Shift Flow Profile
283
-------
2000
1000
500
100
50
Q
O
ffi 10
SAMPLE POINT : 330 GREASE DRAIN
0 60 120 180 240 300 360
TIME OF CLEAN-UP SHIFT, minutes
Figure 4. Clean-up Shift BOD Concentration Profile
.284
-------
and the clean-up task was very specific. In the example shown in Figures 3
and 4, the 330 grease drain, the area being cleaned represents several kill
processes and as the clean-up crew moves from one piece of machinery to
another, additional peaks are seen. Each sampling point had a characteristic
clean-up shift profile and these profiles were determined and used to make
estimates of the mass and the flow discharge during the clean-up shift which
are presented as Table 4.
The total amount of flow and pollutant discharge during a day, the production
shift plus the clean-up shift, are given in Table 5 and 6.
A two way comparison of this data is both informative and important. One
comparison is the discharge of one process area or one sampling point with
another. The second is a comparison of the discharge during the production
shift with discharge during the clean-up shift. Figures 5 and 6 graphically
make this comparison for flow and BOD for the Madison plant. The totals for
the entire hog kill are shown at the right hand edge of these figures.
The water use is divided approximately 1/3 to the clean-up shift and 2/3 to the
production shift. The BOD load is much more heavily predominated by the
production shift. About 95% of the combined BOD load originates during the
production shift and only about 5% is discharged during the clean-up shift.
Similar patterns are found for all of the other pollutants which were measured.
CHANGES
These results indicate that process modifications during the production shift
are an important target if water use and pollutant discharges are to be reduced.
The general strategies for conserving water during the production shift is to
identify wasteful sprays and to reduce or eliminate the use of water for trans-
port of material which could be transported in a dry form. The strategy for
reducing pollutant mass during the production shift is generally to be sure
that blood cannot enter the sewer and to be sure that meat scraps are picked
up and not allowed to be flushed down the drain. Water use during clean-up
can often be reduced substantially by insisting that the clean-up personnel
do not leave hoses running unattended, to be sure that they do not unnecess-
arily turn on sprays or forget to turn off process waters. Pollutional loadings
can be reduced by being sure that blood and meat scraps are picked up dry.
With the results of the initial characterization for guidance and with these
general strategies in mind, a long list of potential process modifications
was generated. Many of these changes could not be implemented before the
project ended but many worthwhile process changes were actually installed and
evaluated in the plant. In some cases the process changes were impossible in
the Madison plant, because of physical limitations, space constraints or
production constraints. Changes were made in other Oscar Mayer plants and some
differences in processes in other plants were used as a basis for estimating
the potential value of a change. The kinds of changes installed and evaluated
are discussed individually in a later section of the paper. To put these
individual changes into perspective, Tables 7 and 8 presented to summarize the
reductions which could be accomplished at the Madison plant.
285
-------
ro
00
Table 5. INITIAL WASTEWATER FLOW AND POLLUTANT LOAD CHARACTERIZATION OF THE COMBINED PRODUCTION AND
CLEAN-UP SHIFTS. EXPRESSED AS PERCENT.
Sample point Flow
Bleed area floor g
drain
Bleed conveyor Q
blood drain
Bleed conveyor Q
wash drain
Dehair floor drain 37
Scald tank 1
Rosin stripper 2
Rail polisher drain 2
"660" grease drain/ -j\ 8
Carcass shower
Center grease drain 3
"330" grease drain 7
Head washer W
Stomach washer drain 14
Hasher washer drain 21
Neck washer (3'
Total % 100
Gal/shift 521480
Liters/shift 1973318
Total
Total volatile
solids solids
5 0.3 0.4
2 0.1 0.2
9 0.5 0.6
6 16.2 18.0
.5 -
.4 0.5 0.4
.5 0.7 0.6
4 2.4 1.6
_ _
9 0.8 1.2
7 27.4 4.2
_ _
4 9.4 13.1
4 41.7 59.7
.
100 100
Suspended
solids
0
0
0
21
0
0
1
(0
0
4
15
55
(0
100
.2
.1
.1
.0
_
.1
.2
.0
.1)
.8
.7
—
.9
.9
.1)
Volatile
suspended
solids
0.2
0.1
0.1
21.1
_
0.1
0.2
0.8
-
0.9
3.8
-
17.7
54.8
-
100
Grease
0.
0.
0.
3.
(0.
0.
0.
-
0.
2.
-
27.
66.
-
100
2
02
03
7
17)
1
4
1
2
0
3
Total
Kjeldahl
nitrogen
0.7
0.2
0.9
26.1
-
0.4
0.8
3.7
-
8.0
10.1
-
7.3
41.8
-
100
BOD
0.
0.
0.
15.
_
0.
0.
1.
-
0.
5.
-
12.
63.
-
100
4
1
4
5
2
5
2
6
0
7
4
COD
0.
0.
0.
15.
_
0.
0.
1.
-
3.
3.
-
12.
63.
-
100
3
1
4
6
2
4
2
1
5
1
1
Gal/1000 Ib LWKa 362.7
Liters/1000 kg LWKa 3025.7
Lb/shift
Kg/shift
Lb/1000 Ib LWKa
Kg/1000 kg LWK
15345 9084
6962.3 4117.9
10.674 6.314
10.674 6.314
6125
2778
4
4
.7
.3
.259
.259
4956
2247.9
3.451
3.451
7802
3537.
5.
5.
8
427
427
808
366.7
0.559
0.559
6133.
2782.
4.
4.
6
5
266
266
15548.
7052.
10.
10.
7
3
833
833
aLive weight kill
' 'Carcass shower is .included in 660 grease drain totals.
'2'Head washer is included in 330 grease drain totals.
^3'Neck washer is included in hasher-washer drain totals.
-------
Table 6. INITIAL WASTEWATER FLOW AND POLLUTANT LOAD CHARACTERIZATION OF THE COMBINED PRODUCTION AND CLEAN-
UP SHIFTS. EXPRESSED PER THOUSAND POUNDS OF LIVE WEIGHT KILL.
Sample point
Bleed area floor drain
Bleed conveyor blood drain
Bleed conveyor wash drain
Dehair floor drain
Scald tank
Rosin stripper
Rail polisher drain
"660" grease drain
Carcass shower ^ '
Center grease drain
"330" grease drain
1^3 Head washer ''/
23 Stomach washer drain
Hasher washer drain
Neck washer
Flow
1.9
.8
3.4
136.3
-
8.6
8.9
30.6
( 20.1)
14.3
27.8
-
52.1
78.0
-
Total
solids
.031
.015
.055
1.729
-
.054
.069
.260
( .092)
.088
2.928
-
1.000
4.445
-
Total
volatile
solids
.023
.011
.038
1.135
-
.028
.040
.102
( .039)
.077
.264
-
.828
3.768
-
Suspended
solids
.009
.004
.006
.896
-
.005
.009
.040
( .004)
.033
.201
-
.677
2.379
-
Volatile
suspended
solids
.008
.003
.004
.728
-
.004
.008
.029
( .004)
.029
.132
-
.612
1.894
-
Grease
.009
.001
.003
.202
-
-
.005
.022
( - )
.008
.120
-
1.461
3.596
-
Total
Kjeldahl
nitrogen
.004
.001
.005
.147
-
.002
.005
.021
( .008)
.045
.057
-
.041
.231
-
BOD
.016
.005
.015
.661
-
.009
.020
.051
( .010)
.025
.215
-
.542
2.707
-
COD
.027
.013
.042
1.685
-
.023
.045
.125
( .030)
.377
.360
-
1.315
6.821
-
Total gallons/1000 Ib LWK'
Total lbs/1000 Ib LWKa
Live weight kill
362.7
10.674
6.314
4.259
3.451
5.427
.559
4.266
10.833
(1
Carcass shower is included in 660 grease drain totals.
' Head washer is included in 330 grease drain totals.
' 'Neck washer is included in hasher-washer drain totals.
-------
ro
00
oo
BLEED AREA
PRODUCTION FLOW
(•HMI/IOOOIkLWK
CLEAN UP FLOW
••Itoni/IOOOIbLWK
0.0
I . t
TOTAL FLOW
f«ll«m'fOOOI»LWK 1.9
0.0
.8
1.9
1.5
SCALD
OEHAIRINO
CARCASS
PREPARATION
|
I
i
.
8C
HI
i
I
I
MAIN KILL FLOOR AREA
I
i
|
|
I
I,
!
M
;
&
. 1
I
K
I
1
1
s
s
.• 3.4
TO.3
M.O
136.3
8.6 a*
0.0 0.0
as •.»
21.9 14.3 M.4
•.T 0 11.4
30.6 14.3 27 .•
453 532
3.6 24,8
52.1 78.0
PRODUCTION FLOW 244.0
CLEAN UP FLOW 118.7
COMBINED FLOW 362.7
gallon* XIOOOIbLWK
Figure 5. Wastewater Flow Mass Balance for the Madison Production and Clean-up Shifts.
-------
BLEED AREA
SCALO
PRODUCTION BOD LOAD
Ib/IOOO Ib LWK
CLEAN UP BOD LOAD
Ib/IOOOIbLWK 0016
0.009 0.001
TOTAL BOD LOAD
Ib/IOOO IbLWK
0.016 0.005 0.015
DE HAIRING
0.536
0.125
0.661
CARCASS
PREPARATION
•
ui
m
s
IU
£
3
I
|
MAIN' KILL FLOOR AREA
i
e
i
i
3
a
5
Ul
s
Ul
s
u
X
s
x,
*
o
•«
Ul
x
X
i
te
X
I
tfl
•c
X
«
Ul
a..
I
e
K
O.OO9 OX)20
O.OOO O.OOI
0.009 0.020
0.037 0.025 0.169
0.014 O.OOO 0.046
0.537 Z.700
0.005 0.007
0.05, 0.025 0.2,5
0.542 2.707
CLEAN-UP BOO LOAD
.Zl» Ib/IOOOIbLWK
COMBINED BOO LOAD
4 266lb/OOO IbLWK
rvj
oo
<£>
Figure 6. BOD Mass Balance for the Madison Plant Production and Clean-up Shifts,
-------
ro
UD
o
Table 7. FLOW AND/OR POLLUTION REDUCTION DUE TO CHANGES EXPRESSED IN GALLONS AND POUNDS
Problem Area Water BOD COD TS TSS Grease TKN Solids
Bleed Trough 5 Lbs.
Bleed Trough Clean-Up 5 Lbs.
Bleed Conveyor Sprays 2,670
Hair Chute - Davenport 110,000 506 1302 1366 704 149 107
Rail Polisher Shut-Off
Beardstown 2,400
Final Carcass Shower
Madison 8,753
Beardstown 21,330
Eyelids on Floor 12 Lbs.
Brisket Splitting „ "7 170
_ Vacuum V
Carcass Splitting Clean-Up J
Viscera Pan Wash Sprays 39,427
Viscera Pan Wash Solenoid Valves 29,825
Hasher-Washer Blade Removal
Head Washer 3,260
Neck Washer 3,034 257
Chitterling Washer
Beardstown 51,850
903 790 148
60 103
2948 7505 4180 2906 3275 138
1005 381 1705
Note: All changes were made in Madison except where noted.
-------
Table 8. COST OF CHANGES AND SAVINGS RESULTING FROM CHANGES
Problem Area Cost of Change Annual Savings
ro
iO
Bleed Trough $ 0.00
Bleed Trough Clean-Up $ 3.00
Bleed Conveyor Sprays $ 0.00
Hair Chute - Davenport $22,000
Rail Polisher Shut Off - Beardstown $ 255
Final Carcass Shower - Madison $ 184
Beardstown $ 8S
Eyelids on Floor $ 86
Brisket Splitting ""}
Bone Dust - Carcass Splitting V $ 2,377
Viscera Pan Wash Sprays ,/
Viscera Pan Wash Solenoid Valves $ 1,285
Hasher-Washer Blade Removal $ 275
Head Washer $ 0.00
Neck Washer $17,000
Chitterling Washer - Beardstown $ 78
$ 40
$ 40
$ 260
$19,406
$ 624
$ 853
$ 2,080
$ 4,078
$ 2,907
$42,697 Pollution Reduction
$96,244 Increased Byproduct Recovery
$ 831
$30,000 Reduced Labor
$ 4,858 Reduced Pollution & Water
Consumption
$ 5,070
Note: All changes were made in Madison except where noted.
-------
It is estimated that changes actually installed plus the use of processes
which exist in other Oscar Mayer plants would yield a 41% reduction in flow-
This is slightly more than 200,000 gallons per day, and at 63% reduction in
BOD and suspended solids, this represents nearly 4,000 pounds of BOD and
4,000 pounds of suspended solids per day. Similar reduction would be
accomplished for nitrogen, grease, and the other pollutants measured.
However, other factors such as physical layout, cost and product quality
considerations preclude complete adoption at Madison and other older plants.
The savings in water used and wastewater discharged to treatment can be
translated into very large monetary savings. The cost savings show in Table
8 has been estimated using the value of $0.15 per thousand gallons for cold
potable water input to the process, as sewer service charge with a volume
component of $0.24 per thousand gallons, plus a BOD component of $0.0319 per
pound and a suspended solids component of $0.0264 per pound. The work year
was taken as 250 working days. These unit costs represent the water and
wastewater charges at the Madison plant. The wastewater charges are similar
to those paid to the City of Madison for accepting pretreated industrial
wastewater. It has not been possible to estimate the actual impact of these
changes on the Oscar Mayer pretreatment plant,nor has it been possible to
estimate the changes in operating cost of that pretreatment plant due to the
reduced waste loads. It is felt that these charges give a reasonable basis
for discussion and provides a realistic method of computing savings.
PROCESS CHANGES
Problem: Blood clots build up in the bleeding trough. If the collected mass
becomes heavy enough, it breaks free and drains into the blood
recovery tank too fast for the system to handle it. Some of this
blood spills onto the floor and gets into the greasewater sewer
system. The amount of blood lost this way was not large, only
about 25 pounds/day but this is 5 pounds BODj..
Solution: The sticker now places the front leg of ervery 30th or 40th hog so
that it drags along the trough. This breaks the clots free before
they grow to an unmanageable size.
Problem: After the last hog of the day was stuck, the clean-up man would
use water sprays to sluice most of the blood left in the trough to
the blood recovery system.
Solution: A rubber squeegee, made to fit the bleeding trough, is now used to
dry clean the bleeding trough and recover the blood. After
squeegeeing the trough the drains are changed so that water goes
to the greasewater drainage system. Dry cleaning has reduced the
blood loss by approximately 25 pounds/day and the additional blood
recovered is easier to process because it is drier.
292
-------
Problem: Cold water, 26,670 gallons per average production shift, sprayed
continuously onto the slotted end of the bleeding conveyor to wash
off blood.
Solution: The sprays were eliminated during production, they are used during
clean-up. The continuous washing did not remove all of the blood
so clean-up was always needed anyway. Eliminating the sprays did
not clean-up more difficult.
Problem:
Solution:
Problem:
Solution:
Problem:
Solution:
Hog hair and toenails removed in the dehairing machine are dropped
into the plant manure water drainage system and flushed to the
sewage treatment plant where the solids are removed on a shaker
screen. Because of the bulk of the hair and the fact that it
sometimes becomes matted, 110,000 gallons/day of potable water are
used to flush the hair through the system.
No changes have been made at the Madison plant, however, the two
methods of handling the hair used in the Davenport and Beardstown
plants are being investigated. In Davenport, hair discharges from
the dehairing machine by a chute into a truck body located three
floors below. This completely eliminates the need for transport water.
In Beardstown, recycled water from the grease flotation tank is used
to transport the hair to the sewage treatment plant. This eliminates
the need for potable water for transporting hair.
Clean-up personnel leave the rail polisher sprays on during clean-up.
The water serves no useful purpose in clean-up and 1,640 gallons of
water/hour are wasted.
In Beardstown, an automatic switch turns off the water after the last
hog has passed through the polisher. The switch consists of a 36
inch push bar which is moved by the hog trolley moving along the kill
rail, to activate a micro-switch which controls a solenoid valve on
the water supply line. An emergency bypass valve for use in case
of malfunction of the solenoid valve is located out of convenient
reach of the clean-up man.
The final carcass shower uses too much water.
at start of study was 60 GPM.
Consumption measured
Different sets of sprays were tried in Madison and in Beardstown. In
Madison, four nozzles were installed to spray across the hams and
allow the water to run down the sides of the carcass to sluice off
loose material. Also, a line of six Veejet sprays were placed along
one side of the shower cabinet to scour off small rosin flakes which
are imbedded in the back. These sprays reduced the water used from
60 GPM to 46 GPM. Further work is needed on these sprays since they
produce a fine mist which drifts onto people working nearby. In
Beardstown, two nozzles spray the hams and two whirljet nozzles
(one on each side of the cabinet) spray the sides of the carcass.
This spray system uses 15 GPM of water.
293
-------
Problem: Eyelids and scraps of skin trimmed from the carcasses are dropped
onto the floor. Despite periodic dry pick-up, much of the material
was washed down the drain. Because of labor standards for the
job, it was not possible to have the trimmers put the scraps into
a barrel.
Solution: A combination screen and bridge of stainless steel was built across
the drain to keep the scrap out of the drain so that it could be
picked up for disposal in inedible rendering.
Problem: When the brisket is split, large quantities of blood serum and
blood clots drop out of the chest cavity onto the floor. A nearby
blood drain carries the blood serum to the blood recovery system.
The blood clots get swept to the hasher-washer drain during periodic
mid-shift clean-ups. In the hasher-washer, the clots break up and
flow out with the wastewater.
Solution: An industrial wet/dry vacuum cleaner was purchased to use during
mid-shift and daily clean-up to get blood and meat scraps off the
floor. The vacuum cleaner picked up everything and left the floor
clean and dry. Unfortunately, the machine was too cumbersome,
required too much time, and interferred with production. One
solution is to eliminate the use of the hasher-washer drain as a
handy disposal for all kill floor wastes. Bloody material must be
picked up for disposal directly into inedible rendering by broom
(or squeegee) and shovel, or by vacuum cleaner.
Problem: Sawdust from the carcass splitting saws is carried down the drains,
by the cooling water used with the saws or it is swept down the drains
during mid-shift and daily clean-up.
Solution: At Madison, dry clean-up (broom and shovel) is used. The material
is placed in a barrel for inedible rendering. In Davenport, a
stainless steel catch pan was designed which slopes to a chute
leading to inedible rendering. All of the sawdust can then be
swept directly to inedible rendering.
Problem: Wastewater from drinking fountains, lavatories and overspray from
the treadmill washer flow across blood clots in the gutter near the
eviscerating treadmill and leach out soluble pollutants.
Solution: Water was diverted from the gutter and the blood on the floor was
kept as dry as possible until it could be picked up during periodic
clean-up. Another solution would be to build a stainless steel
trough under the kill chain to catch all blood clots and any tissue
trimmed from the carcasses. This solution was not tried since the
kill chain in Madison is too low. The heads of the largest carcasses
would touch any trough elevated more than three inches off the floor;
a trough this low would interfere too much with clean-up.
294
-------
Problem: The viscera pan washer and the eviscerating treadmill water use
excessive amounts of water. Continuous sprays on both of these
conveyors wash them according to USDA Regulation. Some of the
sprays were 1/8 inch holes drilled in a 1 1/2 inch pipe. Many
of the nozzles used on the treadmill washer were oversized.
Solution: Spraying Systems Co. helped design sprays which would do the
job using the least amount of water. Water use was reduced
from 115 gallons/min. to 40 gallons/min.
Problem: Clean-up workers would leave water running in the viscera pan and
treadmill washers for several hours during clean-up.
Solution: Solenoid valves controlled by an electrical timer were installed
on all of the water lines to the viscera pan and treadmill washers.
During production, water is needed continuously and the timer is
set on manual so the solenoid valves remain open. At the end of
production, the timer is set on automatic and the cabinet is locked.
The clean-up man can use the sprays by pushing a switch on the out-
side of the cabinet which turns on the sprays for 15 minutes only.
This switch can be activated as often as needed. Clean-up
personnel cannot leave the water running for hours just because
they forgot to turn it off.
Problem: The hasher-washer is the largest contributor of pollutional material
from the kill floor. The hasher-washer recieves the condemned
viscera, all intestines, pizzle and piggy bags and the fat and
tissue from the neck washer. When the intestines are slashed open,
their contents are washed out to the drain. The large size of the
openings in the washer screen also allows the small particles of
fat and other connective tissue from the neck washer to pass
through and be lost to the drain.
Solution: The blades were removed from the hasher-washer so that the intestines
went to inedible rendering intact. This increased the daily yield
of meat scraps by 8,500 pounds/day while reducing the BOD loading
from the kill floor by over 50%. There was some lowering in
grease quality when this was done,but this loss was overshadowed
by the increased quantity of meat scraps. A further test, which was
not tried, is to replace the present screen with a fine mesh
dewatering device which will retain small particles of fat and tissue.
Problem: The stomach washing process consists of a slitter - dumper where the
stomachs are slit open and the contents are flushed out and a tumble
washer which uses a continuous flow of water to wash the contents
from the stomachs before the mucosa is stripped. It was felt that
too much water was used and discharge has a high pollutional loading.
Solution: A fully satisfactory solution was not found. Less water in either
the slitter - dumper or the tumble washer fails to remove all of the
contents from the stomachs and made the muscosa unacceptable for
pharmaceutical use. Tests indicate the average weight of a stomach's
295
-------
Problem:
contents is one pound, of which 81% is water. Of the 19% that is
solids, 1.83% is protein. Many of these solids are soluble and
once dissolved, are not recoverable by screening. The portion of the
stomach contents that filtered through a 20 mesh screen analyzed
at 41,000 ppm TS, 37,000 TVS and a BOD of 42,000 ppm. What is
needed is a slitter - dumper which will dump the stomachs dry and
convey the materials to some place for landfill disposal or for dry-
ing and refeeding.
Six shower spray nozzles were used to wash the carcass cavity and
heads before the heads were removed from the carcass for processing.'
Solution: The number of shower nozzles were reduced to three and the pressure was
reduced from 60 psi to 20 psi saving 18 GPM.
Problem: The manually operated neck washer used too much water and scraped off
too much fatty tissue which was lost to the drain.
Solution: In Beardstown and Davenport tanks were placed under the kill rail in
the neck washing area. The tanks are equipped with surface weirs
which discharge the water while retaining floating fatty solids.
The solids are removed at the end of the shift and sent to inedible
rendering.
Tests were run using a vacuum system to remove the blood clots from
the stick wound. However, the clots were too firmly imbedded in
the connective tissue and could not be removed using just the vacuum.
A Chad Neck Washer was installed in Madison. This machine uses an
oscillating set of sprays at 800 psi to scour blood from the stick
wound. The Chad Unit does not remove as much fat and connective
tissue as the manual washers and it uses less water than two manual
washers 20 GPM vs. 26 GPM. Because of the wide spray pattern,
the unit also does a good job of washing the interior of the carcass.
By relocating the neck washer, it is hoped it can be used to wash
the heads and interior of the carcasses and eliminate the use of the
separate head washer.
Problem: Beardstown is the only Oscar Mayer plant where chitterlings are
still saved. The chitterling washers use too much water. Most of
the water was being used through eight shower nozzles to wash workers
hands and the exterior of the chitterlings.
Solution: The shower nozzles were replaced with Spraying Systems Co. 3/8 inch
GG "Fulljet" Nozzles. Meter readings show a reduction in water use
of 51,850 gallons/day for three chitterling machines.
296
-------
CONCLUSIONS
There is no argument that there is a need in the meat packing industry to
conserve water and reduce pollutional loadings in wastewater. Quantification
by metering and analysis is usually not necessary unless information is desired
concerning the savings generated by making a change. Most of the processes
which are heavy polluters or large users of water are readily identifiable
by visual observation and by having a thorough knowledge of slaughtering
processes. Each place in the slaughtering process where water is used must
be examined to determine if the water is really necessary, or if it is being
used in the most efficient manner.
Whether or not water can be reused within the plant will depend greatly upon
the processes the water is to be used in and whether the local USDA inspector
will approve of the use. At Oscar Mayer, reuse water is used in one plant to
sluice hair to the sewage treatment plant, and in another plant to sluice con-
demned viscera to the hasher washer. We are questioning reuse water for these
purposes, however, since the water leaches pollutional material from the wastes
being transported. It is felt that dry conveying would be better.
It is important that any changes which are made to reduce water use or pollu-
tional loading be done without increasing human labor. At the present time,
it is less expensive to pay the extra surcharge cost for over a ton of BOD
rather than add one man day of labor to a process.
297
-------
POULTRY PROCESSING WASTEWATER-
ADVANCED TREATMENT AND REUSE**
by
Daniel T. McGrail*
INTRODUCTION
A water reuse project is currently under investigation by the
Bureau of Sanitary Engineering of the Maryland State Department
of Health and Mental Hygiene. An advanced treatment plant was
designed and built to study the possibility of recycling poultry
processing wastewater. The final report was published in Iferch
19y4. It^demonstrated that the reclaimed water satisfies the
chemical, biological, and physical limits for potable grade
water. It was recommended, however, to continue the study in
order to demonstrate the presence of any health significant
characteristics not demonstrated by existing drinking water
standards. Also,this study is to show the safety of poultry
processed in this reclaimed water.
The funds were renewed and data collection will continue through
August 1976 when the final report and recommendations will be
published. Much of the data already compiled, however, can be
used to investigate the feasibility of constructing a similar
plant to meet the E.P.A. proposed poultry processing wastewater
discharge limitations. As the deadlines for the proposed 1977
and 1983 effluent standards approach (see table 1) most meat
packing and poultry processing plants will require treatment
past the existing secondary treatment. The aerated lagoons,
followed by the advanced water treatment, consisting of micro-
straining, flocculation and sedimentation, and sand filtration
is a system certainly worth looking intoc
* Maryland State Department of Health and Mental Hygiene,
Bureau of Sanitary Engineering.
**This investigation is supported by funds from the Environmental
Protection Agency, Pacific Northwest Regional Laboratory, under
Grant No. S-803325-101
298
-------
TABLE 1. PROPOSED POULTRY PROCESSING PRODUCTS LIMITATIONS 4/24/75 RBB. REG. VOL. 40, NO. 8
ro
to
to
Subcategory
7/1/77
Best Practicable Technology
Hew Sources
7/1/83
Best Available Technology
Chicken Processors;* Max:** 0.92
Aver"*
Turkey Processors: Max:
Ave:
Max:
Ave:
Max:
Fowl Processors:
Duck Processors:
Ave:
BOO,
0.92
0.46
0.78
0.36
1.22
0.61
1.54
0.77
TSS
1.24
0.62
1.14
0.57
1.44
0.72
1.80
0.90
O&G
0.40
0*20
0.29
0.14
0.30
„ ,C-* ,0.15
0.52
0.26
BPT+WU-N
0.40
0.20
0.28
0.14
0.30
% 0.15
0,26
0.13
BOD,
0.60
0.30
0.42
0.21
0.46
0.23. ,
0.78
0.39
TSS
0.68
0.34
0.48
0.24
0.54
^27
0.92
0.46
O&G
0.40
0.20
0,28
0.14
0.30
0.15
0.52
0.26
NH.-J
(rag/:
8
4
8
4
8
4
8
4
Fecal coliform limit is 400mpn/100ml
1983: TKN.4mg/lf TP»2mg/l, N03/N02-
*A11 units are kilograms/1000 kilograms live weight killed unless listed milligrams/liter.
•* Maximum for any one day
•••Average of daily values for thirty consecutive days.
-------
BACKGROUND
The site of this project is the Sterling Processing Corporation
in Oakland, Md. It is presently engaged in the slaughtering,
eviscerating, and processing of broilers. The present plant
capacity is 6000 birds an hour with a daily average of 175,000
pounds live weight killed. The present operational wastewater
lagoon system consists of rotary screening, a primary and
secondary aerated lagoon system and a chlorine contact chamber.
This project was initiated due to the limited water resources
of the area. After strict water conservation measures^ usage
was cut from 11 gallons per bird to less than 7° The town of
Oakland still could not supply Sterling with its needed 300,000
gallons per day. Three deep wells and a water treatment plant
consisting of flocculation and sedimentation, sand filtration,
and chlorination were constructed. Geological studies have in-
dicated that additional wells will not increase the ground water
supply. Periodic water shortages still plague this plant, par-
ticularly during the summer months. The advanced water treatment
plant facilities were designed to allow the reclaimed water to
enter the plant as a raw water source, before the existing water
treatment. (Figure 1)
RAW WASTEWATER CHARACTERISTICS
Composite samplers were used to collect daily wastewater samples
from the processing plant. Table 2 shows the results of these
sampleso Average wastewater flow from the plant is 300,000
gallons per day.
TABLE 2. RAW WASTEWATER CHARACTERISTICS
BOD.
TSS
O&G
*300.000 gallons
No0 of
Samples
148
119
47
per day and
Ave.
7.7
11.9
5.7
175.000 1
•^•••••WM-IWWMIIIM-M-MIIMI-MMMMIM^^
Max.
(k/kkg Iwk
14.8
39.5
66.4
bs. Iwk.
•* 'Hi— -»P^— ».
Win.
)*
2.0
2.5
1.2
^•^•^^^•^^^•K^M^^WltfB
Stan.
Dev.
3.2
6.7
3.4
300
-------
Figure 1.
WASTEWATER TREATMENT
FACILITIES
LAGOON
CHLORINATION
DRIVER
LAGOON
SOLIDS
SLUDGE
MICROSTRAINER
SCREENS
FLO CCULAT ION-
SEDIMENTATION
SAND
FILTER
PROCESSING
PLANT
WELL!
WATER
I
DRIVER
STORAGE
301
-------
WASTEWATER LAGOON SYSTEM
The wastewater lagoon system is an effective economical method
of treatment. BOD$ loadings average 4501bs./acre/day with 93%
removal, grease is loaded at 3301bs./acre/day with 96% removal,
and suspended solids loading averages ?501bs./acre/day with
79% removal.
The primary lagoon is six feet deep, 5^0 feet long, and 1^0 feet
wide. The detention time averages eight working days. It is e-
quipped with 64 Griffith type circulators, a manual grease skimmer
and an overflow weir to the second lagoon. The grease is skimmed
daily.
The second lagoon is six feet deep, 400 feet long,' and 140 feet
wide. It is equipped with 40 Griffith type circulators, a chlorine
contact chamber, and an overflow weir to the river. Air is sup-
plied to the lagoons by two 20 H.P. and one 30H.P. motors with
Suterbuilt blowers at low pressure. The wastewater treatment
facilities were designed by Griffith Engineering of Falls Church,
Virginia.
The Griffith type system has been in operation for ten years with
excellent results. Problems have arisen in the spring of the year
when all ponds and lagoons turn over, on in very extended cold
weather such as 0 degrees Fahrenheit or less for several days.
Fifty consecutive tests between June and December had an average
BODc of 15PPm» suspended solids of 62ppm, and grease 3.3ppm.
Even considering the results during normal operations, E.P.A.
limitations for this plant would be 4lppm suspended solids.
TABLE 3. AVERAGE SECONDARY LAGOON EFFLUENT CHARACTERISTICS
% Samp3.es % Samples
No. of Ave. Max. Exceeding Exceeding
Samples (k/kkg Iwk) 197? ifex. 1983 Max.
Q
BOD5 121 0.50 5.6 16%
TSS 131 1.80 6.4
GREASE 59 0.20 1.2
Table 3 shows the average lagoon effluent characteristics
302
-------
MICROSTRAINER
The microstrainer, part of the original study which did not include
facilities for flocculation and sedimentation, was designed
to handle an average BOD,? of 15 ppm, suspended solids of 60 ppm,
and grease of 8 ppm. It was found that the discharge from the
lagoons exceeded these values during extreme weather conditions.
Due to blockage of the screens it was necessary to change the
30 micron screens with a ?0 micron screen. These screens are
continually backwashed by cold water sprays, and a hot water
spray is available when necessary. Replacement of the screens
did increase the maximum flow through the unit, however, it
cut its treatment efficiency in half. Though the problems have
been greatly reduced, when the worst conditions exist, flow
through the unit diminishes.
Table 4 shows the treatment efficiency of the microstrainer.
TABLE 4. EFFECTIVENESS OF MICROSTRAINER
No. of
Samples
Ave . Max *
(k/kkg Iwk)
% Samples
Exceeding
1977 Max.
% Samples
Exceeding
1983 Max.
BOD5 21 0.42 1.7
TSS 18 1.4 3,0
Grease - -
FLOCCULATION AND SEDIMENTATION ,: ,
Following the microstrainer the water begins the flocculation and
sedimentation process. These facilities were designed by James D.
Clise of the Maryland State Department of Health and Mental Hygiene,
Extensive jar tests were conducted to determine the most effective
dosages of alum and lime. The use of a polyelectrolytic coagulant
aids was studied. The tests were done using one liter samples
of microstrainer effluent. After chemical additions, stirring
rates were consistant. The flash mixing rate was 100 rpm,
followed by flocculation at 50 rpm for 15 minutes, then finally
sedimentation for 30 minutes.
The results of the jar tests showed the best floe formation with
dosages of 100 ppm alum, 35 ppm lime and 1 ppm of HERCOFLOC1 818.2,
a slightly anionic coagulant aid. It is manufactured by Hercules
Chemicals of Wilmington, De. These dosages are checked regularly
303
-------
by daily jar tests, but the levels usually remain constant.
It was found that through the introduction of the coagulant aid,
suspended solids removal increased from 60$ to 70$ in the settling
basin. The jar tests also gave an indication of the sludge pro-
duction rate. In a one liter sample after the normal test procedure
100 ml of sludge settles. This rate of sludge build up has caused
serious operational problems. Normally every two million gallons
treated requires the complete shutdown of the advanced treatment
unit while the sedimentation basin is drained and the sludge is
pumped. The difficulty in sludge removal is due to an insuf-
ficient slope on the bottom of the basin. Though there is a slope
the texture and thickness of this sludge is such that it will
not flow or drain properly.
Flocculation and sedimentation is a very critical step in the
treatment process. Not only is ?0# of the suspended solids
removed and 80^ of the BOD* removed but it is very successful
for the removal of colloids not removed by either sand or dia-
tomaceous filtration.
One more comment on chemical dosages is worth noting. At present
the goal is much higher than a standard waste treatment plant.
Of 46 tests only one exceeded 1977 maximum day limits and 2 ex-
ceeded 1983 limits for suspended solids. All BOD
-------
SAND FILTER
A simple sand and gravel gravity filter follows the floccu-
lation and sedimentation process. This step is necessary to
insure proper levels of suspended solids. Potable grade water
is the result of this process, the water quality is so high that
none of the samples even approach the pollution discharge limits
for 1983.
Something to consider, however, is the length of run of the sand
filter. Presently filter runs average 200,000 gallons between
backwashes, or 16 hours of continuous operation. If the chemical
dosages were cut back this figure would almost certainly drop.
The backwashing procedure is simple. It only takes approximately
five minutes and by using the extra storage capacity in the
settling basin a thorough backwash can be completed without
a plant shutdown.
The effectiveness of the sand filter is shown in table 6.
TABLE 6. EFFECTIVENESS OF SAND FILTRATION
BOD5
TSS
GREASE
No. of
Samples
56
49
27
Ave. Max.
(k/kkg Iwk)
0.016 0.11
0.0?4 0,20
0.035 0.082
% Samples
Exceeding
1977 Max.
Ofo
Q%
Vfo
% Samples
Exceeding
1983 Max.
0<6
Ofo
Q%
DISINFECTION
Chlorination is performed twice throughout the treatment process.
After the second lagoon enough chlorine is added to carry a 0.1
ppm free residual through the sand filter. This is accomplished
by applying a normal dose of 25 pounds per day or 10 ppm. Final
disinfection is then accomplished by another application of 25
pounds per day.
Under no conditions was the fecal coliform limit of 4oOmpn/100ml
exceeded, and since September 1975. when repairs were made in the
chlorine injector, all test results have shown an mpn of less than
2. That is the limit of sensitivity for bacteriological tests at
the state branch laboratory.
Table 7 shows the effectiveness of disinfection.
305
-------
TABLE 7. DISINFECTION
TOTAL GOLIFORMS
FECAL COLIFORMS
No. of
Samples
84
84
Ave.
MPN/lOOml
<2
Max.
MPN/lOOml
8
<2
ECONOMICS
The last consideration, of course, is economics. As of March 19?4
the total annual cost of the lagoon system was $22,500 or $.31/1000
gallons. The advanced treatment costs were determined to be $19,500
per year, for a totalcost of $42,000 per year. These costs include
interest on the investment, chemicals, power, and one full time
operator shared between the two units.
Construction costs of the wastewater lagoon system was $84,000
not including land costs. Construction costs of the advanced
treatment unit totalled $90,000 for a combined cost of $174,000.
Considering the plant production is 40 million Ibs, Iwk per year,
it costs approximately o.l£ per pound Iwk or o.T4<£ Per pound dressed
weight. These costs alone demonstrate the economical feasibility
of a plant of this nature* (Table 8.)
SUMMARY
It has been shown that flocculation and sedimentation followed
by sand filtration is an economically feasible method of upgrading
an existing aerated lagoon system.
The microstrainer, installed before the flocculation and sedimenta-<
tion facilities, is the cause of many of the operational problems
due to the variations in the lagoon effluent.
The difficulty in sludge removal in the sedimentation basin is a
problem that could be overcome in future designs.
The sand filters need for frequent backwashing is a daily maintenance
procedure that does not require a plant shutdown.
306
-------
TABLE 8.
TREATMENT SUMMARY
AVERAGE
PER CENT FLOCCULATE-
REMOVAL LAGOONS STRAINER SEDIMENTATION FILTRATION OVERALL
u>
o
BOD5 93 15 78 82 99.8
T5S 79 21 67 84 99.4
GREASE 96 - is 99.4
-------
ALTERNATIVES FOR TREATING
POULTRY PROCESSING WASTEWATER
Dr. Franklin E. Woodard*
INTRODUCTION
The processing of poultry for the retail trade has been developed as a "water
intense" industry. Water is used in large quantities because of its superior
heat transfer qualities and for reasons of sanitation. Furthermore, this in-
dustry, like most others, was developed during a time when water pollution was
considered far less important than low cost production. Public Law 92-500, or
the 1972 ammendments to the Clean Water Act of 1969, have placed this nation in
a totally different situation. Environmental pollution is now at least as im-
portant as production cost; therefore, it seems reasonable that many of the pro-
cessing methods which were developed under former circumstances should be changed.
Basically, there are two approaches to the problem of decreasing the discharge
of water pollutants from a processing plant. One is to build treatment systems,
and then more treatment systems to treat the effluent from the first treatment
systems, until the desired water quality is finally achieved. The second is
to change the methods used in the processing plant so that less water pollutants
will be generated. Both approaches require substantial technological develop-
ment and a period of testing before satisfactory performance can be assured.
In administering Public Law 92^500, the U.S. Environmental Protection Agency
issues guidelines, essentially legal requirements, which restrict the allow-
able discharge of water pollutants from given industries on a step-wise basis
leading to eventual zero discharge. They are issued as National Pollutant Dis-
charge Elimination System (NPDES) guidelines. These guidelines are based upon
the discharge of pollutants from certain acceptable wastewater treatment systems,
but as written they do not require the actual installation of a treatment system;
only an equivalent reduction in pollutant discharge. Whether this reduction
would best be accomplished by wastewater treatment or process change can be de-
termined only by an in depth cost effectiveness analysis.
Purpose and Scope
The purpose of this paper is to present a discussion of the pollutant reduction
alternatives available to the poultry processing industry, to present data rel-
ative to the technical feasibility of those alternatives and to present a method
for determining the cost effectiveness of each alternative or set of alternatives.
Not included in this presentation is a discussion of alternatives such as dis-
charge to a municipal wastewater system or final treatment on site by a
308
-------
biological treatment system. Since the effluent from neither of these systems
can be recycled and reused without still further treatment, they are not con-
sidered appropriate in a program whose ultimate goal is zero discharge.
DESCRIPTION OF POULTRY PROCESSING
A typical poultry processing plant produces defeathered and eviscerated whole
and cut up chicken as outlined schematically in Figure 1. Birds are received
live in crates and are hung by the feet on a moving conveying system in the
receiving area. After being killed and bled they are soaked in a hot water
bath called a scalder to facilitate defeathering. They are then defeathered
mechanically, washed and are moved on to the evisceration area where the offal,
lungs, giblets and heads are removed. After a final wash they are emersed in
an ice water bath to cool the meat to approximately 37° F (2.78° C). This ice
water bath, referred to as a chiller, consists of two serial operations; the
pre-chiller and the second stage chiller. Water moves counter-current to the
movement of the birds. The chilled'birds are either packed in ice and shipped,
or are further processed (cut up and/or cooked).
There are four major sources of wastewater in a typical poultry processing plant.
Federal regulations require that the scalder overflow at a rate of at least
1 qt/bird/day, and that the chiller overflow at a rate of at least 2 qt/bird/day.
The third major source is the viscera carriage flume, whose function is to,flush
the viscera away from the evisceration stations to a screen. The water in the
viscera flume originates at the final bird wash station, hand wash stations and
side pan wash which constantly washes down the sides of the flume to keep it
clean. Approximately 360 gpm (22.72 I/sec) of wastewater flow from the viscera
flume in a 12,000 bird per hour plant.
The fourth major source of wastewater that must be treated before discharge is
the nightly plant washdown. The quantity of washdown water is usually about
300,000 gpd (1,135, 500 I/day).
Table 1 presents a summary of the pollutional loads from a typical 12,000 bird
per hour chicken processing plant. The values listed in Table 1 are averages
obtained during many days of sampling at two poultry processing plants in Maine.
The total plant flow is listed as 900,000 gallons per day (gpd) (3,406,500"l/day)
for a processing day during which 90,000 chickens (broilers averaging 3% pounds live
weight were killed. Experience has shown that this total flow can be reduced
to as little as 700,000 gpd (2,649,500 I/day) with some reduction in total"
pounds of pollutants; however, the total weight of pollutant discharged does not
decrease in proportion to water reduction. All of the calculations of pollutant
discharge in this paper were based upon the values shown in Table 1.
Table 1 shows that the largest sources of 6005, suspended solids (S.S.), and
grease discharge are the viscera flume and the night washdown. The viscera
flume contributes 1,516 pounds of BODg per day, or 38 percent of the total
plant discharge of 4,002 pounds per day. In addition, it contributes 34 per-
cent of the total S.S. load and 57 percent of the total grease load. The con-
tribution of BOD5, S.S., and grease from the night washdown are 1,274 Ib/day
(578.4 kg/day), 775 Ib/day (351.85 kg/day), and 650 Ib/day (295.1 kg/day),
respectively, amounting to 32, 25 and 26 percent of the total.
309
-------
Receiving
Killing
Bleeding
CO
h»
S.
+| Scalding
Defeather
Wash
Eviscerate
Rotary
Screen
Plant
Wastewater
Effluent
Wash
Chilling
Packing
Washdown at Night
Ice Machine
Screenings to
Rendering Plant
Figure 1. Schematic of a typical poultry processing plant.
310
-------
TABLE 1. POULTRY PROCESSING WASTEWATER CHARACTERISTICS
u>
System
Total
Plant
Vise
Chill
Overflow
£ Dump
Scald
Overflow £
Dump
Night
Washdown
Flow BOD S.S. Grease £ Oil
GPD mg/1 Ib/day mg/1 Ib/day mg/1 Ib/day
900,000 534 4,002 411 3,082 338 2,537
350,000 520 1,516 ,..360 1,050 500 1,458
,?i»
140,000 600 700 500 583 280 327
65,000 600 325 900 487 50 27
300,000 510 1,274 310 775 260 650
Other 45,000 500 187 500 187 200 75
-------
NPDES GUIDELINES
At the present time, the proposed guidelines relative to the chicken sub-
category of the poultry processing industry are as follows (1):
BOD5: 0.46 kg/kkg LWK*
Suspended Solids: 0.62 kg/kkg LWK
Grease: 0.20 kg/kkg LWK
(* LWK = live weight killed)
(kg/kkg LWK = lb/1000 Ib LWK)
As mentioned previously, these guidelines are still only proposed. Whether or
not these guidelines should be changed is still under consideration. It is in-
tended that this paper will provide information which will aid in a rational
decision to change these guidelines.
ALTERNATIVES FOR POLLUTANT REDUCTION
Three systems for reducing the discharge of pollutants from poultry processing
plants are evaluated in this paper. A brief description of each follows:
1. Treatment of the combined wasteflows by chemical coagulation and dissolved
air flotation followed by sufficient tertiary treatment to meet NPDES require-
ments. The tertiary treatment stages considered in this work included sand
filtration followed by activated carbon adsorption.
2. Effluent flow reduction by replacement of one or more of the items of water
using process equipment with non water using equipment. There are three areas
where such replacement could be implemented. The first would be to replace the
ice water chilling process with a cold air blast system, referred to as "dry
chill". A system of this type has been in operation at one of the processing
plants included in this work for several years.
A second possible alternative would be to replace the viscera removal system
with a vacuum system. Several of these systems have been installed at various
plants in the country, with varying degrees of success.
A third alternative which is at present only in the speculative state, would be
to scald the freshly killed birds in a steam chamber rather than in the present
hot water bath.
The resulting total plant effluent after one or more of the above replacements
were made, would have to be treated to the extent that NPDES requirements would
be met. Financial feasibility would be determined by a comparison between the
cos.t of a given installation vs. the cost of treating the wastewater which that
installation eliminated.
3. Treatment and reuse of water at each individual process. A physico-chemical
312
-------
type of treatment system consisting of screening, chemical coagulation, dis-
solved air flotation, followed by sand filtration, activated carbon adsorption
and disinfection is considered in this paper. Heat conservation should also
be employed for reasons of economic as well as bacterial control.
Alternatives 2 and 3 have essentially the same purpose with respect to pollut-
ant discharge reduction. That is simply to eliminate a water using process as
a source of pollutant contribution to the total plant effluent. Table 2 shows
the calculated pollutant discharge levels that could be achieved by "drying
up" (using a non water using process) or "bottling up" (by treatment and re-
cycle at the source) each of the three major water using processes.
The values presented in Table 2 were calculated as follows: It was assumed
that a secondary treatment system achieving 85 percent removal of BOD5 90 per-
cent removal of S.S. and 95 percent removal of grease would treat the final ef-
fluent from each alternative system. It was further assumed that replacement
of the viscera removal system with a dry system would remove 90% of the present
pollutional load from that source, and that "drying up" the chiller and/or the
scalder would remove 100 percent of the pollutional load therefrom.
Since it has been shown that dissolved air flotation incorporating chemical co-
agulation (DAF) is the most effective and economical system for removing 85 to
95 percent of the pollutants from poultry processing wastewater (2) (3) (4),
DAF was assumed as the secondary treatment system. However, any secondary
system producing 85% BOD5 removal, 90% S.S. removal and 95% grease removal
would yield the same results as those presented in Table 2 .
Table 2 shows that, based on the pollutant discharge levels presented in Table
1, the discharge of BOD^ from a plant having all conventional water using pro-
cessing equipment and having a secondary treatment system (as specified above)
would be 1.91 kg/kkg LWK, 315 percent in excess of the amount allowable by the
proposed NPDES requirements. The calculated discharges of S.S. and grease are
0.98 and 0.40 kg/kkg LWK, respectively, and are 58 percent and 100 percent in
excess of the proposed NPDES requirements. Table 2 shows further that insta-
llation of a dry viscera removal system in conjunction with secondary treat-
ment of the remaining total plant effluent would result in achievement of the
proposed grease limitation, but that the S.S. discharge would be slightly ex-
cessive and the BOD discharge would be 1.91 kg/kkg LWK, 1.45 kg/kkg LWK great-
er than the proposed limit of 0.46 kg/kkg LWK. Still further drying up by in-
stallation of a dry chill system would result in satisfaction of both the S.S.
and grease requirements but the BOD5 discharge would still be excessive. In
fact, Table 2 shows that even after all of the major water using processes
have been replaced by either a dry processing system or a treatment, recycle
and reuse system, the proposed limitation for BOD- could not be met by ordin-
ary secondary treatment of the remaining effluent. This fact strongly suggests
that the proposed guidelines should not be promulgated.
The results shown in Table 2 do not suggest that the installation of dry pro-
cessing equipment (or treatment and recycle at individual sources) is an in-
feasible approach. If the presently proposed guidelines were to be adopted,
further wastewater treatment would have to be installed along with any one of
the alternative systems listed in Table 2 . The optimum total system would
373
-------
TABLE 2. EFFLUENT DISCHARGE LEVELS
Alternative Systems
No. 1 DAF Treatment
alone
(Total Discharge (kg/kkg LWK)
BOD5 S.S. Grease £ Oil
1.91 0.98 0.40
No. 2 DAF Treatment
£ dry vise
1.26
0.68
0.19
CO
No. 3 DAF Treatment
dry vise £
dry chill
0.92
0.49
0.14
No. 4 DAF Treatment,
dry vise
dry chill £
dry scald
0.77
0.46
0.34
0.62
0.14
0.20
Proposed NPDES Guideline
* Assuming that the final effluent from each alternative system is treated in a secondary
plant which achieves 85 per cent removal of BOD5, 90 per cent removal of S.S. and 95 per
cent removal of grease £ oil.
-------
depend upon the lowest overall cost. For instance, even though the discharge
from alternative No. 3 would have to be further treated, the total cost might
be less than the cost of tertiary treatment of the larger quantity of effluent
from a totally wet, conventional processing plant.
TECHNICAL FEASIBILITY
Results relating to the technical feasibility of each of the alternative
systems have been published elsewhere (3) (5). Selected results are presented
in this paper to illustrate (1) that achievement of the proposed NPDES guide-
lines is technically feasible and (2) that the expense of activated carbon
adsorption is extremely high.
Tertiary Treatment of Total Plant Effluent—Table 3 shows the maximum concen-
trations of BOD5, S.S. and grease which could be discharged from a plant pro-
cessing 90,000 chickens per day averaging 3.5 pounds (1.59 kg) and discharging
900,000 gallons of wastewater per day (3,406,500 I/day). Figure 2 shows that
the requirement for S.S. (< 26 gm/1) can be met by sand filtration. Figure 2
shows that at loading rates of both 2 and 4 gpm/ft^ (81.4 and 162.8 1/min sq m)
effluent suspended solids concentrations were well below 26 mg/1. Figure 3
shows a column exhaustion curve obtained by passing DAF effluent through a
column of activated carbon. A search of nine brands of activated carbon showed
that the brand used for the exhaustion curve shown in Figure 3 produced the
best exhaustion characteristics and the highest carbon loading rate (5). This
carbon loading rate, however, was extremely low, and Figure 3 shows that an ef-
fluent concentration of more than 20 mg/1 of TOC was observed after only 6
hours of operation at a contact time of 22.4 minutes. The cost for carbon
alone for this system would be in excess of $50.00 per thousand gallons treated.
Even if a carbon could be developed which was 10 times more effective for this
wastewater, the cost would exceed $5.00 per thousand gallons, ten times greater
than the $.50 per thousand gallon cost for DAF treatment.
Treatment and Recycle of Scalder and/or Chiller Water—Figure 4 shows a sche-
matic of a physico-chemical system for treating and recycling chiller and/or
scalder overflow. Here again, the first treatment step should be DAF (using
chemical coagulation), not only because of its high performance and favorable
efficiency, but also because the sludge produced can be harvested and used as
a poultry or animal feed additive. The effluent from the DAF unit would be
filtered through a sand, mixed media or diatomaceous earth filter for solids
removal, then activated carbon adsorption would be employed to remove dissolved
organics. Figures 5 and 6 show the levels of S.S. and TOC that could be achie-
ved with this type of treatment process. Figure 5 shows that the suspended
solids could be expected to be held below 15 mg/1 and Figure 6 shows that the
TOC could be held below 25 or 30 mg/1. Figure 6 also illustrates the poor
efficiency of activated carbon adsorption for this application as far as carbon
loading rate is concerned. As was the case for activated carbon treatment of
the pretreated total plant effluent, the carbon loading was shown to be less
than 0.01 Ib of impurity per pound of activated carbon, resulting in a cost of
more than $5.00 per thousand gallons.
The results presented in Figures 5 and 6 illustrate the approximate water qual-
ity that could be achieved with the physico-chemical treatment system shown in
315
-------
TABLE 3. POLLUTANT DISCHARGE LIMITS*
Pollutant Maximum Cone, for
900,000 GDP Flow
BOD5 19
S.S. 26
Grease 8
316
-------
CO
O
CO
Q
O
CO
Ul
0.
CO
r>
CO
z
UJ
Co8 34 MG/L
2GPM/FT2 (81.4 L/MIN/SQ.M.)
•4 " (162.8 " )
3 4
VOLUME (LITERS)
FIG. 2: SAND FILTER PERFORMANCE ON DAF EFFLUENT
-------
60
50
40
a
X.
O
30
o
o
20
10
12 18
TIME (HRS.)
24
30
FIG.3: COLUMN EXHAUSTION CURVE FOR DAF EFFLUENT
CONTACT TIME = 22.4 MINUTES
318
-------
co
vo
BIRDS IN
WASTEWATER
OVERFLOW
CHILL
OR
SCALD
BIRDS OUT
RECYCLED WATER IN
HOT WATER
MAKE UP
HEAT
EXCHANGE
TREATMENT PHASE
Fig. 4. Schematic of a Physico-chemical Treatment System
-------
CO
l\i
o
°SS(RAW)=38|M6/L ASS(RAW)= 34M6/L °SS(RAW)= 35 MG/L
FLOWRATE = 2 GPM/SF FLOWRATE = 4 GPM/SF FLOWRATE * 4 GPM/SF
CHILLER WATER CHILLER WATER SCALDER WATER
345
FLOW VOLUME (LITERS)
FIG.5: SAND FILTER PERFORMANCE (CHILLER 8 SCALDER)
-------
CO
O
O
60
50
40
30-
u 20
10
oCARBON A
•CARBON D
6
12 18 24
TIME (HRS.)
30
36
42
FIG.6: EXHAUSTION CURVES FOR A FLOWRATE OF 4.0 GPM/FT*(DAF EFFLUENT)
-------
Figure 4. A final decision on technical feasibility could not be made until
the effect of this type of system on the quality of the poultry being pro-
cessed was established.
COST ANALYSIS
Once the technical feasibility of a set of alternatives has been established
a cost analysis must be carried out. to determine the optimum choice of al-
ternatives for any given processing plant. Variables which will affect the
total cost of a system include cost of potable water, cost of wastewater treat-
ment and cost for electricity as well as capital cost and interest rate. A
systematic procedure has been developed to efficiently execute such an analy-
sis based upon network analysis and linear programming.
The basis of this systematic cost analysis method was a network diagram, shown
in Figure 7, and a linear program for solving the network diagram in an ef-
ficient manner using a digital computer. This method is very similar to that
developed by Ward et al (6) except that wastewater treatment and/or treatment,
recycle and reuse are considered in the analysis presented in this paper. A
more complete presentation of this method of analysis has been published else-
where ( 7 ).
The network diagram presented in Figure 7 consists of a series of nodes num-
bered 1 through 22. Each node represents a state in the progress of poultry
processing. Money must be spent to progress from one node to another. For
instance, Figure 7 shows that in order to progress from node 10 to node 12,
money to operate the chiller must be spent. The lines which connect the nodes
are called arcs. Each arc represents a "stage" or process for which money
must be spent. Therefore, each arc has an associated unit cost which is the
cost of processing one gallon per minute of water and an associated flowrate.
The network diagram thus simulates the flow of water through a poultry pro-
cessing plant. Each arc shown in Figure 7 has a name which identifies the
process which it represents. There is an arc for each existing and proposed
process in the plant. The numbers below some of the arc names shown in Figure
7 identify the rate of flow of water through the process which that arc rep-
resents. To illustrate, the arc named SCALDINP between nodes 2 and 3 repre-
sents the pipe used to fill the scalder. There is a cost coefficient assoc-
iated with this arc which reflects the cost of heating this water. The arc
between nodes 3 and 16 named SCALDER represents the scalder itself. The num-
ber 50 which appears below the name represents the required overflow from the
scalder of 50 gpm (189.25 1/min). The arc named SCALDTRT between nodes 16
and 17 represents the treatment system shown in Figure 4. The flowrate through
this arc would equal 50 gpm (189.25 1/min) minus the assigned flow from node
16 to node 15. The cost coefficient associated with SCALDTRT would reflect
the total cost of installing and operating the treatment system. This cost
would include capital cost, interest charges, chemical, electrical, mainten-
ance and manpower costs.
A linear program has been developed to "solve" the network diagram for each
different set of cost coefficients and assigned flow rates. The objective
function of the linear program minimizes the total system cost based on the
cost coefficient and flowrate information for a given set of conditions. The
322
-------
r
4) (3
SCADRCYL
x
SCALDER
50
-L ' - T /
SW DEHEHWSH /
N|f
\ GIBOISCH
JP \ EVISFLOW
CHILLDIS ;S _^
"\1T VACVISFL
\®*-^~
PMCLNDIS \ FETHRCYL
"\ARN ^
FEATSCRN
PMEFFRCL
"6-, I
v TRETMENT
| CLPAJRCH.
0.01
Fig. 7. Network Diagram for
Cost Analysis
323
-------
output from this program is the total cost for the set of alternatives which
is specified in the input. For instance, if it is desired to compare the cost
of running the processing plant with all conventional, water using equipment
(including wastewater treatment costs) with the cost of running the plant with
a dry viscera removal system, the appropriate cost coefficients are supplied
as input and the computer will calculate the total cost associated with both
alternative systems.
Example Cost Analysis—To illustrate the use of the cost analysis method, two
example applications are presented. The first involves a determination of the
minimum cost set of alternatives for a processing plant which is required to
pretreat its wastewater prior to discharge to a municipal treatment system.
Restrictions placed on the effluent discharged to the municipal system include
only a maximum of 100 mg/1 of grease and a surcharge for BOD5 and S.S. concen-
trations in excess of 300 mg/1. Cost data included a 10 year useful life for
major pieces of equipment and a 5 year useful life for pumps, 10 percent in-
terest charge, 3.6C/KWH for electricity, 5t/100 gallons for fresh water and
34<:/gal for fuel oil.
This method of analysis requires that the total cost for wastewater treatment
be calculated separately. It is desirable to calculate the total annual cost
for operating a treatment system including both amortized capital costs and op-
erating costs. This total annual cost is then reduced to a daily cost, then
the cost per day is divided by the number of gallons which will be treated in
the system in a day's time. The result is a cost per gallon of water treated
(usually stated as cents per gallon). This number is the cost coefficient as-
sociated with the arc named TREATMENT between nodes 20 and 21. When this cost
coefficient is multiplied by the flow of water through that arc in gallons per
minute, the result is the cost of treatment in cents per minute. When this
cost is added to the costs for all the other arcs appropriate to one alter-
native system, the sum will be the total unit cost (in cents per minute) for
that alternative system. This result is compared to the total unit cost for
all other alternative systems to identify the least cost alternative.
Table 4 presents an example of the results obtainable using this cost analysis
method. Table 4 shows that if the cost of water treatment is $.4-2 per thou-
sand gallons, as it was for a DAF system for this particular application, the
least cost solution was to build a DAF system to treat the total discharge from
the plant keeping all the conventional water using equipment. Table 4 shows
that installation of a dry viscera removal system would increase the total
unit cost from 128.39 cents per minute to 132.38 cents per minute, even though
the flowrate of water to be purchased and also to be treated decreased from 922
gpm (3489.77 1/min) to 697 gpm (2638.15 1/min).
One of the peculiar characteristics of the situation associated with the results
shown in Table 4 was that the cost for fresh water was relatively low at $.05
per thousand gallons. The results presented in Table 5 show what effect higher
costs for fresh water would have on the optimum choice of alternatives. The
results shown in Table 5 show that at the current cost of $.05 per thousand
gallons the increased cost resulting from use of a dry viscera removal system
over the conventional water trough system would amount to $4-,800 per year.
If the cost for fresh water rose to $.15 per thousand gallons and every other
324
-------
TABLE 4.
RANKING OF THE ALTERNATIVES
TREATMENT COST = 42C/1000 GALLONS
ALTERNATIVES
FLOW, GPM
COST, C/MINUTE
# 1 Wet Sclad, Wet Vis
#2 Wet Scald, Dry Vis
#3 Dry Scald, Wet Vis
#7 Dry Sclad, Dry Vis
#4 Wet Scald, Wet Vis
100% Dry Chill
#3 Wet Scald, Dry Vis
100% Dry Chill
#5 Dry Scald, Wet Vis
100% Dry Chill
#6 Dry Scald, Dry Vis
100% Dry Chill
922
697
884
659
842
617
804
579
128.39
132.38
138.89
142.88
167.83
171.82
178.34
182.32
325
-------
TABLE 5.
EFFECTS OF INCREASING
COST FOR FRESH WATER
Modifications
Fresh Water Cost
C/1000 Gallon
Yearly Amount, $
Effect
Dry Viscera Removal
Dry Scalding
100% Dry Chilling
5
15
25
5
15
25
5
15
25
4,800
2,100
600
12,700
12,200
11,700
47,500
46,600
45,600
Increase
Increase
**Decrease
Increase
Increase
Increase
Increase
Increase
Increase
326
-------
cost stayed the same the increased cost for a dry viscera removal system would
drop to $2,100 per year- If the cost for fresh water rose 10 additional cents
to $.25 per thousand gallons., the alternative including the dry viscera re-
moval system would then be less expensive than the conventional system and
would save $600.00 per year- Table 5 further shows that up to a cost for
fresh water of $.25 per thousand gallons3 the alternative systems which in-
clude dry scalding and dry chilling remain more expensive than the conven-
tional all wet system. A sweeping generalization to this effect cannot be
made, however, since there are so many variable costs involved in the analysis
which will change from the situation at one plant to another.
The results presented in Table 6 involve a different poultry processing plant.
These results are included in this paper to illustrate the change in optimum
system combinations caused by high costs for wastewater treatment, such as
would result from the installation of a tertiary treatment system. Table 6
shows that if the total cost for wastewater treatment were $2.00 per thousand
gallons, the least cost set of alternatives would include a dry viscera re-
moval system and a conventional ice water bath chiller. Installation of a dry
chilling system would result in an increased annual cost of $36,420. However,
if the cost of wastewater treatment were to total $5.00 per thousand gallons,
installation of the dry chilling system would result in an annual savings of
$22,800.
Again it is emphasized that the results presented in Table 6 , or any set of
results obtained using this cost analysis method, cannot be extended to gen-
eralized conclusio'ns. A solution must be generated for each specific set of
circumstances. .
SUMMARY
The generation of water pollutants in a poultry processing plant has bean dis-
cussed and an analysis made of the quantity of pollutants contributed by each
major processing step. Several alternatives for reducing the discharge of
pollutants from a poultry processing plant were presented, including tertiary
treatment, replacement of wet processing equipment with non water using equip-
ment and treatment, recycle and reuse of chiller and chiller overflow water.
The technical feasibility of both tertiary treatment to meet proposed NPDES
guidelines was demonstrated but was shown to involve extremely high cost. In
addition, the characteristics of treated chiller water using DAF, sand fil-
tration and activated carbon adsorption were presented. A systematic method
for determining the least cost set of alternative methods for reducing pol-
lutant discharge to a given level was discussed. Example solutions showed
that local situations involving the cost of fresh water and wastewater treat-
ment costs greatly influence the choice of optimum system configuration.
327
-------
TABLE 6.
EFFECTS OF HIGH
COSTS FOR TERTIARY
TREATMENT
Rank
Combination
Total Cost" Total Flow
£/min gpm
Tertiary Treatment cost = $2.00/1000
1
2
3
4
Vacuum Viscera
£ Water Chill
Vacuum Viscera
£ Cold Air Chill
Existing Viscera
£ Water Chill
Existing Viscera
£ Cold Air Chill
230.935
261.284
292.101
322.130
Tertiary Treatment cost = $5.00/1000
1
2
3
4
Vacuum Viscera
£ Cold Air Chill
Vacuum Viscera
£ Water Chill
Existing Viscera
£ Cold Air Chill
Existing Viscera
£ Water Chill
410.384
429-385
566.630
585.951
gal.
661.5
497.0
979.5
815.0
gal.
497.0
661.5
815.0
979.0
Annual Cost
Above 1st
Ranked Comb.
$
—
36,420
73,400
109,440
-
22,800
187,495
210,680
328
-------
CONCLUSIONS
1. As presently proposed, the NPDES discharge guidelines for the chicken
processing industry can be met only by:
a. Tertiary treatment of the entire wastewater flow
using sand filtration and activated carbon adsorption.
b. Replacement of all of the major processing units
with non water using equipment.
c. Extensive biological treatment.
All three of the above alternatives must be preceded or accompanied
by dissolved air flotation incorporating chemical coagulation.
2. A cost analysis has shown that:
a. The cost for tertiary treatment using sand filtration
and activated carbon adsorption is unreasonably high.
b. Unless, the cost for potable water and wastewater
treatment are relatively high, the only feasible
dry replacement process is dry viscera removal.
3. The cost feasibility of dry chilling depends upon a higher market
value of the product.
4. NPDES guidelines should be modified to allow use of cost effective and
energy conservative treatment commensurate with receiving water quality.
5. Dissolved air flotation incorporating chemical coagulation is capable
of producing a relatively clean effluent which will not result in
significant water quality degredation in many discharge locations.
The cost for DAF treatment is minimal compared to the present cost
for systems which would produce a significant improvement in pollutant
discharge level. Therefore, dissolved air flotation in conjunction
with sufficient chemical coagulation and screening to produce an overall
removal of 90% BODg, suspended solids and grease should be adopted
as best practical treatment (BPT) technology for the poultry processing
industry.
329
-------
LITERATURE CITED
1. Personal Conversation with personnel at Region I, USEPA, J.F. Kenndey
Federal Building, Boston, Massachusetts.
2. WOODARD, F.E., SPROUL, O.J., HALL, M.W. and GHOSH, M.M., "Abatement
of Pollution From a Poultry Processing Plant," Journal Water Pollution
Control Federation, Vol. 44, No. 10, pg. 1909 (1972).
3. REED, S.W. and WOODARD, F.E. , "Dissolved Air Flotation of Chiller
Water Leading to In-Plant Recycle at a Poultry Processing Plant",
submitted for Publication in the Journal of the Water Pollution
Control Federation (1975).
4. WOODARD, F.E., SPROUL, O.J., HALL, M.W. and GHOSH, M.M., "New
Concepts in Treatment of Poultry Processing Wastes," presented
at the llth Annual Environmental and Water Resources Engineering
Conference, Vanderbilt University, Nashville, Tennessee, June
1972.
t-
5. BERRY, L.S., LAFAYETTE, P.F., REED, S.W. and WOODARD, F.E. , "Lab-
oratory Studies into the Reduction of Pollution from Poultry Pro-
cessing by In-Plant Recycle," Proceedings, 29th Purdue Industrial
Waste Conference, Purdue University, Lafayette, Indiana (1974).
6. WARD, R.C.; LINK, DAVID, A.: and CROSSWHITE, WILLIAM, "An Ap-
plication of Network Theory to Water Management in Poultry Pro-
cessing" , Water Resources Bulletin, American Water Resources
Association, Vol. 8, No. 3, (June 1972).
7. FROST, R.E., and WOODARD, F.E., "Analysis of the Effects of
Wastewater Treatment Costs on the Feasibility of Major Processing
Changes Including In-Plant Reuse of Wastewater in a Poultry Pro-
cessing Plant" Proceedings, 30th Purdue Industrial Waste Conference,
Purdue University, Lafayette, Indiana (1975).
330
-------
PROCESSING EGG BREAKING PLANT WASTE
by
J. M. Vandepopuliere*, H. V. Walton*, W. Jaynes* and 0. J. Cotterill*
INTRODUCTION
The egg breaking industry is faced with an increasingly serious waste disposal
problem. The Egg Products Inspection Act (84 Stat. 1620 et. seq. , 21 U.S.C.
1031-1056) was enacted December 29, 1970. On July 1, 1972 phase two of this
act became effective, controlling restricted shell eggs (checks, dirties,
leakers, incubator rejects, inedibles and loss eggs). All of these types of
eggs except checks and dirties must be denatured or destroyed at the point of
segregation to eliminate them from consumer food channels.
There are approximately 150 egg breaking plants in the United States. These
plants yield some 50,000 tons of waste annually. The current method of dis-
posing of these wastes is either landfills or farm land pastures. It is
becoming very difficult to dump this type of material in landfills or spread
on pastures due to the potential pollution problems.
Basic work has been reported by Walton et_ al. on the chemical composition
of egg breaking plant wastes. This work demonstrated that there were signi-
ficant levels of calcium and protein present. The nutritional value when fed
to laying hens, was reported by Vandepopuliere et_al_.'^) to be comparable to
the nutrients that were replaced from feedstuffs normally used in laying diets.
METHODS
Funds were obtained from the Environmental Protection Agency and the Missouri
Agricultural Experiment Station to set up a waste processing system at a
commercial egg breaking plant (Figure 1). Various types of dehydrating and
^University of Missouri, Columbia, Missouri
**This investigation was supported by funds from the Environmental Protection
Agency, Food and Wood Products Branch under Grant # S803614-01-0 and the
Missouri Agricultural Experiment Station.
331
-------
cooking equipment were investigated. A used Heil SD M-5-12 became available and
it was purchased. It is a triple pass rotary drum dehydrator with controls to
modulate the gas flame for control of the furnace temperature and the final
product temperature. The unit was installed in line in an egg breaking plant
(Figure 2).
BULK EGGS
GRADING
EGG BREAKER
EDIBLE FRACTION
INEDIBLE
SEPARATOR
SHELL WASTE
1
LIQUID—
SHELL
SPIN
INEDIBLE
FRACTION
SPUN SHELL
DEHYDRATOR
•*- VIET SHELL DISPOSAL
(OPTIONAL)
EGG SHELL MEAL
Figure 1. Egg Breaking Plant Material Flow Diagram Showing Arrangement for
Shell Waste Dehydration Study.
332
-------
Figure 2. Triple Pass Rotary Dehydrator
Figure 3. Cyclone Collector with Diverter to Dehydrator
333
-------
During the first few days of operation a problem developed with liquid egg
adhering to the hot drum. It was necessary to obtain additional equipment to
remove the superficial liquid from the wastes. A shell spin was obtained
from Seymour and placed in the production line. A special liquid egg (tanner)
separator was designed and constructed to handle the liquid from the inedible
eggs that could not be conveyed to the shell spin.
There were numerous logistic problems involved in setting up to process egg
plant wastes in a plant that was.on stream. It was necessary to set up dual
systems and diverters to permit the wastes to go directly to the disposal
carrier or be diverted to the dehydrator. This was accomplished by installing
extra augers and slides.
To prevent the wet egg shells from sticking to the hot drum it was necessary
to blend in some of the dried egg shell meal. An auger was used to blend egg
shell meal with the unprocessed waste shells and convey the product to the
dehydrator. The cyclone collector was equipped with a diverter (Figure 3)
to regulate the amount of dehydrated egg shell meal that would be returned for
the blending operation.
DISCUSSION AND SUMMARY
The wet egg shell waste contained approximately 17,9 X 10 microorganisms
per gram. Samples were collected asceptically after dehydration. The total
number of microorganisms (Trypticase Soy Broth) surviving each process temper-
ature is shown in Figure 4. The total count decreased logarithmically with
increased temperature. The counts for the two trials after processing at
140°F averaged about 5.6 X 10^ per gram. At an exhaust temperature of 260 F
the counts decreased to about 70 per gram. All dehydrated samples were
salmonellae negative (AOAC, 1975).
TABLE 1. COST OF DEHYDRATOR OWNERSHIP BASED ON AN INSTALLED COST OF $35000
Cost Item
Annual depreciation
Annual interest
(8% of avg. invested)
Maintanance and Repair
(2% of cost, annually)
Insurance and Taxes
(2% of cost, annually)
Annual Cost of Ownership
Cost/day of operation
(250 days/yr)
10 yr
$3500.00
1400.00
700-00
700.00
$6300.00
$ 25.20
Depreciation
20 yr
$1750.00
1400.00
700.00
700.00
$4550.00
$ 18.20
Period
30 yr
$1166.67
1400.00
700.00
700.00
$3966.67
$ 15.87
334
-------
cc
LU
CQ
co
Oi
O
o:
3
CO
CJJ
O
3
A
\
=- \
I
I
MO 180 220 260 300
PROCESS TEMPERATURE (°F)
Figure 4. Destruction of Microorganisms in Egg Breaking
Plant Waste during Dehydration at Various Temperatures.
335
-------
TABLE 2. COST TO PROCESS ONE TON OF DRIED EGG SHELL MEAL BASED UPON ACTUAL
OPERATING COSTS (MISSOURI FIELD INSTALLATION) COMBINED WITH DRYER
OWNERSHIP COSTS
Depreciation Period
Cost Item
Dryer ownership
Labor
(4 hr/day @ $3-65)
Electricity
(58 KWH/day)
Gas
(5100 ft /day)
Total Cost per ton
10 yr
$9.16
5.30
.57
1.76
$16.79
20 yr
$6.62
5.30
.57
1.76
$14.25
30 yr
$5.77
5.30
.57
1.76
$13.40
The economics of owning and operating an egg shell waste processing system
will vary depending on the type and size of equipment and plaiit. The minimum
dehydrator installation cost would be approximately $35,000. Daily ownership
cost, including depreciation; interest, maintenance, insurance and taxes would
vary from $15.87 to $25.20 depending on the depreciation period (Table 1).
The operating cost to produce one (1) ton dried egg shell meal in the
Missouri field installation breaking 1400 - 30 doz. cases/day was $7.63
(Table 2). The total cost was broken down as follows: labor $5.30, electri-
city $0.57, gas, $1.76. Combining ownership and operating expense the total
cost to produce one ton dried egg shell meal ranged from $13.40 to $16.79
depending on the depreciation schedule. These cost figures could be different
at each location due to different cost factors and number of cases of eggs
processed daily.
Dry egg shell meal yield was 3.85 Ib per 30 doz. case. This plant processing
1400 cases per day produced 5390 Ib. or 2.7 tons egg shell meal per day.
Egg shell meal processed from centrifuged wastes has an estimated feeding
value of $35.00 per ton.
336
-------
1. WALTON, H. V., COTTERILL, 0. J. and VANDEPOPULIERE, J. M. Composition
of shell waste from egg breaking plants. Poultry Science 52: 1836 (1973).
2. VANDEPOPULIERE, J. M., WALTON, H. V. and COTTERILL, 0. J. Nutritional
value of egg shell meal. Poultry Science 54: 131 (1975).
337
-------
WATER USAGE IN POULTRY PROCESSING - AN EFFECTIVE
MECHANISM FOR BACTERIAL REDUCTION
F. A. Gardner and F. A. Golan
Poultry Science Department
Texas Agricultural Experiment Station
College Station, Texas
INTRODUCTION
During the past decade, consumer groups, federal regulatory and
inspection agencies, industry groups and private organizations have contin-
uously stressed the need for strict control of poultry product quality
from areas of production to the final marketing of the finished product.
Product quality includes chemical, physical, functional organoliptic and
microbiological characteristics of the product which are of concern to
the poultry user. Awareness of product quality has generated a critical
need for efficient and economical program components for quality assurance
programs. A sound quality assurance program is now an essential part of
all poultry processing operations and can no longer be considered an ex-
traneous process operation. Sound, efficient and effective quality assur-
ance programs must be financially rewarding and of necessity must be based
on a thorough understanding of fundamental attributes of poultry proces-
sing. An effective assurance program must include selected microbiological
characteristics of the product as essential components of the overall
program. Recent concern for what some have considered to be excessive
water use in poultry processing has forced an evaluation of the need for
water in all areas of processing.
OBJECTIVES
The study reported in the paper was designed (1) to determine the
effects of selected processing functions on the bacterial content of poultry
carcasses, (2) to establish what might be considered microbiological "working
normals" for poultry processing, (3) to determine the effects of water
usage in processing on the microbiological characteristics of poultry and
(4) to develop processor awareness of the microbiological concepts involved
in processing poultry.
EXPERIMENTAL DESIGN
The data presented in this study was obtained from five chicken
broiler processing plants and four turkey processing plants in Texas.
Samples were collected from each of the plants on two separate days. On
each sampling day selected water samples were taken prior to the start
of the day's operation and again after four or five hours of continuous
processing. When sampling from the turkey processing plants, selected
equipment surfaces were sampled prior to the start of the day's processing
and again after five hours of continuous processing. Broiler and turkey
carcasses were taken at pre-determined stations and were sampled by swabbing
338
-------
a defined area of the breast surface. Water and equipment samples were
analyzed to establish the total bacterial plate count (35°C incubated for
2-3 days), and total coliform count. In addition, samples from the turkey
processing plants were analyzed for mold and yeast. Carcass samples were
subjected to analysis for these parameters and for Salmonella. The sampling
schedule used in turkey processing plants is summarized in Table 1. All
samples were held in a slush ice environment for transportation to the
Product Technology Laboratory at Texas A&M University where all micro-
biological analyses were performed.
RESULTS
Initial studies conducted at several participating turkey plants in
this study have indicated that surface bacterial concentrations were
highest on the initial carcasses processed each day and decreased during
the first hour of processing (Figure 1). No further changes in bacterial
concentration were noted during an additional four hour sampling period.
As a result of this finding, carcass samples were not taken during the ;
initial 60 minutes of processing.
The data which follows presents only overall means obtained from
equipment, water and carcass sample analyses. No attempt has been made
to present an overall plant comparison or to compare individual plants .to •
plant processing variations.
WATER SAMPLES
Analyses of water samples taken prior to the start of the day's
processing and again after five hours of continuous processing indicate
a rather extensive buildup of mesophiles, coliform, mold and yeast in
almost all turkey processing plant sampling areas. Exceptionally large
increases were noted in the scald water and in the picker drain water
samples indicating the effective carcass washing action of both systems
(Table 2). It should be noted that scald water containing bacterial popula-
tions in excess of one million, organisms per milliliter retained its ability
to effectively wash bacteria from the turkey carcass. Scald and picker -f.
water samples taken prior to the start of processing contained relatively
high bacterial concentrations indicating the need for more effective down-
time sanitation programs.
Coliform concentrations in scald water samples raemained at relatively
low levels during processing due primarily to the effects of the high
scalding temperature generally used (Table 3). Coliform concentration
increases obtained in all chill waters during processing indicates the effec-
tive washing action of these systems. The relatively high concentration of
coliform obtained initially from the picker drain water samples again
suggests the need for additional emphasis on down-time sanitation programs.
Mold and yeast in water samples were detected in low concentrations
only (Table 4). Small, but significant increases were obtained in the mold
and yeast content of both chill tank samples and in the picker drain
water samples after five hours of processing. These results again suggest
339
-------
TABLE 1. SAMPLING SCHEDULE FOR EACH SAMPLING DAY
Carcass Sampling Stations
Pre-scald Pre-wash
Post-scald Post-wash
Post-picker Pre-chill
Enter-evisceration Post-chill
Pre-viscera pull Post-drain
Post-viscera removal Pre-shrink
Equipment Sample Location*
Dressing-Evisceration Transfer Chute
Trussing Table
Chill Tank Exit Chute
Post Drain Table
Giblet Belt
Carcass Bagging Belt
Hater Samples Location*
Scald Tank Water
Chill Tank #1 Water
Chill Tank #2 Water
Giblet Chill Water
Picker Drain Water
*Samp1es taken initially and after 5 hours of
continuous processing.
340
-------
CM
• 6
y
a>
•u
u
CO
CD
00
3
Figure 1.
Surface Bacterial Counts as Affected
by Hours of Continuous Processing
(Samples taken Pre-Truss)
4.8
4.6
4.4
4.2
4.0
3.8
3.6
3.4
3.2
12345
Hours of Continuous Processing
347
-------
TABLE 2. MESOPHILIC BACTERIAL CONTENT OF SELECTED WATER SAMPLES
TURKEY PROCESSING*
Water
Sample
Hours of
Continuous Processing,
0 Hours 5 Hours
Scald
Chill Tank #1
Chill Tank #2
Giblet Chill
Rough Picker
3,220
388
129
430
970
1,510,000
31,600
10,200
20,100
149,000
*Number of mesophiles per mini liter.
TABLE 3. COLIFORM CONTENT OF SELECTED WATER SAMPLES TURKEY
PROCESSING*
Water
Sample
Hours of Continuous
0 Hours 5
Processing
Hours
Scald
Chill Tank #1
Chill Tank #2
Giblet Chill
Rough Picker
0.6
1.2
0.6
0.7
55.7
1.5
151.4
59.7
23.9
830.1
^Number of coliform per milliliter.
TABLE 4. MOLD AND YEAST CONTENT OF SELECTED WATER SAMPLES
TURKEY PROCESSING*
Water
Sample
Hours of Continuous
0 Hours 5
Processing
Hours
Scald
Chill Tank #1
Chill Tank #2
Giblet Chill
Rough Picker
0.4
1.7
1.2
12.3
9.5
1.7
93.4
27.4
9.1
47.0
*Number of mold and yeast per milliliter.
342
-------
the effective carcass washing action of these systems. Mold and yeast
concentrations in the scald water samples remained at low levels due
primarily to the adverse effects of the scalding temperatures used. The
relatively low concentration of mold and yeast detected in giblet chill
water after five hours of processing indicates that the giblets do not
serve as a source of these microorganisms.
Scald water samples obtained from the five broiler processing plants
indicated that in most cases clean-up during the overnight "down" period
did not effectively reduce bacterial numbers (14,900 per milliliter imme-
diately post clean-up). Continuous processing for a four hour period
resulted in an increase in total count and a build up of fecal coliform
(Table 5).
Samples taken from both the first and third chill tanks in each of
the plants indicate a relatively effective clean-up operation during the
normal "down" periods (Table 6). Fecal coliform were found in concentra-
tions of less than one per milliliter and total mesophilic bacteria were
found in concentrations of about 400 per ml prior to processing. As
expected concentrations of both fecal coliform and mesophilic organisms
increased during processing. The combined effects of the rinsing action
of the chill waters and the normal counter flow directional system of the
Chill tanks water on bacterial concentrations can be seen by comparing
the first and third tank data following four hours of continuous processing.
Bacterial concentrations of water samples from the first chill tank are
double those obtained from samples taken from the third chill tanks.
EQUIPMENT SAMPLES|!
Analyses of equipment samples taken in turkey processing plants
prior to the start of the day's processing suggest that in almost all
cases microbiological concentrations appear only in limited numbers
(Tables 7, 8, and 9). The primary exception to this generalization is
the dressing-evisceration transfer chute where relatively high con-
centration of mesophiles, mold and yeast were obtained indicating the need
for additional down-line sanitation emphasis in this area. During the five
hour processing period, the concentration of mesophilic bacteria, coliform,
mold and yeast generally increased on surfaces where water was not used
as a rinse or from normal tank overflow. This increase was particularly
noticeable on the dressing-eviscerating transfer chute, the trussing table,
and to a lesser extent on the post drain table. These results again suggest
the need for improvements in equipment design which would permit a contin-
uous surface washing action. Bacterial numbers on the chill tank exit
chute and the giblet belt remained at relatively low levels during the five
hour processing period. This effect can generally be attributed to the
constant flow of water over the equipment surface and/or the low bacterial
concentrations generally detected on the product at these locations.
CARCASS SAMPLES
Swab samples taken from the breast area of broiler carcasses at
selected processing stations were analyzed for both total mesophilic
343
-------
TABLE 5. BROILER PROCESSING SCALD WATER ANALYSIS*
Fecal
Collform Mesophiles
Pre-processing 1.0 14,900
4 hours continuous processing 1310 2,680,000
*Geometric mean - bact/ml
TABLE 6. BROILER PROCESSING CHILL WATER ANALYSIS*
' Fecal
Coliform Mesophiles
Pre-processing
1st chill tank 1.0 394
3rd chill tank 1.0 441
4 hours processing
1st chill tank 857 24,100
3rd chill tank 372 12,300
*Geometric mean - bact/ml"
344
-------
TABLE 7. MESOPHILIC CONCENTRATIONS ON SELECTED EQUIPMENT SURFACES*
Sample
Location
Hours of Continuous Processing
0 Hours 5 Hours
Dressing-Evisceration
Transfer Chute
Trussing Table
Chill Jank Chute
Post Drain Table
Giblet Table
Bagging Belt
264
15
54
62
112
143
1,150
163
15
108
156
164
*Number of mesophiles per
TABLE 8. COLIFORM CONCENTRATION ON SELECTED EQUIPMENT SURFACES*
Sample
Location
Dressing-Evisceration
Transfer Chute
Trussing Table
Chill Tank Chute
Post Drain Table
Giblet Table
Bagging Belt
Hours of Continuous
0 Hours
21
21
12
12
25
21
Process ina
5 Hours
357
483
14
49
19
19
TABLE 9. MOLD AND YEAST CONCENTRATIONS ON SELECTED EQUIPMENT SURFACES*
Sample
Location
Dress i ng-Evi scerati on
Transfer Chute
Trussing Table
Chill Tank Chute
Post Drain Table
Giblet Table
Bagging Belt
Hours of Continuous
0 Hours
73
30
38
29
31
33
Processing
5 Hours
24
96
21
75
30
42
*Number of mold and yeast per 100
345
-------
bacterial concentration and for fecal coliform content. Overall results
indicate a statistically significant reduction in total mesophilic counts
associated with the processing operation (Figure 2). Separation of means
by specific processing function indicates that reduction in bacterial con-
centration in all cases can be attributed to the rinsing effect of water
used in processing (scald, wash, chill). The most significant reduction in
surface counts can be attributed to the washing action of the chill system.
Fecal coliforms were obtained from carcass surface samples at all pro-
cessing stations. Increase in fecal coliform surface counts were associated
primarily with the mechanical pickers in the dressing operation and with the
viscera pulling operation during evisceration. However, the data suggest
that the fecal coliform are only loosely attached to the skin since they
were effectively removed by both the washing and chilling operations.
Analyses of turkey carcass samples obtained from the twelve processing
stations show clearly the total bacterial reduction effects of the scalding,
packing, washing and chilling operations (Figure 3). These data indicate
that microbiological reductions on carcass surfaces are obtained only
at processing stations where water is in constant use either as a spray
or a dip. Total mesophilic bacteria on the turkey carcass were reduced
in the dressing operation as a result of the scalding action, the washing
action of picking and the on-line washers generally found prior to carcass
transfer to the eviscerating line. Bacterial concentrations on the carcass
increased during the eviscerating operation and during the handling of
carcasses normally associated with carcass trussing, draining and packaging.
The on-line washing operation and the liquid chilling operation significantly
reduced bacterial numbers.
Coliform numbers on carcass surfaces were almost eliminated by the
temperatures normally associated with the scalding operation (Figure 4).
However, during picking and evisceration the concentration of surface
coliform increased at a fairly constant rate. The increase can in all
probability be associated with the extensive carcass handling required by
the eviscerating operation. The data suggests, however, that coliform
organisms are only loosely attached to the carcass surface since they
were effectively removed by both the washing and chilling operations.
A slight increase in coliform numbers can generally be attributed to the
handling necessary for carcass trussing. However, the liquid chilling
operation effectively reduced the coliform content to negligible numbers
and only a slight increase was obtained during the draining and bagging
operations. These results suggest that emphasis on individual worker
hygiene programs would be one of the more effective methods for reducing
coliform concentration on the turkey carcasses during evisceration.
Additionally, the need for equipment improvements which would permit more
effective removal of viscera is indicated. As stated earlier, a continuous
equipment surface rinsing for the trussing operation appears warranted.
Mold and yeast on the carcass surfaces were generally detected in
relatively low concentrations (Figure 5). The scalding, washing and
chilling operations all contribute to reductions in saprophytic concen-
trations on the turkey carcass. Operations which require carcass handling
346
-------
5.2
5.0
4.8
4.6
NB *'*
I 4"2
f
I *'°
*M
O
^ 3.8
5 3.6
o
u
•rt
M
a
(XI
o
i-i
«•
3.0
2.8
Figure 2
Mesophlllc and collform bacteria/cm of surface area from
carcasses sampled at each of 11 processing stations -
"working normal"
! •
..200 .
HOD
1000
800
700 el
i
••>
600 I
v*.
500
*«-
c
40U -s
100
200
100
•o
•3
«
*
1
•^
?
u
I
!
s
«s
•
-H
T
!
P<
!
I
-s
i
01
-s
i
S
£
Processing Station
347
-------
Figure 3
Mesophilic Bacteria obtained from turkey carcasses
sampled at 12 processing stations
M
0)
u
o
Cfl
m
14-1
o
o
55
01
-C
CO
00
o
.J
5.6
5.4|
5.2
5.0
4.8
4.6
4.4
4.2
4.0
3.8
3.6
3.4
3.2
3.0
2.8
2.6
2.4
2.2
\
\
\ Mean 4- Std. Dev.
Mean - Std. Dev.\
v,.
Mean
CO
o
CO
0) CO
CO
o
CO
4J
CO
o
PL,
a
I
0)
I
w
CO
•H
01
rem
CO
CO
CO 0>
CO
CO
CO
o
o
I
QJ
o
I
4-1
CO
o
PM
CO •!-(
•a
I
4J
CO
o
CO
01
348
-------
Figure 4
Number of Coliform/cm2 of surface area from turkey carcasses
sampled at 12 processing stations
i
I 4J
(U CO
PX
CO
4J
CO
r-1
i— 1
1-1
O
1
-------
Figure 5
Mold and Yeast obtained from turkey carcasses sampled at
12 processing stations
CO ?
I ? I
CO
£
-------
and in which no water is permitted, generally result in increases in the
concentrations of mold and yeast on the breast surface. Generally, pre-
scald sample analyses indicated that mold and yeast are present in about
equal numbers. The scalding operation effectively reduced the relative
concentration of mold so that during the latter stages of processing,
yeast constituted about 90-95% of the saprophytic population. It should
also be noted that the handling and carcass transport required for draining
and bagging produces a small but significant increase in concentration
of mold and yeast.
Salmonella on the final turkey product was detected at very minor
concentrations—generally about two Salmonella per 1000 cm2 of surface
area (Figure 6). The washing action of the scalding and the picker
operations effectively reduced carcass Salmonella concentration. Only
minor changes in Salmonella numbers were noted during the eviscerating and
packing operations.
4
Calculations to characterize the extent of bacterial 'reductions which
can be associated with specific processing functions indicate a reduction
of 95% of the initial mesophilic organisms is obtained during the overall
processing operation (Table 10 and 11). Almost identical results were
obtained in both the broiler processing and the turkey processing studies.
However, differences were noted in the bacterial reductions associated with
the dressing and eviscerating operations of broilers and turkeys. A
significantly greater percentage of bacterial reduction was obtained in
the turkey dressing operation than in the broiler dressing operation.
This difference can probably be attributed to the much smoother breast
skin on the turkey carcass than on the broiler carcass. This difference
in skin texture would make it much more difficult to remove organisms
from the broiler carcass. However, a significantly greater bacterial reduction
was associated with the evisceration of broilers than with the evisceration
of turkeys. The extra carcass handling of turkeys for the draining and
bagging operations resulted in a relatively large increase in carcass
mesophiles. This post-chill handling effect was not noted in processing
broilers since the post-chill operations in broiler processing are, to a
large extent, automated.
Characterization of the bacterial reduction properties of scalding,
washing and chilling operations, indicate that these three processing
functions contribute significantly to the reduction of bacterial numbers
during processing (Table 12 and 13). When the broiler and turkey data
are combined reductions of 67%, 63% and 85% in surface bacterial numbers
can be attributed to the scalding, washing and chilling operations respectively
(Table 10). It is interesting to note that in poultry processing, bacterial
reductions are obtained only from those operations in which water is in
continuous use.
Chemical analysis of the water effluent from several broiler processing
areas indicate that water discharge from the scalding, washing and chilling
operations are not necessarily the major contributors to the final plant
effluent content of either BOD or COD (Table 15). Chill tank water effluent
35]
-------
Figure 6
Salmonella obtained from turkey carcasses sampled at 12
processing stations
cu
u
CO
CO
4-1
CO
cu
CM
§
o
o
o
CO
1-1
cu
CO
c/>
o
z
32
30
28
26
24|
22
20
18
16
14
12
10
8
6
4
2
0
Mean
I Mean + Std. Error
X1
Mean - Std. Error
T3
r-t
CO
U
CO
cu
-o
T—I
CO
u
CO
4-1
CO
&
cu
u
•H
a
(U
i
0)
u
w
I
cu
a
(U
1-1
CO
fi
CO
CO
a)
CO
CO
3
i
u
CO
0
OH
O
CU
•H
.C
O
I
en
£
to
J-l
-o
I
.s
to
I
OJ
a,
Processing Stations
352
-------
TABLE 10. PROCESSING EFFECTS ON THE BACTERIAL CONTENT OF BROILER CARCASSES*
Operation Pre-Treatment Post-Treatment % Change
Dressing 32,400 10,500 -67.6
Eviscerating 10,500 1,550 -85.2
Overall 32,400 1,550 -95.2
*Number of mesophilic bacteria per cm? of breast surface
TABLE 11. PROCESSING EFFECTS ON THE BACTERIAL CONTENT OF TURKEY CARCASSES*
Operation
Dressing
Eviscerating
Overall
Pre-Treatment
54,300
4,500
54,300
Post-Treatment
4,500
3,240
3,240
V
% Change
-91.7
-28.0
-94.0
^Number of mesophilic bacteria per cm^ of breast surface
TABLE 12. THE EFFECT OF SCALDING, WASHING AND CHILLING ON THE BACTERIAL
CONTENT OF BROILER CARCASSES*
Operation
Scalding
Washing
Chilling
Pre-Treatment
32,400
12,300
11,800
Post-Treatment
15,900
4,790
1,230
% Change
-5.19
-61.1
-89.6
*Number of mesophilic organisms per cm^ of breast surface
353
-------
TABLE 13. THE EFFECT OF SCALDING, WASHING AND CHILLING
ON THE BACTERIAL CONTENT OF TURKEY CARCASSES*
Pre-Treatment Post-Treatment % Change
Scalding
Washing
Chilling
54,300
7,130
3,940
12,100
2,450
847
-77.7
-65.6
-79.5
*Number of mesophilic organisms per cm? of breast surface
TABLE 14. THE EFFECT OF SCALDING, WASHING AND CHILLING ON
THE BACTERIAL CONTENT OF POULTRY CARCASSES*
(BROILER AND TURKEY DATA COMBINED)
Pre-Treatment Post-Treatment % Change
Scalding
Washing
Chilling
41,900
9,360
6,800
13,900
3,430
1,020
-66.8
-63.3
-85.0
*Geometric mean of the number of mesophilic organisms per
CITP of breast surface
TABLE 15. COMPOSITION OF SELECTED EFFLUENTS FROM BROILER PROCESSING
Sample
No.
1
2
3
4
5
6
7
8
9
10
•
Description
1:32 PM Inline Birdwasher Drain
1:35 PM Chill Tank #1
1:37 PM Chill Tank #2
1:40 PM Picker Drain
1:50 PM Bird Washer #1 Picking Room
7:30 AM Scald Tank H?0
1:40 PM Scald Tank Overflow
6:55 AM Combined H?0-Post-Settling
2:30 PM Combined HpO-Post-Settling
9:14 PM Combined H^O-Post-Settling
BOD
COD
(mg/lit) (mg/lit)
206
128
258
644
94
119
321
225
446
239
1557
580
995
2843
435
384
1818
557
1474
731
Total
Suspended
Solids
425
374
256
296
108
*
1003
111
— « •.
---
354
-------
contained about 250, inline birdwasher drain about 200 and the scald water
overflow about 325 mg BOD per liter. Final plant effluent contained
approximately 450 mg BOD per liter indicating another major contributor
to effluent BOD. Our studies indicate that giblet and neck chill water and
the process water are potentially significant contributors to final plant
effluent BOD.
CONCLUSIONS
1. The scalding, washing and chilling operations are the most effective
processing functions as measured by overall reduction in surface micro-
biological concentrations.
2. Scalding and chilling continue to be effective carcass surface cleaning
'systems even though bacterial concentrations in each system may increase
substantially during processing.
3. Water use in poultry processing appears to be the only effective means
presently being used to reduce carcass surface bacterial concentration.
4. The scalding operation effectively reduces the number of Salmonella
on poultry carcasses.
5. A 95% reduction in bacterial numbers on the poultry carcass is effected
by current methods of processing poultry.
6. Scalding reduces bacterial numbers by 67%, washing reduces bacterial
numbers by 63% and liquid chilling reduces bacterial numbers by 85%.
7. Scalding, washing and chilling effluents are not the major contributors
to total plant effluent BOD.
8. The results of these studies indicate that changes in equipment design
and in methodology which may result from regulation and/or pollution
concerns for less water use should also be thoroughly characterized
with respect to their effects on the microbiological properties
of poultry products.
355
-------
TREATMENT OF PACKINGHOUSE
WASTEWATER BY SAND FILTRATION
by
Dr. M. L. Rowe*
INTRODUCTION
The technology to be used for the purpose of meeting wastewater discharge require-
ments is an item of concern to the managers of processing plants for red meat
and poultry. Previous papers by Mr. Jack Witherow and Dr. Anthony Tarquin have
devoted attention to some methods of treatment available to the meat and poultry
industry. However, there is a need for a variety of proven treatment methods
which can be reveiwed by the plant manager, as each plant manager must consider
cost, volume and concentration of wastewater, land requirements, climatic condi-
tions, etc. in order that the best available method can be selected for each
individual plant.
Funds for the planning, construction and evaluation of this intermittent sand
filter project were provided by Environmental Protection Agency Grant Number
803766 and the facilities of the W. E. Reeves Packing Company in Ada, Oklahoma
were used for the demonstration site. The intermittent sand filter treatment
facility discussed in this report is in the early stages of investigation and
is not intended to be used as design criteria until further evaluation shows
that it can be a successful and feasible method of treating packinghouse waste-
waters . A final report will be prepared and distributed upon completion of the
research project.
HISTORY OF SAND FILTRATION
The use of sand filtration for the treatment of water and wastewater is not a
recent innovation in the United States. A survey of the literature reveals the
existence of sand filtration for the improvement of drinking water supplies in
the United States as early as 1828 (1). Much of its use and further development
immediately following the introduction of the sand filter took place in Europe.
*Director, School of Environmental Science, East Central Oklahoma State
University, Ada, Oklahoma.
**This investigation was supported by funds from the Environmental Protection
Agency, Food and Wood Products Branch, Corvallis, Oregon, under Grant Number
803766.
356
-------
However, the accelerated growth rate in the United States, especially the
eastern United States, created a demand for larger volumes of drinking water
and around the turn of the century, a number of slow sand filter units were in
use in the United States for the treatment of drinking water supplies.
Just as the accelerated growth rate in the United States had created a demand
for methods of treating larger volumes of drinking water, the need arose for
methods of treating the increasing volumes of wastewater produced by the munic-
ipalities. An experimental intermittent sand filter unit for the treatment of
domestic wastewater was built in Lawrence, Massachusetts in 1888 (2). The
operation of the intermittent sand filter unit proved successful. However, a
rapid increase in the number of sand filter units for the treatment of sewage
was not experienced in the United States until the 1940's. The limiting factors
for the increased usage of intermittent sand filter units were apparently the
availability of natural sand sources meeting the desired specifications and
the requirement of large tracts of land.
Following World War II, the rapidly increasing number of subdivisions, mobile
home parks, and resort areas in Florida created a need for economical and
practical treatment systems which would produce an effluent of suitable quality.
This need for simple and economical sewage treatment units for small volumes of
wastewater led investigators at the University of Florida to test various designs
for intermittent sand filter units (2). Much of the present knowledge concerning
intermittent sand filters has come from these early studies at the University of
Florida.
The early designs of intermittent sand filters have seen very little change over
the years that they have been in use. The units usually consist of an under-
drain bed of tile and gravel which is covered with a filter media of sand.
Designs are usually determined by sand specifications, volume of wastewater,
and the nature of the wastewater to be filtered. In the design of an inter-
mittent sand filter unit, emphasis must be placed on sand specifications and
suspended solids concentration of the influent. These factors must be considered
in order to design a filter unit capable of experiencing a feasible period of
operation before clogging of the filter media occurs.
In the past few years, workers have demonstrated the effectiveness of inter-
mittent sand filters in reducing the suspended solids values of domestic waste
which has received prior treatment in lagoon systems. Evidence of the effective-
ness of intermittent sand filters for the reduction of suspended solids can be
found in the published works of Reynolds (3), Marshall (4), and Walter (5).
Other supportive evidence for intermittent sand filters as a means of lowering
suspended solids values can be found in reports by Grantham (6) and Furman (7).
PILOT SCALE OPERATION
A review of the work by the authors cited above led Witherow and Rowe to
consider the use of intermittent sand filters as a means of polishing the waste-
waters from the lagoons at the W. E. Reeves Packing Company in Ada, Oklahoma.
Other investigations were being conducted at the treatment facilities of the
Reeves' plant at that time and the site served as a suitable place for the
operation of two pilot scale intermittent sand filter units.
357
-------
Two pilot scale units were constructed in close proximity to the lagoon system
at the Reeves' facility. Each unit was formed by welding two 55-gallon barrels
end-to-end. Effluent lines were installed in the bottom of the units and each
unit was filled with gravel and sand to serve as the filter media.
The pilot scale filter units contained 18 inches of gravel, ranging from 0.25
to 1.25 inches in diameter, with the larger gravel particles in the bottom of
the unit. The gravel was then covered with 36 inches of sand which has an
effective diameter of 0.2 millimeters (mm). Investigations were then conducted
by applying wastewater from the extended aeration lagoon and the secondary stage
lagoon to the filter units at a rate equivalent to 0.5 million gallons per acre
per day (mgad).
Analysis of the influent and effluent samples collected during the pilot scale
study revealed a 5-day Biochemical Oxygen Demand (BOD5> removal of approximately
70 per cent and a suspended solids reduction of approximately 50 per cent. The
concentrations of BODs and suspended solids in the effluent from the pilot units
were converted to lbs/1000 Ibs of Live Weight Kill (LWK) and compared to the
Best Practical Treatment (BPT) and Best Available Treatment (BAT) guidelines for
small meat processing plants. The results showed that the effluent from the
filter units could meet the suspended solids limits for BPT and BAT guidelines.
All BODs samples from the pilot units were below the BPT values and only one
was slightly in excess of the BAT guidelines. The results of the pilot scale
study led to the development of the large scale project which is currently being
conducted at the W. E. Reeves' facilities.
PRELIMINARY DESIGN
Since the treatment facilities at the W. E. Reeves' plant had been used for
previous investigations by Mr. Jack Witherow and the author, information per-
taining to the daily volume of wastewater from the plant and the concentration
of various pollutants in the wastewater was readily available to the investigator.
The investigator also had access to information pertaining to the average
suspended solids concentration and average BODs values of the effluent from the
extended aeration lagoon and the secondary stage lagoon in use at the treatment
facilities. Other information pertaining to sand and gravel specifications,
recommended loading rates onto the filters, filter design, etc. were available
from the literature and the pilot study previously conducted by the investigator.
All of this information was used by the investigator in developing the preliminary
design of the project which is now in progress.
The preliminary design for the filter project was based on an average wastewater
flow of 18,000 gallons per day with a maximum daily flow of 30,000 gallons, as
these volumes were known to be characteristic of the flow from the W. E. Reeves'
plant. During the development of a preliminary design, a site adjacent to the
existing lagoon system was selected for the proposed sand filter units so that
loading from either the extended aeration lagoon or the secondary stage lagoon
would be possible by gravity flow. The proposed site for the sand filter units
also had the advantage of providing a means of diverting the flow from the
filters to the existing lagoons in case of an accidental spill or necessary
filter repairs.
358
-------
During the planning of the project, it was decided that instead of one filter
unit, two small units of unequal size should be constructed. This would offer
the advantage of having two different surface areas for the purpose of investi-
gating different hydraulic loadings and would have the added advantage of
providing one operable filter to be used while the second was being cleaned.
A grant proposal containing the results of the pilot investigations conducted
by the investigator and the proposed preliminary designs were submitted to
the Environmental Protection Agency for review. The review of the proposed
project was favorable and the project was funded. After the approval of the
grant proposal, the development of a detailed design which consisted of drawings,
survey reports, and material specifications was completed.
CONSTRUCTION
The final plans called for the construction of two filter units. The filter
units were to be formed by clay embankments on three sides, and the fourth side
of each filter unit would be formed by a common concrete wall between the two
units. The purpose of the common wall was to reduce the amount of land required
for the filters. Thus, the construction would be similar to that required for
the building of a lagoon. Each unit would consist of 36 inches of sand over
18 inches of gravel and each filter unit would have a separate underdrain system.
Specifications called for the bottom of the filter units to be formed of com-
pacting six inches of clay. Plans also specified that the bottom of each filter
was to be sloped toward the effluent drain to insure proper drainage from the
filters. A slope of 6 per cent was recommended.
The embankments were to be constructed of clay soil and were to be built so that
the interior of the dikes were sloped at a ratio of 2.5 horizontal units to 1.0
vertical units. The elevation from the bottom of each filter unit to the top
of the dike was to be a minimum of 9.5 feet, thus allowing a freeboard of at
least 5 feet above the surface of the sand. Specifications called for the tops
of the embankments to be at least 8 feet wide so that vehicles and machinery
could be used at the treatment site. Further design information is given in
Figure 1.
2.5:1 slope
wall
9.5'
14' 28'
Figure 1. Cross-Section of Filter Unit
359
-------
A factor which determined the dimensions of the bottom of the sand filter
units was the space required for the normal operation of heavy equipment, such
as a dozer, during the construction phase of the project. The final specifica-
tions for the bottom of the structure called for a width of 12 feet and a length
of 42 feet. After the bottom was compacted, the 6 inch concrete wall was to be
poured to form two filter units with bottom dimensions of 12 feet by 14 feet and
12 feet by 28 feet respectively.
!
By following the recommendations for the bottom dimensions, slope of interior
dike, and depth of sand and gravel, the completed structure would consist of
two filter units with sand surface areas of 880 and 1220 square feet. If an
average wastewater flow of 18,000 gallons were loaded to the small sand filter,
the hydraulic loading would be approximately 0.90 mgad, and if one-half of an
average days flow were loaded onto the small filter, the resulting hydraulic
load would be approximately 0.45 mgad. Corresponding loading rates of 0.64 and
0.32 mgad could be examined by using the larger filter.
The underdrain network of each filter was to consist of a series of 5-inch
diameter perforated pipes on the bottom of the filter bed. Spacing of the
pipes was to be approximately 3 feet, and each lateral line was to be connected
to a 6-inch diameter pipe serving as the main drain from the filter unit. The
main drain line was to project through the walls of the dike and empty into a
sample collection box. Effluents from the filters would be discharged from the
collection box to the stream that runs through the Reeves' property. Figure 2
shows the layout of the drain system of both filters.
-6" perforated line
•5" perforated line
r
i
fcM» T..-I_
1
1 — _
1
1-
L_J 1
' J~3'~ "
§L
1- 1
1
1 — 1 — __ —
X
collection box
s 6" line to stream
Figure 2. Underdrain System
360
-------
The lateral lines of both filter units were to be covered with 18 inches of
gravel with diameter size ranging from 0.5 to 1.5 inches. Specification called
for the gravel to be covered with 36 inches of washed sand. Sand specifications
called for a sand source having an effective diameter of 0.2 mm and a uniformity
coefficient of 3. A cross-section of the filter units is presented in Figure 3.
8'
26'
'>v---^
>^e —
36'
o
^^"
^ 18"
36" sand
ravel
Figure 3. Cross Section
The filters were to be located so that the effluent from the extended aeration
lagoon or the secondary stage lagoon could be loaded onto the filter units by
gravity flow. However, the construction plans only required a system for
loading the effluent from the extended aeration lagoon to the filter units. The
plan was to add a system for loading the effluent from the secondary stage
lagoon to the filters at a later date, only if the filter were found to be
incapable of handling the wastewater from the extended aeration system. The
location of the filter units is shown in Figure 4.
Manhole
I | from plant
extended
aeration
lagoon
manhole
--- ["""I
03
•u
•H
§
M
-------
A manhole already existed between the extended aeration lagoon and the
secondary stage lagoon and the plans called for utilization of the existing
manhole. The plans called for the installation of a six inch line from this
manhole to a distribution box to be constructed in close proximity to the sand
filters. A six inch distribution line could then be installed from the distri-
bution box to each of the filters. By using this scheme, the wastewater from
the extended aeration lagoon could be discharged to the existing secondary
stage lagoon or to the proposed distribution box for the filters. If the waste-
water from the extended aeration lagoon was discharged to the distribution box,
the wastewater could be loaded onto only one or both of the filters.
Construction of the filter units was contracted by the W. E. Reeves' Packing
Company and the project has been completed, except for the bermuda grass
coverage of the dikes. Construction began in October and the project was
completed in approximately 60 days. The construction time was longer than expected,
due in part to unfavorable weather conditions. Also, the site which was
selected for the sand filters required the removal of several feet of rock in
order to reach the desired depth for the filter units. This required extensive
blasting followed by time-consuming rock moving operations. A final inspection
by the principal investigator revealed that the construction phase of the project
had been completed and that the building specifications were met.
COST
The total contract price for the construction of the sand filter units was
$12,850. This included the construction of the two filter units, the distribu-
tion system, and the effluent collection system. This price also included the
cost of providing a bermuda grass cover of the dike, which has not been completed
at the present time.
According to the contractor, the units could have been constructed for less
money if extensive blasting operations had not been required. Also, all earth
for the construction of the filters had to be purchased and hauled to the
construction site, since soil of the desired specifications was not available
at the construction site. These figures do not include the cost of the extended
aeration system, but cost figures relating to this system are given in the report
by Mr. Jack Witherow.
OPERATION
While the intermittent sand filter units were being constructed, four pilot
scale filters were operated in the laboratory of the School of Environmental
Science, East Central Oklahoma State University. These pilot scale units were
constructed of 6 inch plastic pipe and the units contained 18 inches of gravel
and 36 inches of sand of the same specifications required for the field filter
units. The pilot scale units were operated in the laboratory for three weeks
at hydraulic loading rates of 0.2 mgad. The influent to the pilot units was
settled wastewater from the extended aeration lagoon with an average suspended
solids concentration of 76 mg/1. The effluents from the filters had average
suspended solids concentration of 26 mg/1. Clogging of any of the filters did
not occur during the three weeks of operation.
362
-------
The construction of the sand filters at the demonstration site was completed
in early December, and operation of the filters was scheduled to begin at that
time. However, mechanical difficulties had been experienced with the aerator
in the extended aeration pond and the aerator was removed from the pond for
major repairs. After the aerator was returned to the extended aeration lagoon,
several weeks of aeration were required before the principal investigator felt
that the wastewater in the extended aeration system was suitable for use in the
sand filter investigation.
In the latter part of January, the operation of the sand filters began. At that
time, electrical outlets had not been installed at the manhole between the
extended aeration lagoon and the distribution box for the sand filters and
therefore, an automatic sampler could not be used at the site for the collection
of a composite sample. The principal investigator felt that loading the filters
at this time would be beneficial to the project even though composite sampling
wasn't possible at that time. The only index, therefore, to the quality of the
wastewater loaded onto the filters was obtained by examining a grab sample of
settled wastewater from the extended aeration lagoon.
A 7-day clock timer was used in conjunction with the aerator, so that operation
of the aerator would cease at midnight. Another 7-day clock timer was used to
control the hydraulic valve at the manhole between the extended aeration lagoon
and the sand filters. This system was designed so that the valve would open at
2:00 a.m. and remain open until 6:00 a.m. This scheme allowed a two-hour settling
period after the shut-off of the aerator and a loading time onto the filters of
4 hours. This early morning loading was selected in an attempt to reduce the
length of time that the wastewater on the filters would be exposed to the sun-
light.
The two filters were loaded simultaneously. It was obvious that the filters
were beginning to clog after only 2 days of operation because the wastewaters
added on one day was requiring more than 24 hours before filtering through the
system. Therefore, the operation of the filters was stopped after only four
days. By the fourth day several feet of water had accumulated on the filter
surface and damage of the interior dikes due to erosion had occurred. The grab
samples of settled wastewater from the extended aeration pond indicated suspended
solids values of only 66 mg/1. The suspended solids concentration of the effluent
from the filters were less than 20 mg/1 and turbidity values of less than 10
JTU's were observed. After the loading of the filters had been stopped, a
period of six days was required before the filter had drained and was dry
enough to clean.
An examination of the filters after the surfaces were dry revealed an accumulation
of slude approximately 0.5 inches thick. The sludge layer was removed and the
sand surfaces were cleaned and raked. Visual examination indicated that the
sludge had not penetrated the sand more than two inches. The small number of
days of operation before clogging of the filters and the long drying time in-
dicated that the suspended solids concentration loaded to the filter must have
been much greater than the analyses of the settled lagoon sampler indicated.
This belief was strengthened by examining the pilot study data.
363
-------
It was learned that sludge had not been wasted from the extended aeration
system for several months. This led the investigator to believe that the
sludge concentration in the extended aeration lagoon was too high and that
settling was not occurring properly. The theory was proposed that the suspended
solids concentration in the effluent from the pond was increasing as the
effluent was discharged from the lagoon. Therefore, 16,000 gallons of sludge
were removed from the pond and applied to the fields of the W. E. Reeves' Company.
It was hoped that the settling of the solids in the extended aeration pond
would be improved by the reduction of the sludge.
After the 16,000 gallons of sludge had been removed, two weeks of aeration
of the lagoon were allowed before placing the filters in operation again. Then
in mid February, loading of only the small filter was begun. The same settling
and loading times as those used in the first investigation were used during
this phase of the study. The small filter was clogged after two days of opera-
tion. Loading was then begun on the large filter. This filter appeared to be
clogged after four days of operation. The sludge accumulation to the sand
surfaces during this phase was similar to that experienced in the first attempt
at using the filters.
During this phase of the investigation, composite samples were collected at the
manhole between the extended aeration lagoon and the filters. The samples were
composited during the four hours that the filters were loaded. Analysis of
these samples revealed suspended solids concentration of more than 200 mg/1.
These values were much higher than had been expected and were also higher than
those seen at this same sample point during previous investigations conducted
at the Reeves' site.
These results led the investigator to re-examine all components of the electrical
system controlling the aerator and the hydraulic valve. This investigation
revealed that malfunctions had occurred with the timing system and that the
aerator had been in operation part of the time during the four hour loading
period, thus agitating the sludge and increasing the amount of sludge loaded
onto the filters.
These malfunctions were corrected and it was decided that a longer settling
period in the pond would be advantageous. Loading of the filters resumed on
March 23 > 1976, and the filter units appear to be functioning in a satisfactory
manner.
The timer system for the aerator has been adjusted to shut off the aerator
at 11:00 p.m. and to remain off until 9:30 a.m. on the following morning. The
timer for the hydraulic valve was changed to open the valve at 5:00 a.m. and
close at 9:00 a.m. This schedule allows a settling time of 6 hours and insures
that the aerator does not resume operation until after loading of the filters
has been completed.
Composite samples taken at the manhole between the extended aeration lagoon
and the filter indicate that the influents to the filters have contained average
suspended solids concentration and BOD5 values of 65 and 37 respectively.
364
-------
Effluents from the large filter have shown average BOD5 and suspended solids
values of 8 and 13 mg/1. The average BOD5 and suspended solids values of the
effluents from the small filter have been 8 and 12 mg/1 respectively. These
values represent a 5-day period of operation for each filter.
A visual examination of the sand surface after this short period of operation
of both filters has shown very little accumulation of sludge. Also, the period
of time required for all the wastewater applied to the surface beds has been
small. All wastewater has been dissipated from the sand surface in less than
3 hours after the last wastewater is applied to the filters.
SUMMARY AND CONCLUSIONS
An intermittent sand filter for the treatment of packinghouse wastewater is
being evaluated at W. E. Reeves' Packing Plant in Ada, Oklahoma. The wastewater
from the plant is treated in an extended aeration lagoon, settled, and then
discharged to an intermittent sand filtration unit. A number of difficulties
have been encountered with the demonstration project, but current operation
appears to be successful. Average BOD5 and suspended solids concentrations in
the effluents from the filter units have been 8 and 12 respectively.
Although performance of the filter appears to be acceptable, the study is still
in the early stage of investigation and a much longer study period is required
before definite conclusions can be made. The project plan will require the
operation of the filter unit through warm and cold weather conditions with a
comprehensive analysis of raw wastewater from the plant, the effluent from the
extended aeration lagoon, and the effluent from the sand filter.
Various hydraulic loading rates will be examined and the length of filter-run
before clogging will be determined. The ability of the filter system to meet
effluent guidelines will be determined after the study is completed. Upon the
completion of the study, a final report relating to construction, cost, and
operation will be prepared.
365
-------
1. DANIELS, F. E. Operation of Intermittent Sand Filters, Sewage Works
Journal 17: 1001-1006 (1945).
2. REYNOLDS, J. H., HARRIS, S. E., HILL, D. W., FILIP, D. S., and MIDDLEBROOKS,
E. J. Presented at EPA Technology Transfer Seminar on Wastewater Lagoons,
Boise, Idaho, November 19-20, 1974.
3. REYNOLDS, J. H., HARRIS, S. E., HILL, D. W., FILIP, D. S., and MIDDLEBROOKS,
E. J. Intermittent Sand Filtration for Upgrading Waste Stabilization Ponds,
Presented at Water Resource Symposium Number Nine, University of Texas,
Austin, July 22-24, 1975.
4. MARSHALL, G. R., and MIDDLEBROOKS, E. J. Intermittent Sand Filtration to
Upgrade Existing Wastewater Treatment Facilities, PRJEW 115-2, Utah Water
Research Laboratory, College of Engineering, Logan, Utah, February, 1974.
•
5. WALTER, C. M. Progress Report, Blue Springs Lagoon Study, Blue Springs,
Missouri. Prepared for presentation at the Symposium on Upgrading Waste-
water Stabilization Ponds to meet Discharge Standards, Sponsored by EPA
and Utah State University, Logan, Utah, August 21-23, 1974.
6. FRANTHAM, G. R., EMERSON, D. L., and HENRY, A. K. Intermittent Sand Filter
Studies, Sewage Works Journal 21: 1002-1014 (1949).
7. FURMAN, T., CALAWAY, W. T., and GRANTHAM, G. R. Intermittent Sand Filters.
Multiple Loadings, Sewage and Industrial Wastes 27: 261-275 (1955).
366
-------
WASTE TREATMENT FOR SMALL MEAT AND POULTRY PLANTS
by
Jack L. Witherow*
INTRODUCTION
The technology herein is presented to help meet the wastewater treatment require-
ments of small plants slaughtering and processing cattle or chickens. For
the convenience of plant managers, abbreviations of scientific terms are
defined in the text and in the appendix.
Solving a wastewater problem can best be undertaken in several interrelated
but distinctly separate steps. These are preliminary design, funding, de-
tailed design, construction, and operation. Each step will involve different
technologies, different individuals, different decisions, and result in different
outputs. None of the steps can be skipped, since the results of one are the
basis for the next. Actually, these five steps simplify a long but necessary
procedure to solve a problem at minimum cost. By grouping those tasks which
lead to a recognizable and defined output, the chances of a costly mistake
or omission are reduced.
PRELIMINARY DESIGN
The preliminary design is the most important of the five steps in solving
the problem and minimizing cost. The output of the preliminary design is the
selection of the in-plant controls, the treatment processes, and the site.
These are obviously major decisions and a number of physical facts must
be gathered and requirements satisfied prior to making these selections.
Not only must these selections be based on physical and regulatory constraints,
but also on cost. Minimizing cost is the major function of preliminary design.
Discharge Regulations
The wastewater discharge limitations are the obvious initiation point.
Governmental regulatory groups will be involved. If the plant's discharge goes
into a sewer, the municipal government will have restrictions and sewer charges.
Meeting their requirements will be considerably different than meeting discharge
requirements by the state or federal government under a National Pollutant
Discharge Elimination System (NPDES) permit. However, the treatment processes
described herein can apply to either discharge situation.
* Food and Wood Products Branch, Corvallis Field Station, Industrial Environ-
mental Research Laboratory-Cincinnati, U.S. Environmental Protection Agency
367
-------
Municipal restrictions, that are of concern to meat and poultry plants, usually
involve oil and grease concentration,, feathers, viscera, paunch manure, bone,
hair, hides and fleshings. These restrictions are to prevent obstruction
in the sewer system or treatment plant. Most small processors meet these re-
strictions by in-plant control and pretreatment with screens and grease traps.
All but a few municipalities do, or will, have sewer charges based on flow and
concentration of pollutants. The charges for flow may be reduced if the
wastewater discharge is delayed to reach the treatment plant during the night.
Pretreatment can reduce the concentration and thereby the sewer charge. The
installation of an anaerobic or aerated lagoon with a controlled discharge may
pay a high return in terms of reduced sewer charges.
Plants that discharge to surface water are required to meet the limitations
in their NPDES permit. Limitations in these permits are in pounds per day
for 5-day Biochemical Oxygen Demand (BOD,-) and Total Suspended Solids (TSS).
Other limits are pH not less than 6.0 ana not greater than 9.0 standard units
and fecal coliforms equal to or less than 400 most probable number per
100 milliliter (mpn/100 ml). Ammonia (NH ~) and fats, oils or greases (FOG)
discharges are restricted in either Ibs/day or milligrams per liter (mg/1).
The two bases for discharge limitations are receiving water quality and available
technology. Water quality limitations are used where the discharge would be more
restricted than with technology limits, but only where necessary to protect
the established uses of the watercourse. Discharge limitations based on
water quality vary depending on location of the plant. Technology limita-
tions are based on National Effluent Guidelines and vary with plant size
measured in terms of live weight killed (LWK). The National Effluent
Guidelines have limitations to be met by July 1, 1977, limitation for
new sources of discharge, and more strict limitations to be met by July 1,
1983. These limitations are further categorized by types of plants within
the meat and poultry industries. Table 1 shows the limitations that would
apply to almost all small plants. These limits are averages of daily
values within 30 consecutive days. Maximum limits for any one day are twice
those shown in Table 1. The 1977 limitations are based on Best Practical Control
Technology currently available (BPT) and the 1983 limitations are based on Best
Available Technology economically achievable (BAT).
There is also a 1983 ammonia as nitrogen (NH3-N) limitation of 4.0 mg/1 for
30-day average value and 8.0 mg/1 maximum for any one day. A recent
court decision has declared that these limitations are insufficiently supported
and a new study to establish an ammonia limit is underway. Future national
limitations on ammonia will likely require treatment processes that discharge
low concentrations of ammonia. Several States require discharges to have
2.0 mg/1 or less of NH.-N. In preliminary design, the treatment system should
be planned to incorporate processes by 1983 which discharge a low level of ammonia.
The effluent limitations shown in Table 1 for poultry and fowl processors
were proposed on April 24, 1975. Final guidelines have not been published
but are expected in the near future.
368
-------
Table 1. National Effluent Guidelines
(lbs/1000 Ibs LWK)
Parameters
Slaughter-
houses
Packing-
houses
Chicken
Processor
Fowl
Processor
1977 Limitations (BPT)
BOD .12 .17 .46 .61
TSS° .20 .24 .62 . .72
FOG .12 .08 .20 .15
New Source Limitations (include the above plus NH,-N)
NH.-N .17 ?24 .20 .15
1983 Limitations (BAT)
BOD, .03 .04 .30 .23
TSS° .05 .06 .34 .27
FOG 10* 10* .20 .07
*Value for both maximum day and 30-day average in mg/1.
Monitoring Regulations
NPDES permits and some municipalities have monitoring requirements. Measure-
ments are required of the physical, chemical, and sometimes the biological
characteristics of the waste discharged. These measurements can cost as much
as $100 for each sample depending on the number of analyses required. The
frequency of sample collections and the number of analyses required should
be reviewed as carefully as the discharge limitations.
Municipalities often base their sewer charge on flow, BOD5, TSS, and FOG.
To calculate the charge, measuremeats on these parameters are usually made
prior to each billing. Municipalities frequently have the measurements made
by their personnel, but plant managers often have the results verified
if unusually high charges are levied. Flow can be checked by water meter
readings, or a wastewater flume and gauge can be installed. If a portion of
the sample collected by the city is sent to a commercial laboratory for
verification of BODr, TSS, and FOG, a charge of about $50 can be expected.
The charge for sample collection, preservation, and transportation will be
in proportion to the manhours required and the shipping distance. Laboratories
can be expected to report up to 20 percent difference in concentration. The
usual cause of increased concentrations of these pollutants is unnecessary
material losses when the sample was collected.
The requirements for monitoring on the NPDES permits vary depending on where
the permit was issued. The permit group at EPA's headquarters has recommended
that for small plants, BODg, TSS, Fecal Coliforms, and flow be monitored once
a month and that oil and grease, orthophosphate, ammonia, pH and temperature be
measured twice a year for small plants with discharges under 50,000 gallons per day
(gal/day) and quarterly for plants with flows between 50,000 and 1,000,000 gal/day.
369
-------
Two economical means of meeting these monitoring requirements are: (1)
Plant personnel would collect, ice and ship the samples to a commercial laboratory
for analysis; they would take measurement of flow, temperature and pH and then
prepare, file and mail the reports. (2) An employee would be trained to do the
chemical analyses and fulfill all the monitoring requirements. The first
option is generally the most economical when monthly monitoring is required with
expenditures per discharge point of about $15000 per year.
Where weekly monitoring of these parameters is required annual expenditures
will be about $5,000 per year for option one; and most managers will select
option two. With option two, there would be costs for analytical equipment,
supplies, and laboratory space, but the major cost would be manpower. Analy-
tical equipment costs would be about $2,000, but an employee would need
to devote part of two days per week to monitor the discharge. Option two
offers the advantages of having better control over the waste treatment system
and also provides analytical facilities to run quality control on the meat products.
The guidelines or regulations used to establish and enforce discharge limitations,
sewer charges, and monitoring requirements are interpreted by regulatory per-
sonnel. All of these items should be thoroughly discussed with these people
to prevent any misunderstandings or unnecessary requirements. Only after a
plant manager fully understands the discharge restrictions, monitoring techniques,
and the relationship of costs, can he begin to select the most economical means
of waste treatment and control.
In-Plant Waste Reduction
Control of in-plant wastes is heralded by industry as the most cost effective
means of reducing a wastewater problem. The saving of material previously
lost to the sewer has two economic benefits: (1) It can be sold as product
or as a by-product, and (2) it reduces the cost of wastewater treatment.
In-plant waste reduction involves changes in water usage, process equipment,
and process operation and may require new equipment to collect, transport or
process the material saved. These items are associated with capital expenditures
so evaluation of the potential savings is necessary.
A determination of the potential for in-plant control should be made in con-
junction with a survey of the wastewater to be treated. Both the volume of
flow and concentration of pollutants can be compared with average and mini-
mum discharge conditions in the industry. This comparison will indicate
the potential for reduction in material losses and treatment costs. For small
plants only limited chemical analyses are practical, but flow measurements
and accompanying reduction in flow can be easily undertaken. Thus water
usage is of special value to these plants and is given in Table 2.
370
-------
Table 2. Typical Wastewater Characteristics
(per 1000 Ibs LWK)
Parameters
Flow Ave.
(gal)-Min.
BOD. Ave.
(IbS)-Mln.
TSS Ave.
(Ibs)-Min.
FOG Ave.
(Ibs)-Min.
Slaughter-
houses
640
160
6.0
1.5
5.6
0.6
2.1
0.2
Packing-
houses
940
240
8.1
2.3
5.9
0.6
3.0
0.8
Chicken
Processors
9.3*
4.2*
9.9
3.3
6.9
0.1
4.2
0.2
Fowl
Processors
12.9*
2.9*
15.2
11.8
10.1
6.1
2.3
0.7
*gal/bird with chickens @3.4 Ibs & fowls 05.1 Ibs.
Treatment costs are proportional to both flow volume and quantity of pollutants
(e.g. BOD5, TSS, and FOG). The costs attributed to flow are between 50 and 80
percent of the total cost. A reduction in water use will reduce the quantity
of pollutants. Excess water use removes body fluids and tissue from the
product and floor scraps thus increasing the amount of pollutants. Therefore,
a reduction of water usage should be stressed. The use of automatic closing
valves, pressure regulators, quick open and close valves, and nozzles are
simple means of reducing water use. Clean water, i.e. air conditioning water,
steam condensate, jacket-cooling water, should be segregated from wastewaters.
Reuse of these waters can reduce the wastewater volume needing treatment.
There are several operations where USDA accepts nonpotable water use, e.g.,
as in feather-flow-away flumes or rendering condensers.
Next to reducing overall water volume, reducing unnecessary water contact with
the product, by-products, and waste materials should be stressed. Typically
this encompasses such things as the use of nozzles to more effectively use
the water to wash the product, the transporting of by-products without a water
carriage, and the dry clean-up and disposal of scraps prior to washing the
equipment and floors.
Blood is the major pollutant in wastewater from poultry and meat slaughtering
plants, and where it enters the sewer can be easily detected by color. The use
of containers and collection troughs have been demonstrated effective in keeping
the blood out of the sewer from the bleeding operation and from those cutting
operations which release large quantities of blood. In beef slaughtering plants,
dry collection and transportation of the paunch contents can reduce the waste
load by some 20 percent. A dry clean-up of the poultry receiving area or the
cattle holding pens prior to washing can reduce the waste load significantly.
Land disposal of these segregated wastes is common for small plants. The
general concept in reducing the quantity of pollutants is to keep waste material
off the floor and out of the sewer.
371
-------
In-plant waste reductions that are built into the equipment usually remain
effective; those that require more labor, such as turning off washing sprays during
breaks and lunch periods, often slip unless direct responsibility is assigned
to an individual.
There are a number of reports describing technology that was successful in
reduction of waste discharges from meat and poultry plants. For the manager
who needs to fully explore in-plant control opportunities, reports available
from EPA or NTIS, 5285 Port Royal Rd.f Springfield, VA 22151, are:
"Water and Waste Management in Poultry Processing," Roy E. Carawan,
William M. Crosswhite, John A. Macon, & Byron K. Hawkins, Environmental Pro-
tection Technology Series, EPA 660/2-74-031, May 1974.
"Characterization and In-Plant Reduction of Wastewaters from Hog
Slaughtering Operations," Paul M. Berthouex, Donald 0. Dencker, David L.
Grothman, & Lawrence J. P. Scully, Proceedings of Seventh National Sym-
posium on Food Processing Wastes, EPA, April 1976.
"Workshop on In-Plant Waste Reduction in the Meat Industry," Jack L.
Witherow and James F. Scaief, Environmental Protection Technology Series,
EPA (In Review).
"In-Process Pollution Abatement-Upgrading Poultry Processing Facilities
to Reduce Pollution," EPA Technology Transfer Seminar Publication, July 1973.
>
"In-Process Modifications and Pretreatment-Upgrading Meatpacking Facilities
to Reduce Pollution," EPA Technology Transfer Seminar Publication, October 1973.
"In-Plant Waste Control," W. H. Miedaner. The National Provisioner,
Vol. 167, No. 8. August 19, 1972. pp. 22-28.
Wastewater Characterization
Characterization of the waste flow is necessary to select the treatment process,
to size these processes, and to estimate cost. Treatment plant design for new
processing plants must be based on intended production and industry wide waste-
water characteristics as shown in Table 2. However, the four fold increase
between minimum and average conditions shown in Table 2 clearly shows the
advantage of a wastewater survey prior to design. Expenditures can be several
times that required without an accurate wastewater survey.
The variations of flow volume and waste concentration from hour to hour and
from day to day are large and sampling should be timed to incorporate typical
conditions. The usual survey samples the wastewater discharge and measures
the flow every hour during a three day period. Though these three days need
not be consecutive, they must be when the amount of slaughtering, processing,
and water use are normal in the plant.
372
-------
Hourly samples are composited over one day in proportion to the flow at the time
the sample was taken. (Thus if flow drops to near zero during the night as
is the case in small plants, samples need not be taken). The samples are kept
refrigerated and analyzed the next day.
The information gathered for the preliminary design of the demonstration
facilities at the W. E. Reeves Packing Co. is given as an example. This small
plant slaughters about 500 cattle and 700 hogs per month and processes
about 30 items. The wastewaters come from the slaughtering area, the meat
processing areas, the lavatory, the hide storage cellar, and the holding pens.
The major waste load is from slaughtering which is a one-shift operation,
Monday through Friday. Meat processing is extended until noon on Saturday.
The hide cellar is emptied and flushed once a month and the holding pens
are dry cleaned and then flushed on Saturday.
Both wastewater flows and concentrations were measured in the survey. Eight
hourly samples were taken and composited during slaughtering/processing on
May 19, 1970, and on June 9, 1970. A grab sample was taken on June 19, 1970.
The samples were analyzed according to Method for Chemical Analysis of Water
and Wastes, EPA 625/6-74-003. The data are tabulated in Table 3.
Table 3. Wastewater Chemical Analyses for Preliminary Design
P-arameter May 19, 1970 June 9. 1970 June 19. 1970
BOD,- 2680 1352 1165
Grease 1823 434
The BOD- and grease data exhibit the large variation found in slaughterhouse
wastes. A BOD concentration of 2000 mg/1 was selected for design purposes.
The temperature of the wastewater was 25° to 28°C. To measure flow, a Parshall
flume with a 3-in. throat was installed with a recording flow meter. The
flow measurements are shown in Table 4. A flow of 15,000 gallons/day was
used for design purposes.
Tnis survey proved to underestimate trie flow and overestimate the BODg.
During the following year, weekly measurements showed an average BOD5 and
flow of 1247 mg/1 and 17,800 gal/day, respectively. Compositing samples
taken every half-hour from 6 a.m. to 6 p.m. (the total time personnel were
in the plant) would have improved the measurement of BOD. Comparing long-
term water meter readings with those when the wastewater volume was measured
would have improved the estimate of flow.
373
-------
Table 4. Wastewater Flow Analyses for Preliminary Design
Date Flow (gallons) Operating Period (hrs)
6/18/70 Thursday 19,152 8
6/19/70 Friday 15,072 8
6/20/70 Saturday 2,495 3
6/22/70 Monday 14,040 9
6/23/70 Tuesday 15,498 9
6/24/70 Wednesday 14,274 10
Total 1 week 80,531 gallons 47 hours.
Only the parameters used in design need to be measured. Flow and BODg are used in
design of all biological treatment processes. Grease measurements ari important
to determine the need for a grease trap. There are other parameters that are
needed for specific treatment processes. The most important of these are sulfate
(if an anaerobic lagoon is being considered) and sodium, chloride, and total
nitrogen if a land disposal system is being considered. The source of the sulfate
is the water supply and the concentration can be obtained from the municipality
or by analysis of the well water. Sodium, chloride, and nitrogen need to be
measured in the wastewater stream because their major sources are slaughtering
and hide curing.
Site Selection
Selection of the site and selection of the processes to be used in the
treatment system are usually done concurrently. However, for small plants
available land and operational requirements narrow the alternatives. The first
choice is to build or buy a treatment system. Where sufficient land (1 to
20 acres) can be used for a treatment plant site, then the technology reported
herein can be utilized. Where land is not obtainable or where zoning is
restrictive, then purchase of a compact package treatment system can be a
solution. This option is not covered herein, but information can be obtained
from the manufacturers of such plants. If suitable land is available, the
treatment processes which plant personnel can construct are lagoons, sand filters,
and facilities for land application. The physical requirements of these pro-
cesses should be used to judge the suitability of the available sites.
The Reeves plant is located on 64 acres of rolling land, one mile west of the
city limits of Ada, Oklahoma. An unnamed creek flows north through the property.
The prevailing winds are from the south. North of the plant are cattle ranches
with the nearest residence located 1.5 miles from the plant. South and south-
west of the plant the nearest residences are 0.25 mile distant. In addition to the
processing plant, cattle pens, storage areas for machinery and materials,
and parking areas are located on the grounds.
374
-------
An inspection of the 64 acres revealed several potential sites for a treatment
system. The sites adjacent to the plant were used for holding and feed pens,
and storage of materials and equipment. These were eliminated as potential
sites because of relocation costs and priorities of the ongoing uses. The
site selected was a 22 acre pasture and solid waste disposal area which was
300 feet west of the plant. This site permitted a gravity flow system elimi-
nating the necessity for a pumping station. The disadvantages were land slope
and a conglomerated rock layer immediately under the surface. The site
had sufficient land available for selection of lagoons or an irrigation
system. The trade off of land value for the costs of concrete and steel
in more compacted systems was possible, and plant personnel were capable of
constructing a lagoon or irrigation system.
Process Selection
The first consideration in process selection is reduction of the quantity of
pollutants to levels meeting discharge requirements. This includes reduction
of BOD5, suspended solids, and grease if the discharge is to a municipality.
If the discharge is to a surface water, additional limitations exist on pH
and fecal coliforms. Ammonia limits are applied to new sources of discharge
or where water quality standards control and are expected on all discharges
to surface waters by 1983.
Before looking at the capabilities of the numerous treatment processes,
some special needs of small plants need to be considered. These needs reduce
the number of processes to be considered. Because the use of full-time waste
treatment personnel is not practical for small plants, minimum operational
requirements are more important than minimum capital or power costs. For example
an annual cost of about $6,000 was incurred on each of several different treat-
ment systems demonstrated at the Reeve's plant which included about $2,000
for operation and monitoring. These data indicate operation and monitoring costs
to be an extremely high percent of total cost for small plants, and the impor-
tance of minimizing manpower requirements.
Processes to be considered for small plants are limited to those that require
less than eight hours per week for operation and maintenance, that minimize
and automate mechanical equipment, that prevent undetected failures, and that
allow simple construction to reduce the initial capital requirements. These
requirements encompass the anaerobic lagoon, the aerobic lagoon, an unusual
extended aeration lagoon, intermittent sand filtration, and land application
processes. The last two types of processes are described more fully in reports
by Rowe and Tarquin in this publication. The effectiveness of all of these
processes has been established by recently completed or ongoing EPA Research
and Demonstration projects.
Anaerobic Lagoons
The anaerobic lagoon is the treatment process most widely used at meat and
poultry plants. The anaerobic process is especially suited to the concentrated
hot wastewaters from these plants. The process utilizes anaerobic bacteria,
which function in the absence of free oxygen to break down organic wastes.
The lagoon is usually deep and covered with a blanket of floating sludge.
375
-------
The anaerobic lagoon has obtained greater than 90 percent BODg reduction,
with highest removals during the hot summer. The process has the desired
minimum capital and operating costs, is simple to operate and shows visible
treatment results; mechanical equipment is not necessary and the treatment
processes can withstand the shock loadings common to the industry. The anaerobic
lagoon has given high removal efficiency with and without sludge recirculation.
The benefits of sludge recirculation were outweighed by maintenance problems
at the Reeves plant. Elimination of sludge recirculation results in a desirable
reduction in manpower and capital requirements for the small plants.
Two potential problems must be evaluated prior to selection of the anaerobic
lagoon. These are odor emissions and discharges of ammonia. The sulfate in
the water is reduced under anaerobic conditions to hydrogen sulfide, and the
protein in the wastewater is reduced to ammonia. In some cases, a sulfate
concentration of over 200 mg/1 in the water supply has resulted in such objection-
able odors that the process was abandoned. Even without sulfate in the water,
some odors may occur. Location of the facility at least 1/4 mile from a
single home or 1/2 mile from a residential area is suggested.
Ammonia is toxic to fish, and fish kills due to ammonia in meat plant discharges
have been documented. However, reduction of protein to ammonia is a reaction
that cannot be avoided in biological systems. The problem is that the concen-
tration of ammonia in anaerobic lagoon effluents requires the selection of
additional treatment processes that remove the ammonia.
Anaerobic lagoons are designed with a low surface area to volume ratio to conserve
heat and minimize surface reaeration. Depths of 10 feet or more are desirable.
Economic considerations and maintaining several feet above the groundwater
usually limit depths to less than 18 feet. The volume is based on organic
loading and designs range from 12 to 25 pounds of BOD5 per 1000 cu. ft. with
15 Ibs. BODc/1000 cu. ft. being common. The anaerobic lagoon at the Reeves
plant had a water depth of 10 feet and an organic loading of 12 Ibs. BOD^/IOOO
cu. ft. Removal efficiencies for BOD5, TSS, and FOG were 92, 84, and 95apercent,
respectively. The consistency of the effluent (Anaerobic Effluent) concentra-
tion for BODj., Chemical Oxygen Demand (COD), TSS and NH_-N can be seen in
Figure 1. These frequency vs. concentration graphs display the reliability
of the anaerobic pond for treatment of these wastewaters.
As previously stated, the parameters used in determining municipal sewer rates
are generally BOD, TSS and Oil and Grease. Figure 1 shows that the BOD and TSS
will be below 200 mg/1 except on rare occasion. Oil and grease analyses made
on the effluent show the concentration is well below the 100 mg/1 limit commonly
used by municipalities. The high percent removal and consistency of effluent
concentration from the anaerobic lagoon result in an effluent which would meet
common limitations for discharge to a municipal treatment plant.
Discharge of an effluent with high hydrogen sulfide concentrations to a municipal
system will result in damage to concrete sewers and structures unless precautionary
devices are installed. Hydrogen sulfide, with the characteristic rotten egg
odor, can be detected at low concentrations. At the demonstration facilities,
the sulfate concentration in the water supply was 4 mg/1 and a hydrogen sul-
fide odor could not be detected. Other septic odors could only be detected within
50 feet of the anaerobic pond in the downwind direction.
376
-------
GO
*vl
vj
20 30 40 50 60 70 80 90 100
% OF OBSERVATIONS I VALUE SHOWN
FIGURE I - ANAEROBIC LAGOON - FREQUENCY VS. CONCENTRATION
-------
The increase in NH--N is displayed in Figure 1; most of the effluent concen-
trations were between 65 mg/1 and 85 tng/1. The conversion of protein to
ammonia increases the concentration of'ammonia threefold through the anaerobic
pond.
The oil and grease concentrations in the raw waste and effluent averaged
514 mg/1 and 16 mg/1, respectively, which shows limited loss in the packing-
house and high reduction in the anaerobic pond. A grease cover did not form
on the anaerobic pond which indicates the grease was being digested in the
pond. Some consideration was given to mixing grease and straw on the surface
of the lagoon to form a cover to reduce heat loss. This was not done and
temperature reduction through the pond was as high as 60 percent during part
of the winter when water temperature dropped to around 10°C. Insignificant
changes in removal efficiencies were noted during this period. In a colder
climate there are greater advantages for a cover to reduce heat loss and main-
tain biological activity.
The anaerobic lagoon can produce an effluent suitable for discharge to a
municipality, but it will not produce an effluent suitable for discharge into
a surface water without further treatment in an aerobic process. Aerobic
processes are those that maintain dissolved oxygen in the water.
Aerobic Lagoons
Aerobic lagoons frequently called oxidation ponds or stabilization ponds,
have been widely used by the meat and poultry industries. The lagoon is similar
to a shallow lake having a light green color in the water due to algae.
Frequently the discharges from an aerobic lagoon exceed the new more stringent
NPDES limitations. Consequently, the process must be upgraded or replaced.
Upgrading can be accomplished by either operational schemes or subsequent
processes. Both methods are under development and will be mentioned later
in this publication.
The aerobic lagoon has been successfully used in series with an anaerobic or
aerobic process to obtain additional reduction of pollutants, to increase the
dissolved oxygen content sufficiently to maintain fish life in the receiving
waters, to dilute unusual concentrations of pollutants due to upset in the
prior treatment processes, or to temporarily store the wasteswaters. Storage
allows matching the discharge with conditions in the receiving waters or with
weather and crop conditions where the wastewater is applied to the land.
Storage can be used to enable batch chemical treatment for reducing suspended
solids and fecal coliforms or to limit discharges to periods of low alaae
levels in the lagoon.
The disadvantages of the aerobic lagoon are the limited and inconsistent
removal of pollutants, the production of algae (which"increase the suspended
solids concentration and the pH, sometimes to greater than 9.0 for several
months of the year), and the large increase in volume and pounds of pollutant
discharges following rainfall at the site.
378
-------
The design of aerobic lagoons is based on daily pounds of BOD,, per acre
of water surface and a minimum 3 foot water depth to prevent weed growth and the
subsequent production of mosquitoes. Frequently two or more aerobic ponds
are used in series. The generally accepted loadings range from 20 to 40 Ibs
BODg/day/acre which is similar to design of the process for municipal
sewage. However, aerobic operation has been achieved with loadings three
times the above in lagoons treating meat packing wastewaters.
Because of the strength of the wastewater from meat and poultry plants, the
aerobic lagoon must follow prior treatment to meet the design requirements. Thus
the reduction of BOD,, in the prior process must be estimated. State regula-
tory agencies and designers sometimes insist on a conservative estimate to
provide a margin of safety. For example, state regulatory agencies in Illinois,
Iowa, Nebraska, Tennessee, Pennsylvania, and Minnesota allow 60 percent BODj-
removal when the prior treatment is an anaerobic lagoon loaded at 15 Ib.
BODC/1000 cu. ft.
b
The demonstration project was a two phase study. In the first phase, the two
aerobic lagoons were operated in series following an anaerobic lagoon. In the
second phase, the two aerobic lagoons were operated in series following an
extended aeration process. The organic loading of the first and second aerobic
lagoons averaged 52 and 8.5 Ibs. BOD^/acre/day, respectively, when preceded by
an anaerobic lagoon; when preceded by the extended aeration unit the loadings
were 8 and 5 Ibs. BODg/acre/day, respectively. Detention times in the first
and second lagoons wefe 30 and 84 days, respectively.
The consistency and degree of removal of BOD- and TSS for the two aerobic
lagoons in both phases of the project can be seen in Figures 2 and 3.
Following the anaerobic lagoon the first aerobic lagoon reduced the concen-
tration of BODg and TSS by 53 and 25 percent, respectively. The first lagoon
decreased the average discharge of these pollutants sufficiently to meet
30 day average BPT limitations for 1977; however, the maximum day limitations
were exceeded. The opposite occurred during the second phase of the study.
Following extended aeration, the first aerobic lagoon generally increased
the 30 day average concentration of BODg and TSS, but reduced the highest
daily concentrations of TSS from the extended aeration process. This reduction
of TSS resulted in a discharge that meets the maximum day limitation for 1977.
The 1977 limitations shown on Figure 2 are based on average flow and LWK values.
The second aerobic lagoon resulted in minor changes in BOD5 concentration in
both phases of the project (Figure 3). In the first phase, the TSS was signi-
ficantly increased in the second aerobic lagoon. With respect to the National
Effluent Guidelines, the most significant effect was the increased discharge
volume after rainfall which on a few occasions increased the quantity of
pollutants discharged to twice the maximum day limitation for TSS.
National Effluent Guidelines on pH and fecal coliform were established between
the first and second phase of the demonstration project. In the second phase,
the maximum pH limit of 9.0 was exceeded by both aerobic lagoons. The discharge
from the first aerobic lagoon exceeded the pH limitation three out of 45 measure-
ments. These three samples were all taken in April. The discharge from the
second aerobic lagoon exceeded the pH limitation 15 out of 46 measurements.
379
-------
CO
00
CD
250
200
150
in
Q
O
m
Q
O
co
100
50
PHASE I
BODC
10 20 30 40 50 60 70 80 9O IOO
10 20 30 40 50 60 70 80 90 100
100
90
80
70
60
50
40
30
20
10
max. day
1977 limit
PHASE H
BOD.
30 doy ove
1977 limit"
600
5OO
400
3OO
180
160
140
120
<2 IOO
8O
60
40
20
o»
E
PHASE H
TSS
0
IO 20 30 40
50 60 7O 8O 9O IOO 0 10 20 30 40
% OF OBSERVATIONS * VALUE SHOWN
50 60 70 80 90 IOO
FIGURE 2- FIRST AEROBIC LAGOON - FREQUENCY VS. CONCENTRATION
-------
CO
00
25O
200
o>
E
in
Q
O
DO
150
100
in
Q
O
CD
10 20 30 40 50 60 70 80 90 100
% OF OBSERVATIONS I VALUE SHOWN
FIGURE 3 - SECOND AEROBIC LAGOON - FREQUENCY VS. CONCENTRATION
-------
All 15 samples were taken during the warm months, July through October.
There were no discharges with a pH of less than 6.0, the lower guideline
limitation. None of the fecal coliform measurements on the discharges from
the lagoons met the guideline limitation of 400 mpn/100 ml.
The quantity of pollutants discharged from aerobic lagoons have a log normal
distribution. The significance of this is that the maximum discharge can be
5 to 10 times the average discharge; however, the occurance of these levels
of discharge are infrequent and are commonly caused by heavy precipitation.
Reduction of these large discharges to meet maximum day effluent limits can
be obtained by restriction of the volume of flow discharge and temporary storage
in the lagoon. Control of the discharge via temporary storage may be the
most valuable function of the aerobic lagoon. There are simple outlet designs
that will automatically restrict the flow during peak discharges.
Extended Aeration Lagoon
The extended aeration lagoon is a treatment process that was first demonstrated
on the research project at the Reeves plant. The process, an adaptation of
the activated sludge process, was designed to meet the needs of small meat and
poultry plants. The process incorporates in one lagoon the common sequence in the
activated sludge process, i.e., growth of a bacterial floe, mixing and aeration
of the floe and wastewater, separation of the floe, discharge of the clarified
wastewater, and return of the bacterial floe to the aeration basin. In appearance
the process is a small-deep aerated pond.
For small plants the extended aeration process has several advantages over other
variations of the activated sludge process. It produces a higher quality effluent
and smaller volumes of sludge requiring disposal. It has the ability to handle
shock loads and can both_nitrify ammonia (NH3~) to nitrites (N02~) or nitrates
(N03~) and denitrify N02~ and N03~ to nitrogen gas (N2). It has lower manpower
and capital requirements. The disadvantages are the increased power and
equipment cost for aeration.
Meat and poultry slaughtering plants usually have one operating shift and one
cleanup shift with little or no flow during the night. This intermittent flow
is a disadvantage in most process designs. In the extended aeration lagoon
design, the period of no flow was used advantageously. An anaerobic lagoon
was converted to extended aeration by the addition of a floating aerator
and an automated outlet valve. Both the aerator and valve were controlled
by timers. The valve remained closed except from 2 a.m. to 6 a.m. The aerator
was operated from 6 a.m. until midnight; during the period when the aerator
was off, the activated sludge floe settled. The clear supernatent was dis-
charged between 2 a.m. and 6 a.m. This batch Operation enabled both aeration
and settling to occur in one vessel without a mechanical sludge collection
and handling system.
The extended aerated lagoon has been operational for three years at the Reeves
plant. A year of evaluation data shows that the process meets the 1977 limitations
except for the maximum day limits on suspended solids and fecal coliforms.
Disinfection will be necessary to meet the fecal coliform limit. The suspended
solids violations occurred during the first two months of evaluation because
382
-------
of unusual operating conditions which included a new operator, freezing weather,
sludge wasting and excessive flow due to a broken water line in the plant.
The number of analyses (n), mean concentrations of pollutants in the influent
and effluent, and removals by the extended aeration lagoon are given in Table 5.
The removal efficiencies based on mean concentrations of BOD,, COD, FOG,
NH3-N, and Total Kjedhal Nitrogen (TKN) exceeded 90 percent. Removal of TSS
was 96 percent when based on median concentrations. After the first two months
of the study period the TSS in the effluent dropped from around 200 mg/1
to about the median value of 23 mg/1. Removal of Total Phosphate (T-P) was
lowest at 71 percent; however, no T-P was removed when this unit was operated
as an anaerobic lagoon. The maximum discharge of NH3-N was 7.0 mg/1, except
for 4 days when the aerator was off due to vandalism; Both NHL-N and TKN concen-
trations decreased in the unit with only 2 percent of the decrease accounted for
by the increase in NO?+N03-N. Denitrification occurred, most likely during the
6 hours each day when the aerator was off. Within the pH range of 7 to 8 only
20 percent of the ammonia was in a form that could have been lost to the
atmosphere. The odors of ammonia or other objectionable compounds, e.g. hydrogen
sulfide or mercaptans, were not detected.
Table 5. Concentrations and Removals - Extended Aeration
Parameters
BOD,
COD
TSS
FOG
NH--N
NO,+NO,-N
TKN J
T-P
n
42
46
45
10
44
44
46
46
Influent
(mg/1)
714.8
1630.2
535.8
138.6
12.5 '
0.4
79.0
11.0
Effleunt
(mg/1)
17.0
121.6
65.4
11.9
1.9
2.6
7.8
3.3
Removal
(X)
98
93
88
91
95
— «
90
71
The extended aeration unit was organically and hydraulically loaded lower than
usual in this country. The loadings are common for oxidation ditches used
in The Netherlands. The food to microorganism ratio (F/M) averaged 0.06
Ib BOD5/lb Mixed Liquor Volatile Suspended Solids (MLVSS). The mean hydraulic
detention time was 9.8 days and sludge retention time (SRT) averaged 64 days.
Sludge was wasted five times during the 12 month study period. Mixed Liquor Sus-
pended Solids (MLSS) averaged 3350 mg/1 and the Sludge Volume Index (SVI) averaged
217.
A previous study evaluated the extended aeration process at a packinghouse in
Iowa. In that study the F/M averaged 0.26 Ib BOD5/lb MLSS and the detention
averaged 3.6 days. The effluent concentrations for BOD,, TSS, FOG, and NH3-N
averaged 70, 142, 21 and 18 mg/1, respectively. Both tne loadings and the
effluent concentrations were several times greater than in this study. With
those higher loadings the extended aeration processes exceeded .the 30 day
1977 limitations for BOD, and TSS five and seven fold, respectively.
383
-------
Aerated Lagoons
The aerated lagoon is described because the process is widely used in the meat
and poultry industries. The process can achieve average BOD reductions of
50 to 60 percent. The process is similar in appearance to the extended aeration
lagoon previously described but does not have a controlled discharge or
scheduled aeration. Thus the process does not separate the biological floe
from the effluent and is usually followed by another process to capture the
suspended solids which are created. The ability to capture the biological floe
in a several acre aerated lagoon by batch operation, demonstrated for small
extended aeration lagoons, may not be possible. Thus the aerated lagoon has
more potential value for large plants than it does for small plants. The
aerated lagoon has been used for pretreatment prior to discharge to a muni-
cipality or as a process between an anaerobic lagoon and an aerobic lagoon.
Some aerated lagoons are aerobic lagoons upgraded by the addition of aeration
equipment, but the usual design is for a deeper lagoon—between 8 and 15
feet. The increased depth gives more efficient use of the aeration equipment
and reduces energy and aeration equipment costs. However, usually the controlling
design factor for aerators is the requirement for mixing the biological floe
rather than the oxygen demand of the lagoon contents. Detention times ranging
from 2 to 10 days have been used. The process was not evaluated for its
application to small meat or poultry plants at the demonstration plant and
will not be described further.
Intermittent Sand Filtration
The intermittent sand filtration process is used by more than a sozen small meat
plants in Ohio and has had widespread use following aerobic processes treating
domestic sewage from institutions and subdivisions. The process requires
simple and infrequent maintenance which is important to small plants. Intermittent
sand filtration has been used for nearly a hundred years for treatment of
sewage; however, the increasing volumes to be filtered and the rising cost of
land have reduced its use. For large plants, rapid sand filtration preceded by
chemical treatment would be more practical due to the savings in land and manpower,
but rapid sand filtration with chemical treatment would be likely to fail
unless a full-time operator is employed.
The intermittent sand filter has the capability of upgrading discharges from
several treatment systems to meet the 1977 National Effluent Guidelines for
suspended solids. Suspended solids and fecal coliforms are the two limitations
that most of the present treatment systems fail to meet.
The intermittent sand filter resembles a lagoon filled with sand. A layer
of gravel with drain pipes is under the sand to collect the filtered wastewater
and discharge it through the lagoon dikes. The filter is designed on the basis of
hydraulic loading per unit of sand surface area, and on effective size of the sand
particles. Loadings of 0.2 to 1.2 million gallon/day/acre (mgad) and effective
sizes of 0.2 to 0.7 millimeter (mm) have been used. Pilot scale experiments at
the demonstration plant used a loading of 0.5 mgad and sand of 0.2 mm. Removal of
BOD5 and TSS were about 70 and 50 percent, respectively. Further description
of the design and effectiveness of a full-scale filter is presented by Dr. Rowe
in this publication.
384
-------
Land Application
Land application actually encompasses three treatment processes which have
been utilized by food processors for several decades. The use of land
application is expected to increase dramatically by 1983 when Best Available
Technology Economically Achievable guidelines are applied or sooner where
similar limitations are required because of water quality standards.
The three treatment processes are infiltration, irrigation, and overland flow.
Selection of a process is primarily governed by the drainability of the soil
at the available site, which largely determines the hydraulic loading and
the required land area. The comparative characteristics of the three pro-
cesses are presented in Table 6.
Table 6. Characteristics of Land Application Processes
Design Factors
Loading (in/day)
Land area
(acre/mgd)
Drainability,
soil type
Application
Technique
Infiltration
0.6 to 19
Less than 60
Rapid, Sands
to sandy loam
Spray or
surface
Irrigation
0.1 to 0.6
60 to 370
plus buffer
zone
Moderate,
Loamy sand
to clay
Spray or
surface
Overland Flow
0.25 to 1.0
50 to 150 plus
buffer zone
Slow, Clay loam
and clay
Spray
The infiltration process depends upon the applied wastewaters percolating through
the soil and eventually into the groundwater. The process is restricted to
rapidly permeable soils and to wastewaters which will not plug the soil.
Meat and poultry processing wastewaters may require prior treatment to remove
particulates and grease.
The irrigation process involves application of wastewater to maintain or increase
crop production. The objective of most meat and poultry processors is only to
maintain crop production while maximizing hydraulic loading. This operation
requires smaller storage facilities and irrigation area. The soils often
contain clay and the sodium content of the wastewater can cause plugging of
the soil. High chloride in the wastewater can also cause damage to the crop
growth. When using an irrigation process, in-plant process control to reduce
salt loss should be incorporated.
Overland flow is a process in which wastewater is applied at the top of
smoothly-sloped, plant-covered land and collected at the bottom of the slope
for discharge. The process depends mainly upon bacteria on the slope to reduce
385
-------
the pollutants in the wastewater and upon loss of volume due to evaporation.
The concentration of salt in the wastewater may cause the same problems as
in the irrigation process.
Because of the potential health hazard caused by the addition of nitrites (NCL")
to the groundwater, some States put limits on the amount of nitrogen per
acre per year that can be applied to land. The usual concentration of nitrogen
in meat and poultry wastewaters can exceed these limits when applied at the
loadings shown in Table 6. In these States, partial removal of nitrogen from
the wastewater may be necessary prior to land application.
Both the irrigation process and infiltration process can achieve "no dis-
charge to surface waters." Thus not only could these processes meet the 1977
and 1983 National Effluent Guidelines, but there would be no monitoring expen-
ditures unless required by a State agency. The overland flow process produces
a surface discharge and was, therefore, evaluated at the demonstration project.
The process was evaluated on a raw wastewater receiving screening and aerated
storage, on a wastewater treated in an anaerobic lagoon and on a wastewater
treated in an extended aeration lagoon. The overland flow process preceded by
an anaerobic lagoon met the 1977 limitations for BQD5 and TSS but did not
meet the 1983 limits because discharges of NFL-N averaged 20 mg/1. The
overland flow process preceded by either screening or extended aeration met
the 1977 National Effluent Guidelines and could meet the 1983 limits if provi-
sions are made for control of the discharge volume with partial recycle to the
irrigation area during intense rainfalls. Thus the process could be designed to
meet both 1977 and 1983 National Effluent Guidelines. Fecal coliform measure-
ments showed that the discharge from the process needed to be disinfected.
Further description, design and construction consideration of land application
processes will be presented by Dr. Tarquin in this publication.
Disinfection
Disinfection is a process which eliminates microorganisms to prevent the
spread of disease by protecting public water supplies, aquatic life, bathing
beaches, etc. National Effluent Guidelines include a limit on fecal coliform
of 400 mpn/100 ml. Fecal coliforms are an indicator organism for the presence
of bacteria of a human origin. None of the treatment processes evaluated at the
demonstration project meet this limitation. Consequently, the addition of a
disinfection process is necessary where the limitation is imposed.
Treated wastewaters may be disinfected by several chemical or physical processes;
however, chlorination is used almost exclusively. Chlorination of treated
wastewater consists of the addition of 5 to 15 mg/1 of chlorine in a basin
having about 30 minutes detention. The amount of chlorine required varies
with the degree of prior treatment. Of special concern with meat and
poultry wastewaters is the concentration of ammonia in the discharge. Chlorine
combines with ammonia and reduces the disinfection ability of the chlorine.
For every mg/1 of ammonia, several additional mg/1 of chlorine will be needed
to reduce the fecal coliform to the required level. Chlorine requirements are
usually established in a laboratory by measuring the remaining fecal coliforms
or the chlorine residual 30 minutes after addition of the chlorine.
386
-------
Dosages of chlorine are usually kept constant during periods of discharge
for small plants. The cost of extra chlorine does not justify the cost of
automatic equipment to change chlorine dosage with chlorine demand. The three
methods of application for small plants are the addition of chlorine from a
pressure cylinder, addition of hypochlorite with a dry chemical feeder, or addi-
tion of chlorine generated with an electrolytic cell and brine solution. The
chemical feed and electrolytic cell require electrical power and can be automated
to operate in synchronization with controlled releases. For small plants the
most common method is the use of liquid chlorine from 100 to 150 Ib steel containers.
Cost Estimates
Minimization of cost while meeting the discharge, monitoring, and site constraints
is the primary purpose of preliminary design. The selection of treatment pro-
cesses determines not only the capital cost but also the operation, maintenance
and monitoring costs. The selection of processes which substitute land for
building materials, mechanical equipment, operational requirements, and even
monitoring requirements is common with food processors, especially small plants.
More stringent discharge requirements can be expected with time and increased
population. The present National Effluent Guidelines are scheduled to become
more stringent in 1983. Selection of treatment processes and site should be
based on the ability to meet more stringent requirements. There should
be the capability of adding facilities or upgrading processes. For example, the
anaerobic lagoon can be converted to an extended aeration process. The aerobic
lagoon can be converted to an intermittent sand filter or to a storage facility
for application of the wastewater to land.
The nine treatment processes which have been described are especially suited
to the needs of small plants. All of the processes described could be built
utilizing plant personnel and earth shaping equipment. The processes described
can meet the common requirements for discharge to a municipality or the require-
ments under the National Effluent Guidelines for 1977 or 1983. At the demonstration
plant, two treatment systems have been shown capable of meeting each of these
three discharge requirements. For a discharge to a municipal sewer either
the anaerobic lagoon or the extended aeration lagoon could be used. For a
discharge to meet 1977 National Effluent Guidelines either an anaerobic
lagoon plus overland flow and disinfection or an extended aeration lagoon
plus intermittent sand filtration and disinfection could be used. For a
discharge to meet 1983 National Effluent Guidelines either screening and aerobic
storage plus overland flow and disinfection or extended aeration lagoon plus
overland flow and disinfection could be used. Other combinations of these
processes and additional processes, especially land application by irrigation or
infiltration, are also capable of meeting these requirements but are not covered
here.
The major costs involved with constructing the processes described are in de-
creasing order of investment: land, earth movement, the needed materials (pipe
timers, pumps, motors, valves, sand, gravel and concrete) and aeration equipment.
The last two items could be determined accurately regardless of location but
the first two, land and earth movement, are highly dependent on the site selected.
387
-------
Thus cost estimates for selection of a treatment system or for obtaining
funding must be made for an individual site. However, capital costs and
land requirements have been determined on these systems at the Reeves plant
and are shown in Table 7.
Table 7. Costs at Demonstration Site
Facility
Sewer & Manholes
Anaerobic Lagoon
Aerobic Lagoon
Extended Aeration Lagoon
Intermittent Sand Filter
Overland Flow (estimate)
Capital Needed
($)
2,000
6,000
12,000
12,000
13,000
6,000
Land Needed
(acres)
0.1
0.3
1.5
0.3
0.3
5.0
The capital costs in Table 7 do not include cost for land but do show cost
incurred for construction of the full-scale process at the Reeves plant.
Overland flow is the exception and is an estimate based on pilot-plant data
and site conditions at the Reeves plant. Preliminary costs can be estimated
by comparing size or design factors of another plant with the Reeves plant.
At the Reeves plant the wastewater flow averaged 18,000 gal/day and BODg
averaged 185 Ib/day. There was 550 ft. of 6" diameter sewer pipe and tRree
manholes. The maximum centerline height of the dike in the anaerobic lagoon
was 13 feet. The aerobic lagoons hold four months of flow and would be
similar to the cost for aerobic storage needed in northern climates to limit
land application (overland flow) to days when the temperature is greater
than 32°F and there is less than 0.5 inches of rainfall. The extended aeration
lagoon was previously the anaerobic lagoon with an automated 10 Hp aerator and
outlet valve. The intermittent sand filter was designed at 0.5 mgad with 3 ft.
of freeboard to handle heavy rainstorms. The overland flow was based on 5-day
a week application with winter application of 0.25 in/day. The amount of land
needed, «s shown in Table 7, has been increased to include not only that for
the process but also for auxiliary items, i.e. roads, pipelines, pump, control
structures and fences.
The processes shown in Table 7 were selected for minimum maintenance and
operation costs. The anaerobic lagoon and aerobic lagoon had less maintenance
and power costs because of the absence of mechanical equipment (the aerator
for extended aeration and the pumps for overland flow). Cleaning of the
sand filter resulted in extra operational costs.
A comparison of the relative costs of a process or of a combination of processes
in Table 7 will aid in minimizing capital and land needs. Using a ratio
of the size of the demonstration facilities to those needed at another plant
and data in Table 7 will permit an estimate of the funding requirements.
388
-------
FUNDING
There is a difference between small meat plants and small poultry plants that
becomes increasingly significant when funding is discussed. Poultry processing
has become more highly automated. The typical small plant is a single-line
plant which slaughters 20,000 to 30,000 chickens per day and has a wastewater
discharge of about 200,000 gal/day. The small meat plant is commonly one that
slaughters less than 25,000,000 Ibs annually and has a wastewater discharge of
less than 75,000 gal/day. A poultry plant processing 28,000 chickens/day will
slaughter slightly less than 25,000,000 Ibs annually but will discharge about three
times the wastewater volume as a meat plant with the same annual kill. Though
small meat and poultry plants have about the same quantity and kind of pollu-
tants to treat, because of the larger volume requiring treatment the poultry
processors funding requirements will be two to three times larger.
Funding is more than obtaining the money, though for small businesses this
may be a critical item. Governmental regulations which reduce the business1
taxes or which lower the interest on the borrowed funds can be a significant
part of funding.
The financial opportunities of using depreciation and investment tax credit
and the relative cost of bank financing, Small Business Administration (SBA) loans
and governmental low interest bonds are described in an EPA Technology Transfer
Publication entitled, "Choosing the Optimum Financial Strategy—Upgrading Meat
Packing Facilities to Reduce Pollution."
Conventional bank financing is the most common source of funds for small busi-
nesses. However, the Federal Water Pollution Control Act has authorized the
Small Business Administration to assist small business concerns in adding
to or altering their equipment, facilities, or methods of operation in order
to meet water pollution control requirements established under the act.
The loans are to provide relief to small businesses who might otherwise suffer
substantial economic hardship without some financial assistance. SBA has three
loan options available. They may guarantee up to 90 percent of a loan,
enter into an immediate participation basis loan with a bank or make a direct
loan. Within limits, the private lender sets the interest rates on the first
two types of loans but on direct loans the interest rate corresponds to the
average annual interest rate on all U.S. obligations. The environmental
Protection Agency must perform a technical review of the application to
determine that the proposed additions or alterations are necessary and ade-
quate to meet the NPDES discharge requirements. For additional details contact
either the nearest EPA Regional or SBA District Office.
State governments often have programs which affect funding. Some states exempt
pollution control equipment or facilities from sales, use, and property taxes.
More often states have a financing program which raises the required funds
through state sponsored revenue bonds. The bonds bear a lower interest rate
because the interest is not subject to the Federal income tax. For small
plants, with capital requirements ranging from $25,000 to $250,000, the use
of these bonds may not be practical.
389
-------
DETAIL DESIGN OF LAGOONS
In this section, as in others, several lagoon based processes and their appurtenance;
are stressed. The detail design of intermittent sand filters and land appli-
cation processes are covered in this publication by Rowe and Tarquin, respectively.
Detail design means different things to different people. To a consulting
engineer it means extensive plans and specifications suitable for control of
a construction contract. To the manager of a small plant it means sufficient
instruction to direct plant personnel or a subcontractor to construct a portion
of the facilities. Drawings of the layout of the sewers, lagoons, inlets and
outlets are helpful in preventing errors in elevation, alignment, and size.
The drawings need not contain the details necessary for bidding by a contractor,
but to eliminate drawings would be a serious error. Some States require industries
to submit drawings from a licensed engineer for approval prior to construction.
Location
*"m" m"m I
To determine the exact location of the lagoons at the demonstration plant a
topographical map of the selected site was made. The map showed two foot contours
and soil depth. The topography, and underlying rock of the selected site con-
trolled most features of design. The limited soil at the site was mainly
composed of decayed paunch contents and hair and was unsuitable for dike
construction. A suitable clay soil that was being stripped to obtain
underlying sand was purchased from a quarry approximately one-half mile from the
site. The conglomerate rock at the site was from three inches to two feet under
the surface. Dike height and the underlying rock were the controlling factors in
location. To provide a minimum of three feet depth in the aerobic ponds
removal of rock was necessary. The ponds were drawn on copies of the topography
map at several elevations to calculate minimum cost of rock cut and dike full.
As a result, the ponds were located as far up the slope as possible without
eliminating a gravity flow system. The maximum depth of rock cut was 1.5
feet and maximum centerline height of the dike was 13 feet. A layout of the
treatment system drawn to scale is shown in Figure 4.
Fortunately one of two potential locations for the deep anaerobic pond was
in a draw which allowed construction of a pond with eight to ten feet of water
depth without excavation of rock. Normally location in a draw would not be
safe because of erosion problems from the intermittent stream flow. In this
case a stock watering pond existed above the site. The dam on the pond was
heightened two feet and the spillway structure was enlarged and shaped to
divert any overflow past the lagoon location.
Size and Shape
The size of the lagoon in terms of volume, depth, or surface area is based
on the design criteria given in the preliminary design section, i.e. organic
loading, F/M ratio and minimum depth. Site conditions will influence the size
within established limits of design. Construction techniques will also influence
size. For example, the use of a bulldozer requires a minimum dimension of
about 20 ft. in the lagoon bottom and this accompanied with the necessary dike
slope and depth results in a minimum volume and dike length.
390
-------
CO
10
FIGURE 4-LAGOON LAYOUT
-------
Cost and construction technique usually result in square and rectangular
shapes. Although the boundaries and topography of the site must be utilized
for economical design, the shape should be designed to prevent dead spots
and short circuiting. The location of single or multiple outlets and interior
dikes or baffles can form shapes which prevent dead spots and short circuiting.
In large lagoons the shape should minimize length in the direction of the
prevailing winds to reduce wave size. Wave build up can also be prevented by
interior dikes.
Pi kes
The most important and most expensive item to obtain is a water tight basin.
Excess seepage through the dikes or bottom is a common problem. Dike failures
also occur because of erosion due to wind and wave action or flow over the
top of the dike. The dikes must also keep storm runoff out of the lagoon.
The design of the dikes will be similar to designs used in construction of local
farm ponds. The extension agent experienced with local soils can determine
the suitability of the soil at the site, the need for the addition of clay
or moisture to the soil and safe dike slopes. Suitable soil will support
a dike slope of three horizontal to one vertical and sometimes up to a 2 to 1
slope can be used. Freeboard of at least two feet and often three feet above
the lagoons peak operating water level are used to prevent overtopping.
The width at the top of the dike is a minimum of eight feet. This not only
provides stability but is the width of most construction equipment and allows
the dike to be used for an access road. To prevent erosion the dikes are planted
with grass. On lagoons where wave action is expected to cause severe erosion,
the interior dike slope, one foot above and below the water levels, is protected
with broken rock or concrete. Occasionally a wooden bulkhead, asphalt strip
or cast-in-place concrete is used. The dikes at the demonstration plant are
shown in Figures 5 and 6.
The soil in the dike must be compacted to be stable. For satisfactory compaction
the soil must be moist. During the usual summer construction season this may
require the addition of water to the soil. Satisfactory compaction is obtained
by placing 0.5 to 1.0 ft. of soil, adding water, if necessary, and rolling the
layer with heavy equipment. Frequently the bulldozer used to spread the material
is also used for compaction.
Inlets and Outlets
The design of inlets and outlets must consider such factors as capacity, elevation,
location, and control. The size and slope of the sewer line must be sufficient to
carry the peak flow volume. Usually experience with existing plant sewers will
determine the necessary pipe diameter and slope. For example, 6 and 8 inch
diameter sewers flowing full with a slope of 1 ft/1000 ft will discharge about
100 and 220 gallons per minute (gpm), respectively. If the slope is decreased to
0.5 ft/1000 ft. these discharges will drop to about 70 and 150 gpm. Outlet
elevations are used to control direction of flow and water levels in the lagoons.
The size, slope and elevations of the inlets and outlet pipes should be care-
fully set to prevent undesired restrictions during peak and minimum flow condi-
tions. The inlet, outlet and all sewers, should be designed with access for
cleaning or unplugging operations. This requires a manhole at each change in ^
sewer direction and smooth alignment between manholes.
392
-------
FIGURE 5 - ANAEROBIC OR.EXTENDED AERATION LAGOON
FIGURE 6 - FIRST AEROBIC LAGOON
393
-------
At the demonstration site a manhole was placed at the edge of the lagoon.
The sewer line grade board marks the location in Figure 5. The manhole outlet
invert was set 0.1 feet lower than the inlet invert and a difference of 0.5
feet was set between the outlet in the manhole and the water level in the pond.
These energy drops were to prevent plugging of the 6-inch pipe by the large
particles in the raw wastewater. The end section of this inlet pipe was
set on a small pad of concrete to prevent scarring of the bottom and slippage
of the pipe down the long slope of the dike.
The discharge of the inlet is typically near the center of the lagoon to obtain
mixing. When mechanical aeration is provided the inlet discharge should be in a
zone of turbulence caused by the aeration. In aerobic lagoons where mixing is not
desired, the inlet may be located far from the outlet and/or in the deepest
portion of the pond. At the demonstration plant the 6-inch inlet pipe to the
first aerobic pond was extended to discharge to the deepest portion of the
pond (Figure 6). This deep discharge was designed to reduce odor emission,
to contain the anaerobic zone under the aerobic zone, and to provide storage
for settleable solids in the wastewater.
Control of the flow in the inlet design is needed to divert the flow around the
lagoon during periods of repair and to divide the flow where two or more
lagoons are used in parallel. Figure 7 is a detailed drawing of an inlet manhole
used to by-pass the first lagoon at the Reeves plant. By use of a wooden gate
in the manhole, flow can be diverted to the second lagoon. The square shape
and other design details were to simplify construction.
Design of outlets requires the same considerations, i.e. capacity, elevation,
location and control. The design of the outlet can incorporate several different
desirable features depending on the lagoon process, i.e. scum retention, flow
measurement, or restriction of the volume.
A second common outlet for lagoons is a drain which is located in the deepest
portion of the lagoon. The drain pipe is designed to withstand the pressure
of the dike and construction equipment and to accept a valve or plug. The
pipe must have a water tight seal with the dike. The drain line can be
eliminated if a large capacity pump, portable pipeline and power can be
readily available to pump the lagoon.
The required outlet capacity is not necessarily the maximum gpm used in
the plant. The lagoon will dampen the maximum flow to near the average used
during the day's operation. Precipitation on surface of the lagoons will establish
the maximum discharge possible. Usually this maximum discharge is not desired
and is restricted by the outlet and use of the freeboard for temporary storage.
The simplest of restrictions is the size of the outlet. Thus capacity, elevation
and control in outlet design frequently interact to obtain a desired discharge
volume.
The invert elevation of the outlet pipe controls the minimum water depth in
the lagoon and must be set to obtain the desired depth. The outlet is located
near to the next process or receiving waters to reduce pipeline cost and away
from the inlet to reduce short circuiting.
394
-------
2"x8" Slide
Gate To Block
Flow
5" ClayT
From
Plant
Hand Formed Channel In Floor-
To Obtain Smooth Flow
TOP VIEW
4'-0"
_Welded 1/4" Plate Steel Top
/ Hinged To Wall Of Manhole
6" Pipe-,
-Concrete
Elevation O.IO'LowerJ.
Than Inlet
2"x 8" Slide Gate To-
Elevation 0.10' Block Fl°*
''Higher Than Outlets
i i
! I
I I
Floor Poured Seperately To Grade
7—c
O
-Steel Strap
Handle
ro
o
I
V
6"Pipe
*!
FIGURE? - INLET 8 BYPASS MANHOLE
2"Channel Iron Driven
Into Ground Concrete
Poured Around
395
-------
The design of the outlets at the demonstration site are presented as examples.
The outlet for the anaerobic lagoon was located two feet below the water surface
to allow formation of a scum blanket and to prevent blockage of the pipe. The
outlet was a 6-inch clay pipe extending through a small concrete pad on the
interior slope of the dike. The pipe entered a manhole, similar to the one in
Figure 7, one foot below the water level to reduce the release of odorous gases
which can occur at overflow weirs used with anaerobic lagoons. The water level
in the lagoon was controlled by the elevation of the outlet in the manhole.
The outlet from the first aerobic lagoon was a steel pipe laid during construc-
tion of the dike. The pipe was located one foot below the water surface and
at the furthest point from the inlet to reduce discharge of floating materials
and short circuiting, respectively. Extending the 6-inch pipe one foot beyond
the dike slope would have prevented plugging which occurred twice.
The outlet from the second aerobic lagoon discharged to the receiving stream.
Sample collection and flow measurement at this point was required. To
accomplish these tasks with the accuracy required in a demonstration project
an outlet manhole was designed. The manhole was located at the maximum
distance from the inlet to reduce short circuiting. Details of the structure
are shown in Figure 8. This concrete structure was designed with a submerged
inlet and a variable weir elevation to reduce discharge of floating materials
and permit controlled discharges. The length of the structure was designed
to allow dry access from the bank and to provide room for flow measuring and
sampling equipment. As depicted in Figure 8, a large plywood box was constructed
and installed over the manhole to protect the composite sampler, flow recorder,
and the sample container.
The NPDES permit for small plants may allow grab samples and flow measurements
over a few minutes. If such is the case a simpler and less expensive design
would be an outlet pipe at the desired elevation protected by a wooden baffle
with a submerged opening. Flow measurement can be determined by the height of
flow in the pipe.
The outlet for the extended aeration lagoon was designed to limit the discharge
to a four hour period each day. The outlet was a 6-inch pipeline. An automated
6-inch valve was placed on the line within a manhole where maintenance could be per-
formed. An accumulation of one day's flow in the lagoon raised the water level
6 to 9 inches. This small head allowed use of a rubber seated butterfly valve.
Either an air activated or a motor operated valve was suitable in conjunction
with an electric timer to open and close the valve.
Aeration Equipment
Design of an aeration system deserves an effort in proportion to its cost.
The aeration equipment at the demonstration plant equalled the cost of constructing
the lagoon. Design consists primarily of determining the needed capacity and
selecting equipment from one of many manufacturers. ~Lengthy delivery times might
be expected and design, selection and purchase should be done prior to starting
construction.
The available equipment can be grouped into compressed aeration systems and
mechanical aeration systems. The simplicity of installation and low initial cost
396
-------
_2 Angle Iron Boltsd
To Walls of Manhole
1/4 Plate Sfsel
(Variable Height
v"Nofch Weir
Made of 8 Gage
Sheet Metal
Screwed To 3/4
Marine Plywood
6 Discharge Pipe
Formed With 3/4 Plywood
. Plywood Top
II
Weir Height Con Be
Varied2fl.By Changing
Height of Plywood
Wood Bex Constructed of 2 x 4 Studs
And 1/4" Ext. Plywood Siding To
Inlorge Work Area In Manhole And
To Protect Equipment
1/4 Plate Steel Floor To Support
Sompiing, Equipment
FIGURE 8-FINAL EFFLUENT MANHOLE
397
-------
are advantages of the mechanical aeration systems which can be mounted on
fixed or floating platforms. The changing water levels in the extended aeration
lagoon eliminate fixed platform mechanical aerators. In climates where
extensive ice formation is common, the floating aerator can malfunction due
to ice build up on or around the equipment. The compressed air system consists
of a compressor or blower and a distribution system. The compressor or
blower needs to be housed and the distribution system needs to be cleaned
occassionally. There are several types of diffusers used in distribution
systems for lagoons. The static-draft tubes are designed for low maintenance.
The advantage of the system is the air is introduced at the bottom of the
lagoon and is not subject to problems from ice formation.
A number of manufacturers produce aeration equipment and can be contacted for
aid in selecting equipment. Annual repair costs of 5% or more of the initial
cost may be expected. Consequently, the manufacturer's service capabilities
and parts inventory should be determined.
The needed capacity of the aeration system is described in terms of pounds
of oxygen per day. There are several design formulaes for computing the oxygen
requirements. For small plants a simplified formula using the BOD_ of the
wastewater in Ibs/day and desired Mixed Liquor Volatile Suspended Solids (MLVSS)
in Ibs is:
Oxygen Needed = a1 (BODg) and b1 (MLVSS)
The BODr of the wastewater was determined in the preliminary design. The pounds
of MLVSS is the BODr divided by the F/M ratio which was 0.06 at the demon-
stration plant. The demonstration plant operated satisfactorily with a
concentration of MLVSS between 1,000 and 4,000 mg/1. The terms a1 and b1
are coefficients to describe the oxygen needed for synthesis and the oxygen
needed for endogenous respiration, respectively. These coefficients can be
determined in the laboratory, and this is done where large aeration needs
exist. For small plants, use of coefficients developed on^similar wastewaters
are commonly used. The coefficients used at the demonstration plant were
0.48 for a1 and 0.03 for b1, which were developed in an extended aeration
plant treating a food processing wastewater.
Following is a sample calculation using the wastewater characteristics at the
demonstration plant for average conditions:
BOD5 = 1247 mg BOD5/1 (17,800 gal/day) x 8.34 lbs(10"6 1/mg) = 185 Ibs BOD5/day
MLVSS = 185 4 0.06 Ib BOD5/day/lb MLVSS = 3083 Ibs
Oxygen needed = 0.48 (185) + 0.03 (3083) = 181 Ib/day
With the planned 18 hours of aerator operation per day and the generally accepted
2 Ibs of oxygen transferred per horsepower-hour (Hp-hr), aerator requirements were
5 Hp. for average conditions.
To allow for above average waste loads and inaccuracy of the coefficients,
A 10 Hp. Peabody Welles floating aerator with a variable oxygen transfer
valve was selected. This valve when open reduced the power consumption along
398
-------
with the oxygen transfer. The motor was equipped with a heater to keep it
free from condensate moisture during the 6 hrs/day the aerator was off.
The aerator supplied sufficient oxygen to operate the process even when operated
at 8 Hp.
During the first 18 months the aerator ran as programmed except for a three-
day inoperative period due to vandals cutting the mooring'ropes. During the
second 18 months, the aerator malfunctioned five times and was off for a total
of two months. One malfunction was due to vandalism, three required rewinding
the motor, and one was due to a problem in the electrical control panel. The
motor was rewound twice due to poor alignment of the shaft and impeller when
reassembling the aerator.
The amount of non-operation due to repair, and the infrequency of maximum
oxygen requirements shows that dual aerators in the basin would have given
superior performance. The aeration system, in addition to providing sufficient
oxygen, must keep the biological floe mixed with the wastewater. Mechanical
floating aerators have a conical shaped mixing zone, and two aerators would
more effectively prevent sludge deposits in a rectangular basin. Thus selection
of the aeration system and the F/M ratio may significantly affect lagoon design.
CONSTRUCTION
Construction begins with planning the sequence of events necessary to complete
the facilities. This requires the coordination of equipment, manpower and
materials for each event. Guidelines for satisfactory construction are presented
under the subheadings of Preparation, Lagoons, Pipelines and Manholes, and
General Items. Construction of the demonstration plant is described to in-
dicate the planning required.
Preparation
The builder should determine the availability of transportation, handling,
storage and disposal of materials; availability of labor, water, electric
power, roads; and uncertainties of weather, river stages, or similar physical
conditions at the site. He should determine the conformation and conditions of
the ground, the character of equipment and facilities needed preliminary to and
during the prosecution of the work and the character, quality, and quantity
of surface and subsurface materials to be encountered.
The builder should be prepared to furnish all materials, equipment, tools,
labor, supervision and other services necessary to construct a waste treatment
system.
The builder should prepare for construction by clearing all timber, logs,
brush, rubbish, large rocks and unsuitable soils from the lagoon site. Areas
to receive fill or compaction should be scarified to a depth of six inches,
and the moisture content of the material to be compacted should be adjusted
to obtain near maximum density.
The builder will need to establish the center lines of principal structures,
roads, pipelines, and facilities, set slope stakes and bench marks to establish
the basic layout.
399
-------
Lagoons
The lagoon should be excavated to the dimensions shown on the plans with special
attention to maintaining the given inside and outside dike slopes.
Material for the fill is from the required excavation or designated borrow
areas. The material used is to be free from vegetable matter and other
deleterious substances and not contain rocks or lumps having a diameter
larger than six inches.
The fill material is placed in uniform layers which when compacted shall not
exceed six inches. Each layer shall be thoroughly mixed during the spreading
to insure uniformity of material and moisture in each layer.
When fill material includes rocks, no large rocks should be allowed to nest
and all voids should be carefully filled with earth, and properly compacted.
No large rocks should be closer than twelve inches below the finish grade
except on the inside slope of the dikes.
Compaction should be by sheepsfoot rollers, multiple-wheel pneumatic-timer
rollers or other types of suitable compaction equipment. Compaction should
be accomplished while the fill material is at a satisfactory moisture content.
Compaction of each layer should be continuous over its entire area and the
compaction equipment should make sufficient trips to insure that the desired
density has been obtained.
Compacting equipment should be operated so that the full width of the fill is
covered. Coverage shall be as one continuous trip from end-to-end and shall
overlap previous coverage by not less than three inches. For pipelines laid in the
fill, construct fill surface to at least an elevation two feet above the top of
proposed pipeline prior to starting trench excavation for installation of
pipelines. Sprinkle fill material with water as necessary to produce satisfactory
compaction. If materials is too wet for proper compaction, aerate by discing.
Upon completion, grade surface to proper elevations and cross sections.
Dress side slopes as indicated.
The fill operation should be continued in six-inch compacted layers, as stated
above, until the fill has been brought to the finished elevations.
All excess excavation, not required or suitable for backfill or filling,
shall be disposed of in the designated waste area. The waste area shall be
uniformly graded to conform to existing contours, left with a neat appearance,
and be free-draining.
After grading is completed, topsoil should be spread over the entire graded
area, excluding surfaced area covered with gravel or rip rap and surface areas
to be inundated by water, and planted with native grass.
Pipelines and Manholes
Clay pipe where used should be standard strength clay sewer pipe and conform
to American Society for Testing and Materials (ASTM) standards. When pipe is being
400
-------
placed the lower 90 degree arc of the barrel of the pipe should be in firm
contact with undisturbed earth. Excavations should be made for the bells, but
should be no larger than necessary to clear the bell. Where clay pipe is to be
laid on rock or where the surface is unsmooth and irregular a four inch layer
of sand, crushed rock or small aggregate should be placed under pipe for complete
support. Clay pipe should be laid with correct alignment and slope. Joints
must be watertight to hold infiltration to a minimum. Trenches should be kept
water-free during jointing and for a sufficient period thereafter to allow the
jointing material to become fully set and completely resistant to water penetration,
Obstructions to the construction of the trench, such as tree roots, stumps,
abandoned structures, and debris of all types, should be removed. Minimum
width of trenches in which pipe is to be laid should be 18 inches. The align-
ment and grade of the trench should be established by a surveying instrument
with proper allowance for pipe thickness and for base or special bedding when
required. Removal of rock from the trench and the equipment and grade boards
used at the demonstration plant are shown in Figure 9.
Where pipes pass through the embankments, no granular pipe base or pipe zone
material should be used. Instead, the pipe should be laid on the trench invert.
Steel pipe joints should be welded. The trench should be backfilled with
selected material.
If the manhole is cast-in-place the form's exterior exposed surfaces should be
plywood. Form all vertical surfaces. Trench walls, large rock or earth is
not a suitable form material. The concrete should be a high strength mix.
Precast manhole sections conforming to ASTM Standards, with circular reinforcement,
may be used. The diameter at the base should not be less than four feet.
The concrete base should be constructed so that first section of precast
manhole has uniform bearing throughout the full circumference. Sufficient
grout should be deposited on base to assure watertight seal between base and
manhole wall or the precase section of manhole can be placed in concrete base
before the concrete has set. The section should be properly located and
plumbed. If material in bottom of excavation is unsuitable for supporting a
manhole, excavate two additional feet and backfill to required grade with
three inch minus, clean, pit-run material. Water should be removed from the
excavation prior to pouring the base.
Do not backfill around concrete structures until the concrete has obtained
sufficient compressive strength. Remove all form materials and trash from
the excavation before placing any backfill.
Do not operate any heavy earth-moving equipment within 5 feet of walls of concrete
structures for the purposes of depositing or compacting backfill material.
Compact backfill adjacent to concrete walls with pneumatic tampers or other
equipment that will not damage the structure.
General Items
The builder should surround the waste treatment site with a permanent stock-
tight fence and provide appropriate signs to designate the nature of the
facilities.
401
-------
FIGURE 9 - TRENCH CONSTRUCTION FOR SEWERLINE
FIGURE 10 - MOWING GRASS AROUND LAGOONS
402
-------
The builder should take necessary precautions for the safety of employees on
the work site and to prevent accidents or injury to persons on, about, or
adjacent to the premises where the work is being performed. The builder should
provide and maintain at all times during the progress or temporary suspensions
of the work, suitable barricades, fences, signs, signal lights, and flagmen as
necessary to insure the safety of the public and those engaged in the work.
The builder should construct the necessary ditches, provide the necessary
pumps, and take such precautions as are required to protect the work. Divert
or pump and streamflow and drain the construction area so the work may be
carried on in a satisfactory manner. Drain or otherwise dewater all excavation
areas as required to permit satisfactory operation.
The builder should construct an access road to the work site during construction
suitable for latter access by operators.
The builder should keep the property on which work is in progress and the ad-
jacent property free from accumulations of waste material or rubbish caused
by employees or by the work.
•
Demonstration Plant
Construction of the demonstration plant shown in Figure 4 was begun in December
and required four months to complete the pipelines, manholes, and three ponds.
The anaerobic pond and sewer were completed first and wastewater was turned into
the pond on February 1. The two aerobic ponds were completed next and received
wastewater on April 15. The fence and auxiliary items were then installed. The
last construction item was sprigging of the dikes with bermuda grass which was
delayed until just prior to the growing season.
Weather, diversion of manpower to other tasks and slow delivery of mechanical
equipment were the causes of the major construction delays. Construction
equipment breakdowns and availability of construction materials were minor
problems.
Construction of facilities shown in Figure 4 was mainly accomplished by three
maintenance personnel at the packinghouse under the direction of the owner.
Two others aided in operating dump trucks, front-end loader, and bulldozers
during the earth moving phases of construction. Other part-time manpower
consisted of a survey party and an engineer. No other manpower was required
except that used by the electric company to relocate power pole and to
install service to the project. The use of packinghouse personnel resulted
in delays in construction but minimized cash outlays.
The equipment used on construction of the facilities consisted of two bulldozers,
one 16-yard earthmover, three dump trucks (1-12 yard and 2-6 yard), one
backhoe, one front-end loader, one pickup truck, two air compressors, and a
1/2 sack concrete mixer. The earthmover was inoperable for several weeks
and most of the soil was transported by the trucks. The owner purchased
some of the equipment and leased the rest.
The major materials purchased were soil (a silty clay loam), vitrified clay
sewer pipe, steel pipe for drains, cement, gravel and sand for concrete, form
lumber and dynamite.
403
-------
OPERATION
Preliminary design, funding, detail design and construction are only undertaken
so the treatment system can be operated. Operation need not be difficult or
time consuming, but it must be daily. For small plants, a daily inspection to
catch malfunctions early is the most important part of operation. Something as
simple as seeing an abnormal discharge volume or water height will allow
correction before a major problem occurs. The first and most important
part of operation is to assign responsibility for the system to a capable
and conscientious employee.
After an operator has been selected and his work schedule arranged to permit
daily inspection and operation, he should be instructed to compile the available
information and tools to also undertake maintenance. All manufacturer's speci-
fications, shop drawings and maintenance manuals should be read and kept
as a permanent record. A manual on operation of treatment plants should be
obtained and used as a reference. Operation of Wastewater Treatment Plants,
Manual of Practice No. 11, is available from the Water Pollution Control Federa-
tion, 3900 Wisconsin Ave., Washington, DC 20016 for $3.00. This manual contains
information on the causes and cures of many operational problems. Those reports
listed in the appendix of this publication will give further information on
operation and maintenance of specific wastewater treatment systems. *
Care of the treatment site is also necessary. The fences, gates and access
road will need to be maintained. Mowing of the grass in the land intensive
system as described herein can be quite time consuming. A simple but effective
alternate is to have cattle graze the grass and occassionally mow the weeds
and brush as is done in most pastures (Figure 10).
Other than the necessary daily inspection the amount of operation and maintenance
will vary considerably depending upon the amount of mechanical equipment,
season of the year and detail design. A system of routine preventative maintenance
should be developed especially for mechanical equipment. Some of the routine
maintenance that can be expected by type of facility follows:
Pipeline and manholes will need cleaning of grit, grease, and plant
growth to prevent plugging and odors.
Pumping stations will have similar cleaning requirements and the
pumps will need to be lubricated and the packing gland checked.
The motors, belt drives and water level control will need adjustments.
The pumps used should be alternated weekly.
Screening facilities and grease traps need to be inspected daily
and cleaned when required to obtain effective solids-liquid separation.
The collected solids can be disposed of by burying.
Chlorination facilities need to be checked daily to determine feed
rates, availability of chlorine and leaks.
The extended aerated lagoon has an aeration system, an automated
control panel, and a power operated valve. Daily inspection is
necessary to determine satisfactory operation. Collecting a sample
of mixed liquor and determining the solid-liquid separation in 30
minutes will show if excess sludge is being carried out with the
404
-------
discharge. The extended aeration lagoon needs to have sludge with-
drawn when the MLSS concentration goes above 5000 mg/1. These settling
and concentration measurements on the mixed liquor indicates one
use of an analytical monitoring program for operation. A monitoring
program especially designed for small packers and the use of this
information will be given by Rowe later in this publication.
The sludge from an extended aeration process has been highly treated
and if disposed on land can be a simple operation for small plants.
On five occasions during the year at the demonstration plant several
thousand gallons of settled sludge were removed through the drain
pipe and placed in the surrounding pasture.
A plant manager may wish to establish a maintenance and operation
form to remind the employee of daily, weekly and monthly inspection,
maintenance, and monitoring duties and to establish a record of-l '^
diligent operation. A form that covers one months activities ^r
been found most convenient.
405
-------
LIST OF ABBREVIATIONS AND SYMBOLS
BAT
BOD.
BPTD
COD
F/M
FOG
gal/day
gpm
Hp-hr
LWK
mgad
mgd
mg/1
ml
MLSS
MLVSS
mm
mpn/100 ml
n
N2 .
NH,
NH--N
NO^-
N0,~
NPDES
SRT
SVI
TKN
T-P
TSS
Best Available Technology Economically Available
Five day Biochemical Oxygen Demand
Best Practical Control Technology Currently Available
Chemical Oxygen Demand
Food to Microorganism Ratio
Fats, Oils or Greases
gallons per day
gallons per minute
Horsepower-hour
Live Weight Killed
million gallons per acre per day
million gallons per day
milligram per liter
mi 11i1i ter
Mixed Liquor Suspended Solids
Mixed Liquor Volatile Suspended Solids
millimeter
most probable number per 100 milliliters
number
Nitrogen as a gas
Ammonia
Ammonia as Nitrogen
Nitrite
Nitrate
National Pollutant Discharge Elimination System
Sludge Retention Time
Sludge Volume Index
Total Kjeldhal Nitrogen
Total Phosphate
Total Suspended Solids
406
-------
BIBLIOGRAPHY
Anonymous. Packinghouse Sewage Disposal and Grease Recovery. 1946-
Meat Packers Guide. The National Provisioner.
Baker, Darrell A., Allen H. Wymore and James E. White. Treatment of
Packinghouse Wastes By Anaerobic Lagoons and Plastic Media Filters.
Environmental Protection Technology Series. EPA-660/2-74-027, April 1974.
Bartsch, Eric H. and Clifford W. Randall. Aerated Lagoons - A Report on
the State of the Art. Journal Water Pollution Control Federation. Vol. 43,
No. 4. April 1971.
Beefland International. Elimination of Water Pollution by Packinghouse
Animal Paunch and Blood. U.S. EPA Water Pollution Control Research Series.
12060 FDS, 11/71.
Caldwell, D. H., D. S. Parker and W. R. Uhte. Upgrading Lagoons. EPA
Technology Transfer Seminar Publication. August 1973.
Dawson, F. M. and A. A. Kaliske. Symposium on Grease Removal - Design and
Operation of Grease Interceptors. Sewage Work Journal, Vol. 16, No. 3.
May 1944.
Eldridge, E. F.. The Meat Packing Plant Waste Disposal Problem. The Na-
tional Provisioner. Feb. 23, March 9, March 30, April 27, and May 18, 1946.
Eye, J. David, David P. Eastwood, Fernando Requena and David P. Spath.
Field Evaluation of the Performance of Extended Aeration Plants. Journal
Water Pollution Control Federation. Vol. 41, No. 7. July 1969.
Griffel Associates, Inc. Waste Treatment (Poultry Processing Facilities).
EPA Technology Transfer Seminar Publication. July 1973.
Gold, Donald D. Summary of Treatment Methods for Slaughterhouses and
Packinghouse Wastes. Engineering Experiment Station Bulletin No. 17.
The University of Tennesee. 1953.
Klassen, C. W. and W. A. Hasfurther. Treatment of Wastes from Small
Packinghouses. Sewage Work Engineering. March 1949.
Kountz, R. Rupert. Treatment of Waste from Small Slaughterhouses. Proceedings
of Industrial Waste Conference. Purdue University, 1954.
Larson, K. D. and D. A. Maulwurf. Evaluation of Polymeric Clarification of
Meat-Packing and Domestic Wastewaters. Environmental Protection Technology
Series. EPA-660/2-74-020, April 1974.
Lee, John W. Tertiary Treatment of Combined Domestic and Industrial Wastes.
Environmental Protection Technology Series. EPA-R2-73-236, May 1973.
407
-------
McGrail, D. T. Poultry Processing Wastewater - Advanced Treatment and
Reuse. Proceedings Seventh National Symposium on Food Processing Wastes.
EPA. 1976.
McVaught, J. B. and J. A. Home!, Jr. A Meat Packers Solution to Meeting the
1983 Effluent Requirements. Proceedings Seventh National Symposium on
Food Processing Wastes. EPA. 1976.
Neil, J. H. The Use and Construction of Oxidation Ponds for Industrial
Waste Treatment. Proceedings Seventh Ontario Industrial Waste Conference.
Paulson, Wayne L. and Lawrence D. Lively. Oxidation Ditch Treatment of
Meat Packing Wastes. Environmental Protection Technology Series. EPA.
1976.
Reich, J. S. Effective Grease Recovery Shows In Profits and Treatment
Efficiency. Water and Sewage Works. January 1970.
Rollag, D. A. and J. N. Dornbush. Design and Performance Evaluation of an
Anaerobic Stabilization Pond System for Meat-Processing Wastes. Journal
Water Pollution Control Federation. Vol. 38, No. 11. Nov. 1966.
Steffen, A. J., Dan Lindenmeyer, M. E. Ginaven, Robert Johnson and Charles
Grimes. Pretreatment of Poultry Processing Wastes. EPA Technology Transfer
Seminar Publication. July 1973.
Steffen, A. J. Waste Disposal in the Meat Industry, A Comprehensive
Review. Proceedings, Meat Industry Research Conference. American
Meat Institute Foundation, University of Chicago. March 1969.
Stiemke, Robert E. Disposal of Wastes from Small Abattoirs. Proceedings
of Industrial Waste Conference, Purdue University. 1948.
Summerfelt, Robert C. and S. C. Yin. Paunch Manure as a Feed Supplement
in Channel Catfish Farming. Environmental Protection Technology Series.
EPA-660/2-74-046, May 1974.
Symore, A. H. Trickling Filters for Small Meat Packing Plants. Sewage
Works Engineering, 20:7. 1949.
U.S. EPA. Development Document for Effluent Limitations Guidelines and
New Source Performance Standards for the Red Meat Processing Segments of
the Meat Products Point Source Category. EPA-440/1-74-012a, February
1974.
U.S. EPA. Development Document for Effluent Limitations Guidelines and
Standards of Performance for the Poultry Processing Industry. 1976.
U.S. Public Health Service. An Industrial Waste Guide to the Meat Industry.
U.S. Public Health Service. Publ.#386. 1954.
408
-------
Wells, W. James, Paula B. Wells, Charles A. Haas, Steve L. Hergert and
Steve 0. Brown. Waste Treatment (Meat Packing Facilities). EPA Technology
Transfer Seminar Publication. October 1973.
Wells, W. James, Paula B. Wells and Darryl D. Alleman. Treatment Capabi-
lities of and Extended Aeration System Following Anaerobic Lagoons Treating
Meat Packing Wastes. Proceedings Sixth National Symposium on Food Pro-
cessing Wastes. EPA. 1976.
Witherow, Jack L., Mickey L. Rowe and Jimmie L. Kingery. Meat Packing
Wastewater Treatment by Spray Runoff Irrigation. Proceedings Sixth National
Symposium on Food Processing Wastes. EPA. 1976.
Whitehead, W. K. Evaluating and Treating Poultry Processing Wastewater.
Proceedings Seventh National Symposium on Food Processing Wastes. EPA. 1976.
Woodard, F. E. Treatment Alternatives for Poultry Processing Wastewaters.
Proceedings Seventh National Symposium on Food Processing Wastes. EPA. 1976.
Wymore, A. H. The Design and Operation of a Waste Treatment Plant for a
Small Packing Plant. Sewage and Industrial Wastes, 24:7. July 1952.
Wymore, Allan H. and James E. White. Treatment of a Slaugherhouse Waste
Using Anaerobic and Aerated Lagoons. Proceedings Industrial Waste Conference,
Purdue University. 1968.
409
-------
EVALUATING AND TREATING POULTRY PROCESSING WASTEWATER
W. K. Whitehead
INTRODUCTION
The U.S. poultry processing industry slaughters over 3 billion chickens annually
(1) and produces a tremendous amount of waste that must be treated before it can
be released to a stream. Waste treatment represents an ever increasing cost to
both processors and consumers. Recent performance standards (2) and other
regulations require that the wastes be sufficiently treated to reduce pollution
effects on our environment.
The quantity of water used for processing poultry and the amounts of wastes
generated in the processing areas have been reported. Data from the Gold Kist
Study (3) showed that the average water use for a typical processing plant was
46.5 liter/bird or about 28 liter/kg LWK (live weight killed). As recently as
1971 poultry processing plants in Georgia were using an average of 40.5
liter/bird for slaughter, eviscerating and further processing (4) and for plants
with a municipal water supply the average water rate was $0.12/m3. Based on
these data, water costs the poultry processing industry about $14 million
annually or about $4.86 per thousand chickens slaughtered.
One recent study (5) reported average raw waste characteristics for chicken
processors. BODs averaged 9.9 kg/k kg LWK; suspended solids, 6.9 kg/k kg LWK;
grease, 4.2 kg/k kg LWK; and COD, 19.7 kg/k kg LWK. This study also reported
that the principal sources of waste water in the plants surveyed were the
feather and offal flow away systems, but other areas also contribute to the
waste load. Hamm (6) reported that the bird chiller, giblet chiller,
eviscerating and scalder-defeathering areas produced significant amounts of
COD, fat and solids. Carawan £t al. (3) reported that over 51% of the total
600$ load originated either in the blood collection tunnel, clean-up operations,
or in other miscellaneous activities. That report showed that the final bird
washer and the giblet and bird chillers accounted for about 1.5 kg BOD5/k kg
LWK. Drainage from the offal truck was another large contributor. The overall
plant average BODs load was 15.72 kg/k kg LWK.
If poultry processors are to meet recent regulations governing the amounts of
wastes they may discharge to natural streams, investments of time and money
must be extended. The performance standards (2) eventually limit discharges to
0.30 kg BODs/k kg of LWK, 0.34 kg TSS/k kg of LWK, 0.20 kg oil and grease/k kg
Agricultural engineer, Richard B. Russell Agricultural Research Center, Southern
Region, Agricultural Research Service, U.S. Department of Agriculture, P. 0. Box
5677, Athens, Georgia 30604.
410
-------
of LWK and 4.0 mg of ammonia/liter of effluent. Compliance with those levels
is not impossible, but most poultry processing waste treatment systems do not
consistently meet those standards. One recent analysis (7) estimated invest-
ment and operating costs for model processing plants. A large broiler plant
with a capacity of 76,800 birds/day and a wastewater volume of 2700 m3/day
would require an incremental investment cost of $161,000 and an annual operat-
ing cost of $25,000 to meet the 1977 best practicable technology currently
available (BPT) treatment level. This same plant would require an investment
of $203,000 and an annual operating cost of $52,000 to achieve the 1983 best
available technology economically achievable (BAT) guidelines. An investment
of $578,000 and $62,000 annual operating cost would be necessary to meet the
new source performance standards (NSPS).
Several alternatives are available for treating poultry processing wastewater.
Pretreatment with screens, filters and/or flotation is necessary before dis-
charging the wastewater to either a municipal or private waste treatment system.
These procedures can reduce BOD5 and concentrations of suspended solids and fat
by removing the larger solids from the wastewater. About 65% of the Federally
inspected poultry processing plants discharge their final waste to a municipal
treatment system (8). This may be the most economical means of handling the
wastewater, but for the remaining 35% not having access to municipal sewers an
efficient private waste treatment system is necessary to comply with the
discharge standards.
Most plants with private waste treatment systems utilize some form of lagoon,
either anaerobic, aerobic or a combination of types. Extended aeration, land
application, the rotating biological contactor and activated sludge may also be
used. However, in 1972, Vertrees (8) reported that only two poultry processing
plants used activated sludge and therefore, little information is available on
the application of activated sludge to poultry processing wastewaters.
Since its conception, the activated sludge process has been widely used for
treatment of both municipal and industrial wastewaters. Various modifications
of the process have been developed, but some basic principles are common to all
the processes. The conventional activated sludge process can remove 95% of the
BODs and suspended solids. Although the process produces solids that must be
disposed of, it is very stable and reliable and, when compared to the alternatives
of lagoon or land application systems, its greatest advantage is the small land
area required.
The major physical structures of an activated sludge process are the biological
reactor and the solids-liquid separator. The inputs to the reactor are waste-
water and concentrated activated sludge from the separator and air. Microorga-
nisms in the return sludge react with the organic pollutants in the wastewater
to produce more activated sludge, carbon dioxide and water. The supplied air
provides a mixing action and oxygen for the reaction. The process effluent is
the clarified overflow from solids-liquid separator. The net production of
sludge must be withdrawn and further processed or dryed for disposal.
411
-------
The objectives of this study were to: (1) identify areas of waste generation
in a poultry processing plant and characterize and quantify the amounts of waste
produced, (2) monitor an existing lagoon-type waste treatment system and
determine its efficiency for treating poultry processing wastes, and (3) design
a laboratory-scale conventional activated sludge system and operate it under
controlled conditions to determine its feasibility for use to treat poultry
processing wastewater and determine optimum operating conditions.
PROCEDURE
The first phase of this project was to identify the amounts of waste produced
in each area of a poultry processing plant. A survey of commercial processing
plants showed that several areas seemed to be major contributors of waste, the
scalding process, the picking area, the eviscerating room and the chillers.
The wastewater from one plant, processing 9600 broilers per hour, was sampled
and analyzed for a period of about 10 weeks. Samples were collected weekly,
during normal plant operations, from the scalder overflow, the eviscerating
troughs, the whole-bird chillers, the giblet chillers, the offal truck discharge
and the total plant final effluent. Samples were analyzed for 8005, COD, TOC,
fat, total and volatile solids, and total and volatile suspended solids (9).
The total amount of water used in the processing plant was recorded and the
amounts used in certain areas were measured. The amounts of overflow from the
scalders and the chillers were fixed by inspection regulations at 0.95 and 1.9
liter per bird, respectively. The total amount of water used in the eviscerating
room was that measured with a 90°V-notch weir plus the amount used in hand-wash
stations, which was calculated by collecting and timing the discharge.
In the second phase of the project, we monitored an existing poultry processing
plant waste treatment system. This system (Fig. 1) is a four-stage lagoon
process with aeration in the first two ponds. The first pond has a surface
area of 14,600 m^ (3.6 acres) and is equipped with three aerators. Discharge
from this pond flows to a second pond with 1080 m^ (0.27 acres) surface area
equipped with one aerator which keeps the wastewater in a highly mixed condition.
The wastewater then flows to the first of two settling ponds. Neither of these
ponds is aerated or mixed. Surface area of the first is 8820 m^ (2.2 acres)
and of the second, 7130 m^ (1.8 acres). The effluent from the final pond is
discharged through a weir to a small stream.
The plant processed about 90,000 broilers/day, 5 days a week during the sampling
period. The amount of water discharged to the stream was measured by the weir,
recorded continuously and averaged about 13,225 m^ (3,500,000 gal) per week.
Grab samples taken while the plant was operating and 24-hour composite samples
of the raw wastewater and the treated effluent were collected weekly for 18
weeks during the summer and early fall. Samples were collected on Wednesday or
Thursday of each week and analyzed for BODs, COD, TOC, fat, total and volatile
solids, total and volatile suspended solids and total (Kjeldahl) nitrogen (9).
412
-------
co
RAW WASTEWATER
/FROM PROCESSING
PLANT
POND 2
AERATOR
POND I
3 AERATORS
TREATED
EFFLUENT
DISCHARGED
TO STREAM
* SAMPLE POINT
Fig. 1 Layout of a poultry processing plant's stabilization ponds.
-------
To determine if activated sludge is a feasible process for high strength poultry
processing wastewaters, we operated a lab-scale activated sludge unit under
controlled conditions for several months. The system was located in a controlled
environment room and all tests were conducted at 21°C (70°F).
The main components were a reactor and a solids-liquid separator (Fig. 2). The
reactor was constructed of a 20.3 cm I.D. by 35.6 cm long clear plastic tube
and held a volume of 7 liters. Compressed air was supplied to the reactor
through a 7 mm I.D. plastic tube made into a ring with 0.8 mm holes drilled on
13 mm centers. The flowrate of the air was not measured, but was controlled by
a pressure regulator to provide sufficient oxygen for the reactor and to help
keep the mixed liquor stirred. A mechanical stirrer was also used for agitation.
The solids-liquid separator was constructed of clear plastic with a 30.5 cm by
30.5 cm top opening and a 3.2 cm by 3.2 cm bottom and held a liquid volume of
17 liters. The bottom was drilled with a 13 mm hole fitted with a plastic
tubing that served as the solids recycle line. The clarified effluent flowed
from the separator into a collection tank.
The raw wastewater was obtained, normally twice a week, from a local commercial
processing plant. The wastewater was stored in plastic containers at 2-3°C
until ready for use. The water was homogenized in a commercial blender, then
placed in glass feed bottles. Blending was necessary to keep the waste from
fouling the small plastic feed lines. Each time wastewater was obtained, it
was sampled and analyzed for COD, total and volatile solids, total and volatile
suspended solids, fat, total (Kjeldahl) nitrogen, ammonia nitrogen, nitrate
nitrogen, nitrite nitrogen, total phosphorous and pH (9).
Raw wastewater was fed into the reactor with a variable speed peristaltic pump.
The flowrate was controlled at 28 I/day (1.17 1/hr) and was pumped continuously.
The feed water was stored in glass bottles and was maintained at 2-3°C by a
refrigerated-water bath.
On initial start-up, 20 liters of activated sludge from the return line of a
3 MGD municipal activated sludge plant was used as seed for the reactor. The
system was operated for several days at low hydraulic flow rate (6-10 I/day)
to acclimate the sludge to the poultry processing wastewater.
After about 3 weeks, the feed rate, Q, was increased to 28 I/day. This feed
rate gave a hydraulic detention time, 6 , in the reactor of 6 hrs and was
maintained throughout the remainder of the study. The recycle rate, QR, was
not controlled, but normally was greater than 4 Q.
The sludge age, or mean solids residence time (0g) was the variable operating
parameter, and was controlled by wasting the solids from the reactor daily.
The mathematical expression for sludge age is
•try
es - rnr (i)
414
-------
(Jl
REACTOR
EFFLUENT
SOLIDS-LIQUID
SEPARATOR
SOLIDS RECYCLE
COMPRESSED AIR
Figure
2. Laboratory-scale activated sludge unit.
-------
where :
V = volume of the reactor
X = organism concentration in the reactor,
Q = amount of sludge wasted daily, and
X^ = organism concentration in the sludge recycle.
The accuracy of the determination of sludge age could be increased by including
the mass of organisms which are unintentionally wasted each day in the effluent.
If solids are wasted by removing liquid from the reactor (this liquid is called
the mixed liquor), instead of the recycle line, then one less solids
determination is necessary. The expression for sludge age then becomes
fl
e
s - QWX+(Q-QW)XE
where :
V = volume of the reactor,
X = MLVSS, mg/1,
Q = amount of sludge wasted daily from the reactor,
Q = hydraulic flow rate,
XE = effluent VSS, mg/1.
We selected sludge ages of 1, 2, 3.5, 7, 10 and 14 days. These ages were
accomplished by wasting solids from the reactor at the rates 1.0V, 0.5V, 0.29V,
0.14V, 0.10V and 0.07V, respectively. However, the actual sludge ages were
calculated with Eq. II and were always less than the set values.
We evaluated the performance of the system by analyzing the mixed liquor and
the effluent. The pH, dissolved oxygen and temperature of the reactor mixed
liquor were measured daily except on weekends . The mixed liquor volatile
suspended solids were determined three times per week and the total volatile
solids, once a week. COD, total volatile solids and volatile suspended solids
of the effluent were determined 3 times a week. Effluent COD was determined
on filtered samples and, therefore, the values reported are based on the soluble
material in the wastewater. Once a week a COD determination was made on an
unfiltered effluent sample for a comparison. Concentrations of fat, total
(Kjeldahl) nitrogen, ammonia-, nitrite-, and nitrate-nitrogen and phosphorous
in the effluent were measured weekly. All analyses were performed according to
current practices (9). The settling characteristics of the sludge in the
solids-liquid separator were noted after the system had reached steady state.
Only data acquired after the system had reached steady-state conditions were
used for the analysis of the system. The system was operated at least three
times the set sludge age before any steady-state data were taken. However,
this was not the only requirement for determining steady state. The system
416
-------
was determined to be in a steady-state condition when the concentration of
the mixed liquor volatile suspended solids and the calculated sludge age became
relatively constant from day-to-day; uniformity of the settling properties of
the sludge in the solids-liquid separator also was a good indicator.
RESULTS AND DISCUSSION
Waterflow measurements showed that the eviscerating room area accounted for 26%
of the total plant water use and about 33% of the process water. Analysis of
the wastewater indicated that the eviscerating room was responsible for 37% of
the total plant BOD5, 42% of the COD, 32% of the suspended solids and 90% of
the fat (Table 1). Since many processing plants have problems with fat in their
effluent, these data indicate that the eviscerating room is the major contributor
and special attention should be given to the processes in this area to reduce
the fat load in the water. Any process modification that reduces the amount of
fat entering the water will also reduce the BOD5, COD and suspended solids
concentration.
The scalder and whole-bird chiller were regulated at 0.95 and 1.9 I/bird,
respectively, which combined accounted for about 7% of the total plant water
use and about 9% of the process water use. The combined wastes from these
two areas accounted for about 9% of the total BOD5, COD, suspended solids and
fat. The amounts of water are fixed and little, if anything, can be done to
reduce the wastes generated in these areas. However, one allowable solution
is reuse of the discarded chiller water in the scalder. This may not be
economically feasible because of the cost of heating the chilled water and
because these areas combined account for only 9% of the total waste; efforts
to reduce the pollution load probably could be more profitable elsewhere.
The offal truck discharge, with a flow of only about 35 1/min, was highly
concentrated and was responsible,for about 5% of the total BOD5 and COD, 2%
of the suspended solids and nearly 3% of the fat. These data indicate that
the liquid waste portion of the offal on the truck or receiver should be
contained so that it is not released into the water. In addition, tanks could
be mounted on the offal truck to receive the blood. If pneumatic waste handling
systems are used and excess water were excluded, the offal truck could be
eliminated as a source of pollution.
The data from monitoring a processing plant's lagoon treatment system are
summarized in Table 2 and Figures 3 and 4. The average concentrations of the
wastes in the effluent from these stabilization ponds were: BODs, 18 mg/1;
COD, 103 mg/1; suspended solids, 40 mg/1; and fat, 6 mg/1. Based on kg BODs/k
kg of LWK (kg per 1000 kg of live weight killed) the mean BOD5 during the
sampling period was less than the BPT (best practicable control technology
currently available) average in the EPA proposed performance standards for
poultry processing products. The mean BOD5 exceeded the BAT (best available
technology economically achievable) average and the high BOD5 value equalled
the BAT maximum.
417
-------
Table 1. Summary Of Data For Raw Wastes From One Commercial Poultry Processing Plant
oo
Area
Eviscerating Room
Whole Bird Chiller
Gib let Chiller
Scalder
Offal Truck
Final Effluent
BOD5
mg/1
1522
830
1251
1044
7050
1116
COD
mg/1
2389
1175
1570
1678
12052
1691
TOC
mg/1
262
175
269
396
2231
216
Total
Solids
mg/1
1047
933
1216
1843
8223
999
Total
Suspended
Solids
mg/1
430
365
236
216
1286
368
Volatile
Suspended
Solids
mg/1
423
361
234
208
1260
362
Fat
mg/1
962
335
170
68
1433
266
Values are averages of 6 weekly determinations,
-------
IO
Table 2. Summary Of Data From A Commercial Poultry Processing Plant's Lagoon Waste Treatment System
BOD
Mean -1
Low
High
mg/1
18
10
34
kg/kkgLWK-/
0.32
.18
.60
mg/1
103
72
160
COD
kg/kkgLWK
1.80
1.26
2.80
Suspended Solids
mg/1
40
4
94
kg/kkgLWK
0.69
.07
1.65
Fat
mg/1
6
4
10
kg/kkgLWK
0.10
.07
.18
— Values are means of 18 determinations.
2 /
— Values are based on 0.7 MGD discharge and 90,000 broilers processed per day.
-------
o
<
-la.
I\3
o
o
z
o
o
I-
z
UJ
UJ
160
140
120
100
80
60
40
20
\/\BOD
0
JULY
AUG
31 30 31
-+++SEPT+++-OCT »U
30
NOV
Fig. 3 BODc, COD and total suspended solids effluent concentrations from
a commercial poultry processing plant's stabilization ponds.
-------
ro
IOO
o
UJ
90
Li.
UJ
O
2
UJ
(T
UJ
80
70
L/
u
31
31
30
31
JUL.Y
SEPT
30
OCT
NOV
Fig. 4 Waste removal efficiency of a commercial poultry
processing plant's stabilization ponds.
-------
For oil and grease, both the measured mean and high values were less than the
proposed performance standards for both BPT and BAT. In fact, at no time during
the sampling period did the amount of oil and grease for any one sample exceed
the proposed average value of 0.20 kg/k kg of LWK. This indicates that the
stabilization pond waste treatment system was biologically reducing the fat in
the wastewater to a level that met proposed EPA standards.
The total suspended solids in the treated effluent were significantly greater
than the values proposed by the EPA standards for either BPT or BAT. Both the
measured mean and high values for suspended solids were more than twice the
allowable BAT values. Apparently aerated lagoon waste treatment systems would
not meet the proposed standards and would not, without an extensive amount of
modification, be an acceptable waste treatment alternative for poultry processors.
Aerated lagoons are relatively efficient and maintenance free, but probably are
not the best treatment alternative. The system monitored in this study had a
pond surface area of about 3.2 ha (7.9 acres) and required a total land area of
more than 6 ha (15 acres). For processors located in rural areas, this land
might be available, but for facilities near an urban area, limited land probably
would require some other type of waste treatment system.
The data collected from the laboratory-scale activated sludge unit, after the
system had reached a steady-state condition at each set value of sludge age, are
summarized in Table 3. The effluent soluble COD was not dependent on sludge age,
6g, and COD removal averaged 93% for all values of 65 (Fig. 5). The total
suspended solids in the effluent increased with sludge age. The regression
equation (Fig. 5) shows that the concentration of suspended solids was more than
three times as great at a sludge age of 8 days as at a 0g of 1 day. However, at
the sludge age for observed optimum sludge settling characteristics, 3-3.5 days,
the concentration of suspended solids was only about 50-60% greater than that at
a sludge age of 1 day. Based on the concentration of soluble COD and total
suspended solids in the effluent, the performance of the system differed little
among sludge ages.
The efficiency of removal was more variable for fat than for COD or for total
suspended solids, but was not dependent on sludge age. Considering all sludge
ages, the activated sludge system removed 91% of the fat in the raw waste and
effectively degraded the fat in the wastewater.
The concentration of the biomass in the reactor ranged from 1208 to 4687 mg
MLVSS/1 and varied with sludge age. The average pH of the mixed liquor was
6.0-6.3. The average reactor dissolved oxygen concentrations and mixed liquor
temperatures varied, but were not dependent on sludge age (Table 3).
The activated sludge process removed an average of 88% of the total nitrogen
in the wastewater, but removal efficiencies ranged from 65 to 98%. In general,
the total nitrogen concentration in the effluent decreased with increases in
biomass concentration in the reactor and in sludge age. A plot of the steady-
state data (Fig. 6) showed that concentration of total nitrogen in the effluent
decreased exponentially with increased biomass concentration. The concentration
of phosphorous in the effluent showed just the opposite trend. Phosphorous
422
-------
Table 3. Summary Of Steady-State Data Collected From The Laboratory-Scale Activated Sludge System
Solids Wasting
Rate
I/day
IV
0.5V
.29V
.14V
.10V
.07V
ro
to
Solids Wasting
Rate
I/day
.IV
0.5V
.29V
.14V
.10V
.07V
Sludge
Age
Day
0.98
1.93
3.26
5.91
8.11
8.25
Total
Inf.
mg/1
75
73
82
89
94
86
Biomass
Reactor Conditions
Concentration
MLVSS, mg/1 pH
1208
2248
2306
3120
4687
2271
(Kjeldahl) N
Eff.
mg/1
7
10
7
3
2
30
6.0
6.2
6.2
6.0
6.2
6.3
D.O.
mg/1
7.6
5.8
5.2
5.8
4.4
6.3
Ammonia-N
Inf.
mg/1
2.3
1.6
1.7
2.0
1.3
3.6
Eff.
mg/1
5.5
3.1
3.0
2.0
7.2
1.8
Temperature
°C
20.0
20.3
20.8
20.1
20.5
19.9
Nitrate-N
Inf. Eff.
mg/1 mg/1
0.8 8.8
.1 0.6
.5 10.6
.8 9.2
.4 5.9
.2 1.8
COD
Inf.
mg/1
1126
1097
905
1216
1582
983
Eff.
mg/1
68
119
39
76
116
65
Nitrite-N
Inf.
mg/1
0.04
.01
.04
.05
.05
.03
Eff.
mg/1
0.2
.1
.4
.7
.6
1.1
T.S
Inf.
mg/1
355
320
297
292
438
248
.s.
Eff.
mg/1
8
17
15
26
30
31
Fat
Inf . Eff .
mg/1 mg/1
174 5
178 5
246 5
80 23
154 6
109 13
Phosphorous
Inf.
mg/1
4.2
5.1
5.8
7.1
8.1
4.6
Eff.
mg/1
0.9
2.0
2.9
3.2
4.6
3.0
-------
ro
V)
CD
LU
cr
LU
o
a:
UJ
a.
15
10
TSS
COD
TSS
o
C
OD
S/S0 =0.0234-0.0094 9S
R2=0.70
I
I
I
I
I
23456
SLUDGE AGE, 6S, DAYS
8
Fig. 5 Percent effluent COD and total suspended solids concentration, S/S as a
function of sludge age, 85> in the laboratory-scale activated sludge unit.
-------
ro
01
.70
<% .60
CD" .50
< .40
LJ
* .30
o
.20
.10
p
TKN
S/S0= -1. 55 + 0.25 LN(X)
= 0.85
S/S0=0.25 EXP (-0.00054X)
R2=0.72
o
I
I
1000 2000 3000 4000 5000
BIOMASS CONCENTRATION (MLVSS), X, MG/L
Fig. 6 Fraction of phosphorous and total nitrogen remaining
in the effluent as a function of biomass concentration
for the laboratory-scale activated sludge unit.
-------
concentration increased logarithmically with biomass concentration and sludge
age (Figs. 6 and 7) and ranged from a low of 0.9 mg P/l at 1208 mg MLVSS/1 to
a high of 4.6 mg P/l at 4687 mg MLVSS/1.
The concentrations of ammonia-nitrogen, nitrate-nitrogen and nitrite-nitrogen
were all higher in the effluent than in the untreated wastewater. The ammonia-
nitrogen concentration in the effluent was about double that of the untreated
water. The average nitrate- and nitrite-nitrogen concentrations increased to
about 12-15 fold those of untreated waste.
Based on overall performance, at the sludge ages for optimum sludge settling
characteristics (2-3.5 days), the activated sludge system could be expected
to reduce concentration of total nitrogen about 90% and of phosphorous about
50-60%. However, at this operating condition, the ammonia-nitrogen concentra-
tion would double, nitrate-nitrogen would increase 12-15 times and nitrite-
nitrogen in the effluent about 10 times the concentration of the untreated
waste.
The kinetic coefficients of the activated sludge system for poultry processing
waste were also determined. Trial-and-error solutions for the Monod equation
(Fig. 8) showed that the specific growth rate, y, was between 1.5 and 2.0 day"-'-,
and Kg, the saturation coefficient, was between 20 and 24 mg COD/1. However,
the specific substrate utilization rate, U, was essentially independent of
substrate COD concentration and is best expressed as a constant value for all
sludge ages of 1.5 mg COD per day per mg MLVSS. This constant indicates that
the activated sludge system was capable of daily utilizing or degrading 1.5 gm
of soluble COD per gm of mixed liquor volatile suspended solids in the reactor.
At a sludge age of 3.26 days, the system was capable of utilizing daily about
3.5 gm of COD per liter of reactor volume.
A regression analysis on the data showed that the observed yield declined
exponentially with sludge age (Fig. 9). The yield curve indicates
that daily yield was between 0.3 and 0.1' mg MLVSS per mg of soluble COD
removed from the wastewater.
A regression analysis on the growth rate, 1/63, plotted against the specific
utilization rate, U, (Fig. 10) shows that the yield was 0.39 day~l (mg COD/day
per mg MLVSS) and the specific organism decay rate, k^, was 0.34 day"*. Based
on this analysis the specific organism decay rate was about one-fifth of the
specific growth rate.
In summary, this laboratory-scale study showed that the conventional activated
sludge process is an effective method of treating poultry processing wastewater.
At a hydraulic detention time of 6 hours, the system removed 93% of the soluble
COD and more than 90% of the suspended solids and fat. The process reduced the
concentration of the total nitrogen and phosphorous, but significantly increased
the concentration of ammonia-, nitrate-and nitrite-nitrogen. The sludge age had
a notable effect on the operation of the system and sludge settling characteristics
were best at sludge ages of 2-3.5 days.
426
-------
ro
.70
co -60
^.
CO
o" -50
| -40
W r,^
o: .30
.20
o: .10
o
0
AR
® TKN
o
I
1
i
L
23456
SLUDGE AGE,6S, DAYS
8
Fig. 7 Fraction of phosphorous and total nitrogen remaining in the effluent
as a function of sludge age for the laboratory-scale activated sludge unit.
-------
IN3
C»
r: en o.v
^^
°s 3.0
(9 O
UJ
* 2.0
_j
<^J
^^
!S
UJ
-------
CO
CO
o
IU
Id
K
O
O
O
CJ>
O
UJ
O
UJ
a:
UJ
en
CD
o
0.4
0.3
0.2
O.I
— e
o
I
I
Y0 = 0.38EXP (-O.I9es)
= 0.95
I
I
I
23456
SLUDGE AGE,0S,DAYS
I I
Fig. 9 Observed yield, YQ, as a function of sludge age, 6S, for
the laboratory-scale activated sludge unit.
8
-------
CO
o
1.0
I
- 0.8
0>
iu 0.6
jE 0.4
0.2
o
o:
H
LU
1/08 = 0.39 U-0.34
R2=0.89
G
1
I
I 2 3
SUBSTRATE REMOVAL RATE,U, DAY'1
3.5
Fig. 10 Relationship between sludge age, 6S, and the substrate utilization
rate, U, for the laboratory-scale activated sludge unit.
-------
REFERENCES
1. U. S. DEPARTMENT OF AGRICULTURE. Agricultural Statistics 1974 (1974).
2. FEDERAL REGISTER. Environmental Protection Agency. Poultry Processing
Products. Proposed Performance and Pretreatment Standards. Vol. 40,
No. 80, Part III, April 24, 1975.
3. CARAWAN, R. E., W. M. CROSSWHITE, J. A. MACON and B. K. HAWKINS. Water
and Waste Management in Poultry Processing. EPA-660/2-74-031 (1974).
4. KERNS, W. R. and F. J. HOLEMO. Cost of Waste Water Pollution Abatement
in Poultry Processing and Rendering Plants in Georgia. ERC-0673.
University of Georgia (1973).
5. REID, R. J., J. P. PILNEY, R. J. PARNOW and E. E. ERICKSON. Development
Document for Effluent Limitations Guidelines and Standards of Performance
for the Poultry Processing Industry. North Star Research Institute.
EPA Contract Number 68-01-0593 (1974).
6. HAMM, D. Characteristics of effluents in ten southeastern poultry
processing plants. Poultry Sci. 51: 825-829 (1972).
7. ALLWOOD, J. K. and R. J. COLEMAN. Analysis of the Economic Impact of
Proposed Effluent Limitation Guidelines for the Poultry Meat Processing
Industry. EPA-230/1-74-040 (1974).
8. VERTREES, J. G. The Poultry Processing Industry: A Study of the Impact
of Water Pollution Control Costs. Marketing Research Report No. 965,
U. S. Dept. Agr. (1972).
9. AMERICAN PUBLIC HEALTH ASSOCIATION. Standard Methods for the Examination
of Water and Wastewater. American Public Health Assoc., New York, N. Y.
(1971).
431
-------
A MEAT PACKER'S SOLUTION TO MEETING
1983 EFFLUENT REQUIREMENTS
by
Joseph A. Home!, Jr., P.E.* and Jack McVaugh, P.E.**
INTRODUCTION
The 1983 National Pollution Discharge Elimination Standards (NPDES)
discharge standards will force dischargeers of industrial wastes to perform
virtually complete wastewater cleanup, paving the way toward total water
reuse. For a meat packer to meet these standards for stream discharge, bio-
chemical oxygen demand (BOD,-), suspended solids (SS), grease, and nutrients
may have to be reduced in some cases by 99+%.
Many meat packers in urban areas have the opportunity to shift part
of the responsibility for meeting 1983 standards to the local municipality.
The municipality agrees to accept the wastewater and to comply with the state
regulatory authority's treatment standards. The packing company essentially
agrees to pay a hookup and user fee and promises to remove substances from
the wastewater which would otherwise cause damage to the sewer and treat-
ment plant, or that would upset the treatment process.
For other meat packers, however, it is neither economical nor practical
to discharge into a municipal sewer system. These packers must build, maintain
and operate their own primary, secondary, and tertiary treatment systems.
Ideally, the common objective of treatment plant design is to incorporate
a series of complimentary processes, which are economical with regard to
capital, operating and maintenance costs. Each process must not only be
designed to operate at the desired level of efficiency during normal operation,
but must contain sufficient conservativeness in design to "take up the slack"
should the preceding process or processes fail.
The most critical process may then be the last major treatment unit in
the sequence - the "polishing" unit. At present, however, perhaps the least
amount of attention is given to this unit. An often used polishing unit is
the detention pond. Wisconsin regulations, as well as those of several other
states, require that biological secondary treatment systems be followed by
detention ponds of up to 30 days capacity. Because of the high nutrient
content of their wastewaters, however, many meat packers have discovered
that these ponds often generate their own waste load in the form of algae.
*Vice President, Foth & Van Dyke & Associates, Inc., Green Bay, Wisconsin.
**Industrial Process Engineer, Envirex Inc., Waukesha, Wisconsin.
432
-------
This report describes the design and operating experience of a meat
packing plant waste treatment system. The system is currently meeting present
permit requirements and investigations are ongoing which will allow the eff-
luent to meet 1983 standards. Each process in the treatment sequence was
designed to accept surge loadings, and the final polishing step has shown
promise in overcoming some of the disadvantages of the detention pond.
BACKGROUND
Hillshire Farm Inc., a division of Consolidated Foods Inc., operates
a meat packing plant in New London, Wisconsin. The plant presently slaughters
approximately 1500 hogs/day, and conducts complex processing of the hog car-
casses as well as purchased beef carcasses.
Early in 1972, the State of Wisconsin Department of Natural Resources
issued a statement requiring Hill shire to reduce its wastewater BODc and SS
concentration to maximums of 35 mg/1 (a 90+% reduction), and reduce total
phosphorus by 85%. At that time, wastewater treatment consisted of collection
of kill floor, processing plant, and domestic waste, and passage through a
septic tank system.
Discharge to the municipal waste treatment system was first investigated
as a possible means of waste disposal. In order to discharge to the city's
system, first the New London city limits would have had to be extended outward
to incorporate the plant site; and secondly, a pump station as well as
extension and enlargement of the main interceptor sewer would have been re-
quired to handle the significant increase in flow. Total cost (1972 dollars)
for the lift station and interceptor construction was estimated to range
from $680,000 to $750,000 and was expected to be financed by Hillshire
Farm. Moreover, the company would have realized increased 1972 property taxes
in excess of $40,000 by incorporation with the city, and yearly sewage
surcharge fees would have approximated $25,000 (1972 dollars).
In the spring of 1972, a consulting engineer was retained to recommend
possible methods of onsite wastewater treatment for stream discharge. At
that time the flow was measured, and working closely with the plant management,
future flows were projected. The measured flow rates, as well as projected
flow rates which were used for plant design, are compared with present-day
flow rates in Table 1.
TABLE 1. MEASURED (1972), PROJECTED, AND PRESENT (1976) RAW WASTEWATER
FLOW RATES (6PM)
Measured Projected Present
Average 173 254 230
Peak 273 400 400
433
-------
A wastewater sampling program was carried out on raw wastewater,
septic tank effluent, and on grab samples taken from various locations along
the creek into which the wastewater flows. (The septic tank was the only
existing means of treating the domestic and industrial waste waters in
1972.) The stream winds through marshy farmland and residential areas before
discharging into the Wolf River. Results of analyses conducted on these
samples are shown in Table 2.
TABLE 2. RESULTS OF ANALYSIS OF SAMPLES TAKEN 5-3-72 (mg/1)
Location
Septic Tank Influent
Septic Tank Effluent
150 yds. Downstream
2.5 mi. Downstream
5.0 Downstream
*12 hr. composite samples
t24 hr. composite sample
m
735*
390*
300
16
5
SS
513
373
260
52
25
Total P
12.25
12.5
10.75
11.25
11.5
Grease
548*
130t
33
„
—
Composites were also made of 12 hr. composite "workday" samples with
24 hr. composites which included production plus cleanup periods. Table
3 shows the results of analyses on composite samples of septic tank influent
and effluent.
TABLE 3. RESULTS OF ANALYSIS OF SEPTIC TANK COMPOSITE SAMPLES (mg/1)
12-hr, "workday" comp.
BODC
SS 5
Total P
Grease
Influent
735
590
157
548
Effluent
390
500
27.0
— -
24-hr.
Influent
545
485
94.9
-—
comp.
Effluent
368
475
25.8
130
SELECTION OF UNIT PROCESSES
The present wastewater treatment plant at Hillshire Farm consists of
a rotary screen on kill flow wastes, a lift station, dissolved-air flotation
434
-------
complete mix activated sludge, chlorine contact, and final detention in a
pond filled with marsh vegetation. Because of the difficulties in separating
various process sewer lines in the older sections of the plant, the various
streams are combined before being pumped into the flotation unit. The streams
emanate from the kill floor, processing area, washrooms, garage and main-
tenance area. Blood is recovered, hides and renderable material are processed
offsite, and stormwater runoff does not enter the treatment system.
The following is a discussion of the selection criteria and design
of each of the unit processes. Figure 1 shows the process flow scheme. <
Screening
In meat packing plant wastewater the greatest source of gross suspended
solids consists of fleshings and paunch from the kill floor. Because of the
large quantity of marketable by-product to be removed from the wastewater
at this location, a rotary drum screen had previously been installed on the
kill floor wastewater stream. The drum is 6 ft. diameter with a 10 ft.
face width, and is covered with a 20 mesh screen. Lesser amounts of gross
solids are collected in wire baskets in each of the floor drains through-
out the processing area.
Pumping
Pumping is often required to introduce wastewater into a treatment plant
with several unit processes. Once pumped, the main process stream flows
by gravity through each of the unit processes and into the receiving stream.
A wastewater lift station usually consists of a wet well or sump, into which
the sewer or sewers discharge, and a dry well containing several pumps.
A level control device in the wet well commands each of the pumps to start
and stop, maintaining the wastewater level between maximum and minimum values.
The configuration of flow leaving the lift station, then, is characterized
by surges, the significance of which is very great for two or three pump
lift stations. These sudden surges may have a disturbing effect on the unit
process immediately following.
At the Hillshire Farm treatment plant, the unit process into which the
wastewater is pumped is the dissolved-air flotation unit. Although designed
to provide maximum hyudraulic stability, the unit performs best when not subject
to intense hydraulic surges. A large equalization basin, out of which the
collected wastewater is pumped at a constant rate, has been used to equalize
many types of industrial wastes. However, packing plant wastes would require
intense mixing to prevent separation of grease and settleable solids in areas
where they cannot be mechanically removed.
A compromise plan was reached through the use of variable speed pump,
operating from a proportional level control signal in a large sump. By
automatically matching the pumping rate to the rate at which wastewater enters
the sump, the system takes advantage of the flow rate dampening effect of
the sewer lines and the diverse schedule of water usage inherent in a large
complex packing plant.
435
-------
GENERAL FLOW DIAGRAM
HILLSHIRE FARM COMPANY
NEW LONDON , WISCONSIN
HILLSHIRE
FARM
COMPANY
WASTEWATER FLOW
R.A.S.
W.A.S.
FIGURE 1 - GENERAL FLOW DIAGRAM
436
-------
The lift station consists of two variable speed horizontal trash pumps
of recessed impeller design. The control system consists of an SCR type
infinitely variable speed controller commanded by a bubble tube level
sensing device in the wet well.
Pissolved-Air Flotation
Primary treatment at Hi 11 shire Farm is accomplished by dissolved-air
flotation. Its purpose is to remove free grease and suspended solids
which would only slowly degrade during subsequent biological treatment.
The initial design was approached in two ways, 1) being a unit without
chemical addition or flocculation, intended to remove all free grease and
50% of the SS, and 2) being a unit with chemical addition and flocculation for
removal of essentially all free and emulsified grease and up to 90% of the SS.
the former design approach was chosen, with facilities reserved, however, for
the future addition of chemical feed equipment.
The unit installed is of rectangular deesign with reaction-jet dished -
baffle influent distribution and energy dissipation device. This influent
arrangement distributes the flow over the full width and depth of the flota-
tion compartment, thereby maximizing hydraulic stability through plug-flow
configuration. The unit consists of a steel tank 8' wide by 36' long with
4' of effective water depth. The tank has a hopper bottom with independently
operated surface skimmer and longitudinal screw bottom sludge collector.
Skimmings and sludge flow by gravity into a common sump, where they are
pumped by means of heavy-duty positive displacement pumps into a sludge storage
tank. The combined sludge then awaits land spreading.
The flotation unit uses the recycle pressurization method of inducing
dissolved air into the waste stream, thus maximizing controllability and power
usage efficiency. The unit is sized for a peak raw flow of 400 GPM and
a recycle flow of up to 150 GPM, pressurized to 40 psig. At this peak flow
rate, the tank provides an effective detention time of approximately 14
minutes.
Secondary Treatment
Secondary treatment was obviously required to meet the effluent quality
needs for stream discharge. Physical-chemical means, including activated
carbon were briefly surveyed, but the costs indicated by the results of a
bench scale testing program were prohibitive. Likewise, land spreading of
the primary effluent was ruled out because of the incomplete technology of
the day, and because of the cold Wisconsin climate.
Because of the successful documented operating experiences, biological
treatment was deemed the most practical means of accomplishing secondary
treatment. An economic evaluation of four candidate methods was conducted,
the results of which are summarized in Table 4. Also, design criteria and a
cost summary for each of the competing candidate processes are presented
in the appendix. Processes considered included rotating disc contactor,
contact stabilization, anaerobic-aerated lagooning, and complete mix activated
437
-------
sludge. Capital costs included those of the secondary treatment process and
a holding pond, with detention time as dependent on the process which preceded
it and equalled 10 days for all candidates except the anaerobic-aerated lagoon
system. That process required a 30 day detention time.
TABLE 4. CAPITAL AND OPERATING COST AND MANPOWER ESTIMATES FOR BIOLOGICAL
TREATMENT METHODS. (1972 DOLLARS) ADD AIR FLOTATION AND CHLORINA-
TION COST WHICH IS COMMON TO ALL ALTERNATES
Rotating Disc Contactor
Contact Stabilization
Anaer-Aerated
Activated Sludge
Capital Cost
Add Air Flotation
($)
197,190
17Q.880
200,663
122,428
Operating
Cost
($/yr)
7,800
12,400
12,700
12,400
Manpower
Required
(manhours/day)
3
5
3
5
The activated sludge method was selected because of low capital invest-
ment. When considering both capital and operating costs, the rotating disc
contactor method appeared attractive, but could not be recommended without
extensive pilot testing to verify design assumptions. The use of anaerobic
lagoons was ruled out, not only because of price but because of the potential
for odor in the surrounding community.
The activated sludge system consists of a 275,000 gallon epoxy coated •
concrete aeration basin with diffused air furnished through a fixed header
system. Because of the basin configuration and inlet arrangement, the system
is considered to be completely mixed. Air is supplied to the 78 diffusers
at a rate of 14 CFM/diffuser by two 100 HP centrifugal blowers. At a BOD.
removal rate of 1000#/day and the design mixed liquor volatile suspended solids
(MLVSS) concentration of 3000 mg/1, the foodrmass (F:M) ratio is calculated
to be .15. A 30 foot diameter final clarifier with scraper type sludge
collection mechanism and skimmer is used to clarify the mixed liquor.
Sludge collected in the clarifier sludge sump is pumped at a maximum rate of
275,000 GPD, to a splitter box after which it is returned to the aeration
basin.
Excessive activated sludge may be wasted from the splitter box into
the sludge storage tank by opening a gate valve and permitting gravity flow.
The waste activated sludge then joins the air flotation scum and sludge in
the sludge storage tank to await land spreading. The storage tank is
10 ft. diameter x 19 ft. side water depth, and is equipped with the necessary
gas safety devices such as pressure relief valve, vacuum breaker and flame
resistor.
438
-------
Chi on'nation System
The state regulatory agency requires chlorination wherever pathogenic
organisms may be contained in the wastewater stream. In the Hillshire Farm
system, a baffled chlorine contact chamber follows secondary clarification.
The chamber is sized to provide approximately 20 minutes detention time at
the peak flow of 400 GPM.
A manually controlled, vacuum-type, tank mounted chlorinator is used
to deliver up to 60 Ib/day chlorine into the contact tank.
Control Building
A concrete block building is used to house the lift pumps and SCR control
system, rotary blowers, dissolved-air flotation unit, nominal laboratory
facilities, and provides a future site for chemical feed equipment. Although
in the initial design it was not intended to house the flotation unit, it
was felt that the maintenance benefit of enclosure would outweigh the additional
building costs.
Effluent Polishing
The effluent polishing unit in the Hillshire Farm treatment system
consists of a detention pond filled with marsh vegetation. Observing how
effective the wastewater was treated by the existing stream and marsh vegeta-
tion originally located on the plant property (See "150 yards downstream"
in Table 2), some interest was generated in exploring this method as a means
of secondary treatment. A demonstration grant was sought in order to explore
the concept on a four season basis, as treatment efficiency was suspected
to be sufficient at low winter temperatures. Demonstration grant funds
were not available however, and thought was abandoned as to using this method
for secondary treatment. Effluent polishing was a different factor, however.
At the time of process selection at Hillshire, several studies on bull-
rushes for wastewater treatment were being proposed in this country,
apparently inspired by some successful pilot work on phenolic wastes in
Germany. In that country, the Max-Planck Gesellschaft and Gottingen Uni-
versity had spent several years researching the mechanism by which higher
plants, such as the bull-rush Scirpus locustris heterotrophically metabolized
organic material. Even toxic organics, such as phenols, were tolerated
and metabolized in high concentrations according to tracer studies. A
bactericidal effect was also attributed to the bull-rushes, allegedly through
the production of organic "phytoncides". The research resulted in a pilot
study, using a cascade growth of bull-rushes to further treat municipal secon-
dary effluent. The study indicated virtually complete BOD,, and SS removal
at loading of 7.5 GPD of wastewater per ft* of bull-rush pond. Removals
during winter conditions were reduced by approximately 40%.!»^»3
Interest in the process led to construction of a marsh growth vegeta-
tion pond downstream of the chlorine contact tank. The pond is essentially
an unlined earthen lagoon approximately 220 ft. square and 10 ft. in depth
containing a natural assortment of higher plants native to the locale.
439
-------
Water level in the pond is controlled by a concrete block effluent structure,
which directs the final discharge into a cast iron pipe.
Some of the marsh grass species growing in the pond have been identified
as Common Millein (Verbascum thapsus), Cattail (Typha latifolia), and
grasses of the family Graminear.
UNIT PROCESS PERFORMANCE EVALUATION
A sampling program and study was undertaken recently to determine the
effectiveness of each unit process in removing BOD,., SS, oil and grease,
and phosphorus from Hillshire Farm wastewater. The sampling consisted of
grab and composite samples taken over the period between October, 1974
and January, 1976. After collection, each sample was packed in ice and
transported within three hours to the consultant's lab for analysis. Veri-
fication of results was often made by two independent laboratories.
Table 5 summarizes the average values of the wastewater characteristics
measured during the sampling program, and indicates removal efficiencies
of primary, secondary, and marsh vegetation treatment. Influent samples
were taken from the influent wet well just prior to pumping. Primary and
secondary effluent samples withdrawn from the air flotation and final clarifier
effluent wet wells, respectively, and marsh effluent from near the effluent
structure in the marsh growth vegetation pond.
TABLE 5. AVERAGE VALUES OF WASTEWATER PARAMETERS (mg/1) AND PROCESS
PERFORMANCE EFFICIENCIES
Primary Secondary Marsh
Influent Effluent Effluent Effluent
BOD5 800 (24%) 610 (98%) 10 (30%) 7 (99% overall)
SS 480 (40%) 290 (95%) 14 (42%) 8 (98% overall)
Phosphorus (as P) 24(38%) 15 (53%) 7 (0%) 7 (70% overall)
Oil and Grease 261 (70%) 80 6 (98% overall)
Table 6 shows average concentrations of solids and oil and grease in
the various sludges throughout the plant.
440
-------
TABLE 6. AVERAGE SOLIDS CONCENTRATIONS (% BY HEIGHT)
Air Flotation Air Flotation Mixed
Scum Sludge R.A.S. Liquor
Total Solids
Suspended Solids
Oil and Grease
5.2
4.6
1.1
4.1
2.4
.2
18
.7
14
.3
Influent Pumping
The effect of the variable speed influent pumping on the unit process
following could not easily be evaluated, since the SCR control system cannot
be altered to pump in the conventional "on-off" manner. However, data was
generated to determine how effectively the pumping rate matched the flow
rate into the influent wet well.
Readings from the plant's nozzle-type flow meter, installed ahead of
the wet well, were compared to head-over-weir measurements made on the
air flotation unit's straight edge effluent weir. Of course, the flotation
pressurized recycle flow rate, measured by a differential pressure orifice
plate and manometer on the recycle line, was subtracted from the flow rate
value calculated by the head-over-weir measurement. Results showed only a
slight dampening effect between the flow meter nozzle and the air flotation
unit effluent weir. This could easily be attributed to inaccuracies in
head-over-weir measurement, and to the dampening effect of the air flotation
unit itself.
It was therefore concluded that the SCR control system paces the
pumping flow rate very closely with the influent flow rate.
Dissolved-Air Flotation Unit
Table 5 indicates average BOD^, SS and oil and grease removals of
24%, 40% and 70% respectively. Moreover, floated solids are thickened to
a concentration averaging 5.2% by weight. Scum solids concentrations are
maximized by operating the skimmer on a timed cycle. At present the skimmer
operates for 5 minutes and is off for 10 minutes. The bottom sludge screw
conveyor is operated manually once or twice per day just before and during
flotation sludge wasting. The unit appears to be operating efficiently
although it is operating at peak capacity.
Secondary Treatment
Table 5 indicates that the activated sludge system is producing a high
quality effluent, with average BOD5 and SS concentrations in compliance
441
-------
with most state standards for stream discharge. BOD5 and SS concentrations
were 10 and 14 mg/1, respectively, while the effluent phosphorus concentra-
tion (as P) was 7 mg/1. The 70% phosphorus removal fell short of the NPDES
requirement for 85% removal. Sufficient data on oil and grease removal was
not available to establish a meaningful average.
Analysis of mixed liquor samples, as shown on Table 6, indicates that
the mixed liquor suspended solids (MLSS) is being maintained at the design
level of 3000 mg/1 and that the return activated sludge (R.A.S.) SS concen-
tration averages 7000 mg/1. At the present day average flow of 330,000
GPD, an F:M ratio of .24 is calculated (based on MLSS concentration, BOD5 =
610 mg/1). At this F:M, a good settling biological floe is produced and the
clarifier effluent is clear. Infrequent upsets to the system do occur but
are usually corrected within a few days.
Effluent Polishing
Analyses of grab samples taken near the effluent structure of the
marsh growth vegetation pond, revealed average BODg, SS, and grease concen-
trations of 7, 8, and 6 mg/1, respectively. Corresponding removals for
these substances across the pond were 30% and 42% for the BOD and SS, respectively.
The marsh vegetation apparently had no significant effect on phosphorus removal.
Table 7 presents the analytical results of the pond effluent semimonthly
sampling program over the period between November, 1974 and December, 1975.
BODj. and oil and grease analyses were conducted in the consultant's laboratory
witn frequent verification by two independent laboratories. SS analyses
were conducted at the Hi 11 shire Farm plant site. It is difficult to see any
significant trends in effluent quality which can be related to cold weather,
but higher values of BODg, SS and oil and grease were evident in November,
1974 through March, 1975 samples. (See Table 7). If truly temperature
related, such a trend should then repeat itself in 1976. Nonetheless, all
effluent parameters were in compliance with the existing NPDES permit.
Moreover, at no time of the year did effluent SS concentrations indicate a
significant production of algae, which would be expected in a conventional
detention pond richly supplied with nutrients.
PHOSPHORUS REMOVAL STUDY
In order to meet Wisconsin's stream discharge requirements in 1977,
Hi 11 shire Farm must increase its phosphorus, removal levels from the present
70% to 85%. An intensive literature search by Hi 11 shire Farm's consultant
revealed that of the three inorganic chemicals commonly used for phosphorus
removal, lime, alum, and ferric chloride, researchers in general found the
ferric salt to be the most effective.
The various studies also indicated three possible addition point loca-
tions for the inorganic and polyelectrolyte coagulant aid. One possible
location will be the pipeline or influent wet well preceding the air flota-
tion unit. Another possibility will be at the head end of the aeration basin,
and the third will be approximately two-thirds of the way down its length.
442
-------
TABLE 7. MARSH GROWTH VEGETATION POND EFFLUENT QUALITY (mg/1)
Date
11-7-74
1-10-75
1-31-75
2-14-75
2-28-75
3-14-75
3-31-75
4-14-75
4-25-75
5-29-75
6-4-75
6-20-75
7-1-75
7-18-75
7-30-75
8-15-75
8-29-75
9-17-75
9-26-75
10-16-75
10-28-75
11-19-75
11-26-75
12-29-75
12-31-75
BODC
b
39
19
9
10.5
15
14.5
10
4.5
12
23
30
10
6.8
6.1
2.4
1.6
4.5
2.1
3
12.9
3.1
8
4.2
3
3
SS
53
20
8
34
1
3.2
15.2
8.4
16.4
31
—
--
5.6
3
34
9
—
__
7.2
--
3
4.4
7.5
—
— —
Oil & Grease
9.5
9.9
14.7
5.6
8.9
10
4.3
7.9
5.7
4.7
1.6
2.5
5
3.5
1.8
2
1.9
5.8
5.2
5.4
2.4
5.4
2.1
Bench scale coagulation studies on Hillshire Farm's clarifier effluent
indicated that the state's requirements could be met by the addition of
150 mg/1 of ferric chloride and 0.5 mg/1 of an anionic polyelectrolyte.
Chemical coagulation tests and bench scale dissolved-air flotation tests
were also conducted on raw wastewater to estimate the potential for BOD5 and
phosphorus removals in the air flotation unit.
Preliminary results from flotation test work were inconclusive. Because
of the apparently low alkalinity of the raw wastewater, both ferric chloride
and alum in dosages between 100 mg/1 and 300 mg/1 depressed the pH below
5.0. This is the commonly accepted lower limit for the optimum coagulation
of the hydroxides of aluminum and iron. This indicates, therefore, that
on a full scale basis a three-chemical feed system - coagulant, polyelectro-
lyte, and possibly lime for pH adjustment, would be required for chemical
addition ahead of the air flotation unit. Bench scale flotation test re-
sults are shown in Table 8 using 200 mg/1 alum and 2 mg/1 of an anionic
polyelectrolyte* at a pH of 4.8. The test was run with 50% recycle and a
*Betz 1120A; Betz Laboratories, Inc., Trevose, Pennsylvania
\
443
-------
detention time of 7 minutes. The bench scale apparatus is sketched in Figure
2. Although an aluminum hydroxide floe had been formed, colloidal grease
particles and blood cells had apparently not been adsorbed onto the floe
and removed efficiently. More experimentation will be required to determine
the optimum chemical, dosage, and pH.
TABLE 8. RESULTS OF BENCH SCALE DISSOLVED-AIR FLOTATION TESTS WITH AND
WITHOUT CHEMICAL COAGULATION
BOD5
ss
Phosphorus
Oil and Grease
Raw Flotation
Influent w/o Chemical
720 585
233 110
2
110 23
Flotation
w/ Chemical
490
95
—
14
CAPITAL AND OPERATING COSTS
The Hillshire Farm wastewater treatment plant was installed in 1973
at a total cost of $350,000. Annual operating cost in 1975, including
electricity, chlorine, 40 man-hours per week for operation and less than
5 extra man-hours per week for maintenance totalled $25,000.
CONCLUSIONS
This report describes the design and operation of a meat packing plant
wastewater treatment system, discharging its effluent into a small stream.
In this instance, the packing plant realized capital cost savings of up to
$400,000 and yearly expense savings of up to $40,000 by installing their own
treatment system, rather than discharging to the local municipal sewer system.
The treatment system is producing effluent of a quality greater than
the discharge permit requires. Analyses of each of the unit processes indicates
that each are performing well under maximum design loadings. In addition
to the present BODc, SS, and oil and grease removals, increased phosphorus
removal appears feasible through the addition of ferric chloride to the
aeration basin.
444
-------
-Pressure Ciouge
Volve^
Compressed \
Ai, U. J^
Air *——— — «•>*<
Source Nut &
t
Air *\r*flricr -^
T
1
1
Pressure Chombsr !
Support Assembly-\^ j
f
"*\ \V
is
j
-
.. ''
•f==-
f- :
1 ' "
1 ; ;(
', ' 1
f
: i:
i~T^:
:'.::: i
i
%°
H'A
ijx-.f
.
t .
' •
•' :
:. :
F.;<^
1
? 5
/-Air Bleed
Vo IvP
\
srj?
t
I v
n
-IOOO-
-5OO—
(
M^ttiMr
Groduole Cylinder
(For Mixing Woste-
Woter ond Sludge
V/ith Pressurized
Liquid)
Pressure Chamber
(1.5-2 Lifer Capacity)
FIGURE 2
BENCH SCALE DISSOLVED-AIR FLOTATION TEST APPARATUS
445
-------
REFERENCES
1. Czerwenka, W., Seidel, K., "New methods for goundwater enrlchemnt in
Krefeld." GWFJ06 (30), 828-833, (1965).
2. Foth & Van Dyke and Associates, Preliminary Engineering Report for
Wastewater Treatment Facilities at the Quality Packing House, Inc.', in
New London, Wisconsin, Green Bay, June 1972.
3. Althaus, H., "Biological waste treatment with bullrushes", GWF, 107,
(18), 486-488, (1966).
APPENDIX
ALTERNATE NO. 1
BIO DISC PLANT
I. Bio Disc Design Conditions (the same for all alternates)
Design flow-275,000 GPD (The owner provided the projected future packing
house production.)
12 hour work day flow-250,000 GPD
Design average flow-254 GPM
Design peak flow-400 GPM
Raw waste contain: (before air flotation)
735 mg/1 B.O.D. 1685 #/da
2050 mg/1 C.O.D. 4702 #/da
13.0 mg/1 Total "P"
68.4 mg/1 Organic Nitrogen
4800 mg/1 Total Solids 11,009 #/da
590 mg/1 Total Suspended Solids 1353 #/da
Assuming we receive a reduction of 40% in B.O.D. and 50% in suspended
solids from the air flotation unit, we shall design the secondary unit
for the following:
Design flow-275,000 GPD
12 hour work day flow-250,000 GPD
Design average flow-254 GPM
Design peak flow-400 GPM
The primary waste effluent will contain:
441 mg/1 B.O.D. - 1011 #/da B.O.D.
1230 mg/1 C.O.D. - 2821 #/da C.O.D.
2400 mg/1 Total Solids - 5504 #/da Total Solids
295 mg/1 Total Suspended Solids - 676 #/da Total Suspended Solids
446
-------
Less than 100 mg/1 Grease - 229 #/da Grease
Required Effluent 35 mg/1 B.O.D.
25 mg/1 S.S. (assumed)
2 mg/1 Total Phosphorus
Less than 100 mg/1 Total Grease
Assuming we receive 40% B.O.D. removal through air flotation, 93% removal
through bio disc system will be required to meet effluent standards set forth
by the Department of Natural Resources.
II. Bio Disc System with Clarifier Capital Cost (1972 Dollars)
Sub-Total (Add Air Flotation Cost) $197,190.00
No enclosure included for the above equipment.
III. Bio Disc System with Clarifier Estimated Operating Costs (1972 Dollars)
Annual power costs (est. connected HP 52) $ 6,812.00
(Based on buying power for $0.02 per KWH)
In addition the system would require approximately three hours of
maintenance per day.
Chemical cost (chlorine) $2.75 per day (Estimated)
ALTERNATE NO. 2
CONTACT STABILIZATION PLANT
I. Design Conditions (Also see design conditions for Alternate No. 1)
Raw B.O.D. is 1685 pounds per day less 40% B.O.D. removal in the air
flotation unit, leaving 1011 pounds per day B.O.D. in the primary
effluent. Therefore, a B.O.D. removal of 93% is required through the
secondary portion of the plant to meet the required effluent standards.
II. Contact Stabilization Plant Capital Cost (1972 Dollars)
Sub-Total (Add Air Flotation Cost) $170,880.00
No enclosure included for the above equipment.
III. Contact Stabilization Plant Operating Costs (1972 Dollars)
Annual Power Costs (based on 87 connected HP) $ 11,397.00
It is estimated that 5 hours a day would be required for plant maintenance.
Estimated Chemical Costs (Chlorine) $2.75 per day.
447
-------
ALTERNATE NO. 3
AERATED LAGOON PLANT
I. Design Conditions (Also see Design Conditions for Alternate No. 1)
Raw B.O.D. is 1685 pounds per day less 40% B.O.D. removal in the air
flotation unit, leaving 1011 pounds per day B.O.D. in the primary
effluent. Therefore, a B.O.D. removal of 93% is required through the
remaining portion of the plant to meet the required effluent standards.
II. Anaerobic Lagoon Design
It is proposed to construct two anaerobic basins (lined with existing
clay) and induce a load of 15 pounds per day of B.O.D. per 1000 cubic
feet of tank space. It is assumed that a 50% B.O.D. reduction will oddur
through these lagoons. Install two lagoons each 66 feet square by 15
feet deep.
III. Aerobic Lagoon Design
It is proposed to construct two lagoons lined with existing clay.
Each lagoon shall be 300 feet square by 10 feet deep. Aeration in the
proposed lagoons will be supplied by a subsurface method. The two lagoons
will have a combined detention time of 40 days at design flow.
IV. Lagoon System Capital Costs (1972 Dollars)
Sub-Total (Add Air Flotation Cost) $200,663.50
No enclosure for above equipment.
V. Lagoon System Operating Costs (1972 Dollars)
Annual Power Cost (JBased on an estimated 90 $ 11,700.00
connected HP)
It is estimated that 3 hours a day would be required for plant maintenance.
Estimated Chemical Costs (Chlorine) $2.75 per day.
ALTERNATE NO. 4
AERATION BASIN PLANT
I. Design Conditions (Also see Design Conditions for Alternate No. 1)
Raw B.O.D. is 1685 pounds per day less 40% B.O.D. removal in the flota-
tion unit, leaving 1011 pounds per day B.O.D. in the primary effluent.
Therefore, a B.O.D. removal of 93% is required through the secondary
portion of the plant to meet the required effluent standards.
448
-------
II. Plant Capital Costs (1972 Dollars)
Sub-Total (Add Air Flotation Cost) $122,427.80
No enclosure for above equipment.
III. Plant Annual Operating Costs (1972) Dollars)
Annual Power Costs (Based on Estimated 87 $ 11,397.00
connected HP)
It is estimated that 5 hours a day would be required for plant maintenance.
Estimated Chemical Cost (Chlorine) $2.75 per day.
449
-------
EVALUATION OF VIBRATORY BLANCHER-COOLER FOR
SNAP BEANS AND LIMA BEANS
John L. Bomben*, W. C. Dietrich*, J. S. Hudson*, E. L. Durkee*,
R. Rand**, J. W. Farquhar*** and D. F. Farkas*
INTRODUCTION
The Vibratory Blancher-Cooler resulted from research aimed at reducing
the waste load from blanching and cooling and at improving the design and
heat efficiency of steam blanchers. The pilot plant, whose design and
performance is described here, combined a series of developments of the
work done at the Western Regional Research Center on steam blanching
and cooling vegetables prior to freezing. Each of these developments was
tested to the small pilot plant stage, and the Vibratory Blancher-Cooler
incorporated all of these developments into a. prototype, which was large
enough to demonstrate the suitability of these previously developed concepts
for production use in a vegetable freezing plant.
The use of vibratory conveyors for steam blanching was developed to
reduce the size and improve the heat efficiency of steam blanchers (2, 3).
The spiral or stacked vibratory conveyor allowed for a more compact
design than did the conventional belt conveyor in steam blanchers. In
comparison to water blanchers, steam blanchers are large and have a low
heat efficiency. The lower waste load of steam blanching (4) has not been
sufficient in some cases to justify its use in place of water blanching,
because for the same processing capacity steam blanchers are much larger
and cost more than water blanchers. Until recently, little attention has
been paid to heat efficiency, but the scarcity of energy should make this
an important criterion now.
Using steam blancher condensate as a spray during air cooling was another
development incorporated into the Vibratory Blancher-Cooler. This
technique reduced the waste load of both blanching and cooling (Z, 5). By
using air, instead of flume cooling, the hydraulic waste load of cooling was
reduced enormously, and the organic waste load produced by the leaching of
solids in the flume was eliminated. By using the steam blancher condensate
as a spray during air cooling, the waste load of blanching was elminated,
*USDA, Western Regional Research Laboratory, Berkeley, California.
**California State University, Chico, California (employed by American
Frozen Food Institute for the project).
***American Frozen Food Institute, Washington, DC.
450
-------
and the waste load of both blanching and cooling was reduced to the unevap-
orated and unabsorbed liquid leaving the cooler.
The technique of Individual Quick Blanching (IQB) was included in the Vibra-
tory Blancher-Cooler since it gave an additional means of reducing the size
of the blancher. This heating and holding technique in addition gives a
uniform blanching of vegetables and reduces the waste load of a steam
blancher (1).
Pilot Plant Design
A schematic diagram of the pilot plant is shown in Figure 1, and a photo-
graph of it, as it was installed at Patterson Frozen Foods (Patterson,
California), is shown in Figure 2. The pilot plant occupied a floor area of
1Z ft by 10-1/2 ft, and it had a height of 13 ft-11 in. It was constructed by
the Vibrating Equipment Division of Rexnord Corporation, Louisville,
Kentucky. ->The nominal design capacity was one ton per hour with a one
pound per ft conveyor loading. The pilot plant had a shipping weight of
19000 Ibs. It consisted of the following units: heater, holder, cooler, air
blower with filter, and condensate spray system.
The heater was a vibratory spiral enclosed in a double wall insulated (1 in.
air spacing) housing having one access door (Figures 2 & 3). At the bottom
of the housing there was a 2 in. diameter drain for condensate from the
inside walls. The spiral conveyor, consisting of three 11-15/16 in. -wide
flights, had a 12 in. pitch. Steam was distributed through the central tube,
which had 3/4 in holes 30 in. apart and 3 in. above the spiral conveying
surface. The spiral and housing were made of stainless steel (Type 304,
No. 2B finish on spiral and No. 3 finish on housing). There were two feed
spouts; one was at the beginning of top flight and the other was at the
beginning of the middle flight. Only the top feed spout was used in the
experiments described here. Two 1-1/2 hp electrical motors (220V, 900
rpm), which had eccentric weights on their extended shafts and which were
mounted at 45° to the horizontal, vibrated the spiral at a frequency of 890
cpm. The drive motor mounting was suspended with cables and springs to
isolate the vibration from the structural framework. The amplitude of
vibration could be varied from 0 to 3/8 in by shifting the position of the
eccentric weights on the motors. The direction of the vibration moved the
vegetables down the heater spiral, and the amplitude of vibration determined
their conveying velocity.
The vegetables leaving the heating spiral dropped into the holder (Figures 2
and 3), which was a horizontally vibrating conveyor (5 ft long x 1 ft x 1 ft)
enclosed by double wall insulation (1 in. air space). At the bottom of the
holder near the feed end there was a 1 in. drain for the condensate
produced by the vegetables on the heater spiral. A screen of perforated
stainless steel was installed over the drain to prevent it from being plugged
by vegetables. The holder had a 2° incline toward the feed end to drain
liquid to the condensate pump. The holder was vibrated at 890 cpm by a
1-1/2 HP electrical motor (220 V; 900 rpm) with a doubly extended shaft
on which two series of eccentric weights were mounted. By changing the
number of weights the amplitude of vibration could be varied from 0 to 1/4
451
-------
HEATER & HOLDER
en
ro
FEEDER
RAW
VEGETABLES
COOLER
k BLANCHED
& COOLED
VEGETABLES
Figure 1 - Schematic diagram of Vibratory Blancher-Cooler.
-------
HEATER
HOLDER
COOLER
Figure 2 - Photograph of Vibratory Blancher-Cooler Pilot Plant.
453
-------
T--
Figure 3 - Detailed drawing (elevation) of Vibratory Blancher-Cooler
Pilot Plant.
454
-------
in. In addition, the holder motor was equipped with a tinier that controlled
the period the motor was on and off to get a desired residence time. The
vibration was isolated from the supporting framework by suspending the
holder with cables and springs.
The blanched vegetables leaving the holder were discharged to the cooler,
where the vegetables moved up the vibrating spiral conveyor. The cooler
spiral conveyor (Figures 2 & 3) had the same design as the one in the heater
except that it had 6-1/2 flights. The drive unit for the cooler was of the
same design as that in the heater, but the motors were mounted so that the
vibration in the cooler moved product up the spiral. Air was distributed
over the product in the same way steam was distributed in the heater. The
housing for the cooler was constructed of stainless steel (Type 304, No. 3
finish). It had two access doors and it had a 2" drain leading to an effluent
collection tank.
The blower (Buffalo Forge, No. 40 MW),used for moving air over the pro-
duct in the cooler had a rating of 5300 ft /min of 70° F air at 6 in. of water
pressure. A filter (Continental No. 3P439M Side Access Conopac, 90% effi-
ciency) was installed at the inlet of the blower. A damper on the outlet
of the blower was used to control the amount of air flow.
The condensate spray system consisted of a pump (Waukesha Sanitary
Pump, No. 10), approximately 15 ft of one inch diameter sanitary pipe and
four nozzles (Spraying Systems Co. , Unijet Nozzle No. 2503). Condensate
was sprayed on the product in the cooler at the first, third, fifth and sixth
flights. The condensate from the drain in the holder was filtered through
three layers of cheesecloth to prevent particles from clogging the nozzles.
Temperatures at the inlet and outlet of the heater, holder and cooler were
measured by thermocouples and recorded on a multi-point temperature
recorder. A manometer measured the pressure at the blower, and the
amount of air flow was determined from the blower performance curve
supplied by the manufacturer. A rotameter (Fischer-Porter, Model No.
10A1152, 100% of scale = 199 scfm air), which had been calibrated by
condensing and weighing the steam flowing through it, was used to measure
the flow of steam to the heater. Condensate was removed from the steam
supply by passing it through a purifier (V. D. Anderson Co. , Model LC-
150). A pressure regulator (Spence Engineering Co. , 3/4 in Type ED)
maintained a constant pressure at the throttling valve installed in the steam
line after the rotometer. The relative humidity of the air at the inlet of the
filter was measured with an electronic humidity indicator (Humi-Chek,
Beckman Instruments). A hot wire anemometer (Alnor, Thermo-anemo-
meter) was used to measure the air velocity over the product in the cooler.
Experimental Runs
Raw vegetables for runs were collected into bins (4 ft x 4 ft x 4 ft) from
the production line at the stage where they were ready for blanching. The
snap beans (Galagreen variety) used in the pilot plant were collected at the
point in the production line where they were washed, sorted and cut to 3/4
in lengths. Lima beans (Kingston, S4 and Bridgeton varieties) were
collected after they had been washed and flotation graded in a 13-14% brine
455
-------
solution in the production line. Because the final rod-reel washer and
the water blanchers were so closely connected in the production line, it was
not possible to collect the lima beans for the pilot plant runs as thoroughly
cleaned and washed as they were for the production line blanchers. In
some cases it was necessary to collect beans as they left the brine sorter
without any subsequent washing; other times it was possible to collect the
lima beans after only a brief washing following the brine separator.
The testing of the pilot plant was done with three kinds of experimental runs:
preliminary, batch and continuous. Preliminary runs were used to establish
the feed rate, blanching time and cooling time and to observe the operating
characteristics of the pilot plant for each vegetable. Each batch run con-
sisted of blanching and cooling approximatey 2, 000 Ibs of raw vegetables and
determining the yield of blanched-cooled vegetables, the solids lost from the
raw vegetables and the waste load. In the continuous runs the pilot plant was
operated for a longer time than the batch runs; the same quantities as in the
batch runs were measured, but in addition samples were taken for micro-
biological analysis and sensory evaluation. Since the pilot plant -was not in-
stalled soon enough, there was no time for continuous runs with snap beans.
Residence times in the heater, holder and cooler and the flow rates of steam
and air were varied in the preliminary runs to determine the best conditions
for processing the vegetables. Product velocities on the conveyors were
measured by bundling approximately ZOO g of vegetables in cheesecloth and
timing their passage on the conveyors. With the steam to the heater off,
preliminary runs were also used to observe the product flow on all the
vibratory conveyors. With the steam to the heater on, the vegetables
leaving the cooler were tested for peroxidase (6), and their temperature
was measured. If the peroxidase test was negative and the product tem-
perature was between 80 to 90°F, no further adjustments were made, and
the series of batch and continuous runs were begun, using the conditions
established in the preliminary runs.
For the batch runs one bin of raw vegetables was weighed and dumped into
the feeder-elevator. After the temperature in the heater reached 212° F,
the vegetables were started through the pilot plant. Samples (150 g) of raw
vegetables and of vegetables leaving the cooler were taken every fifteen
minutes. In the snap bean runs a 1. 5 kg sample for sensory testing and
chemical analysis was taken from the pilot plant and production line since
there were no continuous runs. The vegetables leaving the cooler were
analyzed for peroxidase several times during the run. About every 15
minutes a 500 g sample of product leaving the cooler was collected into a
beaker, and the average temperature was measured with a dial thermo-
meter. All the vegetables leaving the cooler were collected into a bin and
weighed. All the cooler effluent was collected from the start of the run to
the time when the vegetables were no longer being discharged from the
holder. The effluent was weighed, and three 500 g samples were taken in
polyethylene bottles.
For the continuous runs with lima beans essentially the same procedure as
used in the batch runs was repeated, except that two to four bins (4000-8000
Ibs) of raw vegetables were fed to the pilot plant giving runs from 2 to 5. 5
456
-------
hours. Samples (150 g) of raw vegetables and those coming from the cooler,
as well as the temperature of those coming from the cooler, were taken
every- hour. Every hour, samples (50-100 g) for Total Aerobic Count and
samples (1.5 kg) for sensory evaluation and chemical analyses were taken
of the vegetables leaving the cooler. For comparison, 1. 5 kg samples of
blanched and cooled vegetables were also taken every hour from the produc-
tion line. All vegetable samples were immediately frozen. The 1. 5 kg
samples were later made into a single composite. The blanched and cooled
vegetables were collected into bins and weighed. The effluent was collected
and weighed about every hour, and from this quantity a 9 kg composite
sample was taken with the remainder being discarded. From the composite,
three 500 g samples were taken in polyethylene bottles.
A series of analyses were done on the samples of the raw vegetables,
blanched and cooled vegetables, and effluent. One sample bottle of effluent,
taken as described above for each run, was refrigerated at 34°F and a
Suspended Solids (SS) (7) analysis was done within 24 hr; analyses of Total
Solids (TS) and Total Organic Carbon (TOG) (8) were done on this refri-
gerated sample within a week of the time of the run. Another sample bottle
of effluent was frozen, and after about one month it was given to a water
analysis laboratory (R. W. Hawksley Co. , Richmond, California) for
analysis of Biological Oxygen Demand (BOD). A spare sample bottle of
effluent was kept frozen for checking any results that were questionable.
All vegetable samples were taken in polyethylene bags. The 150 g samples
of raw and blanched-cooled vegetables were kept frozen and they were
analyzed for Total Solids (9) within two months of the run. The 50-100 g
samples taken for Total Aerobic Count were kept at -20°F (for less than
one month) until they were analyzed (10). The composites of 1. 5 kg samples
were also stored at -20°F; part (50-100 g) of the composite was used for
analysis of peroxidase, ascorbic acid and chlorophyll conversion (6) while
the remainder was used for sensory evaluation by the Duo-Trio test (11)
within 3 months of the run.
The operating conditions of lima bean blanchers and coolers in the produc-
tion line were also measured. The blanchers in the production line were
screw conveyor water blanchers. Cooling was done by either a combination
of air and flume or only flume. The feed rate of vegetables and effluent
discharge rate were measured by collecting and weighing all vegetables
going into the blancher or all the effluent leaving the blancher or cooler for
short periods of time (0. 25 or 0. 5 min). Samples of raw vegetables (150g),
blanched-cooled vegetables (150 g), and blancher and cooler effluents (500g)
were taken for each measurement. The effluent samples were analyzed for
TS, TOC and BOD, and the lima bean samples were analyzed for TS.
Operating Conditions
The operating conditions used for the nine batch and five contiguous runs
made with the pilot plant and the data measured on the lima bean production
line are shown in Table 1.
The residence time for lima beans in the heater and holder were slightly
longer than for snap beans (Table 1) at the same vibration amplitude;
457
-------
TABLE 1. OPERATING CONDITIONS
Run Number
& Vegetable
Variety
Run
Time
(min)
Feed
Rate
(Ib/hr)
Heating
Time
(min)
Holding
Time
(min)
Holder
Temp
(°F)
Steam
Flow
(Ib/hr)
Vegetable
Temp
(-F)
PILOT PLANT
Snap Beans
SB-1 Galagreen 45 2220
SB-2 Galagreen 50 2160
SB-3 Galagreen 49 2110
Lima Beans
LB-1 Kingston 45 2750
LB-2 Bridgeton 35 1230
LB-3 S4 78 1200
LB-4 Kingston 83 2590
LB-5 S4 50 2700
LB-6 Bridgeton 57 2500
LB-7 Kingston 126 1940
LB-8 Kingston 149 1850
LB-9 Bridgeton 272 1850
LB-10 Bridgeton 331 1550
LB-11 Kingston 274 1640
Lima Beans
LBP-1 Kingston 9280
LBP-2 Kingston 6580
LBP-3 Kingston 12400
1.0
1.0
1.0
(a)
(a)
(a)
1.0
1.0
1.0
fb)
(b)
(b)
PILOT PLANT
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1
1
1
1
1
1
1
1
1
1
1
(a)
(a)
(a)
(a)
(a)
(a)
(a)
(a)
(a)
(a)
(a)
1. 3
3.5
1. 3
1. 3
1.3
1.3
1. 3
2. 5
2. 5
2. 5
2. 5
(b)
(c)
(b)
(b)
(b)
(b)
(b)
(d)
(d)
(d)
(d)
PRODUCTION LINE
4.5
4.5
4.5
(h)
(h)
(h)
164+20
185^5
186+10
186+3
209
209
1 96+_l 0
195+10
199+6
200_f5
195+5
(i)
285
285
285
197
197
225
285
328
242
242
242
235
240
(j)
86
90
89
76
75+2
73+1
80_+1
77+1
84+2
83+_3
79+4
83+3
(1)
109+1 \
103+3(m)
(a) Vibration amplitude = 3/16 in.
(b) Vibration amplitude = 1/8 in; on continuously.
(c) Vibration amplitude = 1/8 in; on 5 sec., off 10 sec.
(d) Vibration amplitude = 1/8 in; on 10 sec. , off 10 sec.
(e) Cooler residence time = 2. 9 min. ; vibration amplitude = 1/4"
(f) Cooler residence time = 4. 8 min. ; vibration amplitude = 1/4"
(g) Cooler residence time = variable, vibration amplitude = 1/4" for LB 9-11.
(h) Blancher temp. = 210°F.
(i) Steam flow was erratic since this was first run done with the pilot plant.
(j) There was no measuring device for steam flow on the production line blanchers.
(k) Cooling time: 0. 7 min in air with water sprays and 0. 2 min in water flume.
(1) Cooling time: 0. 3 min in water flume.
for SB 1-3
for LB 1-8.
(m) Cooling time: same as (k)
458
-------
otherwise both vegetables gave reproducible residence times, and they
flowed uniformly down the conveyor spiral. The amplitude of vibration in
the holder was kept at 1/8 in. , and the residence time was controlled by
the timer. In the cooler the conveying velocity for lima beans was much
slower than for snap beans, and the flow was nonuniform at times.
The steam flow was adjusted to maintain a temperature of 190°F in the
holder. The heater was always kept at, 212° F.
Condensate from the heater was pumped immediately to the nozzles in the
cooler. Its total residence time from discharge from the holder to the
nozzles was calculated to be less than 1. 5 min. Although no difficulty was
experienced in spraying the snap bean condensate, the lima bean condensate
sometimes plugged up the nozzles. The unevaporated and unabsorbed
condensate in the cooler drained down the outer edge of the conveyor
surface to the effluent collection tank. :
The air flow was highly turbulent in the cooler and only an average velocity
could be estimated. An air flow of approximately 3500 cfm was used for
most of the runs, and this gave an average velocity of 300 ft/min over the
vegetables in the cooler. The wet bulb temperature of the air entering the
filter was 55-60°F for most of the runs.
Waste Load and Product Quality Results
Table 2 shows the product yield, solids loss and the liquid waste load
obtained for all the experimental runs. The yield is the weight of blanched
and cooled vegetables when compared to the weight of raw vegetables: %
yield = 100 x (weight of blanched cooled vegetables)/(weight of raw vege-
tables). The solids loss measures the solids which are leached from the
vegetables in blanching and cooling when compared to the amount of solids
originally present in the raw vegetables: % solids loss = 100 x (hydraulic
load) x (TS in effluent)/(TS in raw vegetable). The other quantities in Table
2 (hydraulic load, BOD, TOC, SS) are calculated from the feed rate,
effluent rate and analyses of the effluent, and they are reported on the basis
of 1, 000 Ib of raw vegetables.
Table 3 gives the results of sensory comparisons (Duo-Trio) of product
from the pilot plant with that of the production line. The results are
reported as the number of judgements (% correct) correctly identifying the
sample that was identical to the control. The control for a comparison
could be either the pilot plant or production line sample. Judges were also
asked to state which sample they preferred. Probabilities are from the
Binomial Probability Table.
Table 4 summarizes the results of the analyses for total solids, peroxidase,
ascorbic acid and chlorophyll conversion of pilot plant and production line
samples. The results are averages of 2 or 3 determinations done on 50 to
100 g of vegetables. The data on ascorbic acid are reported without
accounting for differences in moisture content between the vegetables of the
production line and those of the pilot plant.
459
-------
'TABLE 2. PRODUCT YIELD, SOLIDS LOSS AND LIQUID WASTE LOAD
Product Hydraulic
Run Yield Solids Loss Load BOD TOC SS
Number (%) (%) (lb/1000 Ib) (lb/1000 Ib) (lb/1000 Ib) (lb/1000 Ib)
PILOT PLANT
SB-1
SB-Z
SB-3
LB-1
LB-2
LB-3
LB-4
LB-5
LB-6
LB-7
LB-8
LB-9
LB-10
LB-11
86.5
98. 5
97. 2
89. 2
83.2
90.9
94. 1
92. 1
93.2
92.9
92.1
92.0
0. 709
0. 707
0. 704
0.494
0.0516
0. 154
0.202
0. 372
0.148
0. 794
1.22
51.4
32. 1
29.3
47. 1
41.7
39. 0
28.3
4.
7.
63
89
11.3
19.6
8. 19
40. 1
60. 2
0.802
0. 520
0.439
0. 885
0. 651
0. 772
0.832
0. 0982
0. 236
0. 794
1. 87
0.617
0.313
0. 289
0.674
0.575
0. 536
0. 514
0.0637
0. 159
0. 200
0.387
0.150
0.992
1.51
PRODUCTION LINE
0.0369
0.0440
0.503
0.216
0.423
0.403
0.0411
0. 137
0. 167
0.347
0. 122
1.80
LBP-1 (a) Jj
T.BP-2
_I_J J-J .L ™ L*
LBP-3
j_jj_> i « ^/
1.
1.
1.
0.
69(B)
515 (C)
61 (B)
20 (C)
62 (B)
480 (C)
548
536
460
780
373
313
(B)
(C)
(B)
(C)
(B)
(C)
3.
0.
3.
1.
3.
1.
18
9i:
77
03
17
00
(B)
L (C)
(B)
(C)
(B)
(C)
1.
0.
1.
0.
1.
0.
78(B)
379 (C)
69 (B)
409 (B)
75 (B)
438 (C)
0.
0.
0.
0.
0.
0.
849 (B)
150 (B)
846 (B)
150 (C)
683 (B)
121 (C)
Hydraulic load, BOD, TOC, and SS are reported per 1000 Ib of raw vegetable.
(B) refers to blanching.
(C) refers to cooling.
(a) Yield could not be accurately measured in the production line.
460
-------
TABLE 3. SENSORY EVALUATION OF PRODUCT BY DUO-TRIO TEST
Total % Preference
Run Number of for Pilot Plant
Number Judgments % Correct Sample
SB-1
SB-Z
SB -3
LB-7
LB-8
LB-9
LB-10
JLB-11
36
36
34
41
40
41
38
40
61
61
53
66 *
65*
80**
63
63
42
44
68
54
67*
63
32*
55
NS = Not Significant.
*Probability 0. 05
**Probability 0.01
461
-------
IABLE 4. ANALYSIS OF VEGETABLES
-PS.
01
Run
No.
SB-1
SB-2
SB-3
LB-1
LB-2
LB-3
LB-4
LB-5
LB-6
LB-7
LB-8
LB-9
LB-10
LB-11
LBP-1
LBP-2
LPB-3
Raw TS
(%)
10.76
10.75
10.75
36.12
--
__
41.29
40.65
38.46
40.86
40.72
41.60
35.92
36.09
36.08
35.35
35.37
Product TS
Pilot Prod.
Plant Line
(%) (%)
11.20
10.75
10.40
37.66
--
41.37
39.93
41.71
39.32
39.74
40.13
40.71
35.62
36.07
10.35
--
-_
_ _
-_
__
__
34.60
__
36.49
36.01
38.03
34.80
34.82
35.81
34.15
35.32
Residual Peroxidase Ascorbic Acid
Yield
(%)
86.5
98.5
97.2
89.2
83.2
90.9
94.1
92.1
_.
--
93.2
92.9
92.1
92.0
_ _
--
~ ~
Solids
Loss
(%)
..
0.709
0.707
0.704
__
__
0.494
0.052
0. 154
0. 202
0.372
0. 148
0.794
1.22
2.21
2.81
2.10
Pilot
Plant
(%)
2.1
1.1
1.5
0.0
-_
0.0
0.0
0.0
0.0
0.0
Prod.
Line
(%)
1.6
0.5
0.5
0.0
-_
0.0
0.0
0.0
0.0
0.0
Pilot
Plant
mg/lOOg
14.1
15.1
14.1
18.7
--
26.7
27.2
20.6
19.1
24.6
Prod.
Line
mg/lOOg
13.9
13.4
13.4
16.7
-.
23.8
24.0
19.7
20.3
25.2
Chlorophyll
Pilot Plant
(%)
16.8
13.8
13.8
13.0
-_
11.9
17.2
15.6
19.1
18.9
Conversion
Prod. Line
(%)
16.8
14.1
14.1
14.5
-_
13.9
10.8
13.0
11.3
8..0
Ascorbic acid analyses were not corrected for differences in the amount of water in the vegetables.
-------
Figure 4 gives the results of the Total Aerobic Count on the continuous runs
of lima beans. Since there were only batch runs for snap beans, no micro-
biological analysis was done for that vegetable.
Discussion of Pilot Plant Design and Operation
The pilot plant makes a large reduction in hydraulic waste load and organic
waste load when compared to the conventional processing simulated in
previous work or when compared to the conventional processing of lima
beans as measured on the production line in this work. The effluent
leaving the cooler is a low volume highly concentrated waste stream:
16000 ppm BOD for snap beans and 16000-29000 ppm BOD for lima beans.
A waste stream of this concentration would be much less expensive to treat
than the high volume, low concentration waste stream typical of the present
blanching and cooling systems. The product yield and waste load obtained
in the pilot plant (Table 2) are similar to those of the previous small scale
experiments for snap beans and lima beans (5). Table 5 compares the
results of the previous'work with the range of results obtained in this work.
The floor space required by the pilot plant was low when compared to a con-
ventional steam blancher of the same capacity. The conveying length of the
heater spiral was 25 ft. Since vegetables could pile up in the holder, its
length was only 5 ft. and it provided residence times equal to (snap beans)
or twice (lima beans) the times in the heater. If a single layer belt loading
of 1 Ib/ft is assumed, a belt steam blancher of conventional design would
have to be 50 ft. long and 1 ft. wide for snap beans and 75 ft. long and 1 ft.
wide for lima beans to provide the same conveying capacity and blanching
times as the pilot plant.-, Since the floor area taken by the heater and holder
was approximately 20 ft (5 ft. x 4 ft. ), the entire blancher of the pilot plant
would occupy 1/2 to 1/3 the space of the 50-75 ft area taken up by the equi-
valent conveyor belt of a conventional steam blancher, excluding the housing
and supporting framework.
The heat efficiency of the pilot plant blancher was very high when compared
to existing steam blanchers. A nylon chute was used for feeding the raw
vegetables to the first flight of the heater. When the chute was filled with
vegetables, it formed a seal, and when it was empty it collapsed allowing
very little steam leakage. The discharge end of the heater led directly into
the holder, whose discharge was sealed by a canvas chute. This elimination
of steam leaks and the insulation of the exposed surfaces combined to give
the blancher an efficiency as high as 85-90%, which was calculated by com-
paring the steam measured by the rotameter to that calculated to be neces-
sary to give the vegetables an average temperature of 190°F (5). Steam
blanchers with water-sealed ends report efficiencies of 25% (12), and
blanchers without seals would be even less.
The Total Aerobic Count of the samples taken from the pilot plant was very
low when compared to the acceptable limit of less than 500, 000 colonies per
gram for frozen vegetables given by Sharf (10) or to the count of product
taken from the production line (1000-2000 colonies/g). There was no
increase in microbial count with time of pilot plant operation for a period
of over five hours (Figure 4), demonstrating that the cooler did not have
463
-------
120
o>
to
110
100
Z 90
2 80
o
V. 70
Z 60
D
O 50
u
uj 40
I—
^ 30
Q_
20
10
140
O LB 7
A LB 8
O LB 9
• LB 10
A LB 11
I
1
01234567
RUN TIME (MRS)
Figure 4 - Total Aerobic Plant Counts on Continous Runs with Lima Beans
-------
TABLE 5. COMPARISON OF PILOT PLANT WITH PREVIOUS WORK AND PRODUCTION LINE
Yield Solids Loss Effluent TOG
(%) (%) (lb/1000 Ib) (lb/1000 Ib)
Snap Beans
Vibratory Blancher-Cooler
Pilot Plant 86.5-97.2
Previous Work: Steam
Blanch & Air Cool
w/Condens. (5)
Previous Work: Steam
Blanch & Water Cool (5)
92.0
94.3
0.71
1.1
3.9
29-51 0.29-0.62
50
5080
0.37
1.5
Lima Beans
Vibratory Blancher-Cooler
Pilot Plant 89.2-94.1
Prevous Work: Steam
Blanch & Air Cool
w/Condens. (5)
Previous Work: Steam
Blanch & Water Cool (5)
Production Line: Water
Blanch & Air-Water Cool
93.9
100.6
0.052-1.2 46-60 0.06-1.5
0.38
1.5
40
5070
0.60
2.8
2.1-2.8 690-2200 2.1-2.2
Previous work is described in reference (5).
465
-------
areas promoting microbial growth. The samples of product taken from the
production line were taken just prior to freezing, and so their higher micro-
bial counts were probably due to the fact that the vegetables had been exposed
to the plant environment on the inspection belts; whereas, the product from
the pilot plant had no such exposure.
The sensory evaluation of the blanched and cooled product showed that there
was essentially no difference in the appearance, taste or texture of the
product due to processing in the pilot plant as compared to processing in the
production line. Some of the small differences (Table 3), found in the lima
beans were probably due to differences in saltiness because the raw
vegetables for the pilot plant could not be collected after the final wash
in the production line. Preferences of the panel were not consistent and
alternated between lima beans produced in the pilot plant and those taken
from the production line (Table 3).
The solids loss (Tables 2, 4 and 5) is a measure of the amount of nutrient
lost by leaching from the vegetables. Although the vegetables processed in
the pilot plant had lower solids losses than those taken from the production
line (Tables 2 and 4) or those processed by a simulated conventional process
(Table 5), the differences in ascorbic acid did not necessarily correlate with
those in solids loss. Generally, the ascorbic acid content was not much
different when samples taken from the production line were compared to
those of the pilot plant. The slightly higher ascorbic acid of the snap beans
processed in the pilot plant could be accounted for entirely by the water
evaporated in air cooling; similarly, the differences in ascorbic acid for the
lima beans of runs LB 5, 7 and 8 could be accounted for by differences
in the amount of evaporation.
Ascorbic acid is lost both by leaching and thermal degradation, but chloro-
phyll conversion is solely a measure of thermal degradation. Since the
cooler conveying velocity was much slower in runs LB 10 and 11 (Table 1),
owing to conditions described later, the heavily loaded conveyor gave an
excessively slow cooling rate. The slower cooling rate was responsible
for the higher chlorophyll conversions, as compared to those of the
production line, found in lima beans processed in these pilot plant runs
In those runs (LB 5 and 7) where the cooler conveying velocity was faster,
there was essentially no difference in chlorophyll conversion between pilot
plant and production line. In contrast, the snap beans showed no tendency
to have a higher chlorophyll conversion in the pilot plant runs. The faster
conveying velocity of the snap beans gave less of a conveyor loading and thus
rapid cooling.
The large variation in solids loss, hydraulic waste load, TOG, BOD and
SS shown in Table 2 for the pilot plant runs with lima beans appeared to
result from variations in the maturity of the raw beans used. Comparison
of Tables 2 & 4 shows that the lower waste loads (LB 5, 6, 7, 8 and 9) were
measured when the raw beans had a high TS (dry, mature beans) and the
higher waste loads (LB 10 and 11) were measured when the TS of the raw
lima beans was relatively low. The drier beans probably absorbed more of
the condensate in the heater and cooler (13).
466
-------
The pilot plant was designed to blanch and cool 2000 Ib/hr. Because itwas
not installed soon enough in the season, it was not possible to test the
maximum capacity for snap beans. The conveyors did not appear to be
fully loaded at the feed rate used (2200 Ib/hr) with snap beans, and a
higher feed rate appeared to be possible. Since lima beans are generally
blanched for longer times than snap beans, and since, at the same amplitude
of vibration, the lima beans conveyed more slowly than did the snap beans,
the pilot plant should have a lower capacity for lima beans than snap beans.
The 2500-2750 Ib/hr feed rates in runs LB 1, 4, 5 and 6 were easily con-
veyed by the heater and holder, but they exceeded the capacity of the cooler.
Thus, the continuous runs had to be made with lower feed rates (Table 1).
Lima beans were not conveyed up the cooling spiral as well as the snap
beans. Under some conditions the conveying velocity steadily decreased
until after about one hour of operation the lima beans stopped moving
altogether. Increasing the stroke of the conveyor to the maximum 3/8
in. stroke did not convey them faster, but it did make their flow erratic,
giving a circulation pattern--lima beans on the outer edge were conveyed
very rapidly and those in the center slowly or not at all, and as the average
conveying velocity decreased, the lima beans in the center moved downward
in the opposite direction to those on the outer edge. If the conveyor was
washed clean of fragments, leaves and stems, which had stuck to the
conveyor surface, the lima beans moved again with their initial conveying
velocity. This conveying problem in the cooler occurred sooner in those
runs -where there were more leaves, stems and bean fragments in the raw
lima beans, a condition aggravated by the inability to get completely cleaned
and washed beans from the production line. In those cases where the raw
lima beans were clean, even after only a preliminary wash, they were
continuously conveyed by the cooler, processing over 8, 500 Ibs in Run
LB 10. Since the lima beans showed uniform, steady flow in the heating
spiral, arranging the equipment so that lima beans flowed down the cooling
spiral would give better control of the cooling time.
Preliminary runs were tried •with Brussels sprouts, broccoli and cauli-
flower. Although it was not possible to convey these blanched vegetables
up the cooler, they were conveyed down the heating spiral evenly and with
reliable residence times. Thus, it appears that the cooler could operate
satisfactorily if it were arranged to convey downward. This arrangement
would also give lima beans a faster conveying velocity in the cooler thereby
giving better cooling conditions. In the pilot plant this would require raising
the discharge of the holder by 5 ft. and thus making the top of the heater
assembly a total of 19 ft. from the floor.
The advantages of reduction in waste load and solids loss from the
vegetables cannot be achieved economically by using air cooling because
frozen vegetables are marketed by total weight. Under these marketing
conditions the evaporation in the cooler translates into a loss of product,
which cannot be balanced by the savings in steam and effluent disposal costs
(2). For lima beans, where brine sorting follows cooling, the loss of
weight is probably regained by soaking in the brine; however, this .regaining
of weight is at the expense of the solids leached from the vegetables in the
brine solution in the same way as flume cooling. The full advantages of air
467
-------
cooling for liquid waste reduction can be achieved only if frozen vegetables
are marketed by some standard other than total weight.
No estimate of the capital cost of a full size blancher-cooler has been made
based on the pilot plant. The pilot plant cost $61, 500. Since the cost
required for installation of the plant would vary considerably, depending on
location, it is difficult to establish that amount, but as a guide the installa-
tion cost of the pilot plant was about 10% of the equipment cost. Since most
conventional steam blanchers are custom-built there is no established cost
with which to compare.
ACKNOWLEDGMENT
This project was carried out under an EPA Demonstration Grant, which
was administered and additionally supported by the American Frozen Food
Institute. The authors wish to express their appreciation to Harold
Thompson, EPA Project Officer, for his guidance on this project. In
addition, we would like to thank personnel of the American Frozen Food
Institute--Joanne Cox, Jean Bohannon, Elaine Carter, and Ray McHenry,
who provided administrative help and John Swartz (University of California,
Berkeley, undergraduate hired for the summer), who served as a technician
in setting up and testing the pilot plant. We also wish to thank those
personnel of Patterson Frozen Foods and Tom Rumsey of the Western
Regional Research Center who provided assistance for this project. The
vibratory conveyors were designed by John Reinders, Project Engineer,
Rexnord Corporation; the photograph was taken by George George,
Corporate Photographer, Rexnord, Corporation.
Reference to a company or product names does not imply approval or
recommendation of this product by the U.S. Department of Agriculture to
the exclusion of others which may also be suitable.
468
-------
1. LAZAR, M. E. , LUND, D.B., and DIETRICH, W. C. IQB: A new
concept in blanching. Food Tech., 25: 684 (1971).
2. BOMBEN, J. L. , BROWN, G. E. , DIETRICH, W. C. , HUDSON,
J. S. , and FARKAS, D. F. Integrated blanching and cooling to" reduce
plant effluent. Proceedings of the 5th National Symposium on Food
Processing Wastes, Monterey, CA. Environmental Protection
Technology Series, EPA-660/2-74-058: 120 (1974).
3. BROWN, G. E. , BOMBEN, J. L. , DIETRICH, W. C. , HUDSON, J.S. ,
and FARKAS, D. F. A reduced effluent blanching-cooling method using
a vibratory conveyor. J. Food Sci. 38:89(1974).
4. RALLS, J. W. and MERCER, W. A. Low water volume enzyme deacti-
vation of vegetables before preservation. Environmental Protection
Technology Series, EPA-R2-73-198 (1973). Oi'
5. BOMBEN, J. L. , DIETRICH, W. C. , HUDSON, J. S. , HAMILTON,
H. K. , and FARKAS, D. F. Yields and solids loss in steam blanching
cooling and freezing vegetables. J. Food Sci. 40: 660 (1975).
6. DIETRICH, W. C. and NEUMANN, H. J. Blanching Brussel sprouts.
Food Tech. 19(5): 150 (1968). ' "
7. EPA. Methods for Chemical Analysis of Water and Wastes.
Environmental Protection Agency 16020--07/71 (1971).
8. APHA. Standard Methods for the Examination of Water and Waste
Water, 13th ed. , p. 257. American Public Health Association, New
York (1965).
9. AOAC. Official Methods of Analysis, 10th ed. , Associaiton of
Official Agricultural Chemists, p. 308, Washington, DC (1965).
10. SHARF, J. M. Frozen fruits, vegetables, and precooked frozen foods.
In "Recommended Methods for the Microbiological Examination of
Foods, " 2nd Ed. , p. 97. American Public Health Association, New
York (1966).
11. ASTM. Manual on sensory testing methods. American Society for
Testing Methods, STP No. 434(1968).
12. LAYHEE, P. Engineered FF line yields 5 big production benefits.
Food Eng. 47(2): 61 (1975).
13. BOMBEN, J. L., DIETRICH, W. C. , FARKAS, D. F. , HUDSON,
J. S., DE MARCHENA, E. S. , and SANSHUCK, D. W. Pilot plant
evaluation of Individual Quick Blanching (IQB) for vegetables. J.
Food Sci. 38: 590 (1973).
469
-------
TREATMENT OF MEATPACKING PLANT WASTEWATER BY
LAND APPLICATION
By
Anthony J. Tarquin
INTRODUCTION
The application of municipal and industrial wastewaters to land
for treatment and or disposal is not new. A textbook on wastewater treat-
ment which was published in 1903 contained complete chapters on sewage
irrigation and sewage farming (1). The land application method of waste-
water treatment has been demonstrated not only in the United States, but
in many foreign countries as well. Liquid wastes with a wide spectrum of
physical and chemical characteristics have been treated successfully by
soil application, including municipal sewage, cannery wastes, pulp and
paper mill wastes, dairy wastes, vegetable and food processing wastes,
wood distillation wastes, poultry wastes, and many others (2, 3).
With the constant annual increase in per capita consumption of beef
and meat products, the need for technological advances in the treatment
of meatpacking wastes has never been greater.
In 1967, the potential daily BOD from the slaughterhouse and meat-
packing industry was estimated at 2.17 million pounds or a population
equivalent of 13 million people (4). The U.S. Department of Agriculture
places the meatpacking industry second to only the Pulp and Paper Industry
in terms of potential five day Biochemical Oxygen Demand pollution. In
the food and kindred products industry, meatpacking ranks first in daily
pollutional discharge.
Interest in the land application method of wastewater treatment for
meatpacking plant wastes has been stimulated by recent federal regulations
requiring a greater degree of wastewater treatment. Of particular impor-
tance in this regard is the increased emphasis that is being placed on
residual carbon, nitrogen and phosphorous removal. Some land application
systems have effectively removed almost all of the carbon and phosphorous
and, under controlled operating conditions, a large percentage of the
nitrogen. In addition to the high degree of waste treatment which is
achieved, of course, some of the valuable plant nutrients are recovered
in the process. The purpose of this report is to present the experiences
gained during pilot scale treatment of meatpacking plant wastewater by
land application. Most of the results reported herein were obtained from
pilot studies at the Peyton Meatpacking plant located in El Paso, Texas.
TYPES OF LAND APPLICATION SYSTEMS
There are three general types of land application systems as follows:
irrigation, overland flow, and infiltration-percolation. All three types
470
-------
can be used for treating meatpacking plant wastewater, depending on the
topography, wastewater strength, soil type, etc.
Irrigation refers to the use of the wastewater for the primary purpose
of growing crops. Accordingly, the hydraulic application rates are rather
low (i.e., less than 4 inches per week) and the wastewater quality is
usually equivalent to secondary effluent. Since meatpacking plant wastes
frequently have very high nitrogen concentrations, the nitrogen loading
rate must be controlled if nitrogen build-up and/or ground water contamina-
tion are to be avoided. For this reason, crops having high nitrogen uptakes
are usually preferred.
Overland flow, as the name implies, involves application of the waste-
water to a sloping area (2-6% slope) with subsequent collection and final
disposal of the effluent. This method is used on clay-containing imper-
vious soils or where topography precludes the use of one of the other
methods.
Infiltration involves percolation of the wastewater into the groundwater.
Hydraulic loading rates for these systems range from a few inches per
week to several feet per week, depending on the quality of wastewater applied.
Loamy or sandy soils are preferred for infiltration-type land application
systems.
WASTEWATER CHARACTERISTICS
Most meatpacking plant wastes are quite strong when compared to domestic
wastewater strength. Table 1 shows the effluent concentrations from the
catch basin of the Peyton Packing Plant.
Table 1. CATCH BASIN EFFLUENT CHARACTERISTICS
Parameter Effluent Cone, mg/1 ab Range, mg/1
PH
BOD
COD
Grease
Phosphorus
Kjeldahl - N
Total solids
Suspended solids
a. Avg. obtained
b. Except pH
7.3
2600
5300
1280
17
110
3,450
1,600
from flow proportional
6.6
200
350
110
3 -
19 -
540
15 -
composite
- 8.1
- 15,150
- 38,890
- 8,165
35
680
- 13,975
10,350
471
-------
The high concentrations for the values shown in the Table are typical
of a packing plant effluent from the catch basin. When other types of
treatment systems are employed, such as oxidation ponds, dissolved air
flotation, or biological treatment systems, the strength of the waste is
lowered considerably. However, when only a catch basin is used for waste-
water treatment, the high nitrogen concentration is particularly troublesome
for land application systems because of the close control required in order
to achieve a high degree of nitrogen removal.
SITE CHARACTERISTICS
The characteristics of the treatment site determine, to a large extent,
the type of land application system which should be designed. Some of the
characteristics which must be taken into consideration in the design of
a land application system follow.
Climate
The climate is one of the most important factors to consider when
designing a land application system. Perhaps the most favorable climate
from the standpoint of system operation is that of the Southwestern
United States. Mild winters and arid conditions are very desirable charac-
teristics because of the minimum holding facilities required. In the northern
areas of the U.S., ponds capable of holding up to six months' flow must
be included in the design of the land application system. Similarly,
where large amounts of rainfall are common, some type of ponding area
would be required for storage of the wastewater during rainy periods.
These problems are not insurmountable but do add to the land require-
ments of the treatment system.
Soil
The type of soil at the treatment site determines, to a large extent,
the kind of land application system which must be used. Soils containing
even small amounts of clay are not well suited for infiltration systems
because of the unfavorable sodium adsorption ratio of most meatpacking
plant wastewaters. The sodium has a sealing effect on the soil which can
reduce the infiltration rate considerably. Overland flow systems are
preferred in clayey soils, therefore. On the other hand, the use of in-
filtration systems is favored on sandy soils. This is because the high
strength of most meatpacking plant wastes requires low hydraulic loading?
with uniform distribution. This is best achieved through high or low
pressure sprinkler application on relatively flat land. The unfavorable
sodium adsorption ratio of most packing plant wastewaters also becomes
less significant when sandy soils are available.
Topography
The topography influences the type of system selected primarily because
of the initial investment cost. Unless the soil contains clay, areas which
are mainly flat favor infiltration systems while areas which have rolling
hills favor overland flow. On the other hand, it is possible to create
472
-------
an overland flow area from land which is flat when soil conditions dictate
such a system. A slope as shallow as 1-1/2% has been used in order to
minimize earthmoving costs. Normal overland flow slopes range from
2-6%.
Grouadwater
The quality of the groundwater in the vicinity of the treatment area "
will have a major influence on the type of system constructed. Where
high quality groundwaters are prevelant, infiltration-type systems for
meatpacking plant wastewaters are not highly recommended because of the
likelihood of groundwater deterioration. Even with close system management,
a considerable amount of nitrogen could be expected to reach the ground-
water in infiltration systems. Additionally, most meatpacking plant waste-
waters are extremely high in total dissolved solids, creating further
potential problems where potable groundwaters are nearby. Thus, when
the groundwater quality is high, infiltration systems for most meatpacking
plant wastewaters are not too feasible.
Vegetation
In all types of land application systems, some type of soil cover is
preferable. For infiltration systems, the vegetation helps maintain the
infiltration rate and prevents compaction of the soil during sprinkler appli-
cation. For overland flow systems, vegetation prevents channeling and
soil erosion and enhances the establishment of the mulchy layer at the
air-soil interface.
The kind of vegetation which is selected should ideally possess the
following characteristics: long-growing season, high nutrient uptake,
high moisture tolerance, high salt tolerance, and non-leguminous. Some
grasses which have worked well in high strength wastewater land application
systems are Reed Canary, Kentucky Tall Fescue, Jose Wheatgrass, Blue
Panicum, and NK-37 Bermuda. Other grasses may be acceptable in particular
parts of the country. Leafy plants should be avoided when sprinkler
application is used because of leaf burn due to the high total dissolved
solids concentration in most meatpacking plant wastewaters.
DESIGN CONSIDERATIONS
The design of a land application system must include several considera-
tions. These include loading rates, operating pressure, and monitoring.
Loading Rates
For meatpacking plant wastewaters, there are three primary loading
rates which must be taken into consideration in designating a system:
the hydraulic loading rate, the organic loading rate, and the nitrogen
loading rate.
The hydraulic loading rate is rarely the limiting loading rate for
meatpacking plant wastewaters because of the high Biochemical Oxygen Demand
473
-------
and nitrogen concentrations usually present in the plant effluent. Therefore,
the hydraulic loading rate for most meatpacking plant wastewater would
be less than 4 in/wk. Such a low hydraulic loading rate would dictate
an infiltration system with sprinkler application in order to achieve
a uniform distribution. Mhen the wastewater has been treated beyond
primary treatment, however, overland flow or infiltration via flooding
can be considered. Witherow (5) has shown that a high degree of treat-
ment can be attained through overland flow of meatpacking plant waste-
water which had been previously subjected to treatment in anaerobic lagoons.
Flood infiltration basins can also be used when relatively high quality
wastewater is available and the soil and topographic conditions are accep-
table.
The organic loading rate is usually the controlling factor for meat-
packing plant wastes. The organic loading rate should not exceed 200 Ibs
BOD/acre/day when odor prevention is of primary concern. Higher rates
can be used with good treatment results when the application site is
away from populated areas.
The nitrogen loading rate must also be controlled when prevention of
groundwater deterioration is of primary concern. The wastewater cannot
be applied in amounts that would provide more nitrogen than the vegetation
can assimilate. The amount of wastewater that could be applied, therefore,
is dependent on the nitrogen concentration in the wastewater and the type
of soil cover used. For an overland flow system, a travel distance of
at least 200 feet should be used. For high strength wastewaters, a longer
travel distance will improve effluent quality.
Pressure
Land application systems have been operated successfully at both
high (90 psi) and low (10 psi) working pressures. The low pressure system
has the advantage of reduced aerosol drift during wastewater application,
but the disadvantage of increased initial investment cost. The increased
initial cost could be offset by the lower annual operating costs due to
the low pressure.
Various types of low pressure systems can be used. Figures 1 through
4 show the details of a low cost, low pressure (i.e. 10 psi) wastewater
distributor. Figure 5 shows a very simple low pressure wastewater dis-
tributor suitable for an overland flow system. The absence of moving
parts in this design is very attractive from the standpoint of maintenance.
Other deisgns for low pressure distribution systems are commercially avail-
able.
Monitoring
An adequate monitoring system should be an integral part of any type
of land application system. For an infiltration system, this should include
soil samples for measuring the quality of the percolated wastewater as a
function of downward travel distance. In addition, at least one ground-
water sampling well should be placed in the center of the treatment area
474
-------
TOP VIEW
en
SIDE VIEW
Fig. I. LOW PRESSURE DISTRIBUTOR
-------
1/2" PIPE (1.3cm.)
I 1/4" X 1/2" BUSHING
(3.2 XI.3 cm.)
MT PIPE
-PIPE FROM RISER
•EMT PIPE
NOZZLE
Fig-2 DETAILS OF FITTINGS FOR DISTRIBUTOR
476
-------
TO DISTRIBUTOR
5CM. PIPE (sch 80)
LARGE WASHER
ADAPTER
TO
VALVE
CONCRETE
Fig.3 DISTRIBUTOR RISER
(See Fig 4 for exploded view)
477
-------
MT"
i
TO DISTRIBUTOR
•3 SCREWS
3 SET
SCREWS
THRUST BEARING AETNA E20
LARGE WASHER WELDED TO
3" PIPE, BEARING PLACED
INSIDE (7.6cm.)
2" PIPE (5.0cm.)
1 1/4" PIPE (3.2cm.)
ADAPTER 3*X2"
(7.6 X 5.0cm.)
RADIAL BEARING
MRC 6208ZZ
SWIVEL
90*ELBOW
IPPLE
Fifl.4. DETAILS OF CONSTRUCTION OF THE RISER
478
-------
NOZZLE
FRONT VIEW
(3.8cm) I 1/2" NIPPLE
I l/2"-90* ELBOW
(3.8 cm.)
I I/2"RISS
5' LONG
(3.8cm.XI.5m.r
\
FEED LINE-
ISER
DISTRIBUTOR
I l/2"-90° ELBOW
(3.8cm.)
BUSHING
BETE TYPE FF
FOG NOZZLE
Risers art spaced
20 opart (6m.)
DISTRIBUTOR PIPE
SIDE VIEW
Fig.5 LOW PRESSURE RUNOFF AREA DISTRIBUTOR
479
-------
and one placed at the edge of the treatment area in the direction of the
groundwater flow. Well water samples should be taken at least monthly
in order to observe any possible deterioration of the groundwater.
For an overland flow system only a very small amount of applied waste-
water should reach the groundwater table. However, at least one well should
be placed in the treatment area in order to observe any possible contamination
of the groundwater by the treatment system.
Sampling of the wastewater applied should also be done regularly so
that the efficiency of the system can be determined. These results will
also serve as a basis for future expansion.
TREATMENT EFFICIENCY
For a properly operated treatment system, a high removal efficiency
can be expected for most of the constituents of meatpacking plant wastewater.
Table 2 shows the treatment efficiency for the infiltration system operated
at the Peyton Packing Company for 2.5 and 4 in/wk application rates.
TABLE 2. INFILTRATION SYSTEM TREATMENT EFFICIENCY
Percent Removal
Parameter 2.5 inch/week 4 inch/week
Grease
COD
Nitrogen
SS
100
99
72
100
100
98
58
100
The average efficiency values shown in Table 2 were calculated from
the wastewater values shown in Table 3, which represent the average strength
of the wastewater that was applied to the treatment areas. These values
are greater than the 24 hour average values shown in Table 1 because
wastewater was applied only during daytime hours, at which time the waste-
water strength was considerably higher than the average values.
TABLE 3. STRENGTH OF APPLIED WASTEWATER
Parameter Concentration, mq/1
BOD 5,800
COD 9,400
480
-------
TABLE 3. STRENGTH OF APPLIED WASTEWATER (Continued)
Parameter Concentration, mg/1
Grease 2,600
Total Kjeldahl Nitrogen 160
SS 3,140
As shown in Table 2, all of the grease was removed at both the 2.5
and 4 in/wk application rates. This is as expected, since most of the
grease that was present in the wastewater was in the form of coagulated
particles. The filtration of the grease by the soil was evident from
visual observation of the soil surface after several weeks of wastewater
application. However, after drying periods of one to two weeks duration,
periodic soil analysis failed to show measurable amounts of grease in.the
surface soil samples (i.e., 0-1 inch) during the entire time this study
was conducted.
The COD removal was excellent at both wastewater application rates
studied, particularly in view of the extremely high organic loading rates
employed. These rates were 1,975 Ibs COD/acre/wk at the 2.5 inch rate
and 3,293 Ibs COD/acre/wk at the 4 inch rate. These results show
the excellent treatability of meatpacking plant wastewater by land appli-
cation.
The high organic loading rates caused some problems with infiltra-
tion and odors, particularly at the higher loading rate. At approximately
3 week intervals, it was necessary to omit one or more of the application
periods in order to allow adequate drying time. If this were not done,
the reduced infiltration rate would cause ponding with subsequent odor,
and insect problems.
Nitrogen removal was also very good when the high loading rates are
taken into account. The 2.5 in/wk hydraulic loading rate resulted in a
1,750 Ibs/acre/yr loading while the nitrogen loading at the 4 in/wk rate
was 2,914 Ibs/acre/yr. Both values are much higher than the normal
application rates recommended for agricultural fertilization. In terms
of nitrogen removal 1,259 and 1,691 Ibs/acre/yr were removed at the 2.5
and 4 in/wk application rates respectively. The nitrogen removals obtained,
therefore, obviously could not be accounted for in plant uptake alone.
For the overland flow system, the results are shown in Table 4 below.
481
-------
TABLE 4. OVERLAND FLOW TREATMENT RESULTS
Influent Cone, Effluent Cone,
Parameter mg/1 mg/1 % Removal
COD
Kjeldahl Nitrogen
Total Solids
Volatile Total Solids
Suspended Solids
Volatile Suspended Solids
9,440
185
7,430
4,690
4,130
3,825
1,495
108
3,420
1,125
535
420
84
44
54
76
87
89
As with the infiltration treatment system, the overall efficiency of
the overland flow system was excellent when the extremely high organic
loading rate is taken into consideration. The COD removal, for example,
was 1,800 Ibs/acre/day, or in terms of BOD, approximately 900 Ibs/acre/day.
Slight odors were noticeable when the system was operated for more
than three consecutive weeks without resting, but the odors disappeared
quickly after irrigation was stopped.
Although there was a noticeable deposition of greasy looking solids
at the edges of the sprinkler application area, these solids turned black
and dried very well during resting periods. Utilization of a proper opera-
ting and drying schedule, therefore, seems to eliminate any problems which
might be caused by solids build-up.
SUMMARY
The treatability of high strength meatpacking plant wastewater by
land application has been shown to be excellent for both infiltration and
overland flow type systems. With respect to organic carbon removal, both
infiltration-percolation and overland flow systems have been shown to be
very effective. The advantage of higher efficiency obtained with the
infiltration system is offset somewhat by the more expensive and complicated
distribution system involved. Additionally, recovery of the treated waste-
water from an infiltration system would be more difficult if subsequent
reuse were anticipated. There is also less likelihood of polluting potable
water supplies with an overland flow system.
Nitrogen removal was also slightly better with the infiltration system
than with overland flow. However, efficient nitrogen removal depends to
a great extent on closely controlled operating conditions.
The potential for phosphorous removal is obviously greater in an infil-
tration system than overland flow. Where phosphorous removal is of primary
importance, therefore, infiltration systems offer a definite advantage
482
-------
compared to overland flow. On the other hand, when soil conditions are
not favorable for phosphorous removal and chemical treatment must be em-
ployed, there are few considerations which would favor one system over the
other.
POTENTIAL PROBLEMS
There are several potential problems associated with land application
of meatpacking plant wastewater. One of these which requires immediate
consideration is the possibility of the presence of pathogenic bacteria
in the wastewater. Brucella, Streptococcus, Staphlococcus, Salmonella, and
Shigella bacteria have been isolated from the effluent of the catch basin
on several occasiosns. The extent of this problem is not known at this
time but should have a high priority in future studies or in proposed
installations.
The problem of pathogenic bacteria present in the wastewater is
compounded by the drifting of aerosols formed during application of the
wastewater. Under even slight wind conditions (5-10 miles/hour) viable
bacterial cells have been measured as far as 500 feet downwind for the high
pressure system. For the low pressure distribution system, the maximum
travel distance downwind was less than 100 feet. It is obvious, therefore,
that this problem should be given careful consideration in designing a
distribution system, with lower operating pressures preferred.
The high TDS concentration in most meatpacking plant wastewaters is
caused by sodium chloride. As a result, most packing plant wastewaters
have a very unfavorable sodium absorption ratio. This would cause serious
problems with infiltration in clay-containing soils unless amendments
were added. Sandy type soils are generally not affected by unfavorable
sodium adsorption ratios and therefore are generally best suited for ac-
cepting meatpacking plant wastes as they leave the plant.
Finally, the high concentration of nitrogen present in most meat-
packing plant wastewaters presents a potential problem of ground water
pollution. The experience gained with soil treatment systems so far indi-
cates that close control of the treatment system is required in order to
remove greater than 50 percent of the nitrogen. Even then, the high con-
centration originally present could cause significant amounts of nitrogen
to reach the ground water table.
REFERENCES
1. Waring, George. Modern Methods of Sewage Disposal. London, D. Van
Nostrand. 1903.
2. Land Applications of Sewage Effluents and Sludges: Selected Abstracts,
Water Quality Control Branch, Robert S. Kerr Water Research Center;
Ada, Oklahoma. Publication Number EPA-660/2-74-042. June 1974.
248 p.
483
-------
3. Sullivan, R. H., M. M. Coh, and S. S. Baxter. Survey of Facilities
Using Land Application of Wastewater. Environmental Protection Agency,
Washington, DC. Publication Number EPA-430/9-73-006. July 1973.
377 p.
4. Environmental Quality. Joint Task Force, U.S. Dept. of Agriculture
and State University. 1967. p. 26.
5. Witherow, Jack L. Small Meat-Packers Waste Treatment Systems. Pro-
ceedings Fourth National Symposium on Food Processing Wastes, EPA-
660/2-73-031. U.S. Government Printing Office, December 1973.
484
-------
SEVENTH FOOD WASTE SYMPOSIUM
REGISTRATION LIST
C. D. AKRIDGE
Swift Fresh Meats Company
P.O. Box 729
Moultrie, GA 31768
DON ALLEN
GA Dept. of Natural Resources
270 Washington St., SW
Atlanta, GA 30334
JAMES B. ALLEN
MS State University
Box 5465
Mississippi State, MS 39762
GEORGE BARNES
City of Atlanta
302 City Hall
Atlanta, GA 30303
JOHN C. BARNES, JR.
Standard Products Co.
Box 389
Kilmarnock, VA 22482
J. H. BAUER
96 Popular St., NW
Atlanta, GA 30303
MARTHA BEACH
N-CON Systems Co.
308 Main Street
New Rochelle, NY 10801
GEORGE BELIEW
Spring Valley Foods
P.O. Box 3508
Oxford, AK 36201
LAWSON BELL, JR.
Robert & Co. Assoc.
96 Poplar St., NW
Atlanta, GA 30303
JOHN BENEMANN
SERL
University of CA
Berkeley, CA 94720
P. M. BERTHOUEX
University of WI
Madison, WI 53705
D. WAYNE BISSETT
Water Pollution Control Directorate
Environment Canada
Ottawa, Ontario K1A OH3
DAN BLACKSHEAR
J-M Poultry
422 N. Washington
El Dorado, AR 71730
JOHN BOMBEN
USDA-WRRL
800 Buchanan Street
Berkeley, CA 94710
HENRY BOROW
Standard Brands Ltd.
550 Sherbrooke St. W.
Montreal, Quebec H3A 1B9
GARY J. BOTTOMLEY
Holly Farms Poultry Inc.
P.O. Box 88
Wilkesboro, NC 28697
485
-------
WAYNE BOUGH
University of GA
Dept. of Food Services
Experiment, GA 30212
DAVE BRIDGES
SWECO Inc.
8040 U.S. Hwy. 25
Florence, KY 41042
6. M. BROOKS
MS Coop. Ext. Service
82 Plaza
Starkville, MS 39759
CALVIN G. BROWN
Brown Devlin Associates
301 Meyer Road
West Seneca, NY 14224
RICHARD M. BUCKLEY
R. C. Noonan, Inc.
P.O. Box 1388
Greenville, SC 29602
THOMAS P. BUCKLEY
Chemtron Food/Process Systems
5323 South Western Blvd.
Chicago, IL 60609
GREG BUCOVE
University of WA
College of Fisheries
Seattle, WA 98195
ROBERT W. BURNS
J-M Poultry Packing Co.
P.O. Box 1758
El Dorado, AR 71730
CLARK CALLAWAY
NCSU Seafood Laboratory
P.O. Box 51
Morehead City, NC 28557
THOMAS CAMPBELL, JR.
University of GA
Georgia Station
Experiment, GA 30212
GERALD CARLSON
North Carolina State
Raleigh, NC 27207
WALLACE CARPENTER
Environmental Protection Div.
270 Washington St., SW
Atlanta, GA 30334
JOHN F. CHAPPLE
Bio-Viro-D, Inc.
2045 Spafford Ave.
West Palm Beach, FL 33409
MARVIN CHARLES
LeHigh University
Bethlehem, PA 18015
T. C. CHEN
MS State University
Box 5188
Mississippi State, MS 39762
REX E. CHILDS
Russell Research Center, ARS
720 Sunnyside Drive
Athens, GA 30601
H. S. CHRISTIANSEN
Carnation Company
5045 Wilshire Blvd.
Los Angeles, CA 90036
STEPHEN A. COHEN
Illinois Water Treatment Co.
840 Cedar Street
Rockford, IL 61105
CARL A. COLD
Holly Farms Poultry Inc.
P.O. Box 8
Temperanceville, VA 23442
ROY R. COWELL
Mountainaire Corp.
Suite 1050, Plaza West Bldg.
Little Rock, AR 72205
EDWIN COX III
Edwin Cox Associates
P.O. Box 8025
Richmond, VA 23223
486
-------
RICHARD W. GRAIN
Industrial Filter & Pump Mfg.
5900 Ogden Avenue
Cicero, IL 60650
MICHAEL S. CREASON
6A Environmental Protection Div.
270 Washington St., SW
Atlanta, 6A 30334
BARNEY CULVER
Culver Duck Farm
12215 CR 10
Middlebury, IN 46540
BOB D'AGOSTARO
CH2M-Hill
1930 Newton Square
Reston, VA 22090
STEPHEN D. DANIEL
Central Soya
Box 907
Canton, GA 30114
D. 0. DENCKER
Oscar Mayer & Co.
P.O. Box 1409
Madison, WI 53701
LARRY DEWBERRY
Barker International, Inc.
P.O. Box 1308
Marietta, GA 30061
JESS C. DIETZ
Clark, Dietz & Assoc. Engrs.
211 North Race Street
Urbana, IL 61801
STEVE DONATIELLO
Ralston Purina Co.
835 South 8th St.
St. Louis, MO 63188
LEO A. EBEL, JR.
Consoer, Townsend & Assoc.
3470 Hampton Avenue
St. Louis, MO 63139
CRAIG S. EKERMEYER
Brown-Mi Her Co.
P.O. Box 158
Wiggins, MS 39577
THOMAS E. ELLIOTT
Swift Fresh Meats Co.
115 W. Jackson Blvd.
Chicago, IL 60604
S. DAVID ELLISON
CHzM-Hill
1930 Newton Square
Reston, VA 22090
E. E. ERICKSON
Northstar Division MRI
3100 38th Avenue S.
Minneapolis, MN 55406
TERRY ERVIN
Development Engr.
Turbitrol Company
415 E. Paces Ferry Rd.
Atlanta, GA 30305
JOHN B. ESKEW
Gold Kist Research Ctr.
P.O. Box 388
Lithonia, GA 30058
L. CLIFF EVANS
AL State Health Dept.
State Office Bldg.
Montgomery, AL 36130
JOHN W. FARQUHAR
American Frozen Food Inst.
919 18th Street, NW
Washington, DC 20006
WILLIAM J. FARRISEE
International Bakerage, Inc.
3300 NE Expressway
Atlanta, GA 32041
RICHARD T. FERRY
Bio-Viro-D, Inc.
2045 Spafford Ave.
West Palm Beach, FL 33409
487
-------
GUNNAR FINNE
Clemson Univ. Ext. Service
P.O. Box 3158
Charleston, SC 29407
D. W. FLOOD
Maple Leaf Mills Ltd.
417 Queen's Quay W.
Toronto, Ontario M5W 1C7
HAROLD E. FORD
Southeastern Poultry
1456 Church Street
Decatur, GA 30030
& Egg Assoc.
ALAN FOSTER
Canada Packers Ltd.
95 St. Clair Avenue W.
Toronto, Ontario M4V 1P2
R. A. GALLOP
University of Manitoba
Food Science Department
Winnipeg, Manitoba R3T 2N2
FRED A. GARDNER
Texas Agric. Experiment Station
Texas A&M University
College Station, TX 77840
JEAN R. GEISMAN
The Ohio State University
2001 Fyffe Court
Columbus, OH 43210
STEVEN GLASS
Maplewood Poultry Co.
Belfast, ME 04915
K. C. GOEL
E. H. Richardson Assoc.
P.O. Box 935
Dover, DE 19901
JOHN H. GREEN
U.S. Dept. of Commerce
Southeast Utilization Res.
College Park, MD 20740
Ctr.
DAVID L. GROTHMAN
Oscar Mayer & Co.
910 Mayer Avenue
Madison, WI 53704
C. FRED GURNHAM
Gurnham & Assoc., Inc.
223 W. Jackson Blvd.
Chicago, IL 60606
REGINALD L. HANDWERK
Food & Drug Admin.
200 C Street SW
Washington, DC 20204
GEORGE HARRISON IV
Knoxville Poultry & Egg Co.
P.O. Box 3220
Knoxville, TN 37917
ROBERT L. HAYNES
MS State University
P.O. Box 542
Mississippi State, MS 39762
R. L. HAYS
CPC International
Box 345
Argo, IL 60501
LEVEN HENDERSON
Pomona Products Co.
P.O. Drawer B
Griffin, GA 30224
ROBERT C. HOEHN
VA Polytechnic & State Univ.
Civil Engineering Dept.
Blacksburg, VA 24061
CHUCK HOFFMAN
Magnuson Engr. Inc.
1010 Timothy Drive
San Jose, CA 95133
RAY HOLYFIELD
GAMCO
Gainesville, GA
488
-------
JOSEPH A. HOMAL
Foth & Van Dyke & Assoc.
1970 South Broadway
Green Bay, WI 54304
JOHN L. HOPKINS III
VA Egg Producers
McGaheysville, VA 22840
DENNIS L. JOHNSON
Swift Fresh Meats Co.
115 W. Jackson Blvd.
Chicago, IL 60604
G. ROBERT JOHNSON
Popham, Haik, Schnobrich,
Kaufman & Doty, Ltd.
4344 IDS Center
Minneapolis, MN 55402
JOE M. JONES
Peterson Industries
Box 128
Decatur, AR 72722
CHRISTINE A. KAHR
James M. Montgomery Consulting
Engineers, Inc.
1990 N. California Blvd.
Walnut Creek, CA 94596
ERWIN KIEFER
Patrick Cudahy, Inc.
4801 S. Kingan Avenue
Cudahy, WI 53110
MONTE W. KORB
Korb Engineering Co.
609 Beachview Drive
St. Simons, GA 31522
VACLAV KRESTA
New Brunswick Dept. of the
Environment
P.O. Box 6000
Fredericton, NB
ARUN R. KUMBHARE
Environment Canada
P.O. Box 2406
Halifax, Nova Scotia
K. D. KURTZ
Clark S'Vicario Corp.
9620 Executive Ctr. Dr. N.
St. Petersburg, FL 33702
KENNETH LA CONDE
SCS Engineers
4014 Long Beach Blvd.
Long Beach, CA 90807
JOSE S. LAGAHIT III
Libby, McNeil & Libby
200 S. Michigan Ave.
Chicago. IL 60684
JOHN S. LAMICA
The Turbitrol Co.
415 E. Paces Ferry Rd.
Atlanta, GA 30305
WILLIAM LANHAM
Green Acre Farms Inc.
P.O. Drawer B
Sebastopol, MS 39359
JEAN LAPERRIERE -
Services de Protection
de 1'Environnement
255 Cremaxie est
Montreal, Quebec H2M 1L5
LES LASH
ENVIROTECH
Box 300
Salt Lake City, UT 84110
JOE L. LAY, JR.
Lay Packing Co., Inc.
P.O. Box 2447
Knoxville, TN 37901
FERNAND L'ECUYER
Cooperative Federee de Quebec
Montreal, Quebec
JEAN S. LENNON
FMC Corporation
1185 Coleman Avenue
Santa Clara, CA 95052
B3J 3E4
489
-------
SERGE LESSARD
Villina, Fortier & Assoc.
3300 Cavendish Suite 385
Montreal, Quebec H4B 2M8
HUDA S. LILLARD
USDA-ARS
Russell Research Center
Box 5677
Athens, 6A 30601
S. S. LIN
Kramer, Chin & Mayo, Inc.
1917 1st Avenue
Seattle, WA 98101
J. ALEC LITTLE
U.S. EPA
1421 Peachtree St.
Atlanta, GA 30309
EDMOND P. LOMASNEY
Southeast Region, EPA
1421 Peachtree Street
Atlanta, GA 30309
A. W. LOVEN
Engineering Science, Inc.
57 Executive Park South
Atlanta, GA 30329
DONALD MacGREGOR
O'Brien & Gere Engineers, Inc.
1231 Wildflower Drive
Webster, NY 14580
ELTON MADDOX
Wayne Poultry Co.
P.O. Box 69
Pendergrass, GA 30567
RICHARD F. MATTHEWS
University of FL
325 Food Science Bldg.
Gainesville, FL 32611
ROBERT G. MCELROY
USDA-Russell Res. Ctr.
Athens, GA 30601
CHARLES V. McREYNOLDS
Blue Star Foods
1023 4th Street
Council Bluffs, IA 51501
ROBERT E. MEANS
Bouillon, Christofferson
& Schairer
5050 Washington Bldg.
Seattle, WA 98101
WALTER A. MERCER
National Canners Assoc.
1950 Sixth Street
Berkeley, CA 94710
JERRY R. MEYER
Ollie Meyer, Inc.
1584 Tullie Circle
Atlanta, GA 30328
JAMES M. MOREAU
L. A. Frey, Inc.
P.O. Box 51507, OCS
Lafayette, LA 70501
DALE MORRIS
Mar-Jac Inc.
P.O. Box 49
Gainesville, GA 30501
FRANK B. MURPHY
ALCOA
ATC - Bldg. D
Alcoa Center, PA 15069
KEVIN McCANN
Autotrol Corp.
5855 N. Glen Park Rd.
Glendale, WI 53209
DANIEL T. McGRAIL
State of MD Dept. of
Health & Mental Hygiene
201 E. Preston Street
Baltimore, MD 21201
DAVID J. McNAIR
Allied Mills, Inc.
110 N. Wacker Drive
Chicago, IL 60606
490
-------
LAURA McQUINN
American Meat Inst.
P.O. Box 3556
Washington, DC 20007
JACK McVAUGH
Envirex Inc.
Box 1067
Waukesha, MI 53186
CARL E. NALL
Pacific Egg & Poultry Assoc.
5320 W. Jefferson Blvd.
Los Angeles, CA 90016
CARL D. NELSON
AL State Health Dept.
State Office Bldg.
Montgomery, AL 36130
PETER NI
Patrick Cudahy, Inc.
4801 S. Kingan Ave.
Cudahy, WI 53110
LEON A. NOLTING
A. E. Staley Mfg.
East Past Road
Monesville, PA 19067
RAYMOND L. NOTARO
Campbell Soup Co.
Campbell Place
Camden, NJ 08101
G. B. OGLESBY
GA Dept. of Natural Resources
270 Washington St., SW
Atlanta, GA 30334
C. M. PARKS
MS Cooperative Extension
State College, MS 39762
JOHN P. PILNEY
North Star Dlv. MRI
3100 38th Avenue S.
Minneapolis, MN 55406
JEFF PINTENICH
Engineering Science, Inc.
57 Executive Park South, NE Suite 590
Atlanta, GA 30329
NAGARAJA P. PGO
Arthur G. McKee & Co.
10 S. Riverside Plaza
Chicago, IL 60606
MARSHALL W. RAY
CAN TEX Industries
P.O. Box 340
Mineral Wells, TX 76067
GENE REECE
Peterson Industries
Box 128
Decatur, AR 72722
JIM ROBINSON
Sanderson Farms, Inc.
P.O. Box 2937
Laurel, MS 39440
WALTER W. ROSE
National Canners Assoc.
1950 Sixth Street
Berkeley, CA 94710
M. L. ROWE
East Central University
Ada, OK 74820
JOHN RUGABER
Pet, Inc.
St.Louis, MO
LARRY L. RUSSELL
James M. Montgomery Consulting
Engineers, Inc.
1990 M. California Blvd.
Walnut Creek, CA 94596
WILLIAM D. RUTZ
MAPCO, Inc.
1437 S. Boulder Ave.
Tulsa, OK 74119
491
-------
JOHN E. SAUVA6E
Swift Dairy & Poultry Co.
115 W. Jackson Blvd.
Chicago, IL 60604
GEORGE A. SCHULER
University of GA
Athens, GA 30601
WILLIAM SCHULTZ
USDA-WRRL
800 Buchanan St.
Albany, CA 94710
LARRY J. SCULLY
Peat, Marwick & Mitchell Co.
1025 Connecticut NW
Washington, DC 20036
KHEM SHAHANI
University of NE
Dept. of Food Science
Lincoln, NE 68583
ED SHEFFIELD
Lou-Ana Foods, Inc.
P.O. Box 591
Opelousas, LA 70570
ROBERT SHERMAN
Wilson & Co.
4545 N. Lincoln Blvd.
Oklahoma City, OK 73132
A. L. SHEWFELT
GA Experimental Station
Experiment, GA 30212
LACY W. SIMMONS
Simmons Engineering Co.
P.O. Box 157
Roswell, GA 30075
JEFFRY P. SMITH
GA Dept. of Natural Resources
270 Washington St., SW
Atlanta, GA 30334
MIKE SMITH
McGill Grogan & Assoc.
P.O. Box 1454
Gainesville, GA 30501
R. E. SPEECE
Drexel University
32nd & Chestnut
Philadelphia, PA 19104
R. G. STENSGAARD
SWECO, Inc.
P.O. Box 4151
Los Angeles, CA 90051
NATHAN L. STEPHENS
Gold Kist Research Center
2230 Industrial Blvd.
Lithonia, GA 30058
RICHARD W. STERNBERG
U.S. EPA - Permits Division
Crystal Mall #2
Washington, DC 20460
CHARLES STEVENSON
Curtice-Burns, Inc.
P.O. Box 670
Rochester, NY 14602
HERBERT E. STONE
Del Monte Corporation
P.O. Box 3575
San Francisco, CA 94119
RAY STROUP, JR.
Lou-Ana Foods
112 Baudoin
Lafayette, LA 70501
GILMAN SYLVESTER
R. P. General Mgr.
Oakland, MD 21550
A. J. SZABO
Domingue, Szabo & Assoc.
P.O. Box 52115
Lafayette, LA 70505
ANTHONY J. TARQUIN
Univ. of TX at El Paso
Dept. of Civil Engineering
El Paso, TX 79968
492
-------
VIRGINIA C. TAYLOR
Ruel Taylor, Inc.
198 Main Street
Gorham, ME 04038
MORRIS A. THOMAS
J-M Poultry Packing Co.
P.O. Box 1810
Gainesville, GA 30501
JAMES E. THOMSON
USDA-Russell Res. Ctr.
P.O. Box 5677
Athens, GA 30604
NORA K. THUMA
Atlantic Research Corp.
5390 Cherokee Avenue
Alexandria, VA 22314
TERRY C. TITUS
Clemson University
Food Science Dept.
Clemson, SC 29631
FRANK TODD
MFC Services (AAL)
P.O. Box 449
Jackson, MS 39205
WARREN K. TROTTER
USDA-Russell Res. Ctr.
P.O. Box 5677
Athens, GA 30604
RONALD A. TSUGITA
James M. Montgomery Consulting
Engineers, Inc.
1990 N. California Blvd.
Walnut Creek, CA 94596
J. M. Vandepopuliere
University of MO-Columbia
Poultry Dept., T-14
Columbia, MO 65201
BILL 0. VAUGHN
Ralston Purina Co.
835 S. 8th Street
St. Louis, MO 63188
YVONNE VIZZIER
Marshall Durbin Co.
Ill Mill saps Avenue
Jackson, MS 39202
K. S. WATSON
Kraftco Corp.
801 Waukegan Rd.
Glenview, IL 60025
EDWARD K. WELLMEYER
Swift and Co.
115 W. Jackson Blvd.
Chicago, IL 60604
OTTO W. WENDEL
Bio-Viro-D, Inc.
2045 Spafford Ave.
West Palm Beach, FL 33409
W. K. WHITEHEAD
USDA-Russell Res. Ctr.
P.O. Box 5677
Athens, GA 30604
ROGER WILKOWSKE
Extension Service
USDA
Washington, DC 20250
GEORGE E. WILSON
EUTEK
1828 Tribute Rd., Suite H
Sacramento, CA 95815
SAM K. WINFREE
University of TN
P.O. Box 1071
Knoxville, TN 37901
FRANK WOODARD
University of ME
350 Aubert Hall
Orono, ME 04473
MALCOLM E. WRIGHT
VPI & SU
Agric. Engr. Dept.
Blacksburg, VA 24061
493
-------
ARNOLD C. M. WU
University of GA
6A Exp. Station
Food Science Dept.
Griffin, GA 30212
FRED S. YOUNG
Environmental Products, Inc.
P.O. Box 2385
Hickory, NC 28601
ROBERT R. ZALL
Cornell University
Dept. of Food Science
Ithaca, NY 14853
BARRY ZUBKE
Mar-Jac, Inc.
P.O. Box 49
Gainesville, GA 30501
494
-------
TECHNICAL REPORT DATA
{Please read Instructions on the reverse before completing)
. REPORT NO.
EPA-600/2-76-304
4. TITLE AND SUBTITLE
2.
3. RECIPIENT'S ACCESSION NO.
Proceedings Seventh National Symposium on Food
Processing Wastes
6. PERFORMING ORGANIZATION CODE
5. REPORT DATE
December 1976 issuing date
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAIVE AND ADDRESS
Food and Wood Products Branch
Industrial Environmental Research Laboratory
200 SW 35th Street
Con/aliis, Oregon 97330
10. PROGRAM ELEMENT NO.
1BB610
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Researcn Laboratory-Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati. OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
Symposium Proceedings
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
.... .
Additional sponsors include: National Canners Association,
American Meat Institute, Southeastern Poultry & Egg Assoc., and Pacific Egg & Poultry
16. ABSTRACT
The Proceedings contains copies of 26 of the 27 papers presented at the
Symposium. Subjects included: wastewater characterization, product and by-product
recovery, processing modifications, wastewater treatment, and water reuse for many
different segments of the food processing industry. Industrial segments included:
red meat and poultry, seafood, dairy, fruit and vegetable, and beverage.
Attendance at the two and one-half day Symposium was approximately 200 with
good representation by industry, universities, consulting firms as well as state
and Federal agencies.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Industrial Wastes, Waste Water Treatment,
By-Products, Foods
Process Modifications,
Waste Characterization,
Water Reuse
13/B
8. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport/
Unclassified
21. NO. OF PAGES
503
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
£ U.S. GOVERNMENT PRINTING OFFICE: 1977-757-056/5535 Region No. 5-11
495
------- |