WATER POLLUTION CONTROL RESEARCH SERIES
12060—03/71
PROCEEDINGS
SECOND NATIONAL SYMPOSIUM
ON FOOD PROCESSING WASTES
CO-SPONSORED BY:
EPA PACIFIC NORTHWEST WATER LABORATORY -
NATIONAL CANNERS ASSOCIATION
MARCH 23-26, 1971 DENVER, COLORADO
ENVIRONMENTAL PROTECTION AGENCY • RESEARCH AND MONITORING
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes
the results and progress in the control and abatement
of pollution of our Nation's waters. They provide a
central source of information on the research, develop-
ment, and demonstration activities of the Environmental
Protection Agency, through inhouse research and grants
and contracts with Federal, State, and local agencies,
research institutions, and industrial organizations.
Inquiries pertaining to the Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Office of Research and Monitoring, Environmental
Protection Agency, Room 801, Washington, DC 20242.
-------
PROCEEDINGS
SECOND NATIONAL SYMPOSIUM ON FOOD PROCESSING WASTES
Co-sponsored by:
Pacific Northwest Water Laboratory, EPA
and
National Canners Association
March 23-26, 1971
Denver, Colorado
For ial« by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20*02 - Price M.SO
Stock Number 5501-0167
-------
EPA Review Notice
This report has been reviewed by the Water Quality
Office of the Environmental Protection Agency and
approved for publication. Approval does not signify
that the contents necessarily reflect the views and
policies of the Environmental Protection Agency, nor
does mention of trade names or commercial products
constitute endorsement or recommendation for use.
fi
-------
FOREWORD
This Second National Symposium on Food Processing Wastes
was co-sponsored by the Pacific Northwest Water Laboratory
of the U. S. Environmental Protection Agency, and the
National Canners Association. The meeting was held this
year in Denver, Colorado, in keeping with the plan to rotate
the meeting among different geographical locations.
The meeting was expanded to four days in order to provide
time for discussion of air pollution and solid waste programs
as well as the reports on new waste treatment research. It
is hoped that these meetings have set a precedent for a
continuing forum for researchers and industry officials to
discuss questions and answers on environmental protection.
iii
-------
CONTENTS
SYMPOSIUM OBJECTIVES 1
James R. Boydston
ENVIRONMENTAL QUALITY CONCERNS OF THE FOOD PROCESSING
INDUSTRY 3
Alvin H. Randall
EPA'S INDUSTRIAL POLLUTION CONTROL PROGRAM 13
Dr. Stanley M. Greenfield
THE CLEAN AIR AMENDMENTS OF 1970 and AIR POLLUTION ASPECTS
OF THE FOOD AND AGRICULTURAL PROCESSING INDUSTRY 21
Dr. S. David Shearer
ACTIVITIES IN MANAGING SOLID WASTES 41
Jack DeMarco
POLLUTION ABATEMENT AND BY-PRODUCT RECOVERY IN THE
SHELLFISH INDUSTRY 51
Edwin Lee Johnson and Dr. Quintin P. Peniston
SALT RECLAMATION FROM FOOD PROCESSING BRINES 75
E. Lowe and E. L. Durkee
REDUCTION OF SALT CONTENT OF FOOD PROCESSING LIQUID
WASTE EFFLUENT 85
Dr. Jack W. Rails, Walter A. Mercer and Nabil L. Yacoub
PRODUCTION AND DISPOSAL PRACTICES FOR LIQUID WASTES FROM
CANNERY AND FREEZING FRUITS AND VEGETABLES 109
Walter W. Rose, Walter A. Mercer, Allen Katsuyama,
Richard W. Sternberg, Glen V. Brauner, Norman A. Olson,
and Dr. Kenneth G. Weckel
v
-------
PROGRESS REPORT: STUDY OF DRY CAUSTIC VS CONVENTIONAL
CAUSTIC PEELING AND THE EFFECT ON WASTE DISPOSAL 129
Joseph W. Cyr
DRY CAUSTIC PEELING OF TREE FRUIT TO' REDUCE LIQUID WASTE
VOLUME AXD STRENGTH 137
Dr. Jack W. Rails, Walter A. Mercer, Robert P. Graham,
Mark R. Hart, and Harry J. Maagdenberg
PRODUCTION OF POTATO STARCH WITHOUT WASTE 169
Roy Shaw and W. C. Shuey
ECONOMIC ANALYSIS OF ALTERNATIVE METHODS FOR PROCESSING
POTATO STARCH PLANT EFFLUENTS ..... 185
R. L. Stabile, V. A. Turkot, and N. C. Aceto
CONTINUOUS TREATMENT OF CORN AND PEA PROCESSING WASTE WATER
WITH FUNGI IMPERFECTI 203
Dr. Brooks D. Church, Dr. Harold A. Nash, Dr. Eugene E.
Erickson, and Willard Brosz
CANNERY WASTE TREATMENT WITH RBC AND EXTENDED AERATION
PILOT PLANTS 227
R. J. Burm, M. W. Cochrane, and K. A. Dostal
CANNERY WASTE TREATMENT BY LAGOONS AND OXIDATION DITCH
AT SHEPPARTON, VICTORIA, AUSTRALIA 251
Dr. C. D. Parker and G. P. Skerry
BIOLOGICAL TREATMENT OF CITRUS PROCESSING WASTEWATERS 271
Dr. F. A. Eidsness, J. B. Goodson, J. J. Smith, Jr.
TREATMENT OF MEAT PACKING WASTE USING PVC TRICKLING FILTERS . . 287
Darrell A. Baker and James White
VI
-------
DEHYDRATION OF CATTLE RUMEN AND WHOLE BLOOD 313
Dr. Donald J. Baumann
WATER AND WASTE MANAGEMENT IN POULTRY PROCESSING 323
Dr. W. M. Crosswhite, R. E. Carawan, and John A. Macon
REMOVAL AND RECOVERY OF FATTY MATERIALS FROM EDIBLE OIL
AND FAT REFINERY EFFLUENTS 337
W. C. Seng
BIOLOGICAL TREATMENT OF HIGH BOD YEAST WASTES 357
Thomas P. Quirk
THE BETTER WHEY—A DILEMMA 409
Sidney Boxer
413
MEMBRANE PROCESSING OF COTTAGE CHEESE WHEY FOR POLLUTION
ABATEMENT
Dr. Robert L. Goldsmith, Dr. David J. Goldstein, Bernard S.
Horton, Sohrab Hossain, and Dr. Robert R. Zall
ACTIVATED SLUDGE AND TRICKLING FILTRATION TREATMENT OF WHEY
EFFLUENTS 447
Thomas P. Quirk and Joseph Hellman
METHANE FERMENTATION OF WHEY 501
Dr. C. D. Parker
STATE OF THE ART OF DAIRY FOOD PLANT WASTES AND WASTE
TREATMENT 509
Dr. W. James Harper and John L. Blaisdell
yit
-------
ANAEROBIC-AEROBIC PONDS FOR TREATMENT OF BEET SUGAR WASTES ... 547
Dr. William J. Oswald, Dr. Clarence G. Golueke, Dr. Robert C.
Cooper, and Ronald A. Tsugita
STATE-OF-ART SUGARBEET PROCESSING WASTE TREATMENT 597
James H. Fischer and E. H. Hungerford
PRINCIPLES OF NUTRIENT CONTROL FOR AGRICULTURAL WASTEWATERS . . 605
Dr. Raymond C. Loehr
OXIDATION DITCH TREATMENT OF MEAT PACKING WASTES 617
Dr. W. L. Paulson, D. R. Kueck, and W. E. Kramlich
SOLID WASTE MANAGEMENT IN THE FOOD PROCESSING INDUSTRY 637
Henry T. Hudson
REGISTRATION LIST 655
viii
-------
SYMPOSIUM OBJECTIVES
by
James R. Boydston*
We are most pleased with the response you have shown to this our Second
National Symposium on Food Waste Treatment Research. As many of you
know, we started this program last year in Portland with a 3-day sym-
posium primarily concerned with reporting on projects of the Research
and Demonstration Grant Program of the Federal Water Quality Administration.
The enthusiasm shown there and the comments we received prompted us to
schedule these conferences on an annual basis. This Second National
Symposium has already been expanded to include air pollution and solid
wastes and the meeting had to be extended to four days to provide room
for all of the technical papers.
The purpose of these symposia is to provide a forum for formal and in-
formal discussions among all individuals involved in the treatment of
wastes from the food processing industry. The registration shows that
attendees at this symposium are administrative, professional, and tech-
nical people from regulatory agencies, universities, engineering consul-
tants, and private industry. We believe there is no substitute for direct
personal contact such as this to determine research needs and to dis-
seminate information on new research findings.
The food processing industry is one of the major sources of organic
industrial wastes. This industry, perhaps more than any other, has a
wide variety of types of plants due to the many segments of the industry—
ranging through fruit and vegetable canning and freezing; seafood pro-
cessing; the beverage industry, including soft drinks, breweries, wineries,
and distilleries; dairy industry; beet and cane sugar; and grain milling—
to name a few. The treatment requirements for wastes from these segments
are extremely varied.
All industries are currently facing increasing requirements for waste
treatment. Waste discharge permits are now required under the Refuse Act
of 1899 and effluent standards will be imposed. Because of these changes,
there is increasing need to develop better methods of treatment, to develop
by-product recovery methods, to change plant processing to reduce waste at
the source in the plant, and to develop higher degrees of treatment to
permit water reuse. To be of value, information on new methods to accom-
plish these goals must be available to those who can make use of it as fast
as possible.
*Chief, National Waste Treatment Research Program, Pacific Northwest
Water Laboratory, Water Quality Office, Region 10, U. S. Environmental
Protection Agency, Corvallis, Oregon.
-------
The Environmental Protection Agency conducts an active national research
program directed to identifying efficient methods for the treatment of
wastes from the food processing industry. We look upon the dissemination
of research results as important as the research itself. Reports are
published for all in-house and extramural grant projects and made avail-
able for public distribution. We provide technical assistance to regulatory
agencies and to industry directly. We feel, however, that this symposium
provides an excellent opportunity for distributing information on new
research, development and demonstration findings since it gives the
opportunity for direct personal review and discussion of these projects.
Your interest will insure continuation of these annual conferences.
This meeting should not be construed as a government water pollution
research meeting. It is co-sponsored by your industry association,
National Canners Association, and includes technical discussions in the
fields of air, water, and solid waste, and includes reports of research
and development undertaken by industry, universities, engineering con-
sultants, and the Federal Government.
-------
ENVIRONMENTAL QUALITY CONCERNS OF THE
FOOD PROCESSING ItTOUSTRIES
by
Alvin H. Randall
INTRODUCTION
Last year, Mr. James R. Boydston of the Pacific Northwest Water Laboratory,
in opening the First National Symposium on Food Processing Waste, stated,
"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 ... it truly represents a cooperative
coordinated effort between industry and government to mutually solve
troublesome water pollution problems."
Reports coming to me from the Association staff stated that the First
Symposium was most productive. The intensive planning for this Second
Symposium and its comprehensive coverage of research on food waste will
add immeasurably to our knowledge of the treatment and management of
food waste materials.
Speaking as President of the National Canners Association and for its staff
and members, we are pleased and honored to be recognized as a co-sponsor
of this the Second National Symposium on Food Processing Wastes.
As both food processor and citizen, I recognize the importance of the
subject matter you will be discussing. Today, we are all concerned in
varying degrees about our environment. What is the actual state of our
environment? What corrective actions must be taken now, and what long-
range planning should be undertaken? Although the sociological answers
must come from everyone, the technological answers and the implications
and costs for achieving these levels of environmental quality must, for
the most part, come from you the scientists.
We deeply appreciate the excellent cooperation and communication which
exists between the Association, its staff and committees, and the various
staff members of the Federal Water Quality Office. This is a unique and
fruitful relationship for which we are grateful and which could serve as
a model that should prevail between all of industry and all other agencies
of government.
For a number of years, the staff of the Association has worked closely
with state and federal agencies on the development of water quality
*President National Canners Association, Washington, D,C., And Executive
Vice President, United Flav-R-Pac Growers, Inc., Salem, Oregon
-------
criteria and the development of technical and economic data needed for
establishment of reasonable water quality standards. Information on
the nature and volumes of food processing wastes has been developed and
distributed to industry members and all levels of government.
Other relationships between the Association and Federal Water Quality
personnel have included the administration and performance of research
grants and contracts concerned with treatment and management of food
wastes. These cooperative efforts have been of great benefit to the
industry. Not only has new and urgently needed technical information
been developed, but the demonstration aspects of the various projects
have filled a need which would not otherwise have been satisfied.
In the performance of these projects, jointly financed by the Association
with the Federal Water Quality Office, it has been the pleasure of the
Association staff and committees on environmental problems, to know and
work with Mr. Allen Cywin, Director of Applied Science and Technology;
and within that Division, Mr. William Lacy and Mr. George Keeler, both
of the Industrial Pollution Control Branch. Equally enjoyable and help-
ful has been our association with Mr. James Boydston, the Symposium
Chairman, and Mr. Kenneth Dostal of the Pacific Northwest Water Laboratory.
There are many other names which could be added to this list. All have
been unstinting in their efforts to assist in the design and performance
of research and development programs which are deemed important to the
food processing industry both now as well as in the future.
Over and above the high regard and respect we have for the knowledge and
efficient manner in which these men handle their responsibilities, we
appreciate the friendly and personal interest they have shown in the
problems and the welfare of the food industry as a whole.
FOOD INDUSTRY CONCERNS AND COMMITMENTS
The membership of the National Canners Association shares in the national
concern for the quality of the physical environment. The basic ingre-
dients for the industry are wholesome raw foods and adequate supplies of
clean water. We believe that environmental control is a complex and
continuing process which requires constant searching for improved solu-
tions to the problems which arise from waste materials generated in food
growing and processing. We believe that standards to assure the quality
of the environment, exclusive of those measures necessary to protect public
health, must be achieved in an orderly and logical manner. Precipitous
and unnecessary actions should not be initiated. These could disrupt the
economy of the Nation and interfere with our industry's first responsi-
bility to grow, process, and distribute wholesome and nutritious foods
for the consumers of the world.
Perhaps the best way to capsule the food industry's viewpoint is to read
to you a formal resolution adopted by the National Canners Association
at its Sixty-Fourth Annual Convention, held a few weeks ago.
-------
"The canning industry continues to recognize its respon-
sibility for the development of lawful solutions to pol-
lution problems, and that a closely coordinated industry
effort, including continuing research at both the labor-
atory and plant level, is the most efficient and economical
means of developing effective pollution control. The inter-
dependence of farmers, processors, workers, and the com-
munities in which they reside — the magnitude, seasonal
nature, and interstate character of the food industry's
pollution problems — and the overriding interest of
the consuming public in a wholesome, abundant, and econ-
omical food supply — make it imperative that Federal,
State and local authorities not only establish realistic
guidelines and objectives, but that they also participate
actively and substantially in the development of solutions
to these problems. In the necessary overall effort to
promote recognition by the public and by industry and
governmental leaders that effective pollution abatement
programs must proceed in an orderly, logical, progressive
manner that will afford adequate time for the develop-
ment and implementation of technology, any additional
needed research, pilot-scale testing, treatment plant
construction and economic adjustment, the industry en-
dorses cooperative efforts to establish standards that
are both practicable in each specific problem area,
and which avoid blanket requirements that may in their
impact be unrealistic, unnecessary, or unreasonable. The
canning industry endorses the establishment of the Federal
Environmental Protection Agency as a forward step and
pledges its cooperation with that agency. Necessarily,
the producers and processors of our food supply must be
allowed to continue to expand the production of food for
a hungry world, while at the same time adjusting to the
demands of protecting our environment."
This Symposium, in addition to dealing with the research and technical
aspects of specific pollution control problems, and their resolution,
affords us the opportunity to present to the public in general, and in
particular, to the Environmental Protection Agency (EPA) and its ad-
ministrators, matters of concern in executing federal programs both in
the immediate future and long range. Therefore, I will outline our
major concerns in the hope that this will create a platform for present
and future discussion of the issues. We believe that only in a con-
tinuing dialogue can we each profit from the concerns and convictions
of the other.
QUESTIONS ON WATER QUALITY STANDARDS
The Federal Water Quality Act of 1965 recognized that different rivers
and bodies of water have different uses and problems and that pollution
control efforts should be tailored to meet water quality standards and
beneficial uses prescribed for each of these waters. Most of the states
-------
have standards which are based on assigned beneficial uses to each stream
or section of streams. In these cases, all control efforts work back from
these water quality standards.
We understand that a change in pollution control philosophy is imminent
in that national effluent standards will be established through federal
legislation or regulation. We are concerned that rejuvenation of the
Refuse Act of 1899 is a step in that direction and that this could be
contrary to a sound, logical, and progressive approach to resolution of
this pollution problem. We hope that some of my following remarks will
clarify this position.
We would have to agree that there are probably cases in which water
quality standards alone are not enough. Our primary concern with any
future effluent regulations that may set quality criteria for liquid
wastes being discharged from industrial sources, is that the costs of
compliance be weighed against the benefits obtained. Where the economic
impact will unreasonably distort the local economy, will it be possible
to regard water quality as the basic objective rather than to adopt limi-
tations based solely on discharge characteristics and loads? Will the
economic impact be taken into consideration in forming regulations?
Will each case or situation be submitted to cost-benefit analyses?
We heartily agree that there should be no discharge in any stream of
the country, which would cause harm to human health or produce an un-
desirable irreversible effect on environmental quality. In such cases,
cost should not be a factor for consideration. Will the program that
is being evolved for pollution abatement be able to keep, as its primary
objective, the maximizing of net benefits to society?
For example, wastes from the preparation of foods are not considered
toxic in the sense that their ingestion would be detrimental to humans
or animal life. Will the yardstick used to measure the impact on the
environment of the specific quality or characteristics of each effluent
also recognize these differences?
We believe that M. A. Bernarde, in his Our Precious Habitat, has sug-
gested a concept which is realistic and which must prevail, for a time,
if the Nation's economy is to remain viable:
"The present state of development of our technological
society makes it mandatory that we accept a certain
degree of pollution. We are still far from developing
new methods of liquid-waste disposal that will make dump-
ing in rivers obsolete. The fact is that industry and
residential communities alike would be forced to shut
down if unified public opinion demanded pollution-
free waterways. It is, therefore, in the best interests
of the community to support research investigations
aimed at discovering the degree of waste a stream can
adequately tolerate without the initiation of noisome
-------
(malodorous) conditions. Decisions as to what consti-
tutes a tolerable degree of pollution must be reached
after considering the water's natural purification
capacity and the purpose for which it is to be used."
We are concerned that demands are being made for removal of higher and
higher levels of organic pollutants, and that such demands may not be
based on scientifically determined needs for higher-level treatment but
rather on political or other expedient considerations. Can and will
evaluations be made of the cost-benefit ratio between the dispropor-
tionate energy input required to obtain these higher and higher levels
of treatment of organic pollutants and the total impact on the environ-
ment resulting from creation of these increasing energy requirements?
We feel that it is extremely important that priorities be established
within the area of pollution abatement itself, just as the Nation as a
whole must establish priorities between pollution abatement and competing
goals and objectives. In emergencies, and in the case of harm to human
health, there is, of course, no alternative to immediate abatement action.
But in all other cases, priorities must be established. We believe the
result must be a carefully planned and well-integrated approach for both
immediate and long range needs. Recognition should be given to the
fact that abatement of pollution is not an all or nothing proposition.
No longer should we have a program of frequently-changing objectives,
but one which moves with dispatch upon pressing needs and establishes
a well coordinated, long-range program that will create the least
amount of economic upheaval.
ECONOMICS AND ENVIRONMENTAL QUALITY
The Economic Environmental Policy Advisory Committee has submitted to
the Honorable Maurice H. Stans, Secretary of the Department of Commerce,
a report entitled Economic Policy Issues and Environmental Deterioration.
We feel this document well describes our various concerns and we would
encourage, in the execution or enforcement of abatement programs, both
immediate and in the future, that these considerations will be applied.
Use of these criteria and guidelines, by those directly responsible for
execution of all programs would allow the most efficient and equitable
solution to the pollution problem. Although it may be desirable from
the standpoint of enforcement, to seek a simple overall solution, it is
apparent that to avoid wasting manpower and money, abatement efforts
must utilize these criteria and guidelines. Otherwise abatement efforts
cannot be kept on an economically viable course. We would further en-
courage incorporation as regulations, for those making judgments on
specific field installations, the six principles of paramount importance
in the report of the Economic Environmental Policy Advisory Committee.
The principles I refer to are as follows:
"First, like other national objectives, pollution abatement involves
a marshalling of economic resources that are limited. This requires a
choice or allocation of resources between competing objectives — in
other words, a recognition of sensible priorities and a recognition of
-------
trade offs. These priorities and trade offs will be influenced sig-
nificantly by the time span during which the problem is treated.
Second , the overriding criterion to be used both in determining the
proportion of national resources to be directed toward pollution abate-
ment and in determining how costs should be distributed between the
public and private sectors is to seek the path or methods leading to
the maximum net social benefit (or conversely to the lowest net social
cost).
Third, in determining the lowest cost solution, pollution abatement
objectives must be subjected to comprehensive yet penetrating cost-
benefit analyses. In such analyses both costs and benefits must be
viewed in their broadest framework. One of the striking aspects of
the pollution problem and a serious limitation, however, is the dearth
of adequate factual information. Clearly, the nation has an obligation
to move as expeditiously as possible toward the development of a body
of information relating to the pollution problems that would enable
policy makers and responsible government officials to make cost-benefit
analyses in an intelligent manner.
Fourth, major reliance should be placed on the competitive market system
and its demonstrated capacity as an efficient allocator of resources
in seeking the lowest cost solution to pollution. Every effort should
be made to tap and implement the ingenuity, initiative, and incentives
that are an intrinsic part of the system.
Fifth, departures from the principle of major reliance on the competi-
tive market system may be dictated by three factors: (1) crises sit-
uations ; (2) research and development needs; and (3) extreme hardship
cases.
Sixth, recognition of the above principles leads naturally to certain
criteria that can help evaluate proposed methods of financing pollution
abatement and also the relative emphasis that should be placed on the
private and public sectors. In the realization of these principles, the
fundamental economic criteria that has guided the development of this
report is that the costs of pollution abatement should generally be
borne by the private producer/consumer sector. Ultimately, the prices
of products and services flowing into the economic stream must reflect
these full costs."
CONCERNS ABOUT LONG-RANGE RESEARCH
It is apparent that present problems and conditions will not be totally
eliminated because of the limitations of existing technology and the
limitations inflicted by economic pressures. There is a need for outgoing
research geared for long-range study programs. We would like a better
and clearer understanding of the EPA's policy and funding for both short-
term and long-range research. We have a deep concern that assignment of
regional laboratories to monitoring programs will result in limited and
meager research budgets rather than greatly expanded research programs
8
-------
that must meet both immediate and future needs. It has been most dis-
appointing over the past few years to see an erosion of research funds
because of reduced priorities in this area, and a resultant failure to
expand the regional laboratories to their full potential. This is a
great loss, not only for the present but also for the future decade.
We in the industry recognize that there are numerous technological
barriers to the solution of some of our problems and we strongly urge
increased research effort to find the practical answers. It has been
estimated that on the average approximately ten years are required for
new research to have full and applied commercial realization.
We hope that future research funding for the regional laboratories can
follow the comments of the National Goals Research Staff: "To the ex-
tent that society insists that basic scientists do work that is more
relevant to present social needs — scientists will be less able to
work where nature appears willing to answer their questions. They may
be required to work on relevant questions that perhaps cannot be an-
swered at all at present, or can be answered only with uneconomic use
of resources. Thus excessive efforts to make science more productive
in terms of immediate social goals may actually make it far less pro-
ductive in the long run."
This concept further shows the need to avoid broad-based applications
of restrictions. Rather, each water system and each case should be
evaluated on its own unique conditions. In this connection, we believe
that government should sponsor research which will study our environ-
ment not only from the standpoint of eliminating unnecessary pollutants
but also with regard to learning how to utilize non-toxic materials in
the environment. We, as food processors, feel that the future may find
us dependent on learning to farm our streams, lakes, and oceans. An
acceptable concept, in our opinion, would utilize non-toxic, organic
wastes, such as those from the processing of seafoods, fruits and
vegetables, for enrichment of the marine environment to the end that
a greater abundance of fish and aquatic plants are produced. To what
extent is there planning and funding to expand our knowledge of the
use of these materials in the aquatic environment?
Most of the deep ocean is, in many respects, a vast desert that yielda
very small quantities of food life in comparison to the potential with
suitably provided nutrients. Can we be assured that plans are being
made for research to determine acceptabe ways of discharging food
processing wastes and, in fact, municipal food wastes to the oceans in
order to yield a more abundant fish life?
I have suggested in preceding comments that although elimination of
many compounds presently discharged to our waterways should be ex-
cluded or reduced, there is the possibility that in many places we
can learn how to better manage what are presently considered as ob-
wateiwavt 'OmP°unds> such as phosphates and nitrates, in these
waterways. To this end, we hope that a major stream, having all*the
variations that could be considered typical of the different cross-
-------
sections of life, could be studied in depth to achieve this more
desirable management goal. In such a stream basin, would it be
possible for the federal government to subsidize, immediately, the
cost for the best technology available for elmination of pollutants
in the effluent? This would permit us to learn whether, indeed, the
programs being projected for the entire country will net all the
benefits we anticipate. The results could show that a greater effort
should be made to perfect utilization of the various residuals dis-
charged to streams by adding to the stream other materials to achieve
the "fish hatchery" approach.
Previously, I briefly mentioned aquatic plants. We believe that a
broad field of study should be instituted on the use of aquatic plants
to remove undesirable materials and nutrients from streams and lakes
by the eventual harvest of these plants as a potential food supply,
In too many areas, marshes and swamps have been drained within which
normal growth of aquatic plants was performing a very valuable service.
We believe that it is time to give consideration to re-establishing
such conditions in rivers and lakes. Where this could be established
under proper management, it may be possible to gain even more benefit
than came from the original marshes. Consideration could be given
to harvesting the aquatic growths, thus short-cutting nature's cycle
of normally returning the nutrients to the environment when the
plants die. Under proper water management, a crop that is beneficial
to society may be harvested. We hope that fertile areas, such as
these, will receive attention and funds for research in the EPA re-
gional laboratories and in universities. If so, then the next decade
will be able to benefit from today's progressive thinking on managing
our waterways.
OTHER INDUSTRY CONCERNS
It has been recognized that, basically, pollution is a people-problem.
Government spokesmen have assured us that industry is not to be made
the whipping boy in pollution abatement programs. Yet, we understand
that the Federal government has said that no industrial representatives
and only conservationists should be placed on water quality and water
resources boards. We would be greatly disturbed if recognition is
not given to the fact that one person may be both an industrialist
and a conservationist. Surely, it can be recognized that one who
totally ignores conservation, or industry's welfare, or any other
factor, is equally unacceptable.
As individuals and as an industry, we urge abandonment of the "crash"
psychology that seems to believe that all social ills can be cured by
massive doses of "instant money" and that massive government spending
will produce an instant pollution-free tomorrow. How can we obtain
a well-informed people who, given an understanding of the total picture,
will, in turn, prevent improper legislation arising from superficial
knowledge, lack of study, and subjective irrational thinking? We would
like recommendations on improving communications between government
and industry, and, in particular, to the general lay public. As an
10
-------
industry and as individual citizens what should our role be in as-
suring that all decisions are based on facts — not conjectures,
arbitrary opinions, or emotions? How do we avoid confrontation and
still maintain economically-viable programs that do not cause major
industry and employment dislocations?
The food processing industry is concerned with the frequently-
changing and unpredictable nature and degree of state and federal
environmental control programs. Decision making by the industry,
leading to compliance with present and probable pollution abatement
requirements, is made difficult on both a regional and national basis
by these major uncertainties. In preparing information for the
National Industrial Pollution Control Council, we have listed eight
environmental matters of general concern to the food processing
industry and would appreciate EPA's candid comments on each of these
issues.
A. Establishment of standards which take into consid-
eration the nature and treatability of each type
of waste and which are flexible depending on the
degree of treatment needed, in each river-basin
or regional area.
B. Establishment of a relatively uniform system of
enforcement to avoid gross inequities which could
affect different members of any given segment of
industry.
C. State and Federal regulations that are changing so
rapidly that we are never sure we are complying
with latest requirements in all phases of pollution
control. It appears that the criteria established
by different states vary and this presents prob-
lems in establishing and projecting budget re-
quirements for plans to satisfy all codes.
D. The need for cooperative industry-government ef-
forts to organize and disseminate to the public,
industry, and all levels of government, an orderly,
logical, and progressive schedule for pollution
abatement. The schedule should include a time-
table with goals for achieving an improved en-
vironment. The schedule must also provide time
for needed research, for trial and error testing,
for construction of the treatment plant or other
abatement facility, and for economic adjustment.
E. The need for public affairs programs to adequately
and objectively inform all concerned about the
realities of environmental quality control and
the total consequences of extreme measures which -may
not be necessary now, if at all.
11
-------
F, The related need for a general understanding of
the total pollution problem, which involves the
whole population in all of its activities and
which cannot be eliminated simply by imposing
strict controls on industry,
G. The non-productive costs of managing waste^
waters and solid wastes may, as a result of sudden
changes in the economic marketplace, cause an
adverse tipping of the balance against older or
obsolescent plants. As an example, the Depart-
ment of the Interior regulation, concerning grants
for treatment plant construction*, can arbitrarily,
without taking into consideration other social and
economic benefits, cause unnecessary economic hard-
ship.
H. Many food processing plants presently utilize
municipal treatment facilities to handle waste
discharges from the preparation of these foods.
Due to possible increasing pressure to upgrade
the effluent quality, industry may be refused this
service and may find it economically or techno-
logically difficult to continue to be a generator
of economic benefits within the community.
In conclusion, we cannot over-emphasize the excellent cooperation
existing among the various branches of the government dealing with
research, and the Association and its membership. This communication
is invaluable from many standpoints and emphasizes the need for this
same approach all across industry. It should be amplified and im-
proved wherever possible in all industries. We would appreciate com-
ments as to how this relationship could be expanded and enhanced for
our own area. Would it be appropriate for industry and government re-
presentatives to meet and explore developments in this area?
We are particularly interested in realizing this same relationship exist-
ing with our government groups that are involved in the environmental
area. We recall that the National Technical Task Committee on Indus-
trial Wastes was an important adjunct to industry and government and
that a great deal was lost upon its discontinuation. Public clamor
raises a furor on single issues. We, industry, and government, must
be dedicated to an on-going program. The greater the inter-change of
information and communication, the better will be the possibility of
realistic and economic programs for improving the environment.
*Federal Register, Vol. 35, No. 128, pages 10756 - 10757, July 2nd, 1970.
12
-------
EPA'S INDUSTRIAL POLLUTION CONTROL PROGRAM
by
Dr. Stanley M. Greenfield*
President Nixon has said "We are determined that the decade of the 70's
will be known as the time when this country regained a productive har-
mony between man and nature." Our problems with the environment have
become critical and we must act now if we are to preserve this harmon-
ious relationship with nature. We have already taken major significant
steps to reverse our degradation of the environment. I am most pleased
to report to you today on the progress made by the Government to date on
this endeavor and to discuss our needs and plans for the future.
Our problems with pollution are not new. Indeed, man has abused the
environment as long as mankind has been on this planet but in recent
years population growth has exceeded the capacity of our air and water
to absorb the municipal and industrial pollutants and we can no longer
turn our backs on the problems we have created. In the past we have
not fully appreciated what we are doing to our air and water resources.
We have ignored the fact that these are finite and limited resources,
although, our demands upon them have increased at alarming rates. Con-
sider that since 1900 the population of the United States has doubled
but in this same period of time the needs of industry for water have
increased ten times. It is estimated that the next doubling of the world
population will occur in 35 years rather than 70 and we can anticipate
similar industrial growth. It is clear that we must take immediate ac-
tion to restore the quality of environment we expect and demand.
We have made progress. New technology has been developed for treatment
of air pollutants and liquid wastes but some times the treatment of one
has been to the detriment of the other. Further, our population and
industrial growth have outstripped our technology in many cases and
in spite of the installation of required treatment facilities, we have
found a decline in stream quality in many parts of the United States.
Recognizing the need for a better directed effort to control all forms
of pollution, President Nixon last year through reorganization created
the Environmental Protection Agency. This was done to consolidate all
environmental pollution control activities of the Federal Government into
a single agency. EPA consists of the water pollution control activities
formerly in the Department of the Interior, the air pollution control
activities and solid waste function of the Department of Health, Education,
& Welfare, and pesticides and radiation control activities formerly in
a number of different agencies.
^Assistant Administrator for Research & Monitoring, U. S. Environmental
Protection Agency, Washington, D. C.
13
-------
Let me speak for a moment on the reasons for the reorganization to form
EPA. In the past the organization of the Federal Government to deal with
pollution has suffered from two obvious problems. First, for many par-
ticular kinds of pollution, a number of different Federal agencies have
had overlapping or closely related responsibilities. For example, three
Federal Departments, Agriculture, HEW, and Interior were directly involved
in regulating pesticides. And similarly, a number of agencies have some
responsibilities for radiation problems. Second, the organizational basis
for controlling pollution was not consistent or adequate. The two largest
agencies, the Federal Water Quality Administration and the National Air
Pollution Control Administration, were organized on the basis of the media-
air or water—through which the pollutants traveled. The other pollution
control programs on the other hand generally were organized on the basis
of particular pollutants, pesticides, radioactive materials, and solid
wastes. Confusion has resulted about the extent to which air and water
pollution control agencies were responsible for radioactive materials
and pesticides when these materials appear in air or water.
The programs to deal with pesticides and radiation were developed in
part because these two kinds of pollutants did not fit neatly into the
categories of air and water pollution. Pesticides and radiation are
found in both air and water and on the land. We can expect that pol-
lution control problems of the future will be increasingly of this kind.
They will involve toxic chemicals and metals which are found in all media
and which run counter to the previous type of organization of air and
water pollution control in the government.
Some pollution problems remain unrecognized because of gaps in agency
jurisdiction or because no one agency had clear lead responsibility. The
Environmental Protection Agency will overcome this handicap because of
its broad responsibility for environmental pollution control.
Another problem of past Federal organization has concerned the fact that
in some cases agencies which have responsibility for promoting a par-
ticular resource or activity also have responsibility for regulating the
environmental effects of this activity. Two clear examples of this
potential conflict of interest were agriculture's regulation of pesticides
and the atomic energy's regulation of radiation levels. Regardless of
how good a job these agencies did, the public was increasingly question-
ing the assignment of promotional and regulatory powers in the same agency.
The Environmental Protection Agency has been given these regulatory func-
tions and this should restore public confidence in our ability to control
pollution from these sources.
The existence of a unified pollution control agency will clarify the
Federal Government's relations with state and local governments and with
private industry. More than half the states already have a single agency
responsible for all forms of pollution. A number of others are consider-
ing establishing such an agency.
The reorganization was affected by an Executive Order of the President
creating the Environmental Protection Agency. Under the law, such a
14
-------
reorganization plan cannot create any new legal authorities or functions,
therefore, EPA simply functions under the same legislation which origin-
ally created its constituent parts, however, this new agency will be
able to perform the existing functions better and as new legislation is
provided will be able tp undertake new activities not easily done under
the previous structure.
The key functions in pollution control are standards setting and enforce-
ment. Standards provide the goals of the control program, the basis for
enforcement actions, and the measure of the program's progress. Standards
should be based upon the total amount of a given pollutant to which humans
or some element of the environment are exposed even though the standards
apply to a particular medium. Lead, for example, may reach humans through
the air or the water but the key question is how much comes from all
sources together. The organization of EPA will permit standard setting
for pollutants which cut across media lines.
The enforcement function will be improved in several respects. The new
agency will be able to examine the path of a pollutant through the total
environment and determine at what point control measures can be most
effectively and efficiently applied. For example, it may be that in some
cases a pollutant can best be controlled by limiting its entry to the
environment as has been done with pesticides.
Research will be similarly strengthened. Research on the health effects
of pollution will be able to take into account the exposure to a given
pollutant from all sources. Research on ecological effects must con-
sider the interrelated parts of the environment and the impact of man's
activities. It will be far easier to conduct ecological studies in an
agency which is not limited to one particular medium or pollutant.
Likewise, waste treatment research may now consider integrated systems
to control air and water pollution and the ultimate disposal or recycling
of solid wastes.
Briefly, these are some of the reasons for the President's formation of
EPA and its more efficient ways of functioning as a unified agency. The
creation of EPA represents a major step forward in streamlining the
Federal Government's activities in pollution control. It cannot be the
ultimate step, however, in a dynamic society such as ours. Much remains
to be done.
In his message on the environment last month the President asked Congress
to take additional steps to further strengthen our pollution control
activities. He has requested a 1972 budget of $2.45 billion for the pro-
grams of the Environmental Protection Agency—nearly double the funds
appropriated for these programs in 1971. The funds will provide for the
expansion of air and water pollution, solid waste, radiation, and pest-
icide control programs, and for initiating new activities in these programs,
Last year the President requested new legislation to broaden the water
quality standards requirements but Congress did not act on this request.
The Government did move ahead, however, to make use of existing authorities
15
-------
through the Refuse Act of 1899 to require discharge permits as announced
in December. The President has again asked this year for reforms in the
water quality laws regarding effluent standards which would be administered
by EPA. Water quality standards now are sometimes imprecise and unrelated
to specific water quality needs and these do not provide a good basis for
enforcement. The President has proposed that the Federal-State Water
Quality Program be extended to cover all navigable waters and their trib-
utaries, ground waters, and waters of the contiguous zone. He has re-
quested that the standards be revised to impose precise effluent limita-
tions on both industrial and municipal sources. He has asked for standards
to regulate the discharge of hazardous substances, similar to those included
in the Clean Air Amendments of 1970. The President has asked for legisla-
tion on standards to require that the best practicable technology be used
in new industrial facilities to insure that water quality is preserved
or enhanced.
In the matter of pesticides, the President has asked for new legislation
which would provide for a registration procedure designating pesticides
for general use, restricted use, or use by permit only. Pesticides de-
signated for restricted use would be applied only by an approved pest
control applicator. Pesticides designated for use by permit only would
be made available only with the approval of an approved pest control
consultant. This would help to insure that pesticides which are safe
when properly used will not be misused or applied in excessive quantities.
The requested legislation would authorize the administrator of EPA to
permit the experimental use of pesticides under strict controls when he
needs additional information concerning a pesticide before deciding whether
it should be registered. This requested legislation would also authorize
the administrator to stop the sale or use of and to seize pesticides be-
ing distributed or held in violation of Federal law.
The President has also asked for new legislation concerning toxic sub-
stances. As we have become increasingly dependent on many chemicals and
metals, we have become acutely aware of the potential toxicity of the
materials entering our environment. Each year hundreds of new chemicals
are commercially marketed and some of these may pose serious potential
threats. Many existing chemicals and metals, such as pcB (polychlorinated
biphenyls) and mercury also represent a hazard. We need better methods
to assure adequate testing of chemicals to avoid environmental crises of
the future. The President has proposed to Congress that the administrator
of EPA be impowered to restrict the use or distribution of any substance
which he finds is a hazard to human health or the environment. The
President has requested legislation to authorize the administrator to
prescribe minimum standard tests to be performed on these substances.
One of the other areas discussed in the President's Message on the
Environment last month included the problem of ocean dumping. He recom-
mended a national policy banning unregulated ocean dumping of all mater-
ials and placing strict limits on ocean disposal of any materials harmful
to the environment. He recommended legislation to require a permit from
EPA for any materials to be dumped into the ocean, estuaries, or great
lakes and to authorize the administrator to ban dumping of wastes which
16
-------
are dangerous to the marine ecosystem. I believe the key words here are
"unregulated ocean dumping," and "wastes which are dangerous." It has
been suggested that much of the ocean may be a biological desert and
some fertilization through the discharge of controlled amounts of wastes
might increase the productivities of the marine environment. There is
an obvious need for research to define the optimum levels of waste dis-
charged to enhance the productivity. Obviously, waste overloads will
degrade the water quality.
These proposals requested by the President in his Message on the Environ-
ment delivered to Congress last month have not been enacted into law
but I believe they reflect the mood of Congress and the American public
for stricter control of pollution in the environment.
Mr. Ruckelshaus, the Administrator, EPA, has said that we now have all
the tools to do the job. We have the organization in the form of EPA
and we have the three essentials of support. We have the support of
the Administration, the support of the Congress, and the support of
the American Public. We have begun a concerted fight to restore the
quality of the environment.
One of our primary goals is to achieve installation of adequate treat-
ment facilities for all municipal and industrial sources of waste within
the next five years. To accomplish this will require the coordinated
efforts of all aspects of society.
The Federal Research and Development Program is an essential key to the
accomplishment of these goals.
The goal of any research program is to provide through technical investi-
gation the criteria upon which decisions can be based to solve the prob-
lem. In the case of industrial waste treatment research, these criteria
may be design criteria for waste treatment facilities. Note that I did
not indicate that the research need is to set treatment standards since
this may involve social and economic considerations as well as techno-
logical requirements developed by us as researchers. It is our job, as
I see it, to provide the best possible technical input upon which sound
decisions can be based in the standard setting process.
President Nixon, in his message to Congress last month on a program for
a better environment, said, "We have the technology now to deal with
most forms of water pollution. We must make sure that it is used."
This does not mean that our job is done, but rather that a solution is
available. Obviously, Henry Ford's Model T provided a solution to the
transportation problem but fortunately development of the automobile did
not stop there.
The challenge we face is to develop tertiary or advanced waste treatment
methods which will permit recycling and closing of the loop in water
using industries, the development of in-plant processing changes to
reduce waste discharge, and the development of processes to permit by-
product recovery and reduce treatment costs. The food processing in-
dustry is an excellent example of how we intend to pursue a systematic
research program. National responsibility for research on waste treatment
17
-------
in your industry has been assigned to the Waste Treatment Research Program
in the Pacific Northwest Water Laboratory in Corvallis, Oregon. This
laboratory is responsible for planning and conducting a coordinated re-
search program utilizing the talent and resources of the Federal Govern-
ment, universities, and the industry through in-house studies and extramural
grant and contract programs.
The first step has been to obtain a definition of problems and research
needs for the many varied segments of the food processing industry. This
step is well under way. The University of Washington in Seattle was given
an $18,400 grant to conduct a study on the status and research needs for
potato waste waters. This project has been completed and a final report
written. Other state-of-the-art studies nearing completion include sea-
foods, fruits and vegetables, dairies, sugarbeets, and beverages.
All of these studies cited were undertaken through the extramural grants
program to define current practices in the treatment of wastes from
various segments of the food processing industry and to pinpoint most
urgent needs for additional research. With this "handle" on the problems,
we can now direct available resources to those areas which most urgently
need immediate solution. We have undertaken bench-scale studies with
our own scientists and engineers and have funded similar research by
universities and industry. Our research grants program has supported
studies by university researchers on analytical methods, treatment pro-
cesses, and the effects of pollutants on the receiving waters.
In 1966, Congress for the first time authorized grants directly to in-
dustry for the demonstration of new and improved techniques for the treat-
ment of industrial wastes. This program from the outset has received
enthusiastic response from your industry and many companies have part-
icipated in this cooperative Federal-Industry program. Many of these
projects will be reported on during this conference. Let me just say
that there are more than 50 currently active projects under our research
and demonstration grant program with a total Federal grant involvement of
over $10 million in the food processing industry alone. Some of these
involve demonstration grants of several hundred thousand dollars to test
at full-plant scale new and improved systems for the treatment of wastes.
This program will be continued but in order to make broader use of the
funds available, more emphasis will be placed on the technical support
of these projects, such as plant design, analytical studies, operation,
and final report write-up. Participation in the construction of massive
treatment facilities will be less likely to receive full grant support.
It has been our policy to support projects aimed at the development of in-
plant changes to reduce waste loads and flow from the plant. Our support
of the development of the dry caustic peeling of potatoes which can re-
duce the waste load by 50%, is an example of this.
We do not feel that conventional secondary treatment is the ultimate an-
swer for water pollution control by industry. I fully expect that many
18
-------
segments of industry will ultimately have complete treatment systems
with recovery of essentially all dissolved and suspended solids.from
waste streams permitting total recycle of water within the plant and
with the sale or reuse of the recovered by-products.
Much is already being done to study methods for the recovery of by-
products and to develop markets for them. Much of the solid waste
from the food processing industry can be used for animal feed and
we are confident that imaginative research will develop even more
productive recoveries.
I have been speaking primarily about the water pollution control re-
search program since this activity has received the major support
emphasis in the past. However, the establishment of EPA now permits
us to consolidate our research efforts within one program in all of the
media affected by pollutants from industry. That is air, water, and
land. Our reaearch effort can and will be extended to include methods
to remove odors and particulates from air streams. Our research will
extend to the ultimate disposal of solids separated from air and
liquid streams so that we are assured that solution of one pollution
problem will not create a second.
In line with our broadened research program to include air pollution
and solid waste research, we intend to broaden our own research talent
to include engineers and scientists familiar with industrial processes
and others such as economists to permit us to truly look at the total
picture of waste generation and treatment and control from an efficiency
standpoint. I want us to be able to actively study and support changes
in processes to eliminate waste at its source.
The development of new information is useless unless this is applied.
This has been called technology transfer. It is our intent to make
certain that new information developed through our research programs,
particularly in the industrial pollution control field, is made
available to you as rapidly as possible. We will do this not only
through publication of reports but in symposia such as this and through
direct technical assistance to the industry. We expect to work directly
with you in the study of waste treatment problems and to make available
through pilot-scale and full-scale demonstrations the latest techniques
applicable to your industry.
The President has asked for a doubling of the budget for EPA this year
and this level of increase will apply as well to the research and de-
velopment program. Much work remains to be done if we are to find an-
swers to our industrial pollution control problems that will be both
effective and feasible but with your continued support and cooperation
we will achieve our goals.
19
-------
THE CLEAN AIR AMENDMENTS OF 1970
and
AIR POLLUTION ASPECTS OF THE FOOD
AND AGRICULTURAL PROCESSING INDUSTRY
by
Dr. S. David Shearer*
INTRODUCTION
It is a real pleasure for me to be with you at your Second National
Symposium. The three speakers preceding me have focused quite con-
cisely upon the Environmental Concerns of the Food Processing Industry
and upon the direction and character of the Industrial Pollution Con-
trol Program and Regional Activities of the Environmental Protection
Agency.
I would like to bring to you this morning a brief review of the sig-
nificant changes and amendments to the Clean Air Act which was signed
into law by the President on December 31, 1970. Time does not permit
me to cover in detail all of the new amendments but only those which
would be of most direct interest to this group. In addition I will
present to you some general qualitative information on the type and
amount of air pollutants from certain type s of food processing oper-
ations. Only the highlights of the latter will be covered here;
however, the symposium proceedings will contain fuller details of my
brief remarks.
CLEAN AIR AMENDMENTS OF 1970
The 1970 Amendments to the Clean Air Act encompassed a wide range of far
reaching improvements designed to accelerate the restoration of our
polluted air and to better preserve those portions of our country which
are still blessed with acceptable air quality. These amendments pro-
vided for changes to the 1967 Clean Air Act as well as entirely new
legislation in several areas. I have listed below most of the salient
aspects of the 1970 Amendments:
1. Emphasizes, as the 1967 Act did, that the States will continue to
have prime responsibility for control of Air Pollution including pro-
visions to allow them to make the political and social decisions nec-
essary to accomplish the requirements of the new amendments.
*Air Pollution Control Office»Environmental Protection Agency, Durham,
North Carolina.
21
-------
2. Increased financial and technical assistance to State and Local
Air Pollution Control Agencies.
3. Increased monies for research and development directed toward
improved and new pollution control techniques and processes.
4. Requirements for the establishment of national ambient air quality
standards to be achieved in a definite time period. On January 30,
1971 the EPA published in the Federal Register proposed primary and
secondary air quality standards for particulate matter, sulfur oxides,
carbon monoxide, hydrocarbons, photochemical oxidants, and nitrogen
dioxide.
5. Increased emphasis on comprehensive State-wide implementation plans
to achieve the national standards. Such plans must include land-use
and transportation control policies in addition to regulation of fuel
storage and handling if necessary to achieve the standards within
the prescribed time period. In addition, public hearings must be held
before adoption of these State-wide plans.
6. Provides for the establishment of federal performance standards for
new stationary sources of air pollution which reflect the use of the
best system of emission reduction that has been adequately demonstrated.
7. Calls for the establishment of federal emission limitations for
hazardous pollutants that may cause or contribute to an increase in
serious irreversible or incapacitating illness.
8. Makes provisions for State and/or Federal officials to be able to
enter industrial establishments for the purpose of copying records,
making inspection of air pollution control equipment and to sample emis-
sions.
9. Provides for several extensions and new provisions for control of
pollution from new motor vehicles including assembly line testing in
addition to provisions enabling states to require inspection and testing
of motor vehicles in the hands of the general public if necessary as
part of the state implementation plan to achieve national standards.
Id Allows the Administrator, EPA, to regulate, control, or prohibit
the manufacture or sale of fuel or fuel additives if emissions endanger
public health or welfare or impair the performance of motor vehicle or
aircraft emission control devices.
11. Provides increased monies for the development of advanced automotive
power systems directed toward the production of a pollution free pass-
enger car.
12. For the first time the new amendments will require the federal
government to establish emission standards applicable to emissions of
any air pollutant from any class or classes of aircraft or aircraft
engines which cause or contribute to air pollution which endanger health
welfare.
22
-------
13. Requires the federal government to not enter into a contract for
procurement of goods, services, or materials with any industrial, concern
that is in violation of any part of a State-wide implementation plan to
achieve air quality standards, new source performance standards, or
national emission standards for hazardous air pollutants.
14. A final major new addition in the 1970 amendments is a provision
whereby any citizen may bring suit in a Federal District Court against
any industrial concern in violation of requirements of the Act or against
State, Local, or Federal Agencies for failure to carry out the mandates
of the Act.
I have included in the material which will be printed in the proceedings
of this symposium flow charts and time schedules for compliance for five
of the key sections of the Clean Air Amendments of 1970. In addition,
I have prepared a table of the proposed national air quality standards
for the six pollutants mentioned in item 4 above.
Thus it can be clearly seen from the above listed items that with respect
to air pollution control there is a firm mandate from the Congress that
the States, Federal Government and Industry must do more and work harder
to enhance and preserve our precious air resources.
PROPOSED NATIONAL AMBIENT AIR QUALITY STANDARDS
A. Sulfur Oxides (Primary*)
80 jug/m3
350 jug/m3
Annual arithmetic mean
Max. 24-hr, cone, (not to be exceeded
more than once/year)
Sulfur Oxides (Secondary**)
60 jag/m3
260
Annual arithmetic mean
Max. 24-hr, cone, (not to be exceeded
more than once/year)
B. Particulate Matter (Primary)
75 jug/m3
260 jig/nT
Annual geometric mean
Max. 24-hr, cone, (not to be exceeded
more than once/year)
Particulate Matter (Secondary)
60 jig/m;
150 jig/nr
Annual geometric mean
Max. 24-hr, cone, (not to be exceeded
more than once/year)
C. Carbon Monoxide (Primary)
10 mg/m3
Max. 8-hr, (not to be exceeded more
than once/year)
23
-------
15 mg/m Max. 1-hr, (not to be exceeded more
than once/year)
Carbon Monoxide (Secondary)
o
10 mg/m Max. 8-hr, (not to be exceeded more
than once/year)
15 mg/m Max. 1-hr, (not to be exceeded more
than once/year)
D. Photochemical Oxidants (Primary)
125 jjg/m Max. 1-hr, (not to be exceeded more
than once/year)
Photochemical Oxidants (Secondary)
125 jig/m Max. 1-hr, (not to be exceeded more
than once/year)
E. Hydrocarbons (Primary)
2
125 pg/m Max. 3-hr. (6 a.m. - 9 a.m.) (not to be
exceeded more than once/year)
Hydrocarbons (Secondary)
125 jig/m3 Max. 3-hr. (6 a.m. - 9 a.m.) (not to be
exceeded more than once/year)
F. Nitrogen Dioxide (Primary)
3
100 /ig/m Annual arithmetic mean
250 pg/m 24-hr, cone, (not to be exceeded more
than once/year)
Nitrogen Dioxide (Secondary)
o
100 jig/m Annual arithmetic mean
250 jag/m 24-hr, cone, (not to be exceeded more
than once/year)
AIR POLLUTION ASPECTS OF THE FOOD AND AGRICULTURAL PROCESSING INDUSTRY
The processing of food and agricultural products for consumer uses in-
volves a number of various processing steps such as collection, refining,
^Primary Standards are based on health considering a margin of safety.
**Secondary Standards are based on public welfare.
24
-------
Attachment B
SECTION 109 - TIME SCHEDULE FOR NATIONAL AMBIENT AIR QUALITY STANDARDS
(CLEAN AIR AMENDMENTS OF 1970 - PL 91-604)
Enactment
of
Legislation
N>
12/31/70
30
Days
Admr. Publish Prop.
Reg. on Nat. Primary
A. Q. Stda. and Nat.
Secondary A. Q. Stds.
for Pollutants for
which Criteria have been
issued prior to
Enactment.
>, Days
1/30/71
Any
Time
Criteria Documents on
Other Pollutants Issued
From Time to Time Along
with Proposed A. Q. Stds.
90
Admr. Promulgates both
Types of Stds. After
Considering Written
Comments.
4/30/71
NOTE: Primary Stds. Based on Health
Considering Margin of Safety.
Secondary Stds.
Public Welfare.
Based on
-------
SECTION 110 - 1MB SCHEDULE FOR IMPLEMENTATION PLANS
(CLEAN AIR AMENDMENTS OF 1970 - PL 91-604)
12/31/70
Plans Submitted Prior to Enactment
of Act Hay be Approved by Admr. If
Consistent with This Act. If Not
Consistent Adnr. Must Within 90 Days
Notify State of Inconsistencies and
Will Promulgate'Plan If State Does Not
Correct Inconsistencies within 6 Months.
7/29/71 and 1/31/72
4/30/71
States Adopt After
Public Bearing Plans
for Both Primary and
Secondary Stds. And
Submits to Admr. for
Approval.
1/31/72
Admr. Promptly Publishes Proposed
Regulations for Plan If:
A) State Falls to Submit Plan
When Due
B) Plan Submitted Not In
Accordance with this Section
C) State Does not Revise Prior
Submitted Plan Within 60
Days of Notification In
Accordance with Subsection
(a) (2) (H) .
6 -j
Mos. *
Admr. Promulgates
Regulations Within
180 Days unless
State Meets Items
In Previous Block.
If Plan Approved,
State Enforcement.
5/31/72
DOTE: Primary Stds. Must be Achieved
Within 3 Tsars, Secondary
Beds. By * Reasonable Time
Later. Admr. Kay Extend Date
for Achieving Primary Stds.
for Hot More Than 2 Years.
Admr. May Extend Date for Sub.
of Plan for Achieving Secondary
Standard not more than 18 Months
(See Sec. 110(f)(l) For another
Type of Extension the Admr. May
Authorize for Certain Stationary
or Mobil* Sources.)
-------
SECTION 111 - TIME SCHEDULE FOR STANDARDS OF PERFORMANCE FOR NEW STATIONARY SOURCES
CCLEAN AIR AMENDMENTS OF 1970 - PL 91-604)
Enactment
of 90
Legislation Day
12/31/70
No Time j.
Stated
Admr. Publishes (and
From Time to Time
Revises) List of
s Stationary Source
Categories.
120 ^
Days
Admr. Publishes Proposed
Regulations Establishing
Stds. of Performance for
New Sources on Pubished
List.
90.,
Days^
3/31/71 llitlii
After Considering
Written Commente Admr.
Promulgates Stds. of
Performance. (He May
From Time to Time
Revise Stds.). Stds.
Become Effective Upon
Promulgation .
Admr. Prescribes Regulations
Establishing Schedule Similar
to Sec. 110 Under Which State
Must Sub. to Admr. Plan for
Establishing Emission Stds. (as
Well as Plan for Implementation and
Enforcement of such Emission Stds.)
For
Pollutants (Not Covered by
Criteria Documents) from existing
Sources of same type as New Sources
Covered under this Section. Admr.
Shall Set Such Stds. And Enforce
Plan If State Does Not.
No Time -
Stated
States May Develop and
Submit to Admr. Plan
for Enforcing Per. Stds.
And He Shall Delegate
Enforcement to State if
Their Procedure Adeouate.
If not Admr. Will Enforce Std.
10/27/71
NOTE: Admr. Shall From Time to Time
Issue Control Techniques for
Categories of New Sources and
Pollutants Subject to This Section.
-------
SECTION 112 - TIME SCHEDULE FOR NATIONAL EMISSION STANDARDS FOR HAZARDOUS AIR POLLUTANTS
(CLEAN AIR AMENDMENTS OF 1970 - PL 91-604)
NOTE: States May Submit Plan for Implementing
and Enforcement of Std. and Admr. Shall
Delegate Such If Adequate. If not Admr.
Will Enforce.
Applies to Existing
Sources Only.
Enactment
of
Legislation
12/31/70
90 y
Days
Admr. Publishes (and
From Time to Time
Revises) List of Haz-
ardous Air Pollutants
for Which Emission
Stds. Will be Issued.
180 ^
Days *~
Admr. Publishes Proposed
Regulations Establishing
Nat. Emission Standards
and Notice of Public
Hearing Within 30 Days.
9/27/71
Public Hearing
Must be Held
10/27/71
150
Days
Admr. Prescribes
Emission Stds. Unless
Info. At Public Hear-
Shovs that Pollutant
Is Clearly not Hazardous.
V
90 j
Days "
Standards
Become
Effective
fttVitT)
3/31/71
NOTE: Admr. Shall Issue from
Time to Time Control
Techniques for Air
Pollutants Under this
Section.
3/25/72
NOTE: Admr. Hay Grant Waiver up to 2 Tears if
Necessary for Source to Install Controls
and Source Will Take Steps to Insure
Health of Persons Will be- Protected from
Imminent Endangerment.
-------
SECTION 113 - TIME SCHEDULE FOR FEDERAL ENFORCEMENT
(CLEAN AIR AMENDMENTS OF 1970 - PL 91-604)
Enactment
of
Legislation
No Stated
12/31/70
tsi
vo
Time
No Stated
Time
No Stated
Time
Whenever Admr. Finds that any Person
is Violating any Requirement of an
Imp. Plan He Issues a Notice to
Person and the State.
30
Days
Admr. Issues Order
Requiring Compliance
With Plan if Violation
Continues 30 Days
Beyond Notice.
If Violations of any Plan so Wide-
Spread that State is Failing to
Enforce Plan Admr. So Notifies
State any Time He has such Infor-
mation .
30
Days
If Violation Continues
Past 30 Days, Admr. Issues
Public Notice and Enforces
Plan Until State Satisifies
Admr. that it will Enforce
Plan.
When Admr. has Info. That Person
is Violating Section 111 (e), 112 (c),
or 114. He issues Order Requiring
Compliance After Person has bad an
opportunity to Confer With Adminis-
tration Regarding Violation.
-------
preservation, product improvement in addition to storage, handling,
packaging and shipping.
The principal air pollutant emission from the food and agricultural
processing industry is particulate matter although other pollutants
are found in certain particular types of processes. The other pol-
lutants which occur are principally odors (such as trimethylamines
and hydrogen sulfides), nitrogen oxides, aldehydes, organic acids,
hydrocarbons, carbon monoxides, and fluorides. Although we have
essentially no information on trace metal air pollutants from this
industry it is probable that such emissions do occur in certain seg-
ments of the industry. One should be cognizant of potential pollutants
such as biological aerosols and the arsenical and mercurial based pes-
ticides. The latter are often used as defoliants to aid in mechanical
cotton picking operations.
The Air Pollution Control Office has not in the past conducted de-
tailed studies of the food processing industry although at the pre-
sent time we have planned under contract a comprehensive systems
study of the food and grain industry.
The Air Pollution Control Office of EPA currently has in the process
of publication a comprehensive document entitled Air Pollutant Emission
Factors ^'. An Emission Factor is defined as "the statistical average
of the rate at which a pollutant is released to the atmosphere as a
result of some industrial activity divided by the level of that activity."
For example, if it is determined that 26,000 tons of carbon monoxide is
released to the atmosphere during the production of 260,000 tons of
ammonia, the carbon monoxide emission factor for ammonia production
would be 200 pounds of carbon monoxide released per ton of ammonia
produced. Therefore, it is seen that the emission factor relates the
quantity of pollutant emitted to some indication such as production
capacity, quantity of fuel burned, etc.
Emission factors are determined and estimated by a whole spectrum of
techniques. These include detailed source testing involving many
measurements related to a variety of process variables, single mea-
surements not clearly defined as to their relationships to process
operating conditions, process material balances and engineering
appraisals of a given process. One must be cautious in applying
emission factors. In general, emission factors are not precise in-
dicators of emissions from a given source but are extremely useful when
intelligently applied in conducting source inventories as part of local,
community, or nationwide air pollution studies.
The Air Pollutant Emission Factor document mentioned above will present
general information, emissions and control, and emission factors for
some ten major industrial categories encompassing some 85 industrial
processes. With respect to the food and agricultural industry cat-
egory the following processes are covered: Alfalfa Dehydrating,
Coffee Roasting, Cotton Ginning, Feed and Grain Mills and Elevators,
Fermentation, Fish Processing, Meat Smokehouses, Nitrate Fertilizers,
30
-------
Phosphate Fertilizers, Starch Manufacturing, and Sugar Cane Processing.
Presented below is a summary of the air pollution aspects of the food
and agiculture industry. Information on nitrate and phosphate fertilizers
are not given in this paper.
The principal air pollutant emissions from the above mentioned processes
are described in the following sections.
Alfalfa Dehydrating. An alfalfa dehydrating plant produces an animal
feed from alfalfa. The dehydration and grinding of alfalfa to produce
alfalfa meal is a dusty operation most commonly carried out in rural areas.
Wet, chopped alfalfa is fed into a direct-fired rotary drier. The dried
alfalfa particles are conveyed to a primary cyclone, where heavy trash
is removed. A second cyclone discharges material to the grinding equip-
ment, which is usually a hammer mill. The ground material is collected
in an air-meal separator and either conveyed directly to bagging or
storage, or blended with other ingredients.
Sources of dust emissions are the primary cyclone, grinders and air-meal
separator. Overall dust losses have been reported as high as 7 percent
but average losses are around 3 percent by weight of the meal produced.
The use of a baghouse as a secondary collection system can greatly re-
duce emissions. Emission factors for alfalfa dehydrating are presented
in Table 1.
TABLE 1 PARTICULATE EMISSION FACTORS FOR ALFALFA DEHYDRATION
Particulate Emissions Ib/Ton of Meal Produced
Uncontrolled
Baghouse Collector
60
3
Coffee Roasting. Coffee, which is imported in the form of green beans,
must be cleaned, blended, roasted and packaged before being sold. In a
typical coffee roasting operation, the green coffee beans are freed of
dust and chaff by dropping the beans in a current of air. The cleaned
beans are then sent to a batch or continuous roaster. During the roasting,
moisture is driven off, the beans swell, and chemical changes take place
that give the roasted beans their typical color and aroma. When the
roasting has reached a certain color the beans are quenched, cooled, and
stoned.
Dust, chaff, coffee bean oils (as mists), smoke, and odors are the prin-
cipal air contaminants emitted from coffee processing. The major source
of particulate emissions and practically the only source of aldehydes,
nitrogen oxides and organic acids is the roasting process. In a direct-
fired roaster gases are vented without recirculation through the flame.
However, in the -fired roaster, a portion of the roaster gases
are recirculated and particulate emissions are reduced. Essentially,
complete removal of both smoke and odors from the roasters can be ob-
tained with a properly designed afterburner.
31
-------
Particulate emissions also occur from the stoner and cooler. In the
stoner contaminating materials heavier than the roasted beans are
separated from the beans by an air stream. In the cooler, quenching
the hot roasted beans with water causes emissions of large quantities
of steam and some particulate matter. Table 2 summarizes the emissions
from the various operations involved in coffee processing.
TABLE 2 UNCONTROLLED EMISSION FACTORS FOR ROASTING PROCESSES
(pounds per ton of green beans)
Pollutant
Type Process Particulates NO^ Aldehydes Organic Acids
Roaster
Direct-Fired 7.6 0.1 0.2 0.9
Indirect-Fired 4.2 0.1 0.2 0.9
Stoner and Cooler3 1.4
Instant Coffee Spray Dryer 1.4b
a - If cyclone is used emissions can be reduced by 70 percent.
b - Cyclone plus wet scrubber always used and thus this represents con-
trolled factor.
Cotton Ginning. The primary function of a cotton gin is to take raw
seed cotton and separate the seed and the lint. A large amount of trash
is found in the seed cotton which must be removed. The problem of col-
lecting and disposing of gin trash falls into two main areas. The first
consists of collecting the coarse heavier trash such as burs, sticks,
stems, leaves, sand, and dirt. The second problem area is that of col-
lecting the finer dust, small leaf particles and fly lint that are dis-
charged from the lint after the fibers are removed from the seed. From
one ton of seed cotton approximately one 500 pound bale of cotton can
be made.
The major sources of particulates from cotton ginning include the unloading
fan, the cleaner and the stick and bur machine. From the cleaner and
stick and bur machine a large percentage of the particles settle out in
the plant. Thus an attempt has been made in Table 3 to present emission
factors which take this into consideration. Where cyclone collectors are
used emissions have been reported to be about 90 percent less.
Feed and Grain Mills and Elevators. Grain elevators are primarily trans-
fer and storage units and are classified as either the smaller more
numerous country elevators or the larger terminal elevators. At grain
elevator locations the following operations can occur: receiving, trans-
fer and storage, cleaning, drying, and milling or grinding. Many of the
large terminal elevators also process grain at the same location. The
grain processes may include wet and dry milling (cereals), flour milling,
32
-------
TABLE 3 UNCONTROLLED EMISSIONS FROM COTTON GINNING OPERATIONS
Estimated Atmospheric Emission (Ibs/bale cotton^
Total Settled Out Estimated
Process Participates 100 Emission Factor
Unloading Fan 15 0 15
Cleaner 10 70 3
Stick and Bur
Machine 6 95 0.3
TOTAL 31 -- 18
a - One bale equals 500 pounds
oil seed crushing and distilling. Feed manufacturing involves the re-
ceiving, conditioning (drying, sizing, cleaning), blending and pelleting
of grains and their subsequent bagging or bulk loading.
Emission from feed and grain operations may be separated into those oc-
curring at elevators, and those occurring at grain processing operations
or feed manufacturing operations. Emission factors for these operations
are presented in Table 4. Since dust collection systems are generally
applied to most phases of these operations to reduce product and com-
ponent losses, the selection of the final emission factor should take
into consideration the overall efficiency of these control systems.
The emissions from grain elevator operations are dependent on the types
of grain, the moisture content of the grain (usually 10-30 percent),
amount of foreign material in the grain (usually 5 percent or less), the
degree of enclosure at loading and unloading areas, the type of cleaning
and conveying, and the amount and type of control used.
Factors affecting emissions from feed manufacturing operations include
the type and amount of grain handled, the degree of drying, the amount
of liquid blended into the feed, the type of handling (conveyor or
pneumatic), and the degree of control.
Fermentation. For the purpose of this paper only the fermentation in-
dustries associated with food will be considered. This includes the
production of beer, whiskey and wine.
The manufacturing process for each of these is similar. The four main
brewing production stages and their respective sub-stages are: (1)
Brewhouse operations, which includes a) malting of the barley, b) addition
of adjuncts (corn, grits and rice) to barley mash, c) conversion of starch
in barley and adjuncts to maltose sugar by enzymatic processes, d) sep-
aration of wort from grain by straining, and e) hopping and boiling of
the wort; (2) fermentation, which includes a) cooling of the wort, b)
additional yeast cultures, c) fermentation for 7 to 10 days, d) removal
of settled yeast, and e) filtration and carbonation; (3) aging, which
33
-------
TABLE 4 UNCONTROLLED PARTICIPATE EMISSION FACTORS FROM GRAIN
HANDLING AND PROCESSING
(pounds per ton of grain processed)
Type Source
Emissions
Terminal Elevators
Shipping or Receiving
Transferring, Conveying, etc.
Screening and Cleaning
Drying
Country Elevators
Shipping or Receiving
Transferring, Conveying, etc.
Screening and Cleaning
Drying
Grain Processing
Alfalfa Meal Milling
Corn Meal
Soybean Processing
Barley or Wheat Cleaner
Milo Cleaner
Barley Flour Milling
Feed Manufacturing
Barley
1
2
5
6
5
3
8
7
0.2
5
7
0.2£
0.4£
3a
a - At cyclone exit (only non-ether soluble particulates")
34
-------
lasts from 1 to 2 months under refrigeration and (4) packaging, which
includes a) bottling-pasteurization and b) racking draft beer.
The major differences between beer production and whiskey production are
the purification and distillation necessary to obtain distilled liquors
and the longer period of aging. The primary difference between wine
making and beer making is that grapes are used as the initial raw
material rather than grains.
Emissions from fermentation processes are nearly all gases and primarily
consist of carbon dioxide, hydrogen, oxygen, and water vapor, none of
which presents an air pollution problem. However, emissions of part-
iculates can occur in the handling of the grain in the manufacture of
beer and whiskey. Gaseous hydrocarbons are also emitted from the
whiskey aging warehouses. No significant emissions have been reported
for the production of wine. Emission factors for the various operations
associated with beer, wine and whiskey production are shown in Table 5.
TABLE 5 EMISSION FACTORS FOR FERMENTATION PROCESSES
Type Product Particulates Hydrocarbons
Beer (Ib/ton of grain processed)
Grain Handling3 3
Drying Spent Grain, etc.a 5 NA
Whiskey (Ibs/ton of grain processed)
Grain Handling3 3
Drying Spent Grains, etc.3 5 NA
Aging (Ibs/year/barrel of
whiskey stored) -- 10
Wine Neg.b Neg.b
a - Based on section of grain processing.
b - No significant emissions.
NA - No emission factor available, but emissions do occur.
Fish Processing. The canning, dehydration, smoking of fish, and the manu-
facture of fish meal and fish oil are the important segments of fish
processing. There are two types of fish canning operations — the "wet-
fish" method in which the trimmed fish are cooked directly in the can and
the "pre-cooked" process in which the whole fish is cooked and then hand-
sorted before canning.
35
-------
A large fraction of the fish received in a cannery is processed into by-
products. The most important of these by-products is fish meal. In the
manufacture of fish meal fish scrap from the canning lines is charged to
continuous live-stream cookers. After the material leaves the cooker it
is pressed to remove oil and water. The press cake is broken up, usually
in a hammer mill, and dried in a direct-fired rotary drier or in a steam-
tube rotary drier.
The biggest problem from fish processing is odor emissions. The principal
odorous gases generated during the cooking process of fish meal manufac-
turing are hydrogen sulfide and trimethylamine. Some of the methods used
to control odors include activated carbon adaorbersj scrubbing with some
oxidizing solution and incineration. The only significant sources of
dust emissions in fish processing are the driers and grinders used to
handle dried fish meal. Emission factors for fish meal manufacturing
are shown in Table 6.
TABLE 6 EMISSION FACTORS FOR FISH MEAL PROCESSING
Trimethylamine Hydrogen
Emission Source Particulates (CH^)^N Sulfide H?S
Coolers (Ibs/ton fish meal
produced)
Fresh Fish -- 0.3 0.01
Stale Fish -- 3.5 0.2
Driers (Ibs/ton fish scrap) 0.1
Meat Smokehouses. Smoking is a diffusion process in which food products
are exposed to an atmosphere of hardwood smoke, causing vsrious organic
compounds to be absorbed by the food. Smoke is produced commerically in
the United States by three major methods: (1) burning dampened sawdust
(20-40 percent moisture), (2) burning dry (5-9 percent moisture) sawdust
continuously, and (3) by friction. Burning dampened sawdust and kiln-
dried sawdust are the most widely used methods. Most large, modern,
production meat smokehouses are the recirculating type, in which smoke is
circulated at reasonably high temperatures throughout the smokehouses.
The emissions from smokehouses are generated from the burning hardwood,
rather than from the cooked product itself. Based on approximately 110
pounds of meat smoked per pound of wood burned, emission factors have
been derived for meat smoking. These factors are presented in Table 7.
Emissions from meat smoking are dependent on several factors, including
the type of wood, type of smoke-generator, moisture content of the wood,
air supply, and amount of smoke recirculated. Both low voltage electro-
static precipitators and direct-fired afterburners may be used to reduce
particulate and organic emissions. Thus controlled emission factors have
also been shown in Table 7.
36
-------
TABLE 7 EMISSION FACTORS FOR MEAT SMOKING
(pounds per ton of meat)3
Pollutant Uncontrolled Controlledb
Particulates 0.3 0.1
Carbon Monoxide 0.6 Neg.C
Hydrocarbons (Clfy) 0.07 Neg.c
Aldehydes (HCHO) 0.08 0.05
Organic Acids (Acetic) 0.2 0.1
a - Based on 110 pounds of meat smoked per pound of wood burned.
b - Controls consist of a wet collector and low voltage precipitator in
series, or direct-fired afterburner.
c - With afterburner
Starch Manufacturing. The basic raw material in the manufacture of starch
is dent corn which contains starch. The starch in the corn is separated
from the other components by "wet milling."
The shelled grain is prepared for milling in cleaners which remove both
the light chaff and any heavier foreign material. The cleaned corn is
then softened by soaking (steeping) it in warm water acidified with
sulfur dioxide. The softened corn goes through attrition mills, which
tear the kernels apart freeing the germ and loosening the hull. The re-
maining mixture of starch, gluten and hulls is finely ground and the
coarser fiber particles are removed by screening. The mixture of starch
and gluten is separated by centrifuges. After separation from the gluten,
the starch is filtered and washed. At this point it may be dried and
packaged for market.
The manufacture of starch from corn can result in significant dust emis-
sions. The various cleaning, grinding and screening operations are the
major sources of dust emissions. Table 8 presents emission factors for
starch manufacturing.
TABLE 8 EMISSION FACTORS FOR STARCH MANUFACTURING
(pounds per ton of starch produced)
Overall Emissions Particulates
Uncontrolled 8
Controlled3 0.02
a - Based on centrifugal gas scrubber
37
-------
Sugar Cane Processing. The processing of sugar cane starts with the
harvesting of the crops, either by hand or by mechanical means. If
mechanical harvesting is used much of the unwanted foliage is left and
it thus is standard practice to burn the cane before mechanical har-
vesting to remove the greater part of the foliage.
Following harvesting, the cane goes through a series of processes to be
converted to the final sugar product. It is washed to remove larger
amounts of dirt and trash, crushed and shredded to reduce the size of
the stalks and then the juice is extracted by two methods, milling or
diffusion. In milling the cane is pressed between heavy rollers to
press out the juice and in diffusion the sugar is leached out by water
and thin juices. The raw sugar then goes through a series of operations
including clarification, evaporation and crystallization in order to
produce the final product.
Most mills operate without supplement fuel because of the sufficient
bagasse (the fibrous residue of the extracted cane) that can be burned
as fuel.
The largest sources of emissions from sugar cane processing are the
open field burning in the harvesting of the crop and the burning of
bagasse as fuel. In the various processes including crushing, evap-
oration and crystallization some particulates are emitted but in
relatively small quantities. Emission factors for sugar cane pro-
cessing are shown in Table 9.
TABLE 9 EMISSION FACTORS FOR SUGAR CANE PROCESSING
Carbon Hydro- Nitrogen
Type Process Particulate Monoxide carbons Oxides
Field Burninga'b
(Ibs/acre burned) 225 1,500 300 30
Bagasse Burning
(Ibs/ton bagasse) 22
a - Based on emission factors for open burning of agricultural waste.
b - There are approximately 4 tons/acre of unwanted foliage on the cane
and 11 tons/acre of grass and weed all of which is combustible.
38
-------
REFERENCES
1. McGraw, M. J., Air Pollutant Emission Factors, United States, DHEW,
PHS, 1970.
39
-------
ACTIVITIES IN MANAGING SOLID WASTES
by
Jack DeMarco*
SCOPE OF THE PROBLEM
Our American way of life that has given us the highest standard of
living in the world has not been without its cost. Now that our gross
national product has exceeded the trillion dollar level, we are
discovering that increased goods and services come to us at the expense
of our environment. Attempting to assess the blame for the current
state of our environment is futile. Our efforts must be more
efficiently expended in ensuring that further degradation of the
environment does not take place.
It is encouraging to note that this is indeed happening. Public
concern for protecting environmental quality is growing. We all have
just cause to be concerned. Solid waste management practices throughout
the nation have in general been deplorable. An estimated 3.5 billion
tons of solid waste are generated each year from household, commercial,
and municipal, industrial, mining and agricultural activities across
the Nation. Over 2 billion tons of this total are related to industrial,
agricultural, and animal wastes. Annually, over a million abandoned
cars are scattered across the Nation, and a good portion of the 48
billion cans and 26 billion bottles thrown away each year also decorate
our landscape in an unsightly manner.
Unfortunately, the wastes that get into our solid waste handling
systems are not managed in a way that will prevent degradation of
environmental quality. The results of our first national survey of
community solid waste practices revealed that the most common practice
of disposal for our community solid waste is the open burning dump (1).
Over 95 percent of the estimated 15,000 land disposal sites were judged
unsatisfactory as they related to potential air, water pollution, and
vector control problems. Incinerators did not fare much better. We
found that over 75 percent of the estimated 300 incinerators located
across the Nation were judged unacceptable in terms of the air and
water pollution problems that occur from their operation. In many cases
the objective of reducing the solid waste volume was conducted so
inefficiently that large amounts of residue were left to be disposed
of. An estimated $4.5 billion per year have been expended by
municipalities, private solid waste management firms, individuals, or
* Deputy Director, Division of Technical Operations, Solid Waste Manage-
ment Office, U.S. Environmental Protection Agency, Cincinnati, Ohio.
41
-------
private industrial organizations on solid waste collection and disposal.
Although this figure may seem impressive, it has not been enough to
provide satisfactory management systems for the waste of this Nation.
We must all do more to stop the degradation and then ultimately restore
the quality of our environment.
THE FEDERAL ROLE
Until October 20, 1965, there was practically no federal involvement
with solid waste management. On that date, Congress, in passing the
Solid Waste Disposal Act, outlined responsibilities for the Federal
Government that would speed the growth of solid waste technology (.2) .
On October 26, 1970, Congress amended the Solid Waste Disposal Act to
further strengthen and expand the federal role (3). The original act
and its amendments state that the federal role shall be;
1. To promote the demonstration, construction, and application of
solid waste management and resource recovery systems which preserve
and enhance the quality of air, water, and land resources.
2. To provide technical and financial assistance to States and local
government and interstate agencies in the planning and development of
resource recovery and solid waste disposal programs.
3. To promote a national research and development program for improved
management techniques, more effective organizational arrangements, and
new and improved methods of collection, separation, recovery, and
recycling of solid wastes, and the environmentally safe disposal of
nonrecoverable residues.
4. To provide for the promulgation of guidelines for solid waste
collection, transport, separation, recovery, and disposal systems.
5. To provide for training grants in occupations involving the design,
operation, and maintenance of solid waste disposal systems. The Solid
Waste Management Office, of the newly formed U.S. Environmental
Protection Agency, has many activities underway designed to fulfill
the federal role of helping to solve the national solid waste management
problem.
Demonstrations
Our program has funded over 125 demonstration grant projects to date (4),
Some are designed to show the feasibility of new and improved solid
waste management technology. These grants allow for full-scale demon-
stration of methods, facilities, and equipment under actual operating
conditions. Other demonstration grants have been awarded to demonstrate
that sanitary and economic solid waste operations can be accomplished
by regional authorities.
42
-------
There are completed or currently active demonstration grants for all
phases of solid waste management. Some of these projects that deal
directly with agricultural and animal waste problems r.ay be of particular
interest to you. The Metropolitan Sanitary District of Greater Chicago
and the University of Illinois are demonstrating what agricultural
benefits and environmental changes result from using digested sewage
sludge on field crops. The objective of the project is to determine
the practicality of disposing of digested sewage solids on land.
Primary consideration is given to investigating factors relevant to
surface and groundwater contamination; soil pollution by heavy metals,
grease, and organic compounds inherent in sewage sludge; method,
frequency, rates, and times for safe application on sandy soil. Pre-
liminary results indicate that the uptake of zinc in soy beans, reed
canary grass, and grain sorghum plants appeared to increase when sludge
was applied. Additionally, the yield from corn and kenaf plots showed
a favorable increase in yield as the sludge application was increased.
Continuing laboratory and greenhouse experiments will obtain information
concerning the fertility value of digested sludge, the amounts of
supplemental potassium fertilization required for high yields of corn
and soy beans, the accumulation of heavy elements in the soil, and
methods of reducing nitrate accumulations in soil drainage waters.
These investigations should provide a sounder basis for the ultimate
disposal of millions of tons of digested sewage sludge and animal manure.
Another demonstration project is designed to improve present methods of
land disposal of waste sea clamshells. The current disposal methods
have resulted in public health hazards, nuisances, and associated social
and political problems that plague coastal communities. The objective
of this project is to demonstrate a practical use of waste sea clam-
shells as oyster cultch material. The oyster planting areas will be
scientifically sampled and the shells examined to determine the effect
of volume and area of planting and of shell size on intensity of spat-
setting and survival. Data to determine any possible ill effects on
the environment will be correlated with spat-setting determinations to
evaluate overall conditions in these oyster setting areas.
We have two demonstration projects underway on dairy manure management
methods. Both studies involve evaluation of existing and proposed
methods for collecting dairy manures. In Pullman, Washington, the
objectives to be demonstrated are the feasibility of a properly con-
structed and operated anerobic lagoon for low-cost storage of dairy
manure during seasons unfavorable for land disposal; the advantages of
properly scheduled applications of dairy manure to farm lands and areas
having seasonally high rainfall and land runoff problems; and the
feasibility of employing lagoon treatment for dairies not having
sufficient land for solids disposal.
In a study at Cerritos, California and through a subsequent demonstra-
tion grant with Los Angeles County, the relative economic and aesthetic
advantages of a water transport system versus a dry system for collection
and disposal of manure will be compared. The problems related to raising
43
-------
and feeding large numbers of dairy cattle on concentrated feedlots and
corrals located near residential developments will also be studied.
Dairy manure composting will be the principal disposal aspect investi-
gated. The basic data will hopefully aid in establishing the required
size of composting units needed for a given number of cows, the frequency
of turning needed to produce a suitable product in minimum time, and
the aesthetic and vector problems present, if any. Through the grant,
an attempt will be made to evaluate the attitudes of people living near
dairies and also to assess cause of comolaints about the dairies from
sanitarians, city managers, and other officials.
All this is only a small sampling of the many demonstration projects
underway. These projects were selected because of some related interests
they have to this particular symposium. Still other projects deal with
milling, grinding, baling, bagging, and crushing of refuse to obtain
volume reduction or to improve salvage; incineration; more effective
use of land disposal methods; composting, rail and pipeline transpor-
tation of waste, and above-ground fills.
One of our particularly promising demonstration projects is related to
the use of solid waste as a supplementary fuel in pulverized, coal-fired
boiler furnces for generating steam for electric power. The project
may result in a solid waste management technique that will provide:
economic disposal of urban-area refuse, a low-sulfur fuel to assist in
reducing the air pollution burden from utilities, and utilization of
waste material as a resource with the consequent conservation of our
reserves of coal as a natural resource.
Another promising project is being conducted that will include fluid
mechanical separation of solid waste, fluid-bed oxidation of combustibles,
and sanitary landfilling of inerts. The project includes potential
resource recovery aspects for paper fibers, ferrous and nonferrous metals,
and glass.
Our many projects cover the problems faced by small rural communities
as well as those faced by thelarge metropolitan complexes of our Nation.
Research
Our Division of Research and Development has activities underway in all
phases of the solid waste management problems. Our research activities
include intramural projects and projects supported through contracts
and grants (5,6). Three distinct areas that provide a broad framework
for many of our present research activities are: (1) public health
effects of solid waste in various solid waste management systems; (2)
utilization of waste materials; and (3) improved process control
activities.
44
-------
Public Health Effects. This research involves investigation of pathogens
associated with incineration, composting, landfill, and other disposal
methods. It also involves investigation of chemical toxicants and
potential occupational hazards associated with solid waste management.
One of our initial intramural projects was a study of pathogen survival
during the course of incineration (7). It has been generally assumed
that pathogens are effectively destroyed through the incineration
process, but recent in-house studies indicate that these organisms may
survive in significant numbers in incinerator residues. Another in-house
project was designed to study the hazardous material selenium (8), This
naturally occurring substance is concentrated in plants that are the
source of cellulose for paper production. Since large amounts of paper
are present in solid waste, a study was conducted to assess what the
concentrations of selenium in solid waste were at different phases
throughout a processing system. Raw solid waste contained between 0.9
and 4.74 micrograms per gram; incinerator residue, 0.003 microgram per
gram; the fly-ash quench water, 0.023 mg/1; the incinerator stack gas,
0.23 microgram per cubic meter; and newspaper, 8.6 micrograms per gram.
Recommended threshold limits for selenium in air are 0.2 milligram per
cubic meter and in drinking water 0.01 mg/1 (9). If we are to know
whether the components of solid waste are within allowable thresholds,
we must continue further investigations of this sort.
Utilization. These activities involve characterizing various types of
solid waste, and determining their chemical, physical, and microbiological
characteristics so that substances with an intrinsic value can be
identified and returned to the useful product cycle. Agricultural
wastes, in particular, may have a large potential in this respect. An
example of a project directed at utilizing solid waste is one to develop
physical, chemical, and microbial systems for converting the cellulose
in solid waste into useful material. Preliminary tests have been run
with microorganisms and enzymes to determine their relative abilities
to degrade cellulose such as paper, rice hulls, and mixed refuse. The
use of certain effective fungi results in a product with a 15 percent
protein content. In another study, a 2-liter pressure reactor has
been found effective in producing glucose from waste cellulose by an
acid hydrolysis process.
Activities are underway through a research grant that will attempt to
develop microbial protein food from cellulosic solid waste. A pilot
plant is being constructed that will initially utilize sugar cane
bagasse as a substrate for this conversion process. Another research
institution has a grant to pursue the study of chemical transformations
of solid waste to products of useful value. To date, their studies
indicate that the most promising chemical treatment processes appear to
be utilization of waste cellulose by ester and ether formation and
hydrogenation of wastes to oil. The transformed wastes may find use
as construction materials, containers, laminates, and other useful
products.
45
-------
Process Control. In the area of process control, we are seeking
improved and integrated processing systems for storage, collection,
transportation as well as ultimate disposal of solid waste. These
studies rely heavily on pilot engineering investigations. An example
of such activity would be the joint U.S. Public Health Service—
Tennessee Valley Authority Composting Plant at Johnson City, Tennessee
(10). This facility has been used as a combination research testing
and demonstration plant for composting solid waste and sewage sludge.
Another type of process control activity is the fluidized-bed investi-
gations being conducted by research grants in West Virginia.
Technical Operations
Our Division of Technical Operations conducts many activities directly
related to coping with the national solid waste management problem.
These include planning, training, and technical assistance.
Planning. State and interstate planning grants are awarded to assist
States in developing comprehensive State solid waste management plans
designed to protect the public from pollution, disease, and nuisance;
to provide an effective and economic means of managing solid waste; and
to recover resources that can be returned to a productive cycle.
Through these plans, State agencies can encourage and guide both local
and regional efforts. Since the award of the first grant in June 1966,
we have awarded grants to 43 states, 5 interstate agencies, the District
of Columbia, Puerto Rico, and Guam. Actual progress is not measured
by simply awarding grants. More important is the completion and
implementation of the comprehensive solid waste plans. We are pleased
that 22 State and 3 interstate plans are completed and implementation
has begun on many of them. Recent amendments to the Solid Waste
Disposal Act have expanded our planning grants to include local govern-
ment entities.
Training. Training of solid waste management personnel is another
area in which there is still a need. Basic training courses in solid
waste management have been offered in Cincinnati and at field locations
by our personnel. Over 3,500 trainees have attended these short-term
(usually 1-week) courses, which have provided the basic orientation
for many State agency regulation personnel, university professors,
planners, and supervisors and managers of public and private solid
waste management systems.
In addition to our in-house training activities, we now have 13 active
solid waste training grants at the graduate level (6). Through these
training grants and our in-house efforts, we are attempting to fill
the need for solid waste management practitioners. The Solid Waste Disposal
Act amendments have also expanded the list of eligible participants for
solid waste training grants. This expansion should substantially assist
our efforts to provide trained personnel at all levels of operation for
solid waste management activities.
46
-------
Technical Assistance. Much of our past effort has been expended in
providing technical assistance, and we expect this to continue. Although
technical assistance was not defined in the Solid Waste Disposal Act,
I believe the simplest definition might be that "technical assistance
is the provision of existing technology to solve present problems." At
least three basic methods are used to provide technical assistance:
1. Developing and disseminating technical data and information related
to the various aspects of solid waste management.
2. Providing guidelines and standards for acceptable solid waste
management.
3. Providing technical personnel to conduct studies, surveys, and
evaluations that will assist individuals, public agencies, and private
organizations in solving their solid waste management problem.
Technical Data. The technical data we supply includes such information
as Sanitary Landfill Facts, which draws together general information
that can be used by personnel involved with the management of environ-
mental activities who are not technically versed in all aspects of
sanitary' landfill operation (11).
For the more technically oriented personnel, we can provide data on
such matters as the characteristics of solid waste (12) or the physical,
chemical, and microbial characteristics of emissions from solid waste
processing operations such as incinerators. Other types of technical
data have been issued on solid waste management in municipal, commercial,
industrial, and recreational activities (14-17).
Additional data are being developed through industrial solid waste
contracts. A report on our contract on the status of solid waste manage-
ment in the food processing industry will be included in the proceedings
of this symposium. We also have contracts underway or already completed
that will result in technical reports on the polymer, rubber, automotive,
auto dismantling, printing and publishing, industrial chemical, drug,
and household appliance industries.
Guidelines and Models. Our present guidelines are state-of-the-art
documents that delineate the best technical information and methodology
currently available for practicing proper solid waste management at
various types of facilities. One such document, already completed, is
Incinerator Guidelines 1969 (18). A panel of nationally recognized
experts assisted us in developing this guideline. The document is
intended to aid in the design and operation of incinerators by making
known the best methods now available. Another document, now in the
final publication stages, deals with the design and operation of sani-
tary landfill (19). An additional document is on current methodology
for closing open dumps (20).
47
-------
In the next two years, many more such documents will be developed for
all phases of solid waste management and resouce recovery systems. Two
projects on legislation are now in the final stages of development.
One will result in a model State solid waste management act. The other
project will result in model local, county and regional ordinances for
solid waste management. Activities such as these are intended to
provide guides, for use throughout the Nation, for people at all levels
of government and private industry interested in the proper management
of solid waste.
Study, Survey, and Evaluation Assistance. Another method of providing
technical assistance is by having our technical staff directly assist
those individuals, agencies, or organizations that request our advice,
counsel, and opinion. In fiscal year 1970 we received over 750 such
technical assistance requests. To date in this fiscal year, we have
received over 400 requests. Many only require a discussion of technical
aspects of some process of solid waste management. Others require
field site visits, studies, and evaluations by a qualified person or
team from the Division of Technical Operations. Technical requests for
field studies have included short-term testing of incinerators for all
aspects of their environmental impact including the quality of air,
water, and residue emissions from the facility, as well as evaluating
a facility's efficiency for solid waste reduction and its economic
operation. These studies have assisted in evaluating incinerator
operations with the purpose in mind of upgrading them so that they can
more adequately perform their function of reducing the volume of solid
waste while minimizing any insult to the quality of our environment.
CLOSING REMARKS
I have tried to give you a brief cross section of some of the activities
of the Federal solid waste management program. I would like to emphasize
that a large part of our efforts are devoted to the concept of resource
recovery. Many of the projects I've described already embody this
concept. More needs to be done. No longer can we afford to permit
solid waste materials to be single-use items. They must be returned
to the useful product cycle to be used again. I am sure that you in
the food processing industry realize as well as I do that you are front
runners in applying the concept of resource recovery and utilization.
An estimated 70 percent of the process wastes throughout the food
processing industry are converted into byproduct use such as animal feed,
charcoal, oil, vinegar, alcohol, and fertilizer. The Nation would
encourage you to continue your activities along these lines, and we hope
to assist you in continuing to find new uses and new methods for taking
what was once waste, treating them as out-of-place resources, and then
returning them to the productive cycle of our economy.
48
-------
REFERENCES
1. Black, R. J., A. J. Muhich, A. J. Klee, H. L. Hickman, Jr., and
R. D. Vaughan. The national solid wastes survey; and interim report.
[Cincinnati], U.S. Department of Health, Education, and Welfare,
[1968]. 53 p.
2. The Solid Waste Disposal Act; Title II of Public Law 89-272, 89th
Cong., S.306, October 20, 1965. Washington, U.S. Government
Printing Office, 1966. 5 p.
3. The Resource Recovery Act; Public Law 91-512, 91st Cong., H.R. 11833,
October 26, 1970. [Washington, U.S. Government Printing Office,
1970.] 9 p.
4. Sponagle, C. E. Summaries; solid wastes demonstration grant
projects—1969. Public Health Service Publication No. 1821.
Washington, U.S. Government Printing Office, 1969. 175 p.
5. demons, C. A., and R. J. Black. Summaries of solid wastes program
contracts, July 1, 1965—June 30, 1968. Public Health Service
Publication No. 1897. Washington, U.S. Government Printing Office,
1969. 46 p. Supplement (insert). July 1, 1968—June 30, 1969.
12 p.
6. Lefke, L. W., comp. Summaries of solid wastes research and training
grants—1968. Public Health Service Publication No. 1596.
Washington, U.S. Government Printing Office, 1968. 48 p. Reprinted
1970. Supplemental (insert), Jan. 1, 1968—July 1, 1970. 8 p.
7. Peterson, M. L. and F. J. Stutzenberger. Microbiological evaluation
of incinerator operations. Applied Microbiology, 18(1): 8-13. July 1969.
8. Johnson, H. Determination of selenium in solid waste. Environmental
Science & Technology, 4(10): 850-853, Oct. 1970.
9. American Conference on Governmental Industrial Hygienists. Threshold
limit values of airborne contaminants adopted by ACGIH for 1969;
and intended changes. Cincinnati, 1969. 28 p.
10. Wiley, J. S., F. E. Gartrell, and H. G. Smith. Concept and design
of the joint U.S. Public Health Service—Tennessee Valley Authority
Composting Project, Johnson City, Tennessee. [Cincinnati],
U.S. Department of Health, Education, and Welfare, 1968. 14 p.
11. Sorg, T. J., and H. L. Hickman, Jr. Sanitary landfill facts.
Public Health Service Publication No. 1792. Washington, U.S.
Government Printing Office, 1968. 26 p.; 2d ed., 1970. 30 p.
49
-------
12. Klee, A. J., and D. Carruth. Sample weights in solid waste
composition studies. Journal of the Sanitary Engineering Division,
Proc. ASCE, 96(SA4): 945-954, Aug. 1970.
13. Achinger, W. C., and L. E. Daniels. An evaluation of seven
incinerators. In Proceedings; 1970 National Incinerator Conference,
Cincinnati, May 17-20, 1970. New York, American Society of
Mechanical Engineers, p. 32-64.
14. Perkins, R. A. Satellite vehicle systems for solid waste collection;
evaluation and application. Solid Waste Management Office for
release through National Technical Information Service, 1971.
(In press.)
15. DeGeare, T. V., Jr., and J. E. Ongerth. An empirical analysis of
commercial solid waste generation. (Submitted for publication.)
16. Cummins, R. L., W. T. Dehn, H. T. Hudson, and M. L. Senske.
Planning a comprehensive in-plant solid waste survey. [Cincinnati],
U.S. Department of Health, Education, and Welfare, 1970. 9 p.
17. Spooner, C. S. Study of recreation solid wastes for the U.S.
Department of Agriculture Forest Service. Public Health Service
Publication No. 1991. Washington, U.S. Government Printing
Office, 1969. 134 p. (In press.)
18. DeMarco, J., D. J. Keller, J. Leckman, and J. L. Newton. Incinerator
guidelines—1969. Public Health Service Publication No. 2012.
Washington, U.S. Government Printing Office, 1969. 98 p.
19. Brunner, D. R., and D. J. Keller. Sanitary landfill—design and
operation. Washington, U.S. Government Printing Office, 1971.
(In press.)
20. Closing open dumps. D. R. Brunner, S. J. Hubbard, D. J. Keller,
and J. L. Newton. [Washington, U.S. Government Printing Office],
1971. 19 p.
50
-------
POLLUTION ABATEMENT AND BY-PRODUCT
RECOVERY IN THE SHELLFISH INDUSTRY
by
Edwin Lee Johnson and Quintin P. Peniston*
INTRODUCTION
The shellfish industry, herein limited to crustacea species, is of f
considerable, and rapidly increasing, importance in many world localities.
Current domestic landings and imports of shrimp alone in the United States
amount to more than one half billion pounds per year. Characteristically,
all species possess an exo-skeleton which is not used for food purposes.
The yield of edible meat generally amounts to from 17 percent live
weight for small shrimp to about 25 percent for larger crab species. The
remainder is waste and in many fisheries this is discharged into harbor
waters adjacent to the processing facilities.
Food, Chemical & Research Laboratories, Inc. has been active for the past
five years in development of processes for separation of shellfish wastes
into marketable commodities of greater value than crude shellfish meals.
Much of this work has been supported by the U.S. Bureau of Commercial
Fisheries. Work is being continued under joint sponsorship by the
Environmental Protection Agency and the City of Kodiak, Alaska, Project
11060-FJQ.
Process Description
The process as originally conceived involves total recovery of waste
constituents to produce calcium chloride brines, protein and chitin as
by-products. It would consist of two counter-current extraction treatments.
The first would employ waste hydrochloric acid to decompose calcium
carbonate and produce calcium chloride brine. By suitable control of
flow and recycling, brine concentrations of the order of 20 percent can
be attained.
The residue, consisting of protein and chitin, is extracted with dilute
sodium hydroxide solution to dissolve and remove all of the protein in
the waste as a sodium proteinate solution. This extraction would also
be conducted as a counter-current operation in order to obtain high
protein concentration in the extract.
Protein is recovered from the extracts by neutralization with hydro-
chloric acid to the iso-electric point of the protein which occurs at a
*Food, Chemical & Research Laboratories, Inc., Seattle, Washington
51
-------
pH level of about 4.0. The precipitated protein is separated from the
supernate by filtration or centrifugation, washed to reduce the salt
content and dried.
Residue from the sodium hydroxide extraction is essentially pure chitin.
It can be washed and dried and marketed as chitin or can be converted
directly to soluble derivatives such as chitosan. The isolation of
chitin by demineralization of Crustacea shells with acid and removal of
protein with alkali has been known and practiced for many years. To
our knowledge however, recoveries of calcium chloride brine and protein
have not heretofore been considered.
In the final report under Contract No. BCF 14-17-0007-960 it was concluded
that the process as described above was both practically and economically
feasible in locations where waste hydrochloric acid was available at
low cost and a market existed for calcium chloride brines. It was
further concluded that the process without modification probably would
not be economic at Kodiak, due to the high costs for delivery of
hydrochloric acid to that area.
Revised Process for Kodiak, Alaska
In the interim period between completion of Contract BCF 14-17-0007-960
and execution of Contract No. BCF 14-17-0007-984 it was determined that
reversal of the order of demineralization and protein removal was
practical and that a revised process involving only protein extraction
from the waste shell could be economic at Kodiak. The residue, consisting
of a matrix of calcium carbonate and chitin, would have potential Alaska
markets as a fertilizer and soil liming agent or could be exported to
a Puget Sound location for demineralization. It has a very real
advantage over raw shell in being bacteriologically stable by virtue
of protein removal and can be shipped in bulk in a moist condition
without putrefaction.
In consideration of the above findings it was thought desirable to study
protein isolation and its characteristics using other shellfish species,
namely shrimp, Dungeness crab and Tanner crab, as well as King crab
on which most of the data to that date had been obtained. Trends at
Kodiak indicate rising levels of shrimp production with decreasing
production of King crab. Dungeness and Tanner crab may also become
larger factors in the total waste load.
Plan of Work for Present Contract
Accordingly, a request was made for continued support to study the
modified process as it might apply to Kodiak with particular emphasis
on isolation, recovery and characterization of protein from all species
of interest. This resulted in the present contract which was later
amended to include consideration of fish wastes such as those from
processing salmon, halibut and herring.
52
-------
Results of studies under this contract have been presented in progress
reports during the contract period. In the present report above results
are collected and summarized under the following subjects.
1. Waste Composition and Behavior in the Alkali Extraction
Process,
2. Factors Determining Rate of Protein Extraction.
3. Recoverability of Protein by Isoelectric Precipitation.
4. Factors Influencing Protein Quality.
5. Possible Utility of Products.
6. Processing Equipment Indications.
7. Pollution Abatement.
8. Economic Considerations.
DISCUSSION
1. Waste Composition and Behavior in the Alkali Extraction Process
While shellfish waste from all Crustacea species are primarily mixtures
of calcium carbonate, protein and chitin, there is considerable variation
between species as to the relative amounts of these substances and also
as to minor components such as lipids, pigments, phosphates, etc. Further,
butchering wastes from crab processing will contain visceral materials as
well as shell. In addition, a waste treatment plant at Kodiak, to offer
complete abatement of all primary pollution from fishery operations,
should provide facilities for handling fishery wastes such as salmon offal,
halibut trimmings and herring wastes. These will contain flesh, blood
and collagen type proteins, fats and bone as primary constituents.
Differences in physical characteristics of waste types are also of major
importance in designing a treatment facility. It has been noted that
Dungeness and Tanner crab shells are more dense than that from King crab
resulting in slower extraction of protein under the same processing
conditions. Shrimp shell is more papery in texture and while its protein
is readily extractible, "freeness" of the undissolved material is often
reduced by felting of flaky chitin particles. Thus, while a battery
of fixed bed diffuser cells might serve as extraction units for crab
shells, it. is doubtful if shrimp could be handled in such equipment.
Butchering waste and fishery wastes would also be unsuitable for fixed
bed treatment for the same reasons.
The optimum "product mix" produced by the treatment facility is also
largely governed by differences in waste composition. There are a number
of possibilities and these should be combined to provide the greatest
return for the least cost. This will be more fully discussed under 5.
Product Utility and 8, Economic Considerations.
53
-------
In prior reports we have used ternary diagrams to represent waste
compositions and changes resulting from processing. While these disregard
minor components, they are useful to illustrate major differences. In
Figure I the composition of shellfish wastes is presented summarizing
analyses conducted under the present contract and previously obtained
data. The relative percentages of protein and chitin have been calculated
to allow for a lower nitrogen content for shellfish waste protein than
the usual 16 percent assumed. A value of 15 percent agrees more closely
with analytical values on protein and actual chitin yields.
The points on the Figure for King, Tanner and Dungeness crab are for
picking line wastes. Inclusion of butchering wastes would lower the
points on the Figure since relative amounts of protein and fat ("other")
would be increased. A point is shown for red crab (Pleuroncodes
planipes). This is a very small pelagic species occurring in sami-tropical
waters in very large numbers. The point represents composition of the
whole animal.
Lines extended from the protein apex to the opposite side of the Figure
represent the protein extraction process. The composition of the chitin-
CaC03 residue is represented by points at the side of the Figure. It
is indicated that King crab waste would be a preferred material for
chitin isolation; shrimp waste would give the highest protein yields;
while Tanner and Dungeness crab are less desirable from both standpoints.
2. Factors Determining Rate of Protein Extraction
Rate studies have been conducted an extraction of Dungeness crab and
shrimp wastes using two different procedures. In one the rate at which
protein concentration approaches an equilibrium value is measured in
a batch treatment at constant alkali concentration and temperature. This
procedure has been found most useful to characterize the behaviour of
different wastes and to determine effects of pretreatment, alkali concen-
tration and temperature. Calculation of rate constants assumes that
the rate of change of protein concentration at the interface between
shell particles and ambient liquid is proportional to a driving force
equal to the difference between protein concentration in shell interstices
(P g) and that at the interface (P.^) .
d Pi/dt = K(PS - P±)
Further assumptions are that P^ is equal to the protein concentration
in the ambient liquid and Ps is equal to the total unextracted protein
dissolved in a constant interstitual volume equal to the water content
of the moist shellfish waste. Values for Ps can thus be calculated from
those for P. and numerical integration can be employed to obtain that
rate constant. This is equal to the slope of the line: loge (Pg - Pi)/Pgo
versus time. Typical data plots are shown in Figure 2 for shrimp and
Dungeness crab waste using different extraction conditions
54
-------
Mill Nl Itir 1/niM UO.
-------
SIil:] [It
{-[II Hill
fi-f i Ji M-
ili±iitUliM1mi-t
, 1 . :, i . l i ........
t-t i-i-1 i , i . . :
! m!-
ff{±tt±tt
FlH-i-i-H-!-
-------
It is found that both shrimp and crab wastes show an initial period of
very rapid extraction amounting to 30-50 percent of the total protein
followed by a rather sharp break and a slow extraction period which
fits the diffusion mechanism outlined above. This is undoubtedly an
oversimplification of the true mechanism but serves to give a numerical
index of the waste behaviour. It is noted that rate constants obtained
for shrimp waste are considerably higher than for Dungeness crab under the
stame conditions, probably reflecting the density of the shell. It is also
noted that temperature effects are higher than would be expected for
strict diffusion dependence of the rate constants. This suggests
chemical activation such as rupture of bonds between chitin and protein
as being involved in the process. Alkali concentration does not
appear to be an important variable. Sufficient alkali to satisfy
the base binding capacity of the protein is necessary. This appears to
be about 10 grams of sodium hydroxide per 100 grams of protein. Also
a pH level of about 12.5 or higher appears necessary for the extraction
since sodium carbonate solutions with concentrations up to 5 percent are
relatively ineffective. Increases in sodium hydroxide concentration
above about 0.5 percent do not produce proportional increases in
extraction rate. Most recent experiments indicate that addition of
sodium sulfite to the alkaline extraction liquor may have a specific
effect in increasing the extraction rate as well as other beneficial effects.
The other procedure for study of extraction rates has been the determination
of protein levels in eluates from percolation experiments in a fixed
bed diffuser type reactor.
Results from this type of experiment are more difficult to interpret
because more variables have to be considered such as flow rate and
alkalinity. Alkalinity is not constant during the extraction because
unextracted protein absorbs alkali during early stages of the experiment.
A mathematical analyses of the elution process was developed (1, 2)
which would allow prediction of elution rates from rate constants
determined by the "approach to equilibrium" procedure. The treatment
would be generally applicable to any fixed bed extraction process
which is diffusion controlled and may prove to be of value should
such types of processing become indicated.
3. Recoverability of Protein by Isoelectric Precipitation
Results from earlier experiments in July 1969 and prior indicated that
the solubility of alkali extracted protein at its Lsoelectric point
would be about 5 grains per liter (0.5 percent). To attain 95 percent
protein recovery would thus require a protein concentration in the
extract of 10 percent. This placed emphasis on the need for counter-
current treatment in order to build up the protein concentration in
the extract to the desired level.
Subsequent studies have shown that protein solubility at the isoelectric
point (actually, the pH of minimum solubility) is to some degree
57
-------
affected by the severity of the alkali treatment and to a very large
degree is a linear function of the salt concentration. Since there is a
minimum alkali combining capacity of about 10 percent for the extracted
protein equivalent to about 14.6 percent of the protein as salt on
neutralization, it is not possible to increase the protein concentration
without increasing the salt concentration proportionally. (One could,
of course, consider desalting procedures such as dialysis, ultra-
filtration or ion exclusion, but these probably would not be economic
in the process.) In addition, the added time of treatment to achieve
high protein concentrations in the effluent would result in more
degradation and more isoelectric protein solubility.
Above conclusions point to an extraction process in which alkali
concentration, total alkali used,and time of contact are all minimized.
In recent experiments these objectives have been approached and results
have been encouraging. In one experiment an isoelectric protein
solubility of 0.25 percent was attained at a salt concentration of 1.3
percent. This would amount to a protein recovery of 97.2 percent from
an 8.9 percent protein solution, assuming minimum alkali consumption.
The use of sodium hexameta phosphate to complex and precipitate addit-
ional protein from the supernate as suggested by Mr. John Spinelli (3) of
the Seattle B. C. F. laboratory also shows considerable promise. In
one experiment protein solubility was reduced to 0.07 percent representing
a recovery of 96.5 percent from a 2.37 percent protein solution.
Protein content of effluent from the process can be expressed as
population equivalent in terms of biochemical oxygen demand in the
following way:
Assume there is 8,800,000 pounds of 100 percent solids shrimp waste
collected per year out of 65,000,000 pounds of raw shrimp with a five
day BOD of 0.65 pounds oxygen consumed per pound of 100 percent solids
waste. This amounts to 5.7 million pounds of BOD per year for collectable
shrimp waste.
There is 4,800,000 pounds of 100 percent solids in collectable crab
waste with a five day BOD of 0.63 pounds oxygen consumed per pound of
100 percent solids waste. This amounts to 3.0 million pounds of BOD
per year for collectable crab waste.
The total for the two is 8.7 million pounds of BOD per year. With a
300 day season the average BOD per day from the collectable waste is
29,000 pounds or a population equivalent of 144,000 persons, assuming
0.2 pounds BOD per day per person.
A recovery of 95 percent of the collectable solids would reduce the
population equivalent to 7,000 persons by utilizing this process.
This remaining five percent would be treated by conventional secondary
methods further reducing the organic loading on the harbor.
58
-------
4. Factors Influencing Protein Quality
In Table I data were presented on amino acid composition of spray
dried crab and shrimp waste proteins in comparison with casein. This
table is re-presented in the present report. It was found that the
compositions of the two shellfish waste protein were quite similar
although determined by two different laboratories using different
procedures. They both compare favorably with casein in total essential
amino acid content. The percentages of arginine and isoleucine are
significantly higher for the shellfish proteins while casein shows
higher values for valine and leucine.
Feeding tests have been conducted on rats using both the shrimp and
crab waste proteins. Both show marked deficiency in sulfur-containing
amino acids which can be corrected by supplementation with either
cystine or methionine. In all other respects the shellfish waste proteins
were found to be equal to casein in nutritional value and there were
no toxic effects noted.
A sulfur balance on crab waste and on the extracted protein indicated
that most of the sulfur in the waste is still present in the spray
dried protein. (It should be noted however that a very slight odor of
hydrogen sulfide has been detected during neutralization of alkaline
extracts.) Based on assays for cystine and methionine (Table I), only
about half of the sulfur found in the protein can be accounted for. It
thus appears that sulfur-containing amino acids were present before
extraction at about twice the level found by assay on the isolated
protein.
Review of the literature^>5,6,7) on effects of alkaline treatment on
methionine and cystine in proteins suggests that methionine should be
relatively stable in the treatment but that cystine can be largely and
irreversibly converted to lanthionine which presumably has no nutritional
value as a sulfur source. The exact mechanism for the reaction is not
clear but the overall effect is the rupture of disulfide bonds with
elimination of one atom of sulfur and recombination of residues in &
thio ether linkage.
Cys tine: HOOC-CHN^-C^-S-S-CJ^CHNl^-COOH
Lanthionine: HOOC-CHNH2-CH2-S-CH2CHNH2-COOH + S
The form of the eliminated sulfur is not clearly established.
Normally, there is an equilibrium between cystine and cysteine residues
in proteins determined by the presence of oxidizing or reducing conditions.
in the system. The formation of cysteine with free sulfhydryl groups
may be an intermediate step in lanthionine reaction. If it is not, or
if the sulfhydryl group can be blocked from recombining as thio ethers,
the presence of a reducing agent should be beneficial in preventing
lanthionine formation. Sulfite ion, either by shifting the equilibrium
toward cysteine or by blocking recombination through formation of s-sulfo
cysteine groups might serve this purpose.
59
-------
Table I
Amlno Acid Composition of Spray
Dried Shellfish Waste Proteins
Sample
Lysine*
Histidine*
Arginine*
Aspartic Acid
Threonine*
Serine
Glutamic Acid
Proline
Glycine
Alanine
Cystine
Valine*
Methionine*
Isoleucine*
Leucine*
Tyrosine
Phenylalanine*
Tryptophan*
Total
As rec'd
Basis %
5.4
2.21
5.5
10.3
3.57
2.7
12.3
4.33
4.2
4.6
0.24
5.5
1.97
4.7
6.6
4.0
4.07
1.0
Irab Protein
100% protein
Basis %
6.35
2.60
6.47
12.1
4.20
3.18
14.5
5.10
4.94
5.41
0.28
6.47
2.32
5.53
7.78
4.70
4.80
1.18
Q7 <5
Shrim
As rec'd
Basis %
6.24
2.22
6.04
6.46
2.93
3.51
13.3
3.40
5.63
5.35
not detnd.
4.21
1.95
3.87
6.10
2.70
3.78
0.55
Casein
As rec'd
Basis %
6.24
2.22
6.04
6.46
2.93
3.51
13.3
3.40
5.63
5.35
not detnd.
4.21
1.95
3.87
6.10
2.70
3.78
0.55
100% protein
Basis %
8.34
2.97
8.06
8.63
3.91
4.69
17.8
4.54
7.52
7.14
5.62
2.60
5.17
8.14
3.61
5.05
0.73
100% protein
Basis %
6.02
2.31
2.41
4.45
3.81
5.88
21.90
15.71
1.16
1.47
not detnd.
7.91
2.75
3.91
11.07
2.72
5.46
Ca 1.0
104.5
99.94
*essential
-------
Experiments Indicate that the presence of sodium sulfite during
extraction does exert beneficial influences. These were noted in increased
extraction rate (rupture of disulfide crosslinks?) and in improved
appearance of the extracted proteins. Analytical results not available
at the time of the above report now suggest that reduction in extent
of cystine destruction may also have been achieved.
Cystine Vas determined by Vassel's modification of the Fleming reaction(S)
on hydrolyzates of the spray dried Dungeness crab waste protein (see
Table I) and of protein isolated from Diffuser Experiment No. 6.
Results showed 0.23 percent cystine in the Dungeness crab protein and
1.05 percent in the protein extracted in the cystine in the presence
of sodium sulfite. It woyld appear that cystine destruction was
reduced in the sulfite experiment.
5. Possible Utility of Products
Products which might be manufactured from shellfish and fishery wastes
by the proposed process would include the following:
a. Shellfish meal
b. Spray dried precipitated protein
c. Evaporated protein concentrates
d. Protein hydrolyzates
e. Deproteinized shell
f. Chitin and derivatives
g. Bone meal
Wide ranges in specifications are possible for most of these products
depending on differences in raw materials and processing conditions.
Obviously, all of the production from a processing facility will have
to be disposed of in some manner or other and the greatest return will
result from the product mix producing the highest sales at the least
overall cost. The optimum product mix will not be constant since raw
materials will vary seasonally and markets will be subject to their
usual instabilities.
The following observations on individual products will serve to
summarize our present conclusions regarding product utility.
a. Shellfish
The economics of production of crude shellfish meal in Alaska have been
frequently investigated and generally found to be unfavorable. For this
reason it has not heretofore been considered as a desirable product from
a waste treatment facility. However, there are arguments for providing
capability of manufacturing such products. Pollution abatement will
require that the facility receive and process all of the waste produced
61
-------
by fishery operations in the locality at any given time. Provision
should be made for overloads, and unforeseen processing difficulties
through alternative processing procedures. Also rising interest in fish
culture has attached new values to shellfish meals, particularly shrimp
meal due to its astaxanthin content. This carotenoid pigment imparts
a desirable pink coloration to the flesh of trout and other species. It
is possible that limited premium markets for shrimp meal may develop in
this area.
b. Spray dried precipitated protein
This would now appear to be the product offering the greatest promise
for maximum return. Incorporation in canned and processed pet foods offers
a large potential market. Several of the major pet food manufacturers have
expressed interest in these products based on examination of samples.
One manufacturer has indicated that they could absorb the entire production
at Kodiak now estimated to be about 4.9 million pounds per year. The
products might command a premium price over that justified by nutritional
value due to odor and flavor characteristics. Other markets such as
industrial uses and use in foods for poultry and livestock have not yet
been investigated.
c. Evaporated protein concentrate
We currently view such products as lower cost outlets for protein
materials which will not meet specifications of preferred products.
Multiple effect evaporation is inherently cheaper than spray drying and
the facility will probably have to process some types of fish offal not
suitable for premium products. Conversion to a product somewhat like
the concentrated "Fish solubles" of commerce at about 50 percent solids
might be the most economic treatment for such materials. They could be
marketed as feed ingredients or possibly for manufacture of liquid fish
fertilizer which has a large demand in the home garden trade. About 900
tons of 50 percent solubles can be produced from the salmon waste
generated by the 1970 Kodiak salmon pack.
d. Protein hydrolyzates
This represents another possible type of protein product. Alkaline
solutions of protein are preferred media for tryptic enzyme fermentations
and hydrolyzates might be marketable in specialty feed formulations or
as nutrients in antibiotic manufacture.
e. Deproteinized shell
The calcium carbonate - chitin residue from protein extraction of
shellfish wastes would be the largest tonnage product of a waste treatment
facility. It would also probably be a product with a low price potential,
requiring careful control of production costs such as expended labor,
heat, storage and freight to permit marketing at a profit for the enterprise,
62
-------
The preferred market would be as a raw material for chitin production but
this would depend on development of markets for chitin and derivatives
and the competitive position of deproteinized shell from Kodiak with raw
shellfish wastes in localities where chitin production were to be under-
taken. Assuming that there will be a net chemical cost for demineralization,
analytical data presented in Figure 1 indicate that deproteinized King
crab shell would be a preferred material for this market due to a higher
ratio of chitin to calcium carbonate.
Studies conducted by Dr. Laughlin at the Palmer Experimental Station of
the Alaska Department of Agriculture using samples submitted by our
laboratories indicate that deproteinized shell has merit as a liming
agent for Alaska soils. Further testo are scheduled for this year to
evaluate its merits as a nitrogen and phosphate source for growing plants.
No estimates are available at this time regarding the possible extent of
such markets in Alaska or the probable delivered cost of the material in
agricultural communities. One property of the deproteinized shell which
may be an asset in such markets is its microbiological stability. We
have found that unlike raw shell, the deproteinized residue can be left
in moist conditions at ambient laboratory conditons without putrefaction
for several weeks. This might permit bulk shipment and cheaper handling
costs. The amount of chitin-CaCOs resulting from the 1970 shrimp and
crab pack in Kodiak is in the order of 4,300 tons.
f. Chitin and derivatives
Due to its insolubility, no important uses have been developed for chitin
as such. It can however, be converted to soluble derivatives such as
chitosan for which many uses have been suggested. These include formation
of fibers and films, coatings for paper and for glass fibers to permit
dyeing, encapsulation agents for Pharmaceuticals, viscosity control agents
for drilling muds, thickeners for printing inks, textile sizing materials
and many others. We have found thirty or more U. S. Patents covering
derivatives of chitin and uses in above and similar applications. Most
of these patents were never exploited due to unavailability of chitin at
reasonable cost. Until recent years shellfish wastes were not available
in a single locality in quantities sufficient to produce chitin in
volume and processes for its isolation without protein recovery resulted
in too high a cost to compete with alternative material.
One new use of chitin derivatives which promises to supply a large volume
market is as cationic polyelectrolytes for water and waste treatment.
We have found that chitosan (deacetylated chitin) compares favorably with
synthetic polydectrolyte materials (polyacrylamides and polymerized quater-
nary ammonium compounds) as a coagulant and coagulant aid for removal of
turbidity from municipal and industrial waters. Due to current concern
with water pollution and the coming need for water re-use a rapidly
expanding demand for such products can be predicted. Economic considerations
indicate that chitosan could be offered competitively with other polyelectro-
lyte materials. Its non-toxic character and biodegradability should act
in its favor.
63
-------
g. Bone meal
Alkali extraction of protein from fishery wastes such as herring, scrap
fish and halibut trimmings will leave a residue of bone which could be
dried and ground and marketed as bone meal. In our experiments on herring
waste a product was obtained containing about 60 percent tricalcium phosphate
and 25 percent residual protein. Indicated production of such material
would probably be only two or three hundred thousand pounds per year
but it should be readily marketable as a home gardening aid. Possibly
blends with deproteinized shell would be marketable in the same field.
h. Fat concentrates
In processing picking line wastes from crab and shrimp production very
little fat has been encountered. Some fat is emulsified with the
alkaline protein extracts from shrimp wastes but this appears to be
carried down with the protein during iso-electric precipitation. The
treatment of crab butchering wastes and fishery wastes will presumably
increase the amount of fat in the process and means for handling it will
have to be provided. It is believed that centrifugal clarification of
alkali extracts will separate the fats and that they would be dried by
evaporation. Possible marketing as a feed additive should be investigated.
Production should not be greater than 25,000 gallons per year.
6. Processing Equipment Indications
The varied nature of raw materials to be handled by the waste treatment
facility with resultant changes in processing procedures and conditions
suggests that processing equipment should be designed as a collection
of unit operations rather than as a closely coordinated, continuous
process such as would be preferred with a fixed raw material and a standard
product. With this concept in mind, major equipment which would be
needed by the facility would be as follows:
a. Collection System
Separate holding hoppers holding the screened waste would be needed at
individual or groups of processing plants which operate on different
products simultaneously. Shrimp waste should be separated from crab
waste. Fish offal would not be mixed with shellfish wastes. Barges
transporting wastes could be compartmented or could make separate hauls
for different waste types.
b. Receiving Station
Separate dump pits should be provided for different waste types received
simultaneously. Conveyors would deliver wastes to hammer mill type
grinding equipment. Effluent from shellfish grinding would be passed
over graded screens to retain particulate matter and discharge waste-
water. (This will be treated by conventional secondary treatment
before final discharge into recovery waters.) Fish offal slurries from
the grinder would pass directly to an alkali extraction unit.
64
-------
c. Fish Offal Treatment
Separate equipment would be needed for fish offal and shellfish waste.
Fish offal slurries would be batch treated in a tank with slow agitation
adding just sufficient alkali to dissolve proteins and holding at 50-
60°C for about 30 minutes. Liquor would be drained through screens to
remove bone fragments and passed through a centrifugal clarifier to
remove fats after partial neutralization. A 50 percent solids "solubles"
product would be made from the liquor.
Bone would be collected and washed in centrifugal equipment, and then
dried in a rotary drum drier. The same drier could be used for
deproteinized shell or for crude shellfish meal depending on possible
scheduling of production. Drying would be followed by final grinding and
bagging operations.
d. Shellfish Waste Treatment
Ground shellfish waste from collection screens would be fed to a set of
probably five extraction units operated with continuous countercurrent
flow of liquor and shell. Units would be horizontal cylindrical shells
with internal screws designed to mix liquid and solid phases and to move
solids through the unit. Alkali addition, temperature and residence
time would be programmed in each stage to obtain an optimum balance between
extraction rate and severity of treatment. The last stage of solids
treatment would be essentially a washing stage. Most of the alkali
would be added in the first stage of solids treatment to take advantage
of the rapid extraction rate possible for 30 to 50 percent of the protein
in the raw shell. Concentrated liquor from the solids treatment would
be cooled in a heat exchanger, then centrifugally clarified and neutralized
to its isoelectric point. Protein would be collected on a vacuum
filter-washer, reslurried in water and spray dried. Filtrate from protein
recovery would be discharged to the secondary sewage disposal plant.
Deproteinized shell would be given a final neutralizing wash and dried
in a rotating hot air drier. The extent of drying would depend on
requirement for storage and shipment.
7. Pollution Abatement
As mentioned in the introduction, the yield of edible meat from shellfish
processing amounts to from 17 percent live weight for small machine
peeled shrimp to about 27 percent for larger crab species, with the
remainder being waste.
In the City of Kodiak there are 14 plants processing shrimp, King
crab, Tanner crab, Dungeness crab, salmon, halibut, herring and scallops.
Two plants process six of these eight items during the course of their
season with the rest of the plants processing two or more items. These
plants are located along a two and a half mile section of the waterfront
of the City of Kodiak.
Table II shows the amount of shrimp processed by these plants in 1970,
while Table III and IV show the production of crab and salmon processed.
65
-------
Table II
Shrimp Processed - 1970
(Million of Ibs. )
Month
Jan.
Feb.
March
April
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Plant 1
2.00
2.00
2.50
2.60
1.50
3.30
3.90
3.90
3.90
3.00
2.20
1.70
Plant 2
0.71
0.42
0.35
0.58
0.07
0.64
1.16
0.89
0.66
0.58
0.60
0.51
Plant 3
0.80
0.80
0.80
0.80
0.40
0.70
1.40
1.30
1.00
0.80
0.80
0.80
Plant 4
1.19
0.78
0.86
0.91
0.44
1.39
1.90
1.62
1.14
0.78
0.80
0.73
Total
4.70
4.00
4.51
4.89
2.41
6.03
8.36
7.71
6.70
5.16
4.40
3.74
Total
32.50
7.17
10.40
12.54
62.61
66
-------
Table III
Crab Production
for the Kodiak Area
Millions of pounds -
in 1970
by Months
Live weight
(Alaska Dept. of Fish and Game)
Month King
January 1 . 29
February -
•-J March -
April
May
June
July
August 1.69
September 3.69
October 2.47
November 1.73
December 1.22
Kodiak Area-
Tanner
0.71
.1.21
2.73
1.74
0.76
0.18
0.14
-
0.01
0.01
0.12
0.14
Dungeness
-
-
-
0.16
0.73
1.91
1.51
0.78
0.49
0.13
0.02
King
1.03
-
-
-
-
-
-
1.35
2.95
1.98
1.38
0.98
Citv of Kodiak
Tanner Dungeness
0.57
0.97
2.19
1.39
0.61 0.13
0.14 0.59
0.11 1.53
1.21
0.01 0.62
0.01 0.39
0.10 0.10
0.11 0.02
Total
1.60
0.97
2.19
1.39
0.74
0.73
1.64
2.56
3.58
2.38
1.58
1.11
Total
11.81
7.75
5.73
9.67
6.20
4.58
20.47
-------
Table IV
Week Ending
6-13
6-21
6-28
7-5
7-12
7-19
7-26
8-2
8-9
8-16
8-25
8-30
9-6
9-13
Total
1970 Salmon Production
Cases Packed
48 Ibs. each
1,781
1,232
17,508
789
14,774
37,838
34,835
40,006
26,714
10,612
13,131
5,314
514
-
205,048
at Kodiak
Frozen Salmon
Ibs.
-
95,280
74,664
27,270
77,402
257,356
392,827
371,214
275,276
19,829
61,164
562,417
-
399,709
2,614,408
68
-------
As presently practiced this waste is dumped into Kodiak Harbor. As
a result of this present discharge of over 73 million pounds of waste
there is up to four feet of residual organic matter in Kodiak Harbor.
While it has not been definitely established that this waste load is
totally harmful to the environment (some fisheries experts believe
that a part of it is food for other marine organisms in the food chain)
the practice does not comply with the water quality standards of the
State of Alaska.
As you may know these laws state that all wastes dumped into receiving
waters must have secondary treatment unless an engineering report
shows primary treatment is adequate.
To treat this entire waste load by secondary treatment is a gigantic
undertaking as up to 5 million gallons of water per day are consumed
by these 14 processors, spread along two and a half miles of waterfront.
This is complicated by the fact that no land is presently available for
such a plant along this area except an island located 600 feet across
a channel from the nearest mainland. We therefore, have been measuring
the organic loading of crab and shrimp wastes.
Table V shows the estimated loss of solids from crab processing as a
result of our process. This represents the cooking and washing solids
not collected by a 40 mesh screen. The visceral material together
with their wash waters are to be collected in their entirety to avoid
their introduction into the receiving waters. If this is followed only
7.3 percent of the total solids from the crab processing operations
will be lost.
Table VI shows the estimated loss of solids from shrimp processing as a
result of our process. This represents the cooking and washing solids
not collected by a 40 mesh screen. The collectable solids are 64.4
percent of the total waste load with 60 percent of the COD retained.
As all of the waters from the salmon canning operations are to be
collected their contribution to the pollution load will be negligible.
This amounts to 40 percent of the live weight catch or 6.44 million
pounds of wet wastes per year or 1.3 million pounds of solids.
The total pollution abatement by using this process is shown in Table VII.
This shows that 73.8 percent of the total pollution load of 20.2 million
pounds of 100 percent solids waste can be eliminated from Kodiak Harbor
by using this process.
8. Economic Considerations
Based on the 1970 production figures, there will be 13.6 x 10^ pounds
of shellfish solids collectable in Kodiak. Of this, 8.8 x 106 are
shrimp and 4.8 x 10^ are crab solids. Using an average of 40 percent
69
-------
Table V
CRAB
20 MILLION POUNDS/YEAR
LIVE WEIGHT
100% Solids %
LOSSES
Not caught by screening 3.8 x 105 # 7.3
COLLECTABLE WASTE SOLIDS
FOR PROTEIN RECOVERY
69% of catch as picked meat 2.02 x 106 # 38.8
25% of catch as leg sections 1.93 x 106 // 37.2
6% of catch as whole cooked crab 0 0
COLLECTABLE WASTE SOLIDS
FOR SOLUBLES RECOVERY
69% of catch as picked meat .64 x 106 # 12.3
25% of catch as leg sections .23 x 106 # 4.4
6% of catch as whole cooked crab 0 0
TOTAL COLLECTABLE 4.8 x 106 # 92.7
TOTAL SOLIDS 5.20 x 106 # 100%
70
-------
Table VI
SHRIMP PROCESSING
63 MILLION POUNDS/YR, LIVE WEIGHT
# X 106
#COD X 106
LOSSES
Not caught by screening
(100% Solids)
Peeled raw & cooked
Cooked & peeled
TOTAL
COLLECTABLE WASTE SOLIDS
100%
Peeled raw & cooked
Cooked & peeled
TOTAL
TOTAL WASTE SOLIDS
3.8
1.1
4.9
6.1
2.7
8.8
13.7
35.6
%
64.4
4.1
1.5
5.6
5.8
2.6
8.4
14.0
40.0
7.
60.0
LOSSES
Shrimp Waste
Crab Waste
Salmon Waste
COLLECTABLE WASTE
Shrimp
Crab
Salmon
TOTAL
Table VII
TOTAL POLLUTION ABATEMENT
# @ 100%
Solids X 106
4.9
0.4
0
5.3
8.8
4.8
1.3
14.9
20.2
24.2
2.0
0
26.2
43.5
23.8
6.5
73.8
100%
71
-------
protein from shrimp and 30 percent protein from the crab. We can
recover 3.5 x 10 of shrimo protein and 1.4 x 10^ pounds of crab
protein per year by using this process. The indicated market price for
this protein is 15 of solids would produce an additional 1,600 tons
of solubles worth $48,000.
The residual calcium carbonate-chitin complex of 4,300 tons is worth at
least $10. 00 per ton or $43,000 F.O.B. Kodiak for its lime value alone
and worth considerably more when markets open up for chitin as an
industrial chemical. The total worth of these products is $837,000.
The plant will have to convert a maximum of 100 tons of wet shell
residues plus the crab butchering waste and miscellaneous scrap fish
associated with the shrimp into products each day during the height of
the season in August.
It also must process the salmon waste which peaks during this time.
They average 100 tons per day during this period.
For these reasons the plant would have to handle 200 tons of wet waste
during a 20 hour period or 10 tons per hour. Such a plant is estimated
to cost in the neighborhood of $1,000,000 based on the costs of a 25
ton per hour conventional fish meal plant built in the continental
United States for a cost of $1,000,000.
Unfortunately at the time of this paper the total economic study of the
process has not been completed. However, based on the rough estimates,
this process appears to be profitable at Kodiak with the profitability
dependent on the method of financing, interest rates, inflation due
to labor and materials and other factors not yet determined.
CONCLUSION
A final report under EPA Grant 11060-FJQ is forthcoming in the near
future and will at this time have the total economic study with
recommendations as to its implementation. The proposed facility would
collect and treat all of the fisheries waste generated in the Kodiak
vicinity. Without the help of the Environmental Protection Agency and
the City of Kodiak and the fisheries processors this development could
not have occurred. Hopefully, a means has been found to eliminate
pollution at Kodiak which may be also applicable in other shellfish
processing localities.
It may develop that recoveries of these wastes can generate profits so
that a present liability can be a future asset at the same time abating
pollution.
72
-------
References
1. T. E. W. Schumann, J. Franklin Institute, .2081:405 (1929)
2. C.C. Furnas, Trans American Society of Chemical Engineers,
J24:192
3. John Spinelli and Barbara Koury, Journal of Agriculture and
Food Chemistry, 18:284 (1970)
4. Horn, M. J., Jones, D.B. and Ringel, S. J., J. Biol Chem 138;141
(1941)
5. Cuthbertson, W. R. and Philips, H., Blochem, J., 39_:7 (1945)
6. Lindleg, H. and Philips, H., Ibid, 39_:17 (1945)
7. Blackburn, S. and Lee, G. R., Biochemica et Biophysica Acta,
19:505 (1956)
8. Block, R. J., Amino Acid Composition of Proteins and Foods
Springfield, Illinois (1945)
73
-------
SALT RECLAMATION FROM FOOD PROCESSING BRINES
by
E. Lowe and E. L. Durkee*
Pickling is one of the oldest forms of food preservation known to man.
In this age of radiation sterilization, freeze concentration, microwave
drying and other forms of high technology processing, it is interesting
to note that the pickling industry in this country keeps rolling along,
producing anywhere between a billion and a billion and a half pounds of
pickled products a year. Most of this production consists of cucumber
pickles, but other foods are also pickled, including: onions, pepper,
beets, snap bean, cauliflower, tomatoes, crab apple, olives, corn,
cabbage, nuts, fruit peel, meat and fish. The Northern and Southern
states, principally Michigan, Wisconsin and North Carolina produce over
80 percent of the cucumber pack, but California ranks third among the
producing states, behind Michigan and North Carolina.
The production of pickled products involves a salting operation in which
the freshly harvested material is stored in salt brine and allowed to
ferment before further processing. In cucumber pickling, for example,
the brine initially contains about 10 percent NaCl. After 4 to 6 weeks,
the concentration is gradually increased to a final value of from 15
to 20 percent, at the rate of from a half to 1 percent per week.
Each year, the pickling industry generates up to 100 million gallons of
used brine from its salting operations. The disposal of this amount of
saline liquid waste without causing water pollution has become a problem
of increasing urgency to the industry. The corrosive nature of the sodium
chloride salt in the brine, and the fact that the waste stream is a mix-
ture of a non-biodegradable salt and organic solids, makes the disposal
problem a particularly difficult one.
In California, the olive industry has been particularly hard hit by the
demands for pollution control. In response to this situation, we have
been studying the possibility of reclaiming the process salt for reuse.
The method we have devised was designed specifically for the reclamation
of olive processing brines, but the chances are good that it will also
be effective for the treatment of brines from some of the other products
that I have mentioned.
Now olives are salted in much the same way as cucumbers. The concentra-
tion of the salt brine used depends on the variety of olive and ranges
from 3 to 5 percent initially. The concentration is gradually increased
over a period of 3 to 4 weeks to a final level of from 7 to 9 percent,
^Western Regional Research Laboratory, Agricultural Research Service,
U. S. Department of Agriculture, Berkeley, California
75
-------
again depending on the variety. Most olives are stored in brine for a
period of from 1 to 6 months.
Several methods have been suggested for reconditioning the used brine,
but because the used brine is more concentrated than the starting brine,
recovery of at least a portion of the salt as a solid is essential if
there is to be no residual waste stream. This can only be done, of
course, by evaporating the water.
Decontamination of the salt or salt solution is the other major problem
in reclaiming brine for reuse. Interestingly enough, the solution to
this problem is facilitated by the requirement that the salt be in solid
form. Figure 1 shows a salt recovery and decontamination method that
has been under study at the Western Regional Research Laboratory, USDA,
at Albany, California.
EVAPORATION BY SUBMERGED COMBUSTION
In this new process, a submerged combustion evaporator is used to crys-
tallize the salt from the used brine solution. Submerged combustion is
a method for heating liquids by passing the gaseous products of combus-
tion from a burner directly up through the liquid body. Heat is rapidly
transferred from the rising gas bubbles to the liquid through the bubble
interface. The partial pressure of the water vapor in the rising bubble
is less than one atmosphere so that boiling takes place at a temperature
considerably below the normal boiling point for a saturated NaCl solution,
227°F. The observed equilibrium temperature is about 200°F, corres-
ponding to a depressed vapor pressure of about 443 mm, and a heat of
evaporation of approximately 978 BTU/lb.
Conventional evaporators can, of course, be used to crystallize the salt,
but there are several drawbacks to this approach. For one thing, capital
costs are several times greater for a conventional evaporator than for a
submerged combustion unit. This is particularly true for an evaporator
handling a saturated NaCl solution because of the extremely corrosive
nature of the solute. Because there are no heat transfer surfaces in a
submerged combustion heater, equipment cost is comparatively low.
The absence of any heat transfer surface works in favor of the submerged
combustion heater in another way. No surface, no fouling.
The disadvantage of submerged combustion is a lower thermal efficiency
as compared to a multiply effect evaporator simply because it is not
practical to recover the heat energy in the exhaust vapor stream. Theo-
retical combustion shows a calculated thermal efficiency of 86.7 percent
for the crystallization of salt from an 8.5 percent NaCL brine solution
based on the gross heat content of natural gas, and 96.8 percent based
on the net heat content.
DECONTAMINATION BY INCINERATION AND FILTRATION
The slurry leaving the bottom of the crystallizer contains about 57.5 per-
cent solids (see Table 1) of which 6% is combustible organic matter that
76
-------
Recond.
Brine
1.026
10.0
4.0
100.0
0.9
0.4
Protein,%(Ndwbx6.38) 2.25 <.06
C.O.D.,ppm 34,700 25
Sp. gr.
PH
Solids ,%
NoCI,%(dwb)
K,%(dwb)
Ca,% (dwb)
Sulphates, % (dwb)
Feed
Brine
1.062
3.7
9.6
84.1
4.0
O.I
0.5
Salt
Slurry
4.1
57.5
91.3
1.3
O.I
0.6
Incin
Salt
100.0
98.3
2.7
O.I
0.7
Table 1. Chemical Analyses of Brines, Slurry and Salt.
-------
must be eliminated before the salt is suitable for reuse. In Figure 1,
the organic contaminants are destroyed by incineration at a temperature
of approximately 1200°F. After 5 minutes in the incinerator, the salt
is decontaminated except for a small amount of carbon residue from the
incineration. The reclaimed salt is best stored in this form from one
season to the next, thus releasing the wood storage tanks for sweetening
during the off-season by the conventional lime treatment.
Figure 2 shows the solids at various stages of treatment: the slurry
of crystallized salt, the incinerated salt and the carbon sludge which
is filtered from the reconditioned brine.
Brine for the following season's pack is prepared by dissolving the
proper amount of incinerated salt in water to produce a 3 to 5 percent
solution. HC1 is added at this point to neutralize the reconditioned
brine, which otherwise has a pH of 10.0 (see Table 1). The carbon is
separated from the salt solution by simple filtration, leaving a clear
brine (see Figure 3) free of organic matter (see Table 1). If time is
of no great importance, the salt solution can be clarified by allowing
the carbon residue to settle out by gravity. Depending on the depth of
the vessel, this might be a matter of several hours.
COST ESTIMATES
Preliminary cost estimates were made for a plant handling 250,000 gallons
of 8.5 percent brine over a period of 40 8-hr days, using natural gas as
fuel. On the basis of 100 gallons of brine per ton of fresh olives (3),
this represents a plant processing approximately 2,500 tons of fruit a
year.
The estimates indicate a reclamation cost of $42.47 per ton of salt or
$1.60 per ton of olives, which includes a first year capital cost of
$16,500 amortized over a period of ten years, and a fuel cost of $13.34
per ton of salt. This compares with fresh salt at $13.40 a ton, to which
must be added an average shipping charge of $0.43 per hundred-weight, and
a brine disposal charge of $0.20 per thousand gallons (3), for a total
cost of $22.53 per ton of salt delivered, or $0.86 per ton of fruit pro-
cessed.
For the average plant, the additional cost for a substantial contribution
to pollution control is something under $2,000 a year for the first ten
years. After that, instead of an added expense, there is a small poten-
tial savings of about $430 a year. In view of the increasing pressure on
conventional methods of liquid waste treatment, the cost of brine dis-
posal will no doubt increase sharply in the days to come, so that the
dollar difference should continue to improve in favor of salt reclamation.
It is obvious that some additional expense will be involved in reclaiming
the salt for reuse, but the small additional cost per ton of fruit is
perhaps well justified by the greater benefits that accrue to society as
a whole.
78
-------
WATER VAPOR, NON-GONDENSIBLES,
AROMATICS-^
USED BRINE
WATER-FILLED
SCRUBBER
HC*
^
WATER
MIXING
TANK
INCINERATED
SALT
SLURRY OF
CRYSTALLIZED
SALT
UNIFLOW
FILTER—•*!
INCINERATED *5 \
SALT BAGGED
FOR STORAGE
U FILTERED
BRINE
SLUDGE
Figure 1. Flowsheet Showing Process for the Reclamation
of Salt from Food Processing Brines
-------
00
o
SLURRY OF
CRYSTALLIZED SALT
INCINERATED SALT
SLUDGE FROM
FINAL FILTRATION
Figure 2. Slurries and Solids at various stages of Treatment.
^-^ - - . *-* . .
-------
i
'4
.
V
INCINERATED
SALT BRINE
RECONDj
Figure 3. NaCl Brines at Various Stagesj>f^Treatment.
81
-------
STORAGE EXPERIMENTS
Experiments are currently underway to evaluate the quality of the brine
made from salt recovered from the previous year's brine. This will be
followed by larger scale experiments involving the storage of freshly
harvested olives in salt brine reclaimed from the 1970 waste stream.
These storage studies are being conducted in cooperation with the NCA
Laboratory in Berkeley, California.
Tests are also being made with cucumber pickling brines and so far, the
results look very encouraging.
ACKNOWLEDGMENTS
The authors wish to thank E. D. Ducay for the chemical analyses of the
brine, slurry, and salt samples, and J. S. Hudson for the C.O.D. analyses,
82
-------
REFERENCES
1. National Canners Association. Investigations on the effect of
variations in processing on ripe olive flavor, NCA Report No.
D-3051, October 1968.
2. Popper, K., Camirand, W. M., Watters, G. G., Bouthilet, R. J.,
and Boyle, F. P. Recycles process brine prevents pollution,
Food Engineering 39(4): 78-80, April 1967.
3. National Canners Association. Reconditioning of food pro-
cessing brines, NCA Report No. D-2297, May 1970.
83
-------
REDUCTION OF SALT CONTENT OF FOOD PROCESSING
LIQUID WASTE EFFLUENT
by
Dr. Jack W. Rails, Walter A. Mercer, and Nab11 L. Yacoub*
INTRODUCTION
Sodium chloride brines have been used in a number of food processing
operations for many years. Brines are used for preparation of ground
meats, hams, bacon, and corned beef, for storage of fermented foods, and
for quality grading of vegetables such as green beans and peas. The
weight of sodium chloride used in 1960 by the food processing industry
was approximately 1.1 million tons (Heid and Joslyn, 1967); most of the
salt was used by the meat packing industry.
Salt used in foods as a flavoring ingredient provides the major portion
of the approximately 10 grams consumed daily by each person in the United
States. The 1000 tons of sodium chloride excreted dialy in urine by the
population of the United States is eventually deposited in receiving
waters. There is little expectation of changing the gradual increase in
salinity of receiving waters due to the ingestion and excretion of sodium
chloride during normal human physiological processes. Fortunately, there
are promising technological developments which should reduce the potential
salinity increase of receiving waters from industrial operations using
salt.
The fermented food, especially pickles and olives, are unique among food
processing operations using substantial quantities of salt because most
of the salt is separated from the final product and discarded as a liquid
waste. In many areas of the United States, salt brines from fermented
food preparation are discharged to sanitary sewers where the incremental
sodium chloride load which they contribute does not substantially increase
the total dissolved solids level in the effluent from the treatment plant.
In other areas, the total dissolved solids in treatment plant effluent
become excessive due to lack of dilution of the processing brines with
sanitary sewage or industrial wastes low in sodium chloride. One such
area, where potential saline pollution of receiving waters is of concern,
is in the Central Valley of California. Here are located a number of
olive processing plants, usually in small towns, which generate substantial
volumes of liquid waste containing sodium chloride and sodium hydroxide.
The bulk of the salt used in olive processing is contained in the storage
brines for freshly harvested olives. These relatively low volume, high
salt content, brines can be managed by ponding or by reconditioning and
reuse (Mercer, et^ a^., 1970; Lowe and Durkee, 1971). It is the large
volume, lower salt content, processing and rinsing brines which present
a more challenging waste management problem.
*Western Research Laboratory, National Canners Association, Berkeley,
California
85
-------
New processing technology such as in-the-jar fermentation of pickles and
olives (Etchells, j|t_ jil_., 1964, 1966) and salt-free storage of olives
(Vaughn, e_^ a[U , 1969) may provide solutions to part of the potential
saline pollution from pickle and olive production. In the case of olives,
it is still necessary to use sodium hydroxide to hydrolyze bitter olive
constituents, so the problem of management of large volume, low salt
content, processing waters still must be solved.
ION EXCHANGE TREATMENT OF SALINE OLIVE PROCESSING WASTEWATERS
Ion exchange is the most promising method currently available to treat
saline wastes such as olive processing waters which contain dissolved
organic compounds as well as inorganic salts. There are five ion
exchange processes which have been proposed for water desalination.
The characteristics of these processes are tabulated in Table I.
Table I
Comparison of Competitive
Ion Exchange Processes for Water Desalination
Name of
Process
Sul-bi-Sul
Desal
Sirotherm
Asahi-Grover
Aqua-Ion
TDS in
feed, ppm
100-1000
150-10,000
1000
1000
100-10,000
Operation
Scale, mgd
5
5-10
5
5
0.4
Cost
$/1000 gallons
0.29
0.13-0.78
0.25
0.25-0.35
0.13-0.17
It is clear from an examination of the information summarized in Table I
that the Aqua-Ion process has a lower cost (at a much smaller scale of
operation) than the other processes listed. The low cost at small scale
of operation is very important because the maximum output of dilute
saline waste from a single olive processing plant would probably not
exceed 500,000 gpd. Figure 1 is a photograph of the pilot unit which
was constructed by Aqua-Ion to treat up to 10,000 gpd of saline waste
under contract to NCA in an Environmental Protection Agency supported
project. Figure 2 is a schematic representation of the Aqua-Ion pilot
unit; specifications of the unit are tabulated in Table II.
The treatment consists of passage of waste, effluent over a mixed bed of
cation and anion exchange resins. The cation exchanger was in the
calcium form and was a sulfonated polystyrene resin (Duolite C-20).
The anion exchanger was in the hydroxyl form and was an aminated poly-
styrene resin (Duolite A-102-D). The polar constituents of the waste,
shown for simplicity as sodium chloride, react with the exchangers as
follows:
(cation) R2Ca + 2 Nad
(anion) 2 ROH + CaCl2
-> 2 RNa + CaCl2
-> 2 RC1 + Ca(OH)
86
-------
Photograph of the Ion-Exchange Pilot Unit.
37
-------
excess L/ME
WATER OUT
A-F =
DISTRIBUTORS
excess L/ME
SALT TO ftLTCft
J*-« WAST£ AMD
UMC IN
SALT ANQ
SOLUBLE LIME
PROCffSS
Figure 2. Schematic Representation of the Ion-Exchange Pilot Unit.
88
-------
Table II
Specifications of the Aqua-Ion Exchange
Pilot Plant
Higgins Loop Diameter: 1 ft
Height of Desalination Leg: 9 ft
Resin Volume: 17.3 cu ft
Exchanger Ration: 65 percent cationic
35 percent anionic
Resin Movement Per Cycle: 22 in.
Uni-Flow Filters: Primary - 20 hoses
Secondary - 7 hoses
Power Rating: 7 horsepower
Depending on the solute concentration, the calcium hydroxide formed
during the removal of sodium chloride will stay in solution or (if the
concentration exceeds 0.0443 N^ at 25°C) will precipitate. The precipi-
tated calcium hydroxide (and other insoluble salts) can be removed
using an Uni-Flow filter (Popper, 1970). These filters are inexpensive
and simple to use. The slurry enters the filter distributor at the top
and flows down through the individual hoses. The clear liquid passes
through the cloth and runs down the outside of the hose to a collection
point. The sludge moves along inside the hose and is discharged
periodically at the bottom.
The product of the ion exchange operation is a solution of calcium
hydroxide and organic material. Part of the organic material originally
present in the waste is converted to insoluble organo-calcium salts
which can be removed by filtration. The calcium hydroxide can be removed
from the ion exchange effluent by carbonation and filtration of the
resulting calcium carbonate or by ion exchange of the calcium for
magnesium. Formation of insoluble magnesium hydroxide to remove calcium
hydroxide is feasible in locations where either the wastewater or the
water supply contains high levels of magnesium. In locations which have high
bicarbonate hardness, the effluent from the ion exchange unit can be
blended with hard water to produce cold lime softening as shown by the
following equation:
Ca(OH)2 + Ca(HC03)2
89
-------
The resin must be regenerated to convert it into a form usable for
further sodium chloride removal. Regeneration is accomplished with a
solution or suspension of calcium hydroxide in the saline wastewater.
The regenerant effluent is saturated with calcium hydroxide and contains
the salts and part of the organic compounds originally present in the
saline wastewaters. The regenerant is recycled many times in order to
increase the sodium chloride concentration to a level which makes salt
recovery or reuse attractive eonomically. The regenerated resin is
rinsed with tap water to remove residual calcium hydroxide and is then
ready for treatment of saline wastewater. Figure 3 shows a flow
diagram of the complete ion exchange and regeneration operation.
OPERATION AND EVALUATION OF THE ION EXCHANGE UNIT
Effect of Different Influent Salt Levels on Salt Removal
The unit was tested at different influent salt levels to determine the
effect on sodium chloride removal. Runs were made with olive processing
brines having sodium chloride levels of approximately 500, 1000, 2500
and 5000 ppm. During these runs sodium chloride concentration was the
only variable held constant. A complete run comprises desalination,
regeneration, and rinsing. Four composite samples were collected from
each run and designated 1, 2, 3, and 4. These samples represented
(1) composite influent to the unit, (2) composite effluent from the ion
exchange unit (product), (3) composite regenerant influent, and (4) com-
posite regenerant effluent.
The first salt level tested was approximately 600 ppm sodium chloride
in the influent; this concentration was obtained by diluting olive
processing water. During this series of runs the deionized effluent
(product) and the regenerant influent were not filtered. Table III
tabulates the results obtained from the analysis of composite samples
collected.
These runs indicated that the sodium chloride content of olive processing
water could be reduced from approximately 600 to 145 ppm. The amount of
desalinated product obtained from each run was 30 to 140 gallons at a
flow rate of 4 to 6 gpm. The same 150 gallons of regenerant were used
for each run by recycling the effluent as influent for each successive
regeneration. The color of the influent brine was light blue when the
pH was relatively low, and reddish-brown when the pH was high. The final
effluent was usually colorless, but a few samples had a yellow color.
The color of the regenerant influent and effluent was brown.
The second salt level studied was approximately 1000 ppm as sodium
chloride. Filtration was used to remove the insoluble organo-calcium
compounds from the regenerant suspension and the solids from the product.
The usual four samples were collected from the last run on each day of
sampling by the NCA personnel. Table IV tabulates the results obtained
by analysis of these samples. The salt content was reduced to a level
90
-------
Brine
Storage Tank
Brine
Pre-treatment
Tank
Purified
Water
for
Reuse
Figure 3. Flow Diagram of the Complete Ion Exchange and
Regeneration Operation.
91
-------
Table III
Analysis of Composite Samples Collected
During the 600 ppm Salt Level Period
Run No.
600-A*
Sample
Number
1
2
3
4
7.4
10.9
11.6
11.6
NaCl,
ppm
610
0
610
605
SS,
9
128
29
47
C. O. D. ,
PPm
154
2
85
104
600-B
600-C
1
2
3
4
1
2
3
4
9.5
11.4
12.0
12.1
590
113
6370**
6960**
7
113
4244***
5868***
183
101
141
134
11.0
12.1
11.8
11.9
600
330
630
590
14
225
852
5956***
158
150
161
173
* A, B, and C are three different sets of samples collected on three
different days. Samples represent the last run on each of these days.
** These unusually high values were due to residues of hydrochloric
acid used to clean distributors.
*** These high values were due to suspended excess calcium hydroxide.
92
-------
Table IV
Analysis of Composite Samples Collected
Run No.
1000-A
1000-B
1000-C
1000-D
1000-E
During the
Sample
Number
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1000 ppm
pH_
10.6
11.5
12.0
12.0
6.9
11.0
12.0
12.0
6.8
12.0
12. 1
12. 1
7.2
12.4
12.4
12.5
7.4
12.5
12.5
12.6
Salt Level
NaCl,
ppm
1195
190
1180
1117
1190
50
1150
1165
1305
90
1340
1320
1195
305
985
1095
1170
205
995
1070
Period
SS,
ppm
60
3
30
150
2
10
20
200
2
0
0
10
16
0
6
9
19
4
17
11
C. O. D. ,
333
115
340
290
356
131
370
350
265
86
214
187
N. R. *
N.R.
N.R.
N.R.
N.R.
N.R.
N.R.
N.R.
* N. R. - Not recorded.
93
-------
of 168 ppm on the average in this series of runs. In batches 1000-B,
-C, -D, and -E hydrochloric acid was added to the olive processing water
to adjust the pH to about 7. The volume of desalinated product from an
individual run was 45 to 70 gallons at a flow rate of 3 to 4.5 gpm.
Regeneration was accomplished using recycling of the 150 gallons used
for the run 1000-A.
Several runs were completed with a salt level of approximately 2700 ppm
sodium chloride in the influent. The desalinated product averaged 155 ppm
sodium chloride content as shown in Table V. The color of the influent
brine was reddish-brown and the desalinated product was yellow. The
regenerant influent and effluent were both yellow. The volume of product
was 20 to 50 gallons at a flow rate of 3 to 4.5 gpm. Regeneration was
accomplished by recycling 150 gallons of regenerant four times.
An influent brine of approximately 6000 ppm sodium chloride content was
passed through the ion exchange unit as the fourth salt level to be
tested. The results from analysis of eight groups of samples collected
during this part of the project are tabulated in Table VI. The desalinated
product had an average salinity of 790 ppm as sodium chloride. The
volume of regenerant used for this series of runs was reduced to about
100 gallons per run and was recycled in sets of 3 or 4 runs.
Effect of pH on Sodium Chloride Removal
Under the experimental conditions used in this project, pH was found to
have 110 significant influence on salt removal. This observation can be
explained by the fact that the resin was in the calcium or hydroxyl form
or, at times, in a carbonate form. When the influent olive processing
water contacted the resin, the pH was increased to the alkaline side of
7 regardless of the influent pH level. Differences in performance due
to pH changes would be expected if the resin bed had been fully con-
verted to the sodium and chloride forms; no such condition was observed
during the project.
Effect of Chemical Oxygen Demand Level on Salt Removal
No effect on salt removal would be expected from traces of non-polar
organic compounds present in the influent brine. If neutral organic
compounds were present in the influent in large amounts, they could coat
the resin and decrease the salt removal efficiency, A more severe
problem would exist if the organic compounds in the influent were polar
in nature since they would compete for active sites on the resin with
sodium and chloride ions. This competition would reduce the desalting
efficiency of the resin bed. Neither extreme of these two situations
was experienced in this project since the COD content of the
influent did not exceed 1600 ppm. There was evidence that the salt
removal was decreased as the COD of the influent increased, but this
relationship was not rigorously established.
94
-------
Table V
Analysis of Composite Samples Collected
Run No.
2700-A
2700-B
2700-C
2700-D
2700-E
During
Sample
Number
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
the 2700
2iL
8.1
12.4
12.4
12.4
7.8
12.0
12.3
12.4
7.6
11.8
12.3
12.4
8.2
11.7
12.5
12.5
8.0
11.8
12.2
12.5
ppm Salt Level
NaCl,
ppm
2500
295
1775
1975
2750
86
2290
2500
2725
125
2450
2605
2795
160
2290
2300
2750
130
2390
2400
Period
C. O. D.
ppm
1236
735
1093
968
359
210
1258
1176
1367
512
1367
1179
1333
453
1201
1101
562
320
485
440
Ca,
ppm
35
766
470
920
24
210
660
870
24
106
412
884
12
47
590
719
12
51
283
1184
CaCO3
ppm
88
1910
1180
2290
59
520
1646
2170
59
265
1029
2205
29
118
1470
1793
30
130
706
2950
95
-------
Table VI
Analysis of Composite Samples Collected
Run No.
6000-A
6000-B
6000-C
6000-D
6000-E
6000-F
6000-G
6000-H
During
Sample
Number
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
1
2
the 6000 ppm Salt Level Period
ES_
7.0
11.9
12.0
12.1
7.3
12.3
12.4
12.5
7.0
11.9
12.1
12.1
7.1
12.0
12.2
12.2
7.7
11.2
11.7
11.8
7.4
11.7
11.8
11.8
6.7
11.7
11.7
11.8
6.9
11.9
11.9
12.0
NaCl,
ppm
6810
490
5700
4650
6710
750
5610
5610
7050
850
5720
5250
5890
840
5790
5820
4950
1150
2590
3010
5590
490
5420
5310
5290
890
5650
4690
5250
840
5720
3990
C.O. D. ,
ppm
1489
234
1251
1003
1510
602
1219
1111
1436
266
1231
1029
1143
474
1209
1067
1238
797
178
257
1163
176
1069
930
969
367
1014
731
929
346
927
546
Ca,
PP™
12
318
704
1072
12
365
754
1084
12
236
625
1108
18
212
660
2286
22
184
921
1333
12
212
707
1108
12
282
695
1120
12
212
577
1025
CaCO3
ppm
29
794
1705
2675
29
911
1882
2705
29
590
1560
2764
44
529
1646
5703
54
460
2299
3327
29
529
1764
2764
29
705
1734
2793
29
529
1441
2558
96
-------
Work was completed on the establishment of a BOD/COD correlation factor
for olive processing wastewater having sodium chloride levels of 2500
and 5000 ppm. A similar factor was established for the desalinated
product water obtained from the ion exchange treatment of these brines.
The data collected for the calculation of the correlation factors is
tabulated in Table VII. The correlation factor varied with the salt
content and the extent of treatment of the brine samples. The average
value of the BOD/COD correlation factor was 0.35.
Effect of Ion Exchange Treatment of Olive Processing Waters on COD
Level in Desalinated Product
The ion exchange treatment of olive processing water was expected to
remove ionized and ionizable organic compounds. The detailed composi-
tion of olive processing water is not known, but such compounds as acetic
acid, lactic acid, citric acid, saccharic acids, and hydrolyzed pectins
are probably present. The Aqua-Ion technology would remove these
compounds either by binding on the resin (to be released later during
regeneration) or by formation of insoluble calcium salts (removed by
filtration). Examination of the data tabulated in Tables III through
VI indicates that substantial quantities of organic materials in the
olive processing waters are removed during the ion exchange treatment.
In some cases, as much as 85 percent of the initial COD material was
removed by treatment and filtration. The effect of residual COD
materials in the desalinated product water on the reuse potential is
of importance, but was not evaluated in this project.
Effect of Different Salt Levels in Influent Brines on the Salt Removal
at Constant COD and pH
Olive processing brines having an initial sodium chloride content of
approximately 500, 1000, 2500, and 5000 ppm were passed through the
ion exchange unit. Suspended solids were eliminated from consideration
as a variable in this part of the study since the influent brine was
filtered through a Uni-Flow filter before entering the ion exchange
unit. Therefore, COD and influent pH were the only compositional
factors which were adjusted to relatively constant values. The adjust-
ment of the COD content was made by the addition of lactic acid to the
olive brine until a value of about 800 ppm was reached. The pH was
adjusted at about 7.5 by the addition of strong sodium hydroxide
solution. One set of samples was collected and analyzed for each of
the four salt levels; the results are tabulated in Table VIII. The
salt removal was approximately the same until the salt level in the
influent exceeded 2500 ppm.
97
-------
Table VII
Correlation Factor Between Five Day B. O. D.
and C. O. D. for Olive Processing Wastewaters
Waste water
NaCl,
PPm
2500
11
ii
M
it
ii
C. O.D.,
ppm
560
760
700
660
630
610
B. O. D. ,
ppm
ZOO
320
290
260
230
270
Average Value
5000
ii
u
n
u
ii
ii
n
1490
1510
1440
1140
830
1160
970
930
470
470
360
310
340
480
230
260
B.O.D.
C. O. D.
0,36
0.42
0.41
0.39
0.37
0.44
0.40
0.32
0.31
0.25
0.27
0.41
0.41
0.24
0.28
Desalinated Product
C. 0. D.,
ppm
290
340
330
320
370
310
230
600
270
470
430
180
370
350
B.O.D. ,
ppm
130
120
110
120
130
150
70
170
70
140
110
70
100
90
B. O. D.
C. O. D.
0.45
0.35
0.33
0.37
0.35
0.48
0.39
0.30
0.28
0.26
0.30
0.26
0.39
0.27
0.26
Average Value 0. 31
0.29
98
-------
Table VIII
Analysis of Composite Samples of Four Different
Salt Levels with
Approx.
Salt
Level
ppm
500
1000
2500
5000
Sample
Number
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
NaCl,
ppm
590
190
490
1100
1070
250
1150
1350
2380
210
2410
2450
5510
1150
5150
4650
C. O. D. and pH Held Constant
C. O. D. ,
ppm
805
212
29
191
958
151
45
146
797
465
91
137
830
426
1029
816
PJL
7.4
10.6
11.9
11. 6
7. 5
10.8
11. 6
11.8
7.3
10.7
11.7
11.6
6.9
12. 3
12.2
12.3
Ca
PPm
22
44
1323
542
22
33
444
737
22
44
2483
750
12
636
730
2510
CaCO3
ppm
54
108
3300
1352
54
81
1109
1839
54
108
6194
1785
29
1587
1823
6262
99
-------
Establishment of the Maximum Sodium Chloride Concentration Attainable
in the Regenerant Effluent
The maximum sodium chloride concentration possible in the regeneration
effluent is of considerable economic importance in evaluating the
overall usefulness of ion exchange treatment of food processing brines.
Ideally, both the product water and the concentrated regenerant
solution could be recycled in selected stages of the food processing
operation. If both of these objectives cannot be accomplished,
concentrating the salt present in the treated processing water in a
small volume would make further management less costly. To establish
the maximum sodium chloride concentration attainable in the regenerant
effluent, the liquid from each of a large number of resin regeneration
runs was recycled after each individual run. To determine the increase
in salt concentration in the regenerant effluent, a composite sample
was taken from each effluent and the sodium chloride content was
determined. The results of this investigation are tabulated in Table
IX. The average increase in salt content in the regenerant effluent
was approximately 40 percent (difference between original influent
and lost effluent of a recycle series), depending on the salt level
and the flow rate. It was found that the salt increase in the regen-
erant suspension occurs at a slow rate. There was not sufficient
operating experience with any single influent brine composition to
determine the maximum salt concentration attainable in the regenerant.
A test was run using a concentration of approximately 20,000 ppm sodium
chloride made by adding solid salt to an olive processing water. The
use of this solution in regeneration gave an average increase in sodium
chloride in the regenerant effluent of 595 ppm. This result indicated
that it was possible to have substantial salt increases in the regenerant
effluent even at salt levels of approximately 2 percent.
Effect of Cycle Time on Regeneration
The work on regeneration was continued using different cycle times.
Table X tabulates the sodium chloride content of regenerant effluent
and influent at various cycle times. The influent sodium chloride
concentration was 1900 to 3500 ppm in these runs and the flow rate was
4.5 gpm. The regenerant effluent was recycled. The cycle times was
calculated by dividing the gallons of influent used in a run by the flow
rate of 4.5 gpm. The maximum average salt content increase was obtained
with a cycle time of 30 minutes. The difference between the 10 and
20 minute cycle times was not significant.
100
-------
Table IX
Increase in the Sodium Chloride Content of
Regenerant Effluent as a
Run
No.
1 A
2 A
3 A
1 B
1 C
2 C
3 C
1 D
2 D
3 D
4 D
1 E
2 E
3 E
l.F
2 F
1 G
2 G
1 H
2 H
1 I
2 I
l.J
1 L
2 L
3 L
1 M
2 M
Flow
Rate,
gp™
2
2
2
2
2
2
2
2
2
2
3
3
NaCl,
INF
2070
2300
2480
2410
2590
2600
2660
2540
2630
2670
2710
2760
2790
2860
2940
2910
2920
2970
3030
2970
3000
3050
3070
924
895
866
820
1024
ppm
EFF
2180
2360
2720
2820
2910
2750
2930
2770
2890
3040
3190
3030
3140
3090
3310
3370
3220
3220
3110
2990
3070
3220
3730
2451*
708
942
1106
1103
Result of Regenerant Recycling
Run
No.
1 R
2R
3R
4R
5R
1 T
2 T
3 T
4 T
1 U
2 U
3 U
4 U
1 S
2S
3 S
4S
1 0
20
30
4O
1Q
2Q
3Q
4Q
1 K
2 K
3 K
Flow
Rate NaCl,
gpm INF
6 1790
1860
1879
1948
1983
2390
2322
2416
2468
2521
2516
2580
2565
7 2012
2011
2200
2190
8 1229
1132
1188
1429
9 2416
2562
2896
2750
3 20, 534**
20,885
20,124
ppm
EFF
1856
2086
2106
2131
2135
2516
2434
2486
2490
2633
2785
2668
2673
2173
2112
2280
2311
1280
1252
1451
1633
2808
2890
2907
2867
20,622
21,762
20,943
* High value due to cleaning of distributors with HC1.
** Salt content of the influent increased by adding NaCl.
101
-------
Table X
The Sodium Chloride Content of the Regenerant
Influent and Effluent at Variouj^Cycle Times
Cycle
Time,
min.
10
20
30
NaCl, ppm
INF
1983
2318
2306
2516
2580
2565
2321
2594
2615
2650
2714
2732
2998
3033
3117
EFF
2135
2363
2451
2785
2668
2673
2790
2878
2837
2790
2937
2948
3179
3136
3146
Cycle
Time,
min,
40
NaCl, ppm
INF
2738
2738
2750
2770
2785
2785
2790
2799
2808
2828
2843
2858
3001
3352
3472
EFF
2849
2884
2890
2972
2925
2984
2890
3177
2907
2880
3024
3010
3060
3732
3674
Continuous Operation to Develop Treatment Cost Figures
In this phase of the study the time between the various unit operations,
e.g. desalination, regeneration and rinsing was kept to a minimum. Only
20 to 30 seconds were required to pulse the resin between the runs.
The only significant interruption was the time needed for fresh brine
make-up. In preparation for continuous operation, the regeneration
chamber was washed with a solution of hydrochloric acid to remove the
organo-calcium compounds which had precipitated on the plastic beads
holding the resin above the distributor screen. This treatment resulted
in regenerant flow rates as high as 10 gpm. However, after a short
time the flow rate decreased due to the plugging of the distributors by
calcium carbonate and organo-calcium compounds. The sodium chloride
content of the influent brine during this continuous operational period
was 1000 to 1900 ppm and the flow rate was varied from 3.0 to 7.5 gpm.
Table XI tabulates the reduction in sodium chloride content of the
influent brine as the unit was operated continuously at different flow
rates.
102
-------
Table XI
Reduction in Salt Content of Influent Brine Obtained
During the Continuous Operation of the Unit at Different Flow-Rates
Flow INF EFF
Rate, NaCl, Volume, NaCl,
>m ppm gal. ppm
Flow
Run Rate,
No. epm
CA-1
Z
3
4
5
CB-1
2
3
4
CC-1
2
3
4
CD-I
2
3
4.5
4.5
4.0
4. 5
3. 5
4. 5
4.0
4.0
3.5
4. 5
4.0
4.0
•3.5
3.5
4.0
4.0
1761
1854
1878
1053
50
50
30
30
30
30
30
35
30
30
30
30
30
30
30
30
642
900
784
670
690
543
525
562
573
465
453
470
452
294
274
273
CE-1
2
3
CF-1
2
3
CG-1
2
3
CH-1
2
3
CI-1
2
3
5.0
5.0
5.0
5.0
5.0
5.0
7.5
7.0
7.5
7.0
7.0
7.0
7.5
7.0
7.0
INF EFF
NaCl, Volume, NaCl,
ppm gal. ppm
1526
1508
1740
1740
1547
30
30
30
30
30
30
652
623
578
720
641
610
30
30
35
30
30
30
30
30
30
1211
1030
860
525
544
550
1508
1110
878
Hardness in Product Water from Ion Exchange Treated Brines
The use of calcium hydroxide as a regnerant in the Aqua-Ion technology
causes this material to appear in the desalination product and results
in a hard water of limited reuse potential without additional treatment.
Calcium ions concentration was determined on a large number of the
usual set of four samples collected from individual runs under a wide
range of conditions. Table XII tabulates the calcium ion concentration
and hardness as calcium carbonate for typical samples.
DISCUSSION
Salt Removal
In general, the salt removal obtained was satisfactory. The results
demonstrated that desired product quality can be obtained at varying
levels of polar solute concentration when COD and pH are relatively
constant. At 5000 ppm sodium chloride concentration, the salt content
of the product was higher than the target value of 175 ppm. The influent
salt level was not believed to be the cause of the higher level of sodium
103
-------
Table XII
Calcium Ion Concentration and Hardness of the
Influent and Effluent of Both the Desalination
and Regeneration Processes
Sample Ca, Hardness as
Run Number Number* ppm CaCO3, ppm
H-A 1 33 83
2 835 2083
3 713 1778
4 1091 2722
H-B 1 22 56
2 668 1667
3 701 1750
4 935 2334
H-C 1 47 120
2 800 2000
3 870 1940
4 990 2470
* Samples are the same as those in previous Tables.
104
-------
chloride in the product water. Rather, the observed result could be
attributed to intermixing which took place through the resin bed and
caused displacement of the salt front. The intermixing problem was
encountered during several periods of operation of the ion exchange
unit. Near the end of the project a technique was found to minimize
the intermixing in the resin bed. On completion of a regeneration
cycle, valve No. 3 and an additional outlet valve (just prior to valve
No. 2 of Figure 2) were opened. Compressed air was used to push out
the regenerant effluent and the water filling the wash chamber. In
this way, the contact between the regenerant and the final product at
the top of the Higgins Loop was avoided. The quality of the final
product was improved substantially when this procedure was applied.
The salt content was reduced from 1500 ppm in the influent to a range
of 130 to 270 ppm in the desalination product. The unit was operated
at 4 gpm during this use of the revised procedure. It was unfortunate
that more time was not available to study all the variables examined
and reported above which were obtained under less than optimal operating
conditions.
Conventional Regeneration
Slaked lime (Ca(OH)2) performed successfully to regenerate spent resin
used to desalinate the olive processing brines. The recycling of the
regenerant effluent increased the sodium chloride content of the
regenerant. The long-term trend of sodium chloride increase was
apparent from examination of data in Table IX, although the difference
in salt content in any pair of adjacent runs did not appear to be
significant (for example Run No. 1C and Run No. 3C). This was due
to the diluting effect of the wash water filling the void space of the
resin. The void space can represent as much as 40 percent of the total
volume of the resin.
At high flow rates, when distributors tended to plug up, washing with
acid solution was required to open up the flow channels. The residues
of hydrochloric acid solution resulted in sudden increases in the
chloride ion readings for some effluent samples. The preliming and
filtration of the regnerant influent resulted in shorter regneration
times and longer operational periods with fewer acid washings being
required.
Within the time limit assigned to any given phase of the project, the
average increase in the sodium chloride concentration of the regenerant
solution was approximately 40 percent. There was no indication of a
leveling off of the rate of sodium chloride increase in the regenerant
effluent with increasing number of cycles.
The best regeneration cycle time was found to be 30 minutes.
105
-------
Carbonate Regeneration
One of the reasons that higher than expected sodium chloride levels were
observed in certain runs was because the resin was in a calcium carbonate
form rather than a calcium hydroxide form.
The equations for the reaction involved in carbonate regeneration are
the following:
(Cation) R2Ca + 2 NaCl
(Anion) RC° + CaC1
-> 2 RNa
CaClr
-> 2 RC1 + CaCO,
It was observed that salt removal was only about 50 percent of the level
removed when the resin was in the calcium hydroxide form in those runs
where the resin was partially or completely in the carbonate form. The
relatively low calcium content of the desalinated product in these runs
confirmedj^ the carbonate resin mode. The low calcium content of desalinated
product effluent shown in Table XIII was the result of removing most of
the calcium as calcium carbonate. The sludge formed from the filtration
of the desalinated product water was found to contain substantial amounts
of calcium carbonate.
Table XIII
The Effect of the Carbonate Regeneration on
Run
No.
C03 A
COS B
C03 C
C03 D
C03 E
C03 F
C03 G
the Calcium Content of the Product Water
NaCl
1590
2850
2950
1050
1150
1980
1650
INF*
Ca
43
12
12
24
24
35
33
CaCO-}
108
29
29
59
59
88
81
NaCl
690
1190
1290
430
570
590
590
EFF
Ca
54
71
47
6
12
47
43
CaC03
135
176
118
15
29
118
108
* All results are in ppm.
The introduction of carbon dioxide into the ion exchange system is the
reason for the appearance of carbonate regeneration. The carbon dioxide
could have been introduced at the following points:
a. In the compressed air used to move the resin. If this were the
source, it could be corrected by passing the air through an alkaline
solution before compression.
106
-------
b. In the lime used for the preparation of the regenerant influent.
Samples of lime were found to contain large quantities of calcium
carbonate. This could be corrected by storing the lime in tightly
closed containers.
c. In gas transfer through the cloth of uncovered filter hoses. The
plastic wrapping of the filter hose was occasionally torn by strong
winds; a combination of low temperatures and rain wetting the
hoses could promote the uptake of atmospheric carbon dioxide.
The problem of carbonate regeneration could be eliminated by operating
the unit at slower rates or by employing a larger desalination leg.
The second change would be the most economic way to correct for the
possibility of carbonate regeneration in a scaled up production unit.
The filtration of the regenerant and the desalinated product was not
possible in a few runs due to the oxidative degradation of the cotton
filter hoses. The perforation of the Uni-Flow filters on these occasions
was the reason for the high suspended solids content of some samples
and for the very high values for calcium and suspended solids in certain
of the regenerant suspensions. The perforation of filter hoses can be
forestalled by good maintenance; hoses should be replaced near the end
of their expected service life.
CONCLUSIONS
The desired quality of the desalination product was obtained under most
operating conditions. Under certain conditions, the sodium chloride
content would be considered high. However, in some areas of California
the total dissolved solids content of municipal water supply range
from 410 to 1,243 ppm (Anon., 1962).
The desalinated product is a hard water and its calcium content should
be reduced in order to increase reuse options.
The sodium chloride content of recycled regenerant solutions was
increased 40 percent over the influent brine level and evidence was
obtained that a 10 fold increase was possible. Stated in another way,
the sodium chloride present in the original olive processing water was
potentially concentrated in one-tenth of the original volume. At the
same time a volume of desalinated water equal to the volume of the
treated olive processing brine was produced for possible reuse.
The cost of desalting 1000 gallons of olive processing brine was
estimated at $0.26. The estimate was based on extrapolation of data
from the pilot plant operation. The flow rate used in the estimate
was 4 gpm which corresponds to a flow rate of 5.1 gpm/sq ft of resin
bed.
107
-------
REFERENCES
1. Anon. 1962. California Domestic Water Supplies, State of California,
Department of Public Health.
2. Etchells, J. L., R. N. Costilow, T. E. Anderson, and T. A. Bell.
1964. Pure Culture Fermentation of Brined Cucumbers. Applied
Microbiology 12/6), 523-535.
3. Etchells, J. L., A. F. Berg, I. D. Kittel, T. A. Bell, and
H. P. Fleming. 1966. Pure Culture Fermentation of Green Olives.
Applied Microbiology 14J6), 1027-1041.
4. Lowe, E. and E. L. Durkee. 1971. Salt Reclamation from Food
Processing Brines Using Submerged Combustion. Proceedings of the
Second National Symposium on Food Processing Wastes, Denver,
Colorado, March 23.
5. Heid, J. L. and M. A. Joslyn. 1967. Fundamentals of Food
Processing Operations, Chapter 4 (by M. A. Joslyn and A. Timmons),
page 79, AVI Publishing Co., Inc., Westport, Connecticut.
6. Mercer, W. A., H. J. Maagdenberg, and J. W. Rails. 1970.
Reconditioning and Reuse of Food Processing Brines, Proceedings
of the First National Symposium on Food Processing Wastes,
Portland, Oregon, pp. 281-293.
7. Popper, K. 1970. Possible Uses of Uni-Flow Filters, Proceedings
of the First National Symposium on Food Processing Wastes, Portland,
Oregon, pp. 362-376.
8. Vaughn, R. H., M. H. Martin, K. E. Stevenson, M. C. Johnson, and
V. M. Crampton. 1969. Salt-Free Storage of Olives and Other
Produce for Future Processing. Food Tech. 23/6), 832-834.
108
-------
PRODUCTION AND DISPOSAL PRACTICES FOR LIQUID
WASTES FROM CANNERY AND FREEZING FRUITS
AND VEGETABLES
by
Walter W. Rose*, Walter A. Mercer*, Allen Katsuyama*,
Richard W. Sternberg**, Glen V. Brauner**, Norman A. Olson*,
and Dr. Kenneth G. Weckel***
INTRODUCTION
The food processing industry must conform to provisions of a number of
federal and state laws and regulations to protect the environment from
various forms of degradation and pollution. The Refuse Act of 1899 re-
quires a permit for disposal of refuse in a liquid state into any navi-
gable water or its tributaries in the United States. The Air Quality
Act of 1965 and the Clean Air Amendments of 1970 cover air pollution and
control. These legislations, collectively referred to as the Clean Air
Act, include not only emissions from industrial stacks, but undesirable
odors and air pollution from exhausts of autos. The Water Quality Act
of 1905 provides for the establishment of standards by states for water
quality. The Clean Water Restoration Act of 1966 provides for grants
to jurisdictions for development of improved methods of waste treat-
ment, water purification, sewer design, and construction of waste treat-
ment plants. The Water Quality Improvement Act of 1970 implements pro-
tection against pollution, particularly by oil and hazardous materials,
and marine sanitation. The Solid Waste Disposal Act of 1965 authorized
a development for economic disposal of solid waste including trash,
garbage, paper, and scrap metal. In addition, definitive regulations
have been established in a number of states on some or all of these sub-
jects. The operations of all fruit and vegetable process plants are
subject to these acts and regulations.
The effective minimizing of the pollutional loads developed in fruit and
vegetable processing is important to both the. economic and social status
of the enterprise. It is the objective of this report to consolidate and
make available the substance of published information dealing with wastes
from fruit and vegetable processing (excluding potatoes). This informa-
tion will be used to direct attention to potentials in reducing pollution
loads through research. Much of such research will have economic advan-
tages to the operation, and will enable the processors to meet the ob-
jectives of environmental protection programs.
*Western Research Laboratory, National Canners Association, Berkeley,
California.
""^Washington Research Laboratory, National Canners Association,
Washington, D.C.
***Department of Food Science and Industries, University of Wisconsin,
Madison, Wisconsin.
109
-------
Waste Quantities
The fruit and vegetable canning and freezing industry includes operations
in 1,838 plants employing 167,000 persons, resulting in increased value
to the raw crop of some $2.2 billion. This industry utilizes an esti-
mated 99 billion gallons of intake water, recirculates about 64 percent
of the intake volume and discharges about 96 billion gallons (see other
estimates below). The percentages of these values compared to those for
all U.S. manufacturing and for all food and kindred products are, respec-
tively:
Number of Plants, 0.6 and 5.6 percent
Number of Employees, 0.9 and 10.1 percent
Value Added, 0.8 and 8.3 percent
Intake Water, O.o and 12.2 percent
Recirculated Water, 0.3 and 12.4 percent
Discharge Water, 0.7 and 12.8 percent
All of the figures are from preliminary reports of the 1967 Census of
Manufacturers, U.S. Department of Commerce, issued in 1970. An independent
estimate of the quantity of water discharged in canning and freezing fruits
and vegetables made for this study is somewhat lower than the census value:
80 billion gallons, including a very small quanitity used in dehydration
plants and a substantial quantity used in types of manufacturing excluded
from the census figures. The independent estimates are in Table 1.
Table 1 also gives estimates of raw product tonnages, and of BOD, suspended
solids, and solid residuals generated. "Other fruit" and "Other vege-
tables", includes all those not specifically listed. Estimated totals,
for the United States for a recent year, mostly 1968, are:
26.4 million tons of raw product
80 billion gallons of wastewater discharged
760 million pounds of BOD generated
360 million pounds of SS generated
8.9 million tons of solid residuals
Citrus, tomatoes, corn and white potatoes (excluding dehydrated potatoes)
account for 67 percent of the raw tonnage, 61 percent of the waste water,
45 percent of the BOD, 61 percent of the suspended solids, and 71 percent
of the solid residuals.
The raw tonnage estimates are believed to be the most precise. They are
mostly from the U.S. Department of Agriculture urop Reporting Service, but
110
-------
TABLE I
Wastes From Canned and Frozen Fruits and Vegetables
1
Raw
Tons
,000
Susp.
Waste Water
103
tons
apple
apricot
cherry
citrus
peach
pears
pineapple
other fruit
1
7
1
1
,000
120
190
,800
,100
400
,000
400
5
5
2
3
4
4
8
gal/
ton
.0
.0
.0
rail.
gal .
5,000
600
400
.0 23,000
.0
.0
.5
.0
4,400
1,600
500
3,200
BOD
Ibs/
ton
40
60
20
4
60
70
20
20
mil.
Ibs.
40
7
4
31
66
28
20
8
Solids
Ibs./
ton
5
12
6
7
10
20
8
10
mil .
Ibs.
5
1
1
55
11
8
8
4
Solid
Residuals
Ibs./
ton
600
360
300
880
500
660
900
1,000
tons
320
21
27
3,390
270
120
450
80
Sub-Total
Sub-Total
Total Above
white potato
TOTAL
12,250
38,700
200
90
4,680
asparagus
beans, lima
beans , snap
beets
carrots
corn
peas
pump., squ.
sauerkraut
spinach, gr.
sw. potato
tomato
other veg.
120
120
630
270
280
2,500
580
220
230
240
150
5,000
1,300
10.
9.
4.
4.
4.
1.
5.
3.
.
9.
7.
2.
4.
0
0
5
0
0
8
0
0
5
0
0
0
0
1
1
2
1
1
4
2
2
1
10
5
,200
,100
,800
,100
,100
,500
,900
700
100
,200
,000
,000
,200
10
25
30
150
55
25
50
80
15
25
200
12
60
1
3
19
40
15
62
29
18
3
6
30
60
80
7
80
4
50
40
10
10
15
3
10
80
4
30
1
10
3
14
11
25
6
3
1
2
12
20
40
720
280
420
820
960
1,340 1
260
1,240
640
320
(in other
200
--
45
lu
130
110
150
,660
78
150
75
37
veg.)
400
500
11,700
24,000
2,400
26,400
4.0
34,000
73,000
9,600 80
82,000
370
570
190
760
150
240
50 120
360
3,350
8,030
760 910
8,940
111
-------
some are National Canners Association estimates based on canned and
frozen pack statistics. They are for 1968 except that tomato tonnage
was reduced from 7 million to 5 million tons because 1967 was abnor-
mally high for canning tomatoes processed. The other estimates are
based on averages of widely varying figures per ton, mostly published
in the 1960's, but some earlier; a few of the sources are unpublished
data of the National Canners Association. The varying estimates for
a given product are partly the result of real plant to plant differ-
ences. Data from many of the references were converted from values
per case to values per ton using U.S. Department of Agriculture fig-
ures for average cases per ton. Some BOD data were estimated from
COD Figures. Missing data, in particular those for "other fruit",
and "other vegetables" were estimated by comparison with data for
the principal products. The second digit in most numbers in the
table is highly questionable and the first digit is generally in-
secure .
A 1969 report by, Secretary of Agriculture and Director of the Office
of Science and Technology, gave the following estimates for canning
and freezing fruits and vegetables:
1963 Estimate: 71 billion gallons wastewater; 660
million pounds BOD
1972 Projection: 93.5 billion gallons wastewater;
845 pounds BOD
Powers, et al. (1967) estimated that 87 billion gallons of water were
discharged from canned and frozen fruits and vegetables in 1964; and
for 1963 the same quantity of wastewater and also 1,190 million pounds
of BOD and 600 million pounds of suspended solids generated from all
canned and frozen foods. These figures are somewhat higher than the
current study estimates and the additional products included in the
earlier estimates do not seem to account for the differences in BOD
and SS.
Other estimates derived from Powers, et al. (1967) indicate that in
1963 all food and kindred products manufacturing was about 1/10 and
the canned and frozen fruits and vegetables industry was about 1/100
of all U.S. manufacturing as measured by value added. These two seg-
ments of the economy used about 5.4 and .5 percent, respectively, of
the total water, but produced about 20 percent and 5 percent, respec-
tively, of the total BOD. These comparisons reflect the relatively
high strength wastewater discharged by food manufacturing.
The U.S. Department of Interior, FWPCA, publication, Industrial Waste
Profiles, No. 6. (1967) estimated for canned and frozen fruits and
vegetables the following total waste loads, million pounds:
1963 1968 1972 1977
660 785 845 905
750 890 960 1035
710 845 910 980
112
-------
Estimates were made for the current survey of total solids in the waste-
water generated by the industry. They were based on sparser data than
were available for the other items and are, therefore, not listed in
detail. The estimated total was 2.4 billion pounds of total solids,
a very high proportion coming from potato processing.
Effluent Treatment
All of the figures on wastes referred to above pertain to generated wastes
Large proportions are reduced by treatment before final discharge.
The 1963 Census of Manufacturers is the source of the following totals
and percentages of wastewater flows from plants reporting 20 million
gallons per year or more:
Percent To:
Total Flow, Public Surface
Billion Gallons Treatment Ground Water
All U.S. Manufac-
turing 13,200 7 1+ 90
Food and Kindred
Products 688 35 11 51
Canned and Frozen
Fruits & Vegetables 66 38 17 42
The small quantities unaccounted for are transferred to other uses. The
food industries discharged much higher proportions of their liquid wastes
to public sewers and to ground than did manufacturers as a whole. The
degree of purification at public treatment plants varies widely; estim-
ates of their average removals from the U.S. Department of Interior,
FWPCA publication, Industrial Waste Profiles, No. 6. (1967) reports:
75 percent of the BOD, 85 percent of the SS and 14 percent of the TDS.
Discharges to ground are principally by irrigation, mostly spray ir-
rigation, but also by seepage from ponds and by pumping into non-produc-
tive wells. Ground disposal generally removes very high percentages of
the pollutionai load. The high degree of utilization of public plants
and ground discharge by food processors is partly explained by the need
for reducing the relatively high strength °f their wastes.
Most industrial wastewater (including that from food processing) is used
for cooling or for other relatively non-contaminating purposes. Added
heat may be a problem, but not other types of pollution. Discharging
this more or less clean water to surface water is necessary to main-
tain stream flows and provide water for down-stream populations and
industries. Wastewater disposed of in streams and lakes is often treated
before discharge.
A 1965 study by Bower found the following distribution of liquid wastes
113
-------
disposal practices in 80 fruit and vegetable canneries, some of which
were also freezers:
Discharge To:
City Systems
Ponds
Percent of Plants:
Non-Urban
9
28
Spray
Irrigation
Surface Water
0
0
55
9
31
5
Urban plants were defined as those located within cities; non-urban plants,
those located in the country, in small towns or on the outskirts of
cities. The contrast in disposal methods by the two groups reflects the
availability of treatment plants in cities and the high cost of city land.
Table II, mostly 1969 - 1970 data from a preliminary summary of a solid
waste survey of canning and freezing by the National Canners Associa-
tion shows the distribution of liquid waste disposal methods found in
a study of all canned and frozen fruits and vegetables except pineapple.
The same plant might use more than one of the listed methods; i.e., a
plant putting up more than one product was tallied under each of them.
"Holding" and "treatment" ponds were not strictly defined; generally
the removal of pollution would be less in a holding than in a treatment
pond. Some methods of treatment at the plant or of disposal were not
included in the summary. For example, about half the citrus plants used
additional methods, and the clarifiers which remove settleable material
at most potato plants were omitted. Most reliance on city systems was
by tomato, peach, pear, and miscellaneous fruit plants; on treatment
ponds by potato and apple plants; and on irrigation, by corn, apple and
pea plants.
TABLE II
Product
Citrus
Tomato
Corn
Potato
Peach
Apple
Snapbean
Pea
Pear
Other Fruit
Other
Vegetables
Percent of Plants Using;
City Holding
Systems Ponds
12
67
40
43
83
30
58
39
92
67
59
_-*
11
12
14
3
10
6
13
4
10
10
Treatment
Ponds
10
19
28
57
11
40
24
10
8
16
23
Irrigation
24
13
44
21
11
30
27
36
8
20
17
*combined percentage for holding and treatment ponds is 10.
114
-------
Industry Waste Costs
Preliminary data from the National Canners Association Solid Waste Survey,
supplemented by information from Bower (1965), Boyle, et al. (1968), U.S.
Department of Interior, FWPCA (1970), National Canners Association (1959),
and the U.S. Dept. of Interior, FWPCA (1969), are the basis of the
following estimated expenditures for an assumed average fruit and veg-
etable processing plant:
Cost Data, Liquid Waste
Item Cost $ Percentage
Item Capital O&M of.plants
In-plant $14,000 1,000 100
Simple treatment (1) 12,000 4,000 7
Further treatment (2) 44,000 13,000 20
Irrigation 24,000 4,500 20
City treatment 6,000 56
(1) Simple treatment includes holding ponds, primary treatment, etc.
(2) Further treatment includes aerated lagoons, secondary treatment, etc.
The characteristics of an "average" plant were estimated from data on
apple, peach, pear, snapbean, corn, pea and tomato processing. The
following plant size and waste averages were assumed: 15,000 tons of
raw product in a 100 day season; 3,000 gal./ton, or about 450,000 gpd,
or 45 million gal./year; 30 Ibs./ton of BOD, approximating 1,000 ppm
BOD; 10 Ibs./ton of SS, approximating 400 ppm. Component costs from the
cited references were estimated for a plant with these "average"
characteristics; they were converted to 1970 dollars using data in U.S.
Department of Interior, FWPCA, 1970. Some estimates were averages of
figures that varied over a wide range.
"In-plant" includes waste flumes, piping, pumps, tanks, sumps, gutters
and screens. Primary and secondary treatment systems were not well or
consistently defined in the references.
Using unit costs per year (and the assumed BOD removals given in the
following table) overall costs for different treatment efficiencies were
estimated for an "average" plant; the in-plant costs were added to each
of the other systems' costs (figures rounded):
BOD, Percent
Treatment Removed Cost Per Year
In-Plant 5 $ 2,400
Primary 50 7,600
Secondary 80 19,000
Irrigation 100 9,300
City treatment 75 8,400
115
-------
Suspended solids would be removed in higher percentages than BOD by the
simpler systems.
Estimates derived from figures in a report by The California State Water
Resources Control Board, Publication No. 39. (1968), for a .5 mgd plant
are as follows (converted to 1970 costs and rounded):
BOD percent Cost Per Year
Removed Capital O&M
0-10 $1,200 5,000
10 4,200 7,500
50 4,700 9,000
80 8,700 11,000
100 4,000 12,000
In-plant costs included those for screening and for handling solid wastes;
these in-plant costs were estimated to be required in all plants and have
been added to the costs of the other systems.
Waste Problems
While the figures in the foregoing seem to indicate that the water usage
and pollution loads from fruit and vegetable process operations are but
a small part of the total national industrial usage and load, the signifi-
cance is actually considerably greater. During the past 20 years, there
has been a constant consolidation of smaller operations into larger more
centralized process operations, resulting in greater usage of water and
more discharge of wastes per operation. Thus, during the highly seasonal
periods of operation in the industry it is not unusual for a process opera-
tion to utilize much more water and to generate more waste than the com-
munity in which the operation is located. The pollution load from a
plant may be 200 - 300 times that in the community of 500 - 1,000 persons
in which the plant is located; or the food waste may be equivalent to
twice that of the wastes of a city of 50,000 in which the plant is lo-
cated. Thus, the quantities of the solid wastes, which may be 7 - 70
tons per 10 thousand cases of product, or of the pollutional load in the
process effluent involve treatment processes of considerable magnitude
involving large capital and operational expenses. Of significance is
that the waste loads in this industry are generated within a relatively
small harvest period in the year, and consequently treatment systems must
be geared to prevent pollution of outlets at periods when rainfall and flow
are at their minima. Further, where the wastes are channeled into -municipal
systems, often these are already overtaxed in capacity and inadequate for
the community requirements.
The solid wastes produced in processing many fruits and vegetables have
relatively little economic value and are not marketable. Disposal some-
times' consists pf distribution to farmers in the immediate area of the
plant who incorporate it as part of the feed for cattle, swine, or
116
-------
poultry, or disperse it upon the land. However, "waste" material from
some of the products with the largest quantities of solid residuals
are specifically handled for stock feed. Examples are citrus, corn,
and potatoes, all three produced in large tonnages for processing and
in areas where cattle or other stock are available for feeding. The
National Canners Association 1970 Solid Waste Survey indicated that
as a result of this use of these three products, about three-fourths
of all residuals from fruit and vegetable processing are used as by-
products. Solid wastes may range from 5 to 70 percent of the raw
harvested crop and their handling and disposal create pollution and
hygiene problems.
Technological Changes in Food Processing
Several persons active in food processing research, engineering and teach-
ing were asked to predict changes in the technology in the next several
years and the effects of the changes on wastes and pollution. All types
of pollution (solid, liquid and air) were to be included in the predic-
tions. All of the experts responded. They expected changes in all of
the 23 agricultural, harvesting, transporting, processing and waste
handling steps that were specifically listed in the query, and they
generally expected the changes to improve the waste and pollution sit-
uation. A summary of specific comments follows:
Agriculture; New varieties should permit better utilization and, there-
fore, less waste. One expert suggested a gain of 1 percent; another a
gain of 50 percent. Some thought new varieties would not effect wastes.
About half the experts believe fertilizer applications and irrigation
will be better controlled with less waste and pollution, a few antici-
pated worse effects from increased applications. Most thought that pes-
ticide pollution would be decreased by improved materials and methods
and by better controls. Increased use of other chemicals was expected
to increase problems by two respondents, but a majority expected no
change or an improvement.
Harvest and Transport; Opinions were divided on the effects of increased
mechanical harvest; sorting in the field and once-over harvesting could
reduce problems, but more soil and trash could increase them. Similarly,
changes in containers and transportation might improve or worsen the
problems, although overall they were expected to be improved. Most
cited bulk handling or hauling in water; some expected more and some
less product damage, a factor likely to vary with the product and the
hauling method. Most thought that increased sorting and pulping in the
field would alleviate waste problems, but one expert said less control
could be exercised there.
Processing: Nearly all of the many cited examples of changes in in-plant
equipment and methods were expected to reduce waste and pollution. There
was a small minority pessimistic about product packaging, including a
suggestion of increased specialized packaging, but some thought more
functional, more reusable or more destructible packages would come into
use. Listed in-plant improvements included: less water usage, more water
recirculation, more efficient washing, improved blanching and peeling
117
-------
methods, product transportation methods using no or less water, and more
efficient product utilization. The experts expected changes in quality
standards and split on whether they would increase or decrease waste
and pollution.
Waste Handling; Improvements in this field were anticipated by nearly
all the experts. Cited here were better solid-liquid separation methods,
improved or more widely used liquid waste treatment, and the development
of more by-products, although one thought the last unlikely. Most ex-
pected better solid waste disposal methods, but one person was pessimistic
about this. Pollution regulations were expected to be stricter.
DISCUSSION
Characterization of Wastes
The processing of fruits and vegetables, i.e., the conversion of the raw
form into a processed (canned or frozen) form for use at some future
time and other place, involves application of many treatments. In the
processing some losses of organic and inorganic material occur, including
separation of soil and extraneous material, peel, seed, core, cob, fiber,
fines, spillage and solubles. These wastes are collected as discrete
materials and liquid effluent, the latter containing colloidal, sus-
pended, discrete and soluble solids.
The discrete wastes are generally removed physically; the pollution load
of the liquid effluent is dependent upon the concentrations of various
components. The capability of reducing the pollution load in the
effluent depends largely on knowing the point of origin of the wastes
generated during processing, the causes for their occurrence, and their
magnitude and characterization. With such information, means to correct
malfunction or unsatisfactory operation may be applied, or efforts made
to develop alternate systems which minimize the loss of extracted solids
or leachate. More importantly, efforts toward reduction in pollution
load can be directed to those phases of operation where generation of
pollution load is relatively great. Complementary to identifying the
important sources of pollution load in the effluent is the application of
monitoring systems by which these operations can be better defined.
Relatively few studies have been made on the magnitude and characteriza-
tion of the wastes generated during processing; these include studies
on apples, citrus, peach, pear, beets, carrots, corn, peas, snapbeans
and tomatoes. Additional information is needed on the characterization
of the wastes developed in processing these crops and others. Examples
of the potential benefits of such studies have been the development of
fruit peelers which are more efficient in separation of peel and skin
with less loss of solubles (apple, pear, tomato and vegetable peeler,
beet, carrot and potato); characterization of the losses of suspended
and soluble materials during water blanching; improved container fillers;
and isolation of waste streams for conditioning and reuse. Efforts in
the characterization of in-plant wastes should be greatly intensified,
so that alternate processes may be developed. The concentration of
118
-------
pollution loads in effluents from water blanchers makes urgent the de-
velopment and application of other systems of blanching in which there
is less loss of solids and conservation of product.
Water Conservation
A major problem in the conservation of water in many food process opera-
tions is that it is considered the cheapest commodity available, and
hence, more readily expendable in comparison with investment in sophisti-
cated equipment and in labor. Efforts to encourage conservation of water
have not been wholly successful except where it is critical in supply or
where economies are apparent in its conservation.
The conservation of water in processing fruits and vegetables is impor-
tant for several reasons; the volume of water used is directly related
to volume of effluent discharged, which in turn is directly related to
capital and treatment costs for the effluent; water is a vector and
extractant for solubles and suspended solids, either from useable product
or waste product; it is a vector for heat, which is relatively costly;
and finally, it is costly per se, contrary to the popular evaluation that
it is not.
Little data are available on the costs of water used in the processing of
fruits and vegetables, including not only the power and capital costs in
making it available from wells, but also the costs of its distribution
through the factory and of its ultimate treatment; little information
is available on the leaching effects resulting in greater pollution loads
from the excessive uses of water.
Studies should be undertaken to establish norms in the use of water for
processing various fruit and vegetable products to establish means of
conservation where such will be effective in reducing pollution loads,
and to develop better understanding by management of costs of water.
Among subjects in need of study are the following:
1. Develop quantitative data on the loss of soil with various root crops
during mechanical harvesting with the objective of focusing attention on
the magnitude of the problem and the necessity for engineering improve-
ment in harvest machinery. (Mercer, 1967).
2. Develop experimental data on new concepts for field site wash stations
to remove and recover soil from root and other crops, and on the related
transport costs and crop quality. (Mercer, 1967).
3. Develop quantitative data on water usage and costs, including energy
and capital costs in water procurement and in its conveyance throughout
the plant.
4. Develop energy balances in water usage throughout the processes,
including water transport, to evaluate better the costs of water uses.
5. Develop data on the costs of waste effluent disposal in terms of
actual (uncontrolled) water usage and conservative (controlled) water usage.
119
-------
6. Develop experimental data on new concepts for cleaning raw product,
including soak and hydraulic systems, cavitation, detergent wash, air
stream separation, mechanical abrasion, and osmotically balanced
washing. (Holmquist et. al., 1954; and Gould, et al., 1959).
7. Develop improved and new procedures for treating water to permit
greater multiple use for fruit and vegetable processes, and particu-
larly to compare economic values in the use of fresh and treated water.
In this review consideration has been given to the use of water in pro-
curement, transfer and processing of fruits and vegetables. There are
many possible ideas for improvement in the procedures used. Perhaps
the greatest need is for solid factual information that will show:
1. The nature and magnitude of each of the problems.
2. What the economic advantages would be from employment of improved,
feasible procedures for conservation of water in the total process.
These are collective and interrelated problems. The voluntary employ-
ment of the conservation procedures can best be implemented by estab-
lishing economic benefits. Regulatory action, likely, is not feasible
without such information. The current emphasis on pollution control
has stimulated interest by processors in water conservation. It has
given impetus to manufacturers to make improvements in equipment and
in the application of water in fruit and vegetable procurement and pro-
cessing. This interest on the part of food processors in water con-
servation needs to be supported by a document which describes the
economic benefits of changing water use practices in food plants.
In principal, there are several major areas in which research should
be done to achieve these aims:
1. Establish research to determine the feasibility of use of water at
site of harvest to:
A. Pre-wash the crop on the site.
B. Conserve the soil.
C. Reduce transport of non-edible material.
D. Facilitate transport of crop from field to plant.
E. Establish economic data on these potentials.
2. Undertake research on water reuse and water recovery, with major
emphasis on the economic implications of water conservation and recovery
in the broadest perspective.
Reconditioning Water for Reuse
In addition to conservation of water in processing fruits and vegetables,
120
-------
there is developing a great need for reconditioning and reuse of water
for certain phases of the operations. This need is the result of de-
creasing availability of water and changes in water quality. A limited
number of studies have been conducted, principally in processing of
tomatoes, indicating feasibility in the reuse or recycling of water in
certain steps of the processes. Clean and sterile retort water, for
example, may be directed upstream; product rinse water may be directed
upstream to product washing treatments. A number of procedures are
available for reconditioning water: ion exchange, reverse osmosis,
filtration, sedimentation, flocculation, and centrifugation. These
have been given relatively little application to reconditioning water
for reuse in fruit and vegetable processing primarily because water
generally is readily available and because the costs of water in
processing are not known. In some instances, however, the supply of
water is already critical and reuse is essential.
Application has been made in conditioning water for use because
its quality is unacceptable as a medium for foods; conditioning has been
employed to correct effluent waste otherwise unacceptable as waste. The
potentials of reconditioning water, however, have not been adequately
determined in order to more effectively stress potentials in conservation.
Studies are needed on the application of various systems of treatment of
water from various in-process streams and on the determination of their
efficacy and economic status. Evaluation must be made also of the
hygienic acceptability of the conditioned waters, since this is an
important criterion for foods. These studies will further direct
attention to means for conserving water.
Blanching
The blanching of vegetables (and some fruits) is done to achieve
certain properties considered essential for canning, freezing and
drying. A major portion of the vegetable crop is blanched in water;
the balance in steam. The blanching process has three significant
effects on operations: It generates a significant pollution load; it
causes losses of nutrients; it causes reductions in yield. These are
of considerable economic importance. The processes used for blanching
have not been adequately investigated to permit maximum correction of
the adverse effects. It is apparent that suitable determinative end
points are not available to permit application of only the necessary
degrees of blanching (excessive blanching appears to be the rule), and
possible alternative systems have not been adequately investigated. Among
these may be cited the sequential timing of liquid and liquid/steam
thermal processes; use of osmotically balanced blanching media, hot air,
infra-red, micro-wave; and sequences of these procedures. The processes
should be updated for crops now harvested, transported and handled by the
more rapid mechanical procedures.
The relatively low volumes of effluents from water or steam blanching
procedures have high concentrations of pollution load. These effluents
are separable from the total waste stream and can be subjected to treat-
ments such as concentration and drying or removal otherwise, thereby
reducing significantly pollution loads in the total process waste flow.
121
-------
Studies are needed to correct the overall impact of blanching on pollution
loads in vegetable operations and to establish potential operational
parameters in this process.
Solid Wastes
The removal of peel from fruits and skins from root crops creates sig-
nificant organic loads in the effluent streams. The discrete portions
of these wastes are separable and removable. The usual procedures of
separation follow thermal pre-treatments, which result in high con-
centrations of solubles or suspended material in the effluents.
Use of peel waste has been achieved to some extent; for example, in
citrus, pineapple and apple waste. For many peel wastes, however,
there appears to be no market, and their disposal is an operational
burden. At least two phases of research are needed to reduce the pol-
lution loads involved in peeling fruit and vegetable products. The
first is investigation and development of alternate systems of solid
waste separation to reduce concurrent losses in solubles and suspended
solids and to increase product recovery. The second is to characterize
and evaluate the material for recovery of potentially useful materials
or applications. The wastes generally are sources of leachate adding
to pollution loads and creating hygiene problems. Traditional proced-
ures of disposal of the separated wastes include use as feed for animal
or fowl and on land. The possible utilization of these wastes has not
been given adequate consideration in the light of modern food process
operations. There are techniques of conversion of the wastes into
alternate physical forms for use or disposal; there are potential
extractants that may have market value; there are potential markets
for nutrients, chemicals or fertilizers. There is potential in the
consolidation of wastes from adjacent processing plants for economics
in conversion and marketing. The quantities of such wastes can be
extensive, ranging from 5 to 70 percent of the raw product, and add
significant stress to operations and to pollution control.
The recently developed dry caustic peeling of certain root crops appears
to have benefits in minimizing the problem of solid wastes, and the
concept should be extended. Other procedures, including comminution,
concentration and incineration should be evaluated to provide better
disposal systems and reduce pollution loads in effluents.
Because the soil and peel loads in certain crops are large, considera-
tion should be given to in-field washing and peeling during harvest;
this would retain both soil and organic peel materials on the land. Al-
though there appears to be objection to such concepts on hygiene grounds,
these may be less serious than the problems in handling waste effluents.
There are several considerations to keep in mind in evaluating problems
and procedures in the utilization of cannery and freezer wastes:
1. Determine where the greatest pollution load"problems occur in the
handling of fruit and vegetable wastes. This involves determining
122
-------
the volume and composition of the wastes and measuring their pollution
strengths and magnitudes. This in turn would permit an evaluation of:
A. Potentials of possible treatments for the waste.
B. Potentials for use of the waste in light of current
market practices, whether as feed or industrial
derivatives.
C. Potentials for isolation, reduction or retention
of the waste at the point of harvest.
D. Evaluation of possible methods for disposal of
waste to the soil.
E. Determination of methods for reduction of discrete
wastes within the process lines and reduction of
leachates in all transport phenomena.
2. Investigate economic feasibilities in handling wastes consolidated
from several plants where potentials for agricultural or industrial
derivatives exist.
3. Investigate possible markets for new industrial products in light
of changing needs: fertilizers, mulches and soil conditioners,
chemical and nutritional derivatives, feed-stuffs and feed derivatives.
4. Consider establishing centers for research in solid waste utilization
where industry may collectively support investigations and potentials in
waste utilization.
Liquid Waste Treatment
There is a lack of information on the relative efficacy of simple screen-
ing and separation systems for the removal of colloidal and dispersed
matter from effluents. There is potential application of such proced-
ures at various stages in the processing of fruits and vegetables to
remove the suspended and finer materials from the effluent streams.
This will help reduce the pollution load in the plant effluent, and
permit separation of water for treatment and reuse. There are only
limited published data on the use of screens, vacuum systems, centri-
fugal systems and sedimentation systems for the treatment of selected
effluent streams.
There is considerable engineering information on the disposal of effluent
wastes upon soil. Soil type, pollution load, application rate, soil
cover, and other conditions vary widely. The results have depended
upon the experience of the participants in the projects. There is a
dearth of information on the potential effects of the applications on
the character of the soil from chemical interactions. It is evident
that in many operations the effluent loads are being applied at maxi-
mum levels, resulting in some degree of run-off and development of
123
-------
undesirable odors. A compendium of engineering information for land
disposal systems is needed.
The disposal of effluents in lagoon systems is widely practiced. The
lagoons serve not only for biological degradation of effluent, but for
storage as well. Commonly, lagoon operations depend on the natural
flora, the nutrients available supplemented by nitrogen and aeration.
There is a great need for information on establishing the optimum con-
ditions for the biological degradation of the various types of effluents
to achieve maximum effect in the shortest time. The information should
include flora, temperature, concentration, pH, nutrient supplement and
oxygen tension necessary to achieve maximum degradation. This infor-
mation is needed to achieve efficiency in lagoon utilization and to
minimize the hazards, which may occur when the lagoons are operated
improperly. Lagoons are extensively used and are important tools in
the treatment of wastes; the procedure is replete with problems in the
light of increasing social and legal pressures.
The processing of effluents by systems which employ yeasts and fungi
imperfecti and by trickling filters should be further examined to de-
termine potentials in sequential treatment. This information is needed
for the different types of wastes derived in processing fruits and
vegetables to assist engineers in designing treatment systems.
Waste Pollution Evaluation
The concentration of pollution loads in most fruit and vegetable process
wastes are measured by other than simple tests,- such as Chemical
Oxygen Demand (COD), or Biochemical Oxygen Demand (SOD). These are
not simple in the sense that for reliability in the values obtained
trained personnel must supervise their application.
The COD test is the more quickly performed (several hours) and the BOD,.
test requires five days, each after procedures have been standardized.
These tests are used as the basis of acceptability of effluents dis-
posed into natural waterways and in part to determine sewer charges.
A number of laboratory tests and instruments have been devised for
quantifying the properties of waters and wastes. These range from
sophisticated instruments for analysis for various components in water
or wastes, to gross or proximate measurement of floe or precipitate.
Many procedures have been applied in technical studies to characterize
wastes and their degradation. Instrument manufacturers offer simplified
forms of some tests to permit routine evaluation of some qualities of
waters and wastes.
For routine surveillance by fruit and vegetable processors of the pol-
lution loads in various steps of processing, there is a great need for
at least two sets of evaluation procedures. The first is the need for
data on the characterization of pollution load equivalent wastes of
specific product processes as measured or interpreted by regular
employees. Among these may be cited values obtained through use of
laboratory instruments, such as turbidimeters, refractometers, centri-
fuges, conductivity meters, oxygen concentration meters, pH meters,etc.
124
-------
The establishing of values from such instrument readings is highly
desirable and presumably would be effective for supervision of pol-
lution control. The second is the need for simplified "quick and
dirty" procedures for measuring with reasonable reliability directly
or indirectly COD equivalents of waste effluent streams in various
stages of processing of various commodities. These should permit
managers to monitor waste streams and treatments better than at
present and should lead to better control of pollution problems.
Waste Treatment Management and Economics
The costs of liquid waste treatment and disposal are not available in
a form which can be of guidance to individuals in management who must
make final investment decisions. A Board of Directors must have in-
formation on available alternatives in meeting requirements for adequate
treatment and disposal of wastes: In capital requirements, in reliability
of results from the capital to be invested, in the cost effects on the
product, and in competitive advantages in possible solutions to the
problem.
Although much information is available on liquid waste treatment and
disposal, it is nowhere consolidated in a manner which permits specific
conclusions backed by authority, acceptable for making financial de-
cisions for operations at a particular plant.
Management in some plants needs information which will enable it to
meet minimum requirements in pollution control immediately; in others,
action will be committed in a few years. There is, then, an immediate
need for an informational compendium to aid management to act on its
urgent problems. The form should allow expansion without major instal-
lational revisions.
The compendium should include, among others, the following:
1. A listing of the laws and ordinances affecting most processors,
with a summary covering major restrictions.
2. A procedure for determining the cost of water supply, water
handling and distribution, including its handling as waste and its
recovery. It should include graphs for determining water costs.
3. Procedures for evaluating the effectiveness of various methods of
screening of wastes to remove colloidal, suspended and smaller parti-
culate matter; their relative merit for specific applications, and
capital and operational costs.
4. Procedures for determining costs of municipal sewage service, in-
cluding specific situation information with factor data for application
to operations where the service is available.
5. Procedures for evaluating land disposal systems, including engineer-
ing factors, limitations, operational capabilities, capital and opera-
tional costs.
125
-------
6. Procedures for determining costs of solid waste removal, utilization
and disposal.
7. Recommendations for monitoring systems or procedures for control
of pollution loads in wastes and effluents.
The compendium should include, first, a section which would be a con-
cise summary of the problems facing a processing plant, relating
these to economic and environmental problems and current external
controls on how waste may be treated and discharged from the property.
It should include evaluation of trends in enforcement by such external
controls.
A second section should include a summary of current waste disposal
possibilities expressed in costs in graphic form for application by
management to mixes of suspended and soluble materials in the wastes.
The graphs could relate the following:
1. Factors on the. required work for different mixes of dissolved and
suspended solids plus nutrients.
2. Relation between any given factor and investment and operation costs.
3. Relations between factors for various mixes and for different pro-
duct wastes.
4. Relations between investment and operating costs against variable
production levels for available methods of treatment, including mun-
icipal, land and biological degradation systems.
In order to implement the information useful to management in the
treatment of wastes, there must be provided additional information
about their character and further evaluation of systems for their
treatment.
The relationship of the composition of the effluent to its biodegrad-
ability is not clearly established; for example, the rates of degra-
dation of effluents bearing different types of organic solids, such
as different types of vegetable starches, or of different types and
concentrations of fruit acids.
The data should be implemented by information regarding measures to
attain complete and rapid biodegradation of the wastes to the point of
enabling economic treatment and recycling/reuse of the water. New en-
gineering concepts may be feasible in the use of waste energy from
stack gases or thermal processes applied to specific wastes or to
biodegrading systems. Some information of this type has been developed.
The efforts at treatment of wastes with unicellular microorganisms
indicate the need for better understanding of the requirements to
attain greater rates and more complete degradation.
126
-------
REFERENCES
1. Bower, B. T., 1965. Unpublished data.
2. Boyle, W. E. and L. B. Polkowski, 1968. Observations of aerated
lagoons for the treatment of vegetable cannery wastes. Sponsored
by Wisconsin Canners and Freezers Assoc.
3. California State Water Resources Control Board, 1968. Cannery
waste treatment, utilization and disposal. Publication No. 39.
4. National Canners Association, 1959. Unpublished survey of waste
costs in canneries.
5. National Canners Association, 1970. Preliminary summaries of a
survey of solid wastes from canning and freezing. For Bureau of
Solid Waste Management, U.S. Dept of HEW.
6. Powers, T. J. Ill, B. R. Sacks and J. L. Holdaway, 1967. National
Industrial Waste-water Assessment Manufacturing Year 1963. U. S.
Dept. of Interior, FWPCA.
7. Secretary of Agriculture and Director, Office of Science and
Technology, 1969. Control of Agriculture-related Pollution.
Washington, D. C.
8. U. S. Dept. of Interior, FWPCA, 1967. The Cost of Clean Water,
Vol. Ill, Industrial Waste Profiles, No. 6-Canned and Frozen Fruits
and Vegetables.
9. U. S. Dept. Of Interior, FWPCA, 1969. The Cost of Clean Water and
its Economic Impact, Volume I The Report.
10. U. S. Dept. of Interior, FWPCA, 1970. The economics of clean water
Volume I Detailed analysis.
127
-------
PROGRESS REPORT: STUDY OF DRY CAUSTIC VS CONVENTIONAL CAUSTIC
PEELING AND THE EFFECT ON WASTE DISPOSAL
by
Joseph W. Cyr*
INTRODUCTION
The dry caustic peeling system was originally developed by Graham,
Huxsall, Hart, Weaver, and Morgan of Western Utilization Research
and Development Division mainly as an attempt to reduce the waste
generated in peeling potatoes and was later worked into an industrial
process by Magnuson Engineers, Inc. of San Jose, California. Side
benefits of lowered peel loss and reduced caustic usage have been
observed by some processors and are being hunted for by others.
This study is being partially financed by the Environmental Protection
Agency. The study was initially authorized in September 1969.
Installation of the dry caustic peel lines, primary treatment
systems, flow measurement instrumentation, sampling facilities and
the necessary laboratory facilities have taken place during the inter-
vening period.
The objectives of this study are as follows:
1. To determine total capital expenditures and operational costs of
the dry caustic process and the conventional caustic process.
2. To compare the quantity and quality of the waste generated by the
two systems.
3. To determine total water consumption, power requirements, and
maintenance costs of the two systems.
4. To compare the silt removal systems and final clarifiers as to
quality and quantity of influent and effluent.
5. To determine whether the dry caustic sludge would be accepted or
rejected for subsequent study as animal feed material.
I would like to take this opportunity to thank Kenneth Dostal of
E.P.A., Dr. Otis Sproul and Dr. John Vennes for their continual
guidance in planning the study and also for their help in standard-
*Western Potato Service, Inc., Grand Forks, North Dakota
129
-------
ization of the analytical procedures.
SUMMARY
1. Work to date shows about 4 times as much water usage in the wet
as in dry caustic peeling. However, this will become nearly 8 times
when water from the Magnuson washer is recirculated to the scrubber.
2. Pounds of dry solids to be dewatered/ton of potatoes to be peeled
in the wet caustic peel system is 3-3/4 times that for dry caustic
system.
3. Pounds of BOD/ton of potatoes to be peeled is nearly 4-1/2 times
as much in the wet system as in the dry.
4. Capital costs show approximately $102,000 for the dry system vs
$29,000 for the wet/line.
5. Operating costs including water are $0.46/ton for the dry systems in
N.D., $0.49 for the wet in Me. If water costs were common to both
plants at IOC/1000 gallons the comparison would be $0.39 vs $0.49; at
80C/1000 gallons, dry system would be $0.46 and the wet would become
$0.89. The final report will include sludge handling and BOD removal
costs where it is expected. The dry system will show lower costs than
the wet.
DESCRIPTION OF PLANTS
Western Potato Service, Inc. processes roughly 120,000 tons of potatoes/
year vs Potato Service, Inc's., 200,000. Both plants are highly
automated and produce primarily frozen french fried potatoes. Both
also produce several specialty items; namely, frozen hashbrowns,
frozen cottage fries, frozen whole boils, frozen tators (at PSI only),
and instant mashed potatoes. Water usage is roughly 1 mgd at WPSI and
3.5 mgd at PSI.
CAUSTIC PEEL AND PRIMARY TREATMENT SYSTEMS
Potatoes are flumed into the plants from plant storage and are stored
in hoppers for even-flow to the peel lines. From the hoppers, the
tubers are belt conveyed and weighed by Ramsey "Vey-R-Weight's"
conveyor scales to the caustic reels and are then discharged onto
retention belts. At this point, the two peel systems diverge. At
WPSI, where the dry caustic peel system is in use, the tubers proceed
over Magnuson shufflos, into the infra red units, thence to Magnuson
scrubbers and finally to Magnuson washers. At PSI, where the conventional
or wet caustic system is in use, the tubers after leaving the retention
belt go to the roll brusher washers. These consist essentially of a
series of cylindrical brushes with water spray nozzles directed
downward onto the tubers.
130
-------
With the dry caustic peel system, the bulk of the peel loss material
(about 90 percent) is removed as a thick, heavy "peanut butter" like
substance which is trucked off as is. The remainder is screened (20
mesh zreens) and is then pumped to the clarifier. With the conventional
or wet caustic system, the entire peel loss material goes down the
flumes, passed through 4 mesh screens and then pumped to the clarifier.
The primary clarifier is 60 feet in diameter at WPSI and allows for
an approximate 5 hour retention time; at PSI, the clarifier is 100 feet
in diameter allowing for 3-1/2 hours retention.
The Magnuson scrubbers are designed to use a minimum of water as are
the washers. Good operation requires about 25 gpm through each set of
units. The roll brusherwasher units on the other hand require
approximately 500 gpm. Thus, the hydraulic, BOD, and suspended solids
loads for primary treatment are markedly reduced when the dry caustic
peel system is used.
131
-------
TABLE 1
DRY SOLIDS AND WATER BALANCE
Dry Solids
pounds/ton
WPSI
PSI
Water
gallons/ton
WPSI PSI
INPUT:
Fresh Water
Potatoes
Fat
Caustic
SAPP, Sodium sulfite, and Bisulfite
Dextrose
Lime
TOTAL INPUT:
4.8
400.0
59.4
11.2
4.8
.3
3.8
483.5
1.46
387,
66.
22,
3.6
.2
0
1889.2
191.2
1.3
2081.9
OUTPUT:
Peeler Solids 57.8 0
Screenings 16.1 18.7
Clarifier
Effluent 32.2
Sludge 30.5
PRODUCT:
Fries including waste + overage 232.9 242.5
Flakes including .waste 30.5 38.5
Flake mash waste 6.4 6.4
Compressor Operation
Boiler Operation
Fry Loss and Flake Dry Loss
Sanitary Sewerage
TOTAL OUTPUT: 406.4
54.6 0
9.6 11300
1508.4
49.8
67.5 71.2
3.2 1.95
19.3
110.8
25.
32,
20.
26,
1881.1
Table 1: This is essentially a materials balance both for solids and
for water. Potato solids were determined with a National Potato Chip
Institute potato hydrometer. Note, too, that french fry waste and net
weight average have been combined in the figure for french fry product;
and flake waste is included in the figure for flake production. Since
fresh water contained some solids, etc., the analytical results for fresh
water were subtracted from those found in the waste, where consequential,
to give true results.
132
-------
TABLE II
CAUSTIC PEEL AND WASHER WASTES
Water - gallons/ton
BOD
COD
Solids
Total
Total Volatile
Suspended
Volatile Suspended
Settleable
Nitrogen - Total Kjeldahl as N
Phosphorus Total P
pH
Alkalinity
Pounds/Ton of
Raw Material Processed
WPSI
246.4
13.0
23.1
23.6
16.5
17.0
15.1
.84
.27
.11
11.4
3.22
Table IIi The levels of waste generation at WPSI on a pounds/ton
basis except in the case of water which is expressed in gallons/ton.
133
-------
TABLE III
CLARIFIER EFFICIENCY
WPSI
Water - Gallons/Ton
BOD
COD
Solids
Total
Total Volatile
Suspended
Volatile Suspended
Settleable
Nitrogen - Total
Kjeldahl as N
Phosphorus, Total P
pH
Alkalinity
Clarifier
Influent
Pound/Ton
1665.9
40.0
62.7
67.6
57.8
63.9
47.6
.51
1.34
2.25
10.8
13.88
Clarifier
Effluent
Pound /Ton
1648.2
24.1
33.6
41.4
21.3
7.55
5.43
.04
1.06
1.83
6.6
8.58
Reduction
%
1.06
39.8
46.4
38.7
63.1
88.2
88.6
91.5
21.0
18.6
38.2
Table III: This table shows the reduction both as a comparison between
clarifier influent and effluent and as percentage reduction based on per
ton of raw material input.
134
-------
TABLE IV
CAPITAL AND OPERATING COSTS
Caustic
Dry
45,000
.128/ton
.062/ton
.048/ton
,087/ton
*.136/ton
Peeling
Wet
$73,000
18,250
7,300
$21,500
5,400
2,150
Capital Costs
Equipment
Installation
Engineering
Operating Costs
Maintenance
Operation
Caustic
Gas
Power
Labor
Water
*Water at 80
-------
TABLE V
SIGNIFICANT DIFFERENCES
Peeling System
Dry Caustic Wet Caustic
Water Use - Gallons/Ton 170 2800 (80% recirculated)
Clarifier Solids to be
Dewatered pounds/ton 34 126
Clarifier BOD to
Secondary Treatment
pounds/ton 24.1 112
Table V: This is a summation of the significant differences between
the two systems.
136
-------
DRY CAUSTIC PEELING OF TREE FRUIT TO
REDUCE LIQUID WASTE VOLUME AND STRENGTH
by
. Me
Mark R. Hart** and Harry J. Maagdenberg*
Jack W. Rails*, Walter A. Mercer*, Robert P. Graham',
INTRODUCTION
The preparation of tree fruits for preservation by freezing, dehydration
or canning employs a complex technology and sophisticated equipment. The
equipment and technology used in peeling tree fruit was developed over
the years with high rates of production, adequate sanitation and optimum
product quality as the principal design criteria. Until quite recently,
only minor attention was paid to the effects of the peeling technology
on the quantities and strengths of liquid wastes generated in the process
as by-products.
Past and current research (and other water pollution abatement efforts)
have, for the most part, emphasized "end of the pipe" handling and treat-
ment of food processing wastewaters. Much time and large sums of money
have been spent in testing and adapting methods of treatment and disposal
for these liquid wastes. Rarely have any of the treatment systems used
proven completely satisfactory. Variations in the nature and volume of
food processing liquid wastes and inconsistent ability of different
treatment systems to handle these variants have caused many problems.
In-plant surveys made by research teams from the National Canners
Association have shown that high percentages of the total dissolved or-
ganic solids in the composite wastewater originate in unit operations
such as peeling and blanching of raw commodities.
The peeling of apples, apricots, and peaches produces 0.47, 0.16, and
0.42 Ibs of B.O.D. per case of finished product, respectively (Anon.,
1970). For peaches, the peeling operation, including the rinsing after
peel removal, comprises 40.5 percent of the total Ib of B.O.D. dis-
charged in the composite liquid waste from a cannery (Spicher, et al.
1967).
Often individual waste streams representing no more than five percent of
the total wastewater volume may contain 50 to 70 percent of the total
discharge of dissolved solids. Generally, the concentrated waste streams
are diluted by converging streams of comparatively clean waters.
'"'Western Research Laboratory, National Canners Association, Berkeley,
California.
**Western Utilization Research and Development Division, Agricultural
Research Service, United States Department of Agriculture, Albany,
California.
137
-------
It has become increasingly apparent in recent years that it is more
efficient to abate water pollution by changing processing technology
rather than by expanding treatment facilities. There are many po-
tential ways in which foods can be prepared for preservation which
will drastically reduce the volume and strength of wastewaters gene-
rated. An engineering development of the past three years in radical
modification of conventional food processing technology to reduce liquid
waste volume and strength is the "dry caustic" peeling of potatoes. This
process resulted from work at the Engineering and Development Division,
Agricultural Research Service, United States Department of Agriculture,
Albany, California (Graham, et_al.. 1968, 1970) and Magnuson Engineering,
Inc., San Jose, California (Smith, 1969, 1970).
The new process, now known as tne USDA-Magnuson Infrared Anti-Pollution
Peeling Process, uses infrared energy at 1650°F to condition the sur-
faces of potatoes treated with strong sodium hydroxide solutions. The
peel can then be removed mechanically by soft rubber scrubbing rolls
rather than by water as is done in conventional caustic peeling. A
final spray rinse using low volumes of water removes a residual peel
fragments and excess sodium hydroxide. The effluent from the peeled
potato rinsing may be combined with the solid material generated to
produce a thick, yet pumpable, sludge.
Direct comparison of the new process with conventional peeling has de-
monstrated that the strength of the waste discharged has been reduced
by 40 percent (Graham, et al., 1970). This result means that the capa-
city of secondary treatment plants required to condition the effluent
to a satisfactory B.O.D. and suspended solids level (Dostal, 1969)
could be greatly reduced with substantial overall savings in equipment
and operating costs. In situations where the potato processing effluent
is treated in a municipal system in combination with domestic sewage,
the resulting lower loadings allow for population growth in the area
served by the plant without requiring costly expansion.
The tonnage of potatoes processed each year is substantial and reduction
of water pollution caused by this commodity received first attention by
scientists and engineers. However, there are other food commodities'
which are chemically peeled and extension of the "dry caustic" peeling
process to these would provide the potential for significant water pol-
lution abatement. For example, the weight of tree fruit peeled by che-
mical methods was about 1,000,000 tons in 1968. Each ton of apricots,
peaches (Mercer, et a1., 1965) or pears processed produced, on the
average, about 12 pounds of B.O.D. and 9 pounds of suspended solids in
the rinse water used in conventional chemical peeling. If only a 40
percent reduction in strength of wastes from plants peeling these fruits
were achieved by "dry caustic" peeling some 5,000,000 pounds of B.O.D.
and 3,700,000 pounds of suspended solids would be removed from the
effluent waste streams each year.
This report describes results from a collaborative project between the
USDA and NCA to extend the water pollutional abatement potential of
"dry caustic" peeling of potatoes to apricot, peach and pear peeling.
The utility of the new peeling process for these three tree fruits was
138
-------
shown to be promising in exploratory studies (Graham, 1970, Hart, et al.,
1970). The potential for substantial reduction of water polluting liquid
waste volume and strength by the new method of peeling tree fruit was
recognized by the EPA (formerly FWPCA of the Department of the Interior)
and funds were made available to support an investigation.
CONSTRUCTION OF PEEL REMOVAL UNIT
Construction of the peel removal unit, as shown in a general concepts
drawing in Figure 1, was accomplished by the USDA in Albany, California.
An employee of the NCA was assigned as a collaborator in assisting USDA
employees in the construction.
The shaft for each spindle of rubber disks, spacers, and stiffeners were
cut from 1/2 in. stainless steel rod. One-half, 1/4, and 1/8 in. spacers
were cut from clear plastic sheets; all spacers were 1-5/8 in. in diameter.
Nineteen slots for one five ft section and 20 slots for the second sec-
tion were milled from 4 in. by 1/4 in. x 5 ft aluminum sheets to ac-
commodate the spindle shafts. A similar number of matched undersize
slots were milled from 1-3/4 in. by 1/2 in. by 5 ft nylon sheet. One-
eighth in. by 2 in. steel angles were cut, drilled and tapped to form
the support structure; this was assembled by using 3/8 in. by 5/12 in.
round head screws. The peel catching tray was fabricated from sheet
metal (See Figure 2).
The idler gear support bar was constructed from 1 in. by 1/2 in. by 5 ft
aluminum bar stock. Holes were drilled and tapped in the bar as shown
in Figure 3. Spindles were held in position by a clamped 1/4 in. by
3/4 in. by 5 ft aluminum bar. Figure 4 shows the standard gear ar-
rangement using #18 sprocket gears with # 8 idler gears. A differential
gear arrangement used alternate #16 and #20 sprocket gears in place of
the #18 gears of the standard arrangement.
The disk cover was fabricated from #20 gauge stainless steel sheet as
shown in Figure 5. A 1/8 in. clear plastic cover was loosely fitted on
each, metal cover section; these were fitted with two wooden knobs to
facilitate removal.
Two spindle carriers similar to those fabricated from aluminum were con-
structed from 3/4 in. exterior plywood.
The starters and variable drive motors (3/4 H.P.) were positioned as
shown in Figure 1. The starters were first attached to brackets and
then positioned to the welded rectangular frame support. The motor
bases were connected to 3/4 in. exterior plywood and then clamped to
the steel frame support. V-belts and pulleys were connected and gal-
vanized steel belt guards constructed and attached to the units. In
the initial trial runs with peaches, an equipment modification was
needed to prevent the peach halves from spinning in place along the sides
of the disk cover. One-half in., half round wood stock was cemented to
the cover (as indicated by the insert in Figure 5) to overcome this
spinning action.
139
-------
(6
0)
3
o
a M!
3 (TO
r g
ct>
0 H-
CD
O
o
»
n
CD
ill!
i |i;i
i'.'in,.;
i ..in,.;
-------
Figure 2
Peeler Tray
-------
Figure 3
Peeler Spindle Bearing Supports and Knee Action Component
-------
"^^^
Figure 4
Standard Gear Arrangement Showing Idler Support Bar
J
-------
\jjELf)
-------
Another modification, the use of wiper bars in the second tray section,
(Figure 6) was introduced to remove a large amount of tacky material
from pears before they came in contact with the fresh water.
The custom fabricated neoprene rubber disks (See Figure 7) were not de-
livered until after the apricot peeling had started. The disks used for
the apricot peeling experiments were fabricated and assembled by hand.
Four and 1/4 in. diameter by 1/32 in. thick disks were cut from a roll
of Shore A, durometer 50 (Anon., 1958) food grade rubber sheet. One-
half in. diameter holes were punched out of both the rubber disks and
2-3/4 in. diameter Mason jar lids used as stiffeners. The spindles
were assembled on the stainless shafts using the plastic spacers.
Six different sequences of spindles were used for peeling experiments on
apricots, peaches and pears. Three of the sequences used 39 identical
spindles and three of the sequences used a combination of several dif-
ferent spindle configurations. The ten different combinations of rubber
disks, spacers and stiffeners used for spindle configurations have been
assigned numbers and are shown diagramatically in Figures 8 through 10.
The sequences of individual spindle configuration in the 39 slots pro-
vided for the spindle shafts are designated by the letters A through F.
The spindle sequence A consisted of 39 identical spindle configurations
disignated by number 1. The spindle sequence F consisted of 39 iden-
tical spindle configurations designated as number 2. The spindle con-
figuration D consisted of 39 identical spindle configurations designated
by number 4. The spindle sequences using a combination of different
spindle configurations to fill the 39 shaft slots were designated B, C,
and E and are identified in Table 1. The shaft position numbers start
with 1 at the feed end of the peel removal section.
OPERATION AND EVALUATION OF FRUIT PEELER
Peeling of Peach Halves
The major effort in this study of peeling certain tree fruit was concen-
trated on evaluation of peel removal from peach halves. The bulk of
peaches harvested for processing are peeled before preservation.
The work on peach peeling was accomplished in a commercial cannery in
Richmond, California. This was done for several reasons. The cannery
location provided adequate supplies of peaches of commercial grade,
pitted with conventional equipment, and potentially returnable to pro-
duction. Experienced plant personnel were available to provide estimates
of the quality of peeled peach halves. Also, commercial scale equip-
ment was available for use in applying sodium hydroxide solution.
The equipment used is shown schematically in Figure 11. The experimental
peeling of peach halves was conducted in the following manner. Peach
halves from the Filper pitting machines were placed in 5 gal. stainless
steel pans and the peach weight was measured. The peach halves were
dumped into an elevator which delivered them to a cup down shaker. The
oriented peaches were conveyed through a commercial sodium hydroxide
145
-------
Figure 6
Disk Scraper Inserted In Spindle Housing
-------
'O
K4*
CO
cn
-------
1*1
II "
:n
I
c
c
K/-
j>
*/'
!>
+1
C
jj
%
1
•+
c
j)
1)
c
1
$
1
d
!)
!>
c
1
§
1
d
) C
„ ii
-x'?7"
1 It
1) 4
C
d
•>
|
$
1
!>
c
c
j>
;
Figure 8
Peeling Disk Configurations
148
-------
o
I
0
o
6 5
O
o
" II
II II
v^'
f
g
1
*
:
:
i\ it
_?/
II U
8
Figure 9
Peeling Disk Configurations
149
-------
I.
~l
10
Figure 10
Peeling Disk Configurations
150
-------
TABLE I
SEQUENCE OF SPINDLE CONFIGURATIONS USED
FOR THE SHAFT POSITIONS OF PEEL REMOVAL UNIT
Shaft
Position
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
B
Spindle
Configuration*
6
7
8
5
6
5
8
7
6
5
6
7
8
5
6
5
8
7
6
5
6
7
8
5
6
5
8
7
6
5
6
7
8
6
5
6
8
7
6
SEQUENCE
C
Spindle
Configuration*
10
7
8
9
10
9
8
7
10
9
10
7
8
9
10
9
8
7
10
9
10
7
8
9
10
9
8
7
10
9
10
7
8
10
9
10
8
7
10
E
Spindle
Configuration*
3
3
3
4
3
3
3
4
3
3
3
4
3
3
3
4
3
3
3
4
3
3
3
4
3
3
3
4
3
3
3
4
3
3
3
4
3
3
3
*See Figure 8 through 10.
151
-------
—>~
Cup Down
Lye Applicator
Peelremoval Unit
N
^ I
J
r
O
Fresh Water
Overflow Collection Point
Figure 11
Schematic Drawing of Experimental Peach Peeling Equipment
152
-------
solution applicator. The coated peach halves were drained of excess
sodium hydroxide solution and delivered by a sheet metal trough to the
peel removal unit. The peach halves moved through the unit and dropped
down a chute at the exit end into a surge tank filled with water. The
submerged peaches were removed from the surge tank by an elevator with
a 45 degree pitch into tared receivers. One to three spray heads pro-
vided a rinsing of the peaches in the elevator before delivery into the
tared containers. The fresh water introduced at the spray heads drained
into the surge tank as make-up water. The surge tank was fitted with
an overflow pipe which emptied into containers for volume measurement
and wastewater sample collection. The peeled peaches were examined
for peeling quality by visual inspection, weighed, and returned to the
Filper pitter discharge flume to be used in commercial production.
A large number of preliminary experiments of short duration were used to
determine good operating conditions. The major variables studied
during the preliminary runs were: sodium hydroxide concentration,
sodium hydroxide bath temperature, residence time of peach halves in
sodium hydroxide applicator, drain time for coated peach halves, peeling
disk type and spacing, feed rate, peel removal section slope and disk
turning speed.
Table II summarizes the results of longer duration peeling experiments
with peach halves using the best operating conditions developed during
preliminary, short duration experiments. Peeling losses were deter-
mined in 3 min. runs using conditions identical to those used in the
runs of approximately one hour duration.
TABLE II
RESULTS OF PEACH HALF PEELING USING
ROTATING RUBBER DISK PEEL REMOVAL UNIT
Run No.
8-21-2*
8-21-1A
8-21-2**
8-21-2A
Feed
Rate,
Ib/hr
4119
4286
4647
4996
NaOH
Strength
Percent
1.5
1.4
1.5
1.4
Peel
Removal
Unit Slope,
in./lO ft
15
15
9
9
Peeled
Fruit
Quality
Good
Good
Fair
Fair
Peeling
Loss ,
Percent
N.M.***
5.1
N.M.
4.7
^Duration of run was 61 minutes
**Duration of run was 56.3 minutes
***N.M. = Not Measured
Table III tabulates the water volume used in rinsing peeled peach halves
and analytical values measured on wastewater samples.
The wastewater overflow from the surge tank was collected in a 55 gal.
drum. A one gal. sample was removed from the drum at 15 min. intervals
153
-------
TABLE III
RINSE WATER VOLUME AND WASTEWATER
CHARACTERISTICS FOR PEACH HALF PEELING
Run Rinse Water Rate
No. gal. /hr gal. /ton
8-21-1 75 36
8-Z1-2 77 33
Wastewater
Sample*
15 G
30 G
45 G
60 G
61 C
15 G
30 G
45 G
56 G
56C
Wastewater Quality
COD, ppm
30,000
60,000
63,300
97, 400
75,700
30,000
53,900
69, 100
63,300
50,200
SS,ppm
7,200
12,400
7,300
12,480
10,850
7,200
10, 150
12,600
18,400
9,750
pH
9.4
9.7
9.8
9.7
9.7
9.4
9.4
9.6
9.7
9.5
Time in minutes, G = grab, C = composite.
TABLE IV
WATER VOLUME REQUIREMENTS AND WASTEWATER
CHARACTERISTICS FOR COMMERCIAL CLING PEACH PEELER
Fresh Water
Input,
gal. /ton
600
540
580
525
460
515
470
Ave.527
Wastewater
Sampling
Point*
1
2
1
2
1
2
1
2
1
2
1
2
1
2
-
Wastewater
Volume,
gal. /ton
400
200
360
180
387
193
350
175
307
153
347
168
313
157
527
Measurements
C.O.D., SS,
ppm PP3*1
5,300
19, 600
5, 600
16,700
5, 100
6,600
5,900
23,000
5,100
30,000
5,100
25, 100
4,900
31,700
13,550
1,230
2, 430
1,060
1,280
780
1,950
1, 150
3,080
1,320
5, 640
1,540
5, 640
910
5,240
2,375
pH
11.3
10.2
11.6
10.3
11.1
10.1
11.1
10.1
11.0
10.7
11.2
10.7
11.3
10.7
10.8
See Figure 12
154
-------
after mixing Che contents of the drum. A composite sample was prepared
by mixing 0.5 gal. portions of each of the 15 min. grab samples.
For comparative purposes, wastewater samples were collected from the
commercial peach half peeling unit (See Figure 12) and analyzed. The
results are tabulated in Table IV.
Peeling of Whole Pears
The peeling of whole Bartlett pears was conducted at two locations. A
limited number of short duration peeling experiments with pears were run
in the commercial cannery with the sodium hydroxide applicating equipment
used for the peach half peeling. It was too difficult and potentially
dangerous to project personnel to obtain sodium hydroxide treated pears
from the commercial equipment in the cannery. The commercial peeler has
a pressurized section following the sodium hydroxide application zone.
The pears are conveyed through an air-lock from the pressure section to
a lowerator which delivers the partially peeled pears to a flume. Re-
moving pears from the lowerator was judged by the Project Director to
be hazardous due to the restricted clearance between lowerator flights
and sidewalls which could catch and break hands and arms. Beyond this
consideration of safety, the treated pears at this point had lost con-
siderable portions.of their peel and were not the best material for
peeling experiments.
Extensive experimental peeling of pears was conducted at the USDA Labor-
atory in Albany, California due to inadequate clearance space between
belt and cover for the commercial peach sodium hydroxide applicator in
Richmond. A screw drive sodium hydroxide applicator unit was used for
these experiments. The results of three longer duration peeling experi-
ments on pears are tabulated in Table V.
TABLE V
RESULTS OF WHOLE BARTLETT PEAR PEELING USING
ROTATING RUBBER DISK PEEL REMOVAL UNIT
Run
No.
10-2-1*
10-16-4**
10-28-1***
*13.0 minute
**12 . 3 minute
***30.1 minute
Feed
Rate,
Ib/hr
1,700
2,050
1,786
run
run
run
NaOH
Strength
Percent
15.2
17.3
17.0
Peeled
Fruit
Quality
Good to Poor
Good
Good
Peeling
Loss ,
Percent
11.3
20.4
15.8
The rinsewater volume used in Run 10-28-1 of Table V was 89.4 gal. per
ton and the wastewater had a C.O.D. of 10,500 ppm, a SS content of 2550
ppm and a pH of 10.6. The rinse water for the pear peeling consisted of
155
-------
Fresh Water
/v7v/r""A
Halved
Fruit"
Cup Down
Shaker
-i>-
i n i ! R^se
Lye Peeler
} Section
--0--
Water Dip
0
Water
yVAA^
|Flum--
bieam
Blancher
T
••[>
Sorting & Trimming
^7
Figure 12
Schematic Drawing of Commercial Peach Peeling
156
-------
a tank containing 20 gal. of water. Analysis was made on a composite
sample of water from the rinse tank.
For comparison purposes, wastewater from a commercial pear peeling unit
(See Figure 13) was collected and analyzed; the pH was 10.5 and the
C.O.D. was 5,040 ppm. The rinse water volume was estimated as about
200 gal./ton of fresh pears.
Three facets of pear peeling were examined during the project. The
sizing and firmness of pears are important considerations in their peeling
performance by the methods used at the present time. Measurements of
size distribution of pears peeled successfully were made to provide in-
dustry specialists with information which they could correlate with their
experience using conventional peeling equipment. The size distribution
of a typical lot of pears is tabulated in Table VI.
Another important factor in pear peeling by conventional methods is the
firmness of the pears. Firmness is measured by reading the force
necessary to insert a rod into a pear (at a point where the skin has been
removed with a knife) to a standard depth. The results of the penetration
testing of a lot of 33 pears typical of those used in the experimental
peeling are tabulated in Table VII. The results of the firmness testing
indicated that the pears used were in the normal range for successful
commercial peeling.
The peel material, which is wiped from the fruit drops or is spun down
as a sludge to the bottom of the peel removal unit tray and drains (or
can be scraped) counterflow to the peeled fruit. The peeling sludge can
be handled as a solid waste material. As disposal of solid waste in land
fill becomes more difficult in the future due to lack of appropriate sites,
the peel sludge has potential as a component of animal feed. A limited
amount of information on the composition of the peeling sludge was ob-
tained during the project. The weight of the sludge is approximately
represented by the peeling loss. No direct measure of peeling sludge
weight was possible due to the deposition of part of this material on
the walls of the equipment. It was impractical to determine the'weight
of the deposited peel sludge as well as to account for losses of material
spun away from the unit from the peeling disks (by centrifugal force).
Values for the content of sodium hydroxide and sodium carbonate in the
pear peeling sludge is tabulated in Table VIII.
Peeling of Whole Apricots
A number of short term peeling experiments were completed at the USDA
Laboratory in Albany, California using Blenheim and Tilton apricots
obtained from canneries in the San Francisco Bay area.
Due to the delay in delivery of vulcanized rubber peeling disks, time
allocated to the construction of a spray rinsing section was diverted to
assembling peeling disks by hand. The rinsing of peeled apricots was
by simple submersion in fresh water contained in a tank. The waste-
water generated was considered as not representative of wastewater from
a spray rinsing unit. Therefore, no wastewater samples from apricot
peeling experiments were collected.
157
-------
Fresh Water
Bin
Conveyer
Lye Dipper &
. Steam Valve
-0"
Rotary
Washer
Overflow Collection Point
K>
Flume
—£>--Core Equipment
Sorting &
Trimming
Figure 13
Schematic Drawing of Commercial Pear Peeling
158
-------
TABLE VI
CROSS SECTIONAL SIZE DISTRIBUTION OF PEARS
Size, Percent Present
In.
2 3/16
2 1/4
2 5/16
2 3/8
2 7/16
2 1/2
2 9/16
2 5/8
2 11/16
2 3/4
2 13/16
2 7/8
2 15/16
3
Total
TABLE VII
FIRMNESS TESTING OF PEARS
Penetration Number of Pears
Force, Ib Penetrated
2-2.9 4
3- 3.9 4
4-4.9 6
5- 5.9 6
6- ..6.9 5
7- 7.9 3
8-8.9 3
9- 9.9 1
10- 10.9 1
in Lot
3
10
10
16
15
21
6
8
6
3
1
0
0
1
100
Distribution
Percent
12.1
12.1
18.2
18.2
15.2
9.1
9.1
3.0
3.0
Totals 33 100.0
159
-------
TABLE VIII
SODIUM HYDROXIDE AND SODIUM CARBONATE
CONTENT OF PEARS PEELING SLUDGE
Sampling
Date
9-29-1
9-29-2
9-30-1
9-30-2
10-1-1
10-1-2
10-5-1
10-6-1
10-13-1
10-13-2
10-14-4
10-14-5
10-28-1
Solids,
Percent
27.3
27.0
23.6
20.8
21.6
18.4
22,4
22.2
28.5
24.0
24.9
20.2
23.5
NaOH
Wet Wt,
Percent
3.8
2.1
2.0
0.6
1.7
0.3
1.7
1.1
3.8
1.8
2.9
1.2
2.2
Wet Wt,
Percent
2.6
2.4
1.8
1.4
1.6
0.9
1.3
1.4
2.5
2.3
1.8
1.2
2.3
Average
23.4
1.9
1.8
160
-------
The results obtained during the peeling of Blenheim apricots are tabu-
lated in Table IX. The results obtained during the peeling of Til ton
apricots are tabulated in Table X.
TABLE IX
RESULTS OF SHORT DURATION PEELING EXPERIMENTS
ON MIXED SIZE BLENHEIM APRICOTS
Run No.* 6-18-6 6-19-1 6-19-2 6-19-4 6-19-5
NaOH Cone . , %
NaOH Temp. °F
Disk Rotating Speed, RPM
Peeler Slope, in./lO ft
Feed, Ib
Feed Time Sec.
Feed Rate, Ib/hr
Peeled Weight, Ib.
Washed Weight, Ib
Total Peel Loss, %
Gears Used
Peeled Fruit Quality
Rinse Water, gal.
Rinse Water, gal. /ton
NaOH Dip Time, Sec.
25-30
210
266
-4
60
240
900
53.9
42.7
28.8
s**
N.R.
5.0
167
33
25-30
150
266
0
30
90
1200
27.6
25.0
16.7
S
N.R.
5.0
333
70
25-30
140
373
0
30
35
3080
29.3
27.5
8.3
S
N.R.
5.0
333
70
25-30
160
373
4
30
90
1200
26.3
25.3
15.6
D
N.R.
5.0
333
70
25-30
150
373
4
30
90
1200
27.3
26.1
13.0
D***
N.R.
5.0
333
70
*Spindle sequence was F in all runs.
**S - Standard
***D - Differential
N.R. - Not Recorded
DISCUSSION
The operation of the peel removal equipment demonstrated on a substantial
scale (one to two tons of raw fruit per hour) that peach halves can be
peeled efficiently by the low water volume ("dry caustic") peeling process.
The quality of the peeled fruit was as good or better than that of fruit
from commercial peeling units as judged by experienced industry people.
The peeling losses measured for the experimental unit were slightly lower
than the peeling losses determined by measurement of commercial peeling.
The difference in peeling losses was not large enough to provide an econ-
omic incentive based on higher product yield as is the case for the new
potato peeling process.
The maior advantage for the new method for peach peeling results
from the lower water volume requirement and the reduction in the strength
of wastewater generated.
A detailed study of fresh water requirements for rinsing and character-
ization of wastewater was made only for peach half peeling for the reason
161
-------
TABLE X
RESULTS OF SHORT DURATION PEELING EXPERIMENTS
ON MIXED SIZE TILTON APRICOTS
Run No. *
NaOH Cone., %
NaOH Temp. °F
Peeler Slope, In. /10 ft
Feed, Ib
Feed Time, Sec,
Feed Rate, Ib/hr
Peeled Weight, Ib
Washed Weight, Ib
Total Peel Loss, %
Disk Spacing, in.
Gears Used
Peeled Fruit Quality
Rinse Water, gal.
Rinse Water, gal. /ton
NaOH Dip Time, Sec.
Run No.
NaOH Cone., %
NaOH Temp. °F
Peeler Slope, irj 10 ft
Feed, Ib.
Feed Time, Sec.
Feed Rate, Ib. /hr
Peeled Weight, Ib
Washed Weight, Ib
Total Peel Loss, %
Disk Spacing, in.
Gears Used
Peeled Fruit Quality
Rinse Water, gal.
Rinse Water, gal. /ton
NaOH Dip Time, Sec.
6-22-2
25-30
210
8
30
120
900
26.3
26.2
12.6
3/4
D*#
Good
5.0
333
30
6-26-2
25-30
210
16
45
90
1800
41.7
40.1
7.3
1/2
S
Good
5
222
35
6-22-3
25-30
210
8
30
120
900
26.7
26.2
12.6
3/4
D
Good
5.0
333
30
6-26-3
25-30
210
16
45
75
1920
41.2
40.0
8.4
1/2
1/2 D
Good
5
238
32
6-23-1
25-30
210
15
45
120
1350
40.8
40.5
10.0
3/4
D
Good
5.0
222
22
6-26-4
25-30
210
16
45
65
2500
41.7
39.4
7.1
1/2
D
Good
5
222
30
6-23-2
25-30
210
15
60
60
3600
57.1
53.8
10.3
3/4
D
Good
5.0
167
22
6-30-1
12.2
210
16
60
115
1875
53.1
53.0
11.5
1/2
D
Good
5
166
30
6-23-3
25-30
210
15
30
40
2700
28.4
26.7
11.0
3/4
s***
Good
5.0
333
22
6-30-2
11.5
210
16
60
120
1800
54.1
53.2
9.8
1/2
D
Fair
5
167
30
6-26-1
25-30
210
8
45
40
2330
41.3
40.7
9.5
1/2
S
Good
5.0
222
30
7-1-1
10.8
140
16
60
120
1800
56.6
N.R.
N.R.
1/2
D
Fair
5
167
66
162
-------
TABLE X (Cont'd. )
Run No.
NaOH Cone., %
NaOH Temp. °F
Peeler Slope, in. /10 ft
Feed, lb
Feed Time, Sec.
Feed Rate, Ib/hr
Peeled Weight, lb
Washed Weight, lb
Total Peel Loss, %
Disk Configuration
Disk Spacing, in.
Gears Used
Peeled Fruit Quality
Rinse Water, gal.
Rinse Water, gal. /ton
NaOH Dip Time, Sec.
Run No.
NaOH Cone., %
NaOH Temp. °F
Peeler Slope, in./lO ft
Feed, lb
Feed Time, Sec.
Feed Rate, Ib/hr
Peeled Weight, lb
Washed Weight, lb
Total Peel Loss, %
Disk Configuration
Disk Spacing, in.
Gears Used
Peeled Fruit Quality
Rinse Water, gal.
Rinse Water, Gal. /ton
NaOH Dip Time, Sec.
7-1-2
11.1
140
16
60
120
1800
54.5
54.3
9.5
0
1/2
D
Fair
5
167
72
7-9-1
9.8
210
16
60
68
3200
55.9
N.R.
6.8
0
3/4
D
Good
5
166
24
7-1-3
10.8
210
16
60
120
1800
55.2
N.R.
8.0
0
1/2
D
Fair
5
167
27
7-9-2
10.0
210
16
60
68
3200
56.2
57.0
6.3
0
3/4
D
Good
5
166
27
7-1-4
10-11
210
16
60
120
1800
55.0
N.R.
8.3
0
1/2
D
Fair
5
167
27
7-9-3
9.8
210
16
60
68
3200
57.8
56.5
5.9
0
3/4
D
Good
5
166
27
7-1-5
11.5
210
16
60
120
1800
55.5
53.8
10.3
0
1/2
D
Fair
5
167
27
7-9-4
10.8
210
16
60
68
3200
57.6
55.0
7.5
0
3/4
D
Good
5
166
27
7-7-1
11.0
210
16
60
68
3200
55.3
N.R.
8.0
0
3/4
D
Good
5
166
24
7-10-1
10.0
210
16
60
68
3200
55.3
55.0
8.3
0
3/4
D
Good
5
166
27
7-7-2
11.2
210
16
60
68
3200
55.9
N.R.
6.8
0
3/4
D
Good
5
166
24
7-10-2
10.0
210
16
60
68
3200
55.0
54.1
9.8
0
3/4
D
Good
5
166
27
* The disk rotating speed was 373 RPM for all runs and the spindle
sequence was F.
*# S - Standard
#** D - Differential
N.R. - Not recorded
163
-------
described above. The results of comparing commercial peeling and ex-
perimental peeling of peaches showed that the new peel removal method
required only one-fifteenth of the fresh water used in conventional
commercial peeling. The wastewater produced in the new peeling method
was correspondingly reduced due to the use of mechanical abrasion for
peel removal rather than water pressure. The strength of the waste-
water from the experimental peeling of peaches was higher per unit
volume than the wastewater from conventional chemical peeling. However,
the much lower volume of wastewater produced per ton of fruit by the
use of mechanical peel removal makes the Ib of C.O.D. ( and of B.O.D.,
although this was not measured directly) generated about one-third that
of the conventional chemical peeling operation.
The numbers used to calculate values tabulated in Table XI are derived
from Tables III and IV. The average of the two wastewater volumes and
the average C.O.D. and S.S. of the two composite samples were used to
calculate the values for the experimental peeler. The average values
from Table IV were used to calculate figures for commercial peeling.
The contribution of the effluent from the blancher wastewater shown
in Figure 12 to volume, C.O.D. and SS was estimated as less than 0.5
percent of the total and was not corrected for in the calculation.
TABLE XI
COMPARISON OF WASTEWATER CHARACTERISTICS FOR
COMMERCIAL AND EXPERIMENTAL PEELING OF CLING PEACH HALVES
WASTEWATER DISCHARGED
(PER TON PITTED PEACHES)
Volume, C.O.D. SS,
Peeler gal. Ib Ib
Commercial 527 59.5 10.4
Experimental 34.5 18.1 3.0
A comparison of water usage and wastewater characteristics in commercial
and experimental peeling of peach halves are tabulated in Table XI.
Pre-Treatment of Unpeeled Pears
Several treatments of unpeeled pears before the application of sodium
hydroxide solution were investigated to improve peeling efficiency and
to reduce peeling losses,
Dewaxing of whole pears by immersion in 2-propanol held at 165°F reduced
the peeling loss, but the quality of the peeled fruit was poorer due to
residual peel fragments. It was concluded that hot or cold dewaxing of
pears did not improve peeling sufficiently to justify the extra operation
which would be required.
In the final pear peeling experiments it was found that placing 1/4 in.
stainless steel bars across the peeling tray parallel to the spindles
164
-------
wiped a substantial portion of softened flesh from the pear surfaces
(see Figure 6). This modification of the peeler unit had considerable
promise in diverting organic material coating peeled tree fruit from
the wastewater to the solid residue fraction.
Peeling Losses
Peeling losses were determined for the experimental peel removal unit
and a commercial unit for each of the three fruits studied. The data
collected during the peeling loss determinations are tabulated as
ranges of values in Table XII.
TABLE XII
PEELING LOSSES FOR TREE FRUIT
Peeling Loss, Percent
Commodity Experimental Commercial
Apricots
Peaches
Pears
3.7 -
5.3 -
11 -
8.3
7.5
20
6.4
5.5
12
- 9.3
- 8.0
- 15
Cost Estimate
The next logical step in demonstrating the utility of "dry caustic"
peeling of tree fruit would be a five fold scale up to a unit with a
capacity of about 10 tons of fruit per hour. The estimated capital
cost of the peel removal section of a 10 ton per hour capacity unit
is $16,000.
165
-------
ACKNOWLEDGMENTS
The collaboration in this project by Gerald S. Williams, and others
of the Engineering and Development Laboratory, Western Utilization
Research and Development Division, Agriculture Research Service,
United States Department of Agriculture, was instrumental in
developing the results reported.
The cooperation, interest and assistance of many individuals associated
with the canning industry is gratefully acknowledged; we especially
thank E. L. Mitchell, W. L. Doucett, Ernest Johnson, R. Lovelace, S.
Platou, S. M. Anderson, R. Ketcher, and L. K. Taber.
We appreciate the advice offered by Kenneth A. Dostal of WQO of EPA
during this project and especially for help in preparation of reports.
Reference to a company or product name does not imply approval or
recommendation of the product by the Environmental Protection Agency
or by the Department of Agriculture and does not imply that there
are no other suitable products available commercially.
166
-------
REFERENCES
Anon. 1555. ASTM Standard Methods, Part 9, page 1303. Tentative
Method for Indentation of Rubber by Means of a Duroneter.
Anon.. Preiininary Data, National Canners Association, 1970. Food
Canning Pastes
Dostal. K.A., 1969. Secondary Treatment of Potato Processing Wastes,
Final Report, Report No. FR-7, U.S. Department of the Interior,
FVfPCA Mi' Region, Pacific Northwest Water Laboratory, Corvallis,
Oregon 97330, July.
Hart, M,R,, R.P, Grahan:, C.C. Huxsoll and G.S. Williams 1970, An
Experimental Dry Caustic Peeler for Cling Peaches and Other Fruit.
30th Annual Meeting, Institute of Food Technologists, San Francisco,
California, May 25, 1970, Paper No. 59; J. Food Sci. .35 (6), 839-41
(1970)
Grahan, R.P., Huxsoll, C.C., Hart, M.R., Weaver, M.L., and Morgan,
A.I., Jr. 1969 "Dry" Caustic Peeling of Potatoes, Food Technology
2_3 (2), 61-65.
Grahan, R.P., 1970, Proceedings, National Symposium on Food Pro-
cessing Wastes, Portland, Oregon, April, pages 355-358,
Mercer, W.A., Rose, W.W., and Doyle, E.S., 1965. Physical and
chemical characterization of the fresh water intake, separate in-
plant waste streams and composite waste flows originating in a
cannery processing peaches and tomatoes, Res. Report No. D-1612,
prepared for the State of California Water Quality Control Board
(March).
Smith, T., 1970, Proceedings, National Symposium on Food Processing
Wastes, Portland, Oregon, April, pages 359-361.
Spicher, P.G., F.J. Agardy and G.T. Orlob, Proc. 22nd Ind. Waste
Conference, 1967, Part I., page 44, Purdue University, Lafayette,
Indiana.
167
-------
PRODUCTION OF POTATO STARCH WITHOUT WASTE
by
Roy Shaw-, & W. C. Shuey**
ABSTRACT
The use of fine grinding and air classification of deydrated potatoes
with sieving and minimum washing to produce potato starch is described.
A substantial reduction in the amount of waste water over that required
by conventional wet milling procedures was obtained. The procedure
proposed would reduce wastes by 90 percent.
INTRODUCTION
This paper reports on investigations of potato starch production employ-
ing air classification technology, and on refinement of the starch
fraction by carefully controlled washing. Utilization of the small
amount of waste water was also considered.
Potato starch is traditionally produced by wet-milling, which requires
large amounts of water to separate proteins, pulp, and other materials
from the product. Olson^) reported that, in a plant which utilizes
250 tons/day, the waste water was 800 gals/ton, excluding water for
washing potatoes. In a modern plant in Europe, Caransat^} with the
newest processing equipment, predicted needing 792 gals/metric ton.
Disposal of this waste water becomes difficult since Federal and State
laws prohibit disposal in water courses, and disposal through municipal
sewage plants would be expensive. Due to the low concentration of
solids in waste water (1 percent according to Heisler^ , OlsonC-), and
deKoe' '\ little had been done towards recovery of components until
the recent work of Heisler _e_t al (3) .
Since the mid-1950's, air-classifiers have become important devices for
separation of flour by particle size and density (Graham^)). The fine
grinders and classifiers described by Behrens^"' effectively separated
flour into fractions having varying protein contents. Gracza^'' defined
this change in protein content of flour fractions as degree of protein
*Red River Valley Potato Processing Laboratory, East Grand Forks, Min-
nesota. Cooperatively operated by the Eastern Marketing and Nutrition
Research Division, Agricultural Research Service, U. S. Department of
Agriculture; Minnesota Agricultural Experiment Station; North Dakota
Agricultural Experiment Station; and the Red River Valley Potato
Growers' Association.
**Plant Science Division, ARS, USDA, North Dakota State University, Fargo,
North Dakota. Cooperative investigations with the Department of Cereal
Chemistry and Technology, North Dakota State University, Fargo, North
Dakota. Published with the approval of the Director of the Agricultural
Experiment Station, North Dakota State University, Fargo, North Dakota
as Journal Series No. .
169
-------
shift. Peplinski et. al- demonstrated that physical characteristics
of wheat classes and variety influence this protein shift. Kent^
showed that changes in moisture content of wheat or flour also ap-
preciably altered the effectiveness of the protein shift. The number
of fractions obtained was dependent upon the complexity of the flow
employed in making the separations.
If finely ground, dehydrated potatoes have properties similar to those
of wheat flour, perhaps the product could be separated into a "protein
fraction" (suitable for animal feed) and a "starch fraction". If the
starch fraction thus produced was not of commercial quality, it could
be washed under careful conditions to minimize the amount of water used.
The concentration of solids in this wash water might be high enough
that drying would be economical, thereby eliminating the waste problem;
or recovery of valuable components by such processes as ion exchange
could make production of potato starch more economical. Even if neither
alternative was attractive, the amount of waste for disposal would be
greatly reduced over present practices.
MATERIALS AND METHODS
The test potatoes were from sound, washed, random lots of various var-
ieties which had been stored at 10-19°C. They were cut, dipped in
0.1 percent NaHSO., solution and dried as described in each experiment.
The dried samples were ground on an Alpine Pin Mill, Type 160, Alpine
American Corporation, Saxonville, Massachusetts.
Samples were classified on a Midroplex Spiral Classifier, Type 132MP,
Alpine American Corporation, Saxonville, Massachusetts.
Total Sugars. Determined by the Anthrone Method of Ashwell et al
Reducing Sugars. Determined by a modified Nelson-Somogyi Method' *•'.
Starch. Sample was refluxed for 2.5 hours in 10 percent HC1 followed
by neutralization and by determination of reducing sugars, from which
values for the previously determined total sugars were subtracted. A
sample of pure potato starch similarly treated was used as a standard.
Protein. AOAC 9th Edition. Protein as measured by micro-Kjeldahl
nitrogen x 6.25 was selected as a criterion of cleanliness; it being
less soluble than sugars, amino acids, organic acids, etc.
Moisture. Determined by overnight drying in a vacuum oven at 32 °C.
Agtron Color. Determined on an Agtron M-500 00/0-97/100, Magnuson
Engineers, Inc., San Jose, California.
Gelatinization. The Brabender Amylograph was employed in studying the
gelatinization properties of the different potato starches. A 20g.
170
-------
sample of starch (dry basis) in 450 ml. distilled water was used for the
determination. The temperature on the amylograph was programmed from
25-95°C., increasing at the rate of 1 1/2° per minute. Tho temperature
was then held at 95° for 15 minutes and was followed by controlled
cooling to 50°. The initial temperature of gelatinization, peak height,
15-minute-hold height, and 50° peak height were recorded.
RESULTS AND DISCUSSION
Fine Grinding and Air Classification
An initial experiment was conducted by using potatoes which had been
dried at 85° and coarse ground on a pin mill at 9,000 RPM. The ground
potatoes were then classified and the coarse fraction or "starch fraction"
was reground in the pin mill at 19,000 RPM and classified. This initial
study showed that a significant shift in protein content had been ob-
tained by such a technique, and that, similar to wheat flour, potatoes
could be air classified after fine grinding.
The second experiment was conducted to obtain a protein profile-air
classification fractionation curve. The potatoes were dried at 85° and
then run in the pin mill at 19,000 RPM. After each classification, the
starch fraction was recycled through the classifier with minor changes
of equipment settings.
Table 2 shows that potatoes are affected somewhat as is wheat flour
(Pfeifer and Griffin)'1-^. There is a "protein fraction" that can easily
be separated from a "starch fraction" and, as with wheat flour, continual
reclassification of the starch fraction improves separation only up to a
point. It appears (Table 1) that it would be impossible to produce a
starch fraction of less than 4 percent residual protein by air classifi-
cation alone.
Table 1
Sample
Control
Results from Initial Study of Air-Classified Whole Potatoes
Grind Classifier
RPM Settings
Feed Fin
Gate Setting
1st cycle 9,000 15 10'
2nd cycle 19,000 15 10C
Classified Percent of Protein Moisture
Fraction Product Content Content
Protein
Starch
Protein
Starch
100
8.8
91.2
4.2
95.8
8.5
15.2
6.8
22.4
6.3
4.7
4.3
3.9
2.7
4.9
171
-------
Table 2
Effect of Reclassifvine Starch Fraction
to
Cycle
Classifier
Settings
Feed Fin
Gate Settines
15 10°
15 15°
15 25°
15 35°
10 38°
Classified
Fraction
Control
Protein
Starch
Protein
Starch
Protein
Starch
Protein
Starch
Protein
Starch
Pounds
Weight
50
2.3
46.3
1.7
44.0
6.1
37.3
21.1
15.1
7.3
7.6
Percent
Protein
Content
7.9
30.0
5.8
18.1
5.4
8.2
5.3
5.6
5.0
5.3
5.2
Percent
Starch
Content
79.6
30.6
83.7
61.8
82.9
76.3
83.6
76.1
82.9
81.9
80.1
Percent
Sugar
Content
.8
1.7
1.0
1.3
.9
1.0
.9
.8
.9
.9
.7
Percent
Moisture
Content
1.3
1.3
1.7
2.2
1.3
1.2
2.0
1.7
2.8
2.2
1.7
Percent
Others
10.4
36.4
17.8
16.6
9.5
13.3
8.2
15.8
8.4
9.7
12.3
-------
Visual inspection of the fifth cycle starch fraction showed it to be
enriched in coarse brown particles. A simple screening test indicated
that most of these brown particles were retained on a 125-micron screen.
A third experiment was conducted to determine the effect of separating
the ground potatoes into two fractions, first by sieving and then by
additional fine grinding of the coarse, starch fraction after air
classification. For this, potatoes were dried at 60°, ground, and then
pin-milled at 14,000 RPM. Sieving over a 125-micron screen separated
the groundmaterial into two fractions. Material remaining on the screen
was principally coarse, brown peel fragments. That which passed through
the screen was air classified. The starch fraction of the first cycle
was reground at 19,000 RPM and again classified.
Sieving to remove brown peel and fibre fragments before classification
significantly lowered the residual protein in the starch fraction as
shown in Table 3.
A sample of blanched, dehydrated diced potatoes was obtained from a
local source for a fourth experiment to determine if blanching would affect
the protein-shifting potential. The data indicated that even cooked
potatoes exhibit some protein shifting when fine ground and classified.
A fifth experiment was conducted to determine the effect moisture content
might have on air classification of potatoes and on protein shift. Po-
tatoes were dried at 60° and run through the pin mill at 14,000 RPM.
Material passing through 125-micron screenire was exposed as a thin layer
to high humidity overnight. This "high moisture" material was classi-
fied, the "first cycle starch fraction" being reground at 19,000 RPM and
again classified.
Moisture as high as 11 percent does not materially affect air classifi-
cation of finely ground potatoes.
A sixth experiment was performed to determine the effect of drying tem-
perature upon air classification of the potatoes. About one-half of
the potatoes were dried at 82°, the rest being dried at 121°; all were
run through the pin mill at 14,000 RPM. Material passing through a
149-micron screen was exposed to high humidity and classified. The starch
fraction was reground at 19,000 RPM and again classified. The feed rate
to the classifier was varied in the second cycle of each run.
The initial drying temperature does not affect classification. The
darker color as reflected by Agtron color (Table 4 and Table 5) indicates
some heat damage, as one would expect. The potatoes were hard and vitreous
after drying at 121°, and did not grind easily; thus, the preliminary
screening was done on a 149-micron screen.
Purification of Starch Fractionation
Previous experiments showed that the starch fraction should be washed
in order to make a commercially acceptable product with less than 1
percent protein. Figure 1 outlines a typical procedure in the reusing
of wash water in a counter-current fashion.
173
-------
Table 3
Effect of Preliminary Sieving before Classification
Classifier
Settings
Grind Feed
Cycle RPM Gate
0 14,000
Fin Classified Pounds
Setting Fraction Weight
+125 micron 1.3
-125 micron 25.8
1 14,000 15
2 19,000 15
Table
Grind
Cycle RPM
0 14,000
1 14,000
2 19,000
Protein
25°
Starch
Protein
Starch
4 Effect of High
Classifier
Settings
Feed Fin
Gate Setting
_
15 25°
15 25°
2.8
22.8
3.3
18.3
Percent Percent
Protein Starch
Content Content
8.7 75.7
6.9 80.1
18.5 44.0
4.9 82.7
11.6 71.6
4.0 86.1
Moisture on Air Classification of
Classified
Product
Starting
Protein
Starch
Protein
Starch
Percent Percent
Sugar Moisture
Content Content
1.2 3.3
0.7 2.5
1.3 3.6
0.5 3.1
1.3 4.3
0.4 2.7
Potatoes
Percent
Other
11.1
9.8
32.6
8.8
11.2
6.8
Percent Percent
Pounds Protein Moisture Agtron
Weight Content Content Color
24.0 8.0
2.5 17.1
21.0 4.7
2.0 11.9
17.3 4.5
11.4 82.5
7.2 81.7
10.9 86.7
6.9 84.8
8.2 90.8
-------
Table 5
Effect of Drying at 82°C. and 121°C. on Classification of Potatoes
DRIED AT 82 °C.
Classifier
Settings
Grind Feed Fin Classified
Cycle RPM Gate Setting Product
0 14,000 - - Starting
Protein
1 14,000 15 25°
Starch
Protein
2 19,000 10 25°
Starch
DRIED AT 121°
0 14,000 - - Starting
Protein
1 14,000 15 25°
Starch
Protein
2 19,000 20 25°
Starch
Pounds
Weight
26.3
4.5
21.5
2.3
18.3
c.
25.0
3.0
21.8
3.8
17.3
Percent
Protein
Content
6.8
14.2
5.1
12.7
4.4
6.6
14.1
5.7
10.7
5.0
Percent
Moisture
Content
6.6
5.1
7.7
4.3
6.1
4.6
3.4
4.8
3.6
4.1
Agtron
Color
77.0
79.5
79.8
79.8
84.9
64.9
70.2
64.6
75.9
72.8
-------
STARCH
FRACTION
CENTRIFUGE
L1 *
L|-CENTRIFUGE
I 1
WATER —UMIXER
' j
L FILTER
DRYER
STARCH
FOR WASTE
TREATMENT
Figure 1, .Production of Potato Starch Without Waste
176
-------
Initial mixing required 3 parts of Liquid to 1 part of starch fraction,
while subsequent steps required 2 parts of liquid. Material in the
brown starch tank contained white starch and pulp as well as dissolved
solids. This "brown starch" liquid was screened in such a fashion that
the white starch and half the liquid was returned to the system, the
pulp and remaining liquid being withdrawn. The amounts of protein in
the washed starch and solids in the wash liquid were determined at each
step. Typical data are presented in Table 6.
Table 6 Typical Analysis of Protein in Washed Starch and Solids in
Wash Water in Countercurrent Washing
Solids
Protein in Washwater
Percent Percent
4th washing .99 .2
3rd washing 1.07 1.0
2nd washing 1.48 2.9
1st washing 1.98 9.1
Initial 4.5
In another experiment the final wash water had a total solids concentra-
tion of 10.85 percent. The analysis of these solids is presented in
Table 7.
Table 7 Analysis of Solids in Final Wash Water
Total Solids 10.85%
Protein (Kjeldahl X 6.25) 28.5
Protein (Biuret Reagent) (4.6)
Free Amino Compounds (as asparagine
using Stein-Moore Reagent) (20.4)
Amino Compounds
Automatic Amino Analyser (18.8)
Total Reducing Sugars 12.4
Total Organic Acid 35.4
Potassium (Flame Photometer) 17.7
The starch fractions of material dried at 82° and 121° were similarly
washed (Table 5). That from the higher drying temperature contained a
partially gelatinized material which required 5 volumes of water to
make into a workable slurry. The gelatinous material was removed with
the brown starch.
177
-------
Brabender aroylograph curves were calculated to determine if the starch
had been heat-damaged and are shown in Figure 2. The temperature of
initial gelatinization was essentially the same for all of the starches
except the commercial. This starch had a slightly higher initial tem-
perature of gelatinization. From an examination of Figure 2 it can also
be observed that the commercial starch granules swell more readily than
do those of the other starches. The highest peak was obtained from
starch dried at 82° and the lowest point was obtained from the Grafton
commercial starch. Variations in peak height amongst the different
starches were observed. A definite breakdown was observed for all of
the starches during the 95° hold period although the breakdown was not
the same in all cases. Likewise, the height at 50° varied somewhat for
the different starches.
An experiment was designed to maximize yields of starch and to deter-
mine the efficiency of a process employing fine grinding, air class-
ification, sieving, and washing. Data obtained from previous experi-
ments indicated that, with the proper combination of the various treat-
ments, it was possible to obtain an acceptable potato starch that would
require considerably less wash water for removal of protein and other
solids.
The potatoes were dehydrated at 60° and put through the pin mill at
19,000 RPM. Screening through an 88-micron screen best removed brown
particles. Material passing through the 88-micron screen was classified
and the starch fraction reclassified until 10-15 percent of the stream
was removed as protein fraction.
Removal of some brown particles from the starch fraction was sieved
with a 53-micron screen and 53/44-micron screening removed a small amount
of inadequately ground particles. Normally, this 53/44 fraction would
be reclaimed by recycling through the mill. The -44-micron material
was used for starch washing steps.
The final, screened starch fraction was then washed by the counter-
current procedure of Figure 1, including the recovery of white starch
in the brown starch fraction. The recovered starch was dried and weighed.
This experiment is summarized in Figure 3.
The final wash water of 10-11 percent solids represents a tenfold con-
centration over normal waste effluent of a conventional wet-milling
potato starch plant, and thus a 90 percent reduction in volume of effluent,
Such a waste effluent could be drum-dried. deKoe' ', working with 1
percent solids effluent from a European starch plant, recommended spray-
drying after preconcentration. Effluent of 10-11 percent solids should
further enhance his suggestion.
Strolle, e_t alA' discussed de-proteinization of simulated starch waste
prior to ion exchange and showed that increasing total solids from 1.5
percent to 2.8 percent significantly increased the efficiency of the
protein removal necessary for ion-exchange. Increasing solids to 10-
11 percnet should further increase efficiency of protein removal.
178
-------
1000
800 -
CO
600 -
oc
LJ
Q
UJ
CD
400-
oc
CD
200 -
0
COMMERCIAL
STARCH
W9ooocoooo
•••••••*
20 40 60
TIME, MINUTES
Fiaure 2. Production of Potato Starch Without Waste
179
-------
WASHED POTATOES
»
SULFITE ^-CUTTER
r" <*•• Ml LL
1 1
100 Ib. *
^ u DEHYDRATOR
8.9 % Prot.
-rl '<+> 4.5 Ib. (12.9% Protein)
1 i t"1 9.6 Ib. (20.7% Protein)
I CLASS
' — SCREE
r itrc
tl(^) 3.7lb (10.1 % Protein)
\
\
X.
• . H ^
< %
_ ^^. fcj 1 V IT Q
•— ^ M IX tn
I 1
(-) HYDRO (+) _ DRUM
^ SCREEN^ ^ DRYER
10.3 Ib. ^
\
1 DECANTING ^ ^_
1 CENTRIFUGE \
1 i FRESHWATER 167 Ib N
• I 4,11" -r- Ist CYCLE WATER—*.
, f " " " —L|St. CUT BROWN ».
ii i no. ^
• PMRIFIPR "^"^rx CUT BROWN ^
| i ruNiMtK _L-,rd.^,,T ODrtultl fc,
• 1
1 |2ndCYCLE
I . WATER
]L.J j
3rdCYCLE i
WATER
L.J
i
«
N.
BROWN
STARCH
TANK
' FILTER
n c u v
>%>> * f * r* .^ UIATCD
i
STARCH
67.5 Ib.(0.4% Protein)
Fiaure 3. Production of Potato Starch Without Waste
180
FEED
28.9 Ib.
-------
CONCLUSION
On a laboratory scale, the process of drying, fine grinding, air class-
ification, and washing can be fairly efficient. The unit processes are
straightforward and use in a larger operation should be feasible.
Use of the drum dryer for wash-water drying would not involve a waste
effluent. Alternately, the liquid could be added to previously sep-
arated protein fraction and redried in a pulp dryer. The animal feed
fraction obtained from the process would have a protein content of
15-20 percent compared to the 6-8 percent protein in the dried potato
pulp obtained from a conventional potato starch plant.
If recovery of valuable by-products by means of ion-exchange appears to
be attractive, the amount of waste effluent for disposal would be only
a fraction of that from a wet-grinding starch plant. Even without
drying or ion-exchange, the volume of waste is only 10 percent and
the quantity of solids only 40 percent that of a conventional plant.
181
-------
LITERATURE CITED
1. OLSON, 0. 0., VAN HEUVELEN, W., and VENNES, J. W. Combined indus-
trial and domestic waste treatment in waste stabilization lagoons.
J. Water Pollution Control Federation 40: 214 (1968).
2. CARANSA, A. Von neue Erfahrungen mit Dorr-Oliver Multizyklonen und
D-6 DSM Bogensleben in der Kartoffelstarke-Industrie. Die Starke
22: 27 (1970).
3. HEISLER, E. G., SICILIANO, J., TREADWAY, R. M., and WOODWARD, C.
F. Recovery of free amino compounds from potato starch processing
water by use of ion exchange. Amer. Potato J. 36: 1 (1959).
4. deKOE, W. J. Protein recovery from potato starch mill effluent.
Water and Waste Treat. J. 12: 55 (1968).
5. GRAHAM, J. C. The use of air classifiers in the flour milling
industry. Milling 144: 215 (1965).
6. BEHRENS, D. Neuere Fein-und Feinstaprailmuhlen. Die Muhle 100:
3 (1964).
7. GRACZA, R. The subsieve-size fractions of a soft wheat flour
produced by air classification. Cereal Chem. 36: 465 (1959).
8. PEPLINSKI, A. J., BURBRIDGE, L. H., and PFEIFER, V. F. Air
classification of leading varieties in U. S. wheat classes by
standardized fractionation procedure. Amer. Miller and Processor
93: 7 (1965).
9. KENT, N. L. Effect of moisture content of wheat and flour on
endosperm breakdown and protein displacement. Cereal Chem. 42:
125 (1965).
10. ASHWELL, G. Colorimetric analysis of sugars. In "Methods of
Enzymology", Vol. Ill, p. 73. S. P. Colowick and N. 0. Kaplan,
Eds., Academic Press, New York (1957).
11. NELSON, N. A photometric adaptation of the Somogyi method for
determination of glucose. J. Biol. Chem. 153: 375 (1944).
12. ASSOC. OFFIC. AGR. CHEMISTS. Methods of analysis, 9th ed., 643:
items 38.009 and 38.011 (1960).
13. PFEIFER, V. F. and GRIFFIN, E. L., JR. Fractionation of soft and
hard wheat flours by fine grinding and air classification. Amer.
Miller and Processor 88: 15 (1960).
14. STROLLE, E. 0., CORDING, J., JR., ACETO, -N. C. and DELLA MONICA,
E. S. Recovering proteins from potato starch factory effluents.
Presented at the 20th Natl. Potato Utiliz. Conf., Riverside,
Calif., July 31, 1970.
182
-------
ACKNOWLEDGMENTS
We thank Dr. B. D'Appalonia, North Dakota State University, for the
starch amylograms; Mr. E. G. Heisler, Research Chemist, Eastern
Marketing and Nutrition Research Division, Philadelphia, Pennsylvania,
for analysis of waste water; and Messrs. R. Maneval, Crops Research
Division, ARS, U.S.D.A., Fargo, North Dakota, and Gerald A. Baumann,
Red River Valley Potato Processing Laboratory, East Grand Forks,
Minnesota, for technical assistance.
183
-------
ECONOMIC ANALYSIS OF ALTERNATIVE METHODS FOR
PROCESSING POTATO STARCH PLANT EFFLUENTS
by
R. L. Stabile, V. A. Turkot and N C. Aceto*
ABSTRACT
The results of a preliminary economic analysis of alternative methods
of treating potato starch plant waste effluent in order to conform to
present government pollution regulations are presented. The alterna-
tives discussed consist of one biological treatment method and four
methods involving recovery of one or more by-products. All of the
alternatives involve considerable capital investment and operating
costs. The only alternative that appears economically feasible at
this time is concentration of the protein (or fruit) water by
multistage evaporation and use of the concentrate in animal feed.
SUMMARY
Potato starch processing plants must now treat their waste water ef-
fluents beyond primary treatment in order to conform to government
regulations on pollution concentration limits and water quality.
Meeting these regulations requires considerable capital investment
and considerably increased operating costs as a result of the instal-
lation of equipment and facilities needed to treat the waste water
for removal of pollutants. Preliminary capital and operating cost
estimates were made on five possible alternatives in order to de-
termine which were economically feasible and worthy of further study.
The alternatives included one conventional type biological treatment
method plus four processes yielding by-products, as follows: 1. bio-
logical treatment, 2. protein recovery with biological treatment,
3. concentration by evaporation, 4. protein recovery and concentration
of protein-free waste, 5. protein recovery, ion-exchange and biological
treatment. The estimates were based, in part, on laboratory and pilot
plant data for Alternatives 2, 3, 4 and 5; Alternative 1 was based on
data from the literature. Only one alternative - concentration of
the protein water by multistage evaporation - appears commercially
feasible at this time. The concentrate would be used in cattle or
poultry feed. This process is under futher study at our Laboratory,
and the results of that study will be published.
^Eastern Regional Research Laboratory, Agricultural Research Service,
U. S. Department of Agriculture, Philadelphia, Pennsylvania.
185
-------
INTRODUCTION
The present requirements for the limits of pollutants in processing
plant waste water are necessary in order to preserve what is considered
good water quality for natural streams. The effect of these require-
ments on potato starch plant operations is to place an economic burden
on an industry which is already only marginally profitable. In present
starch-making technology, practically all of the soluble components
of the potato are released into the plant waste water. This yields an
effluent which has both a high BOD level and a large daily flow. These
two characteristics result in high sewage charges, that is, assuming
the local sewage plant will accept the discharge at all. If the starch
plant must build its own biological waste treatment plant, the costs -
both capital and operating - will be substantial. An alternative
method of treating the effluent is to recover the water-soluble con-
stituents from the waste stream in usable form and sell them as by-
products. However, since the waste stream is dilute, recovery pro-
cesses will have high operating costs and require considerable capital
investment in relation to existing starch plant valuation. Nevertheless,
a by-product recovery process could be justified if the selling price
for the product or products resulted in a reasonable return on invest-
ment .
A preliminary economic evaluation of several potential waste treatment
processes was made in order to see if any were commercially feasible.
Since little pilot plant data had been obtained, this evaluation was
primarily to see which process or processes deserved further inves-
tigation on the pilot plant scale. The estimates are based on
treating the waste water from a 30 ton per day starch plant operating
16 hours per day, 150 days per year. Five alternatives are compared.
One alternative considers biological treatment of the waste with no
recovery of the components of the waste as by-products. The other
four alternatives involve recovery of waste components and yield one
or more by-products. The waste stream was considered to come from a
starch plant using current technology for starch recovery. The pro-
tein water waste flow used as a basis was approximately 104,000 gallons
per day at a 2 percent by-weight dissolved solids concentration. This
basis was considered an approximation of average starch plant opera-
tion using some improved water utilization.
Briefly, the five alternatives compared are: 1. biological treatment,
2. protein recovery with biological treatment, 3. concentration by
evaporation, 4. protein recovery and concentration of protein-free
waste, 5. protein recovery, ion-exchange and biological treatment.
Let us look at the technology involved in each of the alternatives.
Alternative 1 Biological Treatment of Waste.
The treatment requirements for the waste treatment plant include at
least 85 percent removal of BOD and suspended solids and complete
removal of floatable and settleable solids. Other factors that must
be considered for each plant location are the effect of the effluent
discharge on the dissolved oxygen content of the receiving river or
186
-------
stream and the necessity for disinfection by chlorination. Chlorination
is probably not required for starch plant waste treatment because path-
ogens are not present in the source of the waste.
The biological waste treatment process chosen is the activated sludge
type. In this process, shown in Figure 1, the raw waste is screened
to remove fibrous solids which are used as animal feed. The liquid
containing suspended and dissolved solids is fed to the primary
clarifier where settleable solids are removed. The solids removed
from the primary clarifier are used in animal feed. The overflow from
the primary clarifier goes to the aeration tank where biological de-
gradation of the waste occurs. The effluent from the aeration tank is
pumped to the secondary clarifier where the biologically active sludge
is settled. Part of this sludge is returned to the aeration tank; the
remainder passes to a digester where it is converted into gases and
final solids by biological action. Final solids from the digester are
assumed to be disposed of by land fill.
An activated sludge treatment system was selected because the design,
operating procedures, and costs of this type system are well known, and
also because it will consistently give high BOD removal.
The plant is assumed to process 625,000 gallons per day of waste water
containing 11,320 Ibs of BOD and 8,000 Ibs of suspended solids. This
total volume includes the combined protein water, wash waters from
purifying the starch, and water used to flume and wash the raw potatoes.
The treated waste sent to the river is approximately 600,000 gallons
per day, containing 1,100 Ibs of BOD and 1,000 Ibs of suspended solids.
Solids are removed from the screen and primary clarifier at a rate of
about 5,000 Ibs per day.
Alternative 2 Protein Recovery and Biological Treatment.
Protein recovery from protein water has been investigated by Strolle
on a pilot plant scale'^-'. The results of this investigation were
used to design a full scale plant. The process is shown in Figure 2.
The protein water effluent from the starch plant is preheated, using
the heated protein water from the steam injection heater, in a plate
type exchanger. After preheating, sulfuric acid is added to an agitated
tank which feeds the steam injection heater. The pH of the protein
water is lowerd to approximately 3.5, and the exit temperature from
the heater is 210°F. The precipitated proteins are removed from the
slurry using a continuous rotary filter. The wet protein solids, con-
taining about 87 percent water, are dried on a double drum dryer to
about 5 percent moisture. The dried cake is ground and packed in 100
pound bags.
After de-proteinization the waste stream is sent to a biological treat-
ment process for removal of 80 percent of the remaining BOD, to give a
final BOD of about 1,100 Ibs per day. The biological process would
be the same as described under Alternative 1 except that incoming BOD
would be reduced to 8,300 Ibs per day because of the removal of the
protein. Flow would be almost the same as Alternative 1, and there
would be no suspended solids. Therefore, a lower cost would be incurred.
187
-------
RAW WASTE
0.625 MGD
11320 PPD BOD
8000PPD SS
AIR
SCREEN
H
PRIMARY
CLARIFIER
COARSE
SOLIDS
SOLIDS
oo
oo
AERATION
TANK
SECONDARY
CLARIFIER
RECYCLE SLUDGE
n
5000 PPD
EXCESS SLUDGE
5-6% SOLIDS
TREATED
WASTE
0.600 MGD
1100 PPD BOD
1000 PPD SS
ANIMAL
FEED
9.3 TPD
SOLIDS
DIGESTER
POTATO
PULP
13600 PPD
SOLIDS
Figure 1. Activated Sludge System (Alternative 1).
SOLIDS
TO LAND
FILL
GASES
-------
PROTEIN
WATER
104,000 GPD
98%
00
\0
I40°F
PREHEATER
1580 PPD
I40°F
STEAM
60,200 PPD
w
210 F
JET MIXING
HEATERI
CONTINUOUS
FILTER
DEPROTEINIZED LIQUID
108.000 GPD
16,400 PPD SOLIDS
WET SOLIDS
31,500 PPD
STEAM
33,000 PPD
DRUM
DRYER
DRIED
CAKE
GRINDER
BAGGING
PROTEIN
PRODUCT
50-6070 PROTEIN
2,460 PPD "
Figure 2. Protein Recovery from Waste Water (Alternative 2).
-------
Alternative 3 Concentration by Evaporation
The basis for studying this process was to determine if it would be
feasible to concentrate the entire protein water stream and make a
profit by selling the concentrate for feed use.
The protein water is evaporated in a triple effect evaporator to a 60
percent solids slurry. The slurry is mixed with the dried potato pulp
from the starch process and the mixture is sold as animal feed (Figure 3).
Approximately 320,000 pounds per day of steam would be required for the
evaporation.
The capital and operating costs are based on the evaporation step only.
The cost for pulp drying is considered to be recovered through the
sales value of the pulp constituent in the mixed feed product.
Alternative 4 Protein Recovery and Concentration of Protein Free Waste.
This alternative was investigated because it was anticipated that a
market might exist for both the protein and protein-free solids.
Figure 4 shows a schematic flow sheet of the process. The protein water
goes first to the protein recovery process as described under Alternative
2. The de-proteinized liquid is evaporated in a triple effect evapora-
tor as described under Alternative 3.
Alternative 5 Protein Recovery, Ion-Exchange and Biological Treatment.
This alternative consists of the combination of the protein recovery
process already described under Alternative 2, an ion-exchange process
which recovers potassium salts, amino acids and organic acids (both as
ammonium salts), and the biological treatment process described under
Alternative 1. Figure 5 shows these three sections of the process
combined to form this alternative.
The protein water is first sent to the protein recovery process' ', It
is necessary to remove the proteins first because they will precipitate
on the ion-exchange columns if their concentration is 180 ppm or more.
After protein removal, the waste is treated by the ion-exchange columns
which remove mainly potassium ions, amino acids and organic acids from
the waste stream(3,4,5). A final biological treatment removes most of
the remaining dissolved solids, which are chiefly sugars.
Figures 6, 7 and 8 are schematic flow sheets showing the processing
involved in the ion-exchange part of this alternative. In each of the
three sets of ion-exchange columns, two columns are in series at one
time performing adsorption and one column is being eluted and regenerated.
Also, each by-product solution resulting from ion-exchange is evaporated
to 60 percent concentration before drum drying to 4 percent moisture.
Figure 6 shows the potassium ion removal by cation exchangers and recovery
of solids by evaporation. After absorption and elution with sulfuric acid,
an acidic solution is obtained which is neutralized with ammonia. The
potassium and ammonium sulfate solution is evaporated and dried yielding
the mixed salt solid. 190
-------
VO
PROTEIN WATER
104,000 GPD
18,380 PPD SOLIDS
STEAM
3I9,000"PPD
TRIPLE
EFFECT
EVAPORATION
n
CONDENSATE
TO REUSE
OR WASTE
60% SOLIDS
CONCENTRATE
30,600 PPD
SOLIDS FROM
STARCH PROCESS
9.3TPD SOLIDS
DRYING
BLENDING
ANIMAL FEED
MIXTURE
I8.5TPD SOLIDS
27TPD MIXTURE
Figure 3. Protein Water Concentration bv Evaporation (Alternative 3).
-------
vo
PROTEIN
WATER
.104 MGD
18,400 PPD
SOLIDS
STEAM
HZS04
PROTEIN
RECOVERY
PROCESS
PROTEIN
— FREE
LIQUID
STEAM
CONDENSATE
WATER TO
REUSE OR
WASTE
TRIPLE
EFFECT
EVAPORATION
PROTEIN PRODUCT
2460 PPD PROTEIN
4100 PPD SOLIDS
PROTEIN FREE
CONCENTRATE
60% SOLIDS
16,000 PPD SOLIDS
Figure 4. Protein Recovery and Concentration of Protein-Free Waste (Alterative 4)
-------
PROTEIN
WATER
104,000
GPD-t
—STEAM
PROTEIN
RECOVERY
NH.
108000.
6PO
PROTEIN PRODUCT
2460 PPD PROTEIN
4100 PPD SOLIDS
SALT MIXTURE
K2S04
-------
PROTEIN
FREE
WATER
108,000 6PD
—WATER
-H2S04
7060 PPD
NH
1500 PPD
H-
vo
CATION
EXCHANGERS
(3)
DILUTE
SOL'NS.
STEAM
6500 PPH
_
1
NEUTRALIZATION
EVAPORATION
8 DRYING
110,000 GPD
TO AMINO ACID
REMOVAL COLUMNS
MIXTURE:
K,
4800 PPD
. 5880 PPD
Figure 6, Potassium Ion Removal - Ion Exchange Process (Alternative 5).
-------
H2 S04
5900 PPD
WATER
POTASSIUM FREE
PROTEIN WATER
110.000 6PD
I r
NH OH
20OO
CATION
EXCHANGERS
(3)
vO
Ul
J
PPD(NH3)
TO ORGANIC ACID
COLUMNS
112,000 GPD
DILUTE
SOL'NS.
SURGE TANK
1
NH3
220 PPD
NEUTRALIZATION
EVAPORATION
8 DRYING
STEAM
3700 PPH
7400 PPH
AMINO ACID MIXTURE
4680 PPD (AS NHjSALTS)
SOLIDS (200 LBS. NH3>
I
EVAPORATION
a DRYING
(NH4)2S04
7910 PPD SOLIDS
Figure 7. Amino Acid Removal - Ion Exchange Process (Alternative 5).
-------
WATER
FROM AMINO AClD
REMOVAL COLUMNS
112,000 GPD
9360 PPD SOLIDS
VO
STEA
5100 PPH
M J
»PH H
NH4 OH
2480 PPD NH
1
ANION
EXCHANGERS
(3)
i
L
SURGE TANK
I
EVAPORATION
a DRYING
TO BIOLOGICAL
TREATMENT
112,000 SPD
4200 GPD SOLIDS
DILUTE
SOL'NS.
ORGANIC ACID MIXTURE
7130 PPD (AS NH4+SALTS)
SOLIDS (2000 LBS. NH3)
.Figure 8. Organic Acid Removal - Ion Exchange Process (Alternative 5).
-------
Figure 7 shows the amino acid removal from the potassium-free stream
by cation exchangers. The adsorbed amino acids are eluted using am-
monium hydroxide. The columns are regenerated with sulfuric acid.
The amino acid and ammonium sulfate solutions are evaporated and dried.
The amino acids are obtained as the ammonium salts.
Figure 8 shows the organic acid removal from the amino acid free stream
by anion exchangers. The adsorbed acids are eluted using ammonium
hydroxide solution. The organic acid solution is evaporated and dried
to obtain the ammonium salts of the organic acids.
Costs
Capital and operating costs were calculated for each of the alternatives
in order to determine which ones should receive further study.
Table I lists the capital costs of the alternatives in order of in-
creasing fixed capital. Concentration of the protein water by
evaporation requires the least fixed capital with biological treat-
ment next. Alternatives 2 and 4 are next, both requiring almost the
same investment; alternative 4 being higher by about 10 percent.
Alternative 5 requires an investment that is outside the range of the
other four alternatives at $2,550,000.
Table 1 Fixed Capital Costs
Alt. No. Alternative Fixed Capital-$
3 Concentration by Evaporation 514,000
1 Biological Treatment 550,000
2 Protein Recovery 4- Bio. Treatment
(Protein Rec. = $382,000) 807,000
4 Protein Rec. + Cone, by
Evaporation 881,000
5 Protein Rec. + Ion Exch. + Bio.
(Bio. = $350,000) 2,550,000
Table II shows the operating costs for the alternatives, again in the
order of increasing costs. The biological treatment process involves
the least operating cost. Among the by-product alternatives, concen-
tration by evaporation incurs the Least operating cost. The difference
in operating cost between each consecutive alternative, as listed, is
considerable, and much greater than, the probable error involved in es-
timating the costs in this category.
Table III shows the uses and estimated probable sales prices for the
products obtained from the alternatives. The protein would be used for
animal feed or possibly human food. The concentrate from the concen-
tration by evaporation step is mixed with the potato pulp from the
starch process in the proportions in which they are produced, and used
as cattle or poultry feed. The concentrate without protein, Alterna-
tive 4, is also used as a feed additive - the price for this concentrate
197
-------
Alt. No.
Daily
Deprec.
Op. Ex.
Total
Yearly
Bio.
238*
367
605
Table II
3
Cone .
227**
764
991
34,000
114,500
Operating
2
Costs
4
Prot. Rec. Prot. Rec.
+ Bio. + Cone.
347
974
1,321
52,100
146,000
383
1,494
1,877
57,500
224,000
5
Prot. Rec. 4- Ion
Exch. + Bio.
1,538
3,500
5,038
230,750
525,000
Total 90,700 148,500 198,100 281,500 755,750
*Bio, Amortized at 7 percent for 20 years.
**Depreciation for Alts. 2, 3, 4, 5: Str. Line, Bldgs-20 yrs. , Equip.-15 yrs.
Table III
Uses and Estimated Prices for Products
Alt. No.
1
2
Alternative
Bio
Protein Recov.
+ Bio.
Cone, by Evap.
Product
None
Protein
Concentrate with
Protein
Protein Recov. + A. Protein
Cone, by Evap. B. Concentrate
Protein Recov. +
Ion Exch. + Bio.
without Protein
A. Protein
B. Amino Acid Mixt.
C. Organic Acid Mixt.
D. K2S04-(NH4)2S04
E. (NH4)2S04
Use
Feed or Food
Animal Feed
Feed or Food
Feed
Feed
Feed or Food
Beverages
Fertilizer
Fertilizer
Price-^/lb
12.0
6.7
12.0
5.0
12.0
15.0
31.0
2.2
2.25
being lower than the concentrate with protein. The aroino acid mixture
could be used in feed or food. The price for this was estimated from
current prices for amino acids. The organic acid mixture would be used
in beverages as an acidulant. The price is considered comparable to
similar acid mixtures used for this purpose.
The potassium sulfate-ammonium sulfate mixture and the ammonium sulfate
salts would be used as fertilizer. The prices were estimated from current
prices for these chemicals.
198
-------
Table IV shows the daily and yearly sales using the prices for the
products shown in Table III. The alternatives are listed in order of
decreasing sales dollars. The differences in sales between alterna-
tives as listed is probably greater than error in estimating the sales
figures. Total operating costs from Table II are subtracted from sales
to give the gross income as shown in Table V.
Alt. No.
5
3
4
2
1
Table IV
Alternative
Sales
Daily-$ Yearly-$
Protein Rec. + Ion Exch. + Bio.
Concentration by Evaporation
Protein Recovery + Cone, by Evap.
Protein Recovery + Bio. Treatment
Biological Treatment
2960
1244
1100
295
NONE
444,000
186,700
165,000
44,000
NONE
It should be noted that the operating expense figures do not include an
allowance for a return on the investment. Therefore, federal income
taxes are not inlcuded in the operating expenses. Thus, the net income
after taxes for Alternative 3 would be reduced by an amount equal to the
federal income tax. Also, the loss shown for the other alternatives
would reduce the overall federal income tax of the company by an amount
equal to the loss shown times the tax rate.
Table V shows the gross income or loss for each alternative in order of
decreasing income (or increasing loss). Alternative 3, Concentration by
Evaporation, shows the highest gross income, by far, of all the alterna-
tives listed. Here again, the difference between the figures is greater
than error in calculating the figures shown. After the alternative of
concentration by evaporation, the biological treatment process has a
smaller loss than the remaining by-product recovery processes. From
Table V it is apparent that concentration by evaporation offers the only
possibility for making an income, assuming the estimated selling prices
for the various by-products are reasonably correct.
Alt. No.
3
1
4
2
5
Table V Gross Income or (Loss)
Alternative Daily-$ Yearly-$
Concentration by Evaporation
Biological Treatment
Protein Recovery + Cone, by Evap.
Protein Recovery + Bio. Treatment
Protein Recovery + Ion Exch. + Bio.
255
(605)*
(777)
(1026)
(2078)
38,200
(90,700)
(116,500)
(153,800)
(311,750)
*Note: Parenthesis indicate loss.
Table VI, which shows selling prices for various levels of profitability,
can be used to compare the effects of different prices on the commerical
feasibility of the alternative.
199
-------
Table VI
Alternative
Selling Prices for Various Levels of Profitability
Product Product Selling Price,
Est.
To Equal To Break Market
Bio. Loss Even Price
1. Bio
2. Protein Rec.
+ Bio.
3. Cone, by Evap.
4. Protein Recov. +
Cone, by Evap.
5. Protein Recov.
+ Ion Exch.
+ Bio.
None
Protein
Concentrate with
Protein
A. Protein
B. Cone. W/0 Protein
A. Protein
B. Amino Acids
C. Organic Acids
D. K + NH4 Salts
E. (NH4)2S04
29.1
2.07
13.9
5.8
30.2
37.7
31.0
2.20
2.25
53.7
5.3
12.0
6.7
35.1
6.3
37.7
47.1
31.0
2.20
2.25
12.0
5.0
12.0
15.0
31.0
2.20
2.25
For Alternative 2, the Table shows that the protein must be worth almost
5kt per pound for no loss to occur. This is greater than the estimated
selling price of 12^ per pound. This means that, at current prices for
protein, the process is not commerically feasible. In contrast, Alterna-
tive 3 has a break-even price lower than the estimated price of 6.7^ per
pound. Of course, any price between 5.3*5, the break-even price, and 6.7j£
will involve no loss or some income will be earned. Thus, there is a
likelihood for income for Alternative 3 with current prices for feed.
The nutritive value of the feed consisting of pulp mixed with protein
concentrate was estimated at 20 percent above that of corn at $47 per ton.
On this basis the concentrate alone was estimated to be worth $73.50 per
ton at 60 percent solids or 6.7^ per pound of moisture-free solids. It
is possible that a higher price could be obtained if the concentrate
were fed to non-ruminant animals since the protein is of excellent qualityC2),
Under Alternative 5, only the prices for protein and the amino .acid
mixture were varied in order to obtain the additional income required for
the condition of "break-even" and for the condition of "loss-equal-to-
biological-treatment".
Table VI also shows the selling prices of the products for the condition
where the gross loss for the alternative would equal the loss for the
biological treatment alternative. There is only one alternative shown
where the estimated actual market price exceeds both the break-even price
and the loss-equivalent-to-biological-treatment price, and that alterna-
tive is concentration by evaporation.
CONCLUSIONS
We have seen the results of a preliminary economic evaluation of a number
of possible methods of treating the waste effluent from potato starch plants.
Conventional biological treatment of the waste water appears to have both
rather high capital cost and rather high operating cost.
200
-------
Four of the treatment processes yield products, and revenue from the
sale of these products would help offset the operating costs. Only one
of these processes, however, appears economically feasible at this
time; namely, concentration of the effluent by evaporation. Our Lab-
oratory is, therefore, investigating this process on the pilot plant
scale. These studies will enable us to project commercial feasibility
with greater confidence and also make available samples of the product
for testing and evaluation. When further information on this process
has been obtained, we will publicize the results.
Development of the protein recovery process on a pilot plant scale was
carried out in our Engineering and Development Laboratory by E. 0.
Strolle. The ion-exchange processes for recovery of inorganics, amino
acids, and organic acids were carried through the laboratory scale by
E. G. Heisler, James Siciliano and Joseph Schwartz of our Plant Products
Laboratory. Much of the basic data needed for process design in our
economic analysis were provided by these individuals.
201
-------
REFERENCES
1. "Recovering Proteins from Potato Starch Factory Effluents - Progress
and Prospects", E. 0. Strolle, J. Cording, Jr., N. C. Aceto, and
E. S. DellaMonica, EMNRD, ARS, talk presented at 20th National
Potato Utilization Conference, Riverside, California, July 31, 1970.
2. Unpublished report. J. W. White, Jr.., Eastern Marketing and Nutrition
Research Division, ARS, U. S. Department of Agriculture, Philadelphia,
Pennsylvania 19118.
3. "Recovery of Free Amino Compounds from Potato Starch Processing
Water by Use of Ion Exchange", E. G. Heisler, J. Siciliano, R.
H. Treadway, C. F. Woodward, American Potato Journal, January
1959, Vol. 36, No. 1, pp. 1-11.
4. "Recovery of Free Amino Compounds from Potato Starch Processing
Water by Use of Ion Exchange. II. Large-Scale Laboratory Experi-
mentation", E. G. Heisler, J. Siciliano, R,, H. Treadway, and C. F.
Woodward, American Potato Journal, February 1962, Vol. 39, No. 2,
pp. 78-82.
5. "Potato Starch Factory Waste Effluents. I. Recovery of Potassium
and Other Inorganic Cations", E. G,, Heisler, S. Krulick, J.
Siciliano, W. L. Porter, and J. W. White, Jr., American Potato
Journal. Vol. 47, No. 9, September 1970, pp. 326-336.
202
-------
CONTINUOUS TREATMENT OF CORN AND PEA
PROCESSING WASTE WATER WITH FUWGI IMPERFECT!***
by
Dr. Brooks D. Church*, Harold A. Nash*, Eugene E. Erickson* and
Willard Brosz**
INTRODUCTION
Strains of fungi from the IMPERFECT class were examined for their
utility in digesting food processing wastes because of their reported
ability to grow well on a variety of polysaccharide-containing
materials (1) and because of Gray's findings (2) that the mycelium had
a high content of good quality protein. It was reasoned that the
macroscopic size of the mycelium would simplify its recovery. Further,
its high protein content would make it valuable as an animal feed
and thus allow recovery of at least part of the cost of the waste
treatment.
Laboratory studies reported one year agoO) showed that these hopes
for both efficient BOD removal and high quality protein recovery
might indeed be realized. The work was therefore extended from the
5-gallon laboratory stage to two 10,000-gallon pilot plant units at
the Green Giant Company plant in Glencoe, Minnesota. The present
report treats these pilot-scale studies, which were carried out on
corn and pea canning wastes and corn silage waste.
METHODS
The corn and pea waste waters used in these pilot-plant investigations
were taken from a receiving lagoon at a point near the plant effluent
discharge.
The fungus selected for these studies was JuLchod&ma. vJJii&Q, 1-23.
The fungal choice was based on the previous laboratory screening
studies where the criteria used were COD reduction, growth response,
mycelial yield, quality of protein and the ability to compete with the
natural flora. Two types of aeration systems were used in the pilot
plants: a 2-hp floating aerator was installed in a 10,000-gallon
*North Star Research and Development Institute, Minneapolis, Minnesota.
**The Green Giant Company, LeSueur, Minnesota.
* * *This investigation was supported by funds from the Environmental
Protection Agency, Water Quality Office, under Grant No. 12060 EDZ
and The Green Giant Company, National Canners Association, Wisconsin
Canners and Freezers Association, and the Minnesota Canners and
Freezers Association.
203
-------
plastic swimming pool, and a cage rotor was installed in a circular,
rubber-lined, 10,000-gallon ditch. In each facility the liquid depth
was approximately 3.5 feet. The waste was fed from a constant-head
tank; the flow rate, which controlled the detention time, was regulated
by varying the size of the discharge orifice in the tank. The effluents
from the lagoon and the ditch were discharged through an overflow tube.
A schematic diagram of the pilot-plant flow system is shown in
Figure 1; photographs of the facilities are shown in Figures 2 and 3.
A technician was at the pilot installations for eight hours per day
to make observations, take influent and effluent samples from both
the ditch and the pool, and collect some of the data (e.g., pH, DO,
temperature, feed rate, additions of chemicals).
Effluent samples were filtered through pre-weighed Whatman #4 filter
paper and the mycelium and filtrates were frozen, along with the
unfiltered influent or feed samples, for regular transport to, and
analysis in, the laboratory. Every third day samples of unfiltered
effluent were taken to the laboratory for phase-microscopic examina-
tion of the fungal mycelium.
From our microscopic experience we believe that regular examination
of the fungal morphology can serve as an excellent method to indicate
waste treatment efficiency. At times of start-up, the inoculum
was composed of young, rapidly-growing mycelium containing numerous
long growing tips. As the mass of the thallus increased, some
granulation and vacuolation occurred in the mycelium. During steady-
state digestion of waste nutrients, fungal growth is balanced by loss of
mycelium in the effluent. The continuous growth during steady-state
digestion is marked by numerous young growing elongated hyphal tips
and branches, as shown in Figure 4. A minimal number of lysing
hyphae and spores are observed. The actively metabolizing mycelium
has very fine homogeneously dispersed cytoplasmic structures. Because
the steady-state digestion is near starvation with regard to fungal
nutrient, any rapid and radical operational change (e.g.,stoppage
of feed, pH) will be reflected as early sporulation stages in the
thallus as seen in Figure 5. Sporulation develops as hyphal granula-
tion in the first observed change followed, in several hours, by the
formation of asexual spores in the hyphae and at the hyphal tips.
If an irregularity in the pilot operation can be detected within
eight hours, changes (granulation) can be reversed and
spores formed during this time will germinate (Figure 6) to form
new mycelium. If the irregularity is longer (e.g., 24 hours), much
sporulation will occur and 12 to 18 hours will be required before
optimal growth and efficient BOD removal is reestablished.
Inoculation of the pilot plant was accomplished with mycelium grown
from an initial potato dextrose agar stock culture of TVu.oAocteAma
u^u,de-I-23. The growth to pilot-plant inoculum scale was carried out
in 32-gallon plastic garbage cans. The inoculum was aerated by
passing air through plastic pipe inserts with multiple outlet holes.
204
-------
to
o
Ui
Ptont Waste
Sireom
PH
Controller
a
Recorder
PH
Controller
a
Recorder
(NH432S04
NaH2P04
Figure 1. Schematic Diagram of the Pilot-Plant Flow System
-------
Figure 2. Aerated Ditch Used for Treatment of
Corn Canning Waste
Figure 3. Aerated Pool Used for Treatment of Corn
Canning, Pea Canning, and Silage Wastes
206
-------
Figure 4. Young Growing T. vir>ide Showing
Active Metabolizing Hyphal Tips
207
-------
; t
Figure 5. Early Sporulation, Asexual Spores
and Cytoplasmic Granulation Can Be
Observed
208
-------
\
Figure 6. Germination of Spores of T. wiri.de
Note new hyphal structures
209
-------
The growth medium was ground frozen corn or peas, with appropriate
amounts of (NH^SO^ and Na2HP04. Adequate H2S04 was added to adjust
and to maintain the pH at 3.2 to 3.7. The total inoculum of 100
gallons was used for the pool or ditch containing 2000 gallons of
canning waste and 2500 gallons of water. Thus a 5-percent inoculum
was used with regard to the canning waste. New feed was started
immediately following inoculation.
The fungal mycelium was recovered from the pool and ditch effluent
late in the season by screening through a Sweco Model LC-18-C-333 filter
unit. Several mesh sizes were tried and the #120 size appeared most
efficient.
Measurements made in monitoring the fermentation included regular
determination of influent and effluent COD and BOD5, and occasional
measurement of TOC, temperature, pH, and mycelial mass were determined
regularly on the effluent. Microscopic examination of the fungal
growth was conducted at least three times a week. Phosphate determina-
tions were made by the Fiske and Subarow method, protein determinations
by the Lowry method, carbohydrate by micro-kjeldahl, and ammonium
nitrogen measurements by the Conway diffusion technique.™' Total,
dissolved, volatile, and suspended solids were determined according
to standard procedures.' '
RESULTS
1. Pilot-Plant Operation on Corn Waste
Some of the characteristics of the feed are indicated by measure-
ments made on a composite sample comprised of equal aliquots of
samples from the heart of the operating season from August 19 to
September 20. Results of sample analyses are shown in Table 1.
Table 1. Characteristics of Composite
Sample of Corn Waste*
"ft/.1
COD 2436
BOD 1564
TOC 1632
Total Solids 2372
Volatile Solids 1390
Ash 1070
Suspended Solids 210
Ammonia Nitrogen 90
Phosphate as P 8.5
Acid to titrate to pH 3.7 450
*At pH 6.9.
210
-------
Daily records were kept of the principal variables controlled
during the 50-day corn canning season. One of these, temperature,
remained at approximately 20°C at early afternoon samplings during
the bulk of the season. Temperature dipped to below 10°C on the
36th and 37th days of operation and at the end of the season.
The detention time was set to test the effect of variations in
waste treatment times, and was occasionally altered in response to
other events. It was deliberately long immediately after inoculation.
At day 21, the detention time in the ditch was increased for 2 days
because a failure of the pH control equipment allowed the pH to
drop to levels that impaired culture performance. The pool was
operated at longer detention times than desired during part of the
season because of partial failure of a feed pump. The detention times
together with the COD or BOD levels, governed the feed loading that
the system was required to handle each day. Best performance appeared
to occur at detention times between 22 and 26 hours.
Ammonium sulfate and sodium dihydrogen phosphate additions were
adjusted to provide amounts believed to be required from previous
experience. Alterations were made as feed rates were changed, as the
average COD changed, and as analyses of the effluent indicated that
excesses or supposed deficiencies were encountered. (Nlfy^SOA varied
between 3.1 and 5.4 Ibs per 1000 gallons of feed, and Na2HP04 between
0.1 and 2.3 Ibs per 1000 gallons.
The pH control was set at 3.5, and that pH was maintained within
+0.1 unit, with a few exceptions. A major exception in the case of
the ditch occurred on the twenty-first day, when the acid pump^ failed
to turn off and the pH dropped to 1.8. pH control failures also
occurred in the pool twice during the first 20 days of operation and
destroyed operations during this period. The amount of acid averaged
about 4 lbs/1000 gallons of feed. Titrations of the well water used
in the canning operations showed an acid requirement of about 2.9 Ibs/
1000 gallons to bring the pH to 3.5. Titrations of the plant waste
directly from the corn canning operations showed a similar requirement.
Titrations of plant waste drawn from the receiving lagoon, however,
showed a requirement of 6 Ibs of acid per 1000 gallons.
The performance of the ditch and pool in removing COD and BOD is
shown in Figures 7 through 10. COD removal during favorable periods
of operation of the ditch (for example, between days 25 and 36) was
about 89 percent. The high effluent COD on day 21 followed failure
of pH control with a fall of pH to 1.8. The peak on days 37 and 38
occurred when the temperature dropped to below 10°C. BOD removal
was about 97 percent during periods of favorable operation. BOD
levels in the effluent vacillated between about 40 and 90, with a
mean of about 50, except during upsets caused by extremes of acidity
or temperature. It is to be noted that the decreased COD removal
related to extreme acidity on day 21 is reflected by a peak in effluent
BOD also. The increase in COD levels in the effluent during the
211
-------
3500 -
3000-
2500-
,2000 -
3
o
8
1500
1000-
I
• I ' I
500-
10
20 30
Day of Operation
40
50 60
Figure 7. COD of Corn Feed Effluent Streans of the Ditch During
Operation on Corn Canning Wastes
212
-------
3000
2500
2000
o
£
CO
1000
500
Day of Operation
Figure 8. BOD of Feed and Effluent Streams of the Ditch During
Operation on Corn Canning Wastes
213
-------
3500
3000
2500
2000
A ,\
\ \
\ ! \
\
Feed
\
I
1 /N
I/
K
\
g 1500
o
1000
500
Effluent
A ,fx
V
10
20 30
Day of Operation
_JL
40
Firure 9. COD of Feed and Effluent Streams of Pool During Operation
on Corn Canning Wastes.
214
-------
2500
2000
P 1500
o
o
OD
1000
500
A
Feed
A
« A
0,
\
10
20
Effluent
40
50
Day of Operation
Figure 10. BOO of Feed and Effluent Streams of Pool
During Operation on Corn Canning Wastes
215
-------
period of low temperature on days 37 and 38 is scarcely at all
reflected by a corresponding increase in effluent BOD.
The yield of mycelium filtered from daily effluent samples through
Whatman No. 4 filter paper varied from about 0.55 gram to 0,8 g/1.
Values are shown graphically in Figure 11. The dry weight of
recovered mycelium averages 50 percent of the weight of BOD removed
for both the ditch and the pool. The amount of mycelium recovered
was relatively constant. The spikes from days 6 to 9 for the ditch
are thought to be the result of imperfect sampling procedures.
The microbial pattern in the aerated ditch remained predominately
Fiwg-t Zmpe^eett, but not at all time T. v-ot/tde. Yeasts and bacteria
were always present, but they formed less than 0.1 percent of the
total biomass. A fungus, indentified as a G£.otAA.cim, appeared in the
ditch a few days after inoculation with T. v
-------
1.5
liJ
- 1.0
a*
£X
0.5
o
Ditch
Pool
10
210 30
Day of Operation
40
Figure 11. Dry Weight of Solids Recovered from Effluent
Streams by Filtration During Operation on
Corn Canning Wastes
217
-------
The vacuum filter cake formed from fungal material collected on
the Sweco unit could itself be used as a filter media to clean up the
effluent from the Sweco. A 0.5-inch fungal cake accommodated 20
gallons of Sweco effluent through a 10-inch diameter filter at
about 24 inches of Hg pressure drop, before slowing to a filtration
rate of 0.2 gpm/sq ft.
The dissolved oxygen levels in the ditch never dropped below 3.7 mg/1;
the 5 hp rotor was overpowering. DO levels in the pool, however,
approached zero on several occasions, thus providing a basis for
estimating aeration requirements. If one assumes the floating aerator
in the pool delivered 2 Ib DO per hp-hr, 96 Ib would be delivered/day.
One three-day period (days 35-38), the unit was handling 120 Ib of
BOD/day at DO levels less than 1 mg/1. From the above values, an
oxygen consumption of 0.8 Ib per Ib of BOD is obtained.
Another approach, which we carried out a number of times, was to
measure the rate of change in DO levels in samples removed from the
ditch. We assumed that oxygen consumption of the ditch sample continued
in a closed system that was fed waste nutrient at the same rate of
feeding as in the ditch. Assuming BOD was being removed at the same
rate in the sample as in the ditch, the calculated rate was 0.7 Ib of
02 per Ib of BOD removed.
The overall effectiveness of the fungal digestion on corn waste is
shown in Table 3. The data used in preparing this table was taken
from a 20-day period of stablized favorable performance.
Table 3. Characteristics of Composite Samples of Corn Waste
Feed and Effluent Collected in Mid-Season
(Composites are from the same 20 days in each instance)
Milligrams per Liter
Feed
2436
1580
1608
2372
1490
1070
210
0
90
9
Effluent
305
59
104
1743
521
984
101*
646
23
1
Clarified
Effluent
215
41
81
1536
432
998
0
0
0.2
0.01
COD
BOD
TOG
Total solids
Volatile solids
Ash
Suspended solids
Mycelium
Ammonia N
P04E
*"Suspended solids" in this instance refers to materials not removed by
filtration but which could be collected by centrifugation at 5000 rpm.
218
-------
2. Pilot-Plant Operation on Pea Canning Waste
The aerated pool was operated on pea canning wastes for 35 days.
Temperatures were generally above 18°C. Detention times were
initially 45 hours and were stepped down to 18 hours by the end of
the season. Additions of (NH/^SCty and Na2HP(>4 in Ibs per 1000
gallons of feed were 1.8 to 4.4 and 0 to 0.7, respectively. An
average of 6.5 Ib of 112804 per 1000 gallons of feed was required
to maintain a pH of 3.5 Performance is summarized in Table 4.
These values represent analyses over 20 days of the operating season.
During the last 15 days of the season the effluent BOD was consistently
below 45.
Table 4. Analysis of Feed and Effluent
Pea Wastes
Milligrams per liter
Test Feed Effluent
COD 1650 324
BOD5 772 61
TOC 962 126
Total Solids 6815 6585
Volatile Solids 917 711
Ash 5898 5874
Suspended Solids 264 0.8
Mycelium 0 420
Phosphate as P 13 10
The microbial pattern of the pea waste was one of definite pre-
dominance of fang-l Tmp&ifie.c£i, with bacteria, yeast, and protozoa
being present in small numbers. On about the eleventh day, there was
a shift in fungal type from T. V&Lide., which had been used as an
inoculum, to a TuAayuum. The amount of T. \JJJu.d
-------
of the pea canning season, silage juice from the previous year's
corn processing was mixed with the pea waste so that the silage juice
contributed about two- thirds of the incoming BOD. BOD levels of the
feed were adjusted to 1500 mg/1. Detention time was 18 hours, tempera-
ture was above 20°C, the sulfuric acid requirement was 8.1 Ibs per
1000 gallons of waste, and phosphate and ammonia additions were not
required. COD removal was about 83 percent, and BOD about 95 percent
during the first five days after silage addition. This later decreased
to 80 percent. The decrease in performance was associated with a
drop in DO levels to near zero. Oxygen use was calculated to be about
0,5 Ib per Ib BOD removed. The biomass continued to be dominated by
the FoAOtUun which had become the dominant fungus in the pea waste
fermentation.
DISCUSSION
The general effectiveness of the Fumj-t ImpeA^ecXc digestions on the
three wastes to which it was applied in the pilot-plant studies
is summarized in Table 5. The figures used in preparing the table are
neither the best nor worst that could have been chosen from performance
data, nor are they general averages; rather they represent averages
from periods of stabilized favorable performance. BOD removal was
good and COD removal was fair in the cases of both corn canning wastes
and pea canning wastes. In neither instance was removal as good as
obtained in laboratory fermentations where BOD removal ran above
99 percent and COD removal above 96. The reason for the poorer
Table 5. General Efficiency of Fung-c
Process
Percent BOD removal
Percent COD removal
Percent TOC removal
Mycelium produced per unit
BOD removed
H2S04 use -- lb/1000 gallons
Retention time — hours
Corn Canning
Wastes
96
88
93
0.5
4.0
22
Pea Canning
Wastes
95
81
87
0.6
6.5
18
Silage
Wastes
80
83
85
0.3
8.1
18
performance is thought to lie partly in the finer mycelium produced
in the pilot-plant operations. Samples for analysis were prepared by
filtration of a relatively small volume of effluent through Whatman
No. 4 filter paper. Because the mycelium was fine, this procedure
did not produce an entirely clear effluent. In the laboratory a much
larger mycelium was produced which was retained on the filter with
more facility.
220
-------
Nitrogen and phosphate removal is of particular interest in that it
is desirable that levels in the effluent be very low to minimize
contributions to entrophication. In the 1970 season of operation, we
inadvertently added more ammonium sulfate and sodium phosphate than
required either by calculation or by previous experience. Even so,
the levels of phosphate were uniformly lower in the effluent than in the
influent, and at several periods were essentially zero. Nitrogen levels
were likewise essentially zero at some periods of operation. These
low levels had no apparent effect on BOD or COD removal. These
observations, coupled with observations in the laboratory, make it
appear highly likely that the process can be operated with almost no
leakage of inorganic nitrogen or phosphorus into the effluent stream.
Acid usage was higher than expected on the basis of laboratory
experience. This may only reflect a higher level of water hardness
and a lower BOD concentration. Another factor was the anaerobic
changes that took place in the lagoon from which the wastes were drawn.
Direct titration of fresh wastes and of lagoon wastes showed that
almost twice as much acid was required to titrate the material drawn
from the lagoon as was required to titrate the fresh plant wastes.
The acid actually required in the treatment was about two-thirds
that required to titrate lagoon wastes; thus giving evidence of some
acid production by the digestion itself. It is quite evident that a
variety of changes occur quite quickly in the lagoon. These are
evidenced both by the effect of the lagoon materials on T. vViLdz.
growth and by changes in odor and color. The amount of base required
to return the effluent to neutrality was found to be only five percent
as much on an equivalency basis as required of acid to affect the
acidification.
Low temperatures definitely limited the effectiveness of the process.
During the first season, temperatures below 12CC had a considerably
adverse effect. During the 1970 season, temperatures down to 10°C
were successfully accommodated. Successful operation at still lower
temperatures could probably be achieved if part of the effluent
mycelium was recycled. It is also possible that on prolonged operation
at low temperatures, selection would take place for substrains that
grow rapidly at low temperatures.
ECONOMIC ESTIMATES
Economic estimates in terms of costs per pound of BOD removed are
given in Table 6. Among the factors that would lower operating costs
per pound of BOD to a significant degree are higher BOD concentrations,
longer operating seasons, softer water, wastes containing adequate
nitrogen and phosphate, and a mycelium that filters as readily as did
laboratory materials.
221
-------
Table 6. Cost Estimates on Corn Processing
Cents per
Pound BOD
H2SO, 0.43
(NH4J2SO, 0.42
NaH2P04 0.10
Aeration
Power 0.39
Investment 1.16
Labor 0.83
Filtration
Sweco 0.09
Sand bed 0.46
Vacuum Filter 0.33
Drying
Heat
Investment
Total
Credit for dry solids
Net Cost
In making the estimates, it has been assumed that the BOD
concentration in 1,000,000 gallons per day of feed is 1600 mg/1.
Sulfuric acid cost was estimated at 1.7 C/lb and it was assumed that
3 Ibs would be required per 1000 gallons of feed. Ammonium sulfate
at 2 C/lb has been assumed for a requirement of 2.5 lbs/1000 gallons
of feed. This is less than the 3.5 Ibs used during the bulk of the
operation in 1970, but is adequate to give a mycelium with a protein
content of 50 percent and is in line with previous experience on
requirements.
In calculating aeration costs, it has been assumed that 0.7 Ib of
dissolved oxygen will be required for each pound of BOD removed,
that 1 hp delivers 2 Ib of dissolved oxygen, and that electricity
will be available at 1.5 c/kilowatt-hour. In calculating investment
costs, it has been assumed that $500/hp will pay for both aeration
equipment and auxilliary equipment, including the lagoon; that interest
costs will be 7 percent; that the investment would be amortized over
10 years; and that the unit will be operable for 90 days out of the
year. In calculating labor costs, it has been assumed that one man
can take care of the unit and that $100 a day will be adequate to
cover this item.
222
-------
For filtration, it has been assumed that a combination of a Sweco
unit, a sand bed, and vacuum filter would be required. For the
Sweco unit, it has been assumed that a 40-sq ft unit would be1 required.
Such a unit would cost about $7000. A sand bed to handle one million
gallons a day has been assumed to cost $35,000. A vacuum filter to
dewater 6000 Ib of solids/day has been assumed to cost $25,000. A
drum drier to handle 6000 Ib/day will probably cost about $35,000, and
the power cost for drying from 80 percent down to 10 percent moisture
has been estimated at 0.39 C/lb of dry product. Sale of the product
might be expected to return 3.5 C/lb by analogy with the selling price
of soy meal comparable in protein content and quality.
Operating on a year-around basis would reduce investment costs to
one-fourth those listed, reducing the total cost, before credit is
taken for the product, to about 3.0 C/lb of BOD. Production of a
readily filterable material, as was accomplished in the laboratory,
would eliminate the need for the sand filter and give a cost (on a
year-around basis) of about 2.8 C/lb of BOD. The credit of 3.5 C/lb
of product, or 1.75 cents on a pound of BOD basis, would reduce the net
cost to 1.12 C/lb of BOD treated. Further economies would be possible
if oxygen requirements, or costs, and chemical use, could be lowered.
223
-------
ACKNOWLEDGMENTS
The authors wish to express their appreciation to several persons
who have taken a most helpful interest in the study.
Mr. Dale Bergstedt for his efforts in securing many
items of equipment and for coordinating the efforts
of the Green Giant Company, North Star, and commercial
people who loaned or rented some of the material.
Mr. Clarence Sprague of the Green Giant Company was
most appreciated for his many helpful suggestions and
ideas in the construction and maintenance of the pilot-
plant facilities.
We especially wish to acknowledge the cooperation, timely
help, and friendliness expressed by all members of the
Agricultural Production Department of the Green Giant
Company, Glencoe, Minnesota.
Finally, a special expression of appreciation to Mr,
Donald Hartung for his continued, responsible vigilance
of the pilot-plant facilities throughout the study.
224
-------
REFERENCES
1. Mandela, M. , and Reese, E. T., "Fungal Cellulases and the Mlcrobial
Decomposition of Cellulosic Fabric", Develop. Jin InduAt. MtCA.ofa-t0.fc.,
5, 5 (1964).
2. Gray, W. D., The. IL&e o& Fung*. a& food and Ln food P/u>ae44-twg ,
T. S. Furia, Ed., CRC Press, Cleveland, Ohio (1970).
3. Church, B, D., and Nash, H. A., lite. 0& Fungi Imperfecti -trt Wa&te.
Con&io£, Water Pollution Control Research Series 12060 EHT 07/70,
U. S. Department of the Interior, Federal Water Quality Administra-
tion (1970) .
4. Conway, E. D., MccAodc^tw/con A.naty&temt&i,
American Public Health Association, New York, 12th Edition,
p 422 (1965).
225
-------
CANNERY WASTE TREATMENT WITH RBC
AND EXTENDED AERATION PILOT PLANTS
by
R. J. Burm, M, W. Cochrane, and K. A. Dostal*
INTRODUCTION
In addition to the monitoring of various research grants throughout
the country, the Food Waste Research Section of the Pacific Northwest
Water Laboratory conducts "inhouse" research ranging from small bench-top
studies to the operation and investigation of full-size food waste
treatment facilities.
This paper summarizes the observations made during the operation of two
waste treatment pilot plants at the United Flav-R-Pac cannery in Salem,
Oregon during the 1969 and 1970 canning seasons.
Operation of the two plants during 1970 was somewhat unique because they
were run as part of a graduate fellowship program between Oregon State
University and the Pacific Northwest Water Laboratory. Under the program
the laboratory employs a university graduate student while the student
fulfills his M.S. Thesis requirement by conducting research in an area
of mutual interest to both institutions. A much more complete description
of the operation and daily evaluation of the pilot plants will therefore
be forthcoming when the graduate student completes his research report
and submits it to the University in May or June, 1971.
The main objective of the research was to evaluate the applicability of
two waste treatment principles as a means of pre-treating cannery waste
prior to discharge to a municipal system or as the only treatment for the
waste before discharge to a receiving water.
DESCRIPTION OF PILOT PLANTS
Both pilot plants used aerobic treatment with one using a form of extended
aeration and the other a relatively new principle in this country referred
*R. J. Burm and K. A. Dostal are, respectively, Sanitary Engineer and
Chief, Food Waste Research Branch, National Waste Treatment Research
Program, Pacific Northwest Water Laboratory, Environmental Protection
Agency, 200 S.W. 35th Street, Corvallis, Oregon; M. W. Cochrane is a
Graduate Associate, Oregon State University, Sanitary Engineering
Department.
227
-------
to as a Rotating Biological Contactor (RBC).** Details of each plant are
presented below.
Aeration Tank
The extended aeration pilot plant consisted of a 25-foot diameter aluminum
tank with a water depth of 9 feet and a capacity of 32,500 gallons. Oxygen
was provided with a 1-horsepower floating Welles aerator. Effluent from
the tank passed through a 2.5 x 5-foot tube settler. Tube depth was 2
feet and the tubes were sloped 60° from the horizontal plane. The aerator
had a theoretical (and demonstrated) capacity to prevent any settlement
of volatile suspended solids in the bottom of the tank. The tank there-
fore could be considered a combination of extended aeration and sludge
digestion or long-term activated sludge with 100 percent sludge recycle.
Theoretically, the unit should have the capacity of removing significant
amounts of BOD while the tube settler prevents excessive volatile
suspended solids from being discharged in the effluent. Cell matter
synthesized from the BOD would be reflected in a buildup of the mixed
liquor suspended solids concentrations. In the ideal situation, the
unit would remove the BOD from the cannery wastes throughout the canning
season while accumulating the synthesized cell material, and then
endogenously destroy the cell matter after the close of the season. A
picture of the aeration tank and the top of the tube settler compartment
is shown in Figure 1.
KBC
Rotating biological contactors consist of one or more stages of discs
mounted on rotating shafts. The shafts are positioned in such a manner
that the disc surface area dips into and out of the liquid waste during
rotation. Biological masses attach themselves and grow on the discs and
these masses are aerated by exposure of the discs to the atmosphere during
rotation. Some aeration is also provided to the waste water by the actual
physical rotation of the discs.
During the 1969 canning season attempts were made to operate a small RBC
unit which had been originally used in an earlier research grant involving
domestic sewage. This unit consisted of ten separate sets of discs with
individual shafts and hydraulic drive motors. Operating and maintenance
problems were numerous and no useful data were obtained in 1969.
The unit used during the 1970 canning season was a larger and more refined
model owned by the Autotrol Corporation. The entire package unit consisted
of a feed scoop chamber, two disc chambers, and a final clarifer with a
sludge scoop. The feed scoop and both sets of discs were on a central
**Mention of trade names or commercial products does not constitute
endorsement or recommendation for use by EPA.
228
-------
Figure 1
tsJ
Ki
Aeration Tank Installation - United Flav-R-Pac Cannery, Salem, Oregon
-------
chain driven shaft attached to a half-horsepower motor. This drive system
proved to be far superior to the small RBC unit and few operational
problems were encountered. Specific dimensions and operating data are
as follows:
Disc diameter : 1.75 meters
No. of discs per
chamber : 45 and 46
Disc speed : 4 rpm
Disc thickness : 7/16 inch
Disc spacing : 1/2 inch
Clarifier 2
surface loading : 190 gpd/ft @ 2 gpm
A picture of a similar unit is shown in Figure 2.
WASTE CHARACTERIZATION
Processing at the Flav-R-Pac Cannery begins in May or June with straw-
berries and ends in November or December with carrots and/or squash.
Other fruits and vegetables processed are cherries, prunes, beets,
beans, corn, and a limited amount of potatoes. The cannery operates
intermittently for the first few weeks but operation expands to seven
days a week - 24 hours a day during the peak of the season. A summary
of the observed wastewater characteristics during the August-November
period is presented in Table 1.
Table 1 provides an indication of the overall pattern of the waste char-
acteristics at the cannery during various times of the year. Variations
depend on such factors as condition of the crops, different combinations
of crops, and changes in water usages in the cannery. The table shows a
buildup in the COD of the waste during August with maximum values reached
in September when various combinations of corn, beets, beans, squash
and prunes are being processed. COD levels generally begin to fall
through October and November as the quantity and type of vegetables
diminishes.
Most organic data will be presented in this paper in the form of COD rather
than BOD. This is done because more COD data were generated and also
because the BOD test is not considered as reliable or reproducible.
COD/BOD ratios are shown in Table 1 and it is interesting to note that
the ratio increases later in the season simultaneously with the intro-
duction of squash into the cannery processes.
230
-------
Figure 2
10
Rotating Biological Contactor
-------
Table 1. Flav-R-Pac Cannery—Salem, Oregon. Average Process Waste Water
Values, 1969-1970 Seasons*
Vegetables
Period
8/1-8/15
8/16-8/31
9/1-9/15
9/16-9/30
10/1-10/15
10/16-10/31
Processed
Bn,
Bn,
Bn,
Bt,
Bt,
Bt,
Bt
Bt
Bt
c,
c,
Ca
, c
, c
P, S
Ca, S
, s
COD
Mg/1
750
1260
1800
1430
1330
1200
VSS
Mg/1
80
170
455
245
275
360
pH**
6
5
5
6
6
6
.8
.7
.9
.4
.1
.0
Flow
Mgd
.76
.81
.81
.68
.77
.78
COD/
1
1
1
1
1
BOD
„
.25
.15
.50
.40
.60
11/1-11/15
Ca, S
575
155
5.8
.93
1.45
Bn - Beans
Bt - Beets
C - Corn
Ca - Carrots
P - Prunes
S - Squash
*Does not include silt wash water.
**pH - median values.
Volatile suspended matter did not fall off in October and this is
attributed to the processing of squash as well as organic content in
the soil particles adhering to the carrots.
Total nitrogen and phosphorus values during the processing season were
relatively low when compared to carbonaceous material. The raw waste
mean influent value for total nitrogen was 25.6 mg/1 and for total
phosphorus (as P) was 4.1 mg/1 during the 1970 season. Values were,
in fact, deficient to the point that supplementary nitrogen and phosphorus
had to be added to the pilot plants. Standard deviations were 11.0
and 3.9, respectively.
AERATION TANK OPERATION
1969
The aeration tank was assembled at the Flav-R-Pac location during August,
1969, and was placed into operation at the end of August. The system never
worked satisfactorily during the 1969 season because the sludge rarely
showed any tendency to settle under quiescent conditions. As a result,
the tube settler was ineffective and the mixed liquor volatile suspended
solids concentrations averaged only 320 mg/1 over the entire season.
The poor performance is attributable to the following:
232
-------
1. The unit was started near the peak of the canning season and food
to microorganism ratios were in excess of 1.0 for the initial period of
operation.
2. A period of nutrient starvation occurred in early October when the
nutrient feed system broke down.
3. Large quantities of silt were routed past the aeration tank intake
by the cannery during the latter part of October. This was done because
the canner's silt pond became full and the silt had to be routed in the
new direction. The silt either entered the aeration tank or repeatedly
damaged or clogged the intake pumps and thereby stopped the feed to
the tank.
1970
Operation during 1970 was much more successful because of the following:
1. Process modifications at the cannery kept the silt water away from
the intake.
2. A more gradual start-up earlier in the canning season.
3. Much more diligent operation and maintenance of the equipment.
Nutrients were added during both seasons in the form of dissolved
commercial urea and "9-30-0" liquid phosphorus fertilizer. Close
attention was not paid to the nutrient feed rates except to assure that
at least enough was being added to maintain a BOD:N ratio equal to or in
excess of 20:1 and a BOD:P ratio equal to or in excess of 100:1.
A continuous loading limitation during most of the 1970 season was the
inability of the 1-horsepower aerator to provide sufficient oxygen to the
system. DO levels dropped to 0.2 rag/1 on September 1 and, for the next
two months, feed to the unit was limited to no more than 2.5 gpm
because of the low DO observations. Excessive amounts of a high
strength squash waste were fed unintentionally to the unit on the week-
end of November 7-8 and although no DO measurements were taken, it is
felt that DO was zero in the tank over the weekend. This conclusion
is based on the condition of upset which was evident in the tank there-
after.
Operational characteristics are capsulized in Figure 3. The figure shows
COD data from the middle of August (achievement of stabilization) to the
end of October. Operation after this period became erratic because of
intermittent cannery operation, increased intake clogging problems and the
upset after November 8. COD removal is calculated as total COD in minus
soluble COD out. It can be seen that in the operating ranges encountered,
removal was related directly and essentially equal to the application rate.
233
-------
ro
CO
o
-------
The line shown is a least square line for all the data except one
measurement at a maximum loading of 0.3 Ibs/lb. Theoretically the
percentage removal should start dropping off at higher loadings and this
tendency is manifested in the location of this latter point somewhat
below the least square line. Higher loadings were not possible because
of the limitation in oxygenation capacity.
Discussion
Most discussions concerning the biological kinetics of a waste system
of this type are centered around the Monod relationship (1):.
n _ Lo ~ Le _ F Lp m
~ '
Sa T C+Le
Units are those used in this particular analyses.
R = Reaction rate = waste utilization per unit weight of organisms (days ).
L0 = Influent substrate concentration = total influent COD or BOD if it
is assumed that all suspended volatile matter is biodegradable (mg/1).
Le = Effluent soluble COD or BOD concentration = substrate concentration
in a completely mixed basin (mg/1) .
Sa = Mixed liquor suspended volatile matter concentration, which is
proportional to the biological mass (mg/1) .
T = Hydraulic detention time (days) .
F = Maximum waste utilization rate, per unit weight of organisms, at
high waste concentrations (days~l) .
C = Substrate concentration at F/2 (mg/1).
It can be seen that if Le is significantly less than C, Le in the denom-
inator essentially drops out and:
R = (F/C) Le = KLe (2)
where K is a constant. Algebraic manipulation of equations (1) and (2)
gives the following formula:
L0/Le = KSaT + 1 (3)
Formula 3 and its variations are fairly well known and often used in
designing and evaluating biological treatment systems. It must be remembered,
however, that they theoretically should apply only in situations where
soluble BOD or COD in the effluent is much less than C.
235
-------
Three approaches were taken in evaluating the kinetics of the aeration
tank; one was a failure, one was of questionable worth, and one gave
excellent results for the conditions encountered. The three will now be
briefly described, COD values were used in all approaches.
Approach 1
Since Le levels as high as 125 mg/1 were measured in the aeration tank,
it was felt that an attempt to determine kinetic coefficients would be
more valid if the all-encompassing Monod equation were evaluated rather
than the more limited equations (2) and (3) . The approach involved the
following steps:
1. The basic Monod equation was rearranged and a temperature correction
applied to form the following relationship (t = temperature) .
Lfl _
K
This was further modified to:
(Le) + C/F20 (5)
R
Equation (5) can be plotted as the straight line equation:
y = mx + b
2. Analytical measurements and calculations gave Le, R, and t values for
the various sampling days.
3. 0 values between 0.60 and 1.40 were assumed and, with the use of a
computer, least square lines and correlation coefficients were calculated
using the measured Le, R, and t values together with the various assumed
9 values.
4. The optimum correlation coefficient was determined to be 0.66 and the
corresponding 9 value was 1.130.
The approach became questionable at this point, for when the 9 value was
substituted into equation (5), an F2Q of about 0.3 was produced. This
does not approach the values greater than 4.0 which are generally noted
in the literature for F. Furthermore, an observed R value during operation
was 0.27 at a correspondingly low Le value which further discredits the
0.3 value. A possible explanation is the high sludge age or solids
retention time in the system throughout most of the study; these values
were usually between 200 and 300 days and it is questionable whether
the assumption that Sa is proportional to the biological mass is valid
at these levels.
236
-------
This ce".od uses equation (2): R = KLe. If a te-perature correction is
applied it car. be modified to the following:
R .- Q t— 20 r e. ^
—- = .%20 9 (^6'>
Le
Ecuaticr. (6) is the fcmula for a graphical seni-log plot; R/Le was plotted
asainst tezperature and is shovn in Figure 4. A least square line was
calculated for the points and it yielded a K2Q value of 0.00276 and a
temperature coefficient of 1,107. Although the K and 9 values are
reasonable, the lov correlation coefficient does not lend confidence to the
K and 9 values. If a better relationship cannot be developed, the above
constants could be used as guidelines at least.
Approach 3
This approach sirply involved a comparison of influent unit loading with
renoval rate and the excellent relationship, previously shown in Figure 3,
was developed. The equation of the line relates R directly to a constant
tines a function of the influent rather than the effluent as is usually
done. The data were gathered over a 6°C temperature spread and suggests
that 9 = 1.0 in this range. Detention tines varied between 8.7 and 17.4
days while nixed liquor suspended solids increased fron 400 to 2000 Eg/1
during the data period shown. Influent COD concentration varied between
875 and 3000 ng/1 during the sane period. In view of the above wide range
of variables, the excellent correlation of the data is very remarkable.
If a higher aeration capacity were available, the removal rate would have
been observed to fall off fron the 0.98 relationship. This trend is
glinpsed by the slight fall-off of the point representing the highest
loading of 0.3 Ib/lb-day.
RBC OPERATION
The RBC unit arrived at the Flav-R-Pac cannery during the latter part of
August, 1970, and the unit was being fed waste by September 1. Growth
developed on the discs within a few days and the initial sampling on
September 15 showed that good removals were being achieved.
Feed to the unit was essentially the same waste material as was being
sent to the aeration tank, and urea and "9-30-0" fertilizers were added
to the unit at levels high enough to assure that a BOD:N ratio equal to or
in excess of 20:1 and a BOD:P ratio equal to or in excess of 100:1.
Two restrictions were placed on the operation of the unit. The first
involved the rotation speed of 4 rpm which was not adjustable unless the
sprocket was changed and possibly a larger drive motor installed. The
second limitation involved the DO level in the waste beneath the discs.
237
-------
0.001
0.0005
Q>
0.0001
\ I I I
-^ = (.00276}(I.!07T~20)
Le
"r" = 0.51
I I
i i
5 10 15 20
TEMPERATURE (°C)
Figure 4. Aeration Tank - Cannery Wastes
R/Le vs Temperature
238
-------
Advice was received from Autotrol personnel that some measurable DO should
be maintained in this waste. Input to the RBC was therefore limited by
the DO level as was the feed to the aeration tank. Recent conversations
with Autotrol personnel, however, indicate that the only problem encountered
from operation with no DO under the discs might be the presence of odors.
Figure 5 shows the relationship between COD applied to the unit and COD
removed by the RBC. COD removal was computed by subtracting the effluent
soluble COD from the total influent COD. In order to obtain a common
base for comparison between other RBC units, rates are presented per
1000 square feet of disc surface area. The line drawn through the data
points is the least square line and the correlation coefficient indicates
an excellent relationship. The equation of the line shows that about
95 percent COD removal was obtained throughout the range of loading shown.
As loading is increased, the slope of the line will decrease, but the
DO limitation did not allow loading up to this level.
Discussion
A major obstacle to the kinetic evaluation of a unit of the RBC type is
the inability to determine the amount of active biological mass in the
system. The mass is in a dynamic state because of continuous sloughing
and if a sample is taken from an outer disc, there is no assurance that
the growth in the inaccessible inner disc areas is similar.
An attempt was made to substitute surface area for biomass and subsequently
develop a removal rate versus effluent concentration relationship similar
to that presented in Figure 4. However, no meaningful relationship was
apparent.
A comparison of COD influent loading with removal rate per 1000 square
feet gave the very good relationship shown in Figure 5. The removal
rate uniformity over the 10-17°C temperature range indicates that 9
would be close to 1.0 under these conditions. Detention time during
the data gathering period varied from the minimum of 4.1 hours to a
maximum of 10.0 hours and influent total COD concentration varied from
875 to 3000 mg/1. Variation of the biomass during the period is, of
course, unknown. It can be seen, therefore, that despite a temperature
variation of 7°C, influent COD variation of 1:3 and detention time
variation of 1:2.5, the basic removal rate remained dependent only upon
the amount of COD applied up to at least 10.5 pounds of total COD per
1000 square feet of surface area. Additional observations of the RBC
unit will be covered in the next section.
COMPARISON OF THE RBC AND AERATION TANK
The main objective of this study was to determine the effectiveness of
the two principles as either complete treatment or pretreatment processes
for cannery wastes. Both units have demonstrated their effectiveness
under the loading conditions applied. During the study, certain
239
-------
NJ
4^
O
12
10
8
LU
CL
o
en
o
o
o
CO
GO
-I
1
Q
LU
Ld
o
8 o
T
AVE. TEMP: io.o-i7.o°c
DATE: 9-15-11-17-70
£•*
Y '96X-0.3
II II
r =0.99
«
I
I
I
1
24 6 8 10 12 14
COD APPLIED - LBS /1000 SQ FT PER DAY
16
Figure 5. REMOVAL CHARACTERISTICS CANNERY WASTES
-------
characteristics of the two units were revealed which are best covered
in this comparative section. Depending on what is being considered,
one or the other units seems to be superior.
The U.S. Environmental Protection Agency, in general, and the Pacific
Northwest Water Laboratory in particular, do not have the authority to
recommend a particular type of waste treatment process over another.
This is a decision that must be made by the owner and/or operator of
the waste treatment plant or by his consultant. The final decision will
depend on the answers to many questions, a number of which would be unique
for each special situation. It is hoped that data already presented in
this paper and data which will now be presented will assist the owner
or consultant in a final decision. Comparative data from the study will
now be presented and some questions will be raised in areas in which
inadequate data were developed. It is not claimed that all the necessary
information which would be considered in a final decision between the
two processes is included in this paper.
Detention Time vs. COD Removal
The minimum detention time attainable in both units was dependent on the
DO limitations mentioned earlier. As a result the minimum value attained
in the aeration tank was 8.7 days based on an average flow of 2.6 gpm.
The minimum detention time obtained in the RBC (including the clarifier)
was 4.1 hours at an average flow of 2.4 gpm. Measurement of detention
time in the RBC is complicated by the amount of space occupied by the
biomass but the above value is considered to be within 10 percent of the
true value.
Figure 6 compares detention time with COD removal by each unit. The
lines are least square lines, but correlation coefficients for both lines
were poor. It can be seen that removal shows a tendency to decrease
with detention time in the aeration tank. The trend is harder to see in
the RBC data for the detention time involved. The most significant
feature of the figure is, of course, the fact that the RBC removed the
same amount of COD with detention times of only 1 to 5 percent of those
observed in the aeration tank.
COD Removal/Horsepower-Day^
COD loadings to both units were in the same range throughout the study
and removals were also comparable although a much longer detention
time was required by the aeration tank, as previously mentioned.
Power to the aeration tank was delivered to the 1-horsepower aerator
while power to the RBC was delivered to a half-horsepower motor driving
the discs as well as a smaller motor driving the sludge scoop.
Figure 7 relates the COD applied to the power required to remove the
COD. Since both units removed similar amounts of COD at similar loadings,
241
-------
CO
s
o
I
u
o
H
Z
UJ
h-
UJ
o
20.0
10.0
5.0
1.0
0.5
0.1
AERATION TANK
DATE: 9-15 -10-27-70
RBG
I
f
0 10 20 30 40 50 60
TOTAL COD REMOVED -LBS/ DAY
figure 6. RBC AND AERATION TANK DETENTION TIME
VS COD REMOVAL CANNERY WASTES
242
-------
o
I
Q.
X
a
LJ
o
UJ
a:
Q
o
o
Q
2
ID
O
Q.
75
50
RBC
20 40 60
POUNDS COD APPLIED/DAY
Figure 7. RBC AND AERATION TANK COD ADDED VS COD REMOVED/HP-DAY
243
-------
the two least square lines actually approximate the horsepower ratios for
the two units, i.e. 2:1. It can be seen that the RBC removed the same
amount of COD with about 50 percent of the power requirements for the
aeration tests.
Figure 7 presents the actual field data but it does not represent an
entirely accurate picture of power requirements for the following reasons.
1. Data from the Autotrol Corporation indicate that the half-horsepower
disc-drive motor does not operate at capacity until the unit is operated
over 6 rpm. Rotation during the study was 4 rpm and data indicate that
this speed can be maintained with a. power input of 0.1 hp.
2. The 1-horsepower aerator used in this study delivers about 2 pounds
of oxygen per horsepower-hour under standard conditions. Aerators
above 5 hp capacity will deliver between 3 and 4 pounds of oxygen per
horsepower-hour under standard conditions.
It can therefore be seen that data in Figure 7 are conservative and that
both systems should remove considerably larger amounts of COD per horse-
power in full scale treatment plants.
Effluent Suspended Solids Content
Both treatment devices had clarification capacities which would be con-
sidered very adequate for the flow rate encountered. The tube settler in
the aeration tank had an overflow rate of 230 gpdsf at 2 gpm of flow and
the RBC clarifier exhibited an overflow rate of 190 gpdsf at 2 gpm.
Table 2 gives average effluent suspended and volatile suspended solids
values for the two systems during their simultaneous operation. It can
be seen that the level of suspended matter in the aeration tank effluent
was fairly steady throughout the period from September 15 to November 7.
At that point an upset occurred in the system and mixed liquor solids
settleability was reduced to the point that large amounts of solids
passed through the tube settler; More will be said about this upset in
a subsequent section. Levels of solids in the RBC effluent were more
dynamic during the study period with an overall average suspended
solids value of 56 mg/1 being observed. Average value for the aeration
tank effluent, prior to upset, was 72 mg/1. Neither effluent ever
approached the clarity that one associates with a well-operated
activated sludge plant.
The average values of 56 and 72 mg/1 of suspended solids would not be
acceptable for discharge into many receiving waters in this country
under present or probable future water quality and effluent criteria.
However, these values would be quite acceptable for discharge into a
municipal sewerage system.
244
-------
Table 2. RBC and Aeration Tank Effluent Solids Concentrations
Susp. Solids Vol. Susp. Solids
Period RBC A. T. RBC A. T.
(1970) _J1 _jL £. _x _§. 2. x s n x s n
9/15-9/30 38 31 5 70 16 5 36 28 5 65 16 5
10/1-10/16 93 46 5 72 19 5 65 30 5 63 15 4
10/17-11/7 36 16 5 75 13 3 38 — 2 70 9 3
11/8-11/21 56 45 4 325 143 4 92 — 1 260 104 4
X = mean (mg/1)
s = standard dev. (mg/1)
n = no analyses
Sludge Synthesis
In an ideal situation, no synthesized sludge would be discharged from the
aeration tank during the canning season and most of the excess sludge
would be destroyed by endogenous respiration at the end of the season.
Some leakage of suspended matter will always occur, however, and this
was observed to be the case with the aeration tank. Most of the sludge
was nevertheless observed to be held in the tank until the November
7 upset period. This was evidenced by constantly increasing mixed liquor
suspended solids levels and effluent suspended solids concentrations
always less than 100 rag/1. The unit, therefore, did operate from August
through October as an efficient COD removal system without producing an
immediate sludge disposal problem. Ultimately, some inert material would
have to be disposed of.
In contrast, the RBC continually produced sludge which on a larger scale
operation would have to be disposed of by aerobic or anaerobic digestion,
incineration, or other means that will not cause a public nuisance.
Sludge synthesis coefficients and settleability characteristics for
the RBC sludge were not looked into in great depth. Visual inspections
of the sludge showed that it did exhibit good settling characteristics
but the RBC effluent suspended solids values showed that separation
was not as good as can be achieved with activated sludge.
Temperature Effects
Figures 3 and 5 indicate that both units are not greatly affected by
temperature at the relatively mild temperatures in which they were
operated. No meaningful data were gathered in the 0-10°C range, but it is
anticipated that reaction rates would be significantly lower at this level.
245
-------
Subfreezing temperatures would hamper the operation of the RBC to
the point that it would have to be thermally protected unless the source
of the waste was steady and warm enough to prevent freezing. If the
unit were used to treat warm wastes for an industry in subfreezing
temperatures, operation without insulation could prove satisfactory, but
only until a plant shutdown deprived it of the warm water. The rotation
increases the heat loss of the liquid and the unit under these conditions
would either have to be shut down or it would freeze. Either alternative
would destroy the biomass. In contrast, the aeration tank, especially,
if it were buried in the ground, would be able to withstand some influent
stoppages in subfreezing weather.
Organic Overloads
It would be expected that a short-term organic overload would impair the
operation of the RBC because of its short detention time while having
little or no effect on the long detention time aeration tank. No short
duration overloads were noted but the cannery waste itself changes
considerably from hour to hour depending on processing in the plant.
Neither unit seemed to be affected by these fluctuations.
A relatively long-term overload occurred on the weekend of November 7-8
when carrot processing was stopped but high strength squash waste
continued to be discharged. This caused both units to receive loadings
for at least 2 days which were probably far in excess of anything they
had received to that point. Inspection of the units on November 9
revealed that they both had been affected. The RBC was observed to be
discharging high levels of solid material and the aeration tank was also
discharging solids as well as foaming on the surface. DO level in the
aeration tank was 0.0 at this time. On the following day, Tuesday,
November 10, the routine samples showed that the RBC was functioning as
well as ever with good COD removals and low suspended solids in the
effluent. The aeration tank continued to discharge solids and its
removal rates also dropped off. The aeration tank never recovered and
continued to discharge solids until the seasonal shutdown of the cannery
on November 24. Mixed liquor suspended solids 'values in the aeration
tank dropped from 2310 mg/1 on November 6 to 1210 mg/1 on November 24
despite the fact that the unit was still being fed fairly high strength
squash waste during this period.
The conclusion, of course, is that the RBC demonstrated its ability to
"survive" a severe overload while the aeration tank operation was
significantly impaired.
Capital Costs
Capital cost comparisons will not be made between the units. The Autotrol
Corporation and the suppliers of aerators, liners, etc. will gladly supply
this information to anyone interested.
246
-------
SUMMARY AND CONCLUSIONS
1. The RBC and the aeration-type of treatment are capable of removing
most of the organic matter at the levels applied.
2. The RBC removed the same amounts of organic matter as the aeration
tank with a detention time of only 1-5 percent of that observed in the
aeration tank.
3. The RBC produced sludge which would have to be further processed
while the aeration tank did not.
A. Final effluent from both units contained significant suspended
matter.
5. Power requirements for the RBC were less than the aeration tanks'
requirements.
6. The RBC recovered from heavy shock loadings much more quickly than
the aeration tank.
Final decisions on which unit would be more appropriate in a given
situation would depend on information given here as well as other factors
not mentioned in this paper.
247
-------
ACKNOWLEDGEMENT
Acknowledgement is hereby given to the following organizations whose
contributions of manpower, advice, equipment, and access made the
study possible:
Autotrol Corporation
Oregon State University
United Flav-R-Pac Cannery
248
-------
REFERENCE
1. Monod, J., Recherches sur la croissance des cultures bacteriennes.
Paris: Hermann and Cie, 1942.
249
-------
CANNERY WASTE TREATMENT BY LAGOONS AND OXIDATION
DITCH AT SHEPPARTON, VICTORIA, AUSTRALIA
by
Dr. C. D. Parker & G. P. Skerry*
INTRODUCTION
The canning of fruit and vegetables involves the discharge of large
volumes of highly polluted wastewater. It is characteristic of these
discharges that they are seasonally discontinuous being related to
intense canning activity during the usually short period over which
each of the various fruits or vegetables ripen. Except where consid-
erable diversification has been made, discharges may only occur over
part of the year.
This discontinuity of flow and strength, the high B.O.D. value of many
of the wastes, and the rapidity with which the raw wastes develop odour
creates considerable difficulty in devising methods of treatment and
purification which are nuisance free and effective and at the same time
within the financial capability of the industry to accept.
At Shepparton Victoria in Australia there are two large canneries, one
Campbells Soup Co. (Aust.) Pty. Ltd. and the other the Shepparton
Preserving Company.
Campbell's Soup (Aust.) Pty. Ltd. is a subsidiary of the American
Company, they produce meat and general soup products throughout the
year, citrus products June to December, and have a large operation with
tomato products from February to April.
The Shepparton Preserving Co. is an extremely large canner of apricots,
peaches and pears. The cannery handles over 1000 tons of fruit per day
and the daily wastewater discharge is over 2 million gallons.
The City of Shepparton has a population of 20,000 persons which means
that the problem of purificiation and disposal of the cannery wastes,
far outweighs that of sewage purification.
To the domestic sewerage system there are discharged the flows from a
butter factory with an 80,000 gallon per day milk intake Manufacturing
butter, dry skim milk and casein) an abattoirs slaughtering 2000 sheep
per day, a bacon factory and some dairies. The discharge from each
cannery is conveyed to the treatment site by independent pipeline sep-
arate from the main domestic sewerage system.
^Melbourne Water Science Institue Ltd., Water Science Laboratories
Melbourne, Australia.
251
-------
The S.P.C. fruit cannery has been in operation since 1932 and until
1965 its wastes have been disposed of by flood irrigation of 70 acres
of disposal area.
Campbell's Soup Co. factory was established near Shepparton in 1960
and the Sewerage Authority accepted the responsibility for treatment
and disposal of the waste flow.
The problems of constructing effective treatment facilities for the
Campbells Soup Co. waste and the ever increasing load from S.P.C.,
prompted the Authority to undertake a comprehensive study of means of
improved methods of treatment.
At the time of the commencement of investigations the general approach
to the treatment of cannery wastes was the use of low rate aerobic
holding lagoons with odour control by dosage with sodium nitrate and
by flood and spray irrigation. An unsuccessful attempt had been made
by Norgaard et al (1960) to treat peach and pear wastes by anaerobic
fermentation followed by aerobic treatment.
A large plant was subsequently built at San Jose to treat these wastes
with domestic sewage by conventional activated sludge.
To have amplified the existing conventional sewage treatment plant con-
sisting of primary treatment with heated digesters followed by trickling
filters would have cost many million dollars. Studies were therefore
directed towards low cost methods of treatment by lagoons and oxidation
ditch. Spray irrigation was also evaluated.
These experimental studies have already been described in detail
Parker (1966).
In 1968 a Demonstration Grant was made available by U.S. Environmental
Protection Agency to operate facilities made available by Shepparton
Sewerage Authority for the treatment of both cannery wastes by a
combination of anaerobic lagoon and oxidation ditch.
POLLUTION LOAD AND FLOWS
The present flows and the composition of the major sources of pollution
are as shown in Table 1.
It is characteristic of the fluctuations in load that the peak cannery
load which occurs in January-April does not coincide with the peak butter
factory load which occurs in September-December but is considerably lower
than the cannery peak.
INITIAL STUDIES 1962-66
These may be briefly summarized as follows.
Preliminary laboratory model studies established the principles that
purification of fruit cannery waste could be achieved by aerobic lagoon
252
-------
Table 1 Pollution Load
Flow B.O.D. Load Period of Peak
Source mgd (Ibs./day) Discharge
Sewerage System
Domestic 1.2 3200 Continuous
Butter Factory 0.15 8000 September - December
Abattoirs 0.30 3600 Continuous
1.65 14,800
S.p.C. 2.0 50,000 February - April
Campbell's Soup Co. 1.0 5,000 January - March
4.65 69,800
and oxidation ditch provided it were mixed with appropriate quantities
of sewage effluent.
Effective operation of an anaerobic type lagoon could be achieved pro-
vided the same admixture of waste and effluent were used and provided
a frequent regular seed of digested sludge was added. During the 1962
season experimental anaerobic and aerobic lagoons, an oxidation ditch
and spray irrigation areas were operated at Shepparton with continuous
feed. These experimental facilties confirmed the laboratory established
principles and demonstrated approximate permissible loadings for each
process. It was concluded that treatment by aerobic lagoon alone would
involve very considerable areas, and treatment by oxidation ditch alone
would involve very considerable power costs. However, it appeared that
if the anaerobic lagoon process were coupled either with aerobic lagoons
or oxidation ditch this might prove an effective, acceptable and econ-
omic solution.
During the 1963 and 1964 seasons a much larger scale anaerobic lagoon
(4 acres) was operated in conduction with a larger oxidation ditch. These
studies demonstrated that a large size anaerobic lagoon loaded at about
600 Ibs. B.O.D./ac./day would achieve a 70-80 percent reduction in B.O.D.
without any significant . odour. Secondly that an oxidation ditch treating
an algal laden anaerobic lagoon effluent would achieve effective complete
purification and a satisfactory concentration of solids in the ditch
could be maintained. Operation of the anaerobic lagoon in series with
aerobic lagoons also proved the ability of this scheme to achieve complete
purification.
To confirm these findings on a larger scale there was operated during the
1965-66 seasons a 15 acre anaerobic lagoon in series with two large
oxidation ditch units. Over this period a number of different rotor
designs were evaluated, the performance of one compared with two rotors
per ditch was determined and the ability of the ditch to treat raw
waste as well as anaerobic lagoon effluent was demonstrated. These
253
-------
studies established that by loading the anaerobic lagoon at 600-800
Ibs.B.O.D./ac./day a 70-80 percent reduction in B.O.D. can be achieved
and it appeared that this effluent could be treated to reduce B.O.D.
to 20-30 ppm. The power requirement for rotor operation would be of
the order of 0.30 Kwhr per pound of B.O.D. removed.
Observation of aerobic lagoons into which the excess anaerobic lagoon
effluent discharged indicated that at a loading of 80-100 Ibs.B.O.D./
ac./day the B.O.D. which had been reduced from 3000 to 400 ppm. in the
anaerobic lagoon could be reduced to 20 ppra. in the aerobic units.
U.S. EPA DEMONSTRATION GRANT STUDIES 1968-70
Because of the very short fruit canning season and long detention re-
quired in anaerobic lagoons, the findings of the above studies have
had necessarily to be based on very short periods of observation for
the various conditions of loading and operation.
Detailed studies had also been made concurrently concerning the treat-
ment of vegetable canning wastes in anaerobic-aerobic lagoons but not
by the anaerobic lagoon-oxidation ditch process.
To determine the long term reliability of lagoon and oxidation ditch
processes in the treatment of both fruit and vegetable cannery wastes,
the facilities available at the Shepparton Sewerage Authority were
operated and observed continuously for two years, supported by Demon-
stration Grant 12060 EHS
FACILITIES
The facilities used for this study were as follows:
1. Two anaerobic lagoons of 6 and 9 acres, each 4 feet deep, approx-
imately square in shape with three inlets along one side and three
outlets along the opposite side.
The flow of raw cannery waste from both the fruit and vegetable cannery
and the flow of settled sewage or filter effluent from the sewage treat-
ment plant to the two anaerobic lagoons could be controlled by measured
height of flow over V - notch weirs. Daily addition of digested sludge
and digester supernatant liquor was incorporated in the sewage effluent
flow.
Five other anaerobic lagoon units were available for studies of the
influence of nutrient/B.O.D. ratio on performance and six aerobic cells
were available to study the behaviour of aerobic cells receiving anaero-
bic lagoon effluent.
2. Two oxidation ditches each 120 ft. long, 24 feet overall width with
two cage rotors 12 ft. long, 27" diameter installed in one ditch and
one similar rotor in the second ditch. Water depth in each ditch was
39".
254
-------
One original experimental rotor was constructed in a manner similar to
the original Dutch design but in later units considerable modification
was made to eliminate problems with bearing failure, metal fatigue and
inadequate oxygenation capacity.
Rotors were operated with 5" immersion of the teeth. The mixed liquor
from the oxidation ditch was settled in a conventional circular hopper
bottomed sedimentation tank and the sludge returned to the ditches by
a Mono pump operated continuously. The sludge return to the two ditches
was split according to the waste inflow to each ditch. Excess sludge
could be wasted back to the two anaerobic lagoons.
OPERATION & EVALUATION PHASE - ANAEROBIC LAGOON
During the period of the Grant the lagoons treated mixtures of raw sewage
or filter effluent together with fruit canning waste or vegetable canning
wastes. A summary of the various operations carried out from April 1,
1968 to May 7, 1970, is presented in Table 1.
Table 1
Normal Anaerobic Lagoon Operation
Influent Composition
Nature of Waste Load
Date Treated
4/1/68 - Filter Effluent +
6/1/68 Vegetable cannery
Waste 190
6/1/68 - Raw Sewage +
11/11/68 Vegetable cannery
+ Citrus Waste 360
11/11/68- Filter Effluent +
1/4/69 Vegetable Cannery
+ Citrus Waste 140
1/4/69 - Filter Effluent +
2/7/69 Vegetable +Fruit
Cannery 227
2/7/69 - Filter Effluent +
6/20/69 Vegetable + Fruit
Cannery 250
6/20/69 - Raw Sewage +
12/19/69 Vegetable Cannery
+Citrus Waste 105
1/8/70 - Raw Sewage +
5/7/70 Vegetable + Fruit
Cannery Waste 600
255
Percent B.O.D.
N
P04
Ibs/acre/day Removed
90
80
85
86
75
80
ppm.
123
420
480
630
580
520
ppm. ppm.
55 2,100
Lagoon 1
40
Lagoon 2
25
60
44
37
30
50
36
1.4
11
27
20
23
-------
April 1, 1968-June 1. 1968
During this period filter effluent from the Shepparton Treatment Plant
was combined with vegetable cannery waste and fed to the lagoons at
190 Ibs/ac/day. Influent B.O.D. was 123 ppm. and effluent 11 ppm. a
removal of over 90 percent. The influent contained 25 ppm. total nitrogen
and 1.4 ppm. phosphate which provided adequate nutrients compared with
the organic carbon present. The lagoon functioned satisfactorily at a
moderate loading under autumn temperatures (12°C.). The green algae
Chlorella was present at a level of 400,000 orgs./ml. and dissolved
oxygen was present (5 ppm.) in the effluent.
June 1, 1968-November 11, 1968
The Shepparton Sewerage Authority sludge digesters and trickling filters
were not operated over this period and the lagoons treated a mixture of
raw sewage combined with vegetable and citrus waste. The organic loading
on the lagoons averaged 360 Ibs.B.O.D./acre/day with an influent B.O.D.
of 420 ppm. Over the whole period 80 percent of the applied B.O.D. load
was removed. Water temperatures varied from 11°C. in the winter months
to over 20°C. in the late spring in October and November. An adequate
balance of nutrients was always present with a total nitrogen content
of 60 ppm. and phosphorus 11 ppm. The lagoons performed satisfactorily
with regard to B.O.D. removal although the algal population dropped to
only 40,000 orgs./ml. during July, August and early September, the winter
months. It had recovered to 300,000 orgs./ml. by October, with Chlorella
being the predominant organism. Dissolved oxygen also disappeared over
this period and some sulphide odours developed. In September dissolved
oxygen was present in the supernatant but then disappeared until early
November.
November 11. 1968-January 4, 1969
The conventional treatment plant provided primary and secondary treatment
of the town sewage, so that secondary effluent was combined with vegetable
+ citrus waste and treated in the anaerobic lagoons. The strenth of the
combined wastes was similar to that of the previous period, but the flows
available were split between 2 lagoons giving a much lower individual
loading of 140 Ibs./acre/day with an average influent B.O.D. of 480 ppm.
The total nitrogen was 44 ppm. with 27 ppm. of phosphate. 85 percent
purification was achieved at this loading with an algal population of
350,000 orgs./ml. (Chlorella). Dissolved oxygen was present in the lagoon
throughout this period, and there were no odour problems.
January 4, 1969-February 7, 1969
The composition of the waste over this period was secondary effluent
together with vegetable, citrus and fruit cannery waste. The B.O.D.
loading on the lagoons was 227 Ibs./acre/day with an average removal
of 86 percent of the applied B.O.D, Influent B.O.D. averaged 630 ppm..
Total nitrogen averaged 37 ppm. with 20 ppm. phosphate, which gave a
waste well balanced with regard to nutrient content. The algal popula-
tion in the supernatant during this midsummer period increased to several
million organisms/ml. (Chlorella).
256
-------
February 7, 1969-June 20. 1969
During the main fruit canning season, which extends over the first 4
months of the year with lesser activity for another 2 months, secondary
effluent was combined with fruit and vegetable cannery wastes and fed
to the lagoons. The strength of the waste was variable with a peak B.O.D.
of 2500 ppm. in March, falling to a few hundred ppm. in the latter part
of the season. The average loading was 250 Ibs.B.O.D./acre/day. Influent
B.O.D. averaged 580 ppm. with 75 percent removal by the anaerobic lagoon.
The overall nitrogen content was 30 ppm. with 7 ppm. phosphorus. Algae
were at all times present in the supernatant of the lagoons with an
average population of 4 million orgs./ml. (Chlorella).
June 20, 1969- December 19, 1969
The fruit canning season had finished and the Sewerage Authority conven-
tional plant had been shut down so that the anaerobic lagoons received
a mixture of raw sewage, vegetable cannery and citrus wastes. With only
the vegetable cannery oprating, the flows over the second half of 1969
were not high and the lagoon B.O.D. loading was 105 Ibs./acre/day. The
influent B.O.D. was 520 ppm. with over 80 percent removal of applied
B.O.D. Total nitrogen content was 50 ppm-with 23 ppm. of phosphorus.
The algal population of 3 million orgs./ml. (Chlorella) was maintained
throughout this period indicating that the lagoon was coping easily with
the applied loading.
January 8, 1970- May 7, 1970
The usual practice of the sewage authority with the onset of the peak
fruit canning season in the early months of the year had been to operate
the conventional treatment plant treating the town sewage. Secondary
effluent from the plant would then be combined with the cannery wastes
and the mixture purified in anaerobic lagoons. In this season it was
decided not to operate the treatment plant because of the possibility of
odours occurring causing complaints from nearby residents. The Authority
was also commissioning a much larger lagoon area consisting of 108 acres
of anaerobic ponds followed by 140 acres of aerobic ponds at a new site.
Most of the flow was being diverted to the new lagoon area as the new
ponds were being brought into operation.
For the remainder of the Grant period the composition of the waste treated
was a mixture of fruit cannery and vegetable cannery waste together with
raw sewage. The loading over the whole period was 600 Ibs.B.O.D./acre/
day although the flows were only moderate. The strength of the waste in-
creased from 1500 ppm. B.O.D. in Mid January to between 2500-3500 ppm.
B.O.D. until mid April (average 2100 ppm.). Two anaerobic lagoons were
being used to treat the waste. The raw waste averaged 36 ppm. nitrogen
and 4 ppm. phosphorus. The B.O.D. nigrogen phosphorus ratio was therefore
100/106/0.2 which is not favourable for efficient purification. The B.O.D.
removal in the first lagoon averaged 55 percent over the whole season.
Total nitrogen of the effluent was 26 ppm. with 24 ppm. phosphates.
In order to improve the nutrient balance of the waste, nutrients in the
form of commercial ammonium nitrate and poly-phosphate were added to the
257
-------
second lagoon. 80 Ibs. of nitrogen and 8 Ibs. of phosphate were added
daily to achieve a calculated B.O.D. nitrogen phosphorus ratio of 100/
3.5/0.4 which though still low in phosphate would be more favourable
for treatment. However, this lagoon achieved only a 40 percent removal
of applied B.O.D. Total nitrogen content of the effluent was 33 ppm.
with 21 ppm. phosphate.
The addition of these nutrients to the raw waste did not achieve the
expected improvement in purification of the waste. Algal population in
the first lagoon averaged 1 million orgs./ml. with the predominant
organism being Closterium for most of the period.
The algal population in the second lagoon to which nutrients had been
added was higher, 2.8 million orgs./ml. (Chlorella).
LAGOON SLUDGE CHARACTERISTICS
Sludge samples were collected at the inlet of the lagoon 30' from the
influent pipe, at the middle of the lagoon and 30' from the outlet weir.
Sludge depths were measured at 3 points at .inlet middle and outlet aid
the averaged results are presented in summary in Table 2. The Table
also includes results of laboratory analysis for total and volatile solids,
laboratory purification index and gas yield.
The total solids content of the sludge depended on the depth of sludge at
the collection point. A hand sludge pump was used to collect samples
and it drew sludge from 1-3 inches above the mud bottom of the lagoon.
The amount of volatile solids in the sludge collected varied considerably.
In lagoon §2 the volatile solids were higher at the inlet than the outlet
which would be expected but accumulation of the sludge is influenced by
the wind direction and at times the effluent sludge had a high volatile
solids, indicating movement of sludge from the inlet. The least sludge
was always in the middle of the lagoon.
Lagoon 83 sludges do not show the same pattern. The volatile solids are
low during the low loading period of June 1969 and show a higher volatile
solids content for the inlet in September 1969, but during the high load-
ing period of 1970 there were in fact higher volatile solids in the out-
let sludges, probably again due to wind movement-of incompletely digested
sludge from the inlets.
The purification index indicates the pounds of B.O.D. removed/acre/day
under standardized laboratory conditions by a pound of volatile solids
in the sludge sample.
The figures for lagoon 82 inlet sludge are fairly constant at between
3.5 and 5.0 for most of the samplings apart from a low figure in September
1969.
Outlet sludge shows considerably higher purification capacity than inlet,
with sludges collected from the middle of the lagoon having the highest
purification capacity as measured by the laboratory test.
The purification index figures are generally lower in So lagoon, but the
same pattern is maintained in that effluent sludges show the most activity.
258
-------
Table 2 Sludge Characteristics
Lagoon Inlet Sludge
Anaerobic
Lagoon
(S?)
6/68
11/68
1/69
6/69
9/69
1/70
5/70
(s-Q
6/69
9/69
1/70
5/70
(87)
6/68
11/68
1/69
6/69
9/69
1/70
5/70
6/68
11/68
1/69
6/69
9/69
1/70
5/70
(S3)
6/69
9/69
1/70
5/70
Total
Solids
ppm.
38,200
57,940
130,880
88,150
59,860
41,960
59,610
89,350
45,320
113,250
84,940
13,730
17,430
9,320
24,100
26,490
21,880
1,480
22,130
25,000
65,760
35,120
52,520
62,320
60,390
65,910
51,340
80,000
Volatile
Solids
ppm.
18,900
30,270
30,590
26,380
36,780
25,690
33,060
23,520
25,780
53,270
39,120
6,850
9,220
4,960
14,150
16,400
10,280
7,430
13,900
12,010
16,880
17,110
22,520
29,620
18,790
20,040
30,640
45,000
Volatile
Solids
Percent
50
52
23
30
62
61
55
26
57
47
46
Middle
50
53
53
59
62
47
50
63
48
25
49
43
47
31
30
60
56
Purif .
Index
3.9
4.0
3.8
3.2
1.4
3.5
4.8
2.9
2.2
1.7
4.3
Sludge
11.4
12.5
24
4.1
6.0
16.8
12.4
7.6
10.2
3.8
2.1
4.0
5.5
4.0
2.7
3.0
3.9
Sludge
Depth
ins.
10
12
9
5
8
9
9
4
3
3
3
4
8
7
6
6
10
6
8
Gas Yield
mls.gas/day/gm.
of Vol. Solids
5.3
2.4
1.0
1.0
1.0
1.2
0.6
1.3
2.2
0.9
0.2
0.9
2.0
2.2
3.9
1.5
5.3
2.2
5.0
3.8
2.2
3.8
1.0
0.9
1.2
2.4
1.2
0.3
The gas yield in lagoon 82 was high at the inlet in the first sampling but
then declined to relatively low figures. Several samples of the effluent
sludge exhibited greater activity then the inlet.
Gas yields in lagoon So were lower and did not vary from inlet to outlet.
259
-------
Sludge solids accumulated to some extent during the peak fruit canning
season and were not usually digested until the following spring, but
by late in the year there was not more than 6-8" of solids accumulated
at the lagoon inlet or outlet. These lagoons have been in operation for
up to 7 seasons treating in the main fruit cannery waste, so that sludge
accumulation is not a problem with this type of waste.
INFLUENCE OF B.O.D./NUTRIENT RATIO ON ANAEROBIC LAGOON PERFORMANCE
Complementary to the nutrient studies on the two anaerobic lagoons feeding
the oxidation ditch, further investigations were carried out during the
1969 fruit cannery season with two other lagoons known as T2 and S^ which
were dosed with S.P.C. waste with two different proportions of sewage to
achieve two different ratios of B.O.D. to nutrients.
The results of operation of these two lagoons for this season are shown
in Table 3.
Table 3
Lagoon T?
Area (ac.) 15
Flow (g./d.)
S.P.C. 209,000
Sewage 57,000
Influent load
B.O.D. ppm. 1,658
B.O.D. Ib./day 4,420
B.O.D. Ibs./ac./d. 294
Nitrogen Ibs./d. 36
B.O.D./N. ratio 134
Effluent
B.O.D. ppm. 310
Performance
Removal
B.O.D. Ibs./ac./d. 240
B.O.D. (percent removal) 81
Lagoon
106,000
66,000
1,810
3,120
520
33
87
304
430
82
It will be seen that lagoon S^ loaded at 520 Ibs.B.O.D./ac/d. achieved
the same percentage B.O.D. removal and a removal of 430 Ibs./ac./d. com-
pared with lagoon T£ loaded at only 294 Ibs./ac./d. which only removed
240 Ibs./ac./d.
The B.O.D./nitrogen ratio for lagoon Si was 87/1 or B.O.D./N of 100/1.1
compared with a value of 134/1 or 100/0.75 for lagoon T2-
These results taken with the observations made with lagoons 82 and 83 de-
scribed above, indicate the significant influence of B.O.D./nutrient ratio
in the range 30/1 to 130/1 on performance.
260
-------
So far no attempt has been made to determine whether further improve-
ment in performance can be achieved by further reduction in ratio below
the range observed in these studies.
DETAILS OF OXIDATION DITCH EVALUATION PHASE
During the experimental period an oxidation ditch was run continuously
from June 1968 to August 1968, when rotor failure occurred due to faults
which had developed in the overall balance of the rotor. Funds were not
available for its repair until November 1968 when the damaged section was
replaced, stress points eliminated from the whole rotor, and the balance
of the rotor carefully checked. After being placed back in operation in
January 1969, the rotor then ran continuously until May 1970 apart from
a 3 week shutdown in March 1969 when worn water jacketed bearings were
replaced with grease packed roller bearings. Using the experience ac-
cumulated from the operation of 3 earlier rotors over fruit canning
seasons of 4-5 months each, and of the single rotor after 6 months of
continuous running, a new improved rotor was designed, built, and put
into operation in January 1970. This ran continuously for 4 months
without any operational problems or evidence of significant wear.
The various phases of ditch operation and performance are now described.
April 1, 1968 - June 1. 1968
The ditch treated a mixture of filter effluent + vegetable cannery waste
at a B.O.D. loading of 45 Ibs./day. Influent B.O.D. was 280 ppm. and
effluent 9 ppm. a removal of more than 95 percent over the short period
of testing. The flow was 120,000 g.p.d. giving a retention time in the
ditch of 12 hours. The mixed liquor suspended solids was 1840 ppm. which
settled readily. The effluent contained 6 ppm. nitrogen and 1.5 ppm.
phosphate. Power consumption averaged 100 Kw.hr./day giving a power
B.O.D. ratio of 2.0 Kw.hr./lb. of B.O.D. removed.
June 1. 1968 - June 18. 1968
During this short period the ditch was loaded with raw sewage to build up
mix liquor suspended solids and cannery waste. The load applied was 325
Ibs./day at an average flow of 109,000 g.p.d. giving approximately 12 hrs.
detention in the ditch. B.O.D. removal was over 95 percent with an effluent
B.O.D. of 5 ppm. Suspended solids in the mixed liquor were built up to
3,200 ppm. which settled readily.
The effluent contained 7 ppm. total nitrogen and 1.7 ppm. phosphate.
Power consumption averaged 110 Kw.hr./day giving a power B.O.D. ratio of
0.3 Kw.hr./lb. of B.O.D. removed.
June 18. 1968 - August 5, 1968
As the lagoon effluent had declined in strength the ditch was operated
with a mixture of vegetable cannery waste and lagoon effluent at an
average loading of 180 Ibs. of B.O.D./day over the whole period with an
average removal of 80 percent. Early in July heavy rain occurred at
261
-------
Shepparton causing a sudden hydraulic overload of the mixed liquor
settling tank. A significant proportion of the solids were lost before
the excess flow was cut off and mixed liquor suspended solids dropped
to 640 ppm. with little purification of the waste load. The ditch re-
covered rapidly after flows had been cut back and within 2 weeks the mixed
liquor suspended solids was 1360 ppm. and the ditch was treating a load
of 217 Ibs./day achieving a B.O.D. removal of > 95 percent.
August 5. 1968 - February 7, 1969
On the August 5, 1968 an outer section of the cage rotor fractured and
jammed the rotor. The rotor was repaired by late December, mixed
liquor solids were built up in January and the ditch was placed back
into full operation by February 7, 1969.
February 7, 1969 - May 15. 1969
The anaerobic lagoons treated a mixture of filter effluent and vegetable
and fruit cannery waste. The ditch purified lagoon effluent at a
loading of 120 Ibs. of B.O.D. daily. The flow averaged 90,000 g.p.d.
giving approximately 16 hours detention in the ditch. The mixed liquor
suspended solids content ranged from 1300 ppm. to 4600 ppm. during this
period (averaged 2800 ppm.) and settled readily. During this period
ditch influent B.O.D. averaged 135 ppm. which was comparatively low due
to the effectiveness of the anaerobic lagoon treatment. Effluent B.O.D.
however, was 62 ppm. only slightly better than 50 percent removal. In-
fluent nitrogen averaged 30 ppm. with 5 ppm. phosphate and the mix liquor
temperature was 20°C., so that the unexpectedly poor performance of the
ditch could not be due to lack of nutrients or low temperature. It was
noted however, that the algal population in the ditch was 2 million or-
ganisms/ml, (chlorella) and it was decided to perform B.O.D. tests after
this period on the effluent plus a sample that had been filtered to
exclude B.O.D. due to algal decomposition. Effluent nitrogen was 20
ppm. with 4 ppm. phosphate. Average power consumption was 102 Kw.hr./
day giving a power B.O.D. ratio of 1.7 Kw.hr./lb. of B.O.D. removed which
was not consistent with previous experience.
May 15, 1969 - June 20. 1969
The anaerobic lagoon was still treating filter effluent + vegetable and
fruit cannery waste. Because of the decreasing B.O.D. of the anaerobic
lagoon effluent (69 ppm. on May 15, 1969). Vegetable cannery waste was
added to the ditch to increase the B.O.D. load without hydraulic overload.
190 Ibs./day of B.O.D. was added to the ditch with 65 percent removal
(influent B.O.D. was 120 ppm. and effluent B.O.D. averaged 40 ppm.). The
average flow was 170,000 g.p.d. giving retention time of 8 hrs. in the
ditch. In spite of heavy rain on May 29, 1969 mixed liquor solids were
maintained at an average of 2400 ppm. during this period. Total nitrogen
content of the effluent averaged 20 ppm. with 4 ppm. phosphate. The
average power consumption was 102 Kw.hr./day giving a high ratio of 0.85
Kw.hr. of power used/lb. of B.O.D. removed.
262
-------
June 20, 1969 - December 19, 1969
The anaerobic lagoons treated raw sewage + vegetable cannery waste.
The oxidation ditch treated a mixture of raw sewage and lagoon effluent
at a loading of 250 Ibs. of B.O.D. daily at a flow rate of 99,000 g.p.d.
(14 hrs. detention in the ditch). Mix liquor suspended solids averaged
2600 ppm. over this period. The effluent nitrogen content averaged 38
pptn. with 17 ppm of phosphate. The B.O.D, removal averaged 83 percent
after algae had been filtered from the effluent. Mixed liquor suspended
solids averaged 2600 ppm. over the period. Power usage averaged 87
Kw.hr./day, a much lower figure than on most previous phases meaning
0.42 Kw.hr. were used/lb. of B.O.D. removed.
January 8. 1970 - February 11. 1970
The new rotor had been installed for this period in ditch 1 and both
rotors were operated for the remainder of the Grant period (the original
rotor was in ditch 2) .
The anaerobic lagoons were treating a mixture of raw sewage, vegetable
cannery and fruit cannery wastes. The ditches treated a mixture of raw
sewage and lagoon effluent. The load on each ditch was 150 Ibs.B.O.D./
day. The flow averaged 43,000 g.p.d. giving a detention time of 32
hours in each ditch.
Ditch 1 with the new rotor was operated with a higher suspended solids
content of the mixed liquor (3860 ppm. average) and effluent B.O.D. averaged
110 ppm. a removal of 62 percent. Effluent nitrogen content averaged
20 ppm. with 15 ppm. phosphate. Because of the low B.O.D. load, power
consumed/lb. of B.O.D. remold was 1.1 Kw.hr.
Ditch 2 operated at a mixed liquor suspended solids content of 2050 ppm.
and B.O.D. removal averaged 52 percent with an effluent B.O.D of 165
ppm. Effluent nitrogen content averaged 18 ppm. with 17 ppm. phosphate.
The power/B.O.D. ratio was 1.2 Kw. hr./lb. of B.O.D. removed. Over the
last week of this period the strength of the lagoon effluent increased
sharply as the effect of the high strength fruit cannery waste appeared
in the anaerobic lagoon effluent.
February 11. 1970 - May 7. 1970
Both ditches treated anaerobic lagoon effluent from treatment of raw
sewage, vegetable cannery and fruit cannery wastes. The load on ditch
1 with the new rotor was 250 Ibs. of B.O.D./day with a removal of 70
percent. Because of the extremely high strength of the anaerobic lagoon
effluent (1500 ppm.) only 17,000 g.p.d. was treated giving greater than
3 days detention time in the ditch. The suspended solids content of
ditch 1 averaged 3230 ppm. with the solids settling readily. Effluent
B.O.D. averaged 425 ppm. The nitrogen content of the effluent was 21
ppm. with 24 ppm. phosphate. Power/B.O.D. removed ratio was 0.48 Kw.hr./
Ib. of B.O.D. removed.
263
-------
The load on ditch 2 with the original rotor was 230 Ibs. of B.O.D./day
with a removal of 70 percent. Again the flow was 17,000 g.p.d. giving
greater than 3 days retention capacity in the ditch. Suspended solids
content averaged 2080 ppm. in the mixed liquor and settled readily.
Effluent B.O.D. averaged 436 ppm. with a nitrogen content of 19 ppm. and
22 ppm. phosphate. Power used/B.O.D. removed was 0.52 Kw.hr./lb. of
B.O.D. removed.
DISCUSSION
Performance of Process
The operation of the anaerobic lagoons for two years has shown that a
B.O.D. reduction of 75-85 percent can be achieved with loadings up to
400 Ibs./ac./day. The period of loading at 600 Ibs./ac./day was as-
sociated with a very high B.O.D./nutrient ratio and in view of five
years experience with other lagoons it is considered that provided the
B.O.D./nitrogen ratio is held below 50/1 a load of 600 Ibs./ac./day
with 80 percent removal (480 Ibs.B.O.D./ac./day removed) can be achieved.
Sludge
The capacity of lagoon sludges to remove B.O.D. as measured by the lab-
oratory purification index was reasonably constant for inlet sludges and
did not change much with the nature of the waste load or the season of
the year. Outlet sludges generally possessed greater B.O.D. purification
capacity than those taken near the lagoon inlet. This is in line with
previously reported work.
The gas activity of inlet sludges was highest at the beginning of the
project and gradually stabilized under the relatively constant B.O.D.
loading conditions existing. There was some initial stimulation of
gas yield in outlet sludges but this again stabilized to lower figures
over the latter portion of the Grant period.
Nutrients
While the main objective of the project was to establish the reliability
of the two stage process, some study was made of the influence of nutrients
on B.O.D. removal. Operation of lagoons with a range of B.O.D./nitrogen
ratio achieved by varying the cannery waste/sewage flow ratio gave def-
inite evidence of increased performance as the B.O.D./nitrogen ratio was
reduced below 134/1.
It would appear, although very low ratios were not examined, that the
optimum ratio is of the order of 50/1.
Trials in which nutrient ratio was increased by addition of chemicals
showed no definite response. The algal population was increased but
there was no definite trend with regard to B.O.D. removal.
264
-------
Oxidation Ditch
The oxidation ditch stage of the process was marred initially by
mechanical problems with bearing failure and metal fatigue cracking
of the rotor bars. The newly designed rotor used at the latter stage
of the project gave trouble free operation over the five month period
of use.
The process showed no difficulty in maintaining sufficient solids con-
tent in the mixed liquor. It was found that the algal population of
the lagoon effluent was carried through to the final effluent.
The B.OoD. removal was consistent with earlier observations and under
full load was of the order of 30 Ibs.B.O.D./day/foot length of rotor.
The power requirement was 0.4-0.5 Kw.hr./lb.B.O.D. removed.
The effect of overload en performance was demonstrated during the second
fruit cannery season (1970). Despite an increase in the final effluent
B.O.D. to over 400 ppm. a satisfactory fast settling sludge was maintained
throughout.
The only operational factor which caused upset to the process was very
heavy rain. This raised the level of water over the lagoon effluent
weirs inducing a very considerable increase in flow into and out of the
ditch. This increased flow through the final sedimentation tank caused
sludge to rise over the weir and be lost. However, sludge solids were
rebuilt within a matter of two weeks.
Cost Projections
The Shepparton Sewerage Authority was obliged to make a decision con-
cerning the design of new full scale facilities to treat all of both
raw cannery wastes and all the city sewage including abattoirs and
butter factory after primary sedimentation and sludge digestion at the
existing sewage treatment plant, while the demonstration project was
in progress. Based on all the earlier experience and the early results
of the demonstration project, a decision was made to treat initially by
means of the anaerobic-aerobic lagoon system, in preference to anaerobic
lagoon-oxidation ditch system. With regard to capital cost there was a
slight advantage to the anaerobic lagoon-oxidation ditch but when power
costs were considered the annual charge based on local rates of interest
and amortization and running costs, the anaerobic-aerobic lagoon system
was preferred.
The installation has now been constructed and actual costs are available
for the construction of the anaerobic units to treat this combined
cannery-sewage effluent flow of 4 m.g.d. with a B.O.D. load of 70,000
pounds per day.
They are as follows:
Land «a $3CO/ac.) $ 35,000
Earth work (@ 35 cents/cyd.) $ 40,000
Distribution pipes (inlets and outlets) $ 24,000
TT9,000
265
-------
With regard to oxidation ditch costs, data can be established from those
used in the demonstration project. Based on a performance of 30 Ibs.
B.O.D./day/foot length of rotor and other established design parameters,
there would be required in conjunction with the 110 acres of anaerobic
lagoons, seven oxidation ditch units of similar dimensions to those
observed in the demonstration project.
Concrete work (including channels) $ 65,000
Rotors, motors and gears (14) $ 25,000
Electrical $ 2,000
Sedimentation tank $ 25,000
Miscellaneous $ 7,000
Total $114,000
Running Cost
Power for peak fruit cannery
season only 3 months at 2 cents/Kwh. $ 12,600
Maintenance $ 5,000 p.a.
Conclusion
Overall the process of treating cannery wastes by anaerobic lagoon and
oxidation ditch has demonstrated its reliability and freedom from upset
over a two year period of operation. Only one hour per day of main-
tenance has been necessary.
Cost projections indicate that under Australian conditions an overall
B.O.Do load of 70,000 Ibs. B.O.D. per day could be purified for a cost
of $200,000.
Application
As stated earlier the Shepparton Sewerage Authority was obliged to make
a decision in 1968 concerning the design of a full scale facility. On
annual cost figures and because of some doubts at the time concerning
the mechanical reliability of the rotors for oxidation ditches, a decision
was made to treat the wastes initially with the anaerobic-aerobic lagoon
process. It was recognized that provided the reliability of the oxida-
tion ditch as a second stage could be established, amplification of the
plant to cope with increasing load could be achieved by conversion of
some anaerobic units to aerobic units and the second stage facilities
amplified by the construction of oxidation ditches. Provision for this
development was made during construction.
Discharge to these units first occurred in May 1970 and data is avail-
able from that date to the present concerning the performance of the
installation over this initial period of operation.
The whole 24 acres of lagoon holds 370 million gallons and as at present
there still remains ponds with capacity of 40 million gallons to fill.
Over the period May 1970 to January 1971 the whole of the settled sewage
(containing abattoirs and butter factory waste) and the vegetable cannery
266
-------
waste was discharged first to anaerobic lagoons 1 and 2 and later to
anaerobic lagoons 3 and 4. With all four anaerobic lagoons filled
and with an increasingly active sludge developed in these lagoons
overflow from the anaerobic units filled into the aerobic cells No.
6,7 and 8.
In January 1970 with the commencement of the fruit cannery operation
these new facilities have been operated to accept 2/3 of the S.P.D.
waste, 3/4 of the sewage flow and 100 percent of the vegetable waste.
The overall load on the 110 acres of anaerobic lagoon has been 45,000
Ibs.B.O.D./day or 400 Ibs./ac./d. and this has now been applied over
the last six weeks. The remaining flow has been treated at the existing
installation.
The contents of the individual units have had the composition shown in
Table 4.
It has previously been shown that anaerobic lagoons take at least one
season to reach full activity.
It will be seen that as of the present B.O.D. removal in these new
anaerobic units has been over 90 percent at this loading of 400 Ibs.
B.O.D./ac./day.
It is expected that next cannery season the installation will confortably
accept the design load of 600 Ibs./ac./day with 80 percent removal.
Conclusion
These investigations extending over nine years have established for the
conditions existing at Shepparton, a facility capable of purification
of very considerable cannery load together with substantial loads from
abattoirs, butter factory and domestic sewage (population equivalent
500,000 persons) for a cost for treatment units of under $250,000.
267
-------
Table 4
Shepparton Sewerage Authority - New Lagoon Installations
Oo
Date
6/4/70
6/18/70
6/26/70
7/2/70
7/9/70
7/16/70
7/23/70
8/6/70
8/13/70
8/20/70
8/27/70
9/3/70
9/11/70
9/17/70
9/25/70
10/1/70
10/8/70
10/22/70
10/30/70
11/5/70
11/13/70
11/19/70
11/26/70
12/4/70
12/10/70
12/17/70
1/7/71
1/18/71
1/21/71
2/28/71
2/4/71
2/11/71
2/18/71
2/25/71
3/4/71
DB
B.O.D.
ppm.
500
552
416
430
299
256
282
331
371
233
206
401
363
278
390
326
383
308
726
726
500
1088
406
234
530
448
1192
1430
1520
1000
DiB
pH
6.3
5.6
6.5
6.5
6.4
6.5
6.0
5.7
5.8
5.9
6.6
5.7
6.8
6.7
6.5
6.8
6.0
5.9
7.0
5.6
6.5
5.3
6.3
6.0
6.4
6.1
4.4
4.5
5.2
4.9
B.O.D.
ppm>
33
54
67
94
99
68
94
78
103
77
129
160
107
98
119
129
176
149
134
191
310
131
128
62
45
24
13
27
53
46
72
86
73
75
91
2H
6.3
7.2
6.7
7.2
7.1
7.2
7.1
7.0
7.0
7.0
6.9
6.7
7.1
6.7
6.9
6.7
6.9
7.0
7.0
6.9
7.1
6.9
7.1
6.9
7.1
7.0
7.1
6.9
7.1
6.9
7.7
6.8
6.9
6.9
6.6
D?B
B.O.D.
ppm.
20
54
133
105
178
124
169
123
170
102
189
201
140
147
168
144
204
197
153
149
135
119
43
19
11
23
8
29
35
152
47
92
61
80
81
pH
6.6
7.0
6.5
7.0
7.1
7.2
7.1
6.9
6.9
6.9
6.8
6.5
7.1
6.8
6.9
6.8
6.8
7.1
7.1
6.9
7.1
7.0
7.3
6.8
7.1
7.1
7.3
7.2
7.3
7.1
7.2
6.9
6.9
6.8
6.8
D^B
B.O.D.
ppm.
224
201
109
231
268
284
386
299
128
221
234
204
232
256
184
187
192
112
68
165
76
58
28
29
20
89
5
102
106
89
74
£l
7.1
7.3
7.0
6.9
7.0
7.0
6.7
6.7
7.2
6.9
7.0
7.0
6.7
7.0
7.0
7.1
7.0
7.4
7.1
7.1
7-5
7.0
7.6
7.2
8.0
7.1
8.1
6.8
6.6
6.9
7.0
DAB DsB
B.O.D.
ppm.
294
241
261
141
153
123
128
294
158
139
126
134
196
235
349
26
218
49
27
47
130
15
35
67
63
24
27
41
103
66
74
B.O.D.
pH ppm. pH
7.2
7.4
7.0
7.2
7.4
7.7
7.4
6.7
7.2
7.1
7.2
6.8
6.8
7.0
6.8 146 6.!
7.9
6.8
7.8
7-3
7-2
7.2
8.3
7.5
7.3
7.7
8.1
6.9
6.9
6.9
7.0
6.9
DfiB
B.O.D.
ppm.
391 7.0
65
129
30
20
17
16
9
5
5
13
24
7
6
11
15
50
7.1
6.9
7.2
7.2
6.9
7.2
7.7
8.3
7.5
8.2
7-6
8.3
7.4
7.3
7.3
7.9
D7B
B.O.D.
ppm. pH
18
8.3
-------
BIBLIOGRAPHY
J. T. Norgaard, R. Huks & D. A. Reinsch. "Treatment of Combined
Sewage and Fruit Canning Wastes". WPCF Journ. 1088 (1960).
C. D. Parker. "Food Cannery Waste Treateeht by Lagoons & Ditches
at Shepparton". Proc. 21st. Ind. Waste Conf. Purdue Univ.
284 (May 1966).
C. D. Parker & G. P. Skerry. "Function of Solids in Anaerobic
Lagoon Treatment of Wastewater". WPCF Journal 192 (1968).
L. C. Gilde. "Food Processing Waste Treatment by Surface Filtra-
tion". Proc. 1st. National Symp. on Food Processing Wastes.
Portland Oregon 311 (April 1970).
269
-------
BIOLOGICAL TREATMENT OF
CITRUS PROCESSING WASTEWATERS**
by
Dr. F. A. Eidsness, J. B. Goodson, J. J. Smith, Jr.*
INTRODUCTION
Citrus production, processing, and marketing is a multibillion dollar
industry in the United States. The industry is concentrated largely
in Florida, which produces about 75 percent of the oranges and 80
percent of the grapefruit grown in the nation. More than 1,300 square
miles of the Florida peninsula are planted in citrus trees. The 1967-
68 crop from this area amounted to about 9.0 million tons. Future
yields will increase steadily with maturity of new groves and additional
plantings.
Fifty-two processing plants in Florida in 1965 converted about 82 perent
of the citrus crop to single-strength and concentrated juice. Quantities
of waste materials from processing operations are quite formidable. The
industry has made substantial progress in solving its waste problem
through recovery of such valuable by-products as peel, pulp, citrus
molasses, and essential oils. However, it was reported in 1965 that
the Florida plants discharged about 130 mgd of wastewater with a 5-day
BOD loading equivalent to population of some two million people.
Irrigation is widely used in Florida and elsewhere for disposing of
citrus processing wastewaters. Economical and technological success
for this method of disposal is largely dependent upon favorable site
conditions. Disadvantageous locations of its Leesburg and Auburndale,
Florida, plants led The Coca-Cola Company, Foods Division, to seek
alternate means to eliminate the discharge of untreated wastewaters to
the State's surface waters. It was recognized at the time (1962) that
there was no generally accepted treatment process for citrus processing
wastewaters. Although various biological treatments had been subjected
to laboratory and pilot plant studies by a number of investigators; re-
ported results were generally discouraging. Feasibility of a fully co-
ordinated system had not been tested in a single plant-scale venture.
There was a requirement at Leesburg for improvement of the city's pri-
mary sewage treatment plant to include secondary treatment facilities,
as well as for treatment of The Coca-Cola Company's citrus waste. An
estimate of the situation indicated economic advantages for treatment
of the combined municipal and citrus wastewaters. Further, it was con-
cluded that requirements of the State regulatory agency might be met by
*Respectively, Senior Principal, Principal, and Director of Laboratory,
Black, Crow and Eidsness, Inc., Gainesville, Florida.
**Studies sponsored by EPA, Grant No. WPRD 38-01-67 to The Coca-Cola
Company, Foods Division.
271
-------
treatment of segregated strong wastes only from the citrus plant. A
cooperative program to develop a suitable biological treatment plan
on this basis was initiated in August 1962. Laboratory and pilot plant
studies involving biological filter and sludge systems were conducted
over a two-year period. Based on these studies, a treatment plant of
the extended aeration type was installed to accommodate combined city
and citrus plant wastewaters. The first-full scale facility of its
type in the citrus industry was placed in service during the 1966-67
crop season.
Early in 1967 The Coca-Cola Company, Foods Division, was awarded a
demonstration grant by the Environmental Protection Agency. The scope
covered studies of the Leesburg treatment system as well as design,
construction and testing of a full-scale treatment facility for the
Auburndale citrus plant. At Auburndale, all principal wastewaters from
The Coca-Cola Company plant and a neighboring Adams Packing Company
plant are combined with secondary effluent from the City of Auburndale
sewage treatment plant and local storm water in a common public water-
course .
The Coca-Cola Company owns a relatively large amount of land along this
watercourse about 0.6 mile below the plant outfall. Circumstances at
the time favored a treatment system on this land to treat the large
volume of combined dilute wastewaters withdrawn from the channel. An
agreement with the Adams Packing Company was entered into, and treat-
ment facilities patterned after the Leesburg system were installed.
However, a recognized risk was taken in substituting an earthen settling
basin with a simple sludge collection system for conventional mechanical
clarifiers, which would have been extremely expensive for flows involved.
This full-scale experimental facility was placed in service during the
1968-69 crop season.
Although the Leesburg and Auburndale plants were provided primarily for
BOD reduction, it was recognized by the design engineers (Black, Crow
and Eidsness, Inc.), that accelerated eutrophication in receiving
waters is a problem that must be faced in the future. It was seen
also that the facilities afforded an opportunity for investigation of
nutrient removal, as well as for further studies regarding factors
affecting organics removal.
LEESBURG STUDIES
Objectives
General purpose of studies at Leesburg was to investigate treatment of
concentrated citrus processing wastes in combination with municipal
sewage. Operating parameters, control limits, and the plant design
criteria were of particular interest. Nitrogen and phosphorus relation-
ships were also of interest. Predesign studies indicated that an excess
of these nutrient materials in the-City sewage would be highly beneficial
in compensating for a deficit in the citrus wastewaters.
272
-------
Description of Wastewaters
The Leesburg citrus processing plant produces citrus concentrate, with
by-product citrus molasses, cattle feed, and essential oils. Major
sources of high-strength wastewater segregated for treatment are the
juice extractors and finisher area, where relatively large volumes of
water are used during frequent clean-up periods. Other strong wastes
result from less frequent cleaning of juice storage evaporator, blend
tank, chiller, and packaging areas. All of these wastewaters, compris-
ing 'roughly 5 percent of the total wastewater flow, are screened at the
citrus plant prior to delivery to the treatment plant. Characteristics
of citrus wastes segregated for treatment, after screening, are shown in
Table 2. As described in the table, average wastewater is moderately
alkaline. Momentary pH values are highly variable and range from about
4 to more than 13. Dissolved solids are comprised of readily bio-
degradable sugars. Other constituents include peptizing agents, which
interfere with clarification; and peel oils, which may act as bacterio-
static agents.
Extensive predesign analyses indicated that raw combined citrus and
domestic wastes would contain ample nitrogen for successful biological
treatment, with a deficiency of phosphorus. This evaluation was based
on generally accepted BOD:N ratio requirements of 15 to 20:1 and BOD:P
ratio requirements of 80 to 100:1. Analytical data collected during
operation of the plant indicated average ratios of 31:1 and 57:1,
respectively.
The usual season at Leesburg extends from early December into June,
with a midseason lull between early and late maturing crops. Fruit is
processed on a 24-hour per day basis. Weekend operating periods are
variable and 'dependent upon demands.
Average flow of City sewage in 1964 was about 1.15 mgd. BOD amounted
to about 145 mg/1.
Wastewater Treatment Plant
The plant is a completely mixed activated sludge system providing extended
aeration of unclarified raw wastewaters. A shallow oxidation pond was
provided for polishing of the effluent. The plant is illustrated in
Figure 1, and its various components are described in Table 1.
Plant facilities illustrated across the upper part of Figure 1 are
elements of an older primary sewage treatment plant serving the City
of Leesburg. The thickener is actually the old primary plant clarifier,
with no modifications. As shown in the figure, provisions were made
for wasting sludge from the two clarifiers to the thickener. It was
anticipated that the thickener and digesters of the old plant would
serve well enough to concentrate and store waste sludge prior to de-
watering on the drying beds and/or ultimate hauling to a disposal site.
Aerator basins are lined earthen basins. Each was provided with two
mechanical aerators, mounted on concrete piers.
273
-------
cut .
SEIIGE
COOINUIOR
FlUIE
CITRUS KSlSliliB _
FBO« PU»PING STIIim
AER1TION BASINS
=
j. c c^ n "^ -^
2 o o O - r-
o • < > S ^
- 5- « n *S
2 a; o o = =
z S" z ? 2"
^ y c:
en
I T
tl .1
pH RECORDER
ks>—
CHLORINE FRO«
I-ION CON
CHLORIHlIOR
SPLITTER
901
Cl»fllFI£«S
12)
CHLORINE
IB IDU11
II C
PIBSH1U
FLUIE
1 RECIRCULATE.
kr^
O
0 £ KENHISOH NOme
© * uTomTic siKPtfi
(T) »S»«PLE PU«P
PONO
.TO SLUDGE
. TO
LIKE
-------
TABLE 1
DESCRIPTION OF
LEESBURG PLANT FACILITIES
Sewerage Facilities:
Comminutor
Capacity, mgd 6. 0
Degritter
Type Air
Capacity, mgd 6. 0
Wet Well Pumps:
Number 2
Drive Variable speed
Capacity (each), gpm 2, 100
Head, feet 40
Citrus Wastewater Facilities:
Wet well pumps (r.t citrus plant)
Low service:
Number 2
Type Nonclog wet pit
Capacity (each), gpm 350
Head, feet 15
High service:
Number 1
Type Nonclog wet pit
Capacity, gpm 1, 000
Head, feet 36
Phosphoric Acid Feeding Equipment:
Storage tank capacity, gal 1, 500
Feed pumps:
Number 2
Type Wallace & Tiernan A 747
single head metering
Combined Wastewater Facilities:
Aeration basins:
Number 2
Capacity (total), MMG 3. 26
Detention (design), hours 24
Aerators:
Number 4
Type Mechanical
Horsepower (each) 60
275
-------
TABLE 1
(continued)
Clarifiers:
Number
Type
Diameter, feet
Detention (design), hours
Surface overflow rate (design), gpd/sq ft
Weir overflow rate (design), gpd/lin ft
Sludge Re circulation Pumps:
Number
Type
Capacity (each), gpm
Head, feet
Waste Sludge Pumps:
Number
Type
Capacity (each), gpm
Head, feet
Chlorination Equipment:
Chlorine storage
Feeder capacity, lbs/24 hours
Oxidation Pond:
Depth, feet (approximate)
Area, acres (approximate)
Capacity, MMG (approximate)
Detention, hours (design)
Thickener:
Diameter, feet
Capacity, gallons
Suction (Eimco)
55
2.5
750
10,000
Vertical turbine
1,400
25
Nonclog wet pit
150
25
1-ton containers
1, 000
3
5
4.40
31.5
50
157,000
Digesters:
Large:
Type
Diameter, feet
Capacity, gallons
Small:
Type
Diameter, feet
Capacity, gallons
Sludge Drying Beds:
Area, sq ft
Floating cover - unheated
50
150,000
Fixed cover - unheated
25.5
76,000
8,590
fpw
276
-------
Plant design data are shown in Table 2. Construction cost, including
required modifications only to the old primary plant, amounted to about
$500,000.
Plant Performance
The plan of investigation involved series variation of controllable
operating parameters .during the 1966-67 season and evaluation of results.
Delays in completion of plant construction and solving of usual startup
problems consumed most of the season. Consequently, the plan was ex-
tended to include the 1967-68 season. Full execution of the formal plan
was hampered by operational restrictions imposed on the plant. Influent
BOD loading, which was not subject to control, was highly variable and
frequently exceeded design loading. Primary objective of plant opera-
tion was to provide highest degree of treatment at all times. As a re-
sult of these factors, the investigation could not be conducted freely
as a fully controlled study in the manner of a pilot plant investigation.
Operating data collected during the above mentioned seasons are sum-
marized in Table 2. The table also includes, for comparison, more
recent data reported for the mon.th of January, 1971.
The operating data illustrates the plant's remarkable performance with
respect to BOD removal. Plant BOD load exceeded design 61 percent of the
time during the 1966-67 season, and 20 percent of the time the following
season. Despite gross overloads indicated in Table 2, average overall
BOD removals of 94 and 96 percent, respectively, were achieved. As
shown in the table, average removals exceeded 99 percent recently, with
control of loads to design conditions.
In accomplishing above mentioned BOD removals, average unit BOD loading
ranged to 28 lbs/day/1,000 cubic feet, with loadings to 72 lbs/day/1,000
cubic feet on a daily basis. It is seen that the average removals of
better than 99 percent in January, 1971, were obtained with maximum
unit loadings comparable to the design value of 18.3 lbs/day/1,000 cubic
feet.
Average plant influent flows were significantly lower than the design
flow. Such flows resulted in significantly longer detention periods
in the aeration basins than the design period of 24 hours.
Highly variable BOD loadings and other operating problems during the
1966-67 and 1967-68 seasons prevented close control of mixed liquor
suspended solids (MLSS). However, the concentration was varied with
some degree of success between 1,630 and 6,550 mg/1, as shown in Table
2. It was concluded from the experimentation that the optimum concen-
tration was about 4,500 mg/1, which would yield an inventory of 14 Ibs/lb
BOD under design conditions. Confirmation is seen in the data for
January, 1971, when MLSS concentration was controlled in a range of
4,050 to 4,800 mg/1, and average inventory amounted to 23.2 Ibs/lb
BOD with a minimum of the order of 15 Ibs/lb BOD.
277
-------
SUMMARY OF LEESBURG PLANT OPERATION
OPERATING
DESIGN 1966-67 SEASON 1967-68 SEASON
CITY CITRUS COMBINED CITRUS COMBINED CITRUS COMBINED
SEWAGE WASTE WASTES WASTE WASTES WASTE WASTES
HAW WASTES;
FLOW, mgd:
AVERAGE 2.50 0.85 3.15 0. 14 2.05 0.53 1.64
MAXIMUM 1.12 1.63
BOD. mg/l:
AVERAGE 144 706 287 1.470 782 1.044 409
BOD. lbi/d»Y:
AVERAGE 5.000 5.000 8.000 11.200 12.200 5.050 6.810
MAXIMUM 29.900 J1.400 20.200 22.100
SUSPENDED SOLIDS.
mg/l-.
AVERAGE 144 106 148 183
TEMPERATURE. °F. ,
AVERAGE
pH VALUE: ,
AVERAGE 8. 41" 8.0 8. O1 ' 7.9
AERATION BASINS:
BOD. lb«/d»y/M cu ft:
AVERAGE 18.) 28.0 15.6
MAXIMUM 12.0 51.1
DETENTION, hri
AVERAGE 24 38 48
MLSS. mg/l 2.940 1.630 . 6.550
MLSS. lb«/lb BOD:
AVERAGE 10
SLUDGE VOLUME INDEX 67 - 507..
RETURNED SLUDGE ft 100 ....... 42 . 184
TEMPERATURE °F. ......... 66 - 86
DISSOLVED OXYGEN mg/l 0 1 - 90
CLARIFIERS:
OVERFLOW RATE.
gpd/tq ft 750 447 358
EFFLUENT BOD. mg/l
AVERAGE 29 55
RANGE 4-315
BOD REMOVAL. *
AVERAGE 90 93
EFFLUENT SUSPENDED „. ().
SOLIDS, mg/l 60-210 5-410
DISSOLVED OXYGEN, mg/l
pH VALUE 7 0 - 10 I
CHLORINE REQUIRED.
Ib/dly
POND:
EFFLUENT BOD. mg/l 10-47
BOD REMOVAL. *
AVERAGE 94 96
EFFLUENT SUSPENDED
SOLIDS, mj/l
AVERAGE 32
DISSOLVED OXYGEN, mg/l 0.2 - 31
TEMPERATURE F. ... 61-90
DATA
JANUARY. 197 1
CITY CITRUS COMBINED
SEWAGE WASTE WASTES
1. 37 0. 42 1. 79
1.49 0.53 1.92
169 949 343
1.940 3.370 5.220
2.620 5.560 7.930
193 198 198
73 87
8.6 10. &'" 9- 5
12. 0
18.1
44
4. 050
-4. 800
23. 2
200-240
121
101-125
63-75
2. 2-5. 3
390
Z
1-4
99. 4
1.26
1-2. 2
8. 3-9. 6
40
1-S
99. I
1 i
5. 3-14. 3
57-70
"MOMENTARY I>H VALUE or CITRUS WASTE EXTREMELY VARIABLE . 4 TO u*.
WITH AVERAGE SUSPENDED SOLIDS CONCENTRATION OF 6.980 mg/l.
RANGE OF WEEKLY AVERAGES.
278
-------
As seen in Table 2, relatively high sludge volume indices result from
treatment of the combined wastewaters. This was attributed to filamen-
tous organisms (Sphaerotilus) resulting from high carboyhdrate content
of the citrus waste.
It appeared from the two seasons of study that optimum sludge recircu-
lation rate would be somewhat less than 100 percent of influent flow.
Rates of 101 to 125 percent, with average recirculated sludge solids
concentration of 6,980 mg/1, evidently were satisfactory in January,
1971. First cost economy of the vertical turbine pumps installed for
sludge recirculation was outweighed by operating problems and maintenance.
Outages resulted from fouling of impellers and bearings. Stringy solids,
such as hair, were particularly troublesome with bearings. It became
necessary to substitute more conventional sludge pumps.
Plant influent flows to date have resulted in average clarifier surface
overflow rates significantly below the design rate. Yet solids carryover
from the clarifiers has been a rather persistent and troublesome problem,
as evidenced by clarifier effluent suspended solids data in Table 2.
The problem stems from the characteristic high SVI of the sludge, but
it is aggravated by surging. As indicated in the table, the carryover
was much more pronounced during the 1966-1967 and 1967-1968 seasons,
when treatment control was extremely difficult due to bizarre loading
conditions. Recent loadings are much closer to design and afford more
consistent control. As evidenced by the January .1971, data, carryover
has been reduced to reasonable limits. However, the data indicate more
conservative design criteria for clarifiers in similar systems. In
addition to lower design overflow rates, consideration should be given to
deeper clarifiers and designs to minimize disturbances of the very
delicate floes.
It is seen from Table 2 that the pond was quite effective in compensating
for deficiencies in treatment process control. Near design conditions
in January 1971, afforded effluent BOD values of 1 to 5 mg/1 and high
dissolved oxygen concentrations. A study of pond ecology in 1969 in-
dicated a balanced biota of a type desirable in the receiving lake water.
Limited data indicated production of waste sludge solids amounting to
about 0.4 to 0.6 pounds per pound of influent BOD. It was shown that
waste sludge could be thickened to a solids concentration as high as
3.2 percent with facilities provided. Thickened sludge dewatered well,
with no odor, on the drying beds. However, estimated bed capacity of
7,500 gpd was inadequate for handling all waste sludge. It was found
necessary during the 1967-1968 season and during January 1971, to haul
away for disposal some 20,000 gpd of the thickened sludge.
Supplemental nutrients were found to have no effect on treatment efficiency
and are no longer employed at the plant. Nitrogen and phosphorus removals
through the system averaged 70 and 90 percent, respectively. Waste sludge
removed from the treatment system accounted largely for the removals.
279
-------
Principal Conclusions
L. Activated sludge process of extended aeration type is capable of a
high degree of treatment of combined municipal and citrus processing
wastewaters.
2. Design should be based on a MLSS inventory under aeration of about
14 Ibs/lb of daily BOD, which yielded a MLSS concentration of 4,500 mg/1
at Leesburg.
3. Hi£,h sludge volume indices are characteristic of citrus processing
wastewaters.
4. An extremely conservative surface overflow rate is required for
secondary clarifiers, and design should include extra depth and afford
minimal internal turbulence.
5. Flow equalization and minimization of surging should be accommodated
to the fullest extent in the design.
6. Waste sludge facilities should be designed on the basis of about
0.6 pounds dry sludge solids per pound of influent BOD.
7. A polishing pond is advisable.
AUBURNDALE STUDIES
Objectives
Studies at Auburndale dealt with treatment of total effluent from citrus
processing operations. Initial purpose was to evaluate design and
operation of a Leesburg-type system, modified to afford economical
treatment of the much larger flow of weak wastewaters at Auburndale.
Objectives were broadened later to embrace an investigation of an
aerated lagoon system for treatment of the wastewaters.
Treatment kinetics and supplemental nutrient requirements were considered
in the investigation. Further studies pertained to nutrient removal by
aquatic plants, with ultimate disposal of the plants in cattle feed.
Description of Wastewaters
The Coca-Cola Company, Foods Division, plant at Auburndale produces citrus
concentrate and "Hi-C," beverage. Neighboring Adams Packing Company
plant produces concentrate, single-strength juice, and sectioned fruit.
Both plants produce by-product citrus molasses, cattle feed, and essential
oils. In each plant the wastewaters are screened for gross solids removal
prior to discharge to Lake Lena Run, which accommodates local drainage.
Effluent from the City of Auburndale secondary sewage treatment plant
also enters the watercourse in .'the vicinity of the citrus plant outfalls.
These combined wastewaters are diverted from the run into The Coca-Cola
Company's treatment facilities, located downstream.
280
-------
Wastewater flow from The Coca-Cola Company is of the order of 20 mgd,
with a BOD concentration of about 120 mg/1. Adams Packing Company
effluent amounts to some 10 mgd with a BOD concentration around 190
mg/1. The City of Auburndale sewage flow is approximately 0.76 mgd.
Industrial wastes in the sewage contain appreciable amounts of for-
maldehyde and other toxic materials. Nutrient levels in the combined
wastewater were variable and generally marginal for optimum biological
treatment. Characteristics of raw wastewaters diverted to treatment
during the studies are shown in Table 4.
Whereas the usual season at The Coca-Cola Company plant extends from
December through July, that at Adams Packing Company runs from Nov-
ember through July. Both plants operate on a basis of 24 hours per
day with a period of diminished processing about midseason.
Wastewater Treatment Plant
Treatment of facilities are illustrated in Figure 2 and described in
Table 3. As in Leesburg, the system was designed for extended aeration
of unclarified raw wastewater in a modified activated sludge process.
However, notable exceptions were employed for sake of economics in
accommodating the much greater hydraulic loadings.
All basins and ponds are unlined. Site conditions dictated pumping to
achieve flow through the system. ^he arrangement shown in Figure 2
was selected to minimize the cost for pumping facilities. Wastewater
is diverted by gravity from the stream into the aeration basin. Low
lift, irrigation type propeller pumps mounted above the opposite end of
the basin transfer wastewater, including recirculated flow, to the settling
basin. Flow from this point, including recirculation to the aeration basin,
is accomplished by gravity. Mechanical aerators are float-mounted to
adapt to variations in basin water level.
The earthen settling basin, arranged for gravity collection and transfer
of sludge, was substituted for conventional clarifiers as another
economy measure. It was conceded that its efficiency would not match
that of conventional equipment, but it was believed that it might suffice
for maintaining an acceptable concentration of solids in the aeration
basin. The experiment was recognized as a risk by the owner. A dis-
tribution trough across the end of the basin equalizers flows down the
basin. The floor at the inlet end slopes 15 degrees to form a large
transverse trough. A perforated pipe was provided in the trough for
collection and transfer of settled sludge.
Plant design data are shown in Table 4. Construction cost amounted to
about $485,000.
Plant Performance
Plant performance during the 1968-1969 season is described in Table 4.
Exploratory work involving plant operation was divided into three periods.
The first period was devoted to the modified activated sludge process,
281
-------
to
00
CITY EFFLUENT
AQUA AMMONIA
PHOSPHORIC
ACID
RAW
WASTE
CD
n
m TO -n o
3 O o- O Oi
_ £ * o o
" 3 O >
^loO
0*5"-
* <>
"• ^ §. O o
«. Z J£ Z ^
t/i O Tl
INTAKE
POND
®
AERATORS (6)
AERATION
BASIN
3—,
LEVEL I
CONTROL '
®
RECIRCULATED SLUDGE
TRANSFER
PUMPS (3)
AUTOMATIC SAMPLER
= (PROPORTIONAL FLOW)
= DALL TUBES
AUTOMATIC SAMPLERS
(PORTABLE)
SETTLING
BASIN
POLISHING PONDS
TREATED WASTE
-------
TABLE 3
DESCRIPTION OF
AUBURNDALE PLANT FACILITIES
Aeration Basin (Earthen)
Area (water surface), acres
Water depth, feet
Capacity, MMG
Detention (average), hours
3. 15
11. 7
9.70
15
Aerators
Number
Type
Horsepower (each)
Nutrient Feeding Equipment
Phosphoric acid (85% H3PO4):
Storage capacity
Pump type
Pump capacity, gpd
Ammonia (Aqua - 29% NH3)
Storage tank capacity,
Pump type
Pump capacity, gpd
gal.
Mechanical (floating)
75
40 gal. drums
Wallace & Tiernan A 747
22
8, 000
Wallace & Tiernan A 748
960
Transfer Pumps
Number
Type
Capacity,
mgd
Recirculation capability, % of average flow
Settling Basin (Earthen)
Sludge collection section:
Area (water surface), acres
Overflow rate (average), gpd/sq ft
Water depth, feet
Capacity, MMG
Detention (average), hours
Propeller
Two - 16. 74 (each)
One - 9. 33
174
2. 72
132
14. 2
9. 21
14
Total Basin
acres
Area (water surface),
Capacity, MMG
Detention (average), hours
6. 78
20. 77
32
283
-------
TABLE 3
(continued)
Polishing Pond No. 1 (Earthen)
Area (water surface), acres 6. 85
Water depth, feet 5. 0
Capacity, MMG 10. 55
Detention (average), hours 16
Polishing Pond No. 2 (Earthen)
Area (water surface), acres 6. 85
Water depth, feet 4. 5
Capacity , MMG 9. 55
Detention (average), hours 15
Total Pond System
Area (water surface), acres 23. 6
Capacity, MMG 50. 6
Detention (average), days 3. 24
284
-------
TABLE
SUMMARY OF AUBURNDALE_PLANT OPERATIONS
AVERAGE OPERATING DATA
(1)
DESIGN
RAW WASTE:
FLOW, mgd 15. 6
MAXIMUM 30. 0
BOD, mg/1 142
BOD, Iba/day 18,500
MAXIMUM
SUSPENDED SOLIDS, mg/1
TEMPERATURE, °F
pH VALUE
AERATION BASIN:
BOD, Ibs/day/M cu ft 14. 3
MAXIMUM
DETENTION, HOURS 14. 9
MLSS, mg/1 2, 300
MLSS, Ibs/lb BOD 10
RECIRCULATION, % 174
MINIMUM 43
DISSOLVED OXYGEN, mg/1
SETTLING BASIN:
DETENTION, HOURS 32
EFFLUENT BOD, mg/1
BOD REMOVAL, %
EFFLUENT SUSPENDED SOLIDS, mg/1
DISSOLVED OXYGEN, mg/1
pH VALUE
POLISHING POND NO. 1:
DETENTION, HOURS 16
DISSOLVED OXYGEN, mg/1
POLISHING POND NO. 2:
DETENTION, HOURS 15
DISSOLVED OXYGEN, mg/l
TEMPERATURE, °F
pH VALUE
EFFLUENT SUSPENDED SOLIDS, mg/1
EFFLUENT BOD, mg/1
OVERALL SYSTEM:
DETENTION, DAYS 3. 2
BOD, Ibs/acre/day 784
BOD REMOVAL, %
DEC. 1968 5
FEB. 19691 '
15.0
26. 5
114
14, 200
30, 900
27
79
7. 1
11.0
24. 0
15. 5
66
°'38,c,
65'5>
3. 3
33
55
61
44
0. 9
7. 1
17
0.4
15
1. 4
72
7. 1
31
42
3.4
602
67
MAY 1 - 19,
1969( '
5. 3
6. 1
134
5, 920
11, 400
18
85
7. 1
4. 6
8.8
44
49
0. 67
91
6.0
94
19
86
44
1. 7
7. 3
48
2. 4
43
2.9
79
7. 9
9
7
9. 5
251
95
MAY 20 - J.UNE 19,
1969C4)
9. 5
13. 6
149
11, 800
26, 300
26
92
7. 0
9. 1
20.3
24
68
0. 47
108
4. 9
52
33
78
37
1.3
7. 1
27
1. 5
24
1. 5
84
7. 3
26
24
5.3
500
84
^'UNLESS OTHERWISE INDICATED.
(2)TQTAL WASTE FLOW ADMITTED TO PLANT.
(3)PLANT INFLUENT FLOW LIMITED TO 5± mgd.
(4)PLANT INFLUENT FLOW LIMITED TO 10± mgd.
(5)WITH AVERAGE SUSPENDED SOLIDS CONCENTRATION OF 53 mg/1.
285
-------
in accordance with initial objectives of the investigation. In remaining
periods, the plant was operated as an aerated lagoon system to satisfy
broadened objectives.
Operations during the December-February period, involving treatment of
all wastewater, showed that the plant was incapable of the design mod-
ification of the activated sludge process. Despite a wide range of con-
trolled operating conditions, MLSS concentration did not exceed 200 rag/1
and averaged about 66 mg/1. Equivalent MLSS/BOD ratio was only 0.38.
Failure was attributed to inadequacy of the experimental sludge collection
system in the settling basin. However, average BOD removal through the
settling basin amounted to 61 percent, and overall removal was 67 percent.
Such removals for loadings involved proved to be in the range expected of
an aerated lagoon system and suggested studies in that regard for the
remainder of the season.
During the period of May 1-19, plant influent flow was controlled to an
average of 5.3 mgd. As shown in Table 4, average BOD removal through the
clarifier was 86 percent, and overall removal was 95 percent. From May
20 through June 19, influent flow was increased to an average of 9.5 mgd.
Average BOD removals for above described locations amounted to 78 and
84 percent, respectively.
Kinetic studies on the wastewater, using acclimated sludge, yielded BOD
removal rate coefficients (k20°c) ranSi-nS from 0.80 to 2.80 and averaging
1.46. Temperature coefficient (9) was found to vary between 1.04 and 1.06,
with an average of 1.05.
Water collected during the investigation indicated maximum flow and BOD
loading to be about 30 mgd and 32,500 Ibs/day, respectively. Based on
studies described above, it was estimated that improvements involving 40
acres of aeration basin and 450 horsepower in mechanical aerators would
be required for these maximum conditions. Such improvements would be
expected to yield BOD removals of better than 90 percent.
Raw waste average BOD:N ratio was 37:1 and average BOD:P ratio amounted
to 144:1. It was confirmed by ecological studies during the investigation
that a BOD:N:P ratio of 150:5:1 afforded a well balanced biological system
for treatment of the waste. Nitrogen and phosphorus reductions during
treatment were about 21 and 26 percent, respectively.
Nutrient Removal by Aquatic Plants
A portion of Polishing Pond No. 1 was isolated for controlled studies
pertaining to nutrient uptake in water hyacinths. These studies indicated
that at least five days' detention is required for significant removals.
Best removals were obtained with dissolved oxygen concentrations below
0.5 mg/1. Average removals with five days' detention and dissolved
oxygen concentration of about 1.0 mg/1 amounted to 22 percent for nitrogen
and 36 percent for phosphorus. As a matter of some interest, BOD re-
movals amounted to about 70 percent.
286
-------
Processing of harvested hyacinths by squeezing and drying yielded 34,000
gallons of press liquor per acre of hyacinths. Nitrogen and phosphorus
content of the liquor was 335 and 63 mg/1, re3pectively. Citrus plant
feed mill equipment would require extensive modification for incorpora-
tion of hyacinth processing. Processed hyacinths were similar to alfalfa
hay in food value.
Principal Conclusions
1. The existing low cost Auburndale facility will not accommodate an
activated sludge process due to inadequacy of the experimental sludge
collection system.
2. High volume - low strength wastewaters from citrus processing are
amenable to a high degree of treatment in an aerated lagoon system at
BOD loadings of the order of 550 Ibs/day/acre.
3. Hyacinth ponds affording detention times of the order of 5 days
offer moderate nutrient removals. Although hyacinths show promise as
a cattle food supplement, processing presents substantial problems.
287
-------
TREATMENT OF MEAT PACKING WASTE
USING PVC TRICKLING FILTERS
by
Darrell A. Baker* and James White**
PROJECT DEVELOPMENT
In the summer of 1968 Burns & McDonnell Engineering of Kansas City,
Missouri was approached by Farmland Foods, a subsidiary of Farmland
Industries, Inc., Kansas City, Missouri, a fanner owned cooperative,
for the specific purpose of designing a waste treatment plant for
our Denison, Iowa, pork, operation. Several limitations account for
this particular type of plant, but foremost was the land available
for such a plant. In short, this plant was designed specifically
for the minimal land available.
Shortly after the inception of the plan the Environmental Protec-
tion Agency, then F.W.P.C.A. was approached for possible funding of
the treatment plant construction. The construction phase of the
project began in April, 1969 with E.P.A. participating in the pro-
ject in the amount of $290,000.
PLANT DESCRIPTION
The plant is located N.W. of Denison, Iowa, which has a population
of 6,500, and has the capacity to kill and dress 5,000 hogs per day.
The hog cut operation generally accounts for about 40 percent of
the kill plus two or three hundred head per day from our Iowa Falls
plant. Processing as hams, picnics, bellies (bacon) accounts for
about 46 percent of the 40 percent above. The following line diagram
illustrates the breakdown:
ILLUSTRATION I
BREAKDOWN OF PLANT PROCESSES(Ibs.)
Ki
1,00
LI
0000
Cut
/. nn 000
Processing
hams picnics
38,640
184,000
14,720
bacon
27,600
600,000 (shipped)
Rendering
Fat 6s Bones
7,700
*Farmland Foods, Inc., Denison, Iowa.
**Allen Wymore, Burns & McDonnell Engineering, K. C., Mo.
289
-------
The waste from the plant is typical of most packing house operations,
having high BOD, grease and solids content, with variable pH and tem-
peratures. The waste from the plant is collected in two main lines,
one line accepting all wastes from the kill floor area except the
scald tank; the other line accepting wastes from the hog pens, scald
tank, rendering, blood drying operation, and the domestic waste. There
is no cooling water entry into either line.
Plant Units;
The waste from the No. 1 main sewer line is pumped into an air flo-
tation cell for pretreatment before going into the anaerobic lagoons.
The grease removed from the flotation cell is sent back to be rendered
and sold as brown house grease.
The flotation effluent and the other sewer discharge qombine shortly
before discharging into the anaerobic lagoons which are operated in
parallel. The anaerobic lagoons serve two functions, the first being
biological treatment of the wastes and the second that of acting
as holding lagoons to distribute the flow to the filter plant evenly
throughout the work week. The plant generally operates on a 5 or 6
day work week; for this reason an appreciable difference in flows may
be noted on the weekends.
The effluent from the anaerobic lagoons flows through a control valve
which can be operated manually or automatically; then through a
preaeration tank which serves two purposes; to control odors emanating
from the anaerobic effluent, since the plant is relatively close to
a residential area, and to supply a limited amount of oxygen to the
waste before treatment by the trickling filters. Presently a masking
agent is being used intermittently to control odors in the anaerobic
effluent. The preaeration tank effluent is then pumped to the trickling
filters normally operated in series; then to the final clarifiers,
and then to a chlorine contact basin for disinfection. Sludge removed
from the final clarifiers is recycled to the anaerobic lagoons.
DESIGN CRITERIA
A summary of design criteria for the treatment facilities is shown in
Tables 1 through 8.
290
-------
Table 1 Raw Wastes Criteria
Hogs Killed Per Day 5,000
BOD Loading:
5,000 Hogs Kill at 4.3 pounds 21,500 Ibs./day
Average Waste Flow:
Gallons per Hog 175 gal./hog
Gallons per Day 850,000 gpd
Maximum Daily Flow 1,000,000 gpd
Peak Hourly Flow 1,500,000 gpd
Table 2 Air Flotation Tank Design Criteria
Diameter 22'-6"
Water Depth 12'-0"
Hydraulic Rate 1500-gpm
BOD Removal 40%
Grease Removal 85%
Table 3 Anaerobic Lagoon Design Criteria
Number of Cells 2
BOD Applied, Ibs./day , 12,900
Design loading, Ibs. BOD/1000 ftj 15
Water Depth, ft 14
Surface Area, Total acres 1.97
BOD removal, percent 80
Table 4 Preaeration Tank Design Criteria
Detention, minutes 30
Volume of air, cfm 100
291
-------
Table 5 Trickling Filter Design Criteria
Number of Filters 2
Diameter, ft 39
Media Depth, ft 22
BOD Loading, lbs./l,000 ft.J
First Stage
Second Stage 31
Average 4-9
f\
Hydraulic loading, gpm/ft. Surface Area 0.5
Recirculation None
BOD Removal, percent (includes final clarifiers) 91
Table 6 Final Clarifier Design Criteria
Number of Clarifiers 2
Diameter, ft 26
Water Depth, ft 7
Surface Settling Rate, gpd/ft.2 800
Weir Overflow Rate, gpd/ft 6800
Table 7 Chlorine Contact Basin
Detention, at average daily flow, minutes 49
Max. Chlorine Dosage Capacity, Ibs. Cl^/day 100
Chlorine Dosage Rate, ppm 10
Table 8 Treatment Plant Pumping Facilities
Trickling Filter Pumps - Variable Speed:
Filter No. 1:
Number of Pumps 2
Rated Capacity, gpm 700
Filter No. 2:
Number of Pumps 2
Rated Capacity, gpm 700
Final Clarifier Sludge Pumps:
Number of Pumps 2
Rated Capacity, gpm 85
292
-------
PLANT EVALUATION PROGRAM
Originally, the primary purpose of the evaluation program was to study
the performance of the trickling filter system. However, after the
program was begun, sampling stations were added so that the dissolved
air flotation tank and the anaerobic lagoons could be included in the
analysis of the treatment plant performance. The location of all
sampling stations is shown in Exhibit 1. Table 9 shows which sampling
stations were initially composites and grab samples.
Table 9 Sampling Procedure
Sampling Station Type of Sampling
S-l, Air Flotation Tank Influent Composite
S-2, Air Flotation Tank Effluent Composite
S-3, Anaerobic Lagoon Influent Composite
S-4, Anaerobic Lagoon Effluent Grab
S-5, Trickling Filter Effluent Composite
S-6, Final Clarifier Effluent Composite
S-7, Chlorine Contact Tank Effluent Grab
S-8, Final Clarifier Sludge Composite
S-9, Domestic, Hog Pens, Scald Tank Composite
Grab samples were also taken at sampling stations No. S-5 and S-6 to
determine if there was any significant difference between a grab sample
at these two stations and a composite. The data indicated no significant
difference in the results. For this reason, grab samples were taken
at S-5 and S-6 in lieu of composite samples.
The final clarifier sludge was sampled by hand several times throughout
the pumping cycle. These samples were then mixed together to form a
composite.
Three types of automatic samplers were used throughout the program. They
included: (1) a suction head type sampler with 24 bottles for compositing,
(2) a dip type sampler which dipped a 10-15 ml. sample at a set interval,
and (3) a rotating disc type suction sampler.
None of these samplers worked satisfactorily on the air flotation tank
influent because of the extremely high grease content which continually
caused clogging, and the high moisture content in the atmosphere which
shorted out the motors. This problem was eventually solved by providing
a siphon off the flotation tank influent line which discharged into a
55 gallon barrel. The sample for analysis was then taken from the barrel
after the barrel was properly mixed. All laboratory procedures and
analyses were conducted in accordance with Standard Methods, 12th Edition,
published jointly by American Public Health Association, American Water
Works Association, and the Water Pollution Control Federation.
293
-------
FARMBEST INC.
WASTE TREATMENT FACILITIES
DEMONSTRATION GRANT
PROJECT WPRD 241-01-66
LEGEND
-- k. HCPHH OPERATION 'FILTERS IK SERIES)
:-- FILTERS III P»RilLEL
»- SLUOCE LIKE
T.F. TdlCHLIHC FILTER
A • S1KPLINC POINT
rs? • PUMPS
»»» MSTES
r-cn
•ILL FIG03
SCALD TANK AND DOMESTIC
EFFLUENT TO RIVER
~i~^
S-7
S-3
BURNS & MCDONNELL ENGINEERING co.
KANSAS CITY (RM()) MISSOURI
JOB NO.
SHEET . REV.
DEClKBtR •:. t'M
SHEET / OF I SHEETS
-------
The following analyses were made during the program:
DO Chlorine Residual
BOD Grease
COD Temperature
ph Coliforms
Alkalinity Phosphates
Nitrogen Cycle Sulfates
Complete Solids Hydrogen Sulfides
In general, all analyses were run at least once per week, and usually
two to five times per week.
LABORATORY ANALYSES
Time does not permit the presentation of individual analyses, however,
a few of the average analyses appear in Tables 10, 11, 12 and 15.
RESULTS OF EVALUATION PROGRAM
Operational Data
Raw Waste Flows:
The raw waste flows include the wastes from Sewer No. 1 and No. 2.
The total flow from these sewers is shown with the maximum and
minimum daily flow recorded during each month. The maximum daily
flow for the year occurred in December, whereas, the minimum daily
flow for the year occurred in June corresponding to the seasonal
variation for the industry. The monthly averages throughout the
year actually varied very little, with the annual average being
1,123,247 gallons per day.
Trickling Filter Plant Flows;
The trickling filter plant was designed for a constant flowrate.
The anaerobic lagoons were to act as holding ponds so that the
flow discharged to the trickling filters would be constant seven
days a week. The average daily flow discharged to the trickling
filters for each month is shown in Table 13, designated as an-
aerobic lagoon effluent. There is a distinct difference in the
flows for the first seven months and the last five months. From
January through July, the flowrate to the trickling filters was
controlled to buffer the flow over a seven day week. In general,
this was done satisfactory, except on some Sundays, where the flow
decreased substantially. From August through December, a major
operational change was made. It was decided to not have treatment
plant personnel present on weekends. Therefore, the anaerobic
lagoon was not used as a holding posid, and the major portion of the
flow to the treatment facilities were treated as they came in. Thus,
only a minor flow was discharged to the treatment facilities during
the weekends. Table 14 shows the daily average flow to the filters
during these two different operational procedures.
295
-------
Table 10
Sulfates and Hydrogen Sulfide, ppm
VO
ON
Month
1 1 "
Jan.
Feb.
Mar.
Apr.
May
Anaerobic Lagoon
Effluent
Sulfates
40.6
34.6
40.8
32.3
46.3
Hydrogen
Sulfide
4.4
4.3
5.0
-
-
Trickling Filter
Effluent
Sulfates
52.1
57.5
56.3
82.3
73.3
Hydrogen
Sulfide
0.24
0.09
0.30
-
-
Final Clarifier
Effluent
Sulfates
52.1
63.7
63.3
73.3
66.3
Hydrogen
Sulfide
0
0
0
-
_
Chlorine Contact
Tank Effluent
Sulfates
52.6
64.9
64.8
55.0
57.5
Hydrogen
Sulfide
0
0
0
-
-
Average
38.9
4.6
64.3
0.21
63.7
58.9
-------
Table 11 Chlorides, ppm
Month
Jan.
Feb.
Mar.
Apr.
Aug.
Sept.
Anaerobic
Lagoon
Effluent
573
755
803
819
699
735
Trickling
Filter
Effluent
528
742
793
816
684
870
Final
Clarifier
Effluent
535
737
813
839
684
700
Chlorine
Contact Tank
Effluent
551
731
779
831
683
860
Average 731 739 718 739
297
-------
Table 12
7-Day BOD's
June 19, 1970
July 24, 1970
August 19, 1970
September 16, 1970
October 6, 1970
November 4, 1970
December 16, 1970
Raw
Flotation
Domestic
Anaerobic
Filter
Clarifier
Final
Raw
Flotation
Domestic
Anaerobic
Filter (#2)
Clarifier
Final
Raw
Flotation
Domestic
Anaerobic
Filter
Clarifier
Final
Raw
Flotation
Domestic
Anaerobic
Filter
Clarifier
Final
Raw
Flotation
Domestic
Anaerobic
Filter
Clarifier
Final
Raw
Flotation
Domestic
Anaerobic
Filter
Clarifier
Final
Raw
Flotation
Domestic
Anaerobic
Filter
Clarifier
Final
4022
1919
1060
401
564
195
53
1309
2944
415
294
313
119
35
4326
1814
3767
526
528
123
20
11,343
3684
6538
458
384
90
47
5001
2442
3493
372
298
88
31
2663
2086
2498
402
281
145
16
2974
2113
579
379
362
87
121
298
-------
Table 13
Plant Flows. GPP
ho
vo
VD
Raw Wastes Co
Anaerobic Lagoon
Month
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
High
1,223,150
1,185,780
1,205,580
1,258,660
1,236,660
1,160,820
1,265,800
1,232,780
1,229,160
1,240,800
1,277,580
Low
993,720
973,580
951,580
949,640
866,700
979,820
1,113,680
1,145,500
1,154,683
1,104 ,.183
1,069,560
Average
1,062,893
1,098,396
1,065,085
1,109,410
1,099,576
1,055,530
1,165,440
1,173,178
1,192,389
1,152,119
1,181,707
Final
Clarifter
Sludge
Return
108,000
108,000
108,000
108,000
108,000
108,000
108,000
108,000
108,000
108,000
108,000
Anaerobic Lagoon
Influent
High
1,331,150
1,293,780
1,313,580
1,366,660
1,344,660
1,268,820
1,373,800
1,340,780
1,337,160
1,348,800
1,385,580
Low
1,101,720
1,081,580
1,059,580
1,057,640
974,700
1,087,820
1,221,680
1,253,500
1,262,683
1,212,183
1,177,560
Average
1,170,893
1,206,396
1,173,085
1,217,410
1,207,576
1,163,530
1,273,440
1,281,178
1,300,389,
1,260,119
1,289,707
Anaerobic Lagoon
Effluent
High
940,000
899,000
905,000
1,622,000
1,093,000
—
1,384,000
1,481,000
1,334,000
1,670,000
1,840,000
Low
374,000
353.000
396,000
243,000
824,000
--
638,000
726,000
951,000
740,000
555,000
Average
657,000
652,000
704,000
754,000
974,000
--
1,127,000
1,152,000
1,235,000
1,170,000
1,256.000
Plant Effluent
High
832,000
791,000
797,000
1,514,000
985,000
--
1.276.000
1.373.000
1,226,000
1,562,000
1,562,000
Low
266,000
245,000
288,000
135,000
716,000
--
530,000
618,000
843,000
632,000
447.000
Average
549,000
544,000
596,000
646,000
866,000
--
1.019.000
1,044,000
1,127,000
1,062,000
1.148.000
Average
1,123,247
1,231,247
968,000
860,000
-------
Table 14 Trickling Filter Flows
Actual Average Percent of
Design Flow* Daily Flow Design
607,000 gpd 748,000 gpd 123%
607,000 gpd 1,188,000 gpd 195%
*Based on the 5-day working week flow being discharged to
the filters over a 7-day week.
Raw Waste Organic Loading;
Initially, sampling of the dissolved air flotation tank influent was
not a part of the evaluation program. After the program was begun,
it was requested that this waste stream be sampled so that the air
flotation tank could be evaluated. Therefore, accurate data of
this waste stream and the domestic waste stream (sewer No. 2) is
available for only the seven months of the evaluation program.
Table 15 shows the monthly BOD loadings for the two raw waste streams,
It is evident that the waste characteristics vary considerably from
day to day. Approximately 80 percent of the organic wastes dis-
charge to the dissolved air flotation tank while the remaining 20
percent (from sewer No. 2) discharge directly to the anaerobic
lagoons.
Table 15 Raw Wastes, BOD
Domestic Dissolved Air Flotation Tank Inf.
(Interceptor No. 2) (Interceptor No. 1)
Month
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
High
PPm
1224
1449
3240
5133
3004
2197
1052
Low
£ES
369
112
411
317
378
308
369
Average
ppm
769
655
1260
2058
1402
1362
639
Average
Ibs
2609
2818
4489
6949
4841
4240
2095
High
PPm
6795
2944
3720
6336
4301
2290
7558
Low
ppm
1134
943
2484
1407
971
1206
1125
Average
ppm
3134
1771
3178
3515
1768
1621
3325
Average
Ibs
19,260
10,419
20,256
24,218
11,336
10,186
21,117
Total
Ibs. BOD
21,603
13,237
25,100
31,593
16,076
14,444
23,212
Average 4006 16,684 20,752
300
-------
Operational Data Summary;
Table 16 summarizes the basic operational data for the year. The
production facilties operated at an average daily kill rate of 3458
hogs per day, approximately 69 percent of maximum production rate.
The actual waste flow per hog averaged 186 percent of the anticipated
flow, while the BOD per hog averaged 189 percent of the anticipated
BOD load during the sampling period. Table 17 compares the design
criteria with the actual 1970 operational data.
Table 16
Operational Data
Month
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept
Oct.
Nov.
Dec.
1970
1970
1970
1970
1970
1970
1970
1970
. 1970
1970
1970
1970
Hogs
Head
3,015
3,366
3,216
3,340
3,386
3,382
3,031
3,519
3,876
3,743
4,241
4,149
Killed/day
Pounds
Live Weight
692,000
765,000
731,000
763,000
784,000
774,000
674,000
772,000
869,000
865,000
947,000
960,000
Gallons of Waste Flow
Per Head
-
316
342
320
328
328
348
331
302
321
283
285
Per 1000 Ibs.
Live Weight
-
1,390
1,500
1,400
1,420
1,440
1,570
1,510
1,350
1,350
1,260
1,230
BOD.
Per 1000 Ibs.
Per Head Live Weight
-
-
-
-
-
6.4 27.9
4.4 19.6
7.1 32.4
8.2 36.4
4.3 18.6
3.4 15.2
5.6 24.2
Average
3,458 800,000
325
1,420
6.0
19.7
301
-------
Table 17 Suirjnary of Raw Wastes
Average of Percent of
1970 Data Design
BOD
Ibs./day 21,900 20,752 95%
Ibs./hog 4.3 6.0 1407»
Waste Flows:
Gallons per day 850,000 1,123,247 132%
Gallons per hog 175 325 186%
Plant Data
Air Flotation Tank:
This treatment unit is generally considered to be an inplant recovery
unit. However, analyses were run on the unit from June through Dec-
ember to determine the performance of the unit.
Since it was extremely difficult to obtain a representative sample
of the flotation tank influent, the results are somewhat limited in
value. The main constituents removed in the flotation tank are BOD,
COD, grease and solids. The annual averages are shown in Table 18.
Table 18 Dissolved Air Flotation Tank Performance
Analysis Influent, ppm Effluent, ppm Percent Removal
BOD 2624 1762 33
COD 4591 4106 11
Grease 1484 559 62
Total Suspended Solids 2223 1507 32
Anaerobic Lagoon;
The anaerobic lagoons performed very well during the test year. The
results of the annual averages of the more important parameters are
shown in Tables 19 and 20.
The performance of the lagoons were probably enhanced by the thick
grease cover which acts as an insulator. The minimum temperature
of 60°F. of the lagoon contents occurred in December. The summer
temperatures varied between 70-75°F. and the annual average was 69°F.
302
-------
The lagoons performed as expected, removing an average of 82 percent
of the applied BOD even though the lagoons are loaded much heavier
than design loading. Based on pounds BOD applied per day, the total
organic loading averaged 24,940 pounds BOD per day. Thus the lagoon
loading rate averaged 29.3 pounds BOD per 1000 ft. of lagoon volume.
When compared to the design loading rate of 15 pounds BOD per 1000
ft. of lagoon volume, the lagoons are operating at a loading rate
of 195 percent of design.
It is interesting that only 59 percent of the total suspended solids
were removed by the anaerobic lagoons. Although the actual lagoon
detention during the evaluation program was five days as compared
to an expected detention of 7.5 days, based on design hydraulic flows,
one would expect a higher removal of suspended solids.
As expected, much of the organic nitrogen was converted to ammonia
nitrogen in the lagoons. The pH remained relatively constant during
the year, averaging 7.0.
Table 19 Anaerobic Lagoon Performance
Analysis Influent, ppm Effluent, ppm Percent Removal
BOD 2635 477 82
COD 4396 1403 68
Grease 485 106 78
Total Solids 4094 1955 52
Volatile Solids 2112 663 69
Total Suspended Solids 1402 579 59
Organic Nitrogen (N) 95.9 42.1
Ammonia Nitrogen (N) 42.5 121.6
Sulfates -- 38.9
Hydrogen Sulfide -- 4.6
Trickling Filters:
The trickling filters were operated in series during the entire pro-
gram since parallel operation did not provide sufficient hydraulic
loading. The performance of the trickling filters are shown in
Table 21.
303
-------
Table 20
Anaerobic Lagoon Influent, BOD
Domestic + Air Flotation Tank Effluent
ppm Ibs .
Month
Feb., 70
Mar. ,
Apr.,
May,.
June,
July,
o
** Aug. ,
Sept.
Oct.,
Nov. ,
Dec. ,
70
70
70
70
70
70
, 70
70
70
70
High
5960
3406
5263
4645
2780
2130
3521
4149
2265
2520
2041
Low
3836
1648
2760
2986
1295
1260
1370
1013
818
2251
1421
Average
4868
2392
3340
3830
2102
1672
2176
2440
1453
2386
1731
High
49,394
29,007
45,962
39,760
25,751
19 , 049
37,170
39,905
22,087
26,078
18,206
Low
31,791
13,381
21,903
25,560
13,355
12,198
12,725
10,415
7,977
22,094
14,197
Final Clarifier Anaerobic Lagoon
Sludge Return Influent, Ibs.
Average
40,346
21,123
33,464
36,053
19,930
14,978
21,985
23,5t>9
14,103
24,086
16,201
Ibs.
770
770
770
770
770
770
770
770
770
770
770
High
50,164
29,773
46,732
40,530
26,521
19,549
37,940
40,675
22,857
26,848
18,976
Low
32,561
14,151
22,673
26,330
14,126
12,968
13,495
11,185
8 ,.747
22,864
14,967
Average
41,116
21,893
34,234
36,823
20,700
15,748
22,755
24,369
10,873
24,850
16,971
Average
2635
24,169
24,939
-------
Table 21 Trickling Filter Performance
Analysis
Dissolved Oxygen 0
BOD 477
COD 1403
Grease 106
Volatile Solids 663
Volatile Suspended Solids 418
Total Suspended Solids 579
Organic Nitrogen (N) 42.1
Ammonia Nitrogen (N) 121.6
Nitrate Nitrogen (N) 9.3
Sulfates 38.9
Hydrogen Sulfide 4.6
Total Phosphates 53
Trickling Trickling Final Total
Filter Filter Clarifier Percent
Influent, jppm Ef fluent.. _p_pm Effluent, ppm Removal
2.3
296
1010
73
706
443
602
41.1
103.2
25.2
64.3
0.2
52
3.9
124
372
33
354
83
108
21.3
1CO.O
15.1
63.7
0
53
74
73
69
47
80
80
49
18
100
0
The efficiency of the trickling filter system was not as good as
anticipated. It was hoped that the trickling filters would remove
approximately 90 percent of the applied organic loading. However,
the design organic loading was 2580 pounds BOD per day, whereas, the
actual organic loading was 3850 pounds BOD per day during the test
year, or 150 percent of design loading. This results in an overall
loading rate of 73.4 pounds of BOD per 1000 cubic ft. of filter media.
The hydraulic loading "'ate during the evaluation program avtraged
0.79 gpw per ft:.2 of surface area. The design hydraulic loading rate
was^O.5 gpm per ft.z of surface area, or 158/ overloaded.
The, filter provided ample aeration to the wastes, with the dissolved
oxygen in the filter effluent averaging 3.9 ppm.
Th.e filters removed 100. percent of the hydrogen, sulfide present, .but
did not remove any phosphates, with approximately 53 ppm being
discharged in the final clarifier effluent. The filters and clarifiers
provided good grease removal, but still discharged an effluent with
an average of 33 ppm grease.
Several solids analyses were run on the final clarifier sludge.
analyses are shown in Table 22. It was hopeful that some
These
305
-------
Table 22
Trickling Filter Sludge Solids Analyses, ppm
Date
7/17/70
8/11/70
9/10/70
1/24/70
1/26/70
1/27/70
Total
Solids
6696
2659
10,439
6284
6762
6950
Fixed
Solids
2418
1769
2375
2365
2486
2542
Volatile
Solids
4278
890
8064
3919
4276
4408
Suspended Solids
Total
5422
1424
9183
4988
5472
5633
Fixed
1358
792
1388
1289
1428
1463
Volatile
4064
632
7795
3699
4044
4170
Dissolved Solids
Total
1274
1236
1256
1296
1290
1317
Fixed
1061
977
987
1075
1058
1079
Volatile
213
259
269
221
232
238
Average
6632
2326
4306
5354
1286
4067
1278
1039
239
-------
correlation could be obtained between the applied BOD and solids
produced. However, in comparing the suspended solids in the filter
influent with the suspended solids in the final darifier effluent
and sludge, little correlation can be made. Table 23 shows the
pounds of suspended solids per day in the various waste streams.
It is evident from Table 23 that the average suspended solids loading
discharged to the filters was 5340 pounds per day, whereas, the
total pounds of suspended solids removed as sludge and discharged
in the clarifier effluent averaged 5577 pounds per day, for a net
gain in suspended solids of 237 pounds per day.
Table 23 Suspended Solids, Lbs. Per Day
Trickling Final Final
Filter Clarifier Clarifier
Month Influent Effluent Sludge
January
February 2120 473
March 3527 473
April 3527 512
May 2442 813
June 3430 457
July 2948 663
August 6193 558
September 8204 886
October 8650 904
November 6404 1350
December 13,487 1355
Average 5340 767 4810
Chlorine Contact Basin:
The chlorine contact basin was designed for disinfection of the
final effluent, however, the analyses show that some BOD and sus-
pended solids were also removed in the chlorine contact basin.
Table 24 shows the performance of the chlorine contact basin.
307
-------
Table 24 Chlorine Contact Basin Performance
Analysis
BOD
COD
Grease
Volatile Solids
Volatile Suspended Solids
Total Suspended Solids
Chlorine, Total
Coliforms (per 100 ml)
*Except Coliforms
In studying the BOD analyses of the chlorine contact basin's influent
and effluent, it may be that the chlorine affected the BOD test of
the final effluent. Even though the proper procedure for dechlor-
ination was followed in accordance with Standard Methods. 7.7 ppm of
chlorine can not oxidize 63 ppm of BOD.
Table 25 gives the monthly chlorine usage and coliform destruction
through the chlorine contact basin. Excellent disinfection was
accomplished during the year. It is evident from the data that
the majority of the available chlorine was immediately tied up as
combined chlorine. This would certainly be expected with such high
ammonia nitrogen concentrations in the waste stream.
Summary of Treatment Plant Performance;
Table 26 summarizes the average efficiency of each plant unit.
Basin
Influent, ppm*
124
372
33
354
83
108
7.7
35,300,000
Basin
Effluent, ppm*
61
371
17
348
68
90
1.3
1513
Percent
Removal
51
0
49
2
18
17
--
99.99
Table 26
Analysis
Dissolved Air
Flotation Tank
Anaerobic Lagoon
Trickling Filters
Chlorine Contact Tank
Total Plant Removal,
Excluding Dissolved
Air Flotation Tank
Plant Efficiency Percent Removal
BOD
COD Grease Suspended Solids Coliform
33
82
74.
51
11
68
73
0
62
78
69
49
32
59
80
17
97.4 91.5
96.5
93.5
99.99
99.99
308
-------
Table 25
Chlorine Usage and Coliform Reduction
Chlorine Contact
Tank Influent
Chlorine Contact Tank
Effluent
Coliforms/100 ml
Month
January
February
March
April
May
June
July
Augus t
September
October
November
December
Chlorine
Ibs. /day
50
50
44
50
50
60
50
70
60
90
70
70
Chlorine
ppm
9.1
8.0
8.5
7.9
7.3
7.5
6.3
8.7
7.2
6.7
Free Chlorine
PPm
0.7
0.5
0.4
0.1
0.3
0.1
0.1
0.1
0.1
0.2
0.1
* -. _
Combined
Chlorine
ppm
3.2
0.8
0.2
0.3
0.6
0.9
0.8
0.7
0.7
0.7
2.7
_ _ _
Total
Chlorine
ppm
3.9
1.3
0.6
0.4
0.9
1.0
0.9
0.8
0.8
0.9
2.8
.» * _
Chlorine
Contact Tank
Influent
22,200,000
24,000,000
17,700,000
35,200,000
12,300,000
21,200,000
» -• —
Chlorine
Contact Tank
Effluent
65
125
836
4500
767
1360
Average
60
7.7
0.2
1.1
1.3
35,300,000
1513
-------
OPERATING EXPENSES
Operating expenses were recorded for all treatment units with the ex-
ception of the dissolved air flotation tank. Since the primary purpose
of the flotation tank is to recover a saleable product, grease, it is
considered to be an inplant recovery unit, and not a treatment unit.
Operating expenses include personnel salaries, utilities, chemicals,
repairs, and debt service. Table 27 summarizes the annual operating
expenses for 1970.
Table_27 Annual Operating Expenses^ 1970
Salaries $ 47,893
Utilities 1,443
Maintenance & Operating 10,412
Subtotal $ 59,748
Debt Service $ 50,900
Total $110,648
The daily operating expense was $304. per day. Table 28 shows the total
operating expenses based on different parameters.
Table 28 Operating Expenses, 1970
Per Hog Killed $0.09
Per 1000 Ibs. Live Wt $0.39
Per Ib. BOD in Raw Wastes $ .014
Per 1000 Gallons of Ravo Wastes $0.28
During the latter part of 1970, Farmbest reduced their personnel at the
treatment facilities. This will significantly reduce their annual
operating expenses, but should not affect the plant operation. Table
29 shows the projected operating expenses for 1971.
Table 29 Estimated Annual Operating Expenses.,, 1971
Salaries $10,500
Utilities 1,500
Maintenance 300
Operating . . , .8,100
Subtotal $20,400
Debt Service $50,900
Total $71,300
310
-------
Table 30 shows what the estimated expenses will be for 1971 based on the
same parameters as shown in Table 28. These figures are based on the
assumption that the kill rate, waste flow, and organic concentration of
the waste stream will be similar to the 1970 averages.
Table 30 Estimated Operating Expenses, 1971
Per Hog Killed $0.06
Per 1000 Ibs. Live Wt $0.24
Per lb. BOD in Raw Wastes $.009
Per 1000 gallons of Raw Wastes $0.17
CONCLUSIONS
The primary purpose of this research and demonstration project was to
determine the performance of an anaerobic lagoon system followed by
plastic media trickling filters for treating meat packing wastes. At
the time this project was constructed, the treatment of an anaerobic
lagoon effluent by plastic media trickling filters was untried.
The anaerobic lagoons operated as expected, removing 82 percent of the
applied BOD even though the applied loading averaged 195 percent of
the anticipated loading rate. The lagoons are an excellent treatment
unit for treating packinghouse wastes. High removal rates of organic
materials at a minimum capital and operating expenses are accomplished
with anaerobic lagoons.
The trickling filter system did not perform as hoped for; however, there
are several reasons why the trickling filter system cannot perform at
its highest efficiency. Since the final clarifiers are an integral part
of the trickling filter system, the filters and clarifiers must be
analyzed together in determining the performance of the trickling filter
system.
The data clearly shows that the trickling filters organic loading is
150 percent of design loading and the hydraulic loading is 158 percent
of design loading. The higher organic loading will decrease the ef-
ficiency of the system. It is doubtful that the increase in hydraulic
loading would affect the trickling filter operation. However, the
detention time in the final clarifiers is reduced significantly. The
clarifiers were designed to provide a detention time of 2.3 hours, based
on an average flow rate of approximately 607,000 gpd during a seven
day week. However, the anaerobic lagoons are not being utilized as a
holding pond as originally anticipated. Therefore, the majority of the
weekly flow is discharged to the treatment facility the normal five day
work week. This fact, coupled with higher hydraulic flow, which
averaged 968,000 gpd being discharged through the trickling filter
system for 1970, reduces the detention time to approximately 1.4 hours.
During the last five months of 1970, the flow discharged to the trickling
filters averaged 1,188,000 gpd. The detention time during this period
was further reduced to 1.2 hours. It is obvious that the final clarifiers
cannot operate as an efficient unit under such hydraulic loads.
311
-------
The suspended solids concentration in the final clarifier effluent averaged
108 ppm during the evaluation program. Further reduction of suspended
solids and BOD within the clarifiers would be extremely difficult to ob-
tain at such high hydraulic loading rates unless chemical coagulation
facilities are added ahead of the clarifiers.
Another factor which may be affecting the settling characteristics of
the solids is the grease concentration in the trickling filter effluent.
The filter effluent averaged 73 ppm of grease. It may be that the
grease tended to adhere to the solids and change their specific gravity.
This would create a light sludge with poor sludge settling characteristics.
Flotation of solids and grease is apparent in the basins. Although
skimming is provided on the final clarifiers considerable solids are
being discharged in the effluent.
Another possible cause of poor solids settling characteristics is that
denitrification is occurring in the basins. Although the nitrate data
is limited and quite variable, the trend indicates that denitrification
is occurring.
Based on the above discussion, it is apparent that the extremely high
organic and hydraulic loading rates above what was anticipated has
contributed to the reduction in expected plant efficiency. The anaerobic
lagoon and trickling filters operated very well under these extreme
conditions. However, the data indicates that the final clarifiers were
greatly affected by the high hydraulic loading rates and by some
constituents in the waste stream.
312
-------
DEHYDRATION OF CATTLE RUMEN AND WHOLE BLOOD
by
Dr. Donald J. Baumann*
The following is a progress report of a current project in effect at
Beefland International, Inc., at Council Bluffs, Iowa, which is partially
supported by the Environmental Protection Agency.
The title of this industrial waste project is: "Elimination of Water
Pollution by Packing House Animal Paunch and Blood." The eighteen-month
project known as EPA Project No. 12060-FDS was approved in October 1969,
and runs until April 1971.
Objectives of the Project
Although the general objective of this demonstration project is the
elimination of potential water pollutants of animal paunch (rumen) and
blood, the specific objectives are to determine the total quantity of
rumen and blood that is generated in this beef slaughtering operation,
and the determination of the total biochemical oxygen demand (BOD) and
the total chemical oxygen demand (COD) of these materials expressed in
the proper units. In addition, the cost of the dehydrating process of
the animal whole blood and the rumen in terms of gas and electricity
consumption is to be established. Chemical analyses of the dehydrated
products are carried out because of the actual and potential use of
these materials as legal feeds or feed additives.
Beefland International, Inc.
Construction of the physical plant of Beefland International, Inc., was
begun in 1968 and was completed in 1969. Kill operation was begun at the
plant in January.1970. It will have, at full production, the largest
kill-capacity of any slaughtering operation ever approved by the U. S.
Department of Agriculture. At its maximum kill-capacity of 250 head per
hour, a potential kill of 2,500 cattle will be processed daily. At this
rate, a waste problem of approximately 250,000 pounds of rumen and blood
alone would arise daily.
To solve this problem, Beefland installed one McGehee paunch dehydrator
and one McGehee blood dehydrator in a building adjacent to the beef-
slaughtering plant. Paunch and blood obtained in the slaughtering process
are fed through a series of holding tanks to these dehydrators.
*Professor of Chemistry at Creighton University, Omaha, Nebraska, and
Technical Director of EPA Project, 12060-FDS, Beefland International, Inc.
313
-------
Blood, Rumen, and Waste Water Flow
The blood of the animals flows through several openings on the kill floor
directly to a holding tank below from whence it is periodically blown
over into holding tanks adjacent to the blood dehydrator. Similarly,
the rumen from each paunch is emptied into a hopper on the kill floor
into a holding tank below. From that tank it is likewise blown over
periodically by means of air and steam pressure along an overhead line
to a second holding tank adjacent to the paunch dehydrator in the separate
dehydrator building. Figure 1 is a schematic diagram which shows this
flow of blood and rumen.
The waste water which originates from the washing of animal carcasses and
parts, as well as that which arises from the periodic hosing down of
sections of the kill floor during and at the completion of operations of
the day, flows into a central sump. The waste water from the hosing down
operation of the outside holding pens is similarly collected in this
central sump. It is pumped from the sump into a Sedifloater-Clarifler
plant.
The Sedifloater-Clarifier plant is composed of two sections—a flotation
unit and a pressurizing system. The flotation unit is comprised of an
open-top ateel tank, 50 feet in diameter, and various internals specifically
designed to handle the grease "float," the settled solids, and the
clarified liquid. The pressurizing unit consists of special pumps, an
air injection system, an air saturation tank, and an air-release system
with required piping.
The air-saturated waste liquid flows under pressure to the flotation unit.
Just before it enters the flotation unit, it passes through a pressure
release valve. This sudden reduction in pressure as the waste liquid
passes through the valve causes the air to come rapidly out of solution
in the form of tiny bubbles. These bubbles attach themselves to suspended
particles in the liquid causing them to rise in the Sedifloater-Clarifier.
The fatty "float" thus formed is then removed into a discharge outlet.
Floor scrapers in the flotation unit move the settled particles to a
sludge pump from which they are removed for ultimate disposal. The
clarified liquid is drawn out of the flotation unit from a level below
mid-depth and discharged into an aeration basin or lagoon. The lagoon
has the dimensions of 150 ft x 450 ft with a waste water depth of approx-
mately 10 ft. There, four 50-horsepower electric surface aerators tran-
fer oxygen into the effluent piped from the Sedifloater-Clarifier, thereby
further reducing its impurity content by aerobic oxidation. Finally, the
now clarified effluent is discharged from the aeration basin into the
city's sewage system. Figure 2 shows the waste-water flow and component
systems described above.
Dehydrators and the Dehydrator Building
The building housing the dehydrators is a simple structure with
corrugated metal sheets covering a metal framework. The building is
approximately 40 ft wide, 80 ft long, and 40 ft high.
314
-------
BLOOD AND PAUNCH FLOW
Blood
Drain
\
\
Hold
Tank
Kill Floor
FIRE
BOX
GAS METERS
CONTROLS
Wet
Paunch
x
Dry Paunch
Paunch
s
Dryer
Blood
Dryer
VAPOR
RELEASE
VA COLLECTOR
VA COLLECTOR
"1GUK.; i
VAPOR
RELEASE
WASTE WATER FLOW
, To City Sewer
Kill Floor
Aerators
o o
o o
Sedifloater
Clarifier
Aeration Basin
Lagoon
Holding Pens
Dehydrator Bldg.
FIGURE 2
315
-------
Each of the McGehee dehydrator units has a diameter of 8 ft with a
length of 20 ft. These drums revolve on four sets of ordinary rubber
tires (three tires per set) which are motor driven and located at the
four corners of each drum. A ten-foot-long air-cooled firebox is
located at the front of each drum. An adjustable sleeve located on
the opposite end of the drum aids in the control of the moisture content
of the product. Each dehydrator unit is equipped with a vacollector
and gas release system. Figure 1 shows the arrangement of these
components, as well as the storage facilities for both the fresh rumen
and blood and the dehydrated products.
The gas-fired dehydrators are operated at different temperature rangeg.
The dehydration of the rumen has been carried out in the range of 180 F
to 230 F. The blood has been dried between 130°F to 180°F. Temperature
readings are taken at the point where the dried products enter the
vacollectors.
The first of the dehydrators (for whole blood) was put into operation
about mid-March 1970. The second unit for the dehydration of rumen
became operational by May 1, 1970. Table 1 summarizes the cost of the
dehydrators and building construction.
Table 1. Dehydrator Building Costs
Blood Dehydrator $ 52,000
Paunch Dehydrator 62,000
Construction Costs 24,700
Total $138,700
The Laboratory
When approval of the demonstration grant by EPA officials to Beefland
International, Inc., was received in late 1969, plans for the laboratory
facilities needed to reach the outlined objectives of the project were
begun.
Several suitable sites for the laboratory within the main plant building
were selected. The company administrators elected, however, to house the
laboratory outside the main building within a reasonable distance of all
sampling sites. Although it was recommended that a minimum of 12 ft x
20 ft floor space be provided for adequate floor and counter-top space for
316
-------
the required equipment, a considerably smaller space for the laboratory
finally evolved.
A Behlen corrugated metal flat-topped building was erected on a concrete
slab behind the main plant buildings. The inside dimensions are approx-
imately 10 ft x 17 ft with an 8 ft dropped ceiling. Of the 170.8 sq ft
floor plan, an area of 47.4 sq ft is taken up as counter space. The
building was erected during the month of April, 1970.
Occupancy and use of the completed laboratory facility was not possible
until November 1, 1970. This was due to the unexplained delay in the
delivery and installation of the laboratory cabinetry, counter-tops,
and fume hoods by the vendor. Analytical work on dehydrated rumen and
blood was begun by April first by the chemist employed by Beefland
International. As a M.S. graduate student at Creighton University, he
carried on the analytical work at the university. He continued utilization
of these facilities until the Beefland laboratory was completed.
The total expenditures for the laboratory facilities were $12,300, of
which $4,400 were expended for the construction of the buildings. The
major equipment items ($100 or more/unit) purchased are listed in Table 2.
Table 2. Major Laboratory Equipment
Automatic Read-out Balance
BOD Incubation Cabinet
Distillation Apparatus and Redistillation Kit
Drying Oven
Extraction Heater—6 Unit
Kjeldahl Combination Digestion and Distillation Apparatus—6 Unit
Pulverizer—Hammer Type
pH Meter
Muffle Furnace
Spectronic-20 Colorimeter
Electric Grinding Mill
Analytical Data
Analysis of the dehydrated products began immediately after the dehydrators
317
-------
became operational. The percent moisture of both the dehydrated rumen
and blood ranged between wide limits during the first months of operation
because of the inexperience of the operators and adjustments required on
the dryers. The last sixty determinations of the various analyses made
are given in Table 3. The methods used in the analytical procedures are
those from "Official Methods of Analysis of the Association of Official
Agricultural Chemists," 10th Edition, 1965.
Table 3. Analysis of Dehydrated Products1
Blood
Rumen
Moisture
Protein
Mean-%
5.4
88.0
Std. Dev.
1.9
4.7
Moisture
Protein
Fat
Crude Fiber
Calcium
Ash
P2°5
Carbohydrate^
7.1
12.2
3.2
26.1
0.59
7.1
1.47
39.2
2.0
1.4
0.4
3.6
0.09
0.71
0.25
5.5
total of 30 samples but 60 separate determinations.
carbohydrate calculated by subtracting total percentage of moisture,
protein, fat, crude fiber, and ash from 100% for each sample as advised by
Mr. Whitson of the Iowa Agricultural Laboratory in Des Moines, Iowa.
318
-------
Data on undehydrated blood is summarized in Table 4. The chemical oxygen
demand (COD) determination is that given for high level demand industrial
waste in "Standard Methods for Examination of Water and Wastewater," 12th
Edition, 1965, and in "FWPCA Methods for Chemical Analysis of Water
and Wastes," November 1969. No correction for chloride ion was made in
the calculation of total COD. The biochemical oxygen demand (BOD)
procedure used is the modified Azide-Iodide Winkler method for dissolved
oxygen.
Lagoon water was found to be the best seed in the BOD determination of
the fresh blood. The seed correction, taking the dilution factor into
account, was made on those BOD measurements of seed alone which gave a
40 to 60 percent depletion of dissolved oxygen.
Table 4. Data on Fresh Blood
No. of
Mean Std. Dev. Detm's*
pH 7.40 0.12 26
% Moisture 83.4 3.3 28
COD (ppm) 206,070 30,910 48
BOD5 (ppm) 132,440 47,340 24
BOD5 of Lagoon Water2 (ppm) 139 60 42
*The same as number of samples for pH and % moisture. Two determinations
per sample made on COD and BOD.
^Lagoon water (plant effluent) used as the seed in BOD determination of
undehydrated blood.
Table 5 summarizes the data acquired to date on undehydrated rumen. The
COD and BOD are established separately for the liquid and solid portions
of the material. The sample taken for each analysis is filtered, washed
repeatedly and the washings diluted to 2,000 ml. The solid portion
remaining on the filter is dried overnight in a 105°Foven, thus giving
the composition of the sample.
319
-------
It was ascertained that use of a seed was required for the BOD determi-
nation of the liquid portion of the rumen. Holding pen droppings with a
little fresh paunch added but filtered before use was found to act as
an excellent seed in this determination. With the advent of very cold
weather, the BOD determination results of the liquid portion of the
paunch became erratic. The BOD determination of the dried solid portion
of the rumen has presented some technical problems which have not been
solved to date.
Table 5. Data on Undehydrated Rumen
pH (liquid + wash.)
% Moisture
COD (ppm)
Mean
6.60
86.3
Std. Dev.
0.69
2.9
No. of
Detm's1
23
24
46
Liquid Portion 53,820 9,860
% Liquid (90.4) (3.0)
Solid Portion 1,179,870 87.320
% Solid (9.6) (3.0)
Total COD 161,350 34,710
% COD from Liquid 30.9 6.3
% COD from Solid 69.1 6.3
BOD5 (ppm) 28
Liquid Portion 21,730
Solid Portion no data
-The number of determinations is the same as number of samples for pH
and % moisture, but two times the number of samples for COD and BOD.
Table 6 summarizes the kind of information which is being reported
monthly concerning the operation of the dehydrators. Correlations will
be able to be made at the end of the project period which will give the
expected yields of dehydrated products on the basis of the daily kill.
Costs for labor, gas, and electricity per unit-weight of dehydrated product
will also be known.
320
-------
Table 6. Production and Cost Data—January, 1971
Head of Cattle Slaughtered 32,582
Days of Operation 24
Blood Dehydrator
Pounds of Dried Blood Produced 195,492 Ibs
Drying Time Required 224% hrs
Gas Consumption 1,560.71 MCF
Electricity Consumption 9,216 KWH
Man-hours Required 337% hrs
Paunch Dehydrator
2
Pounds of Rumen Generated 1,824,592 Ibs
Pounds of Dried Rumen Produced 158,438 Ibs
Drying Time Required 2QB%. hrs
Gas Consumption 1,266.8 MCF
Electricity Consumption 23,040 KWH
Man-hours Required 342^ hrs
Total Gas Costs (both dehydrators) $1,387.02
Total Electricity Costs (both dehydrators) $379.01
Based on estimated 6 Ibs/animal.
2
Calculated on basis of 56 Ibs/animal.
3Contents of 22,634 paunch dried (69.5% of total). Weight based on 7 Ibs
dried rumen/animal.
Slides of all tabular data contained in this report were shown during the
presentation. In addition, about 60 colored slides of photos taken both out-
side and inside of the main plant, dehydrator building, and the laboratory were
shown to more clearly represent the details of the project as described.
321
-------
To date, only 40 to 70 percent of the total rumen generated has been
dehydrated each month. A better market for this product must be created.
Plans are currently being studied for the installation of a pelletizing
machine for the dried rumen which would incorporate the recovered fat
from the Sedifloater-Clarifier in the product.
322
-------
WATER AND WASTE MANAGEMENT
IN POULTRY PROCESSING
by
Dr. W. M. Crosswhite, R. E. Carawan and John A. Macon*
The research, development and demonstration project in water and waste
management in poultry processing is being conducted at the Gold Kist
plant in Durham, North Carolina. The purpose of the project is to make
changes in plant equipment and operations throughout the plant. Specific
objectives are to:
1. Install and/or modify process equipment and demonstrate its opera-
tions for water and waste reduction.
2. Evaluate the impact of production methods, technical changes in
equipment, conditioning of water and by-product development on water
use and reuse and pollution abatement.
3. Determine the economic implications for the several water and waste
reduction methods demonstrated in the project.
4. Formulate guides for the management of water and waterborne waste
in poultry processing.
Technical and research requirements in support of the project have been
provided by North Carolina State University. The University has respon-
sibility for:
1. Training key personnel to carry out measurement and control work
within the plant.
2. Supervising the sampling and testing of all process waste water to
determine both quantities and pollutional characteristics.
3. Providing guidance and coordination in the development and fabri-
cation of specialized equipment.
4. Coordinating technical changes in plant processes and installation
of equipment with plant management.
5. Providing systems evaluation and benefit-cost analyses for the several
economic alternatives demonstrated in the project.
*Respectively, Associate Professor, Department of Economics; Extension
Specialist, Department of Food Science; and Research Associate, Depart-
ment of Economics, North Carolina State University, Raleigh, N. C.
323
-------
6 Publishing results and related information developed in all phases
of the project including recommendations and guides for the management
of water and waste in poultry processing.
SPECIAL FEATURES
This project has a number of features which are of special interest.
1. The project was jointly developed and conducted by a federal agency,
Environmental Protection Agency; by industry, Gold Kist; and by an ed-
ucational institution, North Carolina State University.
2. An interdisciplinary research team has worked cooperatively. Mem-
bers have had training in microbiology, food science, engineering and
economics.
3. The project encompasses both water use and waste abatement through-
out the plant, from water intake through final waste water collection
and control. Systems analysis has been applied.
4. There has been a full-time staff in the plant working on the project
without assigned production-related responsibilities.
5. The University has provided supporting biological evaluation in all
phases of the project.
6. A cooperative working relationship with the product inspection staff
has been achieved which has enhanced the effectiveness of the project.
PLAN-OF-WORK
The plan-of-work was organized into three phases: collecting benchmark
information (6 months), technical development (12 months) and evaluation,
formulation of management guides and preparation of the final report
(6 months). A water and waste laboratory was set up in the plant for
use in preliminary studies and laboratory analysis.
Benchmark Information
Benchmark information was obtained on water and waste quantities, waste
water characteristics and biological characteristics of both the product
and water at selected points throughout the plant. A flow chart was
developed for identifying processes, water sources, sampling points,
by-product recovery points, product flows and waste water flows, Figure 1,
Processing waste water characteristics for the feather flume, eviscer-
ating flume, selected processes and total effluent are given in Table 1.
Water used in the plant is given in Table 2.
Technical Development Activities
Water has many uses in poultry processing including scalding, product pre-
paration, cooling the whole birds and parts, transporting wastes and
324
-------
FLOW CHART OF POULTRY PROCESSING PLANT
Potable
Water
r~
Empty!
Coops!
1
N.
_ Truck-Borne
Coops
1
Receiving
Area
1
Killing
Station
A
1
"1
e. Defeathering
1
Whole Bird
8> Washing
1
h. Evisceration
i
Final
Washing
1
k. Chilling
I
Grading ,
1. Weighing,
Packing
i
Refrigerated
"l
4
^ Blood __ J
""* Recovery ^
* T
_ _ Blood 4c
1 I
1
1 X
^ Feather .. „ Feathpr -g
^ ^ Flow Away ' Recovery 1
! i * T
.- 1 I Feathers X
I _ 1
n i
Offal offal _J J
V Flow Away ' Recovery \J,
__J Offal X
1
X
h-J
n
^
P\
Jf
^
Final Waste
m. Water Collec-
mmm*^ Product tion & Control
Delivery Trucks
By-Product
—^ Potable Water
—^ Process Water
Waste Water
X
\
Sewer
Figure 1. Flow Chart of Poultry Processing Plant
325
-------
Table 1. Benchmark data on water and waste, Gold Kist Plant, Durham, N. C., December, 1969
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Scalder Entry
Scalder Exit
Whole Bird Wash
Final Bird Wash
Giblet Chiller
Chiller I
Chiller II
Feather Flume
Eviscerating Flume
Plant Effluent
BOD
1,182
490
108
442
2,357
442
320
590
233
560
COD
2,080
986
243
662
3,959
692
435
1,078
514
722
Total
1,873
1,053
266
667
2,875
776
514
894
534
697
Solids
Dissolved
1,186
580
185
386
1,899
523
331
382
232
322
Suspended
687
473
81
281
976
253
183
512
302
375
— Grease
350
200
150
580
1,320
800
250
120
430
150
-------
Table 2. Measured water use for poultry processing, Gold Klst Poultry
Processing Plant, Durham, N. C., July, 1969
Process
1. Killing Station
2. Scalder
3. Pickers
4. Feather Flume
5. Neck Scalders
6. Whole Bird Washers
7. Defeather Cleanup Hose
(1 @ 1 hr.)
8. "Hang-Back" Belt
9. Eviscerating Trough
a. Hand Wash Outlets
b. Side Pan Wash
10. Final Bird Wash
11. Lung Vacuum Pump Effl.
12. Gizzard Machine
& Gib let Flumes
13. Evisc. Cleanup Hose
(2 @ 30 min. ea.)
14. Gib let Chiller
15. Neck Cutter
16. Chillers
17. Packing Ice
18. Bird Pickup
(10% in chillers)
19. Packing Cleanup Hoses
(3
-------
cleanup. Waste added to the water in these processes generates high-
strength waste waters when compared to normal municipal waste waters.
Most of these wastes are highly degradable by biological and chemical
waste treatment processes.
Relatively low water and sewer costs and the absence of restrictions and
surcharges on waste loadings have resulted in a low priority on research
and development of information. There is little information on in-plant
water and waste reduction methods. More importantly, traditional pro-
duction techniques are often not compatible with economic water and
waste management solutions.
Water reduction developments are outlined in Table 3 for the several
major processes in the poultry plant. Each section of the plant is sep-
arated by walls with the activities well defined in each area in the
regulations.
Further water reductions could be achieved by continued use of process
waters, studies of which are now under way. These two changes would
involve the use of the whole bird wash water in the scalder and the use
of chiller and final bird wash water in the gizzard splitting machine.
The potential savings are 120 gallons per minute.
The application of waste reduction methods has been much more limited
than water reduction methods. An improved blood recovery system using
troughs with high sides has been installed to contain the blood in the
killing area. The blood is shipped to the by-product processing plant
in a tank installed under the bed of the feather and offal truck. This
change has reduced cleanup requirements of labor and water and is ex-
pected to provide for almost complete recovery of blood from the killing
area. Stunners are used to reduce body action and the amount of blood
carried on the feathers into the scalding tank.
A small scale settling basin was installed in the early stage of tech-
nical development. A full scale settling basin will be installed to
provide for effective waste water management and control.
SUMMARY OF RESULTS
The establishment of the close relationship between water reduction and
control and improved efficiencies of pretreatment methods has been the
most important finding in the study thus far. By reducing the flow of
water and the variation in that flow, efficiency of the feather screen,
offal screen and settling basin has been increased. This relationship
is illustrated in Figure 2 which shows water use and waste loading on a
per bird basis.
The following reductions in total water use and waste loadings in the
waste water have been achieved:
Reduction in water use from 850,000 gallons per day to 620,000
gallons per day.
328
-------
Table 3. Water reduction development activities by area of the plant and changes in fresh water
use
Area of
plant
Evisceration
Activity —
Use of improved nozzles
Final bird washers
Hand washers
Cycling of side pan wash
Rearrangement of giblet handling
Reduction in
From
50 gpm
285 gpm
90 gpm
360 gpm
fresh water use
To
30 gpm
100 gpm
30 gpm
320 gpm
Scalding and
defeathering
N.
Use of improved nozzles
in whole bird washers
Substitution of recirculated
eviscerating flume water for
fresh water and new design on
feather flume
Use of chiller water in scalder
to replace fresh water
45 gpm
94 gpm
40 gpm
30 gpm
0
0
Cleanup
New high-pressure cleaning system
with foam
112,000 gpd
46,000 gpd
-------
OO
U)
o
•o
t-t
•rt
(0
C
O
T)
0)
10
0)
4J
CO
13 T-
12
11
10 -
, .05
Waste Discharged
.04
.03
.02
, r
o
N
D
F M A M J J
Monthly Records
0 N
D J F M
8s
en
rt
(D
s-
00
(D
I
o
o.
CO
tfl
o
o
(D
M
w
H-
H
a-
1969-
1970-
1971-
Figure 2. Quantity of Water and Waste Per Bird,
Gold Kist Plant, Durham, N. C.
-------
Reduction in waste load from 4,000 pounds of BOD per day to
1,500 pounds of BOD per day.
Reduction in BOD from 600 ppm to 290 ppm.
Reduction in grease from 200 ppm to 90 ppm.
Blood from the killing room has been effectively eliminated
from plant effluent.
Feathers in the plant effluent have been controlled.
The water and sewer rates were increased by 20 percent in July, 1970,
and a surcharge of $80 per 1,000 pounds of BOD was added. The surcharge
is levied on BOD concentration above 250 ppm. Even with these increases,
the water and sewer service costs for the plant have declined. These
changes are summarized in Table 4. Further reductions are expected as
changes are completed in the blood recovery system and a full scale
settling basin is installed.
Table 4. Water, sewer and surcharge costs for selected months
Item Month
July 1969 December 1970
Water $3,069 $2,157
Sewer 3,377 2,372
Surcharge -- 853
Total $6,446 $5,382
An evaluation of individual changes is now under way to determine the
economic feasibility of each change. A partial budget will be devel-
oped for each change and will include an analysis of capital requirements,
depreciation, maintenance and operating costs for equipment and process
changes. The changes are expected to provide savings in labor and mate-
rials as well as reductions in water and waste. There has been a small
increase in the level of by-product recovery due to efficiencies in
screening, isolation of blood and recovery of grease.
The poultry processing firm can meet restrictions on wastes and reduce
surcharges by employing available technology. Development of improved
methods can enhance the firm's ability to reduce water use and increase
waste abatement.
331
-------
ECONOMIC ASPECTS OF WATER AND WASTE REDUCTION
The reliance on water and air resources for assimilation of wastes has
increased significantly as a result of both a growing population and
rising per capita consumption. We would expect that waste dischargers
would be permitted to use the assimilative capacity of water, air and
land resources as long as these uses are not costly to society. The
cost to society has increased to the point that there has emerged a
national commitment to reduce pollution.
It now appears that individual firms and municipalities will be required
to internalize most, if not all, costs associated with pollution control.
A combination of direct restrictions and effluent charges (surcharges)
appears to be favored at this time as a means for internalizing pollution
control costs.
Water Reduction
The several water reduction methods employed in the plant are listed in
Table 3. The incremental cost of water reduction is represented in
Figure 3 by the curve MC. Incremental costs will rise in a stepwise
fashion as more costly methods are employed. Each method can be eval-
uated for cost effectiveness by determining the average total cost of each
method for reducing water use.
For a firm with incremental cost of reducing water use represented by
MC and using the amount of water represented by qo in Figure 3, water
and sewer costs could be reduced by employing water reduction methods
until water use is reduced by the amount q]_ (at which point water reduc-
tion is qi and water use is q0 - q^). At this point, the cost of re-
ducing water use by one more unit is just equal to the cost of purchasing
that unit of water.
Waste Reduction
The typical water and sewer rate structure with sewer charges levied as
a fixed percentage of the water charge does not provide economic incen-
tives for in-plant management and control of wastes. An increasin0
number of cities are establishing sewer surcharges to encourage indus-
trial and commercial firms to reduce their volume of waste, to distribute
sewage treatment costs more equitably among users and to finance the
expansion, construction and operation of treatment systems. Typically,
municipalities establish sewer surcharges which are equal to ihe average
total cost of waste treatment.
Waste treatment costs could be reduced by employing waste reduction methods
when a surcharge is imposed. An incremental cost curve for waste reduc-
tion, MC, is presented in Figure 4 for methods such as in-plant changes
and pretreatment. The incremental cost of waste abatement rises at an
increasing rate in a stepwise fashion as more costly methods are employed.
Development of measures which provide revenue (by-product recovery) or
lower costs (process and equipment modifications which reduce labor and
332
-------
•H
c
-------
•H
a
W
tJ
cO
o
Q
sc
Surcharge
In-Plant Waste Reduction
•Waste Discharged to Municipal System
Figure 4. Marginal Cost of Waste Reduction
334
-------
other input requirements) may result in net earnings from employing
waste reduction methods.
The marginal cost curve, MC, shows the added cost of removing a unit
of waste by waste reduction methods. The quantity of waste q^ would
be removed in the absence of surcharges if firms had adequate infor-
mation on waste reduction methods which produce revenue or lower costs.
In the absence of surcharges, the remaining quantity of waste q/ - q1
would be discharged to the municipal system for treatment for the fixed
amount of the sewer charge.
When a surcharge is levied, firms would find it profitable to remove
an additional amount q~ - q-i since waste reduction costs are less than
surcharge costs. The quantity of waste q^ - q2 would be discharged to
the municipal system for treatment with the quantity q - q^ treated by
the city for the fixed amount of the sewer charge. The firm would pay
surcharge SC on the quantity q3 - q2 for a total surcharge of SC(q3 - q2).
Combined Water and Waste Reduction
Water and waste reduction are complementary activities. Reducing the
amount of water and controlling the regularity of flow increase the
efficiency of screens and settling basins and provide an important tech-
nical linkage of the interrelatedness of water and waste reduction.
The rate structure for water and sewer services provides an additional
linkage between water and waste reduction. If the combined water and
sewer rate is less than the surcharge, firms will find it economically
feasible to dilute their waste waters and water use will increase. Munic-
ipalities wishing to set up a system of surcharges should examine care-
fully the relative levels of water and sewer rates and the surcharge
rate to avoid encouraging water use.
The addition of a surcharge increases the value of water under a typical
water and sewer rate structure. As reduction in water purchases occur,
less of the waste will be treated as part of the normal sewer charge
with surcharges levied on it. This may be more than offset by increased
efficiencies in waste reduction methods from reduced water use, however.
There are potentially significant gains to society from research and in-
formation programs in water and waste management and control. The typical
water and sewer rate structures have encouraged the dumping of waste and
excess use of water because of low rates and lack of economic incentives
for reducing wastes. Adjustments will occur rapidly as we gain a better
understanding of the changes required for effective management of water
and waste.
335
-------
REMOVAL AND RECOVERY
OF
FATTY MATERIALS
FROM
EDIBLE FAT AND OIL REFINERY EFFLUENTS
by
W. C. Seng*
INTRODUCTION
On July 10, 1968 Swift & Company accepted an EPA Research &
Development Grant (12060-DQV) of $249,307.00 or 70 percent of
Project cost to study the removal and recovery of fatty materials
from edible fat and oil refinery effluents at our Bradley, Illinois
Plant.
The Bradley Refinery is a modern high volume edible fat and oil
refinery engaged in all types of processing. Before the initiation
of the Grant, it was equipped with existing standard sewage treatment
facilities consisting of a large rectangular Skim Tank unit and an
Air Flotation cell comparable in design to a Pacific Separator.
In the United States there are about 250 to 300 plants processing
about 18 billion pounds of edible fats and oils annually. The
effluent of these plants is principally fatty material which is
difficult to treat in present sewage facilities.
The overall objectives of this project were to establish a flexible
and complete effluent treatment facility at the Bradley Refinery of
Swift & Company, and then to use these facilities to study the use
of coagulants, synthetic polymers, cathodic protection devices, proper
pH control and other instrumentation in connection with the Skim and
Air Flotation units to remove the fatty materials from the plant
waste and produce an effluent containing 400 ppm, or less, BOD, ether
solubles, and suspended solids.
It was a further principal objective of this project to study a
centrifugal system to separate and upgrade the quality of the recovered
*Swift & Company Research and Development Center, Oak Brook, Illinois.
337
-------
fatty materials such that a more saleable product could be obtained
which would offset part of the cost of the waste treatment. Finally,
a complete survey of individual plant waste streams was to be made.
Under the grant, full scale new equipment and modifications were
installed costing $150,000 of which $67,000 was for additions to the
basic water clarification system. The remaining $93,000 was for
the centrifugal oil recovery system.
PROCESS DESCRIPTION
Figure 1 is a simplified process flow diagram for both the water
clarification and the oil recovery systems.
Water Clarification System
All the plant waste drains into an existing below ground concrete
sump (upper left hand corner, Figure 1) having an effective retention
of about 3 minutes at the typical plant flow rate of 300 gpm. From
there it is pumped to the Skim unit.
Figure 2 is an over view of the existing Skimmer unit (left side) and
the Air Flotation cell (middle), Figure 3 shows the new waste
treatment building, the two agitated waste grease collection and
treatment tanks, and the storage tank for the dilute acid water phase
from the DeLaval centrifuge in the oil recovery system.
Before reaching the Skim Tank, the waste flow (from underground
header) passes through a "chemical mixing loop" (Figure 4) consisting
of 3 inch diameter pipe arranged as a horizontal hairpin turn. The
inlet (bottom) section is Teflon lined pipe, the rest is type 316
stainless steel. Near the inlet end the water phase from the DeLavel
centrifuge (oil recovery system) is recycled and injected into the
waste stream. Next, 66° Baume sulfuric acid is injected under automatic
control for adjustment of the raw waste pH. Five feet downstream from
the acid addition point, a 1 gpm sample stream is directed through a
Union Carbide pH probe cell (inside the stainless steel metal box).
The pH signal inputs to a Union Carbide Water Monitor Instrument
(Figure 5) located in the new waste treatment building. The pH signal
then inputs to a Foxboro Electronic Controller which, in turn, adjusts
the rate of 66° Baume sulfuric acid from a BIF Simplex Propsuperb
Pump, (Figure 6). The resultant pH is recorded on an Esterline Angus
Multipoint Recorder
The pH control system has succeeded in eliminating the extremes of
pH, that is above 9 and below 5, in the skimmer effluent. This
degree of control has very substantially improved the typical
efficiency of the system. However, closer control would be desirable.
But the waste is essentially unbuffered, so the addition of at least
a 3 to 5 minute surge tank in the raw waste line would be needed,
equipped with a mixer into which acid would be added.
338
-------
U)
C.J
Hv •••••••••••••••••W^* WV •nWPHHWvRnHHlHKIWHnpnWH
Figure 6. BIF Automatic Acid Pump (left center), and Canned Aci
center) and Caustic (right center) Transfer Pumps
-------
LO
JN
o
-------
Fieure 2. Overall View o
-------
(-0
-e-
Figure 3. New Waste Treatment Building.
-------
OJ
-e-
to
•••
_4._ Cheraic^l.Mix Loop wlthpH Probe
, . ..
lagnetl'c'fldwraeter. "
-------
GO
4*
*
Figure 5. Instrument Panel, with Temperature, Dissolved Oxygen,
Turbidity, and pH Indicators (left middle) and Recorder
(top left), Magnetic Flowmeter Recorder (top right)^ and
Controller for Acid Pump (lower right).
-------
The waste flow then proceeds around the chemical mix loop In
turbulent flow and exits through a Brooks magnetic flow meter with
Teflon lining and with a continuous electrode cleaning device. The
output of the magnetic flow meter is indicated, recorded and totalized
with a Brooks recorder.
The waste flow then proceeds to the Skimmer unit (Figure 7) which
has an effective length of 36-% feet, is 10 feet wide, and has a 5-foot
depth for an overall retention time of 46 minutes at 300 gpm. The Skimmer
is equipped with surface scraper blades to skim off the grease and
deliver it via a conveyor to a steam coil heated 400-gallon tank near
the effluent end. No provisions were made in the Skimmer for continuous
removal of settled solids. It must be cleaned out approximately once
a month.
The water Is then pumped to the Air Flotation unit. On the way It
passes through a Mixing Equipment Company line blender (Figure 8, far
right center). Air is injected directly into the bottom of the mixer
under a pressure of 30 to 45 psi at a rate equivalent to at least 4
percent by volume of the water processed. A 20 percent alum solution
is injected 20 feet upstream from the in-line mixer.
The waste stream proceeds through a pressure tank, a manual back pressure
valve, and then into the flotation cell (Figure 9). A pressure of at
least 30 pounds, and preferably 40 pounds, Is maintained in the
pressure tank. A polymer at 0.2 percent solution is injected just
down stream of the manual valve after the pressure tank.
The Air Flotation cell is 13 feet 6 inches in diameter by 10 feet
high liquid height, with a retention time of 36 minutes at 300 gpm.
Skimmings are discharged by gravity into an 850-gallon steam coil
heated steel tank. The clarified effluent discharges by gravity into
an underground sump (Figure 2, lower right) from where it is normally
pumped to the 200-foot by 300-foot aerated lagoon on the premises.
Effluent from the lagoon goes to the Kanakee Municipal Waste Treatment
Plant. However, effluent can be by-passed directly to the Municipal
Treatment Plant.
Two 300-gallon solution tanks for polymer and one 300-gallon solution
tank for alum, with Milton Roy piston type metering pumps for each
were located in the new waste treatment building (Figure 10). All
tanks are of Atlac 382 polyester fiberglass reinforced construction.
A sample stream of effluent from the Air Flotation unit is pumped to
Union Carbide dissolved oxygen and turbidity probe assemblies
and a Total Carbon Analyser inside the waste treatment building. A
similar temperature probe Is located right in the Air Flotation cell
effluent.
The temperature probe has functioned reliably, but the turbidity probe
was found unsatisfactory primarily because the two halves of the
prism came unglued on three occasions. When the unit was operative, It
did track the process quite well,
345
-------
; -
..
'- .- '• "
-------
ft**'' J .W'~ f •'-:'. ' '-.
^T'
' -' ••
4lH
*
- .
-^ . •
•••/--•
Pigure_8. _Air. Flotation Cell, with In-Line Blendp.r (far right center)^
In Supply Header, and Pressure Retention Tank (center).
..
-------
Figure 9. (right to left)~Pressure_Tank, Manual Back
ymer
Addition Pipe, and Air Flotation Cell. Treated Effluent Sample__
Pump (far left center).
-------
10. Alum and Polymer Solution Tanks
-------
The dissolved oxygen device functioned quite well except that it
required cleaning once an hour to obtain reasonably accurate readings.
Usually the dissolved oxygen content of the plant effluent was at
saturation.
Total Carbon Analyzer
The Union Carbide automatic total carbon analyzer was investigated
heavily over a long period of time. However, it was found unacceptable
mainly because its input disc filter and sample measurement orifices
became plugged rapidly with the waste grease.
Oil Recovery System
Skimmings from the Skimmer and Air Flotation units are heated and held
without agitation in the two "side tanks" as long as possible and are
then dewatered to remove about half their volume. The dewatered
skimmings are pumped to either one of the two large agitated treatment
tanks (Figure 11). Each one is 10 feet in diameter by 11 feet, 9 inches
high and has a capacity of 6,500 gallons and can normally accept 24 hours
collection of waste grease. While one is being used to collect the
grease, the other is chemically treated and centrifuged.
The two treatment tanks are constructed of fiberglass reinforced Atlac
382 polyester resin, with a hairpin steam coil and steam-temperature
controls. Each tank has a Mixing Equipment Company Lightnin' Mixer,
with a 5 hp motor and two axial flow turbines of 33 inches diameter
operating at 84 rpm, all wetted parts of type 316 stainless steel. As
the grease is collected, it is heated to 170°F and mixed continuously.
Concerning the steam coils, it became apparent that a heavy build up
of cake would form on these coils, requiring maintenance and cleaning
every few days. Therefore, to enable direct steam injection, 3/16"
holes were drilled in the steam coil, 2 inches apart, directly under
the lower agitator impellers. This change for direct steam injection
has been quite successful.
When a tank is to be treated, 50 percent sodium hydroxide and 66°
Baume sulfuric acid are pumped in and mixed for 15 to 30 minutes, each
in succession, bringing the pH first to 10 and then to 2.5.
After treatment, the waste grease is pumped through a 50 millimeter
Dorr-Oliver ceramic cyclone and then to a DeLaval Model PX-213 bowl
opening, disk stack type centrifuge equipped with automatic cycle
controls (Figure 12). The purpose of the ceramic cyclone is to remove
as much as possible of the sand and grit from the feed to the
centrifuge. The sludge underflow from the centrifuge and the under-
flow grit stream from the ceramic cyclone are pumped into a portable
scavenger bin, having a capacity of 2100 gallons (3 days' operation),
located outside the building.
350
-------
t
w
• ."•-,''-" • •• ' • : • ' i.- • . • -- :" ••• .." ,• "• ' .''•-' <•'" ' - -~ • .-'. "' '"'".'
"Figure I'f. Storage Tank_\le£t) for' DeLaval Centrifuge Water Phase,
-------
The water phase from the DeLaval centrifuge is pumped by the pressure
of the centrifuge to a 4700-gallon tank outside the treatment building
(Figure 11). It is of the same fiberglass and resin materials as the
two treatment tanks, but has no agitator, and has a direct steam
injection sparge header and temperature controls. DeLaval acid water
from this tank is recycled back to the chemical mix loop of the water
clarification system where it partially acidifies the raw waste. The
flow rate is adjusted manually to spread the flow over 24 hours of
operation.
The clarified recovered oil from the DeLaval centrifuge is pumped to
an existing outdoor storage tank.
EVALUATION OF FLOCCULANTS
In general, coagulants such as alum produce a pinpoint sized particle.
The role of the polymer is to produce a further agglomeration of these
pinpoint particles to a size that will be more amenable to separation.
Laboratory Screening Tests
These polymers (Table 1) were screened in laboratory tests using
graduate cylinders to simulate air flotation, along with a Hellige
Turbidometer to measure Jackson Turbidity Units. The polymer dosages
were 0 and 2.5 ppm, and coagulant dosages ranged from 0 to 500 ppm.
Table 1. Comparison of Different Polymer Systems
(Ranked in Order of Performance)
Polymer
Swift X-400
Swift X-700
American Cyanamide P-250
Dow NP-20
Nalco 673
Swift X-420
Swift X-lll
Charge
Anionic
Nonionic
Nonionic
Nonionic
Nonionic
Anionic
Cationic
Mean JTU
512
571
591
593
601
646
685
Anionic and nonionic polymers generally performed better than the single
cationic tested, especially at lower pH levels, however, they all
performed well, lowering turbidity 10 to 50 percent more than when
using coagulant alone. Zinc chloride and alum were both somewhat better
then ferric sulfate.
352
-------
Figure 12
Dorr-Oliver P-5Q Ceramic Cyclone (right center)
and~DeLaval 'PX-'21'3' Centrifuge.
-------
Sodium aluminate was also tested but was not effective because its
addition raised the final pH and increasing amounts resulted in poorer
clarity.
Additional laboratory studies were conducted in which the relationships
between surface charge, pH, and turbidity were investigated. The
effective Surface Potential (ESP) of the clarified phase was determined
on a Water Associates' Streaming Current Detector. A typical set of
curves obtained (Figure 13) is given where the sample was adjusted
initially to 8 pH, then alum treated, followed by addition of 2 ppm
Swift X-400 polymer.
All such data suggested that, as expected, the alum required to
achieve maximum clarity was reduced as the pH was lowered. The data
showed that sulfuric acid alone acts as an effective pre-floc agent
for coagulant aid at pH 4. It showed that the point of maximum clarity
(or minimum JTU) appears as the ESP approaches zero, but is not
necessarily maintained even though the ESP may remain near zero as the
curves move from left to right, or in the direction of increasing
alum dosage. It was also seen that the point of maximum clarity is
often rather sharply defined, particularly in the mid-alum dosage
and pH ranges.
To summarize the relationships between initial and final waste pH,
alum dosage, and turbidity, when using 2 ppm of Swift X-400, the curve
in Figure 14 was constructed from the above data. It connects all
points of minimum turbidity for the corresponding final pH. Tie lines
are used to show the pH of the waste sample before alum addition. This
figure was used as a guide to the best range of conditions to be
explored in the Bradley waste treatment system evaluations.
BRADLEY FLOCCULATION TESTS
Based upon the laboratory screening tests above, four polymers were
selected for field evaluation, namely: Swift X-400, American
Cyanamid P-250, Dow NP-20, and Nalco 670. Except for brief tests
with ferric sulfate, alum was used exclusively as the coagulant. The
tests with ferric sulfate showed that, although, results were good and
required perhaps somewhat less dosage compared to alum, considerable
difficulty was experienced in putting the ferric sulfate into solution
and in handling the sludge remaining. Furthermore, the ferric sulfate
colored the resultant recovered oil red. Zinc chloride was not used
because of concern over potential toxicology questions. Sodium aluminate
was not used because of the negative laboratory screening tests mentioned
earlier.
General Procedure
The tests were set up on a shift basis. Composite shift samples were
analyzed. The manually adjusted Milton Roy metering pumps for the
alum and polymers were set to maintain constant dosages corresponding
354
-------
Figure 13.
EFFECTS OF INITIAL pH AND ALUM DOSAGE
ON CLARIFIED PHASE
JTU
pH 1500
ESP 10 1000
+10 500
9 400
0 300
200 400 600 800 1000
PPM ALUM
-------
Figure 14.
TURBIDITY CURVE, AND INITIAL pH,
FINAL pH.AND ALUM DOSAGE REQUIRED
10.
9.
7.
6.
5.
MINIMUM TURBIDITY
CURVE
0 100 200 300 400 500 600 TOO 800 900
ALUM, PPM
-------
to the total waste in-put flow rates as measured by the Magnetic Flow
meter. The pH of the raw waste was controlled at levels per the
aforementioned Figure 14 to yield maximum treated effluent clarity as
much as possible.
The recovered skimmings were collected over a 24-hour period while
only one polymer was used. Then the skimmings were processed through
the centrifuge. In all, 184 shifts of data were obtained.
During the entire period the cathodic protection devices were in
operation at standard conditions.
Evaluation of Flocculant Test Data
All the data was subjected to statistical evaluations facilitated by
the use of a computer. Overall average data for each shift were not
greatly different. For this talk, average data for the 4:00 p.m. to
midnight shift are shown in Tables 2 and 3, beginning with the raw
waste. Flow rates actually varied from 100 to over .600 gpm, and
contaminant concentrations were up to 10 times the average shown.
Air flotation effluent contained an overall average of 401 ppm suspended
solids, 357 ppm ether solubles, and 741 ppm BOD.
Removal efficiencies for the Air Flotation cell were generally 2.5
times as high as for the Skimmer unit, i.e. 70 to 88 percent compared
with 20 to 40 percent. Overall removals of contaminants were 84.0
to 89.4 percent for suspended solids, 87.2 to 92.4 percent for ether
solubles, and 74.9 to 81.6 percent for BOD. Overall BOD removals
were typically 10 percent lower than for suspended solids and ether
solubles. This is explained in part because 50 to 150 ppm soluble
BOD are contained in the raw waste which the system does not remove.
357
-------
Table 2. Averages, All Data, Afternoon Shift
Bradley Flocculatlt Testa
Raw Waste
Gal. 138,407
gpm 288
pH 9.0
Sus. Solids, ppm 3,679
Ether So., ppm 3,984
BOD, ppm 4,012
Skimmer Effluent
pH 6.5
Sus. Sol., ppm 2,706
Sus. Sol., % removed 32.6
Ether Sol., ppm 3,195
Ether Sol., % removed 25.3
BOD, ppm 2,439
BOD, % removed 38.3
Air Flotation Effluent
Temperature 111
Alum, ppm 452
Polymer, ppm 2.59
pH 5.0
Sus. Sol., ppm 401
Sus. Sol., % removed 84.3
Ether Sol., ppm 357
Ether Sol., % removed 88.3
BOD, ppm 741
BOD, % removed 70.3
358
-------
Table 3. Percent Overall Contaminant Removals
Shift
Suspended Solids
Ether Solubles
BOD
Midnight
84.0
87.2
74.9
Morning
86.7
92.4
77.0
Afternoon
89.4
91.2
81.6
In summary, regression analyses performed on all the data led to the
following conclusions:
1. Best results were achieved for both the Air Flotation cell
and the Skim tank when the final Air Flotation pH was in
the range of 3.5 to 6.0.
2. Concerning the Air Flotation cell, good results generally
were obtained at all alum dosages ranging from 100 to 700
ppm, provided the pH of the stream was consistent with the
3.5 to 6.0 range.
3. The polymers all generally performed essentially equally
well at all dosages used. These data suggest that no more
than 2 ppm need be used as a practical matter.
4. The Bradley system, given the present waste load, is sub-
stantially undersized to produce a waste consistently under
400 ppm of all contaminants, particularly BOD.
COMPARATIVE COST FOR POLYMER FLOCCULENTS AND COAGULANTS
Table 4 lists comparative costs for the four polymers tested at the
Bradley system. Also given is the cost at a dosage of 1 ppm polymer in
500,000 gallons of treated water. Note that there is relatively little
difference in the daily cost for polymers.
Table 4. Costs for Polymers
Polymer
Swift X-400
Am. Cyan. P-250
Dow NP-20
Nalco 670
Pounds
5,000 up
5,000 up
5,000 up
5,000-19,999
Cost/Pound Cost/500,000 Gal.
$1.65
1.40
1.55
1.26
$6.87
5.83
6.46
5.25
359
-------
Table 5 lists costs for various coagulants, and daily costs to
treat 500,000 gallons at 100 ppro dosage. Note that alum is the
least expensive.
Table 5. Costs for Coagulants - Truckload Quantities
Coagulant Cost/Pound Cos t/5001 OOP Gal.
Alum, powder $0.0300 $12.50
Zinc Chloride, granular 0.1415 58.94
Ferric Sulphate, powder 0.0455 18.95
Sodium Aluminate, pulverized 0.1270 52.90
Ferric Chloride, powder 0.0650 27.07
INFLUENCE OF CATHODIC PROTECTION DEVICES
The cathodic protection system was operated during the entire period
when data was collected for evaluation of flocculants.
Summary results were:
1. Corrosion was brought under control in all areas where
cathodic protection was installed.
2. Metal surfaces below the water line remained free of adhering
deposits of fat and scale whereas previously a thick, firm
cake would form.
3. Concerning the effect of impressed current on flocculation
efficiency, the "noise" of the process variables prevented
clear cut results as to a beneficial effect on effluent contami-
nants content. Nevertheless, based upon extensive laboratory
and field work in other and similar applications, particularly
in recent months, impressed current does produce valuable
benefits on flocculation and clarification of waste waters.
OIL RECOVERY SYSTEM EVALUATION
When a tank was to be treated and processed, the pH was adjusted to
10, minimum, using 50 percent sodium hydroxide and allowed to mix
for a period of approximately a half hour. Then the pH was adjusted
to approximately 2.5 by the addition of concentrated 66° Baume sulfuric
acid. Again, it was mixed for a period of up to a half hour. For
several tests, only concentrated sulfuric acid was used to lower the
pH to 2.5.
360
-------
Initial Test Runs
Initial test operations showed that a 122 millimeter ring dam was
too large and would not provide good separation under any conditions.
Additionally, it was found that the machine must be operated with a
rather dilute sludge discharge so as to prevent caking and build-up
of solids within the bowl and serious errosion of the bowl periphery.
It became clear-that the presence of filter aids and asbestos fibers
used in the plant operations plugged up the vertical distribution
holes in the centrifuge disc stack and caused considerable errosion.
Therefore, Dorr-Oliver 50 millimeter and 25 millimeter ceramic cyclones
and a Bauer 3-inch diameter nylon cyclone were installed and tested
for removal of at least part of the fibers and gritty material from
the waste grease feed.
Final Tests
All subsequent tests were made using a 119, 116, and 114 mm ring
dam along with the 50 mm cyclone. The feed was shut off each time
the bowl was opened to "shoot", or discharge, the sludge phase.
The general results were as follows:
1. Neither oil quality nor oil recovery were affected signifi-
cantly by typical range for feed rate and composition. A
pH of 2.5 and a 180°F temperature were found optimum.
2. Equivalent results were achieved without using sodium
hydroxide in the chemical treatment but rather using sulfuric
acid alone. When only sulfuric acid was used, less than
half was needed,
3. Best performance was achieved when using a 114 millimeter
ring dam. Furthermore, no smaller ring dam is indicated.
4. a 45 second "on feed1cycle time was optimum.
Table 6 gives typical operating and analytical data for the oil
recovery system when using the 114 millimeter ring dam and a 45 second
"on feed" cycle plus 15 seconds "shoot" time for a total of 1 minute
per cycle. The recovered oil phase contained an average of 0,8
percent moisture, 98.9 percent ether solubles, 0.3 percent ether
insolubles, and 0.13 percent ash.
Table 6 also shows 88.9 percent of the oil in the feed as ether
solubles was recovered in the oil phase. Only 1.9 percent of the
original ash content went to the oil phase. The water phase typically
contained 1.8 percent of the original ether solubles and 43.1 percent
of the original ash. The total sludge contained 3.4 percent of the
original ether solubles and 50.8 percent of the original ash. About
38 percent of the ash removed in the total sludge was removed by the
361
-------
Table 6. Typical Oil Recovery System Data
Feed Oil Water Cyclone Sludge DeLaval Sludge Total Sludge
Rate, Ib./min.
Analyses
Moisture, %
Ether Sol., %
Ether Insol. , %
Ash, %
Distribution
M Ether Sol, to, %
Ash to, %
114 29
67.0 0.8
28.3 98.9
4.7 0.3
1.7 0.13
88,9
1.9
44
95.0
1.3
3.7
1.9
1.8
43.1
15
90.6
4.4
5.0
2.4
2.0
18.6
26
93.1
2.3
5.1
2.3
1.8
30.9
41
92.1
2.7
5.2
2.4
3.4
50.8
-------
cyclone. It is believed that at least 70 to 80 percent of the coarser
particles were removed in the cyclone underflow, thereby nearly
eliminating errosion.
Actually, over 95 percent of the ether solubles would be recovered
since only one half the sludge is lost after dewatering. The decanted
water from the sludge, plus the water phase from the centrifuge, are
recycled to the process.
Additional quality analyses are given in Table 7.
Table 7
AVERAGE ADDITIONAL QUALITY ANALYSES FOR THE DELAVAL RECOVERED OIL
FFA,% FAG Color Titer Sap. No. Unsap, % Iodine Value
21.9 21 34.9 198.2 2.5 65.2
The indicated value today of this oil for soap or animal feed uses is
in the range of 4-1/4 cents to 6-1/2 cents per pound. At 4-1/4 cents,
to be conservative, and based on a recovery of 7,000 pounds of oil/day
for 250 operating days (excluding weekend operations), a total
1,750,000 pounds annually would be recovered, having a value of $74,000.
ECONOMIC EVALUATION
Direct operating costs for the Bradley Water Clarification are shown
in Table 8, based on 500,000 gallons per day and Monday through Friday
operation. Annual depreciation charges were not included. However,
they would represent a minor percentage.
The total daily direct operating cost for the Water Clarification
System is $328, of which 38 percent is for chemicals, 5 percent for
utilities, 46 percent is for direct labor, and 11 percent is for
maintenance.
The total Waste Treatment System is actually operated 7 days a week.
However, on weekends the waste load from the plant is substantially
lower. However, it is believed that the direct operating costs shown
give a good representation of the situation.
363
-------
Table 8. Direct'Operating Costs For Water Clarification
Basis: 500,000 GPD Waste Flow, Monday through Friday.
Alum, 500 ppm, $0.03/lb. 62 19
Swift X-AOO Polymer, 2 ppm, $1.65/lb. 13 4
Sulfuric Acid, 66° Be. 1700 ppd, $0.03/lb. 51 15
Power, 1370 Kw hr., $0.0094/Kw. hr. 13 4
Steam, 3000 ppd, $1.00/1000 Ib. 3 1
Labor, 1 man/shift, $6.29/hr. (incl. fringes
but not supervision or material handling) 151 46
Maintenance (excluding depreciation) 5 11
Total 328 100
Total direct operating cost for the Oil Recovery System, Table 9, is
$171. of which 5 percent is for sulfuric acid, 29 percent for disposal
of the combined centrifuge and grit cyclone sludge (after removing
approximately 50 percent of its volume by decanting), 10 percent for
utilities, 44 percent for direct labor, and 12 percent for maintenance.
The 7,000 pounds of reclaimed oil obtained each day with a value of 4-1/4
cents per pound would yield $300 per day, or 60 percent of the grand
total waste treatment direct operating costs of $500 per day.
Table 9. Direct Operating Costs for Oil Recovery
Basis: 500,000 GPD Waste Flow, Monday through Friday.
Sulfuric Acid, 66° Be. $0.03/lb 8 5
Sludge Disposal, $150/2100-gal. load/3 days 50 29
Steam, 3000 ppd. $1.00/1000 Ibs 3 2
Power, 1440 Kwhr., $0.0094/Kw hr. 14 8
Direct Labor, 1.5 men/day, $6.29/hr. 76 44
Maintenance (excluding depreciation) 20 12
Total 171 100
364
-------
At the present time, inedible oil and tallow prices are about
25 percent above the average for the past 10 years. However, it is also
felt that the recovered oil may find a market where its value would
approach 6 to 6-1/2 cents per pound at present day prices. In any
case, it is believed a fair assumption that over the years, the
value of the recovered oil can be expected to offset at least this
indicated 60 percent of the direct operating cost of such a waste
treatment facility.
It will be noted that the cost of alum and sulfuric acid represent
a high percentage of the overall daily operating cost. The price of
3 cents per pound for alum was for 100-pound bags in quantity. The
price of sulfuric acid is for 55-gallon drums at 3 cents per pound.
Both costs can be reduced significantly if bulk quantities and storage
are used.
IN-PLANT SURVEY
Turning now to the in-plant survey, all the waste, except sanitary
waste, enters the treatment system via a single hot well. Plant
interior operations-which generate wastes consist of caustic refining,
bleaching, hydrogenation, and deodorization of vegetable oils, mainly
soybean, but also cottonseed, corn, peanut, and palm oils: shortening
and margarine manufacturing and packaging, and a large indoor tank
farm area. The outdoor facilities consist of many storage tanks for
raw materials plus a tank car and tank truck loading, unloading, and
washing area.
Time will not permit a detailed discussion, so to summarize, the
survey showed that the great bulk of waste flow and loading results
from general cleaning operations indoors and outdoors. Small amounts
come from the basic refining, hydrogenation, bleaching and deodorizing
operations.
Relationships between Production and Waste Load
Table 10 gives average and range production data for each production
area in thousands of pounds per 24-hour day, covering 42 days of
operations.
No correlations were found with waste flow or loading for individual
production areas or grand total production.
365
-------
Table 10. Production Data for Bradley Refinery
Production Area
Caustic RefiningA'
Hydrogenation
Bleaching
Deodorizing
Shortening Manufacturing
Margarine Manufacturing
A) All oils processed.
Production. 10QO lb/24 hr.
Average
541
290
630
815
552
156
Range
265-990
165-385
270-980
601-946
449-661
92-205
366
-------
BIOLOGICAL TREATMENT OF HIGH BOD
YEAST WASTES
by
Thomas P. Quirk*
EFFLUENT LOADINGS & CHARACTERISTICS
Source of Effluents
Manufacturing processes at the Production Facility included: production
of yeast, vinegar, baking powder, soups, seasonings, and the bottling of
alcoholic beverages.
Waste effluents were discharged in a variable pattern during the period
from 12:00 a.m. Monday to approximately 3:00 p..m. Friday.
The major portion of the BOD and flow loadings were attributable to the
yeast production processes. These processes included: molasses clarifi-
cation, fermentation and separation, equipment sterilization, and yeast
packaging. Spent beers resulting from the separation of yeast from the
fermentation liquors accounted for well over 90 percent of the total
plant effluent BOD.
Equipment sterilization and molasses clarification began early Monday
morning. The first fermentations of the week's production were completed
late Monday morning, and the first discharge of separator beer followed
immediately. The cycle of fermentation and separation continued until
early Friday morning. Equipment sterilization and plant cleanup began
at the cessation of the last separator beer discharge and continued
until late Friday afternoon. No process effluents normally occurred
from Friday evening until late Sunday evening.
Daily Waste Loadings
Waste loadings during production periods were highly variable. BOD con-
centration variations as high as 5 to 1 were experienced. The unpredic-
tability of daily effluent data required statistical analyses for the
development of process design loadings.
A statistical analysis in terms of loading averages and variations from
the average is shown in Table 1.
*Quirk, Lawler & Matusky Engineers, New York, N. Y.
367
-------
Table 1 Analysis of Daily Waste Characteristics
Mean Value Variation* as a
Item ppm Percent of the Mean
1. Effluent BOD 5300 70 to 130%
2. Effluent COD 7500 65 to 140%
3. Effluent Suspended Solids 780 60 to 150%
*Variation = Range of values encompassed by + or - one standard deviation
about the mean.
The average concentrations of the separator beer components and their
relative contributions to the total plant effluent are shown in Table 2.
Separator beers represent approximately 64 percent of the daily effluent
volume and 94 percent of the daily BOD discharge. Effluent flow rates
varied over any 24 hour period. A range of 200 gpm to 1500 gpm was ob-
served during production hours. During periods of intense rain, effluent
peaks exceeded 2000 gpm for short durations.
Table 2 Summary - Effluent BOD & Flow Components
Percent Percent Approximate BOD Concentration of:
Relative Relative Source Remainder
Source Volume BOD ppm ppm
1. First Separa-
tion Beer 25 78 17,400 1,600
2. First and
Second Sep-
aration Beers 47 90 10,000 1,040
3. First, Second
and Third Sep-
arator Beers 64 94 8,000 900
4. Composite Waste 100 100 5,400
Miscellaneous Waste Characteristics;
Temperatures of the effluent exhibited seasonal variation. Monthly average
effluent temperatures are presented in Table 3. Average temperatures vary
from a minimum of 65°F in January to a maximum of 85°F in July.
Daily temperature measurements exhibited variations of up to 5°F from the
monthly averages shown on Table 3.
Caustic cleanup wastes and formaldehyde sterilizing solutions are dis-
charged. Quantities were small and did not appear to effect the biologi-
cal treatment processes. The composite effluent pH varied between 5.3
to 6 ..8 averaging approximately 6.1. Biological processes normally op-
erated in the neutral pH range (6.5 to 8.0).
368
-------
Table 3 Monthly Average Effluent
Average
Temperature
Month °F
January 65
February 66
March 68
April 71
May 73
June 80
July 85
August 84
September 80
October 74
November 73
December 66
Only limited nutrient analyses of the effluent were made. Total nitro-
gen concentrations of 300 to 800 ppm were indicated for the composite
effluent. Nitrogen values agree with the nutrient content of yeast wastes
reported in the literature.
Phosphorous concentrations are estimated in the range of 50 ppm to 100 ppm.
Unit Waste Loadings
As most of the plant's waste load is attributable to yeast production,
correlations relating molasses consumption and waste loading were pos-
sible. Variations were too large, however, for the development of con-
sistent unit loading relationships on a daily basis. The effects of
daily variations were sufficiently damped, however, through the use of
a weekly time scale.
The relationship between total weekly effluent flow and total weekly
molasses consumption is shown on Figure 1. The slope of the relationship
is indicative of the total weekly quantity of effluent attributable to
yeast production, i.e. 2.26 gallons per pound of molasses consumed.
The intercept indicates the weekly flow resulting from plant processes,
other than those related to molasses consumption.
Weekly effluent, BOD Snd COD loadings were also correlated with molasses
consumption during periods for which effluent samples were available.
The relationship is shown in Figure 2. The data are not of sufficient
accuracy to determine, by extrapolation, the load contribution from
operations not associated with molasses consumption. However, unit
contributions of 0.12 pounds BOD and 0.175 pounds COD per pound of
molasses are considered representative in that the data have shown the
contributions from miscellaneous processes are small by comparison to
yeast production.
369
-------
bJ
o
LU
ID
U-
U.
UJ
7.0
6.0
5.0
o
_J
= 4.0
3.0
2.0
1.0
Figure |
TOTAL WEEKLY EFFLUENT FLOW
VS
MOLASSES CONSUMPTION
0 2 4 .6 .6 1.0 1.2 1.4 1.6 IB 2.0 2.2
TOTAL MOLASSES CONSUMPTION, M, MILLION LBS. / WEEK
370
-------
Figure 2
600
500
UJ
UJ
V)
00
o
o
o
o
o
a:
o
o
o
CO
u
u.
u.
UJ
400
300
200
100
WEEKLY ORGANIC WASTE LOAD
VS
MOLASSES CONSUMPTION
RELATIONSHIPS:
BOD/WEEK = 0.12$* MOL./WEEK^
COD/WEEK =0.175^4= MOL/WEEI^
0 .2 .4 .6 .8 1.0 L2 1.4 1.6 1.8 2.0 2.2
TOTAL MOLASSES CONSUMPTION, MILLION LBS. / WEEK
371
-------
PREVIOUS TREATMENT EXPERIENCE
The concentrations of dissolved organics in yeast plant effluents can
exceed those of average municipal wastes by a factor of. 20 or more, and
present a particular problem from the treatment standpoint.
Yeast plants have traditionally located within municipal boundaries and
have been served by municipal sewage treatment facilities. The resulting
dilutions have obviated any difficulties in stabilizing the concentrated
effluent using conventional treatment methods.
Data available from the engineering literature for yeast effluent treat-
ment are meager and inconclusive from a standpoint of a process design.
Published information generally relates to low effluent flow volumes or
considers process designs whose cost and land area requirements are not
generally applicable to all situations.
A small but extensive treatment facility for a similar effluent was
operated in the midwest. The production facility has been taken out of
operation. Treatment operations included; equalization, anaerobic
digestion, trickling filtration, gravity sedimentation, chlorinatian and
final effluent stabilization lagoons. Plant operating data were found
to be meager and in conflict with check sampling performed by State
regulatory engineers. Recovery and analyses of available treatment
plant operating records did not yield useful information.
Pilot plant studies were undertaken by Anheuser-Busch on yeast effluents
over a five year period. The effluent had a maximum concentration of
about 1/2 the average value shown on Table 2. Treatment processes
studied included: sedimentation, coagulation, aeration and a combin-
ation of digestion and trickling filtration. The digesters were fed
from a batch holding tank at steady rates with periodic sludge return
from a gravity final clarifier. Digester effluent was applied to a
rock filled trickling filter followed by a final clarifier and recir-
culation system. Records of pilot plant operation were destroyed.
Analysis of summary information, available in the literature indicated
the need for effective and continuous sludge recirculation in order
to reduce digester detention times and net sludge yield. Trickling
filter performance demonstrated process feasibility and indicated the
advisability of applying synthetic packing materials in place of the
traditional stone media.
Process design data, developed by comparatively recent university re-
search work on continuous digestion of soluble high BOD wastes, have
demonstrated the capability of achieving high BOD reductions with low
sludge yields while maintaining low detention times.
The lack of suitable literature data or field experience in continuous
treatment of high BOD yeast wastes required that independent pilot plant
studies be undertaken.
372
-------
BIOLOGICAL TREATMENT PROCESS
Compliance with treatment requirements will require BOD removals of 90
percent.
Alternative treatment methods considered include: chemical coagulation,
ion exchange, reverse osmosis and biological oxidation.
The capability of processes other than bio-oxidation were evaluated and
demonstrated that neither sufficient removal nor an economical process
could be obtained. Biological treatment methods were determined as the
treatment process of choice.
Biological treatment processes available included: anaerobic digestion,
trickling filtration, activated sludge and combinations of these flow sheets,
Anaerobic Biological Treatment
The anaerobic process is similar in flow sheet to activated sludge and
includes a digester, a secondary clarifier and secondary sludge re-
circulation. Digester contents may be heated and mixed. Digester
effluent may be degassed prior to secondary clarification. Process
characteristics include: the ability to treat variable and highly con-
centrated BOD effluents with minor operating adjustments ; the ability
to achieve high BOD reductions and absorb load fluctuations.; a minimum
production of biological sludge; minimum power consumption; and high
site requirements for process tankage.
Trickling Filtration Treatment
The trickling filter process would employ towers packed with synthetic
plastic media, recirculation of tower effluent and final clarification
of tower mixed liquor. Process characteristics include: the ability
to treat variable and highly concentrated BOD effluents with minor
operating adjustments; the production of relatively low quantities aof
excess biological sludge; a relatively low power consumption and a
minimum site requirement for process tankage.
Activated Sludge Treatment
The activated sludge process would employ a totally mixed aeration tank,
oxygen supply by sparged turbines aerators or by pure oxygen, secondary
clarification and sludge recirculation. Process characteristics include:
a need for equalization of BOD loading input, the need for special
oxygenating provisions for high demand satisfaction, a maximum produc-
tion of waste biological sludge, maximum power requirement and the
ability to achieve highest BOD removals. Site requirements for process
tankage are intermediate between other processes.
A flow sheet schematic for biological processes is shown on Figure 3.
373
-------
Figure 3
SCHEMATIC FLOW SHEET — Bl OLOGICAL TREATMENT
YEAST PLANT EFFLUENTS
ANAEROBIC & ACTIVATED SLUDGE
PROCESS SCHEMATIC
SLUDGE YIELD
**TO~blSPOSAL~
_ _RECIRCULATED SLUDGE
t
DEGASIFIER
TREATED
•EFFLUENT
^ 1 UNTREATED^
^ EFFLUENT^
MIXED
LIQUOR
SEPARATOR
TRICKLING FILTRATION PROCESS SCHEMATIC
S_LUpG£ YIELD
TO DISPOSAL
BIOLOGICAL
REACTOR
MIXED LIQUOR RECIRCULATION
TREATED
EFFLUENT
-
"
^ MIXED
^JQi
UOR
MIXED
LIQUOR
SEPARATOR
UNTREATED
EFFLUENT
TRICKLING FILTER
374
-------
Anaerobic Digestion Process Design
The decomposition of soluble organics through anaerobic digestion proceeds
as a series of simultaneous steps. These reactions may be conveniently
expressed in two reactions as follows:
BOD+Organisms+Volatile + New + Alkalinity + Gas .... (1)
Acids Organisms (CaC03) (CH4+C02)
Organisms+Volatile+Volatile+Alkalinity + Gas
Acids Acids (CaC03) (Cfy+CC^) .... (2)
The conversion of complex organics to volatile acids is accompanied by
high organism growth rates and may be accomplished at low detention times.
As organism or detention time increases, volatile acids are metabolized
to stable end products (gases and alkalinity) and the organisms undergo
self-decomposition. This self-decomposition is termed endogenous
respiration.
Increased organism age results in increased endogenation and increased
gas yields. Both reactions are dependent upon: pH, temperature, mixing,
BOD concentration, and organism (sludge) concentration.
BOD removal n.ay either be direct by conversion of organic matter to
alkalinity and gas or indirect by conversion of organic matter to organ-
isms. Separation of these organisms from the digester effluent is re-
quired to complete the removal process.
In anaerobic systems without sludge recirculation, the organism age is
low and BOD removal is largely indirect, accompanied by high sludge
yields. In recirculated systems, sludge age is high and BOD removal is
mainly direct and accompanied by low net sludge production.
Analysis of the anaerobic process required a definition of the following
components:
1. BOD removal
2. Sludge production
3. Gas production
4. Solid-liquid separation and sludge compaction
The development of process design models for selected components is
summarized below.
BOD Removal Model
BOD removal is obtained by simultaneous operation of two biological path-
ways. In the presence of organisms, BOD is initially converted to organ-
ic acids which are then processed into the end products of: sludge,
methane, and carbon dioxide. The interaction of both pathways determines
375
-------
the rate and extent of BOD removal, biological sludge production and gas
generation.
Definition of the hydraulic conditions in the digester and a selection of
a suitable kinetic law for the treatment reactions were used to develop
a process design.
Data correlation was achieved using a model for a totally mixed, digester
operating under the kenetics of a first order reaction.
The first order reaction model may be expressed in a number of forms.
Expressions of particular interest relate the efficiency of BOD removal
and the velocity of BOD removal to combinations of the remaining process
parameters. The formulation for efficiency is most convenient for de-
sign computations and expresses removal as a function of the biological
rate constant and the product of mixed liquor Sludge concentration and
digester detention time as follows:
Fo = (k1) (Sn . Tp)
*R (k') (SD . TD) +1 . . . .. .............. U)
Where: ER = BOD removal efficiency expressed as a fraction
k1 = a specific biological rate constant expressed in
units of 1/ppm x days
ST> = digester mixed liquor sludge concentration expressed
in units of ppm
TD = digester detention time, in days, for raw waste flow
For a given value of (k'), specifically for the effluent and temperature
of operation, the solids time product (Sp.Tj)) required to obtain a sel-
ected BOD removal (Eg) may be determined. Detention time is determined
by selection of a mixed liquor solids concentration. Detention time is
minimum for that value of Sp which represents a maximum operating value.
Equation (1) may be represented as a linear relationship on Cartesian
coordinates yielding a graphical procedure for model verification and
determination of the rate constant k'. The linear relationship is
obtained by plotting I/ER as ordinate versus
While useful for design computation, the efficiency relationship has
limited utility in the analysis of pilot plant data when the process is
operating within a narrow range of efficiencies. (SpTD) is an easily
measurable parameter and represents a major pilot plant operating para-
meter of wide 'variation. .Efficiency on the other hand usually experiences
a less dramatic change during pilot plant investigations, especially
when excess sludge wasting is held to a minimum. This was the case in
certain series of these investigations where ER varied from 80 to 95
percent while (SD . TD) varied from 3000 to 14,000 ppm/day. For the
analysis of data of this type equation (1) can be rearranged to incor-
porate less sensitive process parameters as follows:
376
-------
La = I + SDTD (2)
UL k'
where: La = BOD applied ppm (adjusted, as necessary, for changes
in reactor BOD storage).
UL = BOD removal velocity, # BOD/# Sludge/day.
k1 = specific biological constant, 1/ppm x days
The BOD removal velocity UL roay be otherwise stated as the pounds of BOD
removed per day, per pound of sludge solids.
Equation (2) relates the concentration of the BOD applied and the
velocity of BOD removal to the environment of the digester as described
by the solids-time product. The relationship may be expressed in a
linear format on Cartesian coordinates by plotting the ratio (La/UL), as
ordinate, versus the solids-time product (S~ . TD) . The slope of the
correlation is 1.0 and the intercept (1/k1).
Data obtained in the study were analyzed in accordance with these forms
of a first order kinetic model.
Temperature Effects on BOD Removal
The value of (k1) increases with increasing temperature. Correlation
of temperature effects by the Arrhenius relationship yields the following
statement:
(k')t = (k')35° (9) t-35 (3)
Where: (k')t = rate constant at temperature (t)
(k')35° = rate constant at reference temperature of 35°C
9 = a temperature rate constant at 1.07 for anaerobic
digestion
t = digester temperature in °C
The effects of temperature can be significant. For example, a drop in
digester temperature from 95 °E to 75°F (35°C to 23.8°C) can require a
100 percent increase in digester detention time for the maintenance of
equivalent BOD removals.
BOD Removal Rate Constants
Anaerobic treatment data are divided into the following series:
Series #1 - Non-recirculated digestion
Series #2 - Sludge recirculated digestion
377
-------
Experimental equipment for high and low temperature work is shown on
Figure 4.
The BOD removal kinetics observed in each series are presented graphically
in Figures 5 and 6. Data abstracted from studies reported in the litera-
ture on the subject have been included in Figure 5 for comparison purposes.
The data show that significant increases in removal rates can be realized
through the recirculation of digester effluent solids. Reductions in
digester temperatures can also cause significant reductions in removal
rates. Failure of the digester process occurred at mixed liquor tem-
peratures of approximately 55°F.
Significant reductions in biological rates at low temperatures, coupled
with the creation of odor problems as temperatures are reduced, indicate
that the process of choice under this series would incorporate elevated
digester mixed liquor temperatures.
Sludge Production Model
A material balance across a digestion unit may be written for sludge
production in a manner comparable to that used for BOD removal.
Calculation of net sludge production over short balance period is
significiantly affected by solids accumulations within the reactor. The
effect of digester accumulations is pronounced in anaerobic systems in
which net sludge yields are low. Corrections for changes in storage or
operation for prolonged periods was necessary under these conditions.
Pilot plant data were analyzed in accordance with these requirements.
Reactions 1 and 2 depict the role of biological cells in the anaerobic
decomposition of organics. The organisms, initially produced through
the metabolic utilization of organics eventually undergo auto-gasifica-
tion or endogenous respiration as sludge age increases.
The net sludge production from the process is the difference between the
growth of new cells and endogenous respiration of the digester population.
Both reactions occur simultaneously and are related to BOD removal as
follows:
U
Where: Us
Ys UL - U's
(4)
the velocity of sludge production convenient units
are #sludge produced/#sludge present/day.
Ys = the gross metabolic organism yield, #solids/# BODr
UL = the velocity of BOD removal - convenient units are
#BOD removed/#sludge present/day.
Us = the velocity of sludge endogenation, convenient units
are frsludge endogenated/#sludge present/day.
378
-------
QUIRK, LAWLER & MATUSKY ENGINEERS
Figure 4
Main Pilot Plant - 21,000 Gallon
Digester and Surfpac Trickling Filter
View of Bench Scale 3 Liter
Digesters and Gas Collector
50 Gallon Pilot Plant Solonoid Feed,
Mechanical Mixers Effluent Container
and Settling Tank 37g
General View of Waste
Treatment Laboratory
-------
FIGURE
BOD REMOVAL KINETICS
ANAEROBIC DIGESTION SERIES #
NON-RECIRCULATED DIGESTION
YEAST EFFLUENTS
300
280
260
240
220
200
ISO
10 16°
i
0
. 140
*o I2°
100
80
60
40
20
C
MLSS LOAD k' x 10" 4
POINT SUBSTRATE BAS|S MEAS. PPM X DAYS
9 BEEF EXTRACT U) VSSe COO O.I
0 BEEF EXTRACT ID VSSe COD O.I
•*: YEAST EXTRACT(I) VSSe CARBON 1.8
B YEAST EXTRACT(I) VSS8 COO Z.8
©&• I8T BEER VSSe BOD 1 .0
+ TOTAL EFFLUENT VSSe BOD 1.8
(1) ANALYSIS OF LITERATURE DATA
RELATIONSHIP
/
' .
/ '
Vf
f
/
/
f
!TBEE
*#
f*
/
© <
f
^— y
Y/\
j^s~-
f
/€
r!t
OTAL
/
9
<2>
'/
EFFLU
*/
t
f
ENT
/
9
25
20
15
10
5
°C
i
5
0
5
0
0 A
'
^T* ^
^^*^r
?
/
y/
?
f
/»
y
y
/ft
& TOTAL CARBON
B REMOVABLE COP
> 4 8 12 16
1
) 20 40 60 80 100 120 140 160 180 200 220
SDTD • ICT3
380
-------
FIGURE 6
BOD REMOVAL KINETICS
ANAEROBIC DIGESTION SERIES *2
SLUDGE RECIRCULATED DIGESTION
YEAST EFFLUENT
RELATIONSHIP
1.6 •
I.4-
UJ
1.2- •
LEGEND
SYMBOL DIGESTER AVG. OPERATING
VOLUME TEMR °F
e
3 LITERS
50 GAL
21,000 GAL
20 GAL
98°
95°
95°
73°
'/«'
I
K' •= 10 X10"^ PPM-DAY
1.0
2.0
3.0
4.0
, '/PPM-DAY- I04
381
-------
Equation 4 is employed to determine values for the sludge production
constants Ys and Us by graphical analysis of sludge production and
BOD removal measurements. The relationship is linear on Cartesian
coordinates when sludge production (Us) is plotted, versus BOD removal
(UL). Sludge production constants Ys and Us are employed to determine:
the net sludge production from the process, the digester environment
required for minimum sludge production and the digester environment
for maximum BOD removal.
Equation (4) may be rearranged to yield a relationship for net sludge
yield per unit of BOD removed as follows:
= Ys -
(5)
"I
The equation illustrates the reduction in unit sludge production as
BOD removal velocity is reduced. The BOD removal velocity or zero
sludge production, i.e. AS/Lr = 0, is determined as:
(UL) min = U4
(6)
Design for operation at (U^) minimum will provide the minimum sludge
for disposal.
Table 4 summariEed the gross metabolic yield (Ys) and endogenous res-
piration velocity (U^) observed.
Series
1
Table 4 Summary - Biological Sludge Production
Parameters Anaerobic Digestion
Effluent Flowsheet
Non-Recirculated
Digester
Temp.
(F°)
Yc
(#/#)
Us
(//////Day)
a. Yeast Effluent
b. Beef Extract (1)
Sludge Recirculated
Yeast Effluent
Yeast Effluent
98 .30
98 .06 - .10
98
73
,45
.39
0
0
.25
.026
(1) Literature Data
The sludge yield of high temperature sludge recirculated anaerobic sys-
tems is considerably lower than comparable aerobic systems, e.g. an
overall sludge yield of 0.075 #/# of BOD removed can be anticipated
from the 95°F, sludge recirculated digestion. Sludge yields are in-
creased at lower temperatures through the reduction of the endogenous
velocity from 0.25 at 95° to 0.03 at 73°F.
382
-------
Minimum Sludge Production
Minimum sludge production occurs at the point at which sludge production
balances sludge endogenation.
Design conditions for operation at zero solids yield can be approximated
by a combination of Equations 2 and 4 to relate the solids time product
at balanced conditions with sludge production constants and the BOD removal
rate constant as follows:
(SDTD)B = (YC) (La) - 1_ ............... (7)
(us) k'
Where: (SDTD)B = the solids time product in ppm x days at which
biological sludge production is balanced by
sludge endogenation.
La = BOD concentration applied to the digester in ppm.
The need to increase the value of (SDTD) in proportion to increases in
the BOD concentration of the waste applied to the digester (La) is shown
by Equation 7 .
At a solids time products below the indicated balance point value, BOD
reductions are decreased and net production of biological sludge is
evidenced.
Table 5 summarizes the balance point values of digester solids time pro-
ducts for the series of effluents studied.
Table 5 Anaerobic Digestion Solids Time Products Required
For Minimum Sludge Production _
jaeries Temperature
°F In ppm x days
2 Sludge Recirculated Digester
a. Yeast Effluent 95 7,500
b. Yeast Effluent 73 68,500
Average data indicated that 30 percent of raw waste solids were gasified
during digestion at solids time products above 10,000 ppm x days.
Maximum BOD Removal
For a sludge recirculated digester, at a given temperature, maximum BOD
removal will be obtained at balance point operation. That is, at a solids
time product for which sludge production balances sludge endogenation.
At this point, the BOD concentration of the mixed liquor has been reduced
to the level for which the rate of BOD removal is sufficient for the
generation of new sludge in the amount required to balance sludge loss
from the total mass of mixed liquor solids.
383
-------
Utilization of the relationship between BOD removal velocity and BOD
concentration in a first order reaction, allows determination of the
minimum operating BOD as follows:
(L ) min = U^ (8)
Y It'
1$ . K.
Process design Equation 7 and 8 allow determination of maximum process
performance from measurements of: sludge production and endogenation
constants and the biological rate constant for BOD removal.
Table 6 summarizes the minimum BOD concentrations considered obtainable
from application of the anaerobic process to study effluents.
Table 6 Yeast Effluents: Minimum Soluble BOD Concentrations from
Anaerobic Treatment Using Sludge Recirculation
Series Effluent Temperature (Le) Min.
°F ppm
2 Yeast Effluent 95 555
Yeast Effluent 73 170
Solids-Liquid Separation & Sludge Compaction
Mixed liquor solids control, upon which the sludge recirculated anaerobic
digestion systems depends, is inturn dependent on the solids-liquid sep-
aration following the digester.
Separation processes include gravity sedimentation and flotation. Sep-
aration characteristics may often be enhanced by the addition of flocculant
aids such as polyelectrolytes.
Bench scale settling and compaction studies were performed in liter cylinders
equipped with stirring rakes under controlled temperature conditions.
An extensive series of laboratory experiments evaluating the effects of
chemical addition in the form of both primary and secondary coagulants on
sludge sedimentation characteristics were conducted. Test results were
negative.
Settling characteristics of digester solids vary with temperature, concen-
tration and the extent of gas entrainment. Sludge leaving the digester
may contain a large amount of entrapped digestion gases. The degree of
entrainment is a function of digester mixing and mixed liquor solids
concentration. Degasification of the sludge through the application of
vacuum has been shown to substantially increase sedimentation character-
istics .
Effluent suspended solids concentrations averaging 500-600 ppm were ex-
perienced in pilot plants operating without sludge degasification. Sludge
degasification would be expected to decrease clarifier effluent suspended
solids concentrations to below 200 ppm.
384
-------
Laboratory experiments in liter cylinders were conducted on degasified
and undegasified pilot plant sludges.
As the final clarifier will function both as a clarifier and thickener,
clarifier loading values in terms of gallons per day/square foot and Ibs
solids/day/ft . Normal municipal practice is to size clarifiers for
800 gpd/ft^ and thickeners at 10 ppd/ft2. Clarifier overflow rates of
less than 300 gpd/ft2 were indicated in the operation without degasifi-
cation. A sharp decrease in overflow rate was experienced as clarifier
influent solids exceed 5000 ppm.
Design criteria established through these analyses are summarized below:
1. Thickening rates are approximately doubled by sludge degasification.
2. Organisms continue to produce gas in the clarifier. This secondary
gasification may cause eventual solids flotation and contribute to
solids loss. Maximum allowable clarifier detention times to prevent
sludge flotation should not be exceeded.
3. Consistant operation of a gravity separator will require the in-
stallation of sludge degasification.
4. Sludge degasification will be required to attain high suspended
solids removal efficiencies.
5. The critical clarifier influent concentration beyond which hindered
settling occurs throughout the clarifier and overflow rates severely
reduced was indicated at 5200 ppm. Higher mixed liquor concentrations
will require recycle of clarified effluent to the clarifier feed well
to maintain clarifier performance.
TRICKLING FILTRATION PERFORMANCE
The classic trickling filter comprises a bed of stone media over which
attached biological slime growths develop. Removal of BOD is obtained
by aerobic processes at the slime surface and by anaerobic processes
within the slime interior.
Prior to the introduction of synthetic media for trickling filters, 1-1/2
to 3" crushed stone or slag was universely employed as a packing material.
Filter sizing was developed from municipal practice and utilized organic
loading expressed in #BOD/cf/day as a guide for design. Because of the
media plugging which resulted from application of concentrated wastes,
such as yeast effluent, operating experience for these effluents are limited.
Packed tower modifications to the trickling filter introduced plastic
geometric packing media to obtain increased surface area and porosity.
Current practice employs a lattice structure similar to an egg carton
insert. BOD removal performance is related to process parameters using
an analytical model descriptive of the biological relation observed and
the hydraulics of the reactors.
385
-------
Numerous design relationships have been developed for specific applications
of synthetic media. Application of these relationships failed to correlate
pilot data or literature data for similarly concentrated wastes in a consis-
tent manner. Wide variations in tower volume for a specified BOD removal
were obtained as different formulations were applied. Determination of
required tower volume by accumulation of consistent operating data for
major design variables, could not be employed because of the time limita-
tions imposed on study execution. An independent basis of design was
developed and confirmed for use in analyzing study data and in extrapolating
design requirements for filter application.
Experimental Procedures
The packed tower trickling filtration process was studied for treatment
of yeast effluent alone and for combinations of yeast effluent and
municipal sewage. Both waste flow conditions were found amenable to
treatment via bio-filtration.
Experimental studies utilized data from two pilot plant systems:
1. A Surfpac media pilot plant
2. A laboratory simulated trickling filter
The Dow pilot plant was installed at the yeast plant for independent
effluent treatment studies. As the need arose for evaluation of the
trickling filtration process for combinations of the industrial waste
and sewage, it became apparent that the transportation of large amounts
of sewage to the pilot plant for treatment studies would be too time
consuming and costly. Pilot plant operating flexibility under these
conditions would be curtailed.
A small scale laboratory trickling filter was constructed to provide a
high degree of flexibility. The laboratory system utilized the geometry
of an inclined plane over which the effluents were trickled. The system
allowed the observation of slime growth and BOD removal characteristics.
Features of the Surfpac pilot plant and the QL&M laboratory systems are
presented in Figures 7 and 8.
General descriptions of the operation of each of the systems follows:
1. Surfpac Pilot Plant
The 3' diameter, 21.5' high pilot plant was mounted on a concrete
platform in an outdoor area designated for waste treatment pilot
plants at the yeast plant. Pilot plant feed and recycle rates
were controlled by constant head orifice boxes located at the top
of the unit. The unit was insulated and contained provision for
the installation of a blower for supplementary air supply to the
media. Modifications to the manufacturer's original design in-
cluded the installation of a sump temperature control system and
a small metering pump for the maintenance of feed and recycle
386
-------
RAW a RECYCLE
ORIFICE BOXES \
RECYCLE
OVERFLOW
MOTORIZED
DISTRIBUTOR
SHELL—
36" Sch.~40
STEEL PIPE
NUTRIENT
FEED-
TANK
NUTRIENT
PUMP
SUPPORT
GRATING
FIGURE 7
SURFPAC
PILOT PLANT
^ D
RAW OVERFLOW
TREATED WASTE
RAW WASTE
9 X 9 X 12 FOOTING,
1/2" 0 BARS 6"C-C,
TOP a BOTTOM
NOTE: electrical control panel
not shown— dll motors
require 110/220 volts
387
-------
oo
00
TYPICAL LABORATORY SIMULATED
TRICKLING FILTER PLANE
SIDE VIEW
INCLINED PLANE
BAFFLED a WITH
REMOVABLE COVER
FRONT V!ZW
«g HEATING
TAPE FOR
TEMPERATURE
^CONTROL
REFRIGERATOR
iT'OR f-\
PUMP
COLLECTION
FUNNELS
UJ
o
a>
D:
FEED
K
MIXED LIQUOR SUMP
TOTAL SLIV.E
LENGTH WITH
BAFFLES = 9'-0"
THERMAL
REGULATOR
-•-EFFLUENT
0
MAGNETIC STIRRER
EFFLUEtJT
TO REFRIGERATOR
~rt
o
m
oo
-------
flows too low for accurate orifice control. The pilot plant
studies spanned 2 series. Series 1 investigated the use of
trickling filtration as a: polishing process in series with a
non-recirculated anaerobic digester. Series 2 investigaed
bio-filtration as a single process. The pilot plant was fed
four days to conform to the normal yeast production schedule.
Recycle from the pilot plant sump kept the biological slimes
wet and viable during off production weekend periods.
Operation of the pilot plant provided BOD removal data for the
development of a design model incorporating the effects of
waste concentration and hydraulic application velocity.
Prototype design criteria such as gpm/ft , the minimum ap-
plication velocity required to keep the media surface wet,
and the effect of high recycle ratios on BOD removal were
evaluated. The pilot plant operated over a wide range of
organic and hydraulic loading conditions.
Operating experience at low temperatures and with supplemen-
tation of tower air supplies by an external blower were
obtained.
Pilot plant experience demonstrated that approximately 3
weeks of operation were required prior to attaining a
representative material balance for BOD removal and slime
growth at any given loading condition. Because of these
time requirements experimental verification could not be
achieved for all potential design conditions.
2. Laboratory Trickling Filter System
The simulated laboratory trickling filter system was constructed
to provide flexibility in waste loading variation and system
operation. Daily feed volumes were small enough to allow the
entire filter effluent to be stored in a refrigerator for pre-
servation during feeding and collection. Slime surfaces were
large enough to provide a wide range of BOD reductions and
sludge yields.
The laboratory filters were rectangular plane surfaces set at
a 45° incline. The 8 foot length of each filter was baffled
to extend the effective slime length to approximately 9 feet.
The total surface area of slime available per unit was ap-
proximately 1.1 sq. ft.
Plane recirculation ratios, as well as BOD and hydraulic loadings
were scaled to approximate prototype conditions.
BOD Removal Model For Pretreatment
BOD removals up to approximately 70 percent in flow over an inclined
plane or through a full scale packed tower were shown to follow a zero
order reaction and to be related as follows:
389
-------
Inclined plane: E = k / (OL) ............... (9)
Surfpac Packed Tower: E = K / (OL) ............ (10)
Where: k = Specific biological rate constant in units of ppm
x gpm/SF or pounds/day/SF
K = Packed tower rate constant including effects of
media specific surface. Constant expressed in
units of ppm x gpm/CF or pounds/day/CF
OL = Organic loading as unit weight of untreated BOD per
unit area of slime area or filter volume
E = Efficiency of BOD removal
Influence of Temperature
The effect of temperature on reaction rate is introduced using the
Arrhenius relationship.
kt = k209AT
Where: k£ = reaction constant at temperature t
k£o= reaction constant at standard temperature, 20 °C
AT = reaction temperature differential °C-20
9 = constant, usually taken as 1.035
Graphical techniques used in correlating data for plane and tower per-
formance are illustrated on Figure 9.
BOD Removal Performance
During Series #1, the Surfpac pilot plant was not equipped with any
method of temperature control, as supplied by the manufacturer. The
steam operated temperature control system, installed in the field to
serve as an interim control measure, was incapable of maintaining tower
temperatures within a tolerance of + 5°F. Operating instability caused
by temperature variations and by other mechanical difficulties are
incorporated in the data.
During Series #2, an automatic temperature control system was installed.
Temperature variation was reduced to a range of +2°F.
Yeast effluent may vary in strength from 1000 to 15,000 ppm BOD. During
the operation of the pilot plants, facilities which partially equalized
1st beer were in use. Raw waste variations were reduced to a range of
2500 to 8000 ppm.
Organic loadings fluctuated as much as 300 percent for any given hydraulic
loading. BOD removal varied from 25 to 85 percent during these studies.
390
-------
Figure 9
ZERO ORDER MODEL CORRELATIONS
ANALYSIS OF SLIMED PLANE
Equation: (E)(OL) = (k2o°) 6AT
(OL)
ANALYSIS OF PACKED TOWER
Equation: (E)(OL) = (K20°)
(OL)
KT
9AT/E
391
-------
Analyses of plane and tower operating data, using the zero order BOD
removal model, are shown on Figure 10.
The laboratory plane, operating under more controlled conditions than
the pilot plant, was amenable to process design modelling for BOD
removals up to approximately 70 percent. A zero order reaction fitted
the operating data reasonably well. Performance above 70 percent re-
moval while attainable was not amenable to process modelling using any
of the normal methods.
Pilot plant data were widely scattered but were described best by a
zero order model.
For comparison purposes, laboratory plane data were scaled up to con-
form to the packing used in the pilot plant. Scaled up computation
of removals versus organic loading are shown with pilot performance.
Organic loading is expressed in units of ppm x gpm/CF for pilot plant
and laboratory data. Pilot plant data are converted to pounds/day/
1000 CF by multiplication by 12.0. Model prediction are valid for
design for loadings in excess of 200 #/day/1000 CF and removals less
than 70 percent.
Sludge Production and Effluent Settleability
During Series #1, the tower feed system, consisting of non-recirculated
digests: effluent, was high in suspended solids content, ranging from
1500 ppm to 2500 ppm. This extremely high influent suspended solids
concentration served to mask sludge production data. Series #2 data
showed variations of from 0.1 to over 1.0 Ib. suspended solids per
pound of BOD removed in net biological sludge yield, depending on
filter BOD removal efficiency and sloughing conditions.
Net biological sludge yields were observed and measured more closely
in the laboratory.
While correlations based upon the combination- of metabolic yield
endogenous respiration rates were not attempted, a net solids yield
of approximately 0.3 pounds of dry solids per pound of soluble BOD
removal is anticipated for a plant achieving 80_ percent BOD removal.
Sedimentation of the trickling filter effluents reduced suspended solids
concentrations to under 50 ppm. Overflow rates ranging from 600 to 800
gpd/ft are anticipated for secondary clarifiers.
Nutrient Requirements
Pilot plant studies showed that sufficient nutrient nitrogen and phos-
phorous are available in combination of yeast effluent and sewage to
sustain bio-oxidation at a high level.
During Series #1 of the Surfpac pilot plant operation the available
nutrients in the yeast effluent were substantially reduced by anaerobic
392
-------
FIGURE 10
ANALYSIS OF TRICKLING FILTRATION
YEAST WASTE
4. U
3.5
3.0
2.5
2.0
1.5
1.0
.5
0
70
60
50
40
20
1 0
o
A LABORATORY INCLINED PLANE
O FIELD PILOT PLANT
A LAB. PLANE SCALED UP TO FIELD CONDITIONS
NOTE ORGANIC LOADING EXPRESSED AS ppm x gpm/ft2 or /ft3
MULTIPLY BY 12 TO CONVERT TO Ibs/day/ft2 or ft3
^^
^^
^^
A
^*
A
«^
* A
O
«^
A
•**"^ L
0 _
<§>
G
^^
\ A
O
^*
0
0 0
A
A
0
^-^"-^"'
A
*^t
A
(
^-«
•°
0
^^
^^
O
I
^_^--'
A
.&_
*~*^
^^
^^
_^-
1^~^
^^
^20= '74 ppm gpm
^^
kT= 1
1
o
^^
o
5 ppm
f
r-~
+^~^
gpm
•(-3
^^
-^
A
^^
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8
a
H
2
§
M
§
U
393
-------
pretreatment. Nutrient supplement was utilized to insure consistent
performance. Nutrient supplementation during Series #2, produced no
substantial change in BOD removals.
Parallel laboratory planes run with and without nutrient supplementation
on yeast wastes mixed with various amounts of sewage showed no substan-
tial differences in BOD removals.
ACTIVATED SLUDGE TREATMENT
The activated sludge process was evaluated for treatment of combinations
at yeast effluents and municipal sewage. The process was considered
both for utilization as a single treatment process and also for use in
series with other biological processes.
Experimental Procedures
Batch-type investigations were used as the principal source of process
design data. A batch analysis refers to the instantaneous loading of
effluent into a reactor and the observation of the variation of system
parameters with time.
Sludge was continuously acclimated for batch reactor activated sludge
seed. The system employed is shown in Figure 11. Sludge generators
were fed at the rate of 0.1 Ib of BOD per pound of suspended solids.
An internal settling compartment provided continuous sludge recycle
within the generators. A minimum acclimation period of 10-15 mixed liquor
turnovers was used for all effluents. When acclimation was completed, the
sludges were sacrificed for BOD removal, settling and thickening inves-
tigations .
A total of five effluents were investigated in the activated sludge
studies. These included various blends of yeast effluents and sewage
and alternative forms of biological pretreatment.
The effect of packed tower bio-filtration pretreatment of a combined
effluent on activated sludge performance was studied in Series #1. This
process is generally considered efficient from the standpoint of cost per
pound of BOD removal when employed for pretreatment. The pretreated
effluent was generated from the laboratory trickling filter plane.
Series #2 evaluated the effects of anaerobic pretreatment.
Series #3 and #4 utilized combined effluents without pretreatment of two
dilutions with domestic sewage: a 1:3 and a 1:5.5 dilution (yeast flow
to sewage flow).
Characteristics of the batch removal studies are presented in Table 7.
Prototype design criteria required the verification of operating data
for a range of operating conditions. Operating parameters include:
aerator mixed liquor solids concentrations, aerator BOD loading and
detention time, temperature, oxygen consumption rates, and sludge pro-
duction levels.
394
-------
FIGURE 11
y-
CONTINUOUS ACTIVATED SLUDGE
ACCLIMATION SYSTEM
FLOW SCHEMATIC
WASTE STOHAGU
ROOM
TEMP. =
55 GAL V/ASTE
RESERVOIR
CONTINUOUSLY MIXED
GU
1
*
-------
Table 7 Yeast Effluents; Characteristics of Activated Sludge Batch Runs
Series Number
Characteris tics
Number of Runs per series 3332
Volume of Reactors, liters 10 10 10 10
Length of Runs, hours 24 24 24 24
Temperature of Runs, °C 20 20 20 20
Mixed Liquor Suspended Solids, ppm 2,400-3,340 1,440-3,000 2,100-5,800 2,500-4,600
Initial Soluble BOD, ppm 360-418 146-186 600-1,010 238-257
Soluble BOD, ppm After 24 hours Aeration 25-32 22-25 14-82 11-26
Average Oxygen Uptake Rate during first
4 hours, ppm/hr. 61-94 47-82 62-107 57-76
-------
Basis of Analysis
BOD removal in the activated sludge process is dependent upon both the
metabolic activity of the mixed liquor solids and the mixing characteristics
of the aeration tank.
Analytical models were developed to describe the following phases of the
activated sludge process.
1. BOD removal
2. Oxygen consumption
3. Sludge production
Prototype design formulations were also derived for gravity liquid-solids
separation.
The reaction for BOD removal involves the consumption of organics by organ-
isms for conversion to new cells and end products of C0£ and water.
The process reaction was expressed in terms of first order kinetics and plug
flow hydraulics:
Le = (i_E) = e-k'SA.T ................... (12)
^
Where: k' = a biological rate constant in 1/ppm x hr
SA = aerator mixed liquor solids in ppm
T = aeration detention time in hours
La = BOD applied to aerated tank in ppm
E = BOD removal efficiency
Data correlations using Equation 12 were made on semi-log paper, plotting
the log of Le/La versus SAT. The scope of the line of best fit is equal to
k'/2.34.
During batch oxidation, a sharp reduction in rate constant may be evidenced
as the reaction process approaches completion. A significant reactor BOD
concentration may also persist, even after prolonged periods of aeration.
In either case, the course of the reaction can be subdivided into two phases
and a rate constant established for each phase. In many instances the value
of the rate constant for the second or retardant phase may be so small in
comparison to that of the first phase as to be considered zero.
Recognizing the presence of a limiting effluent BOD concentration, the
parameters of Equation 12 are modified to describe the removal of that frac-
tion of the BOD above the limiting value. The modification takes the fol-
lowing form.
397
-------
La =
(13)
= Le _
Where: Lm = limiting BOD concentration, ppm
The biological reaction rate constant for organic removal above the limit-
ing value is then expressed as:
(14)
La
The value of the limiting concentration, Lm, may be obtained by analyzing
the velocity of BOD removal, UL = Lr/sA'T over small time intervals during
the course of the batch reaction. The velocity of removal will approach zero
as the BOD remaining approaches a limiting value. The 30D concentration at
which the velocity equals zero is equal to Ltn.
Temperature Effects
The effects of mixed liquor temperature changes on BOD removal rates are
expressed in terms of the Arrhenius relationship using 9 = 1.02.
The value of 9 = 1.02 has found general acceptance for activated sludge
performance.
During this study all batch removal runs were conducted at a constant tem-
perature of 20°C (68°F). Biological rate constants are reported at a
standard temperature of 20°C. Conversion of study values to prototype
design temperatures may be made through the use of the Arrhenius relation-
ship.
BOD Removal Performance
The BOD, suspended solids and oxygen uptake data obtained from a typical
batch run are shown in Figure 12.
Reactor BOD concentrations persisted after extended period of aeration.
Persistance of BOD concentrations may be attributed to a number of bio-
logical phenomena. The yeast effluents are high in organic nitrogen.
Conversion of the organic nitrogen to ammonia and the high sludge age of
the continuous sludge generators may have provided the conditions necessary
for nitrification. BOD values at the end of the reaction would then in-
clude the demand for oxidation of nitrogen as well as carbon.
Endogenation of mixed liquor organisms is accompanied by a lysis of cellular
material. Solubilization of lysed organics contributes to the presence of
BOD after extended aeration periods. The ratio between mixed liquor solids
and organics remaining in a batch reactor is considerably higher than that
of a continuous reactor operating with sludge recirculation. The BOD of
the organics derived from continuous sludge lysis in a batch reactor can
then represent a significant and persistant concentration of apparently
non-removable BOD.
398
-------
FIGURE 12
s
d.
Q.
A
\/^
LEGEND
\?Ql A 0 BOD
A ID A SUSPENDED SOLIDS
L Gl
^ — -- ^5 X
i
O
X
S
a.
a.
>
(/)
2"7 §
O
en
2-6 S
O
Z
,« UJ
,2.5 a.
c/>
2.4 K
O
3
O
2.3 3
Ul
2.2 X
Z
0 2 46 8 10 12 14 16 18 20 22 24
AERATION TIME , MRS.
X
2 100
o.
Q.
Id" 80
tr
UJ 60
ft^
* 40
Id
0 2O
X
0
^5
B^EI
N.
EJ ^%_^
^*^g~JSL
"~ • — -
— — a
02 46 8 10 12 14 16 18 20 22 24
AERATION TIME, MRS.
399
-------
FIGURE 13
ACTIVATED SLUDGE BOD REMOVAL
SERIES #4-YEAST EFFLUENT : SEWAGE = l'-5.5
20 40 60 80 100 120 140 ISO 180 200 220
BOD REMAINING , Le ppm
400
-------
FIGURE 14
A<
SERIEJ
PRE
"o
_j 1.0 .
N. 0.9
-~\ 0.8 .
« 0.7
-1 0.6
0.5
O
2 0.4
2 0.3.
<
QJ O2 .
cc
0
O OJ .
no °-Oj?
m o.o 8 :
0.07 .
0.06 .
0.05 .
0.04 .
0.03.
0 02.
0.01
(
AERA-
:TIVATED SLUDGE BOD REMOVAL
3 *. 1 YEAST EFFLUENT • SEWAGE = 1
TREATMENT BY TRICKLING FILTER
^^
^^-f*
-•
"b^
K 10
2.34= 2.36
SYMBOL
O
«
•
^V^
^^"^
x 10 /ppm-
LEGEND
UNIT fSA
(ppm)
1 3340
2 2400
3 3100
Lm = 83 ppm
• O
HR a^s
^^X. •
^v^^ (
A
>
N,^
^3
} 2000 4000 6000 8000 10,000
FOR SOLIDS TIME PRODUCT, SA -T, ppm- HOUR
401
-------
FIGURE 15
ACTIVATED SLUDGE BOD, REMOVAL
SERIES * 2 YEAST EFFLUENT : SEWAGE = M3
PRETREATMENT OF YEAST EFFLUENT BY ANAEROBIC DIGESTION
1.0.
09
o:I
0.7
0.6
0.5
04
0,3
02
0.1
o0.09
v-J 0.08
-„> 0.07
-1 0.06
- 0.05
(9
z 0-04
< 0.03
X
HI
" 0.02
o
o
CD
0.01
0.001
/
^
>_
)p
\.
\
0 V
f^.
\
^
0 ^v
o
(1V
°:%
i \
. , dv.
K«-i
" K' ' ^.
' \
i * ^
i
\»
yf
r\
\ <»
\ *
\
\
<
LEGEND
SYMBOL UNIT ,SA
(ppm)
O
3
• •
C'IOX2.34=2
K
\
\
\
Vm
1 3000
2 2400
3 440
_m = 28 ppm
.65xlO'4/,
>pm- f
HOUR
0 5000 10000 15000 20000 25000
VERATOR SOLIDS TIME PRODUCT, SA'T , ppm - HOUR
402
-------
FIGURE 16
AC
SERIES
1.0 -
n 3 .
o.§ .
0.7
0.6
0.5
0.4
0.3
0.2
~b
-I
_x
i °-QS
— ' 0.08
0.07
0.06
2 0.05
Z 0.04
^ 0.03
UJ
a 0.02
O
S °-01
0.001
AERA1
TIVATED SLUDGE BOD REMOVAL
** 3 YEAST EFFLUENT : SEWAGE = 1
"s.
^*v
iv
^
t^ ^^j
Xkjk.
%V
X
K'=K;ox2.
^ .
-, ^
i
34 - 1.26 x
X
6iNc~
^v
LEGEND
SYMBOL UNIT ,SA
(ppm
A
A
A
Lm = 50 ppm
1 580(
2 2IOC
3 290
)_
j
5
J
A
0 /ppm -HOUR
S.
>^
^^^
>
^
i
^
:3
0 5000 10,000 15,000 20,000 25,000
FOR SOLIDS TIME PRODUCT, SA T, ppm -HOUR
403
-------
FIGURE 17
ACTIVATED SLUDGE OXYGEN REQUIREMENTS
SERIES * I - YEAST EFFLUENT : SEWAGE = 1:3
PRETREATMENT BY TRICKLING FILTER
oc
X
o
O
o
_1
UJ
o
O.
V>
z
o
o
UJ
X
o
.05
.04
.03-
.02-
'.01
LEGEND
LA SA
POINT UNIT (PPM) (PPM)
O I 287 3340
O 2 395 2400
• 3 410 3100
.04
.05
BOD REMOVAL VELOCITY, UL, I/HR
404
-------
An illustration of the determination of the value of the limiting BOD
concentration for a typical batch series is presented on Figure 13.
The velocity of BOD removal, AL/SaAt, was determined at 30 minute inter-
vals and related to the average BOD present during the interval. The
first order kinetics of the reaction is illustrated by the linearity
between removal velocity and BOD concentration. The existence of a
limiting BOD concentration (1^) at which removal velocity reaches zero
is also evidenced. Graphical analyses of the type shown on Figure 13
were used to determine the value of (Lm) for each series of investiga-
tions.
Graphical analyses of the laboratory results, in accordance with Equation
14, for the,flow sheets studied are shown on Figures 14 to 16.
Table 8 summarizes the biological reaction rates and limiting BOD con-
centrations observed during each series. The rate constants shown are
compared with a k1 = 1.1 x 10"^/ppm x hours for average municipal sewage.
Biological pretreatment of the yeast effluent resulted in an increase in
the reaction rate.
Oxygen Consumption
Oxygen requirements are divided between growth of new cells and endo-
genation of existing cells.
The total oxygen consumption is related to BOD removal and sludge
endogenation as follows:
U0 = Y0 • UL + U0 (15)
Where: Uo = Oxygen consumption velocity, Ibs. of oxygen used per hour
per Ib. of mixed liquor sludge.
Y0 = Oxygen required for growth, #02 per Ibs. of BOD removed.
UL = BOD removal velocity, Ibs. of BOD removed per hour per ob.
of mixed liquor sludge
Uo = Endogenous oxygen velocity, Ib. oxygen required per hour
per Ib. of mixed liquor sludge.
Activate sludge oxygen requirements for batch studies are illustrated
using data from Series 1 as shown on Figure 17.
Values of YQ and U0 are summarized in Table 9. Values obtained from an
analysis of operating data of 45 municipal activated sludge plants ranging
in size from 1 to 800 mgd by Hazel tine, have been included for comparison.
405
-------
Table 8 Summary of Biological Reaction Rate
Series
Method of Pretreatment
Biological
Reaction Rate
(x 10-Vppm-hour)
3
4
Trickling Filtration of
Combined Effluent
Anaerobic Digestion of
the Yeast Effluent
None
None
2.36
2.65
1.26
1.93
406
-------
Table9 Yeast Effluents: Oxygen Consumption Parameters Batch
Activated Sludge Study
Series # YO U'o
(#02/#BOD removal) (#02/#SA-Hours)
1 0.54 0.011
2 1.0 0.020
3 0.32 0.013
4 0.67 0.010
Domestic Sewage
Average 0.50 0.004
Analyses of the economics of activated sludge application will require
consideration of the power consumption for aeration. Oxygen consumption
and power consumption for BOD removal will vary with the efficiency of
BOD removal and with the effluent combination being treated. Oxygen
consumption per pound of BOD removed may be used to evalute power needs
and is related to process design parameters as follows:
#02 = Y0 + U0
#Lr E-f (16)
02/Lr will reach a maximum at maximum BOD removal efficiencies. Power
requirements for mixing will depend upon tank geometry and equipment type.
Selection of aeration equipment will require consideration of the oxygen
uptake rates to be supplied. As BOD concentration to the aeration tanks
increase, uptake rates will increase to the point where mechanical aeration
or pure oxygen equipment will become desirable. The average aeration tank
uptake rate is expressed in terms of oxygen consumption velocity as follows:
UR = (U0) (Sa) (17)
Where:
UR = Average tank uptake rate in ppm/hr,
Uo = Oxygen consumption velocity in 1/hr.
Sa = Aeration mixed liquor solids concentration in ppm
The distribution of uptake rate throughout the aerated tank may be
determined after tank mixing characteristics have been selected.
Sludge Production
Net biological sludge production in an activated sludge unit is expressed
as the net of the growth and endogenation processes. Sludge yield due
to BOD growth is reduced by endogenation of the sludge mass under aeration.
A relationship similar to that derived for digestion may be written to
describe the net yield as follows:
M = Ys - M
Lr UL
407
-------
Where: A3 = Net sludge produced per pound of BOD removal.
Lr
Ys = Sludge yield in Ibs. of suspended solids/lb. of BOD
removed.
UL = Velocity of BOD removal, (I/day)
Us = Endogenous sludge velocity, (percent/day), Ibs. of
suspended solids destroyed per Ib. of mixed liquor
suspended solids per day.
The net solids yield during a batch removal study will only be in the
order of 100-200 ppm under the high BOD concentrations used in this
study. Mixed liquor concentrations in the order of 3000 ppm were main-
tained in the batch reactors. The yield will be 3 percent to 7 percent
of the total sludge mass under aeration. Laboratory suspended solids
measurements, via the Gooch crucible technique, are not generally within
this degree of accuracy. A 5.5 cm buchner funnel fitted with glass fiber
pads was substituted for the Gooch in an effort to obtain higher accuracy.
The suspended solids versus time data, for a typical batch run shown in
Figure 12 are an average of triplicate buchner funnel determinations.
Even with this amount of care, the suspended solids measurements exhibit
high scatter. Sludge yield values estimated from measurements of the
type shown on Figure 12 range from .8 to 1.5 for Ys and from 0.06 to 0.20
for Us.
Sludge yields may be approximated from the oxygen demand parameters.
Yield values using this estimation vary from .5 to .8 #/#.
Estimates of sludge production from activated sludge may be made using
these ranges as guides.
Solids - Liquid Separation
The use of high mixed liquor solids concentrations require that the solids
liquid separator following the aerator be designed for both effluent
clarification and sludge compaction.
The separator most often employed is a gravity clarifier. The unit is
designed to provide a clarified effluent as well as a thickened underflow
for return to the aerator and for subsequent sludge disposal facilities.
Choosingan aerator solids concentration of 3000 ppm and a thickener under-
flow concentration of 15,000 ppm, a compaction ratio of 5.0 is obtained.
A mass loading of 20 Ibs/SF day is calculated. In terms of clarifier de-
sign parameters an overflow rate of 480 gal/SF day is indicated.
408
-------
THE BETTER WHEY - A DILEMMA
by
Sidney Boxer*
Because of the current wave of environmental concern, the problem of
food-processing wastes has been moved to the front pages of newspapers
from the back benches of the laboratories. The dual pressure from the
public and the various governments has accelerated the need for de-
vising methods of abating the pollution by food wastes. Simultaneously,
the competitive demands of the market dictate minimal cost procedures
or, even better, the recovery of these wastes.
The dilemma faced by all food processors is the same - who goes first?
Unless all the processors in a particular food category are required
at the same time to abate their pollution, then, obviously, there will
be a cost disadvantage to those who comply. The answer might be, as
suggested by some Washington officials, a monetary assessment against
the non-compliers. I would go further and suggest that, if there is
an assessment, the funds be placed in an industry pool for research
and development to resolve the pollution problems of the particular
industry.
The cheese industry has the problem of whey waste, a heavy pollutant
because of its high BOD, and yet a food by-product with excellent nu-
tritive value. However, it costs at least $2,000,000., just for the
plant and equipment to recover the whey.
This poses another problem - the economics of the disposition of whey
clearly indicates that large scale processing is necessary. Of ap-
proximately 700 cheese plants in the country, only 10% produce suf-
ficient volume to justify their own whey drying plant.
DAIRY RESEARCH & DEVELOPMENT CORP., in its brochure to the cheese indus-
try, has suggested the establishment of regional drying plants to service
the smaller producers within a limited geographic area. We are presently
organizing nationwide marketing and distribution facilities which will
provide the smaller producers, as well as the larger ones, with the
opportunity of equal participation.
A brief background of the history of our project may be enlightening,
or at least evoke some sympathy. It all started with "Little Miss
Muffet" and her damnable curds and whey. That nursery rhyme has turned
into 23 billion Ibs. of liquid whey per year, equivalent on a BOD basis
to pollution by lo,000,000 people. To build sewage plants to accommodate
*President, Dairy Research & Development Corp., New York, N. Y.
409
-------
this waste would cost about $800,000,000., with an annual operating
expenditure of about $30,000,000.
Fortunately for us, Washington is impressed by large numbers and, I
must admit, the positive value of converting such a large quantity of
waste into edible food. In December, 1968, DRD was awarded a grant by
the then FWPCA of the Dept. of the Interior, now the WQO of the Environ-
mental Protection Agency, for its research and demonstration project
for elimination of pollution by whey, and utilization of the whey as
a food or in food products.
In developing the project, we had excellent cooperation from DAIRYLEA
of New York. It supplied the plant facilities at Vernon, New York, as
well as a great deal of engineering and similar assistance. DAIRYLEA
is to be commended for its foresight and willingness to cooperate in
this project, which will be of ultimate benefit to the entire cheese
industry.
After much soul-searching, we had decided that the only solution to the
disposal of whey was by the recovery of the whole whey as a food. Fol-
lowing three years of investigation into many systems, the most efficient
and economical method appeared to us to be spray drying. The technical
problem was to attain a large scale drying process that was economical
and yet productive of a fine, non-hygroscopic powder fit for human con-
sumption. We had a further problem, in that we were particularly to
undertake the drying of cottage cheese whey, of higher titratable acid-
ity than the whey from hard cheeses.
From the standpoint of time, we were fortunate that THE DeLAVAL SEPARATOR
COMPANY had been conducting experiments and had built a good size pilot
unit at its River Falls, Wisconsin, plant. Following numerous tests
over a period of two years at the pilot plant, we then commissioned
DeLAVAL to build a commercial plant for our demonstration project. They
also had to incorporate a special system for the purpose of producing a
non-hygroscopic powder from cottage cheese whey.
This unit was installed last July at our facilities in Vernon, New York,
and a number of shake down tests have taken place. We have discovered
a number of problems of interest to the chemical and technical men
amongst you.
The higher acidity of cottage cheese whey, a pH of about 4.5, can inhibit
the proper crystallization necessary to obtain the non-hygroscopicity of
the finished product. Even more, the per cent of solids from the evap-
orator to the crystallizer is critical, as well as the temperature, time
and cooling rate factors necessary to obtain the desired alpha crystals
and control mutarotation. Without a doubt, the evaporator must have
sufficient capacity to bring the liquid whey to a proper percentage of
solids before crystallization and drying. We do believe all the prob-
lems have been overcome and around April 1st, we expect to go into sub-
stantial production runs to demonstrate the feasibility of large scale
drying of cottage cheese whey to a non-hygroscopic powder. In the concept
410
-------
of recycling, whey offers the greatest challenge to remove it as a pol-
lutant, and the unlimited opportunity of utilization of a product of
high nutritive value. DRD is proceeding with research and development
of whey powder as additives, in beverages, baked goods, etc.
Not being an engineer, I will have to leave the more thorough discussion
of the chemical problems of cottage cheese whey, and the technical prob-
lems of drying the whey, to others more qualified at this conference.
I would like to say that the administrative and technical staffs of
both EPA and the Department of Agriculture have been most generous in
their assistance. There is no doubt in my mind that the partnership
of government and industry is necessary to solve the vast pollution
problems of this country and, indeed, the world. With government sup-
plying the "seed," and industry doing the hoeing and plowing, the nation
will reap the ecological benefits.
411
-------
MEMBRANE PROCESSING OF COTTAGE CHEESE WHEY
FOR POLLUTION ABATEMENT
by
Dr. Robert L. Goldsmith*, Dr. David J. Goldstein*,
Bernard S. Horton*, Sohrab Hossain*,and Dr. Robert R. Zall**
ABSTRACT
A two-step membrane process has been demonstrated for the treatment of
cottage cheese whey. The process produces valuable protein and lactose
by-products while reducing the BOD of the whey effluent. The process
has been studied in detail in prototype experiments at Abcor, Inc.,
Cambridge, Mass, and in a 10,000 Ibs/day pilot plant at Crowley's Milk
Company, Binghamton, New York.
In the two-step process, a protein concentrate is first recovered in an
ultrafiltration operation. In the second step, ultrafiltration permeate
(de-proteinized whey) is concentrated by reverse osmosis to provide a
lactose concentrate. The protein concentrate can be further concentrated
and/or dried; and the lactose concentrate can be further concentrated and
lactose recovered by crystallization, or otherwise processed. Alternatively,
the membrane concentrates can be used directly as fluid products.
Operation of the pilot plant was successful and almost troublefree. BOD
reduction of the raw whey was about 97 percent, from an initial value of
about 35,000 mg/£ to less than 1,000 mg/£. Membrane life was excellent,
and membrane fluxes were economically high. Membrane flux, membrane re-
jection, and membrane life for both ultrafiltration and reverse osmosis
sections of the pilot plant are discussed in this paper.
The pilot plant produced protein and lactose products with low total plate
counts, and nil coliform counts. Using the cleaning procedure developed
in the prototype program, total plate counts were typically below 50,000
org/ml, or less than the limit for Grade A milk. Thus, the pilot plant
was of a sanitary design and produced dairy grade products.
Projected capital cost for a 250,000 Ibs cottage cheese whey/day demon-
stration plant is $610,000. This includes both the ultrafiltration and
reverse osmosis sections of the plant, tanks, and a building to house the
plant. Projected operating costs are $196,000 per year. Projected income
from utilization of the protein and lactose concentrates makes the process
profitable and results in an attractive return on investment.
*Abcor, Inc., Cambridge, Mass.
**Crowley's Milk Co., Binghamton, N. Y.
413
-------
During 1971 the full-scale plant will be designed, built and put into
operation.
This project is partially financed by the Water Quality Office of the
Environmental Protection Agency under Grant No. 12060 DXF, to Crowley's
Milk Company.
INTRODUCTION
Background
The manufacture of cheese from either whole or skim milk produces, in
addition to the cheese itself, a greenish-yellow fluid known as whey.
Whole milk is used to produce natural and processed cheeses such as
cheddar, and the resulting fluid by-product is called sweet whey, with
a pH in the range of 5 to 7. Skim milk is the starting material for
cottage cheese and gives a fluid by-product called acid whey, with a pH
in the range of 4 to 5. The lower pH is a result of the acid developed
during or employed for coagulation.
Each pound of cheese produced results in five to ten pounds of raw fluid
whey. The high organic content of whey leads to a severe disposal problem.
However, these organic materials have a high nutrient content and, if
properly recovered, could provide useful products. Table 1 shows typical
compositions of whey and dried whey solids. Over 70 percent of the
nutrients from skim milk show up in acid whey, including soluble protein
and lactose.
The organic nutrients of whey, which go unused, place a costly burden on
sewage systems and waterways. The biological oxygen demand (BOD) of whey
has been noted as ranging from 32,000 to 60,000 ppm(l)2). Most of this
BOD is due to the lactose. Specific BOD values for cottage cheese wheys
are between 30,000 and 45,000 mg/£, depending primarily on the specific
cheese-making process used. Every 1,000 gallons per day of raw whey
discharged into a sewage treatment plant can impose a load equal to that
from 1,800 people. This is partially passed into streams in most cases
because BOD removal is not complete. Every 1,000 gallons of raw whey
discharged into a stream requires for its oxidation, the dissolved oxygen
in over 4,500,000 gallons of unpolluted water(3).
Table 1. Average composition of whey products.
Product Water N1Jj™fer°US ^ Lactose Acid Ash
Raw Cheese 93-94% 0.7-0.9% 0.05-0.6% 4.5-5.0% 0.2-0.6% 0.5-0.6%
Whey
Dried Whey 2-6 12-14 0.3-5.0 65-70 2-8 8-12
414
-------
Recent statistics give BOD's of about 0.2 pounds per pound of cottage
cheese curd. Combining this with production statistics, a total BOD
removal of over 200 million pounds would have been required in 1970
for complete waste treatment. Considering that at the very best only
half of the cottage cheese whey produced is currently put to good purpose
and that curd was water can contain up to 3 percent solids, the disposal
problem is severe.
The waste treatment technique described in this paper utilizes membrane
separation processes to remove organic material from acid (cottage cheese)
whey, thereby reducing its BOD to levels which can be easily handled.
This BOD reduction is accomplished in such a way that protein and lactose
or, as an alternative, concentrated whole whey, are produced; and an
economic credit rather than a cost for the overall process operation is
realized. Although the demonstration has been for acid whey treatment,
the technology is applicable to sweet wheys.
Proposed Membrane Process
A two-step membrane separation process has been developed for the treat-
ment of cottage cheese whey. In this process, whey is simultaneously
fractionated and concentrated to give protein and lactose byproducts.
The final effluent has a low biological oxygen demand (BOD) and is ex-
pected to be suitable for reuse within the cottage cheese plant. One
application, for example, is in curd washing. If the effluent is dis-
charged, a final treatment for residual BOD removal may be required,
depending on local and state regulations.
The two-step whey treatment process is based on the application of ultra-
filtration (UF) and reverse osmosis (RO). Reverse osmosis has been exten-
sively studied during the past ten years under funding from the Office of
Saline Water, the bulk of this work focusing on desalination of brackish
and sea waters. Additional publicity has been given to its application
to whey concentration by the United States Department of Agriculture^^,5).
Ultrafiltration is a variation of this membrane separation technique.
Figure 1 shows the basic concept involved. A semi-permeable membrane
separates water and a solution. In the absence of a hydrostatic pressure
differential, water will permeate the membrane so as to dilute the solution.
A counter pressure can be applied to the solution side to reverse water
transport. The amount of pressure required to achieve a static equilibrium
is termed the osmotic pressure. Reverse osmosis is simply the application
of a pressure greater than the osmotic pressure, which drives water from
the solution side of the membrane to the water side, and permits concen-
tration of the solution. The membrane plays the most critical role in
the process. By varying the properties of the membrane, one can control
the retention or passage of selected solutes. When the solute molecules
are large, for example whey proteins, the osmotic pressure is quite low
and a membrane with relatively large pores can be used at low operating
pressures, e.g., 10 to 100 psi. This is termed ultrafiltration. Referring
to Figure 2, the membrane may be called "loose," that is, lower molecular
415
-------
OSMOSIS
t ofm
\atrn.
WATER
•SOLUTION
•«—»
OSMOTIC
'PRESSURE
I arm.
1 arm.
DILUTED
SOLUTION
WATER PASSES THROUGH MEMBRANE TO
CAUSE DILUTION OF SOLUTION UNTIL OSMOTIC
EQUILIBRATION IS ACHIEVED
REVERSE OSMOSIS OR ULTRAFILTRATION
1 atm
WATER
PRESSURE
>OSMOTIC
PRESSURE Of
SOLUTION
PRESSURE IS USED TO DRIVE WATER FROM
SOLUTION
IN MANY CASES SELECTED SOLUTES ARE
ALSO DRIVEN FROM THE SOLUTION
FIGURE 1
-------
*LOOSE*
SKIN
~.3;
>20A* PORES
POROUS
SUPPORT
SKIN
-.Sji
-4A* PORES
POROUS
SUPPORT
FIGURE 2-FLOW THROUGH MEMBRANES
-------
weight solutes will pass through the membrane and will not be retained
in the concentrate. In reverse osmosis, solutions of small molecules
with moderate to high osmotic pressures are retained, and the required
driving force is considerably higher, ranging from several hundred psi
to over one thousand psi. Higher operating pressures are required
because of the substantial osmotic pressure of salt and sugar solutions,
and also because of the greater resistance to water transport of RO
membranes. Again referring to Figure 2, a "tight" membrane is to be
employed.
Figure 3 shows a simplified flow sheet for the two-step whey treatment
process. Cottage cheese whey, with or without filtration for fines
removal, is introduced into a low pressure UF unit (step 1). In this
operation, whey is concentrated 10- to 30-fold by volume. Ultrafiltration
membranes are used which retain only the whey proteins. Thus, it is
possible to obtain a protein concentrate with a higher proportion of
proteins in the dissolved solids, since lactose, non-protein nitrogen,
lactic acid, and minerals pass through the membrane. Operation is
typically in the pressure range of 10 to 100 psi, and at temperatures
of 60 to 130°F. The protein content of raw whey can be increased from
an initial value of about 0.6 percent up to levels approaching 20 per-
cent in this step. Composition of the protein concentrates on a dry
solids basis is shown as a function of the degree of water removal in
Figure 4. It is apparent that at water removal levels exceeding 95
percent, protein concentrate streams can be generated with a protein
composition up to 80 percent. Concentrates of this composition were in
fact generated in the course of the experimental program.
The permeate (ultrafiltrate) from the UF unit is introduced into a
second membrane step. In an RO operation, this stream is concentrated
from approximately 6 percent solids to 20-25 percent solids. Typical
operation would be in the pressure range, 500 to 1500 psi, and at a
temperature of 60 to 100°F. The membrane in the RO section is chosen
so as to retain as great a proportion of the organic solutes as possible,
resulting in the permeate having a low BOD.
The final effluent from the RO section can either be reused within the
dairy or cheese plant or discharged, with or without treatment for
residual BOD removal. Where pollution control regulations are not
overly stringent, a moderate salt rejection membrane can be used in
the RO section to permit partial desalting and lactic acid removal from
the lactose concentrate. In general, from the point of view of pollution
control, this option will probably not be exercised.
The protein concentrate can either be used directly by incorporation into
food products, or it can be dried. Drying may be preceded by concentra-
tion by vacuum evaporation. The lactose concentrate can be further con-
centrated by evaporation and the lactose can be recovered in a simple
crystallization operation.
418
-------
PROTEIN
CONCENTRATE
WHEY
STEP I
UF
10-100 PSIG
LACTOSE
Hz 0, LACTOSE
NON-PROTEIN N.
LACTIC ACID
SALTS
CONCENTRATE
STEP
RO
>200 PSIG
LOW BOD
WATER
FIGURE 3 -MEMBRANE PROCESS FOR WHEY
TREATMENT —FLOW SCHEMATIC
-------
Protein to
Lactose Ratio
Dry Solids
Water
Removed
0.13-0.15
0.20-0.25
0.40-0.50
1.9-2.2
3.9-4.5
6.5-7.3
7.5-8.3
9.5-10
17-18
24-25
0
60 70 80 90
% Water Removed by Ultrafiltration
FIGURE 4' COMPOSITION OF WHEY PROTEIN CONCENTRATE
420
-------
Another process alternative, which has been only briefly explored in
this program, is the direct concentration of whole whey by RO. This
approach has been examined by others(3>^»5), and appears to be attractive
for cheese producers with small volumes of whey. Concentration on-site
would precede transportation to, and processing at, a central facility.
However, a satisfactory cleaning program needs to be defined prior to
wide use.
Following extensive prototype experiments during a six-month period at
Abcor, Inc., a two-stage pilot plant was built. The unit, designed and
built by Abcor, was shipped to Binghamton, New York for a four-month
period of operation. The unit has since been shipped to and put in daily
operation at Crowley's LeFargeville facility.
This paper reports on the pilot plant operation. Results of prototype
experiments are discussed in detail in the Phase I Report (now in process
of review by WQO), and have been reported in part elsewhere^,7,8,9) m
The prototype experiments focused on the examination of a wide variety
of membrane systems and the effects of all major operating variables
(temperature, pressure, time, concentration level, etc.). The pilot
data are somewhat more restricted in scope, but cover the operation of
a moderate scale facility on-site in a dairy—a good test of the per-
formance and reliability of the equipment.
DESCRIPTION OF PILOT PLANT
A detailed description of the pilot plant and its operation is contained
in the Phase I Report (presently under WQO review prior to publication).
The following paragraphs present the major features of the pilot plant.
Ultrafiltration Section (Low Pressure)
2
The UF section of the pilot plant contains 210 ft membrane area (192 Abcor
HFA-180 tubular membranes). These are arranged in four parallel passes
(modules) contained in a stainless steel cabinet, fitted with sprays for
cleaning and sanitizing purposes. At a typical average flux of 13.5 gfd
for high conversion, approximately 1,000 Ibs of cottage cheese whey can be
processed per hour.
Operation was batchwise; that is, one batch (or more) of whey was charged
to the system per day and concentrated without adding additional whey.
During operation, the feed was volumetrically concentrated 20-fold or
higher, depending on the level of protein desired in the fluid concentrate.
At the end of the run protein concentrate was recovered from the membrane
unit by displacement with water or by draining.
The UF section is shown in Figures 5 and 6, installed on-site in Binghamton,
N.Y. In Figure 5 the stainless steel sanitary cabinet is shown. A closeup
of the tubular membranes inside the cabinet is given in Figure 6.
421
-------
Figure 5
Pilot Plant
Sanitary Cabinet Housing Ultrafiltration Membranes
Figure 6
Pilot Plant
Ultrafiltration Membranes in Sanitary Cabinet
422
-------
A separate clean-in-place system was used to mix and circulate cleaning
and sanitizing solutions through the membrane unit (membranes, pumps,
piping, etc.) and also through the cabinet spray pipes.
Reverse Osmosis Section (High Pressure)
2
The RO section of the pilot plant contains 210 ft membrane area in the
form of 42 American Standard TM 5-8 modules. All contain AS-197 membranes.
The modules are mounted in an open rack (Figure 7) for convenient cleaning
of the exterior surfaces. This equipment is presently manufactured by
Abcor, Inc.
The high pressure RO section was used either to concentrate the lactose
permeate from the UF section or whola whey - in either case producing a
lactose concentrate and a low BOD effluent.
The system was designed such that the feed could be processed batchwise
or continuously. For the latter, feed was taken from an interstage tank
and pumped through the system, continuously removing both a fluid concen-
trate and a low BOD permeate.
Figure 8 shows the control panels and pumps for the three pumping stations.
The interstage tank is shown in Figure 5, just to the right of the UF
membrane cabinet. In addition to accumulating permeate from the UF section
to be used as feed for the high pressure section, the interstage tank was
also used to mix cleaning and/or sanitizing solutions for the high pressure
system.
OPERATION OF PILOT PLANT
General Performance
During the period from July 22 through December 1, 1970, the pilot plant
was operated on a fairly regular basis. Operation was generally three
days per week, with a fresh shipment of whey arriving each Monday. During
this period, pilot plant performance was excellent. Capacity remained un-
changed with time, with the UF section having a capacity of 20,000 Ibs
whey/20-hour day, and the RO section having a capacity of 10,000 Ibs
whey/20-hour day. The BOD reduction observed in the bulk of the program
was from approximately 35,000 mg/£ to about 1,000 mg/£ (fresh, low-acid
whey). Some exceptions were observed, in particular during periods when
one (or more) of the RO high pressure modules developed a leak. In fact,
the sole aspect of the pilot plant operation which was not entirely satis-
factory related to the durability of the RO modules. At the beginning of
the pilot plant operation, several developed leaks and had to be replaced.
In addition, over the program period, four modules experienced tube
ruptures. For the last two and a half months of operation, however, per-
formance of the RO section was entirely satisfactory.
423
-------
Figure 7
Pilot Plant
Reverse Osmosis Section Modules
Figure 8
Pilot Plant
Reverse Osmosis Section Control Panels
424
-------
The system design proved to be sanitary and produced products with micro-
biological counts suitable as food or dairy products. Details of the
operation of the pilot plant are presented below.
Sampling and Analytical Procedures
Analyses listed in Table 2 were performed on a regular basis.
Table 2. Analyses used for Pilot Plant Program
Analysis References
COD (Frequently calibrated with
standard glucose solutions)
BOD
Kjeldahl Nitrogen Standard Methods: Water and
Wastewater, 12th Edition, 1965.
Total Solids Standard Methods: Dairy Products,
12th Edition, 1965.
Acidity (Lactic Acid)
Fat
Conductivity
Ultrafiltration Section
Procedure. All the experiments described involved unfiltered cottage
cheese whey. The whey was stored warm until used, with the storage
temperature ranging between 110 and 120°F. With this storage procedure,
operation was with whey with an acidity ranging between 0.5 and 1.3 per-
cent (expressed as lactic acid). This is evident from the data in Figure 9.
Whey was generally received with an acidity of about 0.5 percent (see
August 10, and 17, September 7, October 5 and 25, and November 30). In
the course of two to five days, the whey acidity increased substantially.
As will be shown in the discussion of the RO data, this increase in acidity
led to an increase in the effluent BOD. This relates to the fact that
lactic acid is only moderately retained by RO membranes.
Batch operation was performed according to the following procedure. After
a run, the membrane system was cleaned with AlcoZyme (Alconox, Inc.) at a
concentration of 1/2 oz/gal, according to procedures developed in prototype
experiments. After cleaning, the system was flushed with water and stored
425
-------
FIGURE 9
RAW WHEY ACIDITY
WHEY
ACIDITY %
1.5
1.0
0,5
n
/ / O / It ^&
/ / 0 rf»\ / / /
oo /Cb 7 o o o
MM) MM) 1 1 M 1 M I M Mill
JULY
AUGUST
SEPTEMBER
OCTOBER
NOVEMBER
-------
until the next experiment. Immediately before startup, the system was
sanitized with a solution of Antibac B (Wyandotte Chemicals Corp.) at a
concentration of 1/12 oz/gal.
Following flushing with whey, a small storage tank and the UF system were
filled with raw whey. Upon startup, whey was concentrated either for a
given period of time or until a given volume of whey had been processed.
During this period the small storage tank was kept full by the addition
of fresh whey. At a preselected time or whey concentration, the supply
of fresh whey to the small storage tank was shut off. At this point,
the system contents were concentrated until the capacity of the small
storage tank was exhausted. This latter operation is referred to as
"cook-down".
Dependence of Flux on Protein Concentration. In Figure 10 are presented
data for flux as a function of the retained protein concentration. The
data presented are for runs on four days. The runs of October 10 and 22
were performed with whey of a high acidity. In addition, during the run
of November 30, at high protein concentrations some acidification occurred
in the whey. It appears on the basis of these data, as well as general
observations on pilot plant performance (not discussed here), that some-
what lower fluxes were observed with high acidity whey. Also shown in
Figure 10 are the temperature profiles for the runs.
Flux decreased with increasing retained protein concentration, as expected.
Data are comparable to prototype data which are also shown (120°F and
13.5 gpm).
Dependence of Flux on Time (Life Data). In Figure 11 are shown flux data
for the pilot plant during the 4 1/2 months of operation. In the bottom
plot is shown an average flux for batch operation, to an average concen-
tration ratio (shown in middle plot) at an average temperature (shown in
the upper plot).
The average concentration ratio is the concentration ratio up to the point
of cook-down, and the average temperature is the temperature during this
operation. In general, scatter in average flux up to cook-down can be
related to variations in the concentration ratio and operating temperature.
Higher fluxes are correlatable with lower average concentration ratios and
higher operating temperatures. The most striking observation that can be
made from the data of Figure 11 is that flux was virtually unchanged over
the entire pilot plant operation. A mean average flux for a four-fold
concentration ratio was 14 gfd at about 108°F.
Additional flux data for the entire batch concentration, including cook-
down, are shown in Table 3. Shown are the date of operation, the protein
concentration obtained in the final concentrate, the total solids content
of the final concentrate, the percent protein on a dry solids basis, the
average flux, and the operating temperature.
427.
-------
130
UJ
no
100
90
FIGURE 10
PILOT PLANT ULTRAFILTRATION DATA
EFFECT OF FEED CONCENTRATION ON FLUX
t \ \ \
0,5
o
°0-00.
I I 1 I 1 M
10
20
15
10
0
0
1 1 1 i
1 I 1 1 1 ... 1
0,5 1
10
20
% RETAINED PROTEIN/ (Cp-Cp)
0 RUN OF 11/30/70
0 RUN OF 10/19/70
A RUN OF 10/22/70
$ RUN OF 7/22/70
15,5 GPM RECIRCULATION RATE
INLET PRESSURE = 60 PSIG;
OUTLET PRESSURE = 20 PSIG
,
<12(o,
428
v
GPM)
-------
FIGURE 11
PILOT PLANT ULTRAFILTRATION DATA'
WHEY FLUX VS, TIME
(15,5 GPM CIRCULATION RATE:
INLET PRESSURE 60 psi; OUTLET PRESSURE 20 PSI)
D
LU
OC
r3
»-
<£
cc
LU
a.
s:
LU
2
o
1—4
^_
< 0
cc -4
l- i-
2 <
LU 0£
U
2
O
O
«
u_
CD
X
3
_J
u.
LU
to
<
o:
LU
>
<
130
120
110
100
6
2
0
20
15
10
c
.?
— o
ooo
O Q
— (AVG.TEMP)O ° o o
wo o ° j o o j °o o o
Mill I 1 1 1 1 1 1 1 ! l°l 9 1 1 1 1 1 1 1 I M
0 0
_(AVG.CONC.RATIOl r»° o ^ o° ° °
• . . • • °« » \
O
"•^
II
1
1 M M 1 1 1 1 M
O o
—(MEAN AVG, Ftux)o °0 o ° o«
0 0 n 0 0° o °
Q O
e ° °
o
__
i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 I 1 1 1 1 1
•
JULY AUGUST SEPTEMBER OCTOBER NOVEMBER
429
-------
Table 3. Pilot Plant Ultrafiltration Data
Final Protein Final Protein Average Average
Date Concentration Total Solids (dry basis) Flux (gfd) Temperature
8/3
7/29
8/17
9/31
11/30
8.1%
6.4
14.6
8.3
9.3
14.6%
12.9
21.1
15.5
15.8
55.5%
50
69.5
53.5
59
14.4
10.7
13.7
14.5
13.0
110°F
100
114
112
114
Concentrates with protein percentages between 50 and 70 protein were
generated, and average fluxes ranged between 10.7 and 14.5 gfd. On the
basis of the data of Figure 11 and Table it is concluded that the
capacity of the UF section of the pilot plant remained constant during
pilot operation.
Membrane Retention Properties
Retention properties of the ultrafiltration membranes have been determined
for all major whey components. These are listed in Table 4.
Table 4. Retention of Whey Components in Ultrafiltration
Component Retention (%)
True Protein (Based on gel per-
meation chromatography analysis) 98+
Fat >90
Lactic Acid 10-20
Lactose 0
Minerals 0
Non-Protein Nitrogen 0
It can be concluded that the ultrafiltration membranes exhibited excellent
fractionation properties. Retained fat and lactic acid probably existed
in the whey as protein complexes and were therefore retained in the concen-
trate.
430
-------
Since these data were obtained at the end of the pilot program and since
no changes were made in the UF section during the program, it is concluded
that protein retention and fractionation were excellent throughout the
pilot program.
Reverse Osmosis Section
Procedure. All RO experiments except one with.whole whey were performed
with ultrafiltration permeate. Operating procedure involved cleanup by
flushing with tap water until the system was well purged. Following
storage but before startup, the system was sanitized with Antibac B at
a concentration of 1 oz/120 gal, and purged with UF whey-permeate.
Results for oxygen demand have been expressed in terms of COD, which was
the analysis generally performed. Separate analyses on a variety of
samples (raw whey, UF permeates, RO permeates) indicated the BOD level
to be about 50 percent of the COD level.
Dependence of Flux and Rejection on Feed Concentration. Flux and COD
rejection were determined as a function of feed concentration level.
Data for a batch experiment are shown in Figure 12. Flux as a function
of feed concentration, expressed as COD or approximate total solids, is
shown in the upper curve. In the lower figure, COD levels in the permeate
are given, and in the middle graph, COD rejection.
These data show an expected decline in flux with increasing feed concen-
tration. COD rejection was excellent, ranging from 98.5 to 99 percent.
Based on these data, the mixed permeate for a four-fold volumetric con-
centration would have a COD level of approximately 1,250 mg/-£. This
corresponds to a BOD level below 700 mg/£.
In general, somewhat higher flux levels were observed during the pilot
program. Figure 13 shows data for flux recorded during the period from
July 22, through October 21, 1970. Data points are shown for continuous
operation in three different temperature ranges. Numbers in parentheses
near data points indicate approximate operating pressure. It is apparent
that both increases in operating temperature and pressure resulted in
increases in flux.
The data have been scaled-up to an operating pressure of 750 psi in
Figure 1A. The scale-up factor is based solely on the ratio of operating
pressures; that is, fluxes are multipled by (750/operating pressure). This
is a conservative scale-up factor, since it neglects the osmotic pressure
of the feed, which is an important consideration especially at higher
feed concentrations. The data in Figures 13 and 1A are considered to be
an accurate indication of the flux-concentration profile observed in the
pilot plant program.
No trend in the data could be related to length of operating time covering
a four-month period, and on this basis it is concluded that negligible
membrane compaction occurred. This is confirmed by the absence of a decline
in membrane water flux (data not reported here) during the pilot program.
A31
-------
1—0
u. -
ID O
CO
X Q-
—J O
U_ CD
uu
LU
CD
O
8
6
2
0
100
99
98
2000
1500
1000
FIGURE 12
PILOT PLANT
REVERSE OSMOSIS DATA
RUN OF 8/31/70 (WHEY ACIDITY = 0,8%)
• Q——Q
UJ
Q_
500
1
1
0 50,000 100,000 150,000
FEED CONCENTRATION, COD (MG/A) OR APPROXIMATE
TOTAL SOLIDS (% x 10*1)
432
-------
10
8
0
FIGURE 13
PILOT PLANT REVERSE OSMOSIS DATA
EFFECT OF FEED CONCENTRATION ON FLUX
(DATA FROM 7/22 THROUGH 10/30)
FROM FIGURE 12, 84°F, 700
88°F-89°F
* 82°F-84°F
D 78°F-80°F
STAGE 1 STAGE 2
n o o
(730)
STAGE 3
I 1
PSI
50,000
100,000
200,000 300,000
FEED CONCENTRATION, COD (MG/*) OR
APPROXIMATE SOLIDS (% x 104)
-------
OJ
.C-
FIGURE 14
PILOT PLANT REVERSE OSMOSIS DATA
EFFECT OF FEED CONCENTRATION ON FLUX
(DATA FROM 7/22 THROUGH 10/30)
10
750 PSi, 88°F
(DESIGN BASE)
D
°87°-89°F
0 78°-80°F
%ATA OF FIGURE 12
(700 PS i,
0
I I 1
50,000
100,000
200,000 300,000
FEED CONCENTRATION, COD (MG/*)
ON APPROXIMATE SOLIDS (% x 104)
-------
Based on these data, a very conservative average flux for concentration
of UF permeate from 6 percent to 25 percent solids is 5.5 gfd, and this
value has been chosen as the design basis for the demonstration plant.
Lactic Acid Rejection. Pilot data showing lactic acid rejection are given
in Figure 15. Shown in the lower plot is the level of lactic acid in the
permeate as a function of the lactic acid in the feed. From the upper
plot, rejection is observed to be in the range of 60-70 percent. On the
basis of these and prototype data it is apparent that substantial lactic
acid passes through RO membranes, and this contributes appreciably to the
level of BOD in the final effluent.
Effect of Time on Rejection (Life Data). In Figure 16 overall oxygen demand
reduction data are presented for the pilot operation. Mixed permeate (final
effluent) COD and BOD levels are shown as a function of time. Also given
are the operating pressure and the feed whey acidity.
On July 22, the first day of operation, the effluent COD was about 4000 mg/£,
corresponding to a BOD of about 2000 mg/£. This was followed by a period
during which performance was unsatisfactory. This was related to visually
observable module leaks (riboflavin appeared in the permeate, giving a
green color). However, following the installation of a few new modules,
at the beginning and toward the end of September, COD rejection became
excellent. After the second replacement, performance continued to be
excellent through the end of November (the end of the pilot program) , at
which time COD rejection for all modules exceeded 97.5 percent.*
During this last portion of the program, variation in oxygen demand of the
effluent (mixed permeate) was related primarily to the whey acidity. Higher
acid levels resulted in higher effluent oxygen demand. BOD levels, on days
when low-acidity whey was used, were in the range 1,000 to 1,500 mg/£.
This corresponds to an overall BOD reduction of 96-97 percent, based on
raw whey BOD of 35,000 mg/£.
It is expected that operating pressure also played an important role,
although these data do not indicate any marked effect.
Earlier losses of rejection efficiency were due to the development of
leaks in a few modules. This is demonstrated by the data of Table 5 ,
which gives rejection data on Sept. 10 for a sampling of modules. It
is apparent that the bulk of the modules had acceptable to excellent
rejection. One module had a gross leak; two had smaller leaks; and one
was "off-spec".
At the end of the program, about 75 percent of the initial modules were
still in operation, and these had demonstrated a usable life of 4 1/2
months without detectable deterioration in flux or COD rejection.
COD - COD "
*Rejection is defined as - COD - permeate x 1QO ^ & given fegd
, . v l*t , CODRaw Whey " CODMixed Permeate
concentration. Reduction may be defined as — * ---
_ _ _.
Raw Whey
X 100 and will always be less than "point" rejection values.
435
-------
&•«
80
70
60
FIGURE 15
PILOT PLANT
REVERSE OSMOSIS DATA
LACTIC ACID REJECTION
RUN OF 8/W70
o
50
1,0
_
3 0.5
LU
-------
FIGURE 16
REVERSE OSMOSIS DATA
PILOT PLANT OXYGEN DEMAND REDUCTION
(RAW WHEY COD - 65,000 MG/*; BOD « 35,000
(* 4-FOLD CONCENTRATION TO 22-25% SOLIDS)
LU
CO
CO
-------
Table 5 COD Rejection Distribution of Modules
Module # COD Rejection, % Observations
1 99.3
2 99.2
3 99.2
4 99.0 good rejection
5 99.0
6 98.9
7 98.8
8 98.7
9 98.3
10 98.3
11 98.1 acceptable rejection
12 97.3
13 97.3
14 96.5
15 91.6 "OFF. SPEC.", unacceptable
16 86 leak,
17 84 leak,
18 50 gross leak
Concentration of Whole Whey. In a single experiment, raw cottage cheese
whey was concentrated in the RO section, and flux and rejection were
comparable to values obtained for the concentration of UF permeate.
Similar results were obtained in several prototype experiments. One
marked difference between performance with whole whey and UF permeate
relates to the ease of system clean-up. As will be seen in the micro-
biological data below, when UF permeate was processed in the RO section,
clean-up and sanitizing was relatively simple and effective. However,
in the experiment with whole whey, substantial difficulty was encountered
in clean-up. Probably the presence of residual casein fines as well as
the whey proteins contributed to both system fouling and the accumulation
of solids at system dead-ends. On the other hand, UF permeate is de-
proteinized and contains no suspended solids. Lactose, salts and the
438
-------
residual low molecular weight organics have high solubilities and can be
removed from the system simply by flushing with water. This difference
between concentrating whole whey and UF permeate by RO appears to be
quite important, in that currently available RO equipment is often
difficult to clean. Processing of UF permeate removes this process
limitation.
MICROBIOLOGICAL DATA
During the course of the pilot-plant program, samples were analyzed for
total viable bacteria and coliforms. In Table 6 are given the data
for different test dates. The dates run from the end of July through
the beginning of October. In the second column are given the dilution
ratios for the samples before culturing, The other columns give counts
for raw whey, UF concentrate, UF permeate (in interstage tank), RO feed
(from RO modules), the final lactose concentrate, and the RO permeate.
Several observations can be drawn. First, at no time did the counts
for either total viable bacteria or coliforms in the protein concentrate
exceed the level for Grade A milk (50,000 org/ml). This was the case even
though counts in the feed whey sometimes exceeded this level (see data
for July 31, August 18, September 29, and October 5).
Evidently , operation in the UF section at 110-120°F for several hours
did not lead to bacteriological growth. In fact, in certain cases
it appeared that decreases in plate counts were observed. These data
confirm the previously obtained prototype data which indicate that
growth of microorganisms in cottage cheese whey is not a factor of major
importance —.
In general the UF permeate, sampled in the interstage tank, also had
low total plate counts. There were some exceptions however, which are
attributed to improper cleaning. Although the counts on August 14,
September 29 and October 1 were high, proper cleaning of the system re-
sulted in a return to low counts. (See for example, data of August 18
and October 5.)
The feed to the RO section was taken from the interstage tank. Corres-
pondingly, on days when the UF permeate was contaminated (August 14,
September 29 and October 1), the feed to the RO section was also contamin-
ated. This explains the high counts on the same dates in the RO feed
samples. Additional data for October 12 and October 28 showed contamin-
ation of the high pressure section even though the UF permeate had low
counts. This is thought to be due to the fact that no sanitizer was
used in the RO system on those days.
In general, plate counts for the lactose concentrate followed the counts
for the feed to theRO system. At times, however, counts in the lactose
concentrate were higher than in the RO system (e.g., July 29), and this
is thought to be due to the use of a Tygon tube to drain off the lactose
concentrate. The tube ran only partially full, and microbiological
growth was observed on the tube's inside surface. Proper cleaning and
sanitising of the tube was not possible.
439
-------
Table 6 Total Plate Counts and E. Coli Counts
DATE
7/29/70
7/31/70
8/10/70
8/12/70
8/14/70
8/18/70
9/2/70
9/29/70
10/1/70
10/5/70
10/12/70
10/28/70
DILUTION
1/1000
1/10,000
1/1000
1/10,000
1/1000
1/10,000
1/1000
1/10,000
1/1000
1/1000
1/10,000
1/1000
1/10,000
1/1000
1/10,000
1/1000
1/10,000
1/1000
1/10,000
1/1000
1/10,000
1/1000
1/10,000
RAW WHEY
TPC E.COLI
12,000
20,000
41,000
111,000
<30,000 ,
30,000
<3,000
585,000
440,000
<3,000
<30,000 1
520,000
370,000
<30,000
<30,000
78,000
100,000
-------
Microorganism counts in RO permeates are based on samples taken from
the module permeate collection pans. These were generally not washed
or sanitized, accounting for the relatively high levels of microorganisms
on many dates. On days when the system was well sanitized (see for example,
September 2), however, counts in the permeates were low.
Samples from all sampling points, except for the UF concentrate, showed
contamination at some time. It is to be noted that proper cleaning and
sanitizing was very effective in lowering these counts. Observe the
data of October 5, which show very low counts, approaching levels for
grade A milk. Similar results are seen for September 2.
The final entry in the table gives representative data for an experiment
conducted over several days. At the end of the run on October 28, the
system was cleaned and sanitized. At this time it was filled with water,
which was sampled over a period exceeding one week. Bacteria counts for
the system for November 7 and November 9 are shown, indicating the pre-
sence of very few residual microorganisms.
It is concluded, therefore, that not only was the system design satis-
factory from the point of view of sanitation, but also that the system
could be stored for at least one week in water without substantial
microbiological contamination.
FULL-SCALE PLANT
During 1971 a 250,000 Ibs cottage cheese whey/day treatment facility will
be built and installed at Crowley's LeFargeville, New York location
facility. Based on the pilot plant data, capital and operating costs
have been developed.
CAPITAL COSTS
Capital cost items are given in Table 7. Costs for the ultrafiltration
and reverse osmosis sections are $180,000 and $310,000 respectively.
Six tanks will be required. This includes three 10,000 gallon tanks for
raw whey storage (one-day capacity). One 5,000 gallon surge tank will
be needed for the UF permeate (feed tank for the RO section). Two CIP
tanks will be needed for storage of cleaning and sanitizing solutions.
Total costs for the tanks, inlcuding piping and valves, is projected to
be $50,000.
Additional capital costs include $40,000 for installation, and $30,000
for a building to house the plant. Total capital costs for the plant
are projected to be $610,000.
441
-------
Table 7 Projected Capital Costs for 250,000 Ibs. Whey/Day Plant:
Equipment^
Ultrafiltration Section $180,000
Reverse Osmosis Section 310,000
Tanks
3 - 10,000 gals, plastic silos for whey storage
(1 day capacity)
1 - 5,000 gals. UF permeate surge tank
2 - CIP tanks for storage of cleaning and sanitizing
solutions
Total for 6 tanks with piping and valves 50,000
540,000
Installation Costs 40,000
Building Costs (3000 ft.2) 30,000
TOTAL CAPITAL COSTS $610,000
OPERATING COSTS
Operating costs are shown in Table 8. Among the items listed, costs
for power and cooling water are based on actual operating data from the
pilot plant. The membrane equipment itself, has been depreciated on two
bases. Membrane modules, less the annual charge for membrane replacement,
have been depreciated over five years. The associated membrane hard-
ware (pumps, valves, controlls, etc.) have been depreciated over ten years,
as has been the remaining capital cost items.
The annual operating costs will be approximately $196,000.
PROJECTED BY-PRODUCT VALUES AND PROCESS PROFITABILITY
Fractions obtained from the membrane process have useable characteristics.
In treating 80,000,000 Ibs. of cottage cheese whey per year, 480,000 Ibs.
of protein (at 0.6 percent in whey) and 3,600,000 Ibs. of lactose (at 4.5
percent in whey) will be recovered. The value for these by-products when
used as food ingredients will be sustantial in dollars. The net annual
profit, the difference between income and operating expenses, is expected
to show a highly favorable return on investment.
CONCLUSIONS AND RECOMMENDATIONS
Membrane processes offer an economically attractive solution for waste
treatment of cottage cheese whey. Pilot operation has shown that a two-step
442
-------
Table 8 Projected Operating Costs for 250,000 Ibs. Whey/Day Plant:
Basis: 320 days/yr or 7680 hrs/yr
Labor and Overhead
Supervisory 3 hrs/day x 320 days/yr @ $5.00/hr $ 4,800
Plant Operator 7,680 hrs x $3.00/.hr 23,040
Lab tester 2 hrs/day x 320 days/yr @ $3.00/hr 1,920
Mechanic 1 hr/day x 320 days/yr @ $4.00/hr 1,280
$31,040
Fringe and overhead @ 40 percent of wages 12,400
Total Labor $43,440
2
Annual Membrane Replacement cost at 5 $/ft 33,400
*Power (§1.2 c/kwh 5,000
*Cooling Water 4,000
Steam 3,000
Cleaning Chemicals (10 $/day) 3,200
Disposables (e.g., pump, gaskets, seals, etc.) 3,200
Professional Services (outside maintenance, costs,
engineering services, consultants) 10,000
Interest and Taxes @ 3 percent of Capital 18,300
Depreciation
Membrane Modules ex membranes @ 20 percent 32,500
Membrane process hardware (pumps, valves, etc.)
@ 10 percent 27,000
Remaining Capital Items @ 10 percent 12,000
Projecting Annual Operating Costs $195,540
*Actual costs based on scale-up of pilot plant utility usage.
443
-------
process successfully produces protein concentrates through whey
fractionation and a high concentration lactose concentrate through
reverse osmosis. The BOD of raw cottage cheese whey can be reduced
from approximately 35,000 mg/1 to less than 1,000 mg/1 by membrane
processing. Protein concentrates containing up to 20 percent protein
(80 percent protein on dry solids basis) can be generated in a single
step ultrafiltration. Lactose concentrates containing 20 percent lactose
(75 percent lactose on a dry solids basis) can be generated by con-
centration through reverse osmosis of the ultrafiltration permeate.
Whole cottage cheese whey can be concentrated to about 25 percent solids
by reverse osmosis without prior deproteinization (by ultrafiltration or
otherwise). Products produced from the two-step membrane process,
protein and lactose concentrates, have low total plate counts and nil
coliform counts. Products produced in the pilot plant operation met
standards for Grade A milk. Concentration of whole whey by reverse
osmosis is complicated by the difficulty of clean-up and sanitizing of
the membrane equipment. Prior deproteinization of whey eliminates this
problem. Installation and operation of a 250,000 Ib/day plant are
expected to yield a high return on investment before taxes, correspond-
ing to a rapid plant pay-out. This conclusion incorporates both capital
and operating costs, as well as information proprietary to Crowley's
Milk Company regarding the utilization and value of protein and lactose
products. Plant capital costs for a 250,000 Ibs whey/day installation
is projected to be $610,000; and annual operating costs, $196,000.
Crowley's Milk Company recommends the installation and operation of a
250,000 Ibs cottage cheese whey/day membrane processing facility at the
LeFargeville, N. Y. plant. This unit is to be constructed in Phase
II of the program.
444
-------
ACKNOWLEDGEMENTS
The support of the project by the Federal Water Quality Office and
the help provided by Dr. William Lacy, Mr. George Keeler, and Mr.
Allyn Richardson, the Federal Grant Project Officer, is acknowledged
with sincere thanks.
445
-------
REFERENCES
1. Kosikowski, Frank, "Greater Utilization of Whey Powder for Human
Consumption and Nutrition", (Our Industry Today), J. Dairy Science,
Vol. 50, No. 8, pp. 1343-1345, (1967).
2. Wix, P. & Woodbine, M., "The Disposal & Utilization of Whey, A review",
Dairy Sci. Abstr., Vol. 20, pp. 537-567 and 625-634, (1958).
3. Marshall, P.G., Dunkley, W.L., Lower, E., "Fractionation and Con-
centration of Whey by Reverse Osmosis", Food Technology, Vol. 22,
No. 8, pp. 37-44, (August, 1968).
4. McDonough, F.E., "Whey Concentration by Reverse Osmosis", Food
Engineering, Vol. 40, No. 3, March 1968.
5. Anonymous, "USDA Studies Reverse Osmosis as Whey Disposal Method",
New Release No. USDA 1396-68, Washington, D.C., May 1, 1968.
6. B.S. Horton, R.L. Goldsmith, S. Hossain and R.R. Zall, "Membrane
Separation Processes for the Abatement of Pollution from Cottage
Cheese Whey", presented at the Cottage Cheese and Cultured Milk
Products Symposium, University of Maryland, (March 11, 1970).
7. B.S. Horton, "Prevents Whey Pollution Recovers Profitable By-Products",
Food Engineering, Vol. 42_, No. 7, pp. 81-83 (July, 1970).
8. B.S. Horton, R.L. Goldsmith, S. Hossain and R.R. Zall, "New Method
for Economical Control of Pollution Caused by Cheese Wheys",
presented at SOS/70, Washington, D.C., (August 14, 1970).
9. R.L. Goldsmith, et al., "Recovery of Cheese Whey Proteins through
Ultrafiltration", presented at SOS/70, Washington, D.C., (August
14, 1970).
446
-------
ACTIVATED SLUDGE AND TRICKLING
FILTRATION TREATMENT OF WHEY EFFLUENTS
by
Thomas P. O^iirk* and Joseph Hellman*
INTRODUCTION
The processing and manufacturing of dairy products is one of the
most widespread industries in the United States. More than 20 million
cows produce over 100 billion pounds of milk yearly. A portion of
the production is manufactured into cheese and other dairy products.
Manufacturing operations are located In small communities in or near
the rural milk production areas. Waste waters from milk operations are
characterized by high putrescibility and high oxygen demand.
Whey from the manufacture of soft (cottage) or hard cheese contains
high concentrations of BOD (30,000 to 50,000 mg/1), and total solids
(50,000 to 72,000 mg/1), while relatively low in volume.
These concentrations, if discharged to a municipal sewerage system
without adequate control, may easily upset the treatment process
particularly if slugging occurs.
Whey rapidly exerts its oxygen demand causing septicity; lowers pH
in the primary settling tank and primary digester; produces a
substantial increase in sludge volume and can seriously reduce the
performance of biological treatment processes.
Recovery by evaporation and drying is the most satisfactory solution
to the problem of waste whey. The recovered whey solids may be
incorporated in foods, or in feeds, or may be used for the manufacture
of by-products. However, recovery is a fractional proposition that
misses about one-fifth of the production which escapes as dilute rinse
water.
When dairy processing operations are located in small communities, the
industrial waste may contribute more organic load than the entire
domestic population. Under such circumstances, domestic sewage treatment
design criteria become inapplicable for sizing facilities to treat
a combined effluent.
The Breakstone Foods Division of Kraftco maintains a large cottage
cheese manufacturing operation at Walton, New York. Considerations
*Quirk, Lawler & Matusky Engineers, New York, N. Y.
447
-------
relative to available land area indicated joint treatment of the
sewage whey mixture. A project for the development of a suitable
treatment process was undertaken.
The Village of Walton was awarded an EPA Research and Development
Grant No. 11060 DUJ, for evaluation of packed tower trickling filtration
treatment. Supplemental support to the project was provided by the
Breakstone Foods Division of Kraftco. Prior studies using the
activated sludge process were supported by Breakstone and the Village
of Walton.
Previous Studies
Although waste treatment problems associated with whey bearing waste
are frequently referenced in milk processing and waste treatment
literature, there have been relatively few published studies of formal
treatability investigations.
Wasserman(37)investigated the utilization of whey as a substrate for
yeast culture. Optimum yeast yields, of 0.57 pounds of yeast per pound
of lactose present, were obtained at pH 4.7 to 5.0, with supplementation
of nutrient nitrogen.
The results of treatability studies with activated sludge have been
reported by several authors. Jasewicz and Porges^^j. 25) attributed
the pronounced tendency of the culture to bulk to nutrient deficiencies
that included nitrogen as well as other unspecified growth factors.
Adamse^-' observed that process response was more rapid at pH 5 than
in the more neutral range from pH 6 to 8.
(21)
Adverse experience was reported by Maloney et al relative to the
introduction of whey into sewage stabilization ponds. Culture changes,
poor performance and odor evolution were attributed to the presence of
whey in the influent waste.
Significant contributions to the waste treatment literature have been
made relative to the application of biological filtration to the
treatment of whey bearing wastes. Schulze(29,30,31) employed mesh
screens for filter simulation and also observed culture growth
characteristics.
Ingram(15,16) operated a deep rock filter pilot plant on whey bearing
waste and advocated hydraulic loadings in excess of 1.0 gpm/SF to insure
complete distribution and to maintain freshness. Results were obtained
showing 67 percent BOD removal at an organic loading of 300 Ib BOD/1,000
CFD. Under the prescribed operating conditions, odor problems were not
encountered.
448
-------
WASTE CHARACTERISTICS
A synthetic waste substrate was formulated for use in all laboratory
studies. The waste substrate was prepared from Breakstone whey,
settled sewage from the city of Yonkersf N. Y., skim milk and tap
water in accord with the following formula:
Whey
Skim Milk
Settled Sewage
Tap Water
99.94%
The formula is a simulation of the waste mixture expected for design.
Representative values of the COD, BOD and suspended solids for the
trickling filter studies were 900, 650 and 100 ppm, respectively.
Corresponding values for laboratory activated sludge study were
900, 750 and 45 ppm.
Table 1 presents a summary of characteristics of each effluent as used
in pilot plant studies. Urea fertilizer was added to the waste as
nitrogen supplement.
It is useful to segregate suspended and dissolved fractions of total
BOD. The concept assists in the visualization of BOD removal by
settling facilities and in the appraisal of the BOD contributions of
effluent suspended solids losses. The results of determinations of
BOD equivalency of the feed and effluent suspended solids from treatment
of whey effluent are given in Figure 1. The BOD equivalencies for the
suspended solids were 0.63 and 0.63 Ib BOD/lb suspended solids,
respectively, for feed and effluent.
Correlations between BOD and COD were sufficiently reliable to enable
useful estimation of BOD from COD within the range of the data.
Generally the BOD was about 60 percent of the COD. The COD remaining
after total BOD removal is estimated at 50 ppm.
BOD-COD correlation relationships are shown on Figure 2.
EXPERIMENTAL FACILITIES
Experimental Scale
Sewage and industrial waste discharges are subject to pronounced hourly
variations in flow, strength and characteristics. Under such unsteady
conditions, the problem associated with the evaluation of the role of
specific variables on process performance becomes greatly magnified. The
evaluation of parameters can be effectively executed under controlled
laboratory conditions enabling isolation of the effects of specific
variables. These investigations employed laboratory techniques except
where practical considerations as to scale dictated field studies. The
evaluation of sludge dewatering characteristics was the principal area
in which scale requirements dictated field studies. A summary of
449
-------
Table 1. Characteristics of Pilot Plant Influent
Whey Effluent Whey & Sewage
Characteristics Aver. Range Aver, Ranee
Suspended Solids, ppm 240 770-41
Total BOD ppm 830 1593-440
Soluble BOD ppm 560 1075-170 241 310-172
Soluble BOD% 76
Suspended BOD ppm 173
BOD Equivalent of Suspended
Solids 0.72
Total COD ppm 1440 2390-840 451 595-307
Soluble COD ppm 841 1490-264
Soluble COD% 74
Suspended COD ppm 300
COD Equivalent of Suspended
Solids 1.25
pH 5.5-7 6-7
450
-------
FIGURE 1
451
-------
I ». WHEY &• SEWAGE i
BOD:- COD RELATIONSHIPS-
I UNTREATED1 EFFLUENT ;
""©":,'W,ITHpUTj NUTRIENTS
I ' t
! i
H—j-700
6J~WITH" NUTRIENTS-
452
-------
experimental scale Is shown on Table 2. Laboratory scale studies were
conducted at QL&M laboratories. Field scale studies were conducted
at Breakstone Foods.
Table 2. Biological Processes Investigated
Whey Plus Sewage
1. Processes Experimental Conditions
1. Fluid Bed Systems Scale Operation
A. Extended Aeration Laboratory Continuous
B. Contact Stabilization Laboratory Batch
C. Aerated Stabilization Laboratory Continuous
2. Fixed Bed Systems
A. Trickling Filtration Laboratory Continuous
Field Continuous
Fluid Bed Facilities
Extended Aeration
The flow sheet for the extended aeration process consisted of a
completely mixed aeration tank followed by a secondary clarifier.
A major fraction of the underflow from the clarifier was returned to
the aeration tank to maintain the desired culture concentration. The
clarifier was contained within a 15 gallon aeration tank. Compressed
air delivered through porous stone diffusers provided the aeration.
The unit was maintained in a controlled temperature room set for 20°C,
plus or minus one degree. Feed pumps were employed to continuously
feed waste from agitated storage tanks to the treatment unit. Nutrient
nitrogen in the form of aqueous ammonia was added to the simulated waste
feed to the extent of 1 part nitrogen per 20 parts BOD. A schematic
diagram of the pilot plant is presented in Figure 3.
Aerated Stabilization Basins
The flow sheet for aerated stabilization basins consisted of a completely
mixed aeration basin without a secondary clarifier. A long detention
time is employed to compensate for the limited concentration of culture
that can be maintained in the aeration basin. Excess culture in a
prototype installation will either settle in the aerated basin or be
discharged with the effluent. Aeration was provided by compressed air
delivered through porous stone diffusers. Six pilot plant units were
employed, ranging in size from about 3.0 to 17 liters. The units provided
detention times of from 1 to A days. Four of the units were fed simu-
lated waste without nitrogen supplement and two units were fed simulated
453
-------
waate with, nitrogen supplement to tha extent of 1 part nitrogen per
20 parts BOD (added as aqueous ammonia). Temperature control was pro-
vided by maintaining the units in a controlled temperature laboratory
at 20°C. A schematic diagram of the aerated stabilization basin process
is presented in Figure 4.
Contact Stabilization
The flow sheet for contact stabilization consists of a completely
mixed aerated contact tank followed by a secondary clarifier. The
underflow from the secondary clarifier is recycled through an aerated
stabilization tank enroute to the contact tank. Thus the process uses
the "adsorptive" capacity of the activated sludge to remove pollutants
in the contact stage and provides time for conversion of the absorbed
material in a separate stabilization stage. The flow sheet normally
permits a reduction in overall tank volume as compared with other
activated sludge flow sheets.
The laboratory experiments investigated only the contact stage of the
stabilization process since information was available from the extended
aeration studies to evaluate the stabilization stage. The experiments
consisted of the formulation of mixtures of known quantities of simulated
waste and activated sludge. The mixtures were aerated and samples were
withdrawn periodically for analysis of process performance in terms of
removal of COD and suspended solids. The operation was performed as
batch tests on a scale ranging from 0.5 to 2.0 liters. Temperature
control was provided by keeping the units in a controlled temperature
laboratory at 20°C. A schematic diagram of the contact stabilization
process is presented in Figure 5.
Sludge Dewatering
A 15 gallon activated sludge pilot plant was employed as a waste
sludge generator for use in sludge dewatering studies.
The pilot plant was maintained in a controlled 20°C environment and
operated as an extended aeration waste treatment plant. The sludge loading
was maintained at about 0.05 Ib of BOD/lb of sludge. The concentration
of suspended solids in the mixed liquor was relatively stationary at
about 2,000 ppm. The aforementioned conditions effected a settleable
culture with a typical sludge Volume Index of 145 ml/g. Samples of
sludge were withdrawn from the pilot plant reactor as required for sludge
dewatering studies.
Dissolved Air Flotation
Sludge dewatering studies using dissolved air flotation were conducted
with a pressure bomb. The unit had a liquid capacity of 1 liter and was
operated at a pressure of 70 psig. Variables evaluated were recycle,
rise time, float volume and effluent suspended solids.
454
-------
FEED
TANX
PUMP
EXTENDED AERATION PROCESS
AIR
AERATION
RETURN SLUDGE
SETTLING
WASTE
SLUDGE
EFFLUENT
o
c:
3)
m
OJ
-------
AERATED STABILIZATION PROCESS
AIR
FEED
TANK
»•/
AERATION
EFFLUENT
PUMP
33
m
-------
CONTACT STABILIZATION PROCESS
FEED
TANK
Q-
= PUMP
AIR
El
AERATION
RETURN
SLUDGE
SETTLING
STABILIZATION
TANK
WASTE
EFFLUENT
01
-------
Bulked sludge was employed for a series of flotation experiments.
The activated sludge was bulked by increasing the concentration and
rate of feed to the unit. The bulking of the sludge caused the concen-
tration of suspended solids in the mixed liquor to decrease from the
unbulked level of about 2,000 ppm to a bulked level of 450 ppm. In all
experiments the suspended solids concentration in the float was
determined gravimetrically so as to eliminate interference due to
entrapped air.
C entrifugat ion
Studies were made using a 10 ml laboratory spin test on sludge
samples with and without polymer conditioning. The polymers employed
for sludge conditioning were Dow A22, N17 and C31. Observations were
made of the volume of the mud layer, the centrate turbidity and the
dosage of conditioner.
Filtration
Samples of sludge, with and without chemical conditioning, were filtered
on a Buchner funnel apparatus to determine the specific resistance to
filtration. Conditioning agents were ferric chloride, ferric chloride
and lime, and Dow C-31 polymer. The scale of the experimentation was
260 ml. Variables observed were time of filtration, volume of filtrate,
vacuum level, sludge solids concentration before and after filtration,
and dosage of conditioning agent.
Purifax Treatment
A sample of sludge was submitted to Purifax Incorporated, for treatment
by the "Purifax Process" (chlorination ) The dewaterability of the
treated product was evaluated from Bucl.ner funnel filtration tests and
by sand bed characteristics.
Laboratory Scale Trickling Filters
Waste was fed to four laboratory test stand units to simulate
trickling filter performance kinetics. The units, illustrated in figures
6 and 7, were rectanglular plane surfaces set at various inclines.
Filter lengths were side channeled to provide a controlled surface area
of slime. Units were constructed to provide flexibility in waste
loading variation and system operation. Feed volumes were stored under
refrigeration during feeding and collection. Slime surfaces were
sufficiently large to enable the investigation of a wide range of BOD
removal efficiencies.
Plane recirculation ratios and BOD and hydraulic loadings were scaled
to approximate prototype conditions. Application of feed and recycle
458
-------
TYPICAL LABORATORY SIMULATED
TRICKLING FILTER PLANE
SIDE VIEW
REFRIGERATOR
PUMP
FRONT VIEW
CHEATING
TAPE FOR
TEMPERATURE
CONTROL
LU
_J
O
LsJ
FEED
MIXED LIQUOR SUMP
WIDTH =j/2
THERMAL
REGULATOR
•EFFLUENT
MAGNETIC STIRRER
EFFLUENT
TO REFRIGERATOR
-------
I
FILTER PLANE TE^T STAND UNITS
-------
was continuous with automatic temperature control. Operating
characteristics are presented on Table 3. A period of 24 hours
was allowed for development of steady state conditions between runs.
Table 3. Laboratory Trickling Filter Plane
Operating Characteristics
Laboratory
Characteristics Trickling Filter
1. Number of Units Run 4
2. Total length (Feet) 9 and 18
3. Available Media (SF) .375, .75 and
4. Volume of Feed Required
per day (Liters) 3-22
5. Temperature Automatic Control + 2°C
6. Sampling Procedure Manual Grab Sample's
7. Duration of Runs 3 to 5 days
8. Operation Schedule 7 days/week
9. Method of Operation
A. Feed Continuous
B. Recycle Continuous
Odor Control
The trickling filter planes were enclosed to deodorize the effluent
air stream. A minimum air flow of 40 ft^/gal was provided. Total
air flow over the four trickling plans averaged 40 ft^/hour. A • Welsbach
Model T-816 laboratory ozonator was chosen as the ozone source. Its
rated capacity of 8 gms ozone/hr using air as the feed gas insured
adequate capacity over and above the 10 ppm dosage recommended by the
manufacturer for deodorizing sewage sludges. Control of ozonation
was accomplished using a Welsbach Model H-81 ozone meter. A minimum
reactor detention time of 5 minutes was provided before exhausting the
deodorized air effluent outside the laboratory.
Solids-Liquid Separation
Several tests were performed on the laboratory trickling filter effluents
to determine sedimentation characteristics. Effluent samples were
settled for various detention times in a standard 500 ml polyethylene
cylinder. Supernatent samples were withdrawn at the 150 ml level.
On-Site Pilot Plant
A pilot plant was used to treat whey effluent and whey effluent com-
bined with settled sewage. The principal function of the on-site
461
-------
pilot plant was the generation of secondary sludge on a scale sufficient
to enable the development of practical dewatering and disposal processes.
The pilot plant flow sheet consisted of a primary settling tank followed
by trickling filtration and batch settling of filter effluent as shown
on Figure 8.
The filter unit was supplied by the Koch Engineering Company and had a
cross-sectional area of 7 sq ft with a media depth of 18 to 20 ft. The
packing media was Koch Flexirings, a 3.5" plastic, webbed cylinder with
a specific surface of 28 sq ft/cu ft.
The primary and final settling tanks were not evaluated for treatment
performance on a continuous operating basis.
The majority of trickling filter samples were continuous composites
for periods of typically 3 to 5 hours. Grab samples were taken for
shorter filter runs.
The samples were analyzed at the QL&M laboratory on the day following
their collection. Preservation was accomplished by acidification and
refrigeration during transit and prior to analysis.
Sludge samples were obtained by collection of effluent into two 350
gallon settling tanks operated on a batch basis. After filling, the
tanks were mixed and allowed to settle quiescently for one-hour. At
the end of the settling period, the supernatent liquid was pumped to
waste and the sludge was collected for shipment. Refrigerated
samples of sludge were transported to the QL&M or other laboratories
for analysis and evaluation.
Sludge Centrifugation
A centrifuge test of the trickling filter sludge was conducted by
Sharpies division of the Pennwalt Corporation.
A 100 gallon sample of a 0.74 percent sludge was used for the
centrifuge test. Jar tests, for chemical aids, were conducted for a
qualitative evaluation of coagulation. Nalco 610 appeared to be the
most effective and was selected for evaluation during the centrifuge
runs.
Two types of centrifuges were evaluated. The first was the Super-D-Canter
P-600, which is a continuously fed, 6 inch, horizontal bowl centrifuge.
The unit was evaluated at various feed rates without the addition of
the chemical aid, and at a selected feed rate with various amounts of
chemical aid.
The second centrifuge was a Fletcher Model 2PP-200, which is an automatic
cylically fed, vertical bowl centrifuge. This unit was evaluated at
462
-------
0-.
LJ
PACKED TO'A/ER BASE a SEWAGE STORAGE
PACKED TOWER, FULL LENGTH
PRIMARY SETTLING TANK (LEFT.
TWO BATCH SECONDARY SETTLING TANKS (RIGHT)
TRICKLING FILTER PILOT PLANT
WALTON, NEW YORK
QUIRK, LAWLER a MATUSKY ENGINEERS
o
!*)
GO
-------
various feed rates without the addition of the chemical aid. The
effect of chemical aid was estimated by assuming the same improvement
as that which occurred with the Super-D-Canter. Samples of the
centra te liquid and sludge cake were measured for suspended and total
solids, respectively.
Figure 9 illustrates the equipment employed.
Vacuum Filtration
Sludge specific resistance measurements were used to select a chemical
conditioner and its optimum dosage. The three conditioners evaluated
were: ferric chloride, a Dow Chemical Company polymer Purifloc C-31,
and lime addition in conjunction with ferric chloride.
Vacuum filter solids loading rates were determined by performing a
series of filter leaf tests on a selected sludge sample.
PERFORMANCE OF FLUID BED BIOLOGICAL TREATMENT SYSTEMS
Extended Aeration BOD Removal
BOD removal performance was found to be excellent with capability of
achieving removals in excess of 95 percent on a consistant basis.
BOD removal rate kinetics were considerably higher than domestic sewage.
Mixed liquor sludge settling characteristics were demonstrated to
control process operation rather than BOD removal requirements.
Contact Stabilization BOD Removal
A good settling sludge from the extended aeration reactor was employed
in the experiments, the results of which are summarized in Table 4.
BOD removals in excess of 95 percent were demonstrated. The high
effluent suspended solids concentrations indicated for the process
discouraged additional investigation in a limited screening study.
However, it is conoivable that with more intensive study an applicable
process could be developed.
Biological Sludge Production
A material balance was constructed to evaluate the net sludge production
resulting from BOD metabolism and endogenous respiration as follows :
Us = AS = YSUL -
SATD
where: us = Velocity of net sludge production
Ib sludge prdduced/lb sludge under aeration/day
464
-------
c
01
. -Gt LAKE
SUPER-D-CANTER
FLETCHER
QUIRK, lAWLER a MATUSKY ENGINEERS
-------
YS
Velocity of BOD removal - Ib BOD removed/lb
sludge under aeration/day
Velocity of endogenous metabolism of sludge -
Ib sludge metabolised/lb sludge under aeration/
day.
Sludge yield from BOD metabolism -
Ib sludge/lb BOD removed
SA = Aeration solids concentration - pptn
TD = Aeration detention time - days
AS = Net biological sludge production = ppm
The data obtained relative to sludge yield were split into categories
of "with" and "without" nutrient supplement and analyzed statistically
by a polyfit computer program. Results are presented in Table 5.
Table 4. Contact Stabilization Performance
Culture
Cone,
SA
£2H
1650
3310
4950
6600
COD of Raw
COD of Raw
Filtered Effluent
COD j>m
589
249
25
25
= 1010 ppm
= 982 ppm
= 98 ppm
Settled Effluent
(3 60 min
Contact
COD
ppm
974
666
278
278
Time
SE*
ppm
482
484
217
235
*SE
Effluent suspended solids concentration
466
-------
Table 5. Biological Sludge Production
Average Parameters
Whey & Sewage
With Without
Parameter Nutrients Nutrients
1. YS - Ib Sludge/lb BOD Removed 0.400 0.700
2. UE - Ib Sludge/lb Sludge/day 0.055 0.065
The relationships obtained are employed for the estimation of the net
biological sludge production under various operating conditions maintained
in the aeration phase. The net production results from the particular
balance between sludge production from BOD metabolism (Ys • UL) ancj
sludge destruction from endogenous respiration (UE) .
Oxygen Consumption
The total oxygen consumption attributable to BOD removal and biological
culture requirements can be estimated from a material balance relation-
ship as follows:
U =
Where: UQ = Velocity of net oxygen consumption -
Ib ^2/lb sludge under aeration/day
UL = Velocity of BOD removal - Ib/lb/day
Ufi • Velocity of endogenous oxygen consumption -
I1?02/lb sludge under aeration/day
OM = Unit oxygen requirement for BOD metabolism
The relationship is employed to determine the total oxygen consumption
under various operating conditions maintained in the aeration phase.
The total oxygen demand results from the particular balance between
metabolism requirements (°M • UL) and endogenous respiration require-
ments
The data collected for oxygen consumption were split into categories
of "with" and "without" nutrient supplement prior to statistical
processing by a polyfit computer program. Results are presented in
Table 6.
467
-------
Table 6. Oxygen Consumption
Average Parameters
Whey & Sewage Bio-Oxidation
With Without
Parameter Nutrients Nutrients
1 °M - lb°2/lb BOD Removed 0.350 0.480
2. Hi - lb°2/lb sludge/day 0.075 0.065
Sludge Settleability
Factors that affect the settleability of the culture include waste
characteristics, temperature, BOD removal velocity, nutrient supple-
mentation and temperature. In the laboratory treatability studies
observations were made of the effects of BOD removal velocity and
nutrient supplementation on the settleability of the culture. Figure
10 gives the results obtained from determination of the settleability
of the cultures prevailing in the seven reactors on a given day.
The settleability is indicated by a Sludge Index which expresses the
wet volume in milliliters, occupied by a gram of dry sludge solids
after settling for a period of 30 minutes. Cultures were temperamental
in respect to settling characteristics. The Sludge Index fluctuated
in a random manner over a range of values of the BOD removal velocity.
The only acceptable value of the Sludge Index was 108 ml/gram obtained
with nutrient supplementation at a BOD removal velocity of 0,087/day.
The other cultures were indicated to be poor settling and difficult
to dewater. The results obtained throughout the studies indicated
that when the BOD removal velocity reached a level of 0.2/day, the
culture developed poor settling characteristics.
Some results obtained relative to the effects of BOD removal velocity
and nutrient supplementation on effluent suspended solids concentration
are given in Figure 11. The general trend of the results suggested that
the loweat effluent suspended solids concentrations were obtained with
nutrient supplementation and with low values of the BOD removal velocity.
Flotation
Laboratory flotation experiments were run with samples of activated
sludge containing 2,000 to 3,200 ppm of suspended solids. Effluent
suspended solids range from negligible concentrations to 25 ppm.
The rise time of the float is indicative of the velocity of separation
of the solid phase. As such, it is the parameter governing the maximum
hydraulic loading on the prototype unit. Empirical experience indicates
468
-------
2000
c
'E
O '600
o
E
o»
^ 1200
1
X*
0 800
z
UJ
CD
Q
13 400
CO
0
(
EFFECT OF BOD REMOVAL VELOCITY ON
SLUDGE VOLUME INDEX
c/
x
/
/
/
^
l
i /
~j{
/
/
/
X.
A
s \
•^
^^
\
\
\
\
O WITH NUTRIENT SUPPLEMENTATION
X WITHOUT NUTRIENT SUPPLEMENTATION
\
\
\
o
D O.I 0.2 0.3 0.4 0.5 0.6 0.7
UL , 1 / DAYS
o
c
•33
m
-------
EFFECT OF BOD REMOVAL VELOCITY ON
SETTLED EFFLUENT SUSPENDED SOLIDS CONCENTRATION
SETTLED EFFLUENT SUSPENDED SOLIDS,
PPM
_ N OJ *
o 8 8 8 g
O WITH NUTRIENT SUPPLEMENTATION
X WITHOUT NUTRIENT SUPPLEMENTATION
O
O-
O
X
X
X
f
/
x/
/
/
/ X
/
X
y
/
/
/
r
/
/
X.
/
/
/
/ ~
/ "
/
/
O
o
/
o
^
o—
3 0.1 0.2 0.3 0.4 0.5 Q6 Q7
UL , I/ DAYS
FIGURE II
-------
that rise times of less than 40 seconds in the experimental unit are
desirable in order to maintain performance stability in the prototype.
Although long rise times may yield floats of improved concentration,
there is a risk that release of attached bubbles may occur with sub-
sequent redispersion of suspended matter. The effect of recycle ratio
on the rise time for the two concentrations of activated sludge is
illustrated in Figure 12. For design purposes, a recycle ratio of 2.0
would appear to be sound for the 2,000 ppm sludge. For design purposes,
a gravimetric float solids value of 1.7 percent is selected as repre-
sentative of prototype performance.
Centrifugation
Centrifuge spin tests were made on samples of activated sludge with
a suspended solids concentration of 4,000 ppm. Poor results, indicated
by cloudy centrate and indistinct separation were obtained on the
sludge alone or when the sludge was conditioned with 100 ppm of Dow
Polymers A-22 and N-17.
A clear centrate and a distinct solids separation was obtained with
sludge conditioned with 50 ppm of Dow C-31. The mud layer amounted
to 1.8 ml after 30 seconds. The indicated solids concentration of the
mud layer was 2.2 percent. Centrifugation of a sludge sample conditioned
with 9 ppm Dow C-31 produced a mud layer of 2.4 mis in 30 seconds and
a slightly cloudy centrate. The indicated solids concentration of the
mud layer was 1.67 percent.
Polymer dosage would be about 5 Ib/ton of solids to yield a mud phase
of 2.0 percent solids.
Vacuum Filtration
Laboratory experiments were made to determine the specific resistance
to filtration of samples of activated sludge conditioned with ferric
chloride, ferric chloride plus lime and a cationic polymer. The
results indicated that ferric chloride alone was more effective than
ferric chloride plus lime.
The practical dosage of ferric chloride for sludge conditioning was
about 7 percent of the sludge dry solids. Solids in the dewatered
product was only 11 percent. The conditioning procedure effected a
3.3 fold increase in filterability.
Results were obtained for sludge conditioned with cationic polymer
(Dow C-31). At a practical conditioning dosage of about 0.5 percent
of the sludge dry solids, the polymer effected a 2.3 fold increase in
filterability. The solids in the dewatered product was 10 percent.
471
-------
-.0 X. 10 TO v'i (MCM 4.6 1
KFU^P'-L jv t*er«? TO
pi;-, .^.ii-.TV..:: trr"
' --:-•-;-• : re.
FLOTATION _Q
-^-:r SLUDGE RISE TIME
—;——;—-fi5
;:_,., .> .; * _^ _..7TJ_
L.QL.-^3200_. PPM snrms
SE::EES^EHE
+ •*—- •- - ~-"~~"~''~~~ " "' " "—-~~-
-------
Based on the experimental results, it is calculated that the vacuum
filter yields for sludge conditioned with ferric chloride and with
C-31 Polymer would be 1.5 and 1.25 pounds dry solids/sq ft/hr respect-
ively at solids concentrations of 10 percent.
Purifax Process
A sample of activated sludge containing 4,500 ppm suspended solids was
transmitted to the Purifax Corporation for processing. The chlorine
demand of the sample was determined to be 78 ppm. The sample was
"Purifaxed" with approximately 98 ppm of chlorine with the result that
within a few minutes the solids floated and were removable by skimming.
The pH after processing was less than 3.0. The residual solids were
rapidly dewaterable on an underdrained sand bed or on a Buchner funnel
filter. The dewatered cake could be lifted from the sand bed after
20 minutes. No decomposition of the Purifaxed sludge liquor or the
dewatered cake was evident on standing for a month. Neutralization of
the Purifaxed sludge liquor with lime prior to filtration adversely
affected dewatering properties.
After degassification, the Purifaxed sludge settled readily to a sludge
concentration of 2.0 percent solids in 30 minutes. Untreated sludge
would settle typically to a concentration of about 0.59 percent solids
in 30 minutes. The improved settling characteristics represented a
70 percent reduction in the volume of sludge to be handled.
The specific resistance to filtration of the Purifaxed sludge was
determined. By calculation, it was estimated that the material would
dewater on a vacuum filter to 9 percent solids at a rate of 0.6 pounds
dry solids/ sq ft/hr.
TRICKLING "FILTRATION PERFORMANCE
Laboratory Simulation
The classic trickling filter comprises a bed of stone media over which
attached biological slime growths develop. Removal of BOD is obtained
by aerobic processes at the slime surface and by anaerobic processes
within the slime interior. Packed tower modifications to the trickling
filter introduced plastic geometric packing media to obtain increased
surface area and porosity. Current practice employes a lattice structure
similar to an egg carton insert. BOD removal performance is related to
process parameters using an analytical model descriptive of the biological
relation observed and the hydraulics of the reactors.
BOD changes (at constant temperature) in flow over an inclined plane or
through a full scale packed tower have been shown to follow a first
order reaction and to be related as follows:
473
-------
Inclined plane: Le = e~k'H/U ........................ (3)
^o
Packed Tower (Surfpac or equal) :
Where: H = Tower or plane height in feet
U = Liquid application rate expressed as:
gpm/LF of plane width for laboratory
units and gpm/SF of tower area for
full-scale units.
i
k = Specific BOD removal rate constant for effluent
treated obtained from laboratory analysis.
KT = BOD removal rate constant for particular
packing media used in full scale tower.
i
The value of k is specific for the effluent treated. The value of
KT varies with the type of full scale packing media used and is computed
using the value of k and the characteristics of the media.
A graphical solution to equation (3) is obtained by taking logarithms.
Temperature effects are adjusted to 20°C by applying the Arrhenius
correction ( g&D to the height variable.
A plot of data on semi-log paper will provide a linear correlation with
slope equal to (k'/2.3) and an intercept of 1.0 at H6^/u' = 0 as
shown on Figure 13. The graphical technique is employed in the analysis
of test data.
Figure 14 is a presentation of biological filter plane results. The
linearity of the plot over the wide range of operating conditions,
verifies the applicability of the kinetic formulation. For the
Conditions represented by the plot, the BOD removal rate coefficient,
k*20, was computed as 1.6 x 10~3 gpm/ft2. The literature was reviewed
for whey data to which the correlation could be applied. The data of
Schultze^31) were analyzed for whey application over a vertical surface
comprised of a screen mesh. The correlation is shown on Figure 14 and
supports a (k') value of 1.6x10-3 gpm/ft2.
Laboratory reaction rates are scaled to full scale conditions for a
media similar to Surfpac using the following relationship (™) :
(KT) = k' . Av . ft • Cw ............................ (5)
474
-------
FIGURE
LU
cc
O
CC
u
O
O
*—t
u
u.
"•v,
rH
U-
o
O
GRAPHICAL SOLUTION OF SLIMED PLANE
BOD REMOVAL PERFORMANCE EQUATION
FIRST ORDER REACTION
I/HYDRAULIC LOADING - HOAT/U' - SF/gpm
475
-------
HGURE
TRfCKLING FILTER 'LANE PERFORMANE
i ; I
WHEY .«! SEWAGE =
WITH NUTRIENTS 3
TEMPERATURE -
BOD REMOVAL -
'SAMPLING-6RAB
1,6 x 10 T GPM/SF
REMOVAL - 25 TO! 80% i
SAMPLING-COMPOSITE
=1,6 X 1C
1000
1200
1400
I/HYDRAULIC LOADING -
SF/GPM
476
-------
Where: ^'20 = BOD removal rate constant
1.6 x 1Q~4 gpm/SF for whey sewage
AV = Specific surface of media
27 SF/CF for Surf pac ;
^t = Factor for slime area reduction due to
slime thickeners
0.8 for whey and sewage
GW = Factor for efficiency of wetting of full
scale media 0.90 for Surfpac
and % = 1.6 x 10~4 x 27 x 0.8 x 0.9
=0.03 gpm/CF
A comparison between the treatability of whey and other industrial
effluents using full-scale packed towers and a media similar to Surf-
pac demonstrates the treatability of whey effluent as follows:
Trickling Filtration of ^
Comparison of Full Scale BOD Removal Rate Constants
Effluent KT- gpm/CF
1. Ragmill .............................. .083
2 . Slaughter House ...................... . 044
3 . Integrated Kraft Mill (Average) ...... .031
4. Whey ................................. .030
5. Boxboard Mill ........................ .027
6 . Canning .............................. .021
Effect of pH
The level of pH influences enzyme activity and culture characteristics
in biological systems. The results of a limited number of data
collection runs at different levels of influent pH indicated a trend
towards improved performance in the pH 7 or below. It is noted that
similar trends were obtained by Wasserman^-'-' and Adamse(4) with other
btologLcal systems using whey as substrate. Data indicate an increase
in rate constant from 1.1 @ pH = 9.8 to 2.8 @ pH = 4.5
Effect of Temperature
Temperature effects are introduced by adjustment of the value of the
BOD removal coefficient, k' in the following manner:
477
-------
kt = k20 Q(t-20) (5)
H
where: kt = BOD removal rate coefficient at T°C
BOD removal rate coefficient at 20°C
6 = Constant
t = Waste temperature, C°
A series of experiments made over the temperature range from 19 to 30°C
is shown on Figures 15 and 16. The value of k increased with temper-
ature from 1.6 x 10~3 gpm/ft2 at 20°C to 2.2 x 10" 3 gpm/ft2 at 29°C. The
average value of 9 was computed as 1.032, as shown on Figure 17, and
was in agreement with the value assumed in current practice.
Effect of Nutrients
A series of experiments were made to determine the effect of nutrient
supplementation upon BOD removal performance. The nutrients selected
for investigation were combinations of ferrous iron and ammonia. Data
analysis indicated that the addition of 1.5 ppm of ferrous iron had
no effect on BOD removal, whereas the addition of 37.5 ppm of ammonia-
nitrogen effected a 40 percent increase in the rate coefficient for
BOD removal. Comparisons of trickling filter plane performance with
and without nutrient addition are shown on Figure 18.
Odor Generation
During the experimental work, odor was detected only at the higher
BOD loadings, above .036 Ib BOD/day-ft2 of slime area. Application of
ozone at an approximate dosage of 10 ppm effectively deodorized the
effluent air stream. No evidence of odor was detected either in the
laboratory or at the air effluent exhaust.
Pilot Plant Performance
Erratic performance was attributed in part to plugging of the media
due to the lack of free-fall and to the extensive slime growth effected
at high loadings. Poor performance, black sludge and poor draft in
smoke tests were evidence that the pilot plant was occasionally effected
by anaerobic conditions. Substantial scatter of performance was evident
but a trend for better than 50 percent removal was exhibited for BOD
loadings of less than 150 ptcfd.
Sludge Production
Biological sludge production data was collected for a period of four
days. The average biological solids production approximated 1.0 Ib of
dry solids/lb of BOD removed. During the period of observation the BOD
removal of the trickling filter averaged 75 percent. Generally, low
yields are associated with high degrees of BOD removal (85-95 percent).
478
-------
FIGURE 15
TRICKLING FILTER PLANE PERFORMANCE
WHEY & SEWAGE, WITH NUTRIENTS
EFFECT OF TEMPERATURE
.7 x 10-3 GPM/SF
200
400
600 800
H/U' SF/GPf
1000
1200
479
-------
FIGURE 16
TRICKLING FILTER PLANE PERFORMANCE
WHLY & SEWAGE, WITH NUTRIENTS
EFFECT OF TEMPERATURE
k29°C = 2.2 x 10-3 GPM/SF : I.h
1200
H/U' SF/GPM
480
-------
FIGURE 17
TRICKLING FILTER PLANE PERFORMANCE
WHEY & SEWAGE WITH NUTRIENTS
EFFECT OF ThMPEKATURE ON BOD RATE CONSTANT
3,0
, i ; I' : < !
Ml!; if;
1 ;kb.latibjiBhiLp_:
j-flog;e
dlilM
1,0
10
12
AT-°C
481
-------
TRICKLING FILTER PLANE PERFORMANCE
WHEY & SEWAGE
DISSOLVED BOD REMOVAL ,,, .
OPERATION WITH NUTRIENTS a F»U = 7,0
FIGURE 18
1-3
>
TRICKLING FILTER PLANE PERFORMANCE
WHEY & SEWAGE
OPERATION WITHOUT NUTRIENTS a PH 7,0
REACTION RATE
i ; i
k'20 = 1.15 x 10~3 GPM/SF
0)
>
0
200
400
600 800
'= SF/GPM
1000
1200
482
-------
Lower degrees of removal normally generate higher sludge yields The
field studies were not extensive enough for the development of a
sludge production model. A value of 0.7 Ib of SS/Xb of BOD removed
is recommended for biological sludge production for the blend of whey
waste and sewage. This value equals the biological yield coefficient
for activated sludge treatment.
Solids Liquid Separation & Compaction
Figure 19 illustrates the supernatant suspended solids after varying
detention times for a given trickling filter plane operating condition
with differing effluent suspended solids.
Projecting the data, a supernatant suspended solids concentration of
between 75 to 100 mg/1 would be expected from gravity sedimentation.
Measurement of pilot plant filter effluent clarification was made,
routinely, by pumping the trickling filter effluent into the holding
tanks. When full, the tank was sequentially mixed, sampled, allowed
to settle for 30 minutes, and resampled near the surface. The suspendad
solids removal averaged 80 percent. This should be considered a
limiting value, i.e., attainable at low overflow rates, unless chemicals
are used to aid settling.
Suspended solids present in the untreated whey and in the trickling
filter effluent have been shown to exhibit a BOD equivalent to 60 percent
of their weight. The achievement of high overall degrees of treatment
will require that suspended solids in the final effluent be limited to
minimum values.
Settling characteristics of trickling filter solids are such as to
require coagulant addition in order to insure minimum suspended BOD in
the treated effluent.
Clarifier underflow is expected to be approximately 1 percent solids
and to thicken to 2 percent at a sludge loading of 6.0 Ib/SF/Day.
Centrifuge Results
Table 7 presents results for all the centrifuge runs. The addition
of 9.3 Ib of chemical/ton of dry solids at a feed rate of about 19 Ib/tnin
raised the percent recovery of the Super-D-Canter to the level of the
Fletcher model. The cake solids concentration of the Super-D-Canter
and the Fletcher were respectively 6 percent and 9 percent.
Both models appear to give reasonable feed rates for acceptable
recoveries of 80 to 85 percent. However, both models' cake solids
concentrations are well below the desired values.
483
-------
FIGURE 10
•r-r-T tTi—TT I"' i TT'T j r'-'-r* l^r-
484
-------
TABLE 7
TRICKLING FILTRATION TREATMENT
WHEY & SEWAGE
CENTRIFUGATION OF WASTE SLUDGE
Slurry
Feed Rate
Ibs/min.
2.1
6.5
9.5
23.5
6.8
9.9
18.9
23.7
18.3
19.2
19.3
Chemical Aid
Feed Rate
Lbs/Ton of
Dry Solids
-
_
-
-
-
-
-
-
9.3
19.6
33.0
Super-D-Canter Fletcher
P-600 2PP-200
% Recovery Cake Solids , % % Recovery Cake Solids , %
99.2 Average
99.4 of
9%
93.7
77.2
98.9 5.5
98.7 6.3
59.0
59.8 5.7
81.7 5.9
82.1
86.7 6.5
-------
Vacuum "Filtration
Ferric chloride and Purlfloc C-31 were compared using dosages of
comparable cost. The results favored the use of ferric chloride
rather than the polymer. Various dosages of lime in conjunction with
a constant dosage of ferric chloride were evaluated. Specific
resistance was found to increase as lime was added to the samples.
For this reason, lime was eliminated from consideration and ferric
chloride was selected as the conditioner for further tests and
final design.
The determination of the optimum dosage of ferric chloride was made
by measuring specific resistance versus dosage on three representative
samples. Seven pounds of ferric chloride/ 100 Ib of suspended solids
was selected as the optimum dosage.
Figure 20 presents a summary of test conditions and results.
Solids dewatering by vacuum filtration could be achieved at an indicated
loading of 1.2 Ib/SF/hr using 7 percent FeCl3 to produce a cake of
20-25 percent solids.
Odor generation from the acidified conditioned sludge will require that
special ventilation facilities be incorporated in the design of sludge
filtration facilities.
TRICKLING FILTRATION & ACTIVATED SLUDGE COMPARISONS
Process comparisons between packed tower trickling filtration and
activated sludge are summarized as follows:
1. In order to maintain an SVI under 200, an organic loading of less
than 0.1 Ib BOD/lb Sludge/day is indicated for activated sludge.
2. Bench scale pilot plant operation indicated a sludge volume index
of 145 when operating at a loading of 0.05 Ib BOD/lb sludge/day.
These low organic loading requirements may be compared with a
nominal value of 0.25 Ib BOD/lb sludge/day for municipal sewage
operation at high degrees of treatment.
3. At an organic loading of 0.1 Ib/lb/day, and using a design mixed
liquor solids concentration of 2500 ppm, an aeration detention time
of at least 10 hrs per 100 ppm of BOD is required for activated
sludge operation at a SVI of 200 or less.
4. At an organic loading of 0.1 Ib/lb/day, and a 2500 ppm mixed liquor,
activated sludge aeration volume would be over 4-1/2 times the
volume of a staged packed tower facility.
486
-------
FIGURE 20
n' ill i ; ! n i
TRI;CKtIflG fldTER TR
| I WHEYS'SEWAGE^ ;
KLUDGE FtiMidN CHARACTERISTlies
487
-------
5. Activated sludge settling characteristics were found to be
periodically unstable at SVI values below 200.
6. Flotation of activated sludge mixed liquor, using bench scale
equipment, indicated a design overflow rate of 600 gals/day/
SF based on untreated waste flow. Flotation would not be
economically competitive in comparison with sedimentation.
7. Activated sludge operation was observed to exhibit significant
sensitivity to pH, nutrient and temperature control.
8. Waste sludge solids from the activated sludge process were not
amenable to dewatering by vacuum filters or centrifuge.
PROCESS DESIGN FOR TRICKLING FILTRATION
Design loading for combined whey and domestic sewage are presented in
Table 8. Whey contributions to the combined loadings are summarized
as follows:
Flow 85%
BOD 97%
Solids 80%
Table 8. Treatment of Whey & Sewage
Combined Loadings for Design
Characteristic
1. Flow - mgd 1.17
- gpm 815
2. Suspended Solids - ppm
Total 162
Volatile* 129
Inert 33
3. BOD - ppm
Total 705
Suspended** 102
Dissolved
*Estimated from laboratory analysis of synthetic waste.
**Based on 0.63 Ib BOD/lb SS measured in laboratory.
488
-------
Suspended solids and BOD concentrations required in the treated
effluent are shown on Table 9 together with the nominal percent
removals required to achieve effluent limitations. Because of the
high proportion of raw waste BOD in the dissolved form and the equally
significant fraction of effluent BOD contributed by suspended solids
discharged, the requirements for removal of dissolved BOD in the
packed tower exceeds that of total BOD removal as measured between
influent and effluent concentrations.
Process design procedures and design alternatives are presented for
full scale packed tower trickling filtration below.
Filter Volume & Geometry
For a given type of packing, tower volume will vary with the following
design parameters:
1. Liquid Application Rate
2. Recycle
3. Tower Height
4. Efficiency of BOD removal
The effects of variations in the first three parameters is dependent
upon the necessity to maintain a minimum wetting rate. In all cases,
an increase in efficiency of removal requires an increase in tower
volume. In general, the effects of design parameters can be described
as follows:
Design Variable Change in Variables Change in Tower Volume
Height - H Increase Decrease or no change
Recirculation ratio - r Increase Increase
Application velocity - U Increase Increase
Efficiency of Removal - E Increase Increase
Structural requirements and hydraulic distribution problems limit
maximum tower height. Heights of 20 feet are common with maximums to
45 feet.
Commercial packing of the lattice structure type appears to require
itdnimun application velocities of 1.0 gpm/sq ft. Operation below the
minimum velocity can result in progressively less utilization of tower
packing.
In order to maintain commonly used heights and provide a minimum appli-
cation velocity, effluent recycle is usually required for high BOD
489
-------
removal efficiencies. The added tower volume required to accommodate
recycle varies with the efficiency of removal sought.
Because of the non-uniform influences of design variables, process
design calculations involve relatively complex manipulations.
Table 9. Treatment of Whey & Sewage
BOD & Solids Removals Required
Characteristic
I. Effluent Required
1. Suspended Solids - ppm 45
2. BOD - ppm
Total 60
Suspended 28*
Dissolved 32
*Based on 0.63 Ib BOD/lb SS as measured in laboratory
II. % Nominal Removals
1. Suspended Solids
Influent - ppm 168
Effluent - ppm 45
Removal - ppm 123
Removal - % 73.2
2. Total BOD
Influent - ppm 730
Effluent - ppm 60
Removal - ppm 670
Removal - % 91.8
3. Dissolved BOD
Influent - ppm 623
Effluent - ppm 32
Removal - ppm 591
Removal - % 94.9
490
-------
Tower Volume
Equation 4 is rearranged to determine the wetting application rate
(U) for tower operation using Surfpac of similar media without
recirculation. Application velocity without use of recirculation is
given the symbol uo and is expressed as follows:
U = (KT) (H) (6)
(2.3) log (1-E)
A. maximum tower height of 42' is selected to minimize potential
recirculation requirements and application rates, (UQ) , are examined
for parallel and series operation of filters. Series operation will
reduce total tower volume requirements but will require additional
pumping.
Application rates and stage removal efficiencies for single and multi
stage designs are tabulated below:
Single Two Three
Stage Stage Stage
1. Efficiency/stage - E% 95 77.5 65
2. Application Rate without
recirculation - UQ gpm/SF 0.42 0.85 1.23
Recirculation would be required on the single and two stage plants.
Recirculation ratios are determined using a material balance for BOD
taken across the trickling filter as follows:
r = (1) - (f)
QTE) (7)
(f - 1)
The value of (f) is determined by substitution into equation (4)
using a minimum application velocity of U = 1.0 gpm/SF.
A tabulation of recirculation requirements versus stage design is then
shown below:
Single Two Three
Stage Stage Stage
1. Efficiency/Stage-E% 95 77.5 65
2. Application Rate - gpm/SF 1.0 1.0 1.23
3. Recycle Ratio Required-r 6.6 0.4 0
491
-------
A modification of the basic formulation is used to determine tower
volume requirements. The modification is achieved by expressing
tower volume requirements as a volume per unit of untreated flow (V)
and relating this unit volume to process variables.
A material balance for flow is used to develop a definition for unit
volume (V1) as shown below:
By flow balance:
(1+r) Q •
and
V' = V = (1+r) (H) (8)
Q (U)
Unit volume requirements when recirculation is not required are
designated by the. subscript, o, i.e. VA, and are determined from
equation(S) by setting r = 0 and substituting into equation^) the
value of application rate (U) from equation'"' which results in the
following design equation:
vo - 2.3 log (1/1-E) (9)
KT
where: Vo = CF of tower volume/gpm of untreated
effluent when tower is not recirculated.
Unit volume requirements for a recirculation tower must be increased
over that for a non-recirculated design because of the plug flow
hydraulics.of flow through the Tower. V values are determined from
equation^ by setting U = umin, e.g. 1.0 gpm/SF, and substituting into
equation(S) the value of r as defined in equation^). The design
equation is stated as follows:
(V) = (H) ' (E) (10)
(Umin) {(1-E) (f-1)}
Using design equations(*) and (10) unit volume requirements for single
and multi-stage tower designs can be compared as follows:
492
-------
Single Two Three
Stage
1. Efficiency/Stage 95 77.5 65.0
2. Application Rate - gpm/SF 1.0 I.Q 1.23
3. Recycle Ratio - r 6.6 0.4 0
4. Unit Volume - CF/gpm 320 116.8 102
5. Total Flow Pumped 7.6Q 2.8Q 3Q
6, Organic Loading*
Ib BOD/1000 CFD 22.8 64.5 73.0
*based on influent raw waste
Because of the high BOD removal efficiency required, a multi-stage
plant is necessary to reduce tower volume. In the multi-stage design
sedimentation of recycle flow prior to tower application is indicated
to reduce media plugging tendency. Increased sedimentation capacity
is then necessary for a multi-stage design. A comparison of the total
capital cost for two and three stage design alternatives for: sed-
mentation; recycle pumping stations; and trickling filters was made as
follows: (ENR = 1540)
1. Two Stage $606,000
2. Three Stage $596,000
The cost saving for a three stage design is within the accuracy of
cost estimation and is thus, not significant. In order to eliminate
a third pumping station as a maintenance center, the two stage design
is selected.
Sludge Disposal Facilities
A material balance for waste treatment plant sludge solids is presented
on Table 10.
Under design conditions 5060 Ib/day of raw and biological solids
from various conventional treatment units require disposal. An
additional 1500 Ib/day of chemical precipitate is anticipated as the
result of lime addition to the final charifier. Lime addition is
included to adjust effluent pH and to increase final clarifier solids
removal by coagulation. Enhanced final clarifier solids removal is
required to insure compliance with treated effluent BOD concentration
493
-------
Table 10. Treatment of Whey & Sewage
Material Balance for Sludge Disposal
in pounds/day
Design
Description Condition
1. Influent Suspended Solids 1580
2. Effluent Suspended Solids -420
3. Difference 1160
4. Biological Solids produced 3900
5. Sub-Total Raw & Biological
Solids 5060
6. Conditioning Chemicals 355
7. Sub-Total Conditioned Solids 5415
8. Lime Precipitate Solids 1500
9. Total Solids for Disposal 6915
limitations. The provision for ferric chloride addition prior to
dewatering increases solids loadings 355 Ib/day. Total solids aggregate
6915 Ib/day.
Waste solids disposal will employ dewatering and land-fill. Thickening
of primary and secondary solids prior to dewatering will be required.
Based upon thickening studies, a solids loading of 6.5 Ib/SF/day can
be used to achieve an underflow solids concentration of 3.0 percent.
A total thickener area of 1070 SF will be required to process the
design solids loadings of 6915 Ib/day.
Sludge dewatering will be obtained by vacuum filtration. Laboratory
and pilot plant studies indicate the use of a filter loading of 1.2
Ib/SF/hr. Selection of a 6 hr/day filtration schedule will require
two, 500 sq ft vacuum filters. Special ventilation facilities will be
included to control odor levels in the filtration room.
Flow Sheet
A schematic flow diagram of the recommended waste water treatment
facilities is shown on Figure 21. The system employs primary
settling, two-stage packed tower trickling filters, final settling,
coagulation, sterilization by chlorination, and sludge dewatering.
Grit removal facilities have not been provided because the raw waste
is anticipated to contain virtually no grit and digesters (requiring
grit protection) are not included in the sludge disposal system.
494
-------
3 II S g
I °5 e . 5
£ I §S g I" Q S
UJ 0 — UI 5 GS2 . * 13 _l =•
3 i* 5"- d 2~ ° « 5° °
- *« u sg « ° £ 3
§ < z i ,sl | "£ g «5 P g 5 s
' | S 1 ^«S 1 S^8"" ^E8!^ a
> t ^ j 1 /^\k_
in f
o '
? ^
w '
' o:
0)
to
a.
>- ^
i» .n^f > ^ n^r in
•*'•' 'U^v y U v y"-,-1
i v — ^ v_^
Q
"^^-PRIMARY SLUDGE PUMPS
— 'REC IRCULATION •—
-^- COAGULAf^
* X 1 1 X
PR1MAKY SLUDGk — ' >— UNAL SLUUb
•* SLUDGE THICKE
/ — SUPERNATANT T
' 4 Ir^ THICKENED RLUDf
SLUUGb CUNUITIUNJNb -/ 1
CHEHICALS / H— SLUDGE STORAGE
1 1 1
L_H SLUDGE TRANSFE
r- FILTRATE
/ •+— UUOK L
VACUUM FILTER — '
••—SLUDGE CAKE
LAND FILL— y t1
§ -.
tn ^ — -i
O 5 £ »- *- ^
SK _ w g 5^
tl 5 § 5 <^
**UJ|_ ^ _. 1 IU ^W
J "5 s I 3 "1
f itu se °* >- ' 3
= 3 £ 5.
•> x -t- 3 :
5 iu a.
t Is"
Li ttf T \
\ y " * *r t
nn ' s I
tA
" A. It
f)— FINAL SLUDGE PUMPS H
i-
z
T 3
E Si
MERS
.£ PUMPS
TANK
R PUMPS
BNTROL
VILLAGE OF WALTON
Kill YORK
FLOW DIAGRAM OF PROPOSED
WASTE-WATER TREATMENT FACILITIES
PACKED TOU-r FILUR
QUlRfc.. LAWLER cv KATuSKY ENGINEERS
NEW YORK. NEW YORK
"3
M
a
NJ
M
-------
The screened influent will be pumped and metered to the primary settling
tank influent distribution chamber and mixed with a recirculation
flow from the 2nd stage packed tower. The combined flow will continue
through the primary tanks. The effluent from the primary settling tanks
is pumped up to the 1st stage packed tower distributor. This 1st
stage effluent is pumped to the 2nd stage packed tower distributor,
about 50 percent of the effluent of the 2nd stage packed tower will be
returned as a recirculation flow to the influent distribution chamber
of the primary settling tanks. The process flow stream continues
through the final settling and coagulation tanks, chlorine contact
tank and is finally diffused in the receiving waters.
Sludge from the primary and final settling tanks and coagulation tanks
is pumped to sludge thickeners.
The sludge thickener underflow will be stored in a sludge storage
tank prior to dewatering. The sludge cake will be disposed in a
sanitary land fill.
The treatment of whey bearing waste and its sludge dewatering is
associated with objectionable odors. Covers will be provided for the
packed towers and odor control ventilation equipment would be provided.
An architectural rendering of the complete treatment facilities is
presented as Figure 22.
Capital Costs
The division of capital cost for treatment of whey and sewage is based
upon their relative contributions of four major design loading parameters
and a detailed allocation of the capital cost of each major treatment
unit among the loading parameters.
The procedure assigns the cost of a unit to the design parameter (s)
which determines the capacity of the particular unit operation. • Allo-
cation of the capital costs of sludge disposal facilities is made
between suspended solids in the untreated waste and suspended solids
generated from BOD removal. Solids generated from final clarifier
coagulation are allocated to BOD.
Treatment plant items which cannot be attributed to a specific design
loading, such as administration building, are pro rated among the
design loading parameters based on the allocations achieved for all
other units. Cost allocations are also presented in terms of capital
cost (ENR = 1540) per unit amount of each design loading i.e.
$139,500 /mgd of peak hourly flow
$360,000 /mgd of average daily flow
$272/lb/day of dissolved BOD
$134/lb/day of suspended solids
496
-------
WALTON
MUNICIPAL
WASTE WATER
TREATMENT
PLANT
FIGURE 22
QUIRK.
LAWLER &
MATUSKY
engineers
New York.NY
497
-------
These unit capital costs do not include credit for a 60 percent
construction grant applicable to the facility.
TRICKLING FILTRATION OF WHEY EFFLUENT - SUMMARY
1. Trickling filtration of whey effluent and whey effluent mixed with
sewage has been shown to be an effective treatment method both on
absolute terms and on a relative basis when compared with activated
sludge.
2. A comparison of whey treatability with that of other industrial
effluents, demonstrates that whey treatability compares well with
that of the average industrial effluent when packed tower trickling
filters are employed.
3. Nutrient additions other than ammonia nitrogen were not productive
of increased rates of bio degradability. The addition of ferrous
iron did not, as anticipated by other investigators, result in an
increased treatability.
4. Nitrogen addition resulted in a 40 percent increase in the rate
of BOD removal.
5. The inherently acidic reaction of a whey effluent was shown to have
no adverse effect on trickling filtration performance using high
porosity media. pH variations from 7.0 to 4.5 were not detrimental
but, rather resulted in an increase of BOD removal rate as pH
decreased. pH increases above 7.0 were shown to result in a reduction
of BOD removal rate. Although these comments are based on limited
data the trend of pH influence appears clear.
6. Trickling filter operating variables of: temperature, recirculation
and hydraulic application rate have been quantified in a verified,
process design model(•*") summarized in this paper. While the number
of implications which can be drawn from a sensitivity analysis of the
model are too extensive to be within the scope of this study, several
qualitative comments can be made as follows: (a) Increased temp-
eratures are beneficial for process performance, (b) Recirculation
should be included only to insure adequate application velocities
and not to provide additional removal efficiency, (c) The
inclusion of recirculation will require provision for additional
filter volumes which will normally out weigh the effects of repeated
application, (d) The need to maintain application velocities
sufficiently high to insure adequate wetting of packing media will
require towers of maximum practicable height or stage treatment to
provide high BOD removals, economically. This latter requirement
is dictated by media characteristics rather than whey characteristics.
-------
7. Trickling filter performance will be sensitive to flow variation
rather than BOD variation. Response times of from 6-24 hours
should be experienced after a significant change in hydraulic
application rate. Under normal operating conditions, filter
performance should be stable.
8. Filter sludge growth will be prolific and will require a high
porosity media of low susceptibility to retention of sloughed
filter slime to avoid plugging.
9. Filter odors should not be offensive at organic loadings
requested for high BOD removals. However, filter installation
should be provided with covers if proximate to odor senstive
areas.
10. Final sedimentation of filter effluent may require coagulation
for production of a low solids effluent and/or for removal of
suspended BOD to insure an overall plant performance above
90 percent BOD removal.
11. Secondary sludge can be thickened using gravity equipment, however,
thickener requirements will be significantly greater than that used
for domestic sewage.
12. Dewatering of secondary sludge can be accomplished by vacuum
filtration. Centrifugation performance would be poor and would not
be recommended. Vacuum filtration characteristics of trickling
filter sludge exceed, significantly, those of activated sludge.
Dewatering requirements may well control the selection of the BOD
removal process.
499
-------
METHANE FERMENTATION OF WHEY
by
C. D. Parker,*
INTRODUCTION
Disposal of wastewater from the manufacture of dairy products, butter,
cheese, casein, powdered and condensed milks in Australia has created
problems of odour nuisance and water pollution.
In 1964 the Australian Dairy Produce Board Research Fund sponsored a
research project at this Institute to study ways and means of developing
low cost treatment of these wastes to obviate the problems created by
their disposal.
The project included a survey of existing methods of disposal, the
identification of the flow and composition of wastes from individual
processing operations and the experimental operation of treatment
facilities to treat general factory waste and undiluted whey by lagoons,
oxidation ditch and methane fermentation.
The disposal of undiluted whey arising from the manufacture of cheese
from whole milk and casein, from skim milk is a problem in many parts
of the world.
The annual production of whey in Australia is 800 million gallons per
year, an average value of 2.2 million gallons per day. It is typical
of the dairy industry in Australia that there is a marked seasonal
variation in milk production and the discharge of whey. In some cases
the peak/minimum intake ratio is as high as 14/1. Peak production
occurs in spring in the period September to December. The production
of whey during the peak period would be of the order of 6 million gallons
per day.
Many of the larger factories which basically produce butter are capable
of considerable flexibility in handling the skim milk. It may be
acidified and casein of various grades recovered and dried, or it may
be dried direct as powder either by spray or roller drying.
By far the greatest source of problems associated with whey disposal
is whey from casein manufacture.
EXISTING METHODS OF DISPOSAL IN AUSTRALIA
In Australia at present there is very little recovery of milk solids
from whey. One or two factories have dried cheese whey and powdered it
*Melbourne Water Science Institute, Australia.
501
-------
on roller driers, an attempt was made to recover lactose but this has
been abandoned.
Almost the whole of the 800 million gallons produced is either dis-
charged with the general factory waste flow, tankered to pig farmers
or spread over areas of waste land.
In some cases factories establish their own piggeries as a means of
disposing of whey. Because of the seasonal variation in whey production
it is not possible to rely on a regular level of supply for pig food
and in any case whey must be supplemented by other solid feed. As a
result it is seldom possible to dispose of the whole of the whey produced
to farmers at peak production. At a typical factory run piggery with a
peak whey production of 65,000 gallons per day over half cannot be used
for pig food and supplemental methods of spray irrigation are required.
In a few cases whey has been spread or spray irrigated over pasture or
bare land but the areas required are considerable. The maximum per-
missible rate of application appears to be on the order of 500-1000
gallons per acre per day and often this needs to be diluted 2-3 times
with factory waste if the pasture is not to be damaged.
Whey utilization for the recovery of milk soli'ds by yeast fermentation,
reverse osmosis and electrodialysis is currently being examined but
so far no experimental or commercial installations have been constructed.
METHANE FERMENTATION OF WHEY
Digestion of sewage sludge solids by the sludge digestion process is a
well established sewage treatment procedure.
In the digestion process solids are fermented first to volatile organic
acids by a variety of heterotrophic bacterial types and these acids are
then converted to methane and 002 by specific methane forming bacteria.
Schroepfer and others have shown how this process can be adapted to the
purification of high B.O.D. wastes from meat processing, in heated
digester type units.
Oswald and others have shown how the same process of methane fermentation
is involved in the reduction of B.O.D. of similar wastes in anaerobic type
lagoon installations.
Our own work (Parker 1967) on fruit and vegetable canning waste has
shown how the same mechanism can be used to reduce the high B.O.D. of
these wastes, also in anaerobic type lagoons.
The B.O.D. of whey is 35-40,000 ppm. very much higher than any of the
above wastes treated in this way, the composition of cheese and casein
whey is shown in Table 1.
502
-------
Table I Composition of Whey
Casein Whey Cheese Whey
B.O.D. ppra. 40,000 36,000
PH 4.3 4.8
Titratable acidity 45 53
(meq./l.)
As part of the research project studies were made to determine whether
whey could be fermented to methane and CO* in admixture with fermenting
sewage sludge or alone.
Fermentation with Sewage Sludge
At the Moe Sewage Treatment Works, there were available two heated sludge
digesters with facilities for measurement of gas yield from each inde-
pendently.
Because of excess digester capacity it was feasible to make one avail-
able for a trial fermentation of whey with digested sewage sludge. The
digester capacity was 42,000 gallons and the normal sludge load 2000
gallons per day. This was reduced to 500 gallons per day and a daily
addition of casein whey added. The casein whey was transported daily by
tanker to the sewage treatment works and stored in a 1000 gallon tank.
A concrete pit was constructed into which the unneutralized undiluted
whey could be measured. By valve the required daily addition could be
allowed to flow into a second pit together with digested sludge drawn
from the bottom of the digester. The mixed contents of the second pit
were then pumped back through the normal raw sludge inlet to the digester.
In this way effective mixing of the whey with the digester contents was
achieved.
Over a period of nine months, daily additions of whey were made, the gas
yield measured and samples of digested sludge and supernatant liquor
examined at weekly intervals.
Over this period, normal gas production continued and analysis of the
supernatant liquor displaced daily, showed that the B.O.D. was 300-400 ppm.
The maximum rate of whey addition achieved was 500 gallons per day re-
presenting a B.O.D. loading of 0.026 Ibs.B.O.D./cu. ft./day, a value
about 2/3 the usual design capacity for digesters. This represented a
detention time of 80 days.
Owing to demands on the plant for normal sewage sludge digester the ex-
perimental loading had to be discontinued before breakdown of the
digestion process occurred.
503
-------
This study showed that provided effective prefixing of the whey with
the digester contents was achieved substantial destruction of whey
B.O.D. could be brought about (99 percent) by methane fermentation.
Analysis of the digester gas produced showed this to be of a normal
value of 65 percent methane. 35 percent C0£.
Fermentation of Undiluted Whey
In view of the success of the Koe sludge digester operation an attempt
was made to maintain a successful fermentation in the laboratory with-
out any sewage sludge feed.
The fermentation was carried out in a one gallon glass jar. The fer-
menter was housed in an incubator held at 95" F. Fermentation was
initiated by using 500 ml. of digested sludge from the Moe digester
and the jar filled to the overflow level with water. Small daily ad-
ditions of unneutralized undiluted casein whey were made and these
slowly increased in volume.
Gas yield was measured daily and the supernatant liquor, displaced daily,
was analyzed once a week.
The maximum daily addition achieved without upset to the fermentation
was 250 ml, of whey per day. This represented a detention time of
L6 days and a B.O.D. loading of 0.13 Ibs.B.O.D./cu. ft./day. This
was maintained for over 3 months without upset to normal fermentation.
Gas yield at this maximum loading rate was 31 volumes of gas per unit
volume of whey (5 cu. ft. per gallon). Analysis of the gas showed
the normal methane content throughout the period of observation. Apart
from the initial addition of seed sludge no further solids addition
was made.
The supernatant displaced by the added whey had the composition shown
in Table 2.
Table 2
Whey Supernatant Liquor
Added Displaced
B.O.D. ppm. 36,000 450
Ammonia-Nitrogen 15 450
PH 4.3 6.9
By prior neutralization of the whey to 7.0 the maximum load could be
increased 10 percent. Cheese whey could be fermented equally well as
casein whey.
To determine the feasibility of the process on an experimental scale a
500 gallon fermenter was constructed and fed entirely with casein whey.
504
-------
The fermenter was heated externally through an electrically heated heat
exchanger. The temperature of the contents wera maintained at 95°F.
The contents of the fermenter were pumped continuously through the
exchanger, whey was introduced at a slow rate continuously by a peri-
staltic pump into the recirculated fermenter contents. Only the lower
third of the fermenter contents were circulated. The fermenter was
initially seeded with 50 gallons of digested sludge from the Moe digester.
It was found that by displacing the outflow from the top water level in
the fermenter very little solids are discharged. There is maintained a
constant volume of solids in the fermenter and addition of sludge solids
is not needed. Provision was made for settling solids out of the outflow
and their return by pump to the fermenter. In fact there was found to be
no need for this recovery of solids as a clear outflow was obtained.
Gas was led off the top of the fermenter and measured through a domestic
type gas meter.
The rate of whey addition was slowly increased. At the completion of
the research project period this had been increased to 60 gallons per
day, a detention period of 6-7 days. Owing to expiring of the grant
the load was not increased to bring about failure of the fermentation
process.
At this maximum rate of whey addition operation was continued for two
months and the composition of the outflow was as shown in Table 3.
Table 3 Experimental Fermenter Outflow
Whey Outflow
B.O.D. (ppm.) 36,000 450
pH 4.2 7.0
It can be concluded that with a detention of 6-7 days and probably with
a detention of about 5 days a B.O.D. reduction of 99 percent can be
obtained if the fermentation is carried out trder the conditions de-
scribed. The gas yield confirmed the laboratory value of 5 cu. ft. per
gallon of whey.
ANAEROBIC LAGOON TREATMENT
Whey can be successfully treated without nuisance by daily addition to an
anaerobic type lagoon in which algal growth is well established ^n tne
upper layers of the pond.
In these investigations it -was found that with a pond 30 ft. square and
5 ft. deep a regular daily addition of undiluted whey could be made at
a B.O.D. loading of 400 Ibs./ac./day. Under these conditions such a pond
could be operated for two years without nuisance.
505
-------
The B.O.D. of the pond contents averaged 250 ppm. and the pH was main-
tained in the range 6.5-7.5.
The detention time with such a loading is of the order of 1000 days
which in practice would mean that there is no outflow. An outflow
could be achieved if the whey were to be diluted with general factory
waste.
DISCUSSION
These investigations have shown that subject to careful observance of
loadings and optimum conditions for the appropriate microbiological
fermentations, the B.O.D. of whey can be reduced substantially by
methane fermentation either in admixture with sewage sludge in a
conventional sludge digestion unit alone at much higher rates by con-
tinuous culture fermentations or in anaerobic type lagoons in the
presence of an algal population.
The costs of treatment depend on the size of the operation and a number
of other factors. To treat a peak whey discharge of 20,000 gallons
per day based on Australian construction costs, by lagoons or in separate
whey fermenters would be as shown.
LAGOONS
Loading rate:
1st. stage - 400 Ibs.B.O.D./ac./day (85 percent removed)
2nd. stage - 60 Ibs.B.O.D./ac./day with 1/1 recirculation
of 2nd. stage effluent to inlet first stage.
Lagoon areas:
1st. stage - 18 acres.
2nd. stage - 16 acres.
Land 40 acres @ $500 per acre $20,000
Earth works involved 13,600 yds.
@ 40 cents/yd. $ 5,440
Pipeline 2700 ft. $ 3,000
Recirculation pump $ 500
$28,940
WHEY FERMENTER
Digester requirements: 100,000 gallon capacity at loading of 20,000
gallons per day.
Process requires continuous addition of raw whey, continuous recircu-
lation of contents maintained at 85°-90°F., and overflow of effluent
from near surface. Raw and recirculation addition to be made at base
of tank inducing circulation of contents. Recirculation draw off from
centre tank one third up from base. Overflow to pass over weir in in-
spection pit on wall.
506
-------
Optimum conditions expected to reduce B.O.D. from 36,000 - 400 PPm. and
produce 5 cu. ft. per gallon. per day "methane" gas.
For full scale plant of above capacity, the digester proposed would be
of mild steel construction, located above ground with suitable insula-
tion, and floating gas holder cover. Raw whey would be pumped from
whey storage to a header tank. The contents would be recirculated by
pumping through a heat exchanger which would be fired by gas evolved
in the process with fuel oil as standby heating fuel.
The gas system required would comprise the gasholder cover with flame
trap and safety valve; pipework and meter to the heat exchanger; pipe-
work, meter, regulator valve and flame trap to a waste gas burner for
disposal of excess gas; and necessary sediment and moisture traps.
Recirculation pump, heat exchanger, switchgear and pressure gauges
would be located within a weatherproof housing.
t - Steel 100,000 gal.
— — floating cover digester
1. Construction of Digester $ 40,000
2. Insulation $ 3,400
3. Heat Exchanger $ 6,000
4. Gas Equipment $ 4,000
5. Recirc. & Delivery Pumps $ 1,000
6. Pipework and Valves $ 1,700
7. Housing $ 200
8. Electrical $ 800
9. Raw Whey Storage
10. Modifications to Digester
$ 57,100
Provisional Sum 10 percent $ 5,710
Engineering & Supervision 10 percent $ 5.710
$ 68,520
It should be emphasized that both the methods propounded for methane
fermentation of whey have only been demonstrated on a very small scale,
Both need to be operated as a large scale demonstration unit before
they can be accepted as a reliable and economic method of treatment.
507
-------
STATE OF THE ART OF DAIRY FOOD
PLANT WASTES AND WASTE TREATMENT*
by
W. James Harper and John L. Blaisdell**
The dairy food industry is an important part of the overall food
industry of this country and contributes materially to fluid wastes
which, must receive treatment prior to their discharge to streams and
other waterways.
The industry has been in a dynamic state for the past several decades
with, major changes occuring within the industry. During this period
of time, there has been a slight increase in the amount of milk shipped
from farms for processing, a material decrease in the number of plants
processing dairy foods, a very significant increase in the amount of
milk processed by each plant, and a major trend towards automation and
mechanization. These changes have significantly increased the waste
load per plant. Under prevailing practices, waste loads per plant can
be expected to range from 2000 to 10,000 pounds of BOD per day.
The trend over the last several years for reduced milk production
appears to be reversing in that the amount of product being processed has
actually increased from 104 billion pounds in 1960 to 108 billion
pounds in 1969, whereas the total production of milk on the farm changed
from 123 billion pounds in 1960 to 116 billion pounds in 1969.
This paper reports a comprehensive study of the present status of dairy
wastes, in respect to composition, control and treatment based on
literature reports and current industrial knowledge and practice.
Little research in respect to dairy food plant wastes and waste treatment
has been conducted in this country in the past decade, except that which
has centered around the visible whey disposal problems that are the
subject of other papers at this conference. An attempt was made to
supplement our literature knowledge with information gained from current
industrial experience. A survey was made of all major proprietory and
cooperative dairy food firms in this country, and information was
obtained on current knowledge of about 697 plants in 38 states. These
plants comprise about 11 percent of the total dairy food plants in the
country, but process more than 65 percent of the total milk supply. In
addition, plant visitations were made to 30 dairy food plants. A major
* Supported by Grant No. 12060-EGU from the Water Quality Office of
the Environmental Protection Agency of the U. S. Government.
**Department of Dairy Technology, The Ohio State University, Columbus,
Ohio.
509
-------
emphasis was placed on the large, modern and automated plant, since
these types of operations are becoming the dominant type of dairy
food plant operation, and these large operations can be expected to
continue to increase. The significance of these plants can be
illustrated by the fact that 10 of the plants visited process about
2 percent of all the milk produced in this country at the present
time.
Industrial Knowledge of Waste Loads
During this investigation, effort was made to assess the knowledge of
the dairy food industry in respect to the knowledge that the industry
has of its waste load in terms of BOD, COD and suspended solids.
Information in respect to number of plants with (a) knowledge of waste
loads, (b) paying surcharge on waste composition, (c) their own
treatment facility and (d) waste water discharges to municipal sewer
systems is summarized in Table 1. Only about 11 percent of these plants
had any knowledge of their BODr in their waste water and only 7 plants
possessed any knowledge of COD. Knowledge of BOD,- loading was directly
related to those plants who had either been cited for recent violations,
had their own treatment facilities or were paying a surcharge on waste
composition. Payment of a surcharge on waste.composition and operation
of its own treatment facility, however, was no guarantee of knowledge
of the BOD^ by an individual plant. One national organization operating
several treatment facilities indicated that it did not measure 6005 on
the raw or treated waste and utili2ed only visual inspection as a means
of evaluation. Approximately 90 percent of all the dairy plants surveyed
discharged their waste water to a municipal treatment facility.
Generally, the dairy food industry knows the problem of fluid waste
and waste treatment will have to be faced, but has not taken the
initiative in solving the problem and with a few exceptions it appears
that the industry will not take remedial steps until forced to do so by
legal and/or economic pressures,
A major reason for the passiveness of the industry is economic. Because
of low profit margins, the expenditure of funds for non-economic terms
has been .avoided. The concept that the dairy food industry must consider
water pollution control as an intrical cost of doing business, like any
other production item, has not been accepted. In addition to economic,
other reasons for industry's passive attitude are (a) a broad lack of
knowledge of the nature and strength of dairy food plant waste, (b) a
failure to understand the potential economic value in recovering wastes
and converting them to usable by-products, and (c) a general lack of
understanding of the technology of waste control and treatment. Numerous
members of the industry who are not currently involved with operating
waste treatment facilities or paying surcharges on BOD or other waste
composition components, have little understanding of BOD, and most have
never heard of COD. As municipalities increase the practice of charging
the industry for its waste on a compositional basis and/or impose legal
510
-------
Table 1. Summary of Knowledge of Dairy Food Plant
Waste Water Loads by the Dairy Food Industry
of plants with
Plant Type
National and
Regional Propri-
etory Companies
Local Proprietory
Companies
Cooperatives
No. of
Plants
320
234
93
Knowledge of
waste loads
(BOD)
12.8
11.7
6.4
Payment on
basis of
BOD
5.0
8.5
8.6
Treatment
plants for
own wastes
9.1
6.8
10.7
Waste dis-
charge to
city
91.0
95.7
51.6
Total
647
11.5
6.8
8.6
87.1
-------
limits on waste composition going to municipal treatment facilities,
the industry will take a more active role in fluid waste control.
An active;nationwide educational program is needed to influence attitudes,
to provide a sound technological basis for controlling the pollution
problems of the dairy industry of the future.
Dairy Plant Wastes
Industrial wastes from dairy plants consist primarily of varying
quantities of water born milk solids from a variety of sources,
detergents, sanitizers and lubricants and domestic wastes. The
quantity and strength of the waste water discharged from the dairy
food plants vary widely depending upon the quantity of water utilized,
the type of process in the operation, and the control management exerts
over various waste discharges.
Information concerning the composition of the dairy food plant wastes
was obtained from both literature and from industrial sources. Prior
to quite recently, the majority of available.information in respect to
dairy food plant waste water composition was limited to BOD, suspended
solids and COD. Very limited information on dairy food plant waste
water is available for temperature, pH, fat, protein, carbohydrates and
other components such as phosphates, chloride, sulphur, surfactants and
sanitizers.
Dairy Food Composition and BOD
The biological and chemical oxygen demand of milk plant waste water will
vary as a function of the products manufactured since differences occur
in the amount of oxygen that is required for the oxidation of different
constituents such as fats, carbohydrates and protein. Various dairy
products differ widely in their relative concentrations of major organic
constituents and an understanding of the composition of various fluid
dairy foods is essential to a full understanding of dairy plant wastes.
The constituents of primary concern in water pollution for various
dairy products are cited in Table 2. These values are based upon current
average compositional values and vary somewhat in previously
compositional data for dairy products in earlier waste guides
Reported BOD values for various milk constituents and other organic
components that may be present in dairy food plant waste waters are.
listed in Table 3. The major three constituents of milk contributing
to BOD are lactose, milk fat and milk proteins with BOD5/lb of component
averaging 0.65, 0.89 and 1.03 respectively. Utilization of these
figures permits the calculation of BOD and the relative contribution
to BOD by various constituents. Calculated values, literature values
and percentage contribution of milk components to BOD in different
products are cited in Table 4 for most common dairy foods. Except
for whey, the calculated and literature values are in close agreement.
512
-------
Table 2. Average Composition of Milk and Milk Products (100 g.)
Product
Skimmilk
2% milk
Whole milk
Half & half
Coffee Cream
Heavy cream
Choc, milk3
Churned
Buttermilk
Cultured
Buttermilk
Sour cream
Yoghurtb
Evap. milk
Ice creamc
Whey
Cot. cheese
Whey
Fat
g.
0.08
2.0
3.5
11.7
18.0
40.0
3.5
0.3
0.1
18.0
3.0
8.0
10.0
0.3
0.08
Protein
g-
3.5
4.2
3.5
3.2
3.0
2.2
3.4
3.0
3.6
3.0
3.5
7.0
4.5
0.9
0.9
Lactose
8-
5.0
6.0
4.9
4.6
4.3
3.1
5.0
4.6
4.3
3.6
4.0
9.7
6.8
4.9
4.4
Total
Lactic Organic
Acid Solids Ca
g • g • mg •
8.56
12.2
11.1
19.5
25.3
45.3
18.5
0.1 8.0
0.8 10
0.75 24.6
1.1 10.8
27
41.3
0.2 6.3
6.7 6.1
121
143
118
108
102
75
111
121
121
102
143
757
146
51
96
P
mg.
95
112
93
85
80
59
94
95
95
80
112
205
115
53
16
Cl
mg.
100
115
102
90
73
38
100
103
105
73
105
210
104
195
95
S
mg.
17
20
19
16
12
9
19
15
17
12
19
39
20
8
8
Total
Ash
g.
0.7
0.8
0.7
0.6
0.6
0.4
0.7
0.8
0.7
0.6
0.7
1.6
0.9
0.6
0.8
Viscosity
(c.p. at 20°C)
1
2
2
7
15
25
15
1
500
10000
3000
30
.4
.4
.2
.5
.0
.2
.0
.5
35.0
1.4
1.3
Organic Ingredients Added
a. Sucrose, 6% & chocolate solids, 1%
b. Fruits
c. Sugar, 15%
-------
Table 3. Reported BOD Value for Various Milk Constituents
And Related Constituents
Pound BOPc / Pound Component
Constituent Range Average
Lactose .42 - .72 .65
Glucose .53 - .78 .66
Lactic Acid .63 - .64 .63
Milk Fat .70 - .95 .89
Glycerol .65 - .83 .75
Butyric Acid .34 - .90 .75
Sodium Butyrate - .41
Palmitic Acid - 1.07
Sodium Palmitate 0.4-1.02 .70
Stearic Acid - .80
Sodium Stearate .45-1.70 1.20
Milk Proteins .60-1.20 1.03
Casein .25-1.17 1.04
Hydrocolloids
Sodium Alginate - 0.36
Carboxymethyl-
cellulose - 0.30
514
-------
Table 4. A Comparison of Literature and Calculated BOD^ Values
and Percentage Contribution of Milk Components to Product BODij
% Contribution to BODq by
Product
Skimmilk
2% milk
Whole
milk
Half and
half
Coffee
cream
Heavy
cream
Choc
milk
Churned
buttermilk
Cultured
buttermilk
Sour Cream
Yoghurt
Evaporated
milk
Ice Cream
Whey
Cottage
cheese
whey
Literature
BODs
67,000
104,000
156,000
206,000
399,000
-
68,000
-
-
-
208,000
292,000
34,000
31,500
Calculated
BOD^
73,000
100,000
99,000
167,000
219,000
399,000
145,000
71,000
64,000
218,000
91,000
206,000
290,000
45,000
42,000
Milk
Fat
6.3
17.8
31.5
62.4
73.1
89.2
21.5
4.2
1.3
73.5
27.8
34.6
30.7
5.9
1.6
Milk
Protein
49.3
43.3
36.4
19.7
14.1
5.7
24.2
48.2
52.2
1.4
37.6
35.0
15.9
20.6
22.1
Lactose
44.5
39.0
32.2
17.9
12.8
5.0
22.4
46.7
39.4
10.7
27.1
30.6
15.2
70.8
68.1
Lactic
Acid
_
-
-
-
-
-
-
0.98
7.1
21.7
7.2
-
—
2.8
10.5
-------
BODg and Waste Water Volume Coefficients of Commerical Dairy Food
Plant Wastes
The strength of dairy food plant waste waters has been reported generally
in terms of part per million of BOD, without regard to total waste
water volume or the relationship of the amount of milk processed; and
waste volumes are often reported irrespective of the volume of product
processed. Waste water volume and strength coefficients have been
calculated from industrial and literature data in respect to the
amount of milk, or milk processed. For the purpose of this paper, the
waste coefficients are defined as follows:
Waste Water Volume Coefficient (WWVC) = <*•*•• waste water X 8'34)
(gal. milk processed x 8.6)
pounds of waste water/pound of milk processed
Waste Water Strength Coefficient = pounds Bpp5/day_
pounds milk processed/day x 1000
pounds BOD5/1000 pounds of milk processed
A summation of literature values for waste water coefficients for
various types of dairy food plant operations are shown in Table 5. The
use of volume as a reference for the coefficients eliminated the
"yield" variable for different products and brings the coefficients into
relatively close agreements for different types of dairy food plants.
Through personal contacts and plant site visitation, information on
dairy plant wastes and waste treatment was obtained for 57 plants in
terms of milk volume processed, gallons of waste water produced, and
ppm BOD5. The pounds of BOD per day, volume coefficient and BOD
coefficients were calculated (Table 6). All of the plants surveyed
used either advanced technology or a mixture of advanced and typical
technology with 30 of the 57 plants manufacturing more than one type
of dairy product. Statistical analyses indicated no significant
difference in BOD or volume coefficients and the type of plant operation;
except that BOD coefficients of the plants processing milk plus cottage
cheese and ice cream had significantly higher coefficients than the
other type of operation. Coefficients were unrelated to plant size
or to the degree of automation. Plants that were fully automated in
respect to processing and CIP systems frequently had higher than average
volume and BOD coefficients. The wide variability of both volume and
BOD coefficients among plants of similar technologies and size and a
lack of correlation between waste coefficients suggests that the controlling
factor in waste volume and BOD coefficients was management related. Based
on site visits for 20 plants, the evaluation was made of management
practices and coefficients. These are tabulated in Table 7. The data
516
-------
Table 5. Summary of Literature Values of Waste Water Coefficients*
Products
MfR.
Milk Receiving
Milk
Butter
Cheese
Condensed
Powder
M. CC. 1C.
Mix Prod.
No. of
Plants
6
11
2
12
2
10
2
5
Waste Volume Coef .
Range
4.6 - 12.5
1.5
1.4
0.3
1.2
0.8
0.8
1.1
- 18.6
- 8.3
- 5.1
- 2.3
- 11.5
- 1.2
- 6.8
Average
6.1
4.9
4.85
2.06
1.75
2.8
1.0
1.2
BODe Coef
Range "
0.02 - 4.8
1.1
0.8
0.2
1.0
0.6
0.6
1.3
- 22.0
- 2.1
- 4.1
- 1.9
- 12.3
- 0.9
- 3.2
•
Average
1.0
5.2
1.46
1.8
1.45
8.9
0.7
1.9
Total 46 0.3 - 18.6 3.9 0.2 - 22.0 2.6
* Volume coefficient in #/# milk processed, BOD5 coefficient in #BOD5/1000
#milk or milk equivalent received.
517
-------
Table 6. Summary of Commerical Plant Survey for Waste Water Coefficients*
Products
Mfg.
Milk
Cheese
Ice Cream
Cond. Milk
Butter
Powder
Cottage Cheese
No. of
Plants
6
3
6
2
1
2
3
Waste Vol.
0
1
0
1
1
0
Range
.1 -
.63 -
.8 -
.0 -
-
.5 -
5.4
5.7
5.6
3.3
5.9
.8 - 12.4
Coef .
Average
3.25
3.14
2.8
2.1
0.8
3.7
6.0
BOD Coef.
0
1
1
0
0
1
Rang
.2 -
.0 -
9 _
2 _
-
.02 -
.3 -
e
7
3
20
13
4
71
.8
.5
.4
.3
.6
.2+
Average
4.2
2
5
7
0
2
34
.04
.76
.6
.85
.27
.0
Cottage Cheese &
Milk 19 0.05 - 7.2 1.84 0.7 - 8.6+ 3.47
Cottage Cheese
Ice Cream &
Milk 9 1.4 - 3.9 2.52 2.3 - 12.9 6.37
Mixed
Products 5 0.8 - 4.6 2.34 0.9 - 6.95 3.09
Overall 56 0.1 - 12.4 2.43 0,2 - 71.2 5.85
*See footnote 1, Table 5
+Whey included, whey excluded from all other operations manufacturing cottage
cheese
518
-------
Table 7. The Effect of Management Practices on Waste Coefficients
Plant
No.
1
42
43
6
36
37
9
26
48
8
10
40
Products
Manufactured
Milk
Milk
Milk
Cottage Cheese
Cottage Cheese
Cottage Cheese
Ice Cream
Ice Cream
Milk
Cottage Cheese
Milk
Cottage Cheese
Milk
Milk
Milk Processed
// / day
400,000
150,000
600,000
300,000
650,000
17,000
34,000
250,000
1,000,000
900,000
//BOD/
10000
Milk Proc.
0.3
7.8
0.2
2.0
1.3
71
32.2
2.1
0.7
8.6
3.3
#Waste
Water ///
Milk Proc.
0.4
5.2
0.1
0.8
4.7
12.4
5.3
0.8
1.0
7.0
1.1
Level
Management
Practices
Excellent
Poor
Excellent
Good
Good
Poor
Poor
Good
Good
Poor
Fair
Cottage Cheese
1,000,000
2.1
1.2
Good
Explanation of Prac.
Rinses saved, hoses off -
out of use, Filler drip
pans
No steps taken to reduce
waste
Rinses saved, returns excld
Filler drip pans cooling
power
Whey excld, Fines screened
out, Wash water to drain
Whey excld, Spilled curd
handled as solid waste
Whey incld, poor housekeeping
Rinses to drain leaks, drips
Water running-not in use
Freezer rinses segregated
Whey & wash water excld
Rinses segregated
Whey excld, many drips
leaks, returns incld
Whey excld. Good water
volume control
Whey excld, rinses saved
-------
Table 7. (Cont.'d)
N3
O
Plant
No.
52
3
30
33
34
44
50
56
Products
Manufactured
Milk
Cottage Cheese
Milk
Ice Cream
Cottage Cheese
Milk
Ice Cream
Cottage Cheese
Milk
Ice Cream
Cottage Cheese
Milk
Ice Cream
Cottage Cheese
Milk
butter
Whey Powder
Milk Powder
Butter
Milk Processed
// / day
765,000
400,000
800,000
600,000
900,000
300,000
500,000
200,000
//BOD/
1000#
Milk Proc.
1.8
3.9
7.7
12.9
9.1
0.9
0.2
3.0
//Waste
Water ./#
Milk Proc.
1.1
1.4
3.5
3.3
2.8
0.8
5.9
2.5
Level of
Management
Practices
Good
Fair
Poor to
Fair
Poor
Poor
Good
Good to
Fair
Fair
Explanation of Prac.
Returns excld, good water
control
Whey & wash water excld
Rinses excld
Whey excld, sloppy oper
spillage, leaks, hoses run
Whey incld, poor housekeeping
pumps, valves, lines leak
hoses run
Whey excld, many leaks,
drips, etc.
Buttermilk excld, few
leaks, dry floor condi
No entrainment losses, all
powder handled as solid
waste, no leaks or drips
Continuous churn hoses
running, numerous leaks
& drips
-------
clearly revealed that there was a direct relationship bewween management
attitudes and practices and the coefficients. Under extremely good
management, coefficients of 0.5 pounds of waste water per pound of milk
processed and 0.5 pounds of BOD per thousand pounds of milk processed
were obtainable. A realistic average of 2.0 pounds of BOD per thousand
pounds of milk processed and 1.5 pounds of waste water per pound of milk
processed appeared to be generally achievable under good management.
Coefficients above 3.0 can be considered to be excessive and an indication
of poor management.
With whey and water milk excluded, the waste coefficients of different
operations are quite similar. If the overall average coefficients of
2.4 for volume and 5.8 for BOD are assumed to be representative of the
national dairy food industry, the total waste water from dairy food plant
operations would be equivalent to 200 million pounds of waste water with
475 million pounds of BOD^ per year. Since observations have indicated
that plants without prior knowledge of their waste loads have coefficients
in excess of three pounds of waste water per pound of product processed and
3.0 pounds of BOD per thousand pounds of product processed, the actual
amount of BOD from dairy food plants - taking into consideration the
50 percent wasting of by-products at the present time, we can estimate a
current overall organic waste load from dairy food of 4.0 billion pounds
of BOD5 per year at the present time.
BOD-COD Inter-relationships
The 1959 Revised Guide for Dairy Plant Waste Treatment (19) recommended
the utilization of COD in preference to BODg as a means of measuring
the strength of the wastes. This recommendation was based on the work
of Forges and co-workers(11) utilizing skimmilk as a model for dairy
food plant wastes. Investigators are in agreement in respect to the
BOD-COD relationships between whole milk, skimmilk and whey. The
average values cited are:
Product BOD-COD ratio
Whole milk 0.69
Skimmilk 0.63
Buttermilk (churned) 0.66
Whey 0.52
Lactose 0.53
Casein 0.46
Whey protein 0.23
Fat 1.28
With the exception of the BOD-COD ratios of the constituents of milk are
much lower and the lowest biologically oxidizable material is the whey
protein.
BOB-COD ratios reported in the literature for industrial dairy food
plant wastes are shown in Table 8. Values vary widely and are generally
less than for milk products themselves, as might be anticipated. Overall,
the BOD-COD ratio ranged from 0.10 to 0.88 with an average of 0.53, as
compared to a BOD-COD ratio of 0.65 for whole milk. Values below 0.6
-------
Table 8. BOD5/COD Ratios for Raw Dairy Plant Wastes
Investigator
Hoover & Forges
Schulz-Fulkenheum
Walholz
Furoff
Schweizer
Rensink
Christansen
Walgren
Bergman, et al
Bergman, et al
Annon
Current plant
Type of
Plant
-
-
-
multi-produce
cheese
butter
-
-
multi-product
multi-product
-
—
Year
1953
1955
1956
1960
1960
1962
1964
1966
1966
1968
1968
BOD s /COD
Range
0.34-0.80
0.30-0.70
-
0.22-0.51
0.31-0.66
-
0.55-0.77
0.11-0.75
0.43-0.60
0.40-0.75
0.22-0.88
0.10-0.76
ratios
Average
-
0.64
0.33
0.45
0.66
0.64
0.47
0.53
0.56
0.35
0.48
survey
Total
0.10-0.88
0.53
522
-------
can be interpreted to suggest less efficient biological oxidation of
milk wastes than pure milk; probably caused by the presence of non-milk
constituents. Possible "toxicity" of dairy plant waste is suggested by
values below 0.4, but no reference to possible "toxicity" has been pre-
viously cited.
In the course of this investigation, several national and/or regional
dairy companies who have had a long experience in measuring the strength
of dairy plant wastes, indicated they had attempted to utilize COD but
abandoned the position several years ago because of wide variations in
BOD-COD ratios and apparent lack of agreement between values suggesting
that COD values were less reproducible and of less value than BOD. Data
was obtained from one plant for BOD-COD ratios in raw and treated wastes
at hourly intervals during a production day. As shown in Table 9, the
ratios varied from 0.12 to 0.90 at different periods of the day. Low
ratios between 1:00 and 5:00 p.m. coincided with the major periods of
equipment process cleaning. These low values suggest either "toxicity"
from detergents or the presence of a large amount of refractory material.
Solids, pH and Temperature
Available information from industry for solids, pH and temperature of
dairy food plant waste water in relation to BOD strength is summarized
in Table 10. The suspended solids content of dairy plant waste waters
from the literature varied between 2400 and 4500 ppm. Based on the limited
literature values available, there was no statistically significant
correlation between the suspended solid strength and the type of dairy
plant operation. The data suggest that over 70 percent of the suspended
Table 10. Solids, pH and Temperature of Dairy Food Plant
Waste Waters (Industry Values)
Waste Component
Suspended Solids, ppm
Volatile Suspended Solids,
ppm
Volatile Total Solids,
ppm
Total Solids, ppm
PH
Temperature,°F
BOD, ppm
No. of Values
24
19
22
27
30
17
11
Range
24-5700
17-5260
57-4700
135-8500
5.3-9.4
55-120
15-4790
Average
1497
2397
7.1
76
2100
523
-------
Table 9. BOD/COD Relationships in Raw and Treated
Dairy Plant Wastes (1967) at Hourly Intervals*
BOD:COD Ratios
Time
8:00 a.m.
9:00 a.m.
10:00 a.m.
11:00 a.m.
12:00
1:00 p.m.
2:00 p.m.
3:00 p.m.
4:00 p.m.
5:00 p.m.
6:00. p.m.
7:00 p.m.
8:00 p.m.
9:00 p.m.
10.: 00 p.m.
11:00 p.m.
12:00
.4
.35
.51
.28
.66
.26
.21
.21
.12
.25
.45
.60
.57
.55
.88
.90
.90
Raw
.4
.35
.51
.28
.66
.26
.21
.21
.12
.25
.45
.60
.57
.55
.88
.90
.90
Treated
.32
.29
.26
.25
.15
.48
.90
.82
.77
.59
.75
.57
.68
.54
.64
.63
.60
Rate of
Waste Water Flow
(mgd)
.025
.020
.025
.010
.035
.035
.035
.050
.035
.045
.035
.014
.020
,020
.022
.010
.005
Raw
COD
Values
3440
3200
3400
4700
2600
3400
5200
5800
9280
6240
7760
3200
4480
2720
2400
1056
886
*Equipment Sanitized 6-8 a.m.
Clean up 1 - 6 p.m., maximum activity about 2-4 p.m.
524
-------
solid is volatile, or organic in nature, in dairy food plant waste
waters. Volatile suspended solids, as percentage of total suspended
solids, ranged from 68 percent to 98 percent with an average of 85
percent. The suspended solids are composed primarily of coagulated
milk, fine particles of cheese curd and pieces of fruits and nuts from
ice cream operation, and contribute from 10 percent to 30 percent of
the total solids. Volatile solids as percentage of total solids,
ranged from 36 percent to 96 percent with an average of 63 percent.
The non-organic total solids is about twice the non-organic volatile
suspended solids. The proportion of non-organic solids in milk is
about 13 percent, which suggests that most of the suspended solids in
dairy food plant waste waters are of a dairy food origin. The increase
of non-organic matter in a total solid reflects the contribution of
non-organic material from detergents, sanitizers and lubricants.
The pH of the dairy food plant waste waters varied from 4.4 to 9.2 with
a median value of 7.2. Although it would be expected that cheese
plants might have a lower pH in their waste water than other types of
plants, this was not found in the data available. The major factor
affecting pH of dairy food plant wastes is the cleaning compound, either
acid or alkali and its relative significance in the dairy food plant
waste water.
Suspended solids to BOD ratios ranged from 0.25 to 0.35 with an average
of 0.31, and the total solids to BOD ratios were essentially identical
for different types of dairy plant wastes with an average of 1.62. The
relationship of solids to BOD appeared to be independent of the type
of dairy plant operation and characteristic of all dairy plant waste
waters.
Only limited studies have been made of the temperature of dairy food
plant waste waters as they leave the dairy plant. Reported values
ranged from 70°F to 115°F (20°C to 40°C). The most comprehensive
study in this area was made by Zall^'4) who showed that waste water
temperature in a small dairy plant could be reduced about 15 percent
by management control of the use of hot water in cleaning, and by
eliminating the practice of allowing hot water hoses to run while not
in actual use.
Chemical Composition of Dairy Food Plant Waste Waters
No systematic complete study of the chemical composition of dairy plant
waste waters has been made to date, and only limited data is available.
The ?iost comprehensive studies have been those of Walgren^O) an
-------
food plant wastes is important because of the role of the various
refractory elements in water quality. The organic composition and
concentration of phosphate and chloride reported in various studies
are presented in Table 11 for literature and industrial survey data.
Limited data indicate the following differences in the ratios of
fat to protein and to lactose.
Source Protein/Lactose Ratio Fat/Lactose Ratio
whole milk 0.64 0.70
dairy waste water 0.56 0.12
dairy waste water 0.43
dairy waste water 0.85 0.57
dairy waste water 1.08 0.42
dairy waste water 1.14
dairy waste water 0.32 0.05
Protein/lactose ratios and fat/lactose ratios in dairy food plant
wastes are significantly different from those for whole milk.
The phosphate values which vary independently of BOD, were all in excess
of 10 ppm and could be considered as a potential source of phosphate
for inhancing algae growth. The values varied widely, apparently
reflecting a difference in the amount of phosphate cleaning compound
used in different dairy operations. Most chloride values in waste
water exceeded 75 ppm which is considered to be the lerel that cannot
interfere with industrial processes. Forty-five percent of the
chloride values given exceeded 250 ppm.
To interpret the data of these and other refractory elements in dairy
wastes; values for calcium, potassium, sodium, phosphorous, magnesium,
chloride and nitrogen in milk wastes containing 0.1 percent (1000 ppm
3005) milk solids would be:
Material ppm
BOD 1000
nitrogen 55
chloride 10
phosphorous 12
calcium 12
sodium 4
potassium 15
magnesium 5
Available data for magnesium, calcium, sodium and potassium are also
summarized in Table 11. The values differ widely in respect to sodium
(60-807 ppm), and potassium (11-160 ppm) and were less variable for
magnesium (25-49 ppm) , and calcium (57-112 ppm). Variations in the
latter two minerals reflected variations in BOD, whereas variance in
526
-------
Table 11. Chemical Composition of Dairy Food Plant Waste Water
_ PPM _
Constituent Range Average
450-4790 1885
Nitrogen 15-180 76
Protein 210-560 350
Fat 35-500 209
Carbohydrate 252-931 522
P 11-160 50
Cl 48-469 276
Ca 57-112 37
Mg 25-49 87
N 60-807 322
K 11-160 67
sodium and potassium were unrelated to BOD and reflected differences in
the sodium and potassium composition of cleaning compounds and water
treatment utilized in the different plants. Based on these data, milk
salts would contribute about 10 to 50 percent, 10 to 75 percent, 1 to
10 percent, and 25 to 100 percent of the magnesium, calcium, sodium
and potassium ions, respectively in dairy food plant wastes. Thus,
90 percent of the sodium, 50 percent of the magnesium and 25 percent
of the calcium appeared to come from non-dairy sources, presumably
cleaning compounds, detergents and lubricants.
Lubricants are not mentioned in the literature as a source of pollutant
in dairy plant waste waters, only limited data could
be obtained during the course of this study concerning specific contribution
of these materials to BOD. Most of the lubricants are salts of fatty
acids and range in BOD from 0.8 to 1.2 pounds of BOD^ per pound of
material.
Detergents and Sanitizers
Selected compounds used in the manufacture of detergents and related
materials and their contribution to BOD5 (Ib/lb of product) are shown
in Table 12. Wetting agents and surfactants varied widely in BOD values
ranging from 0.05 to 1.2. The most commonly utilized surfactant
contained 0.65 pounds of BOD5/ pound of product; non-ionic wetting
agents exhibit a low BOD, with average values of 0.2 pounds BOD per pound;
and acids which are used in dairy food plant detergents, have BOD
values ranging from 0.25 to 0.85. The most commonly used acid
detergent in cleaning food plant equipment has a BOD value of 0.65
pounds BOD^/pound of detergent.
527
-------
Table 12. BODr of Selected Chemicals in Detergents, Sanitizers,
and Lubricants used in Dairy Food Plants
Material //BOD/// Product
Acetic acid 0.65
Duponol B, alkyl alcohol, sufonated 0.45
70% hydroxyacetic acid 0.07
alkyl phenyl condensate of ethylene oxide 0.04
Phenoxypolyloxye thylene 0.00 5
Nacconol NR-Na alkylarylsulfonate 0.004
Neutrony x 600, aromatic polyglycol ether 0.0
Nopco 1111-sulfonated coconut oil 0,96
Nopco 1665-soluble fatty acid ester 0.12
Pine oil 1.08
Tallow 1.52
Triethanolamine 0,01
Ultra Wet DS-soldiutn alkylarylsulfonate 0.0
Linear alkylarylsulfonate 0.65
Ethylene glycol 0.70
Zalon-fatty amide 0.20
The amount of various detergent components used in cleaning various
types of dairy food plant by hand and CIP method were calculated and
are listed in Table 13 in lb/1000 Ib of milk processed. Alkali, as
sodium hydroxide, is the major component ranging from 0,23 to 1.52
lb/1000 Ib of milk processed for hand cleaning and from 0.23 to 0.47
for CIP. Phosphate values ranged from 0.05 to 0.76 and 0.04 to 0.21
for CIP and hand cleaning, respectively. Acid ranged from 0.08 to
0.84 for hand cleaning and from 0.07 to 0.15 for CIP, whereas
surfactant was the component used in the lowest concentrations, with
amounts per thousand pounds of milk processed ranging from 0.02 to
0.11 for both hand and CIP cleaning, respectively. Ice cream plants
utilized the highest concentration of detergents, with the least
differences between hand and CIP cleaning procedures. This reflects
the relative amount of equipment in ice cream plants still cleaned
by hand and the fact that CIP is limited to pipelines and storage tanks.
Only acids and surfactants contribute BOD to the waste water. The
amount of 6005 in the waste water in lb/1000 Ib processed from
materials used in dairy food plant detergents as a function of
concentration is shown in Figure 1. Under average conditions in the
modern milk plant, the amount of BOD contributed by surfactant and
acid detergent could be about 0.1 lb/1000 Ib of milk processed. The
organic acid appears to be the major source of BOD and could be substi-
tuted by an inorganic acid. However, the major inorganic acid of choice
will be phosphoric acid which would add substantially to phosphate levels.
528
-------
Table 13. Coefficients of Components of Cleaners & Sanitizers
#/1000# Milk Processes
Product
Alkali (NaOH)
Phosphate (hexa-meta P04)
Surfactant
Acid
Chlorine
Market
hand
0.77
0.12
0.08
0.84
0.34
Milk
CIP
0.25
0.05
0.02
0.15
0.10
Butter
hand
0.79
0.76
0.025
0.35
0.14
CIP
0.25
0.15
0.08
0.10
0.08
Cheese
hand
0.65
0.38
0.08
0.14
0.08
CIP
0.41
0.07
0.05
0.09
0.03
Condensed
hand
0.23
0.046
0.02
0.088
_____
& Powder
CIP
0.21
0.04
0.02
0.07
____
Ice
hand
1.52
0.22
0.11
0.52
0.37
Cream
CIP
0.47
0.22
0.11
0.15
0.13
-------
00
c
Ul
CjO
O
tr
c
rt
O
l-h
n>
rr
fD
i-l
00
fD
i-t
H-
0>
M
(fl
rt
O
D3
O
a
01
rt
(B
CO
anionic surfactant or acid
(BOD = 0.65#/#)
non-ionic surfactant
O.I 0.2 0.3 0.4 0.5 0.6 0.7
# DETERGENT PRODUCT / IOOO * MILK PRODUCED
-------
Data for cleaning and sanitizing solution in dairy plant waste water
for 27 plants are summarized in Table 14. Cleaning solution in percent
of waste water, ranged from 2.2 to 41.6 percent with an average value
of 15 percent. The sanitizer solution contributed 0.2 to 13.8 percent
of the waste water and averaged 3.1 percent. The lower the waste water
coefficient for a dairy plant, the greater was the percentage of that
waste water made up of detergent and sanitizer solutions. The average
concentration of the detergent in the cleaning solution would be 5000
ppm and for sanitizers the average would be 100 ppm. In terms of the
waste water, the concentration of alkali, sodium hydroxide, ranged
from 32 to 6280 ppm with an average of 500 ppm; phosphate ranged from
30 to 75 ppm with an average of 43 ppm; acid ranged from 10 to 348 ppm
and averaged 66 ppm; whereas chlorine ranged from 6 to 234 ppm and
averaged 70 ppm.
Based on the data obtained during this investigation, it will appear
that detergents and sanitizers may be more significant in dairy wastes
than previously reported'^). They potentially contribute significantly
to BOD, to refractory COD and may be significant in providing toxicity
and poor performance to dairy waste treatment facilities. Their effect
on the dominating microflora of dairy food plant waste treatment systems
and a full determination of their real significance requires further
investigation.
Variation in Dairy Food Plant Wastes
Frequency plots of the variability in day to day BOD,- values for six
different dairy plants are shown in Figure 2. Five of the six plots
follow a normal distribution and show a straight-line relationship,
whereas the sixth plot shows an S-shaped distribution typical of very
poor day to day control. Where normal distribution existed, the maxi-
mum BOD level (at a 1 percent confidence limit) ranged from 1.77 to
2.46 times the mean value, with an average of 2.1. For the plant with
abnormal distribution, the maximum BOD was 3.6 times the mean value.
Variation is a characteristic of dairy food plant wastes which needs
to be evaluated on a plant to plant basis in order to provide information
for both design and proper operation of waste treatment systems. On
a daily basis, control of variables in the proper process can result
in a normal and predictable distribution around the mean. Generally,
maximum values at 1 percent confidence limit can be predicted by
multiplying the mean value by 2.5.
Sources of Dairy Food Plant Wastes
The most visible pollutant in dairy food plant waste waters is whey
from cheese and cottage cheese operations; with about 17 billion
pounds of sweet whey being produced annually in this country from
ripened cheese and about 6 billion pounds of acid whey coming from
cottage cheese and allied processes. The next most significant source
of BODc is derived from every 1000 pounds of milk processed into
cottage cheese from whey and another 8 pounds comes from the wash
531
-------
Table 14. Cleaning and Sanitizing Compounds
in Dairy Food Plant Waste Water
% Waste Water as
Average ppm of Following
Type of Plants
Milk
Cheese
Ice Cream
Cottage Cheese
Cleaning
Range
3.4-21
4.6-22
2.22
__w
Solut.
Aver.
9.4
14.3
12.0
18.0
Sanitizing Solut.
Range
0.6-5.0
5.0-6.2
0.2-3.0
_._
Aver.
2.2
5.8
1.3
10
Alkali
353
43
150
Acid
51
—
33
75
Phos-
phate
175
10
64
Sur-
factant
52
—
5
21
Cl
155
52
16.8
234
Ul
LO
Milk & Cottage
Cheese 7-41 17.0 0.3-13.8 3.4 360
Milk, Cottage
Cheese and
Ice Cream 10-22 15.6 1.4-5.0 3.0 184
55
34
56
73
41
25
102
57
-------
Ln
to
00
C
1-1
O O
(D rr
•1 O
(D
D O
rt 0>
O H-
O O
O- P
0> 3
P
rt DO
cn o
D
01
01
rr
(D
0>
rt
0>
i-t
O
l-h
80
cn 60
Q
cn
ID
O
40
Q
O
GO
20
20
50
90
99
FREQUENCY DISTRIBUTION
-------
water. Under average conditions the total BOD5 from all other dairy
food plant processes combined is about 2-5 lb/1000 Ib of milk processed.
Literature values, for BOD5 from various unit operations which are in
general agreement with each other, have been presented by two investi-
gations^, 18). Values obtained from several modern, automated dairy
food plants manufacturing fluid milk product and cottage cheese are in
Table 15. Exclusive of whay and wash water, the process yielding the
highest BOD5/1000 Ib of milk processed is pasteurization; with the milk
solids coming from the start-up, change-over and shut-down of the
process on water. The next most significant source is the filling
operation, where machine jams in high speed packaging equipment is
frequent. The loss of 5 to 10 gallons of milk in a single machine jam
is not uncommon.
Limited data for ice cream operations indicated that the loss of BODj
in rinsing mix tanks and ice cream freezers ranged from 2-10 Ib of
BODr/1000 Ib of milk processed into ice cream. Because of the high
viscosity of the mix and ice cream, the percentage of product remaining
on the surface after draining can be up to 8.5 percent of the total
present in the processing equipment.
Because of its significance, as indicated in other presentations today,
the cottage cheese operation merits special attention. Cottage
cheese whey BODr values in industrial practice were found to range from
30,000 ppm for filtered whey to 65,000 ppm for whey with a high level of
fine particles present. Thus, in addition to whey and wash water, the
amount of curd shattering can be a significant source of both BOD and
suspended solids in cottage cheese operations. The 6005 for whey,
wash water and fines in one commercial dairy food plant is presented in
Table 16. As indicated, the total BOD value of the fines was lb/1000 Ib
of milk processed into cottage cheese. The first wash water contained
about 60 percent of the total BOD in the wash waters. Effort is needed
to minimize the BOD in wash waters, since the solids level (less than
2.0 percent) is too low for practical recovery. This may be accomplished
by: (a) complete draining of the curd, followed by addition of 10
percent volume of wash water that is drained and combined with the
whey, before the first complete washing, (b) utilization of whey as a
means of cooling the curd, in which whey is removed, cooled and pumped
back on the curd; thus reducing the volume of wash water required, and
(c) counter current washing through a cooling tower. The possibility
exists also for developing new methods of cottage cheese manufacture
from concentrated milk which would reduce the volume of whey and wash
water, but this is in the future from a practical viewpoint.
Waste Control and Treatment
Waste Reduction;
Minimization of waste water volume and strength is necessary to reduce
the waste coefficients found in most plants to the achievable level of
about 0.5 pounds waste water/Ib of milk process and 0.5 pounds of
BOD5/1000 Ib of milk processed. This can be achieved with current
technology, if management makes the necessary effort.
534
-------
Table 15. Industrial Data for BOD Coefficients for Unit
Processes in Two Fluid Milk Cottage Cheese Plants
Ibs of BOD/1000 Ibs of milk received
Plant A Plant B
0.25
0.23
0.15
0.75
0.30
0.25
0.08
Process
Tank truck rinsings
Silo tank rinsings
Tank trucks and storage tank
rinsings
Separation (CIP)
CIP Separator sludge
HTST Start up and Change over
and cleaning
Filling operations-milk
Lubricants
Product Returns
Cottage cheese wash water
Unaccounted
Total Waste Water 2.16 2.01
*Initial start up collected, product change overs not diverted.
Table 16. BOD Coefficients for Cottage Cheese Operations*
ppm BOD
Operation
Whey C. C.
First wash
Second wash
Third wash
Fourth wash
Transfering
Filling*
Fines
3,700
17,000
1,700
2,000
2,200
B
3,500
11,700
2,200
550
185,000
//BOD/1000 Ib. milk proc.
A B
26.6
5.8
1.3
0.7
0.75
25.2
6.3
1.2
0.30
0.027
2.2
*0peration A used 3 washing with each -wash being equal to 75 percent of
whey volume. Operation B used 5 washes, with the 1st, 3rd, and 4th washes
being equal to 50 percent of the whey volume and the 2nd wash being equal
to the. whey volume.
•fValues based on curd spillage. Usual amount of curd loss would range
from 5-10 lb/100 Ib of cottage cheese.
535
-------
To optimize the control of waste In the dairy industry, the following
suggestions are made for plant operations.
1. Segregation of all major sources of waste with separated drains as
required being put into all new plants and remodeled plants.
2. Segregation of recycling of non-polluted water (including water,
etc.).
3. Utilization of the automation system of the plant to eliminate
wates; collecting the first rinses from tanks and cleaning
operations, and saving the water-milk mixtures resulting from
start-up and changeover shutdown of HTST units.
4. Eliminating all product discharged during start-up from product
changeover of HTST pasteurizers.
5. Improvement in the efficiency of GIF and sanitizing operations to
reduce required concentration on these materials.
6. Use of post-cleaning rinses as make-up water for sanitizing and/or
cleaning,
7. Elimination of returns from waste water streams.
8. Collection of lubricants and reclamation for reuse.
Implementation of these practices can reduce BOD^ coefficients to about
1.0 pound in a well managed plant.
Waste Treatment
Dairy food plant wastes are treated primarily by biological oxidation
methods, with over 90 percent of all the dairy food waste water being
treated in municipal systems. Where the municipal systems are receiving
more than 50 percent of their BOD^ from milk wastes or more than 10
percent of the BOD^ from whey, the treatment plant takes on characteristics
of a dairy food plant waste treatment system.
Biological oxidation is a function of the microflora of the waste
treatment system, which in turn is dependent upon the composition of the
wastes. According to Dias and Bhat^-1, the dominant microflora in a
municipal activated sludge plant in decreasing order are Zoogloea,
Comannanmonas, Pseudomonas, Micrococcus, Flavob'acteriuro , Achromobacter,
Alcaligenes, Corynebacterium, Bacillus, Spirillum and yeasts. All of
these have been found in activated sludge of dairy waste treatment.
However, the dominant rricrofloia reported by various investigators for
dairy activated .sludge are quite different from that of domestic activated
sludge. Adarnse^ J found that the adaptation of the bacterial flora
from municipal to dairy wastes required approximately 50 to 60 days.
536
-------
In his investigations, the dominant microflora were Corynebacterium,
of the Arthrblacter type, followed by Achromobacter. Pipes(^' in his
investigations of activated sludge plants found that the dominant
microflora in his particular investigations were Bacillus, Beggiatoa,
Arthrobacter, Actinomycetales, and Sphaerotilus. Similar micro-
organisms have been reported in the microflora from trickling filter
and aerated lagoon systems operating on dairy wastes.
Only limited investigations have been made of the metabolisms of the
microflora of various dairy food plant waste treatment systems. Since
the dominant microflora of dairy food plant waste treatment systems
appear to differ from the microflora of municipal systems, there is
need for further specific information. The most exhaustive study in
respect to the factors affecting the metabolism of the microflora
of dairy.food plant waste treatment systems have been conducted by
Adamse' . In investigation of the growth characteristics of the
major groups of organisms in his system, he found that the Arthrobacter
species was the most active and efficient of the three types, with a
generation time of 2.5 hours at 15°C. In batch experiments with well
aerated activated sludge, the oxygen demand immediately after feeding
artificial dairy wastes was found to exceed the oxygen supply. This
caused a decrease in dissolved oxygen content of the material and a
simultaneous increase in the COD of the effluent. During the
dissimulation of the substrate, acid intermediates from carbohydrate
breakdown were excreted to the system, with sharp drop in pH. This
suggested that the oxidation of the carbohydrate fraction was limited
even though adequate amounts of oxygen were supplied. The decrease in
the pH preceded more slowly when the dissolved oxygen content approached
0, indicating that acid formation was due to activity of aerobic
organisms. From the results obtained, the conclusion was reached that
the dissimulation period for protein fraction preceded at a remarkably
lower rate than dissimulation of carbohydrate fashion. Typical data
are presented in Figure 3. In studying the metabolism of sludge
microorganisms, Adamse found that the drop in pH in the activated
sludge shortly after feeding was due to the accumulation of acetic acid
in the presence of adequate dissolved oxygen, and lactic acid
accumulated when dissolved oxygen was below 0.5 ppm. Both acids were
degraded after lactose had been exhausted, but lactic acid was more
readily oxidized than acetic acid. The apparent degradation rate of
dairy wastes was found optimal at pH 6.5 of the activated sludge
suspension. Of the lactose added to the activated sludge, approximately
10 percent was respired, up to 50 percent was accumulated in poly-
saccharides in the cell biomass and the remainder was used for the
synthesis of cell constitutent and partly accumulated as intermediates
in the sludge mass.
The difficulty in the treatment of whey appears to be related to the
inability of microorganisms to degrade whey protein and the ability
of these proteins to interact with the polysaccharide of the cell mass.
537
-------
I
o
fcE
-------
Incorporation of whey or whey wash water in the activated sludge treat-
ment plants have been known to cause excessive foaming. The foam
has been shown by our earlier studies to be a combination of microbial
cells and major whey protein B-lactoglobulin. Although the BOD/N
ratio of milk wastes approximate those desired (20:1), the available
nitrogen is considerably less because of the binding of the protein
to the cell mass and the inability of enzymes to readily degrade
material. Ammonium nitrogen has been frequently indicated as a
requirement in dairy food plant waste waters. However, many dairy
food plant waste treatment systems and some investigators have failed
to show an increased efficiency in BOD reduction through the addition
of nitrogen; and the exact reason for these observations needs further
investigation.
Waste Treatment Methods
The types of biological waste treatment methods utilized in the dairy
food industry are primarily activated sludge, trickling filters, aerated
lagbons, irrigation and a combination of these.
Characteristics with respect to loading and efficiency of BOD re-
duction are summarized for selected treatment systems in Table 17.
The data are not inclusive of those collected, but illustrate the
differences that have been reported. Text books and many authors
repeatedly state that dairy food plant wastes are easier to treat than
municipal wastes. By this, most mean that because of the relatively
low suspended solids, it is not necessary to go through an extensive
solids removal prior to biological oxidation. However, in practice,
it has been shown that the biological oxidation phase of waste treatment
is much more difficult than for municipal waste systems, primarily
because of the high organic load and because of the variability in
both hydraulic and organic loading. Efficiencies of biological
oxidation treatment systems for dairy wastes are generally reported to
be higher than 90 percent. In the site visits and from those studies
that have been exhaustively reported in the literature, it becomes
apparent that this degree of efficiency is achieved only for a part
of the operating time. Based on some twenty evaluations of treatment
plants for dairy wastes, it was found that the waste treatment
facilities were less than 70 percent efficient about 25 percent of the
time. Figure 4 shows typical frequency plots of the variability of
waste treatment for several systems treating only dairy wastes.
Because of the high BOD load, and in many modern plants up to 4000 ppm,
even 99 percent efficiency does not bring the BOD in the effluent down
to a level acceptable in some states for direct discharge to a stream.
As a result, an increasing number of plants that have their own
treatment facilities are combining two or more biological treatment
methods. It would appear that a two-stage or three-stage biological
oxidation is necessary in order to provide for optimum treatment to
meet stream standards.
539
-------
Table 17. Comparison of Waste Treatment Methods for Dairy Food Plant
BODc in ppm
Untreated
Method of Treatment Range
Ul
•p-
o
Activated Sludge
(10)
Activated Sludge
Trickling Filter
Aerate Lagoons
Irrigation6
84-2,920
180-15,000
130-32,000
75-10,000
_
Aver.
404
1903
2846
286
220
Treated
Range Aver.
7-632
1-1700
10-1400
25-800
_
85
163
81*
191
% Reduction
In BODS
Range
17-99
57-99.8
35-99.5
52-89.5
_
Aver.
79.08
88.7
77.4
76
Loading
Range
0.018-1.10
0.06—485
0.4—49.4
10-395
2500-90,000
Aver.
0.35a
137b
12. 7C
162d
27,700^
Retention Time
(hrs)
Range Aver
1.5-120 21.4
No information
estimate 36-48
168-1080 420
_
avail,
hrs.
(17 da
_
a.lb BOD/lb MLSS; blb BOD/100 cubic feet; clb BOD/cubic yard; dlb BOD/day/acre; eBOD loading generally not given,
the value of 220 is given as maximum loading rate recommended; ^volumetric loading in gallons/day.
-------
100
OQ
n>
4N
W H <
M 1-1 to
C CO >-!
O, pj H-
TO rt (to
-• (D H-
^00
H rt 3
Tl
Hi H-
II O 3
H rt
HOP'
H- to (0
O H-
?7* t~i trJ
I—l *- 3
M n
CD —^
^ !> O
v-x [-1 Hh
II O
H-
> r-h
n> HI
>-( ro
to >-(
it ro
(D 3
Q. rt
TO T)
O (T)
o en
3
... o
II CO
It
> n>
o
to
rt
ro
a-
90
80
o
ID
Q
LU
a: 70
60
50
% PROBABILITY
99
-------
The relative efficiency for BOD and COD treatment under commercial
dairy plant operations obtained from plants surveyed in this study are
presented in Table 18. Values show that in one dairy food plant waste
in which the BOD of the effluent was 15 ppa, the COD of the
effluent was 175 ppm. This may reflect the refractory nature of
certain detergents and chemicals used in the cleaning and sanitizing
of dairy food plant equipment.
Problems encountered in the treatment of dairy food plant wastes
include:
1. Periodic poor performance, and lack of correlation between loading
rate, sludge characteristics and BOD reduction.
2. Poor settling characteristics of sludge in activated sludge systems,
with problems of pin point floe, deflocculation and bulking.
3. Periodic loss of the biomass in trickling filter systems.
4. Low levels of available nutrients, caused by possible complexing
of milk constituents with mfcrobial cells, and possible feed back
inhibition of B-galactosidase by lactic acid.
5. Slow recovery of the biological activity of the treatment systems,
after shock loading, or periods of low feeding.
6. High organic concentrations, high hydraulic and nutrient variabil-
ity and consequent shifts in the mfcroflora of the waste treatment
facility.
7. Possible "toxicity" in some dairy wastes, attributable to sanitizers
and detergents.
8. Stable, high strength foams in activated sludge systems in the
presence of whey or often cottage cheese wash water in addition to
light froth foams associated with over-aeration.
9. Relatively high COD levels in treated waste waters, with BOD-COD
ratios of less than 0.1 in some cases.
10. Frequent clogging of diffuser-type areators and sludge breakup.
11. Poor sludge filtration characteristics.
542
-------
Table 18. BOD5/COD Relationships in Raw and Treated Dairy Plant
Wastes from Survey of Industrial Dairy Plants
Products
Raw Dairy Plant Wastes
After Primary Treatment
After Secondary Treatment
Manufactured
Milk
Cottage Cheese
Milk
Cottage Cheese
Milk
Yogurt
Cottage Cheese
Cottage Cheese
Milk
g Milk
Cottage Cheese
Ice Cream
Milk
Ice Cream
Butter
Cottage Cheese
Milk
Ice Cream
Cottage Cheese
Milk
Ice Cream
Cottage Cheese
Ice Cream
BODC
3300
3700
750
2863
3200
2168
3750
3200
140
COD
3379
7071
850
6500
9000
5300
9000
4150
350
BOD/COD
.97
.52
.88
.44
.41
.40
.41
.51
.40
BOD5 COD BOD/COD BOD5
145 400 .36 15
400 600 .67 80
266
145 430 .33 15
140 350 .40 35
150
35
COD
100
250
700
175
250
560
250
BOD/COD
.15
.32
.38
.08
.14
.26
.14
-------
REFERENCES
1. Adamse, A. D. 1967. Bacteriological studies on dairy waste
activated sludge. Meded. LandbHogesch. Wageningen, 66:(6)1-79.
2. Dairy Effluents Sub-Committee on the Milk and Milk Products Technical
Advisory Committee. 1969. Dairy effluents. Department of Agricul-
ture and Fisheries for Scotland.
3. Dias, F. F. and Bhat, J. V. 1964. Microbial exology of activated
sludge. I. Dominant bacteria. II. Bacteriophages, Bdellovibrio,
coliforms, and other organisms. Appl. Microbiol., 12:412.
4. Federal Water Pollution Control Administration. 1967. The cost of
clean water. Vol. Ill-Industrial Waste Profile No. 9-Dairies. Fed.
Wat. Poll. Cont. Admin., Publ. No. I. W. P.-9, U. S. Gov. Printing
Office, Washington, D.C.
5. Fisher, W. J. 1968. Treatment and disposal of dairy waste waters:
A review. Dairy Sci, Abstr., 30:(11)567-577.
6. Fritz, A. 1960. Determination of water pollution by dairy effluent.
Milchwissenschaft, 15: (12)609-612.
7. Hoover, S. R. 1953. Biochemical oxidation of dairy wastes. V. A.
review. Sewage Ind. Wastes, 25:(2)201-206.
8. Lawton, G. W., Breska, G., Engelbert, L. E., Rohlich, G. A., and
Forges, N. 1959. Spray irrigation of dairy wastes. Sewage Ind.
Wastes 31:923-933.
9. McKee, F. J. 1957. Dairy waste disposal by spray irrigation. Sewage
Ind. Wastes, 29:(2)157-164.
10. Pipes, W. 0. 1968. An atlas of activated sludge. Federal Water
Pollution Control Administration. U. S. Department of Interior.
11. Porges, N. 1958. Practical application of laboratory data to dairy
waste treatment. Fd. Technol., 12:(2)78-80.
12. Reynolds, D. J. 1966. Methods for estimating the strength of dairy
effluents. 17th Int. Dairy Congr., 5:773-780.
13. Sarkka, M., Nordlund, J., Pankakoski, M., and Heikonen, M. 1970.
Water pollution by Finish dairies. 18th Int. Dairy Congr., I - E,
A. 1.2 11.
14. Schulz-Falkenhain, H. 1963. Treatment of dairy waste waters in
oxidation ponds. Dte. Molk.-Ztg., 84:1403.
544
-------
15. Svoboda, M. 1966. On the problem of terminology: Ponds, tanks,
lagoons. Vodohospdarske TEI, 8:223.
16. Svoboda, M., and Salplachta, J. 1956. Some observations on the
purification of dairy effluent by Geotrichum candidum. Cal.
Mikrobiol., 1:(4)176-182.
17. Svoboda, M., Hlavka, M., Salplachta , J., and Stelcova, D. 1962.
Study on the quantity and pollution of dairy effluent in
Czechoslovakia. Sb. vys. Sk. Chem,-technol. Praze, Potravin
Technol., 6:(2)139-168.
18. Trebler, H. A., and Harding, H. G. 1949. United States trends in
disposal of dairy waste waters. 12th Int. Dairy Congr., 3:Sect, 3,
688-697.
19. United States Public Health Service. 1959. An industrial waste
guide to the milk processing industry. U. S. Public Health Service
Publ. No. 298.
20. Wallgren, K., Leesment, H., and Magnusson, F. 1967. Investigations
on irrigation with dairy waste water. Meddn. svenska Mejeriern. Rik-
sforen., 85:20.
21. Walzholz, G. 1967. Dairy effluent. Dte. Molk. Ztg., 88:(42)1722-1724,
(43)1778, (44)1811, (45)1849-1850, (46)1888-1889, (47)1926-1927, (48)
1960-1962.
22. Walzholz, G., and Pester, A. 1960. Engineering problems in the
collection and evaluation of dairy effluent. Dte. Molk.-Ztg.,
81:(51/52)1886-1889.
23. Zack, S. I. 1956. Trickling filter treatment of wastes at two milk
processing plants. Sewage Ind. Wastes, 28:1009-1019.
24. Zall, R. R., and Jordon, W. K. 1969. Monitoring milk plant waste
effluent a new tool for plant management. J. Milk Fd. Technol.,
32:(6)197-202.
545
-------
ANAEROBIC-AEROBIC PONDS FOR TREATMENT OF BEET SUGAR WASTES
by
William J. Oswald, Clarence G. Golueke, Robert C. Cooper, and
Ronald A. Tsugita*
This paper is designed primarily to review the most significant
results from the last two years of a four-year study of a pilot
scale ponding system for beet sugar flume water treatment at Tracy,
California. Results of the first two years of the study have been
presented previously in a technical paper^^ and in more detail in
the first progress reports.(2) The material to be presented herein
is drawn from a more recent technical paper'^' and from the final
engineering report for this study,^ ' in which the study and the
results are described in much greater detail than is possible in
this brief resume.
The system was constructed near the Holly Sugar Factory in Tracy
for the purposes of the study, and consisted of an influent pump, a
DSM screen, a 2-hour mud settling tank, an influent volume meter with
a series of three ponds, a recirculation system designed to draw water
from the third pond of the series and to inject it with the influent
at a metered rate, an effluent meter and pump. The three ponds
consisted of a 1-acre surface area unit with a maximum water depth of
14 feet, a 2-acre surface area unit with a maximum water depth of
7 feet, and a 3-acre surface area unit with a maximum water depth of
4 feet. Although their names did not always reflect such conditions,
the three ponds were referred to as "Anaerobic," "Facultative," and
"Aerobic" in the order described. Each ponding element was designed
so that the water depths could be decreased to about 2 feet below the
maximum so that a range of depths could be explored. The physical
appearance and arrangement of the system is shown in Figure 1. The
system was equipped for automatic monitoring of sunlight and air and
water temperature. A complete chemical laboratory was maintained at
the plant site for analytical work.
During the first two years of the study the three ponds — anaerobic,
facultative, aerobic — were operated in series with variable loading
and recirculation. The results indicated that most of the BOD was
removed in the primary (anaerobic) pond and that the additional ponds
added little to the BOD removal in spite of their greater surface
area. On the other hand, the primary anaerobic pond was always
malodorous, the odor increasing with increased loading between 500
*Respectively Professor of Sanitary Engineering, Research Biologist and
Associate Professor of Public Health, University of California,
Berkeley, and Senior Engineer, James M. Montgomery Consulting Engineers,
Lafayette, California.
547
-------
and 2,000 Ibs of BOD per acre per day. Recirculation from the
aerobic pond was found to be beneficial in decreasing odor and
improving treatment, providing it was not excessive. Recirculation
rates of three or four times the influent rate tended to shorten
detention times, force loading forward, and diminish the quality
of the final effluent from the system.
The major conclusions of the earlier study were that it is not
possible to operate a ponding system for flume water wastes without
odor in the absence of relatively large quantities of molecular
oxygen. Thus, systems involving aeration by mechanical aerators
or by photosynthetic oxygenation required investigation. Controllable
growth of algae in the aerobic pond essential to photosynthetic oxy-
genation is dependent on controlled mixing and a balanced nutrition
for the algae.
Despite generally favorable BOD removal in the system, the three-
pond system as it was applied in the first study would not be
suitable for a routine application to factory wastes simply because
waste treatment was not accomplished to the extent required for an
odor-free operation, or for an operation in which an effluent would
have to meet reasonably strict discharge requirements. The importance
of recirculation is primarily related to conveyance of oxygen,
"seed1,' and nutrients into the primary stages of the system; in di-
lution of the influent waste with recovered water; and, in forcing
stronger wastes forward in the system, thus possibly increasing the
efficiency of secondary units. The conclusion of the initial studies
was, however, that the optimum rate of recirculation changes with
changing conditions in the ponds, and that under some conditions,
rates in excess of one or two times the influent volume may be un-
desirable.
In summary, the work done previously has shown that substantial
reductions in solids, BOD, and nutrients may be attained in simple
ponds, but that land use, effluent BOD, and odor would be excessive.
On the other hand, the first series of studies showed that by using
an anaerobic pond in series with other ponds, land use could be de-
creased. However, effluent BOD and pond odor would still be excessive.
The specific purpose of this second study was to explore mechanisms
of aeration and to improve pond design and operation to increase the
rates of BOD and odor removal and to further explore methods of
decreasing land requirements. Another purpose of the work was to
derive design criteria for systems of ponds which would have predictably
satisfactory performance over a range of environmental conditions, and
which could therefore be applied to meet discharge specifications in
a variety of climates.
During the period reported in this paper, three types of systems were
studied: an anaerobic pond with aeration, an anaerobic pond with
548
-------
e
Q - 48,000 GPM
max
v
^
Ul .
-c- '
4 HIGH BAFFLES (See Fig. 3 for tietoils)
AEROBIC POND 1ST
FACULTATIVE POND 3K
NOT USED IN JULY-DEC. 1968
STUDIES
/!/
MIXING PUMPS/lx^
4-l2,OOOgpnr., I1 TDH
10 H.R eoch (See fig. 2 for defoils)
Clip
FLOATING
AERATOR
ANAEROBIC POND TI
OPERATING SEQUENCE : Aog 1967- H only
Foil 1967 -H plus
Spring 1968- H plus Ttr
Foil 1963- H plus IE
^
^
FIGURE I. SHOV/ING LAYOUT OF PONDS, ARRANGEMENT OF BAFFLES AND MIXING
PUMPS IN AEROBIC POND
-------
aeration in series with a facultative pond, and an anaerobic pond
with aeration in series with a photosynthetic algae pond. Analytical
work in the study was quite detailed. In addition to measuring and
maintaining records of pond depth, detention period, recirculation,
and mixing, the studies also included measurement of air and water
temperatures and light energy input, as well as analyses to determine
BOD, COD, total and volatile suspended and dissolved solids, nitrogen,
phosphorus, magnesium, calcium, sulfate, and sulfide, algae and other
organisms, and algal packed volume. Routine analyses were made
according to techniques given in Volume 12 of Standard Methods for
Examination of Water and Waste-Water,(5) and modified to meet needs
peculiar to the material being tested in the study. Gas production
was studied with Bronson(6) gas collectors (shown in Figure 2) . Data
on methane fermentation were limited to the final series, both
because a clearcut methane fermentation did not become established in
the ponds, and because gas emission measurements were interrupted by
water condensation in the transmission lines from the Bronson collectors
to the gas meters in all but the final runs. Although a standard
method is available for odor evaluation, this was deemed too time-
consuming, and a simplified "direct sniff" method was developed for
expressing odor. This was done by establishing six arbitrarily
designated characteristic odors: 1 - none; 2 - beets; 3 - other;
A - cow dung; 5 - I^S; and 6 - foul. Also, six intensities were
arbitrarily estimated: 1 - being a very low intensity and 6 - being
a high intensity. The arbitrarily assigned numbers were multiplied
together to give an "odor product." The lowest odor product possible
is 1 and the highest is 36. A sample which smelled like l^S and was
moderately strong would have an odor product of 5 x 3 or 15. A strong
beet smell would have a product of 2 x 6 = 12, and so on.
Nutrient addition to the ponds was studied, but the results were
inconclusive.
During the course of the studies, several 24-hour studies were
conducted to obtain information on the diurnal rates of oxidation
and photosynthesis in the ponds.
In the original conceptual design of this project, it was an
objective that the system should be operated under a given set of
experimental variables for a period of time until steady state could
be attained. One of the annoying problems in the study which seemed
to interfere with this objective was a lack of flow control due to
persistent clogging and other problems with the feed system. In
retrospect after viewing the completed data one realizes that because
of load variation hopes of attaining a high degree of control and
for reaching a steady state were unrealistic. The fact is that,
because of drastic but apparently uncontrollable momentary daily
and seasonal changes in factory waste strength no steady state could
have been attained in these studies regardless of flow control, nor
can one expect a true steady state to ever be maintained in a plant
processing the wastes of a beet sugar factory.
550
-------
t_n
P.S.
SCR. HALF COUPLINGS
(PLUGGED)
18 GA. SKIRT
(GALVANIZED)
4- EXTRA HEAVY HEX NUT
SCH. 40 GALV.
STL. PIPE
MAKE GAS TIGHT-W£t,
OR SOLDERED CONSTRUCTION
BREAK ACROSS FLATS
FOR STIFFENING
SCH. 80 STL.
PIPE(GALV. AFTER
FABRICATION)
4 - -5- J- BOLTS
TO MATCH SUPPORT
|- t HOLE
CONCRETE 1
<--j- STL. PLATE
(GALVANIZED)
FIGURE 2
BRONSON GAS COLLECTOR
1
"o
1
K)
I
r.
,
4 — .-if
3'_0" J
X4
-------
EXPERIMENTAL
As noted previously the experimental work involved a study of the waste
assimilated by the anaerobic pond with a 5 hp floating aerator; an
appraisal of the combined action of the aerated anaerobic and facul-
tative pond in series; a study of the oxygenation capacity of the
algae pond without mixing; and the combined action of the aerated
anaerobic and mixed algae pond in series. During the series studies,
it was intended to operate the system with a continuous feed rate and
a continuous recycle rate; but as noted above, because of interruption
in flow and changes in waste strength, it was only possible to meas-
ure the flows and loads as they occurred, and loading rates were
computed on the basis of such flow and BOD data for arbitrarily spe-
cified monthly time periods. These arbitrary loadings, together
with other specified conditions, are presented in each section together
with the results.
Before reviewing the process results, it is worthwhile to present
the available data on the mean value of all of the biochemical
parameters studied for Tracy flume water over the entire four-year
period of the study, 1965 to 1968 inclusive. The detailed analytical
information accumulated on waste composition during the period was
averaged, and the averages tabulated in Table 1. Variance was also
determined for some of the parameters. As is evident from the table,
variance (*"/M) was often in excess of 1000 percent, and this degree
of variance is evidently one of the factors which has led to serious
difficulties in the treatment of beet sugar factory wastes. In
addition to the large random momentary variation and daily variations
in waste strength and composition, the data evidenced an even stronger
seasonal variation which will become evident to the reader when results
are presented for parameters such as COD and nutrients.
Aeration in the Anaerobic Pond
An anaerobic pond is normally designed to avoid the intrusion of
oxygen, but the severe odors in the anaerobic pond encountered
during the earlier studies led us to explore the use of a surface
aerator for odor control. The hope was that surface aeration would not
interfere with methane fermentation. The aerator was a 5 hp
floating unit''' located in the anaerobic pond, as shown in Figure 1.
During a prior series of tests the anaerobic pond had been heavily
loaded; and following several weeks without influent, the residual
BOD was 330 mg/1 and the pond void of dissolved oxygen. Prior to
the study, the odor intensity level of the anaerobic pond was above
12, and was characterized by a foul hydrogen sulfide stench that
had barely improved after standing several weeks. As indicated
graphically in Figure 3, after three days of aeration with the new
aerator, the odor intensity dropped to a level of 4. The odor
quality became that characterized as "cow dung." By the seventh
day, the odor intensity level had dropped to 1.0, which was the
lowest value on the arbitrary scale used.
552
-------
TABLE I
Mean Values for Certain Analytical Parameters of Holly-
Tracy Flume Water - 1965 - 1968
(Following 16 mesh DSM screening and 1,5 hr. sedimentation)
Parameter
Total Nitrogen (N)
Ammonia (N)
Nitrate (N)
Chlorides
Sulfate
Alkalinity (CaC03)
Sulfide
Phosphate (P)
Calcium
tlagr.cciurr,
Sodium
Potassium
BOD (unfilt.)
COD (unfilt.)
COD (filt.)
Suspended Solids
Suspended Volatile
Suspended Ash
Dissolved Solids
Dissolved Volatile
Dissolved Ash
Total Solids
Total Volatile
Total Ash
Sugar
Dissolved Oxygen
Physical factors
pH
Light Penetration
Specific Conductance ji mhos .
Units
mg/1
it
•»
n
n
it
ii
tt
ii
n
n
n
ii
n
u
n
ii
»»
It
II
tl
II
II
l(
II
II
CO
cm
Value
16. A
6.3
2.6
400 *
210.0
538.0***
0.68
3.4
178.0
66.0
222.0**
88,0**
930.0
1601.0
1195.0
1015
360
655
2209
1139
1070
3224
1499
1725
1.25
0.0
7.06
45.6
300
Var i ance
«•••
«••*
*»•
•»••
•• V
••M
•»••
—
--
a D
«••
«••
—
—
--
11269
7140
6348
6016
2764
7754
14979
4456
8328
0.25
1.8
17.8
299
*Based on specific conductivity
**Single values
***By difference
553
-------
The COD and BOD results for this aeration study shown graphically
in Figure 3 indicate that in the presence of oxygen, oxidation of
substances occurred rapidly in the system. As is evident from the
graph, the BOD which was initially 330 mg/1 declined to 18 mg/1 on
August 28; whereas the COD steadily increased from 390 to 590 mg/1
during the first four days, and then declined to 134 mg/1 by August
28. The initial increase in COD was apparently due to the mixing
pump bringing into suspension colloidal material which exerted a
COD but not a BOD. A COD of 100 probably approaches the lowest
COD likely to be encountered with stabilized beet sugar waste.
During the period August 1 to August 16, no dissolved oxygen was
detected in the system. After August 19, dissolved oxygen began to
appear in the water during the afternoons, although it would be at
zero concentration in the mornings. This indicates that after
August 19 photosynthetic activity was beginning to replenish oxygen
in the system even while oxygen was being introduced by the aerators.
During the period August 1 to August 16, the BOD decreased from
335 mg/1 to 100 mg/1. Thus in 15 days, the total reduction was 235
mg/1. Inasmuch as during this period the volume of the anaerobic
pond was about 2.5 million gallons, the total amount of BOD oxidized
was 235 x 2.5 x 8.34 = 4,900 Ibs, or 327 Ibs per day. Assuming
that during this period natural re-aeration contributed 20 Ibs per
day (the surface area being 1 acre), the aerator must have contributed
307 Ibs per day, or 2.54 Ibs per hp-hr. This rate is precisely that
published by the manufacturer^) — namely, 3.2 Ibs of 62 per kw. hr,
at 0 dissolved oxygen.
The rate of change of COD was somewhat higher after the first five
days, but the overall rate, neglecting the initial "hump", corresponds
well with the change in BOD. The initial hump probably resulted from
the disturbance of bottom substances which had a COD but little
measurable BOD. It should be noted that about 20 days were required
to satisfy the pond BOD and attain free molecular oxygen, and that
the odor level had dropped to 1 when about one-third of the time had
elapsed. This corresponded to the satisfaction of about one-third
of the BOD.
During the period of mechanical aeration without loading, the pH
in the pond slowly decreased from 7.6 to 7.1 between August 1 to
August 10, and then increased to 7.9 in the period August 11 to
August 18. The increase probably was due to photosynthetic activity.
Aerated Anaerobic and Facultative Pond in Series
In the next series of experiments, the aerated anaerobic and
facultative ponds were operated in series. Key conditions and re-
sults from this experimental series are summarized in Table 2. No
recirculation was applied. Application of waste to the anaerobic
pond was at the initial rate of 25 gpm. The BOD of the influent
ranged from 1308 to 1639 ppm. During several periods in these
554
-------
eo
o>
E
X
Q
O
CO
x
O
O
O
70
60
50
40.
30
20
10
I I I I I I
5H.R Flooting surfoce cerotor
furnished by Welles Products
Corporation- Roscoe , Illinois
I I
•_» Q
I I
- COD
a—
BOD
DISSOLVED OXYGEN
13 15 17 19
AUGUST, 1967
21
c»
20 .
§
5?
X
O
o£
>
O
5
to
25 27 29 31
FIGURE 3 CHANGE IN COD, BOD, AND D.O. IN ANAEROBIC POND DURING
SUSTAINED AERATION WITH 5 H.R FLOATING SURFACE AERATOR
555
-------
experiments, the waste had a strong foul odor, a brown color, and
a nitrogen content as high as 64 mg/1 — indicating the intrusion
of Steffen waste into the flume water. As the experiments progressed
the effluent BOD from the anaerobic pond began to increase, and
rose from 18 to 53 mg/1 within a week. It reached 129 mg/1 by the
end of the 25 gpm run. During the same time, the COD rose from 134
to 228 mg/1. Odor at first increased and then declined. At the
feed rate of 25 gpm, the average organic load entering the anerobic
pond was estimated to be between 400 and 500 Ib/day/acre, exceeding
by 75 to 175 Ibs the daily aeration capacity of the 5 hp floating
aerator. Frequent interruptions in flow due to influent pumping
failures at first permitted the aerator to maintain a small residual
of dissolved oxygen in the surface layers of the anaerobic pond, and
kept the odor intensity level down to about 5. However, when the
loading rate was increased to 50 gpm, all dissolved oxygen disappeared
from the system and the odor level in the anaerobic pond rose to 9.
According to Table 2, loadings varied from 900 to 0 Ibs/day/acre.
Generally speaking, most of the BOD was removed in the anaerobic
pond and odor levels in the anaerobic pond were not as severe as
they had been without aeration.
Because of the high BOD removals in the anaerobic pond, loadings
to the facultative pond were always less than 180 Ibs per day.
Effluents from the facultative pond varied from 47 to 140 mg/1 BOD,
with little apparent relationship to applied BOD. Odor levels in
the facultative pond were always 3 or less. During this period,
the facultative pond contained a rich culture of the blue green algae
Oscillatoria limosa which remained in suspension, and apparently in
some cases retained or possibly fixed nitrogen to the extent that
nitrogen concentration in the facultative pond was higher than those
in the anaerobic pond. 0. limosa concentration in the facultative
pond reached concentrations approaching 100 mg/1. Nitrogen concen-
trations in the anaerobic pond were always less than those in the
influent — a fact which confirms the high nitrogen removals in the
anaerobic pond reported in the Progress Report of earlier work.^'
The degree of nitrogen removal in the anaerobic pond was, however,
somewhat less than had occurred previous to the introduction of the
surface aerator.
Photosynthesis and Respiration
Prior to the activation of a mixing system in the aerobic pond, a
study was made to determine the rates of photosynthesis and oxi-
dation in the unmixed pond. This determination was made by means of
a 24-hour study. Such 24-hour studies of dissolved oxygen and
temperature may be extremely valuable in systems undergoing oxi-
dation and photosynthetic oxygenation because the data may permit an
evaluation of the in situ rates of oxygen use and the d.n situ rates
of oxygen production. No dark bottles or other devices are needed,
with no disturbance of the system. One requirement for the success
of such a 24-hour study is the attainment of oxygen supersaturation
556
-------
TABLE 2
RESULTS
1967 Aerated Anaerobic Pond Plus Facultative Pond With No Recirculation
Pond
Inf
An
Fac
Inf
An
Fac
Inf
An
Fac
Inf
An
Fac
Inf
An
Fac
Inf
An
Fac
Run
13
it
n
14
i»
(i
15
-------
at some point in time. The method is based on the fact that at the
points in time that the dissolved oxygen concentration of the pond
is equal to the saturation concentration for that temperature, no
exchange of oxygen gas can occur with the atmosphere and consequently
the observed rates of change in oxygen concentration are independent
of atmospheric gas exchange, and are a function solely of the diff-
erence between the rates of photsynthetic oxygen production and
microbiological respiration. In spite of several efforts, super-
saturation was reached in only one 24-hour study, but this was
sufficient to permit the desired determinations. During that study the
applied BOD load to the aerobic pond was about 60 Ibs/day/acre, and the
solar energy input was about 200 calories per cm2 per day.
Results of that run are plotted in Figure 4. From the figure it
may be observed that the exchange independent rate of respiration
was 0.467 mg 02/liter per hour, and the mean photosynthetic rate at
the surface netted 1.16 mg/l/hr. The gross surface photosynthetic
rate including respiration was therefore 1.62 mg/l/hr. From these
data the rate of oxygen use by the pond was 86 Ibs/day/acre. The
net surface photosynthetic oxygen production was 214 Ibs/day/acre,
and the gross surface rate 300 Ibs/day/acre. Inasmuch as, due to
light absorption, the photosynthetic rate decays logrithmically
with depth, the average rate with depth would be about one-third the
surface rate or about 100 Ibs/day/acre. The apparent photosynthetic
efficiency of the algal pond, assuming a sunlight energy input of
200 cal/cm^ per day, was then about 2.3 percent. As is to be expected,
this rate of oxygenation was slightly greater than the rate of
deoxygenation based on the respiration rate, a fact that is also
evidenced by the observation that the oxygen level was continuously
near saturation. One may further conclude that inasmuch as the system
was supersaturated during daylight hours and barely fell below
saturation at night, there must have been a net export of 02 from the
system, indicating that it was underloaded for the conditions obtaining.
A corollary is that a BOD loading of 100 Ibs/day/acre could have been
applied at this time with no deterioration of the system. This
rate, however, probably exceeds the maximum safe design, particularly
in view of the fact that beet sugar wastes frequently contain pigments
and colloids which interfere with the penetration of light.
Aerated Anaerobic Pond in Series with a Mixed Aerobic Pond
In spite of the fact that photosynthetic oxygenation can frequently
provide 100 Ibs of oxygen or more per acre per day in an unmixed
pond and when sunlight, temperature, and waste quality permit, there
are many times when the ponds become stratified, sunlight is
diminished, or the wastes highly pigmented, and photosynthesis cannot
meet the oxygen demand of a heavily loaded system. Under such
conditions a supplementary method of oxygenation is essential if
odors are to be avoided and a good quality effluent to be obtained.
558
-------
SATURATION D.O.
RESPIRATION RATE
AO/At = -'0.467 mg///hr
\
MEAN PHOTOSYNTHETIC RATE
(NET) AO/At =+ 1.16 mg///hr "
POND 3 , 3'DEPTH,
3.1 ACRES, 2.834x10 go!.
I2N
4PM 8PM I2M 4AM
TIME - 20-21 APRIL 1967
8AM
I2N
FIGURE 4 DISSOLVED OXYGEN VS TIME OF DAY
55?
-------
The simplest form of supplementary aeration, and one that is highly
compatible with photosynthetic oxygenation, is a flow-mixing system.
Such a system maintains algae and their nutrients in suspension and
through eddy diffusion provides surface oxygenation when photo-
synthesis is inactive. As noted previously, the final series of
experiments reported herein involved such a flow-mixing system in
series with the surface aerated anaerobic pond. As was shown in
Figure 1, the pond was equipped with flow diversion baffles and flow
was induced by means of four low head-high volume propeller pumps.
These pumps were fabricated by the University shopsmiths from components
readily available on the market. A schematic diagram of the pumps is
shown in Figure 5, and of the entire system is shown in Figure 6.
The system shown was operated on a sustained basis for the entire five
months of the 1968 Fall campaign. Climatological conditions are
summarized in Figure 7, and flow loading and BOD removal conditions
for the system are shown in Table 3. The letters I, E, R, and T
refer to the flow lines shown in Figure 6; the letters An and Al refer
to the anaerobic and algae ponds.
Evidence presented in Table 3 indicates that the overall system
attained BOD removal efficiencies varying from 84 percent to 98
percent. Overall COD and BOD values for the system are presented
in Table 4 as a function of pond and month. Unfiltered BOD values
are plotted in Figure 8. As is evident in the figure, 80 percent or
more BOD removal was always attained. Values for the pH of the waste
and anaerobic and aerobic pond effluents are shown in Figure 9. The
Figure evidences the extreme variation of over 5 pH units observed
for beet sugar flume water.
The aerobic pond was subject to considerable variation in dissolved
oxygen, as is evidenced by the P.M. dissolved oxygen levels plotted
in Figure 10. Because of the presence of dissolved oxygen,the aerobic
pond never had an objectionable odor.
Odors were, however, present in the anaerobic pond in spite of the
aeration applied. That certain of these odors resulted from the
reduction of sulfate, is evidenced by the sulfate value as a function
of pond and month, shown in Figure 1L. Sulfate reduction was directly
related to water temperature, since its rate decreased during the
Fall. The relationship between sulfate reduction and temperature
observed may be expressed as R% = 55 + 2(T-10)(D in which R is the
percent reduction in sulfate and T is the temperature in °C. A
statistical analysis indicated that much of the sulfide released as
a result of sulfate reduction combines with magnesium. This is
probably an organic complex of magnesium sulfide. It characteristically
imparts the dark ink-like color typical of anaerobic beet sugar waste.
Inasmuch as there was evidence of a clear relationship between loading
and odor, an effort was made to correlate the two. The results of this
correlation are shown in Figure 12. The available evidence (with two
exceptions) indicates a positive and linear relationship between odor
560
-------
to
6"
4 -18"PROPELLER PUMPS
0 = 12,000 GPM @ AH = I FT
TOTAL 0 : 48,000 GPM
^x-^7
SCALE :
--|ro
w
^ s
MOTOR = 10 H. P.
900-1000 RPM
DECK 2"x4"
/
CORRUGATED IRON
SHEETS
^
FLANGE
6"x6"
POST
4'DIAMETER x 4' PIPE
HOT DIPPED
FIGURE 5 MIXING PUMP INSTALLATION FOR ALGAE POND-TRACY
561
-------
-RECYCLE PUMPS
EXCESS
WASTE
ANAEROBIC
POND
(AN)
rr
-FEED PUMP
SHUNT TO 8SDF SYSTEM
<-
FACTORY WASTE
TO FACTORY PONDS
AEROBIC POND (AO.
MIXING PUMPS
EFFLUENT
PUMP
FIGURE 6 SCHEMATIC DIAGRAM OF SYSTEM AS APPLIED IN AERATED ANAEROBIC PLUS
MIXED ALGAE POND SERIES
-------
MEAN MAX. AIR TEMP
MEAN WIN. AIR TEMP
AUG
SEP
OCT
MONTH
NOV
DEC
FIGURE 7 MEAN VISIBLE SOLAR ENERGY AND SURFACE WATER
TEMPERATURE , MAXIMUM AND MINIMUM AIR TEMPERATURE
AS A FUNCTION OF MONTH
563
-------
TABLE 3
Monthly Mean Flows BOD Values and Performance Data for
Anaerobic-Algae Pond in Series
Month 1967
I* Flow GPM (mean)
BOD of I mg/1 (mean)
I Load Ibs/day
Ibs/acre/day
R* Flow GPM (mean)
BOD of R mg/1 (mean)
R Load Ibs
An* Load Ibs/day
influent + recycle
T* Flow GPM (mean)
BOD of T mg/1 (mean)
Al* Load Ibs/day
Ibs/acre/day
An Removal Ibs/day
E* Flow GPM (mean)
BOD of E mg/1 (mean)
Load Disch Ibs/day
E * R Ibs/day
Al Rem Ibs/day
Ibs/acre/day
BOD Rem Eff An 7.
BOD Rem Eff Al %
Overall Eff % of
BOD Removal
Jul
88
1180
870
•i
•»•»
73
•»M
__
«
151
•••t
• ••
•»«•
73
WM
•»•»
•••
—
•••
Aug
105
1440
1810
it
148
142
255
2060
100
268
322
107
1738
22
142
38
293
29
10
95.5
9.0
98
Sep
75
1260
1134
t«
148
191
341
1475
183
245
538
179
937
90
191
207
548
-10
- 3
64
-1.8
84
Oct
128
915
1406
it
148
82
147
1553
195
133
313
104
1240
76
82
75
222
91
30
80
29
95
Nov
17
1420
290
n
23
81
22
312
26
176
54
18
258
33
81
32
54
00
00
83
00
89
Dec
82
1410
1380
n
82
85
83
146?
113
114
156
52
1307
36
85
37
120
36
12
89
23
98
*Latters Identified in Figure 6
N.B. All BOD values arc ultimate BOD's, I.e. 5-day BOD/.684
564
-------
TABLE 4
Summary of Monthly Mean COD and BOD Values
as a Function of Pond and Month
Month
Jul1
Aug
Sept
Oct
Nov
Dec
Samp
Inf
An
Al
Inf
An
Al
' Inf
An
Al
Inf
An
Al
Inf
An
Al
Inf
An
Al
Unfilt
COD
mg/1
1482
261
150
1450
389
314
1559
630
544
1570
750
656
1960
715
610
2380
759
579
Flit
COD
mg/1
603
119
142
961
217
112
1128
144
116
1180
137
104
1541
209
150
1969
165
130
COD
BOD
Unfilt
1.85
2.54
3.00
1.48
2.11
3.20
1.81
3.78
4.17
2.50
8.25
11.90
2.01
6.0
11.0
2.47
9. SO
10.00
Unfiltered ECO
5 -day 20° C
ng/1
803
103
50
985
184
98
863
167
130
6V 5
91
56
971
119
55
963
78
58
Ult 20°C
mg/1
1180
151
73
1440
269
143
1260
245
191
915
133
82
1420
174
81
1410
114
85
Ult 20°C
lbs/1000 gal
9.85
1.61
0.61
12.00
2.24
1.19
10.50
2.04
1.59
7.60
1.11
0.68
11.80°
1.45
0.67
11.75
0.95
0.70
Flit
BOD
mg/1
132
103
Values for July are based on single or duplicate tests only
^Single sample
•*M»an of four weekly samples
565
-------
MOO
I2OO
IOOO
o eoo
<
to
o
UJ
xa
> 600
O
40O
200
UNFILTERED BOD 5 DAY 20°C VALUES
INFLUENT
A ANAEROBIC
D ALGAE
A'
D-
•D..
.0-
•A-
•D-
A.
Q.
JUL
AUG
SEP OCT
MONTH
NOV
DEC
FIGURES. MONTHLY MEAN VALUES FOR UNFILTERED BOD AS A
FUNCTION OF POND AND MONTH
566
-------
12
10
KEY
FEED ANAEROBIC NALGAE
POND POND
AUG
SEP
OCT
MONTH
NOV
DEC
FIGURE 9. EXTREMES AND CENTRAL TENDENCY pH RELATIONSHIP
IN INFLUENT, ANAEROBIC AND ALGAE POND AS A
FUNCTION OF MONTH
567
-------
16
14
12
^5
o»
e
sr'°
a:
K)
5
z
LJ
O 8
^
~
X
O
HSSOLVED
o
LJ
4
2
O
—
[
-
1
]
\
\
\
\
\
\
\
\
\
\
\
L \
\
\
\
\
I
1
1
^
"""" ~*(
1
1 1 [
x
"N
1
c
/
/
/
/
/
/
/
/
/
f
/
/
/
/
]
1
\
\
\
\
\
\
\
\
I
I 1
-
_
—
—
JUL
AUG
SEP OCT
MONTH
NOV
DEC
FIGURE 10.MONTHLY MEAN , MAXIMUM, AND MINIMUM DISSOLVED
OXYGEN VALUES FOR ALGAE PONDS DAILY AT 3RM.
568
-------
400
300
o>
E
•h
u
200
u.
100
INFLUENT
A ANAEROBIC
D ALGAE
(I) D.
(I) A.
v
«
^ \
N \
x
•D-
(I) SINGLE VALUE ONLY
A
A
* D
A
JUL
AUG
SEP OCT
MONTH
NOV
DEC
FIGURE 11. MONTHLY MEAN VALUES FOR SULFATE AS A FUNCTION
OF POND AND MONTH
569
-------
2000
-------
product and loading, the approximate relationship being that
OP = 3.5 + .00531A2) in which OP is the odor product and L is the
BOD loading in Ibs/day/acre. As an example, for a BOD loading of
1,000 Ibs/day/acre, the indicated odor product would be 8.8.
Monthly mean value for the unfiltered COD of the waste is shown in
Figure 13, and for filtered COD in Figure 14. It is evident from
the figures that in spite of the fact that the filtered COD varied
from 500 to almost 2,000 mg/1, the effluent COD remained relatively
low. The indication is that as the season progressed, while a greater
and greater fraction of the organic load was converted from insoluble
to soluble form; the insoluble form increased in the ponds, probably
reflecting deterioration of the beets with aging in the Fall and
the growth of increased amounts of microbial material in the ponds.
Nitrogen in the waste similarly increased. A summary of monthly
mean nitrogen values is shown in Table 5. Of interest is the fact
that nitrate removal in the anaerobic pond was consistently high,
whereas ammonia nitrogen removal declined with temperature and an
increased fraction of the nitrogen became incorporated in organic
matter in the anaerobic and algae ponds as temperature decreased.
The indication is that the nitrogen removal mechanisms in the anaerobic
pond are primarily microbial in nature.
Nutrient relationships for carbon, nitrogen, and phosphate in the
system are shown in Table 6. The evidence is that carbon/nitrogen
ratios were too high for ideal microbial growth during most of the
time, and that nitrogen as well as phosphate addition would have
benefitted the system. The evidence is, however, that nutrients
should not be added to the anaerobic pond, since both nitrate and
phosphate are apparently to some extent reduced along with sulfate.
The corollary is that any nutrient addition should be made in the
aerobic pond where they would become incorporated in algal cells and
through recirculation returned to the anaerobic pond in the form of
organic nitrogen.
The quantity of algae in the ponding system was estimated on the basis
of centrifuged packed solids, and the results of these determinations
are shown in Figure 15. All solids data is shown in Table 7. It is
apparent from Figure 15 that the maximum bloom of about 280 mg/1 of
algae occurred in the algae pond during September, after which it
declined. On the other hand, the concentrations of algae in the
anaerobic pond steadily increased, possibly reflecting decreased
breakdown in the anaerobic pond at lower temperatures or some con-
centration phenomenon. During September and October there was a serious
depletion in algae concentration due to invasions of the algae pond
by rotifera and daphnia. These incursions caused a depletion of
dissolved oxygen in the system, and in one case apparently triggered
a bloom of purple sulfur bacteria which in September briefly replaced
algae as the principal, photosynthetic organism in the system.
571
-------
3000
2000
9
E
2
LU
Q
Z
LU
CD
X
o
IOOO
UNFILTERED COD
INFLUENT
.£, ANAEROBIC
•D ALGAE
-A-
A-
^>°-
A
D
-A'..
D
JUL
AUG
SEP OCT
MONTH
MOV
DEC
FIGURE 13. MONTHLY MEAN VALUES FOR UNFILTERED COD AS A
FUNCTION OF POND AND MONTH
572
-------
300O
2000
^
o>
E
•»
O
H
-------
TABLE 5
Summary of Monthly Mean NltroRon Values as a Function of
Species, Pond and Month -All Values Mg Per Liter as N
Month
Jul2
Aug
Sep
Oct
Nov
Dec
Sample
In
An
Al
In
An
Al
In
An
Al
In
An
Ai
In
An
Al
In
An
Al
N03"
2.3
0.2
0.2
2.97
0.36
0.42
1.45
0.42
0.40
3.20
0.51
0.50
4.31
0.48
0.55
3.14
0.45
0.67
NH4*
2.0
0.2
0.1
1.64
1.18
0.68
1.17
0.58
0.30
1.38
0.79
0.54
2.76
1,6?
1.31
1.54
1.30
1.10
N03" + NH4*
4.30
0.40
0.30
4.61
1.54
1.10
2.62
1.00
0.70
4.58
1.30
1.04
7.07
2.11
1.86
4.68
1.75
1.77
Total N
5.25
1.40
0.60
5.60
2.60
1.81
4.85
3.78
4.23
4.77
4.22
4.30
9.28
6.21
5.44
7.53
5.77
6.25
1
Diff
0.95
1.00
0.30
0.99
1.06
0.71
2.23
2.78
3.53
0.19
2.92
3.26
2.21
4.10
3.58
2.85
4.02
4.48
The difference is assumed to be organic N plus nitrite
the latter usually being negligible in magnitude.
July values are singles or duplicates only, all others
are means of 10 - 20 values.
574
-------
TABLE 6
Nutrient Relationships
Month
Aug
Sep
Oct
Nov
Dec
Sample
In
An
Al
In
An
Al
In
An
Al
In
An
Al
In
An
Al
Flit
COD
mg/1
961
217
112
1128
144
116
1180
137
104
1541
209
150
1969
165
130
C1
mg/1
404
91
47
474
60
49
495
57
44
647
88
63
827
69
55
N2
mg/1
5.6
2.6
1.8
4.8
3.8
4.2
4.8
4.2
4.3
9.3
6.2
5.4
7.5
5.8
6.2
C/N
72
35
26
99
16
12
103
14
10
70
14
12
110
12
9
P3
mg/1
1.5
13.9
7.3
0.78
2.04
1.57
0.79
1.25
0.54
1,53
1.43
9.30
0.57
0.43
7.10
10.42 x Filt COD
2
Total Nitrogen in unfiltered samples. No extra
nitrogen was added during the course of these experiments.
^
P was added at the approximate average rate of 5 ing/liter.
Because of poor flow control, phosphorus levels vere sporadic
but in excess of normal requirements.
575
-------
OC
o
x
O
o
in
CO
Q
_J
o
to
O
UJ
O
u_
o:
UJ
o
u.
o
UJ
o
U)
o
2
3.0
2.0
A ANAEROBIC POND
D-—• ALGAE. POND
TO ESTIMATE DRY WEIGHT
IN nig/lifer , MULTIPLY
PACKED VOLUME (ml/lifer) x 140
D
\
A
AUG
SEP
OCT
MONTH
NOV
DEC
FIGURE 15. MONTHLY MEAN PACKED VOLUME OF CENTRIFUGED
ALGAL SOLIDS AS A FUNCTION OF MONTH AND POND
576
-------
TABLE 7
Summary of Solids Data
Monthly Mean Dissolved Solids mg For Liter
Month
Jul
Aug
Sep
Oct
Nov
Dec
Volatile
In
975
1358
1150
1250
1513
2142
An
333
520
477
608
557
606
Al
268
391
441
528
507
523
Ash
In
1039
1128
1118
946
1215
1129
An
1164
1002
922
831
898
884
Al
1149
1099
978
930
935
954
Total
In
2014
2486
2268
2196
2728
3271
An
1497
1522
1399
1439
1455
1490
Al
1417
1490
1419
1458
1442
1477
Monthly Mean Suspended Solids mg Per Liter
Month
Jul
Aug
Sep
Oct
Nov
Dec
Volatile
In
105
241
378
199
199
327
An
84
128
376
452
413
492
Al
94
156
374
414
369
395
Ash
In
148
239
465
447
212
472
An
19
25
82
155
79
152
Al
20
27
88
131
75
98
Total
In
253
480
843
646
411
799
An
103
153
458
607
492
644
Al
114
183
462
545
444
493
Monthly Mean Total Solids mg Per Liter
Mont a
Jul
Aug
Sep
Oct
Nov
Dec
Volatile
In
1080
1599
1528
1449
1712
2469
An
417
648
853
1060
970
1098
Al
362
547
815
942
876
918
Ash
In
1187
1367
1583
1393
1427
1611
An
1183
1027
1004
986
977
1036
Al
1169
1126
1066
1061
1010
1052
Total
In
2267
2966
3111
28^2
3139
4070
An
1600
1675
1857
2046
1947
2134
Al
1531
1673
1881
2003
1886
1970
577
-------
Results of the study on gas production are presented in Figure 16.
As the Figure shows, no gas production occurred below 13.5°C, and
some form of inhibition of gas production occurred above 17°C. The
nature of this inhibition is unknown, but it obviously must have
severely influenced the nature of BOD removal and odor production in
the anerobic pond.
DISCUSSION
The results of the studies reported herein show very clearly the
extensive variation in waste strength, ph, and nutritional characteristics
of beet sugar factory wastes. The seasonal variation in waste strength
are so large that it would be impossible to design an effective
unbuffered, short detention waste treatment system for steady-state
or average conditions. The degree of variability is apparent when
one examines the unfiltered COD data (reference 4 Appendix 3A) in
which values vary from 550 mg/1 in August to 3643 mg/1 in November,
and 3380 mg/1 in December; and when one considers the influent pH
values shown in Figure 9 which vary from 5.7 to 11.6.
No short-detention period biological system lacking in buffer capacity
could withstand sudden shifts in nutrient and pH of the magnitude
found without violent upsets or complete failure in essential microbial
growth. High-rate ponds could tolerate the changes in pH, but could
not tolerate the changes in loading. Activated sludge units could not
tolerate either the change in pH or the change in loading. Trickling
filter units could perhaps tolerate the variable loading but could not
tolerate the variable pH. Thus, activated sludge units, high-rate
ponds, and trickling filters if considered in the design of primary
units of systems for factory waste would have to be designed for that
loading which could not be exceeded at least 95 percent of the time.
They would have to be continuously monitored and protected from changes
in pH. Design criteria such as these would make such treatment extremely
expensive. The obvious corollary to these statements is that a massive
primary buffer system is vital to any successful and economical bio-
logical treatment of beet sugar factory waste. In considering the
design characteristics of a buffering system, the anaerobic pond is an
obvious choice because its simple earthwork construction, large size
and relatively long detention period will, according to our evidence,
buffer almost any extreme variation in pH which might occur in a factory
effluent and would dampen changes in BOD. At the same time it would
be relatively inexpensive.
In view of the necessity of a buffer pond, it is fortunate that in
addition to acting as a buffer, a substantial degree of waste treatment
is attained in an anerobic pond. It was found during the course of
these investigations that in the anerobic pond BOD removal was directly
proportional to BOD loading in the range of 500 to 2,000 Ibs of BOD/
day/acre. At loadings above 2,000 Ibs, removal efficiency is believed
to decline. The removals as found are related to loading approximately
as follows:
R = 0.8 L (3)
578
-------
300
200
•
i
O
O
O
I-
O
O
too
T
ANAEROBIC POND
BRONSON COLLECTOR
SEP
AUG
POINT OF INFLECTION
8 12 16
MONTHLY MEAN POND TEMPERATURE,
20
24
FIGURE 16. RELATIONSHIP BETWEEN GAS PRODUCTION AND
TEMPERATURE FOR VARIOUS MONTHS
579
-------
in which L is the BOD load between 500 and 2,000 Ibs/day/acre and R
is the BOD removed. Of course, both R and L should be expressed in the
same units. Thus, it appears that an anerobic pond loaded at 1,000
Ibs/day/acre will produce an effluent containing 200 Ibs of BOD/day/acre
and one loaded at 2,000 Ibs/day/acre will produce an effluent having
400 Ibs of BOD/day/acre. If BOD loading were the only criterion of
performance, the obvious choice of recommended design loading would
be 2,000 Ibs/day/acre. It is an unfortunate fact that when the study
pond received BOD loadings as high as 2,000 Ibs/day/acre, it was
continuously malodorous; hence some criterion other than Equation 3
is required on which to base a decision regarding an upper limit for
anaerobic pond loading. Two main criteria seem to be available: odor
level and acceptable discharge BOD.
Inasmuch as odor is one of the most urgent problems, odor level will
be discussed first. In the case of the anerobic pond, the data
accumulated during this study were not sufficiently refined for a
final conclusion; but one is left with the general impression that for
normal beet waste, an odor product of lower than 8 to 10 may be barely
acceptable. An odor product much above 10 is definitely unacceptable
because it is always accompanied by sulfide concentrations which are
always detectable. An allowable odor product of 10 would imply very
little carrying power for the odors; or in air, odors of such intensity
would be quickly diluted to subliminal levels as a function of distance
from the ponding site. It seems quite certain that odor products of 8
or less would be acceptable for anaerobic ponds because these levels
were sometimes reported for the aerobic pond, which was never described
as objectionable in pond-site observations.
In the plot of odor product vs loading, there was a straight-line
relationship with two observed exceptions—the October 1967 data
and the November 1968 data. These data come from periods when there
were serious interruptions in flow, and consequently rapid changes in
the environment. Total nitrogen concentrations were also high during
these periods, possibly due to the discharge of Steffens waste..
Steffens waste is, of course, notorious for its ability to produce
vile odors when impounded, and "Steffens waste spills" are always
accompanied by sudden increases in odor in ponding systems^ '' There
was, however, no reported incidence of a "spill" and no report that
Steffens waste had been bled into the effluent. Thus, it can only be
concluded at this time that a load of 1,000 Ibs of BOD/day/acre may be
acceptable in an anerobic pond when mechanical surface aeration is
applied to the extent of meeting one-third of the applied BOD, and
when the effluent is discharged into a functional aerobic or algae pond
from which there is a recycle of about 1 Q and in which algae are
growing and producing oxygen.
With regard to acceptable discharge BOD, effluents from the anaerobic
pond varied from 129 mg/1 at a load of 500 Ibs/day/acre to in excess
of 400 mg/1 at a load of 2,000 Ibs/day/acre. Water of this quality
would be useless and its discharge illegal without extensive additional
treatment.
580
-------
The additional treatment studied involved both facultative and algae
ponding. At BOD levels ranging from 50 to 200 Ibs/day/acre as noted
previously, odor products in the systems studied tended to be 6 or less
and objectionable odors at the pond side were minimal.
With regard to the application of BOD criteria to the facultative
pond and algae pond, an examination of the available data is best
aided by plotting mean final effluent BOD as a function of mean BOD
loading. Such a plot is shown in Figure 17. As is evident from the
Figure, the BOD of unfiltered effluents was not affected by loadings
up to 100 Ibs/day/acre while the BOD of filtered effluents was not
affected by loadings up to 180 Ibs/day/acre, but the unfiltered BOD
both from the facultative pond in 1967 and the algae pond in 1968 would
not be acceptable in the environment under the currently proposed
water quality standards(9).
It is important to note that decreased BOD loadings below 100 Ibs/day/
acre did not appear to influence the effluent BOD, probably because
most of the BOD involved in these samples was of a suspended nature,
either in the form of colloidal sulfides, algae cells, or of bacterial
cells. This was true because the pond was continuously mechanically
mixed and samples were drawn directly from the mixing system. The
suspended nature of the BOD is demonstrated by the fact that the BOD
of the algae pond effluent was reduced from levels of 150 to 190 mg/1
to 10 to 13 mg/1 by filtration. In view of these relationships, it
is apparent that the poor BOD removals attained in the algae pond were
due to removable BOD. Therefore, had the removal been effected,
greater efficiencies would have been attained.
Merely decreasing the BOD loading below 100 Ibs/day/acre did nothing
to improve effluent quality. On the other hand, filtration improved
quality dramatically even at loadings of 180 Ibs/day/acre. An
examination of suspended solids data shows that effluent suspended solids
often exceeded a concentration of 500 mg/1 with over 75 percent volatile
matter. Thus, apparently an improved unfiltered effluent could not
be attained by reduced loading. It should be noted, however, that in
the case of the algae pond, loadings between 100 and 180 Ibs/day/acre
yielded effluents which when unfiltered had BOD levels between 80 and
190 mg/1 respectively. Thus, while a loading of 100 Ibs/day/acre was
associated with unsuitable effluents, higher loadings produced effluents
substantially more unsuitable. The apparent conclusion is that aerated-
anaerobic ponding followed by either facultative or algae ponding
without further treatment did not produce an effluent suitable for
discharge regardless of the degree to which loading is decreased. One
is then, confronted with only two clear alternatives: 1) to use
facultative secondary ponds loaded at 100 Ibs/day/acre or less and to
dispose of the final effluent on land owned by the Factory from which
there is no discharge; or 2) to use a more intensive form of secondary
ponding and to remove suspended solids from final effluents prior to
discharge by filtration or some other method of separation which will
remove the fine suspended solids which contribute most of the effluent
BOD.
581
-------
200
-------
Intensive secondary ponding could involve the use of mixed algae ponds
of the type studied, or an even more intensive form of algae production.
For example, where climate permitted, the algae pond could be optimized
for photosynthetic oxygenation with an average production of about 200
Ibs of oxygen and 120 Ibs of algae/day/acre, Under these conditions,
it is conceivable that the value of the filtered solids could pay for
the cost of filtration. However, this would require much more study.
Allowable loading would then be about 200 Ibs/day/acre, and following
separation the discharge would meet rigorous quality standards. One
nay then logically say, "Why discharge such high quality water particu-
larly in water-short areas?" If the quality of a filtered discharge is
as high as indicated in Figure 17, reuse in the factory by recycle
is a worthwhile consideration whenever filtration is used. Obviously
the less new water brought into a factory the less it will have to
discharge. On the other hand, because of the high water content of
beets, it seems inevitable that factories will always be forced to
discharge or otherwise dispose of large amounts of excess water
regardless of recovery practice.
Based on the criteria of 80 percent BOD removal in a primary anaerobic
pond and a loading of 200 Ibs BOD/day/acre in an algae pond, one can
explore the areal requirements for treatment in an idealized anaerobic-
algae system. To have a basis for calculation, a 4.5-K ton factory
discharging 36,000 Ibs of BOD per day is assumed. It is also assumed
that excessive odors will occur when the aerated-anaerobic pond loading
exceeds 1000 Ibs of BOD/day/acre. If a load of 500 Ibs of BOD were
applied to the anaerobic pond, the area of anaerobic pond would be
72 acres, the discharge would be 100 Ibs/day/acre or 7,200 Ibs, and
36 acres of algae pond would be required for an aggregate area of
108 acres. If a load of 1,000 Ibs/acre is applied to the anaerobic
pond, the total loading of 36,000 Ibs will require 36 acres. The
anaerobic pond effluent will contain 7,200 Ibs of BOD and therefore will
require 36 acres of algae pond. The aggregate area in this second
case would be 72 acres. If a load of 2,000 Ibs of BOD/acre were used,
18 acres of anaerobic pond would be required for anaerobic treatment,
but odors would be severe and a stronger waste would be discharged to
the algae pond. Although according to Equation 2 the effluent BOD
would be 7,200 Ibs, and would require 36 acres of algae pond, in reality
because of bacterial and sulfide turbidity, an area larger than 36
acres—say 40 acres—probably would be required if the effluent were to
be fully oxidized. Moreover strong odors would occur.
A plot of these relationships is presented in Figure 18. From the
Figure, it becomes evident that with the assumptions described, the
minimum area required for odor-free treatment of factory waste from a
4.5 K ton factory by an aerated anaerobic-algae system would be about
72 acres. Higher loadings would be accompanied by severe odors and,
as indicated by the dotted portion of the top curve, would require
more algae ponds for aeration. Lower loadings would, of course, require
more area, the area increasing as the areal loading is decreased.
583
-------
20
oo
o
•o
Ol
crt
.O
O
T3
!•
•o
c
3
.C
f»
O
2
O
O
O
Q.
cr
a:
a
16
12
ADDITIONAL AREA REQUIRED TO PRODUCE OXYGEN
TO OXIDIZE REDUCED INORGANIC SUBSTANCES AND
TO COMPENSATE FOR COLLOIDAL TURBIDITY
MIXED
SECONDARY
OR
ALGAE
PONDS
DESIGN OPTIMUM
PRIMARY
OR
ANAEROBIC
PONDS
( SURFACE AERATED
FOR ODOR CONTROL)
FOR EXPERIMENTAL SYSTEM
I
FIGURE 18.
100 2OO 3OO
POND AREA,acres
MAIN WASTE PONDING ARIEA REQUIRED FOR 4.5 K-TON FACTORY
-------
As an alternative to using facultative or algae ponding as the secondary
system, one should consider the alternatives of more intensive aeration
of primary ponds or the use of aerated secondary ponds. The possibility
of applying additional surface aeration in an anaerobic pond to prevent
odors and to permit higher areal loadings and decreased pond surface
area is worthy of consideration, and was to some extent examined
experimentally. In the study with a 5 hp floating surface aerator,
it was found that as long as the dissolved oxygen remained zero,
oxygen entered the pond at the rate of 307 Ibs/day due to the 5 hp
aerator. Odor emission, however, appeared to be a function of the
amount of unoxidized material remaining, rather than a function of the
rate of aeration. Thus, if one were to go from a loading of 1,000
Ibs/day/acre to a double loading of 2,000 Ibs/day/acre, control of
odor probably could not be attained by aerating at twice the 307 Ibs/
day/acre, i.e., 614 Ibs/day/acre. Instead aeration would have to be
at a rate of 1000 + 307 Ibs or 1,307 Ibs/day/acre. Thus, it is
believed that to double loadings in the system studied, aeration would
have to be increased fourfold to prevent odors. The savings in area
would thus have to be evaluated in terms of the cost of one 20 hp
aerator operating continuously for each acre of secondary pond replaced.
Thus, the possibility of decreasing area by going to loadings higher
than 1,000 Ibs/day/acre is not as attractive as it might at first seem.
With regard to replacement of secondary facultative or algae ponds with
aerated secondary ponds, it will be assumed that the mechanically
induced reaeration rates obtained in oxygen-free ponds do not apply
and that the dissolved oxygen level for discharged or recovered" water
should be about 4 mg/1. Under such conditions, according to the
manufacturers brochure( ', a surface aerator provides 1.8 Ibs of 02
per hp-hr. At this rate, each 5 hp aerator will contribute 220 Ibs
of 02 per day. This is approximately the amount of oxygen contributed
by one acre of high-rate pond. By itself, a 5 hp aerator would
probably be less costly than one acre: of pond, but inasmuch as waste-
water storage for reuse or disposal may be an essential part of any
practical system, a substantial pond area may be required anyway, and
all of the pond costs therefore need not be allocated against the
aeration process.
Aeration due to flow mixing also is worthy of consideration as a source
of supplemental aeration. This form of aeration is often referred to as
eddy diffusion aeration because it is dependent on renewal of the
surface resulting from eddies generated as the water moves past small
discontinuities at the pond bottom. A pond four-feet deep with the
liquid moving around a closed circuit at a velocity of one foot per
second and containing no free dissolved oxygen will absorb about 100
Ibs of oxygen/day/acre through surface reaeration. Although this
amount of oxygen is small compared with that attainable through
mechanical surface aeration, it is a method which is compatible with
photosynthetic oxygenation. If the liquid contains growing algae and
essential nutrients, the algae may produce over 200 Ibs of oxygen/day/
acre through photosynthesis, and the dissolved oxygen level will
585
-------
always be positive and frequently near saturation. Oxygen produced
photosynthetically, of course, does not depend upon an oxygen deficit
for input; and hence, never is accompanied by vile odors. Odors
could result if the pond were so severely overloaded on a sustained
basis that algae were unable to grow. However, if algae either fail
to grow for one reason or another or are killed by toxins or predators,
eddy diffusion aeration in the flow system will provide for the
absorption of sufficient oxygen by the pond to keep it from producing
vile odors of high intensity, providing the loading does not exceed
200 Ibs/day/acre. Thus, flow mixing is an excellent backup for photo-
synthetic oxygenation; but it is not as efficient an aeration system
as is mechanical surface aeration.
Provision for any form of biological treatment requires in addition to
a mild and stable environment adequate nutritional conditions. Flume
water usually contains about 400 tng/1 of carbon as C, about 15 mg/1
of available nitrogen as N, and about 3 mg/1 of soluble phosphate as
P. Rapidly growing bacteria are about 40 percent carbon as C, 10
percent N as N, and 1.5 percent phosphorus as P, and rapidly growing
microalgae are about 55 percent carbon as C, 8 percent nitrogen as
N, and 1 percent phosphorus as P. Based on these percentages, flume
water contains enough carbon to support 725 mg/1 of algae, enough
nitrogen to support 188 mg/1 of algae, and enough phosphorus to support
300 mg/1 of algae. It is clear that compared with carbon, there is
a deficit in both nitrogen and phosphorus, and that 725/188 x 15 or
about 88 mg/1 of N and 725/300 x 3 or about 7.25 mg/1 of P would be
required to permit incorporation of all of the organic carbon into
algae. Thus, 73 mg/1 of N and 4.25 mg/1 of P would have to be added.
A less expensive alternative would be to remove carbon by processes
other than by photosynthesis, and thereby decrease the carbon to a
point at which it is in balance with the nitrogen and phosphorus for
algae growth. By so doing, it would not be necessary to provide
supplementary nitrogen and phosphorus.
Several processes observed in the anaerobic pond during this study are
accompanied by loss of considerable carbon dioxide to the atmosphere.
This is probably especially true in the case of mechanically aerated
ponds. For example, the satisfaction of 307 Ibs of BOD/day/acre by
aeration led to the production of 450 Ibs of C02 per acre. Thus,
probably 122 Ibs of carbon/day/acre was lost to the air. This alone
would constitute a loss of 222 Ibs of carbonaceous AGP. Following
this line of reasoning, if two-thirds of the BOD had been met by
aeration, sufficient carbon would have been lost to permit the balance
to be incorporated in algae with no need to add nitrogen or phosphorus.
Methane fermentation also is a potential method for decreasing the
carbon content of the waste. As noted earlier, methane fermentaion in
the system was extemely low, and did not establish itself appreciably
in spite of the presence of large amounts of carbon, highly anaerobic
conditions, adequate temperatures, and sustained periods of inundation.
586
-------
Under the warmest temperatures encountered in the study, 250 ft^ of
gas were emitted from the bottom of the anaerobic pond. This gas
probably contained about 50 percent methane, of which 75 percent is
carbon. Thus only about 4 Ibs of carbon were lost daily by fermenr-
tation. If methane production rates equal to those in domestic sewage
ponds had been attained, as much as 200 Ibs/acre of carbon would have
been lost daily in the form of methane. However, this did not occur
in the system. As to why it did not occur, it is possible that
methane bacteria found it difficult to survive in a pond in which
surface aeration was in progress. However, even before the aerator
was installed, there was a dearth of fermentation. Thus, further
studies would be required to substantiate aeration interference.
Another hypothesis advanced for poor fermentation is the excess of
H^S or sulfides in the system due to vigorous sulfate reduction in the
anaerobic pond. Sulfide is known to be toxic to methane bacteria even
in moderate concentrations. Thus, carbon elimination by methane
fermentation could have been inhibited by this mechanism.
In order to continue with the discussion of carbon, nitrogen, and
phosphorus ratio and the need for nutrient supplementation, it is
necessary to digress slightly to explore sulfate reduction in some
detail, since this is also a potentially significant way in which
carbon may be lost.
The overall reaction involved in production of H2S from sulfate using
organic matter as an energy source is approximately:
H2S04 + 2CH20 Bacteria^ H2S + 2C02 + H20 (4)
Organics
According to the data on sulfate reduction shown in Figure 11, as much
as 208 mg/1 (average about 140 mg/1) of sulfate was reduced in the
anaerobic pond during the five-month period of observation. Based on
the stoichiochemistry of equation (4), the reduction of 140 mg/1 of
sulfate would have involved the oxidation of 88 mg/1 of organic matter
and the production of 50 mg/1 of H2S and 128 mg/1 of C02 or 35 mg/1
of carbon. A substantial fraction of this C02 was doubtlessly incorp-
orated in the alkalinity of the water in the anaerobic pond.
One may question why, that although 50 mg/1 of H2S were produced, a
maximum of only 5-6 mg/1 of HS~ were recovered. This imbalance may
be explained by the supposition that as soon as HS~ was produced, it
could have reacted in several ways. It could have reacted with
magnesium or other metals in the system to produce insoluble metal
sulfides which of course were not measured in the dissolved sulfide
determination; it could have been emitted into the air by action of the
aerator; or it could have been utilized by sulfur bacteria. Emission
into the air was evidenced by the presence of sulfide odors about the
anaerobic pond. However, the magnitude of emission must have been small
because of the pH of 7.5 to 8, (a level almost always found in the
anaerobic pond) most of the sulfide would be present as the HS~ species
rather than as H2S.
587
-------
Inasmuch as there is no H2S normally in the air, the reaction involved
in the emission of- H2S occurs spontaneously whenever dissolved H2S
exists in solution. The quantity remining in solution is mainly a
function of pH. The reaction is:
H+ + HS- Acid> H2S f (5)
Basic
At pH 4 practically all of the material is in the H s or gaseous form
and emission rates are high. At pH 6, half is in tne gaseous form and
half in the ionic HS~ form; and at pH 7.5, about 90 percent is in the
ionic form and 10 percent in gaseous form and emission rates are low.
At pH 9-5, almost all material is in the ionic form and there is no
emission of H2S. Thus, in view of the pH levels in the anaerobic pond
(Figure 9), the possibility is slight that all H2S would have been
emitted due to aeration during August, September, October, and November
when the pH was higher. Yet, a substantial fraction could have been
lost that way during December when the pH was low. The curves in
Figure 11 give evidence that some sulfides were probably carried over
into the algae pond as a reduced complex which later was oxidized to
sulfate. This is evidenced by the fact that the sulfate concentration
in the algae pond was consistently greater than it was in the anaerobic
pond.
With regard to the loss of sulfide to sulfur bacteria, according to
the data collected there were numerous sulfur bacteria in the pond
from time to time although their biomass was small. However, several
parts per million of sulfide could have been converted to elemental
sulfur by these bacteria.
Returning to the specific question of carbon losses, according to
Equation 4, the overall conversion of carbon due to sulfate reduction
must have been on the order of 128 mg/1 as C02 or 35 mg/1 as carbon.
Overall then, more than one-half of the carbon introduced must have
been converted to C02.
Another check on carbon transformation is provided by the COD data.
According to COD information (see Table 4), the overall reduction in
unfiltered COD for the anaerobic pond averaged 1,150 mg/1. Based on
the classical oxidation equation:
CH20 + 02 > C02 + H20 (6)
For which the combining weights are:
30 + 32 > 12 + 32 + 18
An average carbon release of 460 mg/1 in the anaerobic pond would
have occurred due to oxidation of 1,150 mg/1 of COD. It cannot be
588
-------
decisively stated, however, that the decreases in organic carbon were
accompanied by an increase in alkalinity, because the alkalinity was
not measured routinely. If an actual loss in carbon occurred, and if
nitrogen and phosphate had remained constant while carbon decreased,
there would theoretically have been nearly sufficient N and P to
satisfy the carbonaceous algae growth potential of liquid entering the
algae pond. Even though carbon may have been lost, nitrogen as well
as carbon and sulfate was also lost as the carbon was passed through the
anaerobic system. The greatest loss occurred in nitrate-nitrogen.
Losses in this nitrogen form were as high as 90 percent. Losses in
ammonia-nitrogen also amounted to from 20 percent to 90 percent or
more of that originally introduced, depending on the time of year.
Inasmuch as there was so little nitrogen to begin with, proportional
losses in nitrogen kept pace with or exceeded losses in carbon, with
the result that a severe shortage of available nitrogen prevailed
throughout the series, i.e., in both the anaerobic and the algae pond.
Based on the results, the significant conclusion is that it would be
difficult to amplify the quantity of nitrogen in the algae portion of
the system by adding nitrate or ammonia to the anaerobic pond because
they are apparently simply reduced or oxidized and emitted as nitrogen
gas or ammonia from the anaerobic pond and thus wasted. With regard to
the algae pond, N-fertilization is best accomplished by adding nitrogen
to the algae pond directly. The best nitrogen additive to the
anaerobic pond probably would be organic N. Organic N addition is
potentially provided by settled algae brought in with recirculatant
and algae pond effluent.
When the applied BOD to the algae pond is 200 mg/1, 10 to 15 mg/1 of
NH^-N should be adequate to provide for the nitrogen deficiency in the
anaerobic pond effluent going to the algae pond. Addition of the
nitrogen as anhydrous ammonia probably is to be preferred to adding
the nitrogen as nitrate since nitrate would be quickly reduced.
In the past nitrate has been added to sour anaerobic ponds to provide
some degree of odor control. However, because of the large amount of
nitrate which must be added to provide control(40 percent of the BOD)
and its costs, this method of controlling odors would be about 5 times
as expensive as would be control with floating surface aerators.
Phosphate added to the anaerobic pond and algae pond evidently was
accompanied by its rapid disappearance from the system—probably as a
precipitate in the algae pond and possibly by phosphate reduction in
the anaerobic pond. According to Waksman and Starkey (10) under
anaerobic conditions when organic matter and when phosphates and the
necessary bacteria are present, phosphates are reduced to phosphate (H3P03),
hypophosphites (I^PC^) , and phosphine (PH-j) gas with the release of C02«
Thus, nitrates, sulfates, and phosphates are all potentially reduced in
anaerobic ponds.
Although the methodology of nutrient addition was not exhaustively
explored in this study, the prior discussion is convincing evidence that
the strategy of nutrient addition is at least as important as
589
-------
the nutrient addition itself. On the basis of the limited experience
obtained in this study, it is certain that nutrient addition could
best be studied in a full-scale system.
The question of why algal or facultative ponds following anaerobic
ponds performed poorly in BOD removal is explained on the basis of
several facts derived from the study. First of all, because of the
efficiency of BOD removal in the anaerobic pond, only a limited amount
of BOD remained to be removed in the secondary ponds. Secondly,
nitrogen losses in the anaerobic pond limited subsequent algae growth;
and finally the presence of large amounts of colloidal suspended solids
in the algae and facultative pond effluent imparted to them a BOD of
about 100 mg/1 even when the loading was very low. Had there been
separation of suspended solids from the final effluents, removals
would have been greatly improved.
The greatest difficulties in this aeries of experiments and in the
entire study resulted from the fact that the pilot plant was on a
shunt from the main factory waste, and consequently was subjected to
very frequent failures in the feed system. Inability to control this
factor within the budget provided for the study ultimately led to the
termination of the studies, and to the conclusion by the authors that
any further pilot work with beet sugar wastes should be done with the
entire output of a factory rather than with a shunt system.
The fact that nitrogen and perhaps phosphate usually must be added to
flume water for aerobic treatment following passage through an anaerobic
pond, and the fact that federal and local standards may be established
regulating the quantitites of nitrogen and phosphorus which may be
discharged into the environment suggests that future studies must
involve the control of nutrient additions to give maximum benefit with
minimum residual discharge.
Control of effluent BOD as well as effluent nitrogen and phosphorus
will almost certainly involve a filtration, coagulation, or separation
step to remove suspended solids from final effluents. The development
of adequate filtration or harvesting systems is thus an area of
significant concern which must be further explored. Effective separation
following adequate treatment should permit significant reuse of water
for relatively high purposes within a factory.
Predators were extremely difficult to deal with in the algae pond but
it is not clear whether the same succession of predators would occur
in an algae pond in which there was sufficient nitrogen for algae
growth, and hence, in which carbon is limiting. Results obtained in
recent studies of domestic sewage systems give some evidence that
certain predators such as daphnia cannot withstand the pH changes which
occur in carbon-limited algae ponds. Because of their relatively
large physical size as compared to algae, predators can be removed from
recycled streams by screening. DSM or rotary screens having mesh
openings about 200 to 400 microns are effective in predator removal.
Both screening and carbon limitation in preference to use of pesticides
should be further studied for predator control.
590
-------
There Is no clear explanation for the lack of methane fermentation in
the system studied, because as opposed to the findings in this pilot
study, methane fermentation is frequently observed in primary beet
waste ponds. General observations indicate, however, that the
fermentation is most active and visible in those ponds which receive
a substantial quantity of mud. Whether this mud traps the gas and,
hence causes the release of larger and more spectacular bubbles, or
whether it actually acts as an essential substrate surface for methane
bacteria is not clear. Because primary sedimentation was used, little
mud entered the anaerobic pond of this system. Methane fermentation
is always slow to start in new systems. Yet there were periods during
earlier runs of this series when the ponds showed some evidence of
fermentation more vigorous than that observed during the Fall 1968
campaign. Studies of a pond in which methane fermentation is definitely
established and in which the amount of settleable solids introduced can
be controlled would be required to explore this phenomenon.
The pertinence of these studies to cold climate installations should
be considered. Chemical treatment with lime for pH control and
sedimentation with recycling water, and discharge of excess water to
an aerated-anaerobic pond is the only alternative thus far explored.(H)
This system leads to accumulation of a high carbonaceous load in the
recycled water. The maintenance of a high pH causes precipitation and
removal of essential nutrients for microbial growth. Thus, decomposition
in the anaerobic pond is slow and unbalanced. It is believed that if
covered digestion ponds could be developed, it would be preferable to
pass wastes through an anaerobic pond prior to chemical treatment.
Floating covers for the ponds would preserve factory heat, prevent odors,
and permit a high degree of fermentation to occur. Following this
fermentation, chemical treatment with supplementary aeration could be
applied, and supernatant liquids would be suitable for reuse in the
factory or for storage without odor nuisance. Thus, the development
of inexpensive pond covers would be a worthwhile study for future
investigation.
CONCLUSIONS
Because of the large quantities of mud and extraneous vegetation
contained in beet flume water, screening and short-term sedimentation
are absolute requirements for pretreatment of beet sugar waste. The
degree of sedimentation should be limited in time, however, since
carry-over mud seems to have a beneficial effect on fermentation in
an anaerobic beet waste pond.
Even after screening and short-term sedimentation, beet sugar waste
is so variable in pH, BOD, and composition that short detention period
biological processes such as activated sludge, trickling filtration,
and highrate ponds cannot be effectively employed as the initial process
in a treatment system.
591
-------
With mechanical surface aeration at 300 Ibs of 0^ per day, an
anaerobic pond 14 ft deep and one acre in surface area may be loaded
at a rate of 1,000 Ibs of ultimate BOD or about 700 Ibs of five-day
BOD/day/acre without giving rise to an excessive odor nuisance.
When factory waste is passed through an anaerobic pond, the discharge
has a more uniform pH and a much lower and more uniform BOD than does
the original waste. It is thus more subject to effective short
detention time secondary biological treatment.
Following passage of waste through an anaerobic pond, the pond
effluents are devoid of oxygen, high in a turbidity consisting of
microorganism and colloidal-reduced substances such as metal sulfides,
high in BOD, and are malodorous. Effluents of this quality must be
subjected to aerated treatment before storage, discharge to the
environment or reuse.
Aerobic treatment subsequent to adequate anaerobic ponding may involve
photosynthetic oxygenation, simple ponding in a facultative pond, or
possibly mechanical aeration. The last alternative has not been
extensively explored. In the case of photosynthetic oxygenation,
loadings of 200 Ibs of ultimate BOD/day/acre would be acceptable,
whereas in the case of facultative ponds, loadings of 100 Ibs of
ultimate BOD/day/acre are recommended. In the case of mechanical
aeration, it seems likely that to maintain a 4 mg/1 DO residual, 1 hp
would be required for each 40 Ibs of daily ultimate BOD applied to the
aeration pond.
Recirculation of secondary pond effluent to the anaerobic pond influent
has a beneficial effect upon overall treatment. The effect probably
is related to nutrient return and "seeding." A recirculation rate of
% to 1 Q appears adequate to achieve these benefits.
If recovered waste water from an aeration pond is to be rendered suit-
able for discharge to the natural environment or for recycle and .reuse
in the factory, it should be subjected to filtration or separation to
remove clay turbidity and bacterial and algae cells. The removal of
these substances following anaerobic-aerobic treatment produces a final
effluent of fairly high quality having a BOD of less than 20 mg/1.
A minimum of two anaerobic ponds in parallel are recommended to permit
cleaning of the ponds and to provide load control. Similarly, parallel
aeration or algae ponds are recommended to permit flexibility of
operation and maintenance.
Nutrient supplementation with ammonium-nitrogen at 20 mg/1 and
phosphate at about 10 mg/1 is essential for adequate biological
aerobic treatment following passage through the anaerobic pond. To
avoid losses of nutrients at the anaerobic pond, nutrients should be
added following passage of the liquid through the anaerobic pond.
592
-------
Treatment to remove dissolved nutrients will not be required if
nutrient supplementation is carefully controlled. However, to meet
the quality standards set up by most of the states, suspended solids
will have to be removed from the final effluent by filtration or by
some other separation device prior to discharging the effluent into
the environment.
The problem of odors in beet waste ponds can only be solved by avoiding
overloading of ponds, by providing sufficient treatment area,
sufficient aeration and supplementary aeration, and by providing nutrient
supplementation. Anaerobic pond loadings should not exceed 1,000 Ibs/
day/acre, and aeration should be applied at the surface of the
anaerobic pond to the extent required to prevent odors. Secondary
aerobic ponds should be loaded at not more than 200 Ibs/day/acre,
and supplementary aeration in the form of flow mixing or possibly
surface aerators should be provided. Final effluent must be filtered
or otherwise separated to produce a clear supernatant if it is to
be suitable to meet most discharge requirements.
593
-------
ACKNOWLEDGMENTS
This research was supported by a demonstration grant WPD 93-94
from the Environmental Protection Agency and by matching funds
from the Beet Sugar Development Foundation.
Special thanks are due the personnel of Holly Sugar Company, Tracy,
California, for their interest, aid and support throughout the
course of these studies.
We also wish to acknowledge the efforts of Mr. Henry Gee, Research
Specialist of the University of California, Berkeley, for guiding
the analytical work and preparing the figures for publication.
We are especially grateful to the Welles Products Corporation,
Roscoe, Illinois, and to Mr. John Larson of the E. C. Cooley Company,
San Francisco, California, for furnishing the 5 hp floating surface
aerator used in these experiments.
594
-------
REFERENCES
1. Tsugita, R. A., W. J. Oswald, R. C. Cooper, and C. G. Golueke.
"Treatment of Sugarbeet Flume Waste Water by Lagooning - A
Pilot Study." Journal of the American Society of Sugar Beet
Technologists 15:4:282-297 (1969).
2. Beet Sugar Development Foundation. "Facultative and Algal Ponds
for Treating Beet Sugar Wastes." Report on WPD 93-01-02
WPD 93-03 partial. Beet Sugar Development Foundation, Fort
Collins, Colorado (1967).
3. Oswald, W. J |