WATER POLLUTION CONTROL RESEARCH SERIES
17050 DDY 12/71
A LITERATURE SEARCH AND
CRITICAL ANALYSIS OF BIOLOGICAL
TRICKLING FILTER STUDIES-VOL. 1
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's waters. They provide a central source of
information on the research, development, and demonstration
activities in the water research program of the Environmental
Protection Agency, through in-house research and grants and
contracts with Federal, state, and local agencies, research
institutions, and industrial organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications Branch
CWaterJ r Research Information Division, R&M, Environmental
Protection Agency, Washington, D. C. 20460
-------
VOLUME I
A LITERATURE SEARCH AND CRITICAL ANALYSIS
OF BIOLOGICAL TRICKLING FILTER STUDIES
by
Functional Products and Systems Department
The Dow Chemical Company
Midland, Michigan 48640
for the
Office of Research and Monitoring
ENVIRONMENTAL PROTECTION AGENCY
Project No. 17050 DDY
Contract No. 14-12-474
December 1971
For gale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C., 20402 - Price $2.60
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication. Approval
does not signify that the contents necessarily reflect
the views and policies of the Environmental Protection
Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
ii
-------
ABSTRACT
A compilation of the literature on biological trickling filter studies
and related pollution abatement processes was made. References were
selected for a section of the review and then critiqued.
The review is composed of two volumes. Volume I, the literature review
and critical analysis, is comprised of four parts: (a) Introduction,
Definitions, History and Background Theory of the Trickling Filter Pro-
cess; (b) Plant Design, Materials of Construction, Operation, Maintenance
and Performance; (c) Trickling Filter Research and Development Approaches,
Ecology, and Patents, and (d) Applications of Trickling Filters to Specific
Industrial Wastes. Volume II is the bibliography, a compilation of over
5,000 references in author and journal alphabetical sequence.
Based on the review, several general conclusions were drawn. There is
no well-defined theory of design and operation generally accepted by
the principal investigators in the field. A great amount of published
work was redundant, and European efforts were not readily accepted in
the United States, and vice versa. The literature reflects cycles of
interest in trickling filters. The value of much of the early work was
ignored. Solutions to complex pollution problems will be made by industry
with strong urging and support from local, state, and Federal governments.
The biological trickling filter will be used in high efficiency, modern
wastewater treatment plants. The process is not applicable to all pollu-
tion problems, but its shock survival capabilities and rapid flow-through
time are definite advantages which cannot be overlooked in any design of
a waste treatment facility.
This report was submitted in fulfillment of Project Number 17050 DDY,
Contract 14-12-474, under sponsorship of the Environmental Protection
Agency.
iii
-------
CONTENTS
Section
I
II
III
IV
V
VI
VII
VIII
IX
X
XI
XII
XIII
XIV
XV
XVI
XVII
XVIII
XIX
XX
XXI
XXII
XXIII
XXIV
XXV
XXVI
XXVII
XXVIII
XXIX
XXX
XXXI
XXXII
Conclusions and Analysis 1
Summary of Review 5
Introduction 11
Part I
Introduction 15
Definitions 19
Historical Development 23
Background Theory of Process 27
Part II
Design Criteria of Various Trickling Filter
Modifications 47
Construction Trends 123
Operation of Trickling Filters 155
Maintenance of Biological Trickling Filters 163
Performance Summary of Trickling Filters 173
Part III
Laboratory-Scale Investigations 195
Pilot-Scale Investigations 201
Field- and Full-Scale Investigations 207
Ecology 213
Patents 221
Part IV
Brewery and Distillery Wastes 225
Organic and Inorganic Chemical Wastes 233
Gas and Coke Plant Wastes 241
Food Processing Wastes 251
Institutional and Military Wastes 263
Laundry and Cleaning Wastes 273
Meat and Poultry Wastes 279
Metal Working Wastes 289
Milk Wastes 295
Pharmaceutical and Fermentation Wastes 301
Pulp and Paper Wastes 309
Radioactive Wastes 321
Tannery Wastes 327
Textile Wastes 333
Acknowledgements 341
-------
FIGURES
PAGE
1 SCHEMATIC DIAGRAM OF TRICKLING FILTER PROCESS 17
2 FACTORS INFLUENCING THE USE OF TRICKLING
FILTERS 25
3 SCHEMATIC OF THEORY 29
4 BIOFILTRATION RECIRCULATION DIAGRAMS 9O
5 FLOW DIAGRAMS OF SINGLE STAGE HIGH-RATE FIL-
TRATION 91
6 FLOW DIAGRAMS OF TWO-STAGE HIGH-RATE FIL-
TRATION 92
7 CONSTRUCTION COST PER CAPITA FOR TRICKLING
FILTERS 118
8 CONSTRUCTION COSTS PER CAPITA FOR IMHOFF SLUDGE
DIGESTION PLANTS 119
VI
-------
TABLES
No. Page
1 Design Comparison of Different Rate Filters 55
2 Depth of Medium Trends 85
3 Characteristics of Types of Filter Media (5196) 149
4 Treatment Operational Costs of Sewage Plants 161
5 Summary of Filter Criteria 186
6 Composite List of Trickling Filter Organisms -
Cooke (779)- 214
VII
-------
SECTION I
CONCLUSIONS AND ANALYSIS
The following conclusions and analysis are derived from
the review of the literature on biological trickling filter
studies.
1. Several theories are proposed to describe the perform-
ance of biological trickling filters. Yet in the ninety
years of development, there is no generally accepted
theory.
2. Recent publications demonstrate in-depth understanding
of specific phases of the mechanisms involved in bio-
logical filtration.
3. A lack of appreciation of ecological information is
noted by many workers who regard the trickling filter
process as a "black box" in which the identity of
responsible organisms is unimportant, and only the
overall performance of the process is evaluated.
4. Intermittent dosing at low hydraulic rates is generally
replaced by continuous dosage at high rates which in-
volves design changes to assure an acceptable effluent.
5. The contact time or retention time is significant, and
greater removal is achieved by having an interface
of active organisms, pretreated wastes and sufficient
oxygen for an optimum time interval.
6. Little conclusive evidence is available which illus-
trates inherent advantages of one medium over another,
provided these media conform closely to ideal medium
requirements. However, some porous materials adsorb
certain organics and radioisotopes to a limited extent.
7. Natural ventilation is adequate for most waste treat-
ment conditions. With elevated temperature and high
organic loadings, forced air ventilation is required,
but an oxygen-enriched atmosphere is seldom, if ever,
used for ventilation.
8. Evidence indicates that the temperature of the waste,
not of the air, is significant in determining the
performance of biological filtration.
-------
9. Controversy exists over the effects of recirculation,
and recent reports indicate that (a) recirculation
of settled sludge is detrimental, (b) recirculation
of clarified effluent is advantageous to filter
performance, and (c) intermediate clarification
between two-stage filtration is not necessary.
10. Trickling filters are used successfully in series
operation with other units such as septic tanks and
the activated-sludge process. In the case of the
latter, high-rate filters are effective prior to
activated-sludge treatment and low-rate filters
are effective after activated-sludge treatment.
11. The terminology for the units for performance evalua-
tion certainly can be improved, since percent BOD
removal (commonly used) does not express the actual
pollution load to a receiving body, nor is it easily
correlated with operating expenses reported in
dollars/mgd. :
12. In-depth mathematical investigations have the tendency
of losing their usefulness to the nonmathematically
oriented practitioner, and it is unfortunate that
these valuable studies have not been correlated with
the vast operator experience available.
13. Data derived from laboratory-scale investigations
are used conservatively to define pilot plant
conditions, but are adaptable, with the proper inter-
pretation, to full-scale designs.
14. pilot-scale investigations are used for research and
development purposes as well as definition of design
criteria. An unfortunate temptation must be noted,
for this scale study is a means of delaying the /time
of construction of full-scale facilities.
15. Engineers place more confidence on field- or full-
scale investigations and relationships derived there-
from than on laboratory data, with the recognized limi-
tations of uncontrollable environmental conditions,
variable waste sources, and problems in sampling,
analyzing and interpreting the data.
16. Many industrial wastes are at least partially treated
by biological filtration with most influents requiring
some form of pretreatment to remove suspended solids,
adjust pH, equalize temperature and shock loads, and
maintain a continuous flow. The ability of biological
trickling filters to survive temporary changes due to
fluctuating characteristics of industrial wastes, even
after pretreatment, is of considerable value.
-------
17. Architectural considerations do not govern trickling
filter plant design in this country, but indications
of an awareness that this is desirable are encouraging.
18. It is observed that proper operation by qualified
personnel is imperative to assure that a biological
filtration plant will perform as designed and con-
structed.
19. For a biological filtration plant to be nuisance-
free, the primary maintenance duty is simply clean-
liness. Several approaches using physical, chemi-
cal, and biological means for the control of filter
flies are suggested, with the biological scheme to
be preferred.
2O. A point of interest for future design may be gathered
from the many papers criticizing old waste treatment
plants which do not have facilities for nutrient
removal. However, at the time of their design, the
purpose of their construction was disease prevention
and control of water-borne epidemics. Present and
future designs should, whenever possible, project
the possibility of additions to treatment facilities.
21. The literature reflects competition or a lack of
confidence among nations which results in conflicting
conclusions and duplicated efforts, but also produces
alternative solutions to waste treatment problems.
22. The several trade waste associations and committees
of founder societies are dedicated to the dissemina-
tion, sponsorship, and evaluation of work on waste
water. These groups have made major contributions
to the characterization of various waste waters with
suggested methods of treatment. This is an excellent
example of cooperation to solve complex problems.
23. This review is not intended as an in-depth patent
search, but sufficient evidence is gathered in the
areas investigated to result in the citation of some
issued patents.
24. Cost comparison data are difficult to obtain, and
quite often the information is reported in a nonstahdard
form, further complicating the economic evaluation of
the processes.
-------
25. A significant lesson in design, as it affects the
total economy, appears available based on the experi-
ences during World War II. Under the pressures of
war, treatment plants were intentionally designed
to accomplish just adequate treatment, and, as
populations become more dense and these plants were
modified, further complications developed in opera-
tion and conversion, and added expenses were incurred
to correct early designs.
26. The literature is highly redundant; for example,
reports on the construction of a waste treatment
facility are published in several trade journals.
Many conflicting viewpoints are also expressed, seem-
ingly unaware of the work of other investigators.
27. Examples of rediscovery are noted, such as the use of
dipping contact filters, with little or no evidence
that modifications have been made to correct the
errors of the past which caused the earlier work
to fade in popularity.
28. The apparent lack of discretion by authors to clutter
the literature with similar, if not identical, pub-
lications cannot be overemphasized. Under the title
of "Solve All Problems" -type papers, several people
published rather weakly documented investigations
with only tacit applicability to the solution of
waste treatment problems.
29. It is observed that cycles of interest in biological
trickling filters are experienced in about fifteen-
year increments, beginning shortly before 190O.
3O. Generally, the biological trickling filter has been, is,
and will be used for a considerable period of time. All
facets are not understood, but sufficient experience has
been gained which allows this process to be used effec-
tively in conjunction with other waste treatment opera-
tions. The process is obviously not applicable to all
waste problems, but many investigators outline, in hun-
dreds of pages, the advantages and illustrate the cases
where the biological trickling filter provides the treat-
ment required.
-------
SECTION II
SUMMARY OF REVIEW
In describing the literature on trickling filters, E. Sherman
Chase (658)* stated, "Over the years much has been written
upon trickling filters. A review of the literature shows
that the same thing has been said over and over." With few
exceptions, Mr. Chase's statements, based on 35 years of
experience and written 26 years ago, are still applicable
today.
The purpose of the present effort was established upon a
recognized need to review and critique the literature of the
biological trickling filter. Engineers and scientists must
have available information to aid in documenting their con-
clusions, to draw upon the experience of the past, to edu-
cate the novice, to avoid unintentional duplication of work,
and to keep abreast of developments.
The approach involved a compilation of references which were
then categorized. Major areas of interest were outlined.
The Categories were matched and blended to produce sections
covering specific topics in the major areas. The result of
this effort facilitated a composite of many of the references
on a subject. References were selected, reviewed and critiqued
(Volume I). The bibliography (Volume II) is organized in
author and journal alphabetical sequence.
General information on definition of the biological trick-
ling filter process is followed by a discussion of the
history and background theory. This information provides a
basis for the discussion of the remainder of the text. The
text is prepared in four parts, i.e., Part I - Introduction,
Definitions, History and Background Theory of The Trickling
Filter Process; Part II - Plant Design, Materials of Con-
struction, Operation, Maintenance and Performance; Part III -
Trickling Filter Research and Development Approaches, Ecology,
and Patents; and Part IV - Applications of Trickling Filters
to Specific Industrial Wastes.
Theoretical investigations deal with the effect of suspended
solids, the hydraulic and organic loadings, and rates of
recirculation applied to biological trickling filters. The-
oretical criteria are developed on the chemical and mechanical
properties of the required trickling filter medium. Several
investigators developed a lively dialogue on various theories
of natural and forced ventilation for the biological trickling
filter. Mathematical investigations are noted to predomi-
nate in the literature of the last decade as attempts were
made to put the various theories of biological filtration on
a quantitative basis.
*References are found in Volume II.
-------
In the discussion of design factors, the type of waste is
identified as the principal criterion to be considered for
the application of biological filtration. Several modifica-
tions of the trickling filter have been designed, such as
contact and dipping contact beds, prior to the conventional
form of low-rate and high-rate trickling filters operating
in single- and two-stage systems. Multi-unit systems employ-
ing biological filtration in series with the activated-
sludge process, lagoons, septic tanks, Imhoff tanks, and other
unit operations are reported with their inherent advantages
and disadvantages. Modifications of the basic process,
such as three-stage biological filtration, alternating
double filtration and contact aerators, have application
under specific circumstances. Design criteria for roughing
and polishing filters are well established and developed
for biologically generated solids separation and disposal,
as well as effluent disposal through reuse. Design informa-
tion related to recirculation is shown to have attracted the
attention of the mathematically oriented investigators who
developed expressions relating the hydraulic and organic
load and their effect on the filter.
Pretreatment systems have been designed principally for con-
ditioning the influent to the filter by removing most of
the suspended material, providing a neutral pH, an ambient
temperature, and an equalized flow. Post-treatment applica-
tions have been designed primarily for the solids-liquid
separation and disinfection, as well as some nutrient re-
moval to provide an acceptable effluent. Design factors
for the ventilation and underdrainage systems and filter
enclosures involved detailed studies of the temperature
effect, ventilation rate, oxygen content, carbon dioxide
content, and nuisance control, among other factors. Economic
considerations of design deal with capital cost versus opera-
tional cost, with limited data available.
Construction trends in waste treatment plants are summarized
by characteristics developed over several decades. Construc-
tion of package plant installations has a practical applica-
tion, and data on their successful operation are available.
Materials of construction for distribution, biological film,
support media, walls, and enclosures plus other appurtenances
are discussed in considerable detail. Construction practices
in the economic use of local materials for enclosures, media
and underdrains are also reported. Architectural considera-
tions for waste treatment plant construction are not empha-
sized to any extent in the literature.
Proper operation of biological trickling filter plants is
of utmost importance. The best designed plant, constructed
to exact specifications, can still produce poor quality
effluent, if operated improperly. Many problems, such as
filtration start-up, are adequately solved through good
-------
communication among operators of treatment plants. Operator
training was of sufficient interest that information was
published dealing with the requirements of properly trained
individuals, licensing these individuals, as well as estab-
lishing a wage scale. Operational costs are reported fre-
quently prior to World War II, but later cost data are only
available to a limited extent.
Maintenance has become a major factor in conjunction with the
design and operation of biological trickling filters. As
wages increase, man-hours of maintenance will be decreased.
Information dealing with maintenance and the development of
maintenance-free equipment occurs intermittently throughout
the literature. The use of corrosion-resistant materials
is encouraged, and access to the equipment for maintenance
is incorporated in the design. The avoidance of nuisance
conditions of odors, filter fly production, and other ob-
jectionable conditions occupies a considerable volume of
the literature.
In an attempt to evaluate the performance of biological
filtration processes, many investigators developed design
expressions and mathematical models., More data were avail-
able to investigators after the National Research Council's
studies in the late 1940's. With these data and the model
described by the Council, several investigators proposed
alternative relationships involving various parameters.
An appreciation for statistics was demonstrated, and mono-
graphs were developed for the rapid solution of many of the
derived expressions. The determination of the key parameters
defining biological filtration performance evoked consider-
able debate among investigators. Other filter performance
criteria were expressed in terms of bacterial removal, viral
removal, and disease prevention. Explanations of the per-
formance of biological filters in relation to these factors
were published and data indicating the limitations of the
process are available.
The results of research and development, using the laboratory-
scale biological trickling filter simulation, either in a
miniature filter configuration or by modeling some aspect of
the filtration phenomenon, are extensively reported. Aside
from the economy, certain environmental factors can be con-
trolled by testing at this scale, thereby simplifying per-
formance changes with defined variables. Laboratory simula-
tions, such as inclined flat planes, rotating tubes, verti-
cally suspended spheres, as well as other devices, were used
to develop rational formulas to describe the performance of the
biological trickling filter. Laboratory-scale research
and development are frequently used to study and measure
significant variables from which a pilot plant can be con-
structed, leading to the development of design criteria
for full-scale operation. With proper interpretation of the
data, designs have been made for full-scale installations
based on laboratory-scale experimentation.
-------
Pilot plants have been investigated, with or without prelimi-
nary laboratory' studies,primarily for the development of
design criteria for full-scale installations. The results
from pilot plant operation have been compared with data from
theory developed in the laboratory. Pilot plants have been
used extensively by several interested groups, such as
regulatory agencies, industry, and consulting engineers,
to develop efficient waste treatment facilities. If design
changes were required, considerable savings were obtained
at the pilot plant-scale which would not be feasible under
full-scale conditions nor meaningful under laboratory-scale
conditions.
Many contributors emphasized that field- and full-scale in-
vestigations are the most desirable scale for testing and
development. Very high credibility was given to relation-
ships developed from those data. The investigators demon-
strated an awareness of the limitations of full-scale opera-
tions, such as the expense involved, the duration of study,
and the inability to establish key parameters under controlled
conditions. The conservative approach to research and
development has been to use limited laboratory investiga-
tions to define performance characteristics, extensive pilot
plant investigations to define design criteria, and full-
scale construction and operation evaluations to determine
efficiencies and required modifications.
Various life forms have been regarded as important in the
function of biological trickling filters. The required
environmental conditions for the filter biota were noted,
along with observed organism stratification throughout
the depth of the biological filter. Frequent ecological
investigations were reported by regulatory agencies, design
engineers, and operators for upgrading the performance of
waste treatment plants, detecting malfunction, and indicat-
ing the type of correction which should be instituted.
Occasional reference was made to the use of fungi for speci-
fic waste treatment, but the usual reference dealt with
control by physical, chemical, and biological means.
The interrelationships among algae, bacteria, fungi, protozoa,
insect larvae, worms, nematodes, rotifiers, and other
organisms were reviewed in several accounts. Many cases
were cited of ecological upset when one life form predomi-
nated and usually resulted in a nuisance condition or a
loss of plant efficiency. In an effort to increase treat-
ment plant efficiencies, pure culture was used in the
waste treatment, but the results showed no advantages.
The results of research and development were the many patents
in the area of biological filtration, distribution devices,
filter media, ventilation and odor control systems, post-
treatment devices, trickling filters in combination with
other processes, unique cleaning and maintenance procedures,
and other similar categories.
8
-------
Several specific industrial wastes treated by biological
filtration with various degrees of efficiency were reviewed
and categorized, such as brewery and distillery, chemical
production, coke and gas plants, food processing, military
and institutional, laundry and cleaning, meat packing and
poultry, milk processing, pharmaceutical and fermentation,
pulp and paper, radioactive, tannery, and textile wastes.
Most of the industrial wastes required considerable pre-
treatment to guarantee efficient biological filtration.
Quite often combined units such as the trickling filter
with the activated-sludge process gave the most efficient
and reliable treatment.
Prior to 1930, there was considerable interest expressed by
the investigators in treating various trade wastes admixed
with domestic sewage, but considerable difficulty was en-
countered. In the post-World War II period, with the advent
of high-rate biological filtration, successful treatment of
many of the more difficult trade wastes was accomplished
in combination with domestic sewage treatment. Biological
filtration was shown to have a definite part in the treat-
ment of industrial wastes, but was not a cure-all, as other
systems possessed advantages under specific conditions.
General conclusions and recommendations were made based upon
the reviewed publications. These conclusions dealt with
the quality of the publications, significant points of
agreement and disagreement, and interpretation of informa-
tion in the reviewed material.
-------
SECTION III
INTRODUCTION
The conscientious efforts of many thousand people working on
biological trickling filters at the laboratory, pilot plant
and field scales, along with the subsequent reporting of the
results, have created a problem — a tremendous volume of in-
formation. To delve into the background literature requires
considerable time and effort and is not always practical or
feasible. To bridge the gap of the unavailability of dissem-
inated information to the novice, practicing engineer or
environmentalist, all of these data should be collected, com-
piled and evaluated in one source. Thus, this literature
review and critical analysis were undertaken in an attempt to
provide a source of trickling filter information to all tech-
nically interested individuals in this rapidly expanding field.
It was the intent of this project to assemble, compile, and
categorize the international literature pertaining to the bio-
logical trickling filter, and then review and critique this
comprehensive documentation. Constructive criticism included
the strength as well as the weakness of the information. Care-
ful attention was given to the intent and purpose of the in-
vestigations. Selection of the articles reviewed was based
upon their timeliness and technical excellence. It was found
that many articles were similar; for this review, represent-
ative papers were selected and the others listed as additional
references.
This literature review covers the period from the earliest
beginning of the concept of the trickling filter through 1968.
The primary sources of information were the Chemical Abstracts
and the Water Pollution Abstracts. These sources indicated
the original articles in the trade journals, and many of these
were reviewed and, if sufficiently important, obtained for the
reference file. Government publications were another valuable
source of pertinent information. The personal reference files
of certain members of the technical staff of The Dow Chemical
Company, Midland, Michigan, with background in waste disposal
and trickling filters, were the beginning of this compilation
of references.
A review of the collected information developed a list of over
6OO key words. By using the key words and numbered references
with a computer, the references to specific topics were located
and sorted out, and the sections written with the literature
and literature abstracts on hand.
The review and critique are comprised of two volumes. Volume
I is the literature review and critical analysis of selected
articles. After the summary and general conclusions, Volume
I is organized in four technical parts, i.e., Part I - Intro-
11
-------
duction, Definitions, History and Background Theory of the
Trickling Filter Process; Part II - Plant Design, Materials
of Construction, Operation, Maintenance, and Performance;
Part III - Trickling Filter Research and Development Ap-
proaches, Ecology, and Patents; and Part IV - Applications
of Trickling Filters to Specific Industrial Wastes.
Volume II is the bibliography of 5,665 quoted and additional
references in alphabetical sequence. The reference numbers
are enclosed by parentheses in the text of Volume I.
12
-------
PART I
INTRODUCTION, DEFINITIONS, HISTORY AND BACKGROUND
THEORY OF THE TRICKLING FILTER PROCESS
Part I of this review establishes a background on the subject
of biological trickling filters. After a brief introduction
of waste treatment, a simplified explanation of the trick-
ling filter process is given. The applications of trickling
filters are considered, along with the definitions used in
the literature and in the text. The history and development
of the trickling filter process are followed by the develop-
ment of the theory of the process. This introductory back-
ground defines the scope of the various factors which affect
the biological trickling filter.
13
-------
SECTION IV
INTRODUCTION
At the risk of over-simplification, the biological trickling
filter process may be described as follows: The sewage is
distributed over a suitable solid supporting structure. As
the waste stream falls, intimate mixing of the air, the waste
stream, and a microbial population (which is attached to the
supporting structure) results in an aerobic biochemical re-
action, reducing the organic waste. The trickling filter is
most commonly used after primary sedimentation has removed
settleable solids. Non-settleable solids and soluble organic
material are removed by an adsorption-oxidation phenomenon oc-
curring at the biofilm-waste stream interface. The effluent
from the trickling filter usually contains solids which are
generated by periodic or continuous sloughing of the bio-
logical film from the filter medium which was developed during
the aerobic oxidation. The effluent is then subjected to
further clarification, or other processes, which will handle
the solid-liquid separation.
The number of trickling filter installations in the United
States exceeds 4,5OO. Dosage rates may be from 25 to more
than 9,OOO gal./ft2/
-------
obtained from the waste stream being treated. The normal
life functions of respiration and synthesis of the organisms
convert the polluting materials to carbon dioxide and water
with the development of additional cell mass (biofilm or
sludge). Periodically, the cell mass or sludge detaches or
sloughs from the supporting medium and is separated from the
waste stream in a subsequent clarifier. The generalized
biochemical reaction occurring in the filter is oxidation,
and may be represented as:
(Organics) + (Air) + (Biofilm) Efflufint > (Need Clarifier) + Residual
Organics
and
Inorganics
CHONSP + 02 + Cells H£*° > More Cells + W03~ + C02 + H20 + Oxidized
Byproducts
The design, control, and operation of this seemingly straight-
forward method of treatment have occupied sanitary engineers
and interested scientists for over 9O years. There have
been, many types and arrangements of systems proposed, con-
structed, and operated (and some discarded) for the distri-
bution, reaction, and collection of the waste stream in
the evolution of an efficient biological reactor. The re-
lationships of the supporting medium, the biological film,
the waste stream, and the atmosphere, as illustrated in
Figure 1 (3917), have challenged the mathematical skill
of engineers and scientists for many years. Design equations
and formulations are currently being proposed to explain
some of these relationships.
APPLICABILITY OF THE PROCESS
The application of the trickling filter to wastewater treat-
ment has been found profitable in areas where: (a) personnel
may be limited (2204), (b) small flows exist, (c) an effluent
of from 20 to 3O mg/1 of BOD is acceptable, (d) partial treat-
ment is required in a multi-stage process, (e) land area
requirements dictate height to be increased to achieve the
designed reaction volume, (f) intermittent discharges of
toxic or inhibitory waste create shock conditions, and (g)
a specific treatment may be made on an industrial waste.
Besselievre (308) and Hoak (1964), as well as others, have
reported successful use of trickling filters on industrial
wastes, such as pharmaceutical, radioactive, high carbo-
hydrate such as that found in beet sugar plants, and brewery
wastes. Wastes occurring as a byproduct of coke, food
processing, chemical industries (2351), distilleries, meat
packing, pulp and paper mill, laundry, petrochemical and
petroleum, poultry processing, textile, tannery, and many
16
-------
others have also been treated by this method. The ability
of the trickling filter to sustain itself during temporary
changes in temperature, or organic content, or toxicity of
the influent has made it a valuable process in industrial
waste treatment. The trickling filter was frequently used
in the treatment of military waste (559O), as well as the
more conventional application for treating municipal waste
(1826, 4916) .
BIOLOGICAL FILM
r
ANAEROBIC
STONE
MEDIA
AEROBIC
ORGANIC ACIDS
WASTE LIQUID
AIR
ORGANIC
POLLUTANT
METABOLIC PRODUCTS
AND
EXCESS CELL GROWTH
Fig. 1 - Schematic Diagram of Trickling Filter Process
[by Permission of "Water and Sewage Works"
(3917, p. 100)]
-------
SECTION V
DEFINITIONS
During the evolution of the biological trickling filter on
an international scale, the process became known under many
names. Terms such as biofilter, percolating filter, sprin-
kling filter, sparkling filter, contact filter all refer to
the process generally known as the biological trickling fil-
ter. Throughout the text of this report, these terms will
be used interchangeably, but with the same meaning.
Based on studies by the American Society of Civil Engineers,
the National Research Council publications (3O53, 4916, 559O)
and the Great Lakes-Upper Mississippi River Board of State
Sanitary Engineers (5O11), terminology has been evolved and
recommended. The U. S. Federation of Sewage and Industrial
Wastes Associations (5588) proposed the standardization of
units of expression for design and operation of waste treat-
ment plants.
The biochemical oxygen demand (BOD5) is a five-day
test which has been used for several decades to eval-
uate the strength of the waste and the treatment plant
efficiency (352, 140O).
Depth of the filter medium is the distance in feet
above the underdrains. The minimum distance is five
feet and should not exceed seven feet.
The distributor of the trickling filter may be the
mechanical (rotary or traveling) type used commonly
today, or the fixed nozzle type as used in years
past, most of which were patented (224, 283O, 5574).
Dosing has been used to describe the cycle time
base that the waste stream is applied to the filter.
The ecology (1826) of the trickling filter relates
to viral particles as well as bacteria, fungi, algae,
protozoa and invertebrates.
Efficiency is the percent reduction of BOD of the
primary settling tank effluent by a single stage
trickling filter and the subsequent settling
. , . IQO x (Influent - Effluent)
tank; i.e., —^ r •
Influent
Media or medium refers to the supporting structure
on which microbial film develops such as rock, tile,
wood, plastic, etc.
19
-------
Nitrification is a phenomenon in which micro-
organisms (Nitrobacter, Nitromonas) oxidize ammonia
nitrogen to nitrite and later to nitrate. In the
early history of trickling filters the degree of
nitrification was used to measure the efficiency
of the process (658). with the development of the
BOD test, nitrification was used only as substanti-
ating information and in some cases, if nitrification
has begun, creates confusion (36O).
Pooling, ponding, clogging are terms used to de-
scribe blocking of sections of the media with
biological or other material, causing the filter
to short circuit, impairing the filter efficiency,
and creating nuisance conditions.
Post-treatment is the several unit processes after
the trickling filter treatment. The generated and
sloughed bio-mass and the treated liquid waste are
separated to improve the quality of the effluent
stream to be acceptable to the receiving body of
water.
Pretreatment refers to several processes preceding
the trickling filter. The principal function is
to materially reduce the load of suspended solids
applied to the filter, reduce odors, or other waste
conditioning prior to filtration.
Recirculation by pumping may be applied to the
filter effluent, the clarifier effuent, .the
clarifier sludge, and mixtures of effluent.
Several recirculation configurations have been
developed, some of which are patented (274, 5574).
Recirculation ratio, EL, is the ratio of the volume
of flow recirculated to the volume of the average
raw sewage.
Sloughing is the action of the biofilm releasing
from the supporting medium and subsequently ap-
pearing in the effluent from the trickling filter.
Specific surface area (ft2/ft3) defines the actual
surface area of the medium available for biofilm-
waste stream mixing per unit volume occupied by the
medium (2993).
Suspended solids of importance to trickling filters
are those solid materials which are carried over from
the primary clarifier and, after treatment through
the filter, are discharged to the secondary clarifier.
20
-------
Temperature of the waste as well as of the air
is of concern in trickling filter design and
operation.
Trickling filter loading has been expressed at
various times as million gallons a day; popula-
tion per acre or population per acre foot; pound
of BOD applied per acre foot, pound BOD applied
per cubic yard; pound BOD applied per 1,OOO cubic
feet; cubic foot per pound BOD applied, and pound
BOD per square foot of surface (1O29). To avoid
confusion, there are two kinds of loadings when
referred to trickling filters. One is the hydraulic
loading, or the volume of the waste water, and
the other is the organic loading, or the weight of
oxygen-consuming material in the waste water. Con-
fusion was also created in the units associated
with these two loadings. Consequently, the U. S.
Federation of Sewage and Industrial Wastes Associa-
tion (5588) recommended: hydraulic load - gallons
waste flow per square foot of media surface area
per day (gal./ftVday) r organic load - pounds BOD
per 1,OOO cubic foot per day of filter volume (Ib
BOD/1, OOO ft3/day) which was also endorsed by the
Ten States Standards (5O11).
•s
Underdrains are substructures which support the
filter medium and serve the dual function of
carrying away the treated waste and sloughed mate-
rial while acting also as ventilation ducts.
Ventilation of trickling filters is important to
assure an aerobic environment and may be achieved
by natural draft or forced air systems.
21
-------
SECTION VI
HISTORICAL DEVELOPMENT
The history and development of the "biological trickling fil-
ter may be undertaken with the assumption that an understand-
ing exists of the purpose of sewage treatment (3851, 4317).
Ancient and medieval waste disposal systems were comprised
of collection, some sedimentation, and often land disposal
filtration (1548). As populations become more concentrated,
not enough land was available for filtration or sewage farms
(589). Methods to artifically biologically treat waste were
sought. According to Halvorson (1667), the first trickling
filter, designed by Bailey Denton and built in Birmingham,
England, in 1871, used soil as the filtering medium rather
than rock. Expanded collection systems, such as that of
London (2091) , provided a challenge to the engineers to de-
velop treatment works which would handle the large quantity
of sewage. The application of chemistry and biology in the
treatment of waste water was demonstrated by Rawlinson and
others in England prior to the early 190O's (589, 6O2).
One of the early types of artificial treatment was the con-
tact bed (698, 2984), which was a tank filled with a broken
rock medium, intermittently submerged with sewage, allowed
to soak, and then was drained to rest. The time required
for this operation was usually on the order of two hours
filling, contact for two hours, then draining and resting
for six hours. In this manner, a four-foot deep bed, one-
half acre in area, could handle 5OO,OOO gallons/acre/day,
allowing for resting and maintenance. This rate of treatment
was too slow and required too much area, and the results of
this operation were sometimes a nuisance to the community.
Moreover, the reason for the contact or soaking period puzzled
the engineers for many years (1667).
The method of distributing the waste to be treated over the
medium has also been specifically reported (946). The first
installations of trickling filters with distribution by spray
nozzles were reported to be at Salford, England, by Joseph
Corbett in 1893 (2984) and at the Lawrence Experiment Station
in Massachusetts in 1891 (134, 69O, 1667). The Lawrence
Station has been credited with developing the use of the
coarse grained medium (698, 1166). There was much competition
for funds and manpower for the development of the biological
treatment processes by systems which used chemical and/or
mechanical means of treatment (13O).
Development efforts were further promoted through the estab-
lishment of organizations such as England's Royal Commission
on Sewage Disposal from 1898 to 1915 (589), the Lawrence
Experiment Station (691) in this country, the Imperial Board
of Health and Allied Scientific institutions, 19O8, in Germany
(4723), and later the Water Pollution Research Laboratory in
23
-------
England (5196), and the Robert A. Taft Sanitary Engineering
Center in the United States. Chase (658) commented in 1945
on the 35 years of previous experience using trickling
filters for waste treatment and the slow rate of development.
It was noted by Stanley (4187), as by Chase, that the sim-
plicity of operation of the trickling filter accounted for
its popularity and also for the lack of development of the
process. This period of arrested development was also noted
by- Imhoff (221O) . The trickling filter (1166) was suffi-
ciently popular that, in 194O, 58$ of all plants in the
United States providing secondary treatment utilized trick-
ling filters. Though the percent declined by 1957, the
number of trickling filter plants had increased to 2,682
and later to 3,5O6 in 1962 (85). The number of installations
were indicative of the municipal waste treatment plants
employing trickling filters. At most installations there
were at least two and sometimes as many as 20 or 3O trick-
ling filters. The trickling filter was common also at
military installations and various institutions (559O). As
noted in Figure 2, the rate of increase in the usage of
trickling filters and in the population served has increased
sporadically due to development influences, some of which
have been noted by the various investigators.
Appurtenances for trickling filters were reported (2449) based
on traditional uses and new intensive work which demonstrated
the performance of the biological trickling filter under
higher hydraulic loads (2OO to 1,OOO gal./ft2/day). In 1937,
Jenks (23O6) developed the theory and design for biofiltra-
tion. The use of the concept of higher hydraulic and organic
loading with recirculation by Montgomery (3O74), 3O77),
Nelson and Lanouette (3161), Stanbridge (4168), and Fischer
and Thompson (128O) resulted in a few patented processes.
With the development of the high-rate biological trickling
filters, design criteria were re-evaluated and formulations
derived. The National Research Council (559O) in 1946
proposed a mathematical formulation, and in 1948 Velz (4535)
reported equations relating the concentration of the waste
to physical parameters involved in the biological filtration
process to express the treatment efficiency. Mathematical
description of the trickling filter process and its various
modifications in the last decade and a half, e.g., Schulze
in 1960 (3917) and Kehrberger and Busch in 1969 (2419), have
occupied a considerable volume of literature.
The popularity of the trickling filter was increased by the
development of the high-rate filters and also by the neces-
sity of more accurate design work. Studies on retention
time in filters (3981), the effect of recirculation (2419)
and development of the alternating filtration system (1267,
3294) were among many of the problems attacked. Problems
of controlling filter flies (738, 14O7, I960, 3886), the
effect of medium (928, 1667, 27OO), and other maintenance
(1350, 4143) and operational (134, 1O28, 22OO, 5399) prob-
lems were also investigated at length.
24
-------
CO
*1
ui co 2
JC
o
V)
i
p
t t
t I JJ
« t rt U»_
x^*"* POPULATION
210
trt
ISO d
sol
30
1893
1914
1930
(YEAR)
1946
1962
Examples of Typical Influence Factors
a - Salford, England installation
b - Madison, Wisconsin installation
c - Rotating distributors used in U. S.
d - Activated sludge popularized
e - BOD test verified
f - High rate filter developments
g - Military application of high-rate
h - National Research Council report
i - Velz performance relationship
j - Plastic media used
k - Stack and Howland developments
1 - Bloodgood influences
m - Schulze investigations
n - Behn evaluations
o - Caller and Gotaas studies
p _ 1899-Rivers and Harbors Act (33 U.S.C. 4O7)
q - 1912-Public Health Service Act (P.L. 62-265)
r - 1924-Oil Pollution Act (P.L. 68-238)
s - 1948-Water Pollution Control Act (P.L. 80-845)
t - 1956-Federal Water Pollution Control Act (P.L. 84-660)
u - 1961-Amendments to the Federal Water Pollution Control
Act (P.L. 87-88)
v - 1965-Water Quality Act (P.L. 89-234)
w - 1966-Clean Water Restoration Act (P.L. 89-753)
Fig. 2 - Factors Influencing the Use of Trickling Filters
25
-------
Biological trickling filters have found application inter-
nationally. Imhoff (221O), Ehlgotz (1124) and Husmann (2136)
used this process in Germany and developed it to a high de-
gree (relative to the United States) in the 1930's. In
England, Hurley (211O) in 1938 experimented with the basic
process and cited the principal advantages of the activated-
sludge process and of percolating filters (1199). Datesman
(875) reported the advanced state-of-the-art of percolating
filters. Investigations (5196) have been carried out to
obtain valuable information to optimize this process, partic-
ularly with plastic media. Trickling filters have been op-
erated in the Netherlands (2O24, 2458), in Australia (1O31),
in New Zealand (3704, 37O7), India (773, 4263), many in South
Africa (1685, 2958, 4741), Poland (147), Russia (4254, 484O,
5448) , Argentina (189) , France (3955) , Holland (2455) ,
Uruguay (3842), El Salvador (63O), and Malaysia (5O87), to
mention just some of the countries. Due to the low oper-
ating personnel requirements, developing nations, with their
finances being strained for other purposes, have found many
times that the trickling filter was an economic solution to
their sewage problems.
26
-------
SECTION VII
BACKGROUND THEORY OF PROCESS
This section briefly explores several topics dealing with
the theoretical aspects of the biological trickling filter
process. Theoretical and operational variables and their
importance have been reported in the literature in a random
manner. Where applicable, numerical limits or ranges of the
variables are noted. Subsections are critiqued. This sec-
tion should provide an insight into the thoughts of the in-
vestigators during the development of the theory of biologi-
cal trickling filters.
An in-depth, we11-documented theory of biological trickling
filters has not been generally accepted (363), and continu-
ing discussions of the inadequacies of the theories reflect
the investigators' desires to develop a theory acceptable
to all (2965, 3915) . Reports are still being published
dealing with various factors and their effect on the theory
of biological trickling filters.
The development of the theory appears to have begun with
rather general thoughts and observations. It was observed
by workers, such as Herring (1898) in 19O8, that contact
surface area, suspended solids, air flow through the tower
or bed, liquid contact time, temperature and distribution
of the waste material over the filter^ affected the operation
of the trickling filters.
»
With the need for construction of sewage treatment plants
during World War II, design expressions for predicting the
performance of biological trickling filters required adapting
the theory into mathematical form. By observing the qualita-
tive significance of each of the many variables, independent
investigators developed several diverse opinions. Many
texts summarize the existing theory of biological trickling
filters, typically Metcalf and Eddy (2984), Pearse (3327),
Babbitt and Baumann (112), Eckenfelder and O'Connor (1O8O),
and Rich (3586), to list just a few. The information in-
corporated into these texts was basically derived from papers
similar to those of Herring (1898), Husmann (2145), and Bach-
mann (134). Much of the early experience was reported in
general terms and only occasionally were numerical values
attached to trickling filter variables.
Quite often the theory was reported as introductory informa-
tion (1O75, 2663) on some specific aspect of the operational
problems of waste treatment systems employing trickling fil-
ters. Reviews, e.g., Kleeck (25O4) and Lohmeyer (2756),
referred to the developed theory of trickling filters while
pointing out operational factors in papers aimed at waste
treatment plant personnel. Busch (549) in 1968 criticized
the theory of the biological trickling filter and suggested
27
-------
viewing the system as a film flow reactor (similarly viewing
the activated-sludge process as a fluidized system) under
aerobic or anaerobic conditions, then using further theoreti-
cal refinements.
The separation of theory from the mathematical investigation
is difficult after the efforts of a National Research Coun-
cil (559O) and the development of their formulation. In-
vestigators, such as Velz (4535), Fairall (1232), Stack (4157),
Behn (246), Rowland et al. (2O65), Schulze (3917), Caller
and Gotaas (1395), Germain (146O) , Kornegay et al. (2547),
Mehta et al. (2955), Robertson et al. (3647), Roesler et al.
(3658), Kehrberger and Busch (2419), and Archer et al. (85)
were very much interested in the proper applications of trick-
ling filter theory, but the emphasis of their work was math-
ematical, as will be discussed later in corresponding sec-
tions. The significance in noting their work here is that,
with the availability of the extensive data from the National
Research Council's investigations (559O) plus the increased
usage of trickling filters, these investigators were moti-
vated to include their observations in the form of mathemat-
ical relations. In general, the same variables were being
considered as those which were outlined in the early 190O's.
However, the problem was now being critically investigated
by determining the most significant of the variables.
To aid in the discussion of the theory of biological trick-
ling filters, a schematic view of a trickling filter and the
relationship of the many variables used in the theory rela-
tive to the filter structure are shown in Figure 3. That
portion of the theory to be reviewed here was organized into
the following categories: (a) the waste characteristics of
organic strength and suspended solids, and their influences,
(b) contact time, hydraulic dosing and load and their rela-
tionship on the filter behavior, (c) the filter medium,
depth, surface area and effectiveness, (d) the underdrains
and enclosures such as surrounding walls and roofs and this
relation to ventilation, and their effect on performance,
(e) the use of clarifiers and/or recirculation of the ef-
fluent and sludge, (f) ventilation and the oxygen content,
and (g) interfacial factors relating to the biological film
and the various transfer characteristics.
ORGANIC STRENGTH AND SUSPENDED SOLIDS
The influent quality of the waste being applied to the
trickling filter as a function of organic and suspended
solids content was recognized by Herring (1898) in 19O8 in
the term he described as "liquidity." Buswell et al. (555)
were concerned with the removal of colloids from sewage
and showed that this removal process was biochemical in
28
-------
FOR FIXED NOZZLES)
INFLUENT _J
1
!
l
>
DOSING
•
AIR
OXYGEN OR AIR
NATURAL OR FORCED
COVER
\H-ENCLOSURE
WALLS
BIOFILM
MEDIUM
INTER FACIAL FACTORS OF
FILM AND ECOLOGY
EFFLUENT TO
CLARIFIER OR
RECYCLE
UNDERDRAINS
Fig. 3 - Schematic of Theory
-------
nature. P6*nninger (3413), in developing calculations for
sizing trickling filters, demonstrated that the strength of
the waste and the amount of suspended material affected the
efficiency of the trickling filter. In a laboratory investi-
gation, Atkinson and Swilley (1O1) indicated that suspended
and dissolved salts reduced the efficiency of a trickling
filter. This effect was thought to be due to the reduction
of the active area of the organism-liquid interface by entrap-
ment of solid particles or by adsorption (1O1).
McLachlan (2928) reported that the main action of the process
of purification by the trickling filters was not by combus-
tion of the impurities to carbon dioxide, but rather a trans-
ference of the polluted matter from colloidal suspension to
a coagulated condition. This point was supported by Levine
(27O4), who noted that the evaluation of the trickling filter
efficiency solely on the basis of biological activity does
not completely consider the physico-chemical coagulation
properties of the process.
Popel (3428) also recognized the effect of the organic
strength of the influent waste, determined as the biochemi-
cal oxygen demand (BOD). He found that organic strength is
significant in determining the efficiency of the filter when
he combined this with other factors into a simplified for-
mula for evaluating the process. Velz (4535) reported that
the efficiency of the trickling filter was proportional to
the remaining concentration of the organic matter and this
factor has been the basis of many later mathematical de-
scriptions of efficiency.
While evaluating deep bed trickling filters, Burgess etal. (536)
investigated varying organic strengths and hydraulic load-
ings on trickling filters. Data were obtained with an ex-
perimental filter on reduction of BOD and on nitrification
which supported Velz's (4535) basic law. However, with the
weak sewages examined, high hydraulic and organic loadings
gave results much lower than those predicted by the theoret-
ical formula. Meltzer (2964) concluded, during an investi-
gation of contact time, that the increase in purification
efficiency obtained with multiple stage filters, deep fil-
ters, and recirculation will be less with weak sewages than
with strong sewages.
In developing a mathematical model, Kornegay and Andrews
(2547) assumed that removal of organics is described by a
saturation function which incorporates the effect of diffusion
and growth rate. The effect of organic concentration and
suspended solids is covered in this paper (2547). In analyz-
ing trickling filter variables from a mathematical standpoint,
Galler and Gotaas (1395) used organic loading as one of the
major variables. They contended that the removal of BOD is
affected most by the BOD in the influent as well as to a
significant extent by recirculation, temperature, and depth.
30
-------
Biczysko (322, 323) in 1965 indicated that, within a certain
range of the load of organic matter, the efficiency of the
biological process is independent of this load; when this
load exceeds a certain critical value, anaerobic processes
are responsible for the work of the filter. In developing
a background for the use of plastic media, Pearson (3328)
outlined factors affecting the design and operation of high-
rate roughing filters, one of which is the organic content
of the waste.
Critique
The involvement of suspended solids and organic content or
strength of the sewage was recognized early in the develop-
ment of the trickling filter theory. As the mathematical
model was developed to consider the effect of these factors
on the efficiency of trickling filters, divergent opinions
were expressed. An example of the conflict of the theory is
typified by the statement by Caller and Gotaas (1395), " . .
it was found that the hydraulic rate did not contribute any
significant effect to the BOD removal efficiency," in con-
trast to the statement by Schulze (3917)," ... these re-
sults clearly demonstrated the hydraulic load and depth are
two of the major factors determining the efficiency of trick-
ling filters." Agreement and disagreement are reported with
each of the views and the corresponding views on the in-
fluence of the strength of the sewage. Some light may be
shed upon the matter if one relates the Caller and Gotaas
approach to that of treating municipal waste primarily and
that of Schulze to treating industrial waste principally.
Caller and Gotaas made the comment that, "high hydraulic
rates are indicative of high organic loading" (probably true
for" municipal waste), while Schulze made the remark that
"organic load can be changed without any change in hydraulic
load" (definitely true for industrial waste). It would ap-
pear then that the 322 field observations of Caller and
Gotaas and the controlled model studies with field data
applied by Schulze may actually support one another's con-
clusions.
CONTACT TIME, HYDRAULIC DOSING AND LOAD
An awareness of the significance of the retention time
(contact time) of a waste in the biological trickling filter
was shown by Herring (1898) in 19O8. It was reported in
England that 1OO minutes of contact would give a good quality,
nonputrescible effluent (1898). Blunk (367) and Ponninger
(3406) , using the salt concentration in the effluent tracer
technique, determined that the contact time varied from one
to five hours depending upon the type and size of material
and the amount of film which had developed on the filter sur-
face. Laboratory and pilot plant studies were used by
Sheikh in 1966 (3981) to determine that the most suitable
31
-------
tracer was the radio-isotope Br82 in preference to the salt
technique used above. PSnninger (3406) suggested the follow-
ing retention hypothesis: each drop of water falling on a
particle of the filtering medium spreads over the whole
particle and penetrates the biofilm, mixing with the liquid
already in the film; liquid is forced out of the film, col-
lects in a drop on the underside and falls on the particle
below. This explanation of the long detention times in the
trickling filter was challenged by Stene (42O7).
Contact time has been usually represented as the mean reten-
tion time as determined from a plot of the dye or salt
tracer concentration in the effluent of the filter, as used
by Blunk (367), P5nninger (3406) and Sinkoff et al. (4042).
Sheikh noted (3981) that the mean retention time, as used
by many workers (2O64, 2963), was an inadequate parameter
to define the retention characteristics of the filter, and
suggested that it be replaced by the modal time and the stan-
dard deviation of the log normal distribution to compare and
derive expressions for filter performance. Meltzer (2963)
made an effort to relate retention time in filters with the
rate of flow, the size of the medium, and the height of the
filter. Howland (2O64) proposed a theory, based on experi-
mentation, which indicated that the degree of purification
obtained depended upon the time of flow of sewage over the
slime-holding surfaces on the porous media of a trickling
filter. This theory implied that the time of travel through
the filter varies inversely with the two-thirds power "rate
of flow per unit width of plate." Escritt (1208), in pre-
paring a paper on general considerations, stated that it was
important to achieve a sufficient retention period for the
organisms to effect oxidation. However, in contrast, Levine
(27O4) minimized the importance of oxidation. Escritt (12O8)
also reported that many investigators were of the opinion
that an optimum condition was achieved "when the rate of
filter loading was in the region of 9OO and 1,OOO gallons
per day per superficial yard of the top of the bed" (1OO
gal./ft2/day) .
In laboratory-scale experiments, Bloodgood et al. (364)
determined that the weight of biofilm on a trickling filter
had no clear relationship to the waste contact time. Their
studies demonstrated that the contact time on the laboratory
devices was inversely proportional to the two-thirds power
of the liquid application rate. Eckenfelder et al. (1083),
while investigating retention times on plastic media, re-
ported that the contact time is dependent upon hydraulic
loading and distribution of the biological film. They con-
cluded that the presence of the biological film considerably
increased the mean residence time. Howland et al. (2O65)
investigated the capillary effect of water and the contact
time on a laboratory simulation of trickling filters.
Schulze (3915) proposed that the hydraulic load determined
32
-------
the period of contact and that this in turn determined the
efficiency of treatment. He showed that the efficiency was
inversely proportional to the 0.67 power of the flow, in
agreement with Rowland, Schulze, and Bloodgood et al. It
was noted by Sheikh (3981) that the modal residence time was
inversely proportional to the O.7 power of the flow, which
was in fair agreement with the values of 0.67 proposed by
previously mentioned investigators.
Since the flow path of the waste through a trickling filter
could be very tortuous, Meltzer (2964) proposed that a true
Gaussian normal distribution curve would result when the
lengths of the path were plotted against their number. The
validity of this theory was experimentally verified and he
further suggested that it is also an explanation of the dif-
ference of opinion held by many workers regarding the effect
or organic strength and/or hydraulic load upon the effi-
ciency of trickling filters. Mathematical expressions were
derived by Sheikh (3981) for different sizes of medium, and
the removal of organic matter was reported to occur in three
stages along the depth of the trickling filter with up to
93$ removed in about 30 minutes in the first foot of the
filter. The liquid retention time for the same depth was
98.4 minutes, suggesting that a coarser medium in the first
foot of the filter could be used. He also concluded that
performance data obtained from pilot-scale filters can be
applied to full-scale filters only if all the operating
conditions of the two filters (depth, medium, and hydraulic
surface loading) are identical. Of primary importance is
the time of contact as reported by Biczysko (322, 323) in
relationship to other variables, such as temperature, pH,
type of medium, and amount of air supplied.
As noted historically, work has been reported on the fre-
quency of dosing. Francis (1335) stated that by 1930 it was
well established that the efficiency of the biological treat-
ment depends on the closeness, frequency, and duration of the
interfacial contact between the biologically active medium
and the sewage with adequate available air. There was con-
siderable debate dealing with intermittent versus continuous
dosage, and Halvorson (1667) in 1936 indicated the trend was
toward continuous dosage. Halvorson also supported the posi-
tion of continuous dosing and indicated that rest periods are
not justified for operation of properly designed trickling
filters and that long rest periods are actually harmful to
the performance of the unit. However, Levine (27O4) investi-
gated the various sizes of medium and dosing cycles which
would produce more effective purification. With dosing cy-
cles of 2.5 minutes, the rate of discharge was uniform and
was approximately equal to the rate of application. A much
greater reduction in BOD than normal was effected when the
rate of outflow from the filter was uniform. Hydraulic dis-
tribution and its effect on the operation of high rate fil-
ters were studied by Kehr and Ruchhoft (2418), who found no
33
-------
significant difference in the efficiency due to rate, and by
Jenks (23O5), who stated that up to a recirculation ratio of
six the reduction of BOD in Ib/lOOO ft3/day of filter is
practically proportional to the ratio.
Concepts such as minimum contact time and continuous dosage
were recognized and employed by Jenks (23O6), Halvorson
(1667), and Levine (27O4) as they developed and built high-
rate biological trickling filters. Lanz (2629), in an effort
to determine the highest allowable hydraulic loading rate,
showed that 3O.5 m3/m2/day of settled sewage per filter sur-
face per day (32.6 mgad or 77O gal./ft2/day) with sufficient
aeration produced a nonputrescible effluent. However, the
efficiency of the filter decreased with the increasing loads.
The term high-rate and high-capacity developed some contro-
versy and confusion, as reported by Montgomery (3O71). He
specified that the total quantity of sewage through a filter
in a day, or better still the total quantity of BOD handled
per acre-foot per day was described by its "capacity" where-
as the "rate" of operation is "the quantity of sewage it is
capable of handling per second." However, the application
rate per day (gal./ft2/day) should be considered separately
from the instantaneous hydraulic load, which could be a
steady "rain-like" distribution or a heavy intermittent dose
from the distribution box. Montgomery's comments had direct
bearing on the type of distributor which was being used on
the various high rate trickling filter installations, such
as the "Aero-Filter" (Halvorson and Smith) and the "Bio-
Filter" (Jenks), which were classified, respectively, as low
rate with high capacity and high rate with low capacity.
Covill (8O7) in 1942 used the high capacity designation in a
discussion of trickling filter and other types of biological
treatment. He concluded that high capacity filtration in
its various forms offers distinct advantages in lower con-
struction costs, reduction in odor control and fly nuisance,
nonclogging of filter media, and production of an effluent
(two-stage treatment) equal to that from an activated-sludge
process.
In addition to hydraulic rates being increased, other modi-
fications were described by Tomlinson (4422), such as two-
stage operation of trickling filters with the order of their
dosing being alternately reversed. This alternating double
filtration system could treat 2.5 times as much sewage as a
single filtration system with no particular effluent quality
degradation as judged by the BOD test. In 1958 Nelson and
Lanouette (3161) summarized the investigations leading up
to the design of the Bio-Filter and described the basic
principles of the other patented high-rate treatment methods
such as the Ward, the Aero-Filter, and the Accelo-Filter
methods. Formulae derived from the National Research Council
in 1946 for interpreting results of filter performance were
34
-------
applied to single- and two-stage filter design and the rela-
tionships illustrated.
Later, Meltzer (2965) agreed with Montgomery (3071) that bio-
logical filter efficiency was considered to be a function of
the "intensity of hydraulic surface loading rate." Maximum
efficiency occurs at an optimum rate of flow which is spe-
cific to the type, size and configuration of the medium.
There was no evidence that two-stage filtration or recircula-
tion of the effluent would be any more efficient than single
filtration with deep filters. However, Sorrels and Zeller
(4111) reported on the advantageous use of two-stage trickling
filters without intermediate sedimentation, giving a higher
degree of purification than single-stage filtration, even at
higher hydraulic load. The primary filters were heavily
loaded hydraulically and organically without recirculation
and were actually operated as roughing filters.
Critique
The use of the mean residence time to represent the contact
time in the filter has been shown to be inferior to the use
of the modal time. However, the contact times were very
closely related to the inverse of the hydraulic load raised
to the two-thirds power (2O64, 3915, 3981). Comments (364)
on the relationship of the biofilm to retention time appear
initially to be in conflict. However, one investigator
(1083) noted distribution of the biofilm, and the other
(364) was measuring total film accumulation. The hydraulic
considerations are noted here to demonstrate the awareness
of the investigators. General practice has evolved from the
use of intermittent discharge at high rates to continuous
discharge at low rates. The latter hydraulic application
rates were based upon a better understanding of the signif-
icance of the retention time or contact time in the filters.
The concepts of high-capacity and high-rate very clearly
identify the limitation of the early designed trickling fil-
ters. In the Design Section (Part II) on hydraulic loading,
it will be noted that much higher hydraulic loadings were
made possible by the use of fabricated media.
FILTER MEDIUM
The principle behind the use of a filter medium was to sup-
ply an inert supporting structure to which the biological
growth can adhere. It has been reported (2816) that some
slag media exhibit a sorptive property which was desirable
for industrial waste treatment.
Herring (1898), Husmann (2145), and Kleeck (2504), with
5O years of combined experience, emphasized the importance
of the available superficial or contact area of the medium.
P6*pel (3428) and Meltzer (2965) identified the significance
35
-------
of surface area characteristics of the medium in their re-
spective theories and formulations. A finer grained medium
was reputed (1667) to have greater surface area, but also
would clog more easily. Recognizing that a deeper, more
porous medium was important, investigators, such as Halvor-
son (1667) in 1936, used fabricated media such as tile.
Pearson (3328), almost thirty years later, used plastic
media which were designed for greater bed depth, greater
surface area, greater void space, and other desirable opera-
tional factors. Comparisons of media have been made in
several studies (1O83, 2992, 4463).
Strickhouser et al. (4252) and Walton (4596) have demon-
strated the importance of the depth of the medium. They
stated (4252) that, theoretically, a greater depth of the
filter resulted in greater rate of purification. However,
practical considerations of the medium, due to trouble with
ventilation and sloughing, limited the depth of the filters.
Walton (4596) claimed, in 195O, that the quality of the
filter effluent is dependent upon the depth of the filter
medium rather than its volume. Biczysko (323) and others
(536) reported that the bed height or depth was significant
in determining the efficiency of the trickling filter.
Velz's formula (4535) included a parameter for the depth of
the filter.
Critique
Practically, the design criteria required of a medium for a
trickling filter operation are covered in Part II. The
workers listed above were aware that the medium should be
fairly inert to biological and chemical interactions of the
sewage effluent with only minor changes over a long period
of time. The medium should have adequate space for ventila-
tion and sloughing of solids. The function of the depth was
not clearly understood for some time, but it should be noted
that recirculation and two-stage applications were used so
that, in effect, deep beds were in operation for a long time.
Later, oxidation towers tens of feet high have been used with
and without recirculation. The conservative nature of the
engineers and scientists protecting the public forced them to
rely on media which had been proven over the years. Media
other than rock, such as tile and plastics, are being used
at an increasing rate as the result of extensive efforts of
media fabricators. With better understanding of the desir-
able medium properties, additional materials and designs
will undoubtedly be used.
UNDERDRAINS AND ENVIRONMENTAL CONTROL
The underdrain systems, enclosures of wall and roof, the
temperature control, and nuisance control are all very
closely related. The necessity of providing properly sized
36
-------
underdrains to carry the flow and supply necessary air to the
filter was known by Corbett (787) and Halvorson (1667) and
was well established in the early 1900's (1898). The Trick-
ling Filter Floor Institute (5574), in their technical in-
formation bulletins, published the results of several in-
vestigators along with detailed descriptions and illustra-
tions of underdrain systems. Much of this work is based
upon acceptable practice and is therefore reviewed accordingly.
By putting walls on the trickling filter the fly larvae could
be drowned by flooding (1667). Lohmeyer (2756), in a review
of the operation of trickling filters, discussed the type and
sizing of the underdrain system and the use of walls enclosing
the trickling filter. Complete enclosure of the filter, such
as was reported by Husmann (2145) as a means of controlling
odors and fly nuisances, was not uncommon.
The influence of temperature, as investigated by Lanz (2629),
indicated results which could not be generally agreed upon,
e.g., the temperature of the sewage remains practically un-
changed during the purification process and no noticeable
cooling occurs in the winter. However, Petru (3342), in
discussing the work of many investigators, observed that
some cooling occurred in the effluent from the filters and
this cooling was affected mostly by the individual filter
medium characteristics. Rowland (2O63) proposed a mathe-
matical relationship which considered the effect of tempera-
ture on the biological filter efficiency. The temperature
effects were investigated by Schroepfer et al. (3903); how-
ever, the effect of temperature on the operation of a filter
was acknowledged by Biczysko (323) as being only secondary.
Critique
Brief mention of temperature, nuisance, underdrains, and en-
closures was made to demonstrate the awareness and concern
of the investigators. Information dealing vith the specific
application of these aspects to problems is provided in Part
II of this report. The effect of temperature is, at best,
not fully developed but sufficient work has been done to in-
clude it. Walls surrounding trickling filters are in common
use, and problems dealing with ventilation blockage were
solved. However, with the advent of high rate trickling
filters and fabricated media, the necessity for flooding was
not as great. Therefore, structural walls have become less
important. Nuisance control dealing with odor, flies, and
ice formation has been handled quite well in specific locali-
ties by chemical and/or enclosure techniques. Enclosure of
trickling filters for nuisance control is being continued
due to increased public attention.
37
-------
SECONDARY CLARIFICATION AND RECIRCULATION
The theory behind secondary clarification after the trick-
ling filter and recirculation has progressed through several
stages of development. Clarifiers as part of the trickling
filter process were routine prior to 19OO (1898, 25O4, 2925).
Since the biological trickling filter mechanism was thought
to be adsorption of colloidal material and complexing soluble
organics into settleable solids, there were only a few in-
stallations built without secondary clarifiers. Kleeck
(2381) presented a review of the performance of clarifiers.
Sorrels and Zeller (4111) and others (85, 5574) were interest-
ed in operating two-stage filtration plants without inter-
mediate clarification. It was reported (85) that in the
operation of a two-stage system the intermediate clarifier
had no effect on overall efficiency. In most instances, it
was found highly desirable to follow the second trickling
filter operation with a clarifier (853, 463O). Fischer (1286)
supplied data that were useful in design evaluation of clar-
ifiers and filters.
Recirculation of trickling filter effluent was utilized
prior to 1940 (3406, 4207). Newly developed accelerated
methods of biological treatment took advantage of recircula-
tion (2544). Meltzer (2963, 2964, 2965) described work
covering ten years which was indicative of the continued
interest expressed by several investigators in the use and
effectiveness of recirculation. His studies led him to
believe that an increase in treatment efficiency due to re-
circulation actually does not exist when the total system is
considered. However, Benzie (274) stated that very good ef-
fluents from high-rate filtering plants have been obtained
at recirculation ratios of from O.5 to 1. In the development
of the trickling filter theory, the National Research Council
(559O), Caller and Gotaas (1395), and others (2064) included
recirculation in addition to other variables.
According to Galler and Gotaas, recirculation of effluent can
improve the quality of the final effluent discharged to a
certain extent, but with more than four volumes of recircu-
lated liquid to one of plant influent there is no improvement.
Dreier (1O29) strongly recommended the use of recirculation
based on system evaluations. The only effect recirculation
had, as claimed by Germain (146O), was that one was able to
maintain a constant rate of dosing. Studies at Mississippi
State University (85) using the National Research Council
formula indicated that recirculation improved overall ef-
ficiency of the plant. Leibee et al. (267O) found that for
best filter efficiency a minimum as well as a maximum loading
must be considered, i.e., controlled recirculation. Very
recently, Kehrberger and Busch (2419), using mathematical
models to describe the effect of recirculation, reported the
results of a mathematical analysis and laboratory experiments
38
-------
of the effect of recirculation on removal of total soluble
organic carbon. Two of their conclusions were: "... (A)
For heterogeneous film flow reactors, recirculation has a
detrimental effect on removal of total soluble carbon; and
(B) for pseudo-homogeneous film flow reactors, recirculation
has a beneficial effect on removal of total soluble organic
carbon ..." Pseudo-homogeneous film flow is defined as a
film flow system in which liquid entering the reactor flows
through or contains the microbial mass and reacts at all
points in the liquid phase. Data are given for a glucose
system from which they draw the cited conclusions. Roesler
et al. (3658) prepared a mathematical model for recircula-
tion, in which the cost of recycling was also included.
Critique
The use of clarifiers in various stages of the process has
been adequately investigated. As a design consideration,
intermediate clarifiers may or may not be economically used.
The solids balance of the process and waste are prime fac-
tors. Theory is well established for the solids-liquid
separation. Design standards have been reasonably estab-
lished (5011, 5331) .
The advantage of recirculation, however, has not been well
established from a theoretical basis and the search for a
solution of the problem has gone from laboratory study to
field empiricism and back to laboratory study. It would ap-
pear that Kehrberger's work (2419) lends definite support
to the diverse mathematical and experimental effect of re-
circulation. By the definition of his film flow model, his
conclusions imply that recycle of clarified effluent degrades
the treatment plant efficiency, whereas the recycling of
sludge and some effluent improves the efficiency of the
trickling filter process. This is of interest, if correctly
interpreted, because recycling of sludge into the active bio-
logical reaction system, such as the trickling filter, would
cause more harm than good due to the complexed organics in
the settleable solids being released in a soluble form and
possibly escaping via degradation. It has also been reported
that recycling of effluent dilutes the waste, thereby dropping
efficiency. It will be most interesting when Kehrberger's
conclusions are put into practice with current process tech-
nology and evaluated under field conditions.
VENTILATION OF FILTER
Early in the 19OO1s and before, Nasmith (3147), Corbett (787)
(1667), and others (1626) explained the principles of aero-
bic treatment and the necessity for adequate availability of
oxygen in the biological trickling filters. Levine (2698)
studied the effect of bottom ventilation on the trickling
filters and found that while the mechanism was not explained
39
-------
there was a definite detrimental effect when ventilation was
blocked. Tatham (4296) stated that a measure of pollution
in a waste liquor is dependent upon the oxygen required for
its complete oxidation and that the rate of absorption of
the dissolved oxygen is proportional to the concentration
of the polluting material. He then suggested an "avidity"
constant, a numerical measure of the activity of the bio-
logical oxidation, which is dependent upon the type of waste
being treated and the circumstances under which oxidation is
taking place.
Measurements of the content of carbon dioxide in the gas
phase were made (2629) and it was demonstrated that the dis-
solved oxygen content of the sewage increased slowly as the
waste passed through the filter. However, a tour of a number
of European waste treatment plants by Rudolphs (3741) prompted
the observation that the results of the operations of some
sewage treatment plants did not support the theory that the
biological activities in the trickling filters are limited
by the oxygen content and that the carbon dioxide concentra-
tion in the lower parts of the filter is toxic to the
oxidizing bacteria. McLachlan (2928) added further support
to these observations by discrediting the effect of carbon
dioxide in the biological trickling filter process. However,
concern was still expressed by Husmann (2145) that not enough
oxygen was available in normal ventilation of trickling fil-
ters to maintain the activity of the aerobic bacteria. Since
the upper part of the trickling filter exerted the most bio-
logical activity, he suggested that the efficiency of the
filters could be possibly raised by increasing the available
oxygen by artificial aeration.
Halvorson (1667) reported that forced aeration is essential
during seasons when the air and the water temperature are
approximately equal and the ventilation of the trickling
filters is at a very slow rate. Petru (3342) took exception
to this view on aeration in trickling filters and suggested
that the most favorable conditions for movement of air occur
near the surface, which places shallow filters at an advan-
tage over deep filters.
Heilmann (1860) described investigations and development of
artificial aeration or ventilation for trickling filters as
a method of controlling nuisances and securing better opera-
tion. The case of limited space and a required degree of
treatment was outlined by Kominek (2544), in which he sug-
gested artificial aeration as a method of accelerated waste
treatment with the possibility of using oxygen instead of
air. Sufficient surface area of liquid exposed to the air
plus sufficient turbulence of the liquid to produce the de-
gree of aeration needed were factors described by Escritt
(1208) as being necessary to effect oxidation of organic
material.
40
-------
One of the factors that Biczysko (323) cited as being of some
significance was the air supplied to the Bio-Filter. Pearson
(3328) stated that air velocities of 6 ft/min are more than
adequate to supply the oxygen requirement of filters with
50$ void space. Eckenfelder et al. (1083), in studying the
performance of fabricated media, reported oxygen transfer
capacities of O.O162 Ib/hr/ft at a hydraulic load of 3 gal./
min/ft2 (4320 gal./ft2/day). Confidence in the ventilation
properties of existing technique was expressed (5574) and
documented with full-scale operational data. The deeper bed
installations have been shown to generate sufficient draft
for ventilation (4330, 5637). Problems with high organic
wastes from acid fermentation of molasses have been reported
by Krige (2571), in which normal aeration of the molasses
slop in admixture with sewage before filtration was not
satisfactory.
Critique
Perusal of the literature dramatizes the conflicting evi-
dence in the effect of ventilation on the efficiency of
trickling filters. However, it is common practice today to
build deep bed filters operating under natural draft condi-
tions. Exceptions arise; if an extremely high oxygen con-
suming organic waste is being treated and the filter is
acting as a roughing tower, then forced ventilation may be
used. Even the use of enclosed filters does not necessarily
mean that forced ventilation is necessary. It is recommended
that the underdrain system must be designed and operated to
flow only partially full, allowing free access for air move-
ment above the liquid level. Internal and external tempera-
ture differences will usually generate sufficient air move-
ment to supply adequate oxygen under normal conditions.
Additional information on ventilation is reviewed in Part II,
"Design and Construction."
INTERFACIAL BIOFILM INVESTIGATIONS
Knowledge of the reactions occurring on the surface of the
medium, the biological film and the waste stream at what one
might call the interface (refer to Figure 3), was clearly
demonstrated prior to 1911. Nasmith in 1911 (3147) described
the reactions which were taking place in the trickling fil-
ter as chemical, thus organic matter plus oxygen give inor-
ganic matter and humus. In a discussion of Francis1 work
(1335), Harris pointed out that the important difference
which exists in trickling filters versus other biological
processes is that each stratum of a percolating filter con-
tinuously performs the same biological functions. Rudolfs
and Gehm (3755), discussing colloids in sewage and sewage
treatment, suggested that aerobic biological treatments,
e.g. treatment by the activated-sludge process and in trick-
ling filters, depend on surface activity of the gels of
41
-------
biological origin which are formed. The action may be due to
surface attraction, formation of adsorption compounds, or
concentration of food materials for bacterial growth. Ad-
sorption of soluble matter may be due to base exchange, and
a theory of adsorption of dispersed matter by neutralization
of electric charges was advanced. This adsorption theory
was used to account for the rapid removal of soluble matter
and was supported by bacteriologically identifying the micro-
organisms (555). Similar to this work, about 4O years later,
cultures of Zoogloea were shown (1775) to coalesce with small
floes, suggesting a mechanism of colloidal entrapment by this
life form.
Blunk suggested (367) that biological treatment in the trick-
ling filters occurred by liquid displacement in the film
layer, similar to the hypothesis postulated by PGnninger (3406) ,
and that the Paramecia activity of the biological film was
an indication of the biological activity available. Col-
loidal, biological, and enzymatic theories were discussed
by Theriault (4353) to explain the phenomenon which he called
"clarification." This "clarification" was used to denote the
overall improvement which results from treatment of sewage
for a brief period, but does not provide complete removal of
turbidity in the 3O- to 4O-minute time limit.
Several investigators (364, 12O8, 2064) demonstrated an appre-
ciation of the surface flow regime and the biological film-
waste interface as they function in the overall removal of
organic waste. The influence of turbulence on bacterial
slimes was discussed by Hartmann (1775) who showed that
turbulence is important in the transport of food molecules
from the environment to the surface of bacterial cells. An
attempt was made by Swilley et al. (4277) to use transport
equations to describe simplified biological oxidation sys-
tems.
In the development of the feasibility of using synthetic
plastic materials as a filter medium, Pearson (3328) indi-
cated six basic steps as a biochemical basis of biological
flocculation which included: (a) adsorption of dissolved
and colloidal organic matter to the outside biofilm; (b)
breakdown of large molecules by extracellular enzymes and
absorption of small molecules into cells; (c) growth of the
primary population, food storage and cell division; (d) in-
gestion of primary flora by secondary grazers; (e) endogenous
respiration of carbonaceous matter by all members of the bios
with release of ammonia to biofilm; and (f) oxidation of am-
monia to nitrite and nitrate by autotrophic bacteria.
Biczysko (322) indicated activity of the biofilm as a parameter
of primary importance. Wintz (4776) referred to his thesis
in which he showed that the microbial content of the bio-
film averaged 55O million bacteria per gram of film, 8O£ of
these being gram-negative rods and 12$ being gram-positive
spheres, and the predominant bacteria changing with every
test.
42
-------
In 1966, Sanders (3823), reporting on the oxygen utilization
of slime organisms in the stream environment, concluded that
under the experimental conditions and in a steady state
condition the oxygen utilization of the total system will
remain constant with respect to time. Although the thickness
of the slime mass of the attached organisms increases linearly
with respect to time during this steady state phase, the rate
of oxygen utilization remained constant. The averaged thick-
ness of the slime was 11 p. (O.O11 cm) while the minimum thick-
ness for diffusion of oxygen was 21.2 M. (O.O212 cm). Maier
(2837) showed that increasing the thickness of the slime
beyond 48 M- (O.O48 cm) had no significant effect on the
organic removal under the laboratory conditions studied and
that the mass transfer is the rate limiting step except at
high feed concentrations. Sanders (3824) previously re-
ported that the growth rate of slime bacteria under stream
conditions was limited when the attaching surface became
completely covered with one layer of cells. The maximum rate
of nutrient removal from the substrate occurred when the slime
thickness equaled the limiting thickness for the diffusion of
oxygen.
A mathematical model, called the "cell model," to represent
the trickling filter used for biological treatment of organic
waste was proposed by Busch and Hughmark (548) . This model
assumes that the substrate and oxygen are consumed at the
surface of the inclined plate and that the liquid film can
be divided into a number of rectangular cells. The model
considered the consumption of glucose at the biological sur-
face on the plate and diffusion of glucose and oxygen through
the liquid film to the biological surface. Comparison of
the data calculated from the model with experimental data
for an inclined plate indicated that the liquid film flow was
not laminar. Good agreement was obtained between calculated
and experimental glucose concentrations. The following year,
Kehrberger and Busch (2419) developed a mathematical model
based on three theoretical film flow models which deal with
various concentrations of organisms in the liquid film which
are present in a recirculating trickling filter process.
Kornegay and Andrews (2547) developed a mathematical model
for a fixed film reactor which describes the rate of sub-
strate removal from zero-order at high concentrations to
first-order at low concentrations, which are not usually
covered by current design equations used for deep trickling
filters. The biofilm reached a constant thickness of approxi-
mately 2OO M. (O.2O cm) at which point the shear on the film
surface was sufficient to wash the newly formed organisms
into the liquid phase.
43
-------
Critique
The intent of reviewing the various theories proposed here
was to indicate, briefly, the thinking which has gone into
the development of the theory of biological trickling filters.
To start with general observations, then develop empirical
relationships, to be followed later by rational design and
the theoretical models has been the historical development
for the biological trickling filter process. As shown in the
exploded view of the surface of the trickling filter in Figure
3, there are many reactions taking place simultaneously under
varying conditions. In the sewage plant, these reactions
generate an extremely complex problem which still confronts
engineers and scientists of today. A grave drawback is the
lack of applicability of the usual high level kinetic-
ma thematically oriented papers.
It is very unfortunate that more authors have not followed
the examples of investigators such as Busch to attempt to
relate mathematical theory to the solution of real problems.
Quite often workers, especially in the investigation of the
biofilm, in disciplines other than sanitary engineering,
have not made the attempt to relate their work so that the
practicing engineer can appreciate it. The result is that
the contributions of the scientists are regarded as "blue-
sky" and irrelevant.
44
-------
PART II
PLANT DESIGN, MATERIALS OF CONSTRUCTION, OPERATION
MAINTENANCE, AND PERFORMANCE
This section details the literature of the trickling filter
in five broad categories. The first category concerns design,
illustrating the criteria which have been used during the de-
velopment of trends in providing secondary treatment with
biological trickling filters. The second section describes
the materials of construction for the various components of
trickling filters. The third deals with various operational
techniques which have been employed for biological trickling
filters. The fourth consists of maintenance techniques and
problems resulting from the operation of trickling filters.
The fifth category is composed of a performance summary of
the biological trickling filters and the formulae developed
for process evaluation.
45
-------
SECTION VIII
DESIGN CRITERIA OF VARIOUS TRICKLING FILTER MODIFICATIONS
The current design criteria have been the result of an inter-
national effort. Investigators in England, Germany, France,
Russia, South Africa, South America, the United States, and
many other countries, have been working to develop these cri-
teria (189, 371, 1190, 1690, 3955, 4840, 5378). The design
criteria have been organized and categorized according to the
type of waste and plant, hydraulic and organic load aspects,
combined use of the trickling filter with other secondary
processes, and economic considerations. This organization was
made because the primary differences in the modifications of
the trickling filter are concerned with hydraulic or biological
load. Other factors such as the media used in the filters, the
ventilation requirements, pretreatment, post-treatment, and
solids disposal are, to a degree, similar in each of the modi-
fications and are treated accordingly.
A basic appreciation of the design problems may be drawn from
many textbooks (112, 3O8, 1O8O, 2984). Escritt (1195) observed
that the Fifth Report of the Royal Commission on Sewage Disposal
issued in 1908 has had a great influence in developing the de-
signs involving percolating or trickling filters. Especially
in England (and later in this country), the recommendations of
the Royal Commission were interpreted with almost the same ac-
ceptance as a statute of the land. A classic example was cited
by Escritt, "But at the present time by far the greater number
of percolating filters are being constructed to an exact depth
of six feet," with very little justification other than it had
been used in the past, worked fairly well, and was recommended
in the Royal Commission report. In the design of sewage
treatment plants, arbitrary depths, loadings, retention times,
etc., are sometimes used, but the usage of arbitrary parameters
is less prevalent.
Buswell et al. (555) in 1928 and Bloodgood (363) in 1956 re-
ported that considerable progress had been made in the develop-
ment of biological trickling filters, but that it was apparent
that not all of the information published on the theory and
design was in agreement. Spurts of activity in developing de-
sign criteria have been reported (1326, 464O) but, unfortunately,
as others (658) have noted, new developments in design are
slowly accepted. However, with more operational experience and
the data released by the National Research Council (5590), a
basis for rational design was developed after World War II
(3O76) and specific problems were identified and attacked on
the new processes, such as the high-capacity trickling filters.
Greeley (1564) outlined considerations in the design of high
rate trickling filters as influent, effluent, waste character-
istic, effect of recirculation, localized factors such as
temperature and degree of treatment required, and overall cost
and economy of the process. Mathematical interpretations of
47
-------
the above factors were given more consideration after the
publication of the critical study of the design and operation
of several hundred military treatment plants by the National
Research Council (5590). This study provided design informa-
tion useful for treating sewage uncontaminated by industrial
waste waters. Baker and Graves (150) in 1968 reviewed several
design formulations and concluded that more accurate data are
required to establish the validity of the different formulas
and to determine the effect of filter efficiency parameters on
volume requirements.
There are many divergent viewpoints, e.g., those of National
Research Council (559O), Eckenfelder (1O81), Galler and Gotaas
(1395), Schulze (3916), Velz (4535) and Rowland (2O64) , since
there are many varied applications of the trickling filter and
its modifications which have been developed over many years in
different geographical areas.
TYPE OF WASTE
In the review of design, the influent waste characteristics
are of primary importance. Generally speaking, there are es-
sentially two types of waste streams, municipal waste and in-
dustrial waste. Practically speaking, it is only rarely that
one of these two types of waste is treated solely by itself.
Mixed waste normally exists in practice, but frequently the
mixed waste stream can be treated as either municipal or in-
dustrial waste, depending on the relative proportions in the
waste. For example, a waste stream containing a high phenolic
content with some domestic sewage from a chemical processing
plant would be termed an industrial waste, whereas a waste
stream entering a major metropolitan waste treatment plant with
minor industrial contribution would be regarded as municipal
waste, with a contributory industrial waste influent.
Combined sewers (sanitary and storm) further complicate the
problem of wastewater treatment and consequently the functional
design of biological trickling filters. Approximately 28$ of
the population of the United States which are served by sewers
have combined sewer systems (3954). These systems deliver an
influent to the waste treatment plant which has extreme vari-
ations in flow and organic loadings. The storm water causing
these variations has been a problem overseas (2879, 5549) and
in this country, as summarized by the Federal Water Pollution
Control Administration (3631) . The very frequent use of the
trickling filter to handle municipal waste may well be based
on the fluctuating flow experienced at most small treatment
plants (2687,2970, 3053).
In many municipal plants, the situation where supernatant
liquid from the anaerobic digester is pumped back through
the filters is further complicated (3O76). The design
must take this additional load in consideration. Typical of
48
-------
the problems generated by this recycling supernatant technique
is the development of odor problems (4187) which must be mini-
mized by appropriate design. There are literally hundreds of
articles, as noted in this review and will be discussed later,
in which the biological trickling filter is used in municipal
waste treatment plants. Industrial wastes have been treated
extensively (3O7, 546, 2977, 3312) by biological filtration.
The importance of biological trickling filters to handle
industrial waste was reflected in the literature to such a
degree that Part IV of this review is specifically concerned
with the application of the biological trickling filter to
industrial wastes.
A prime advantage of the trickling filter in industrial waste
treatment applications is the ability of the filter biota to
survive shock loads of toxic waste, elevated temperature,
high oxygen demand, and other rapid changes which may occur
in an industrial waste disposal system. It has been noted by
Busch (546), Eckenfelder (1O78) and McKinney (2925) that it is
not because the microorganisms are any more hardy in a bio-
logical trickling filter than in any other biological system,
but that survival is actually due to the relatively short re-
tention time in the filter (546) and that contacted and killed
organisms are flushed from the filter, exposing another layer
of microorganisms not subjected to the shock load (1O78) .
Critique
The trickling filter emerges as a versatile treatment device
with many advantages. The filter is so constructed that it is
highly resistant to hydraulic overloading and to shock loads
of organic or toxic material. The demands on the operator for
skilled management are nominal or absent so that the trickling
filter works effectively in the hands of unskilled people. The
filter is accordingly favored for small installations.
Since filters may be made with considerable depth and because
they cope effectively with a wide range of difficult waste
problems, the role of the filter is by no means limited to
small installations. It continues to be an effective device,
requiring consideration whenever biological treatment of wastes
is contemplated.
HYDRAULIC AND ORGANIC LOADING
In this section, hydraulic and organic loading for the trick-
ling filter and its modifications are discussed. A very brief
description of many of the modifications is provided along with
the design considerations.
Contact Beds
The contact tank was a step between land irrigation as a waste
disposal procedure and higher rate systems such as the trickling
49
-------
filter and the activated-sludge process (2159). Several texts
have been published (348, 1O40, 1727) which generally related
the advantages, disadvantages and general design criteria of
contact beds. Prince Albert of England (4463) in 185O pro-
posed the use of a system employing charcoal, gypsum and
burnt clay laid up on a false bottom and subjected to an up-
flow of sewage to provide purified water and a source of
fertilizer.
Clark (698) reported on the extensive studies which began in
1894 at the Lawrence Experiment Station. The various factors
affecting contact beds were discussed, such as efficiency of
different materials, double-contact filtration, i.e., recycle
or two-stage, disposal of nitrogen, permanency of contact
beds, composition of organic matter, best methods of opera-
tion, rate of flow and different depths. Adequate informa-
tion was provided in the 193O's by Metcalf and Eddy (2984)
in the third edition of their book on sewage. A description
of the contact bed based on their information was as follows:
A rock medium, 1/3" to 3" in diameter, is piled to a spec-
ified depth (4 to 6 ft in this country and 2.5 ft in England)
and primary influent is applied. The dosing sequence con-
sisted of filling for two hours, contact for two hours, then
draining with 6 hours' rest. For a 4-ft bed covering O.5
acre area, O.5 mgad (11.5 gal./ft2/day), allowing for rest-
ing and maintenance, could be treated. The operation was
flexible, depending on the number of fillings per day. How-
ever, contact time greater than two hours gave inferior re-
sults. Distribution of the sewage was by filling lateral
underdrains and pipes, and troughs similar to underdrains
common in trickling filters. Maturing or conditioning time
of the bed (for anaerobic film build-up) required from a few
weeks to several months. When the medium became clogged,
the solution was removed and the media washed. The contact
bed had fewer odors and flies than the trickling filter.
Preaeration usually guaranteed nitrate formation in the ef-
fluent. Single, double, and triple stages were used (2984).
The literature (1064) reflected successful use of contact beds
for municipal and industrial waste treatment. However, in
the first decade of 19OO, sprinkling filters were preferred
to contact beds because of the greater capacity of the fil-
ters (322O, 5O59). Contact beds were relegated to the solu-
tion of waste treatment problems for remote locations (486,
3982), such as institutions, recreational sites, military
installations (262O), and other low population areas. With
increasing popularity of trickling filters, methods to con-
vert contact beds into storm water tanks, sludge drying beds
or trickling filters were developed (159, 3441, 3951).
The obsolescence of contact beds is attributed to the higher
efficiency of trickling filters. The contact bed is of his-
torical importance only.
50
-------
Dipping Contact
A slightly different variation on the contact bed principle was
given by Allen (38) and Doman (99O) in the late 192O's and early
193O's. Allen (38) reviewed Buswell's work (556) on a series
of paddle wheels supported over a channel of waste. The wheels
were rotated by the flow of sewage, the blades thereby being
alternately submerged and exposed to the air or in "dipping
contact" with the waste. Doman (99O) described difficulties
of anaerobic film build-up on an experimental contact filter
with partially submerged rotating plates. Schmitt (3879) dis-
cussed Kessener's system (2452) in which revolving brushes
placed along one side of the tank dip into the liquid to a
distance which can be varied according to the aeration required,
and concluded that it was a feasible waste treatment device. It
is probable that this system was a forerunner to the brush aera-
tors .
The approach of moving the medium alternately through the waste
stream and the atmosphere was recommended much later by Reid
et al. in 1960 for the treatment of high phenolic-type waste
(3547) . Concentrations of up to 75OO mg/1 phenol have been
treated by this method. Hartmann published a series of papers
(1769, 177O, 1771, 1772, 1773) on the advantages and operation
of the dipping contact filter and its application to several
types of industrial wastes. The advantages of this process are
adaptability to varying loads, concentrations and degrees of
treatment, low cost of construction, ease of operation, and a
greater resistance than the activated-sludge process to dilute
poisons, mineral oils and detergents (1769). He has applied
the method to the treatment of sewage, and mixtures of sewage
and trade wastes, including those from dairies, breweries,
maltings and canneries (1773) . This technique (1771) used
rotating disks which were made from expanded polystyrene
strengthened with unexpanded plastic and separated by metal
spacer disks. The plastic disks acted as the medium to sup-
port the biological growth. The British Patent 930,226 issued
to Hartmann(1772) was based on this design, and the successful
operation of the experimental plants at Stuttgart and at Heil-
bron in Germany has also been described (177O). More recently,
Joost (235O) reported in general terms on a similar system
called the "Rotating Biological Surface" waste treatment proc-
ess. Although he cited the advantages of the system, he did
not give sufficient data for design purposes.
Power .costs for this type of plant have been shown by Allen
(38) and Joost (235O) to be low compared to other biological
systems. Using a two-stage system, Hartmann(1773) stated that
mixtures of sewage and trade waste waters have been treated
with 6O$ removal of BOD by the dipping contact filter. Joost
(235O) concluded that 9O$ of the BOD could be removed from
domestic waste with a retention time of less than 45 minutes.
In an effort to reduce costs and odors, Schreiber (3891) de-
veloped a type of dipping contact bed which allowed the medium
to move for periodic cleaning. In the "Proceedings of the Inter-
51
-------
national Congress PRO AQUA 1961," Stengelin reviewed the design
and operation of dipping percolating filters and their advan-
tages (42O8) .
Critique
A critique of the contact bed information may appear superflu-
ous since this system has essentially been phased out of waste
treatment plant operations due to its low capacity treatment.
However, due to the recent renewed interest, the following
comments regarding dipping contact filters are made.
Dipping contact filters have not enjoyed wide popularity, even
with the work of Buswell in the 193O's, Schreiber in 1940,
Hartmann in the 196O's, Stengelin's review in 1961, and Joost
in 1969. Since the technique was known and studied, it would
appear that conditions were not favorable for its use. Other
systems, such as the trickling filter or the activated-sludge
process, must have definite advantage, since their use is in-
creasing and is preferred. The application of the dipping
contact system has been limited to rather unique problems. The
operating difficulties and the nuisance factors suggest that
the economics of the application be carefully evaluated. In
Mr. Joost's paper, which was written for the purpose of in-
forming designers of the technique, necessary operational data
are missing. For proprietary reasons, the economics of the
technique were not revealed and at his own suggestion the capi-
tal cost of the system, also not given, should include an en-
closure device to protect the biomass on the rotating disk from
interference by windstorms, hail or heavy rain. A highly favor-
able design characteristic is the low personnel requirement of
this system.
Conventional Trickling Filter Systems
The term "conventional" means a treatment plant employing trick-
ling filters in a commonly used fashion, i.e., after primary
treatment and followed by secondary clarification. Thousands
of these plants utilizing trickling filters have been built
(4916) and their design, construction, operation, and problems
described. Typical of the literature are papers such as the
sewage plant for Leominster, Massachusetts, (population 4,OOO)
built prior to 1919 (5314) and recent reports for the plant at
Magna, Utah, (2722), and Greensboro, North Carolina, (5644),
both built in the early sixties, which incorporated trickling
filters in the sewage treatment plant.
Pearson (3328) in 1965 listed design factors which, generally,
are in agreement with those described by Chase (658) in 1945.
These factors are summarized as follows:
52
-------
1. Pretreatment - The solids-liquids separation to reduce
^the clogging of filters in the form of primary clari-
fication by some type of sedimentation, screens, pre-
aeration, prechlorination or other methods is viewed
as extremely important. It is noted that fine mechanical
screens were installed at Akron and Baltimore between
the primary settling tanks and the filters to inter-
cept clogging materials.
2. Loadings - The loadings on the trickling filter are
expressed in the terms of a volume of settled sewage
applied per day and pounds of oxidizable material per
unit area or volume applied per day. Chase expressed
the opinion that the use of BOD as the sole criterion
for filter loading did not seem correct, as it did
not consider nitrification.
3. Depth of filters - The depth of filters was viewed
as significant from the standpoint of nitrification,
and limited land area favored deeper filters. It was
further noted in the case of two-stage high rate
filters that filter depths of only three feet are used.
4. Filter media - One of the most important considerations
in design is the selection and placing of filter media.
Chase's comments, published in 1912 (655), are essen-
tially applicable today, such as the stones used were
nearly cubical with rough (not porous) surfaces and
not susceptible to deterioration. The sizes of the
stone were from 1.5 to 2.5 inches diameter for
standard rate filters and 3 to 4 inches diameter for
high-rate filters. The uniform distribution of the
sewage over the surface of the filter by fixed spray
nozzles contributed to the practical application of
trickling filters to the purification of sewage.
5- Dosing methods - The limitations for using six spray
nozzles are noted. The technique of intermittent
dosage as used in 1945 is significant in aiding the
ventilation requirements of the filter. An advantage
of continuous application is the reduction of the
fly problem.
6. Ventilation - The evidence from forced ventilation
and natural ventilation, as well as freely ventilated
walls and enclosed walls, indicated to Chase (658)
that neither forced ventilation nor porous filter
walls are essential, provided there is adequate
underdrain ventilation.
53
-------
7. Underdrains - Of several types of underdrainage
systems which have been devised, any system is
adequate which provides prompt removal of the
filtered sewage and allows ventilation between
the top surface of the filter and its undersur-
face. The system must have sufficient capacity
to permit easy inflow and outflow of air and
to provide sufficient velocity of the effluent
to prevent accumulations of sludge and grit.
8. Post-treatment - Due to the sloughing phenomenon,
a final clarifier is required to separate the
settleable solids from the purified effluent.
The secondary sedimentation could be intermittent
sand filter, a filter with a settling tank, or
other means. Chlorination is used for bacteria
removal.
9. Recirculation - The evidence available to Chase
on the most advantageous combination for recir-
culation is inconclusive, but the trend has
been established to use recirculation on standard
rate as well as high-rate filters. There is much
evidence that recirculation is of considerable
value.
Papers were published (589, 893, 187O) which are typical of
many indicating the modes and trends of biological trickling
filter usage. Design information is available in Imhoff's
handbook (2176), yet investigators have always demonstrated
a curiosity and a desire to improve their-design procedures
(4793) .
High-Rate Filters
There is no rigid dividing line between the various rate
trickling filters, as may be noted from the literature (274,
658,1563, 3O71, 4916). However, some rather general ranges
are given in Table I to differentiate the various types of
trickling filters by the hydraulic and the organic loadings.
Greeley (1563) reported that high-rate filters are operated
with organic* loads in the range of 5O times the normal load
for the low-rate (standard rate) trickling filters. The de-
velopment of high-rate trickling filter systems generated a
few patented processes (8O7).
54
-------
T&ble 1
Design Comparison of Different Rate Filters
a
Rate
Hydraulic Loading
gallon/min/ft2
gallon/fta/day
million gallon/acre/
day (mgad)
Organic Loading
Ib BOD/1000 ft3/day
Ib BOD/acre-ft/day
Ib BOD/cu yd/day
Low
High
Super'
0.0175-0.07 0.139-0.7
25-100
5-25
220-1,100
0.137-0.68
200-1000
8.7-44
25-300
1-6
1440-8640
62.5-455
126-1512
1,100-13,000 5,560-64,320
0.68-8.1 3.32-39.9
*See references 538, 1697, 49l6 and 5574.
Also referred to as roughing filters.
High-rate filters are of three principal types, depending on
rate of feeding, recirculation or loading. The three types
are the Biofilter, the Accelo-Filter and the Aero-Filter.
Jenks (23O5) described the Biofilter plant as comprised of
essentially one or more detention tanks, a similar number
of filters, relatively shallow and arranged for stage working,
and a recirculation system. A Biofilter (5574) usually has
the media about 4-5 ft deep, and utilizes recirculation.
Organic loadings are of the order of 2,5OO-3,OOO pounds of
BOD per day per acre-foot (57.5 to 69 Ib BOD/1OOO ft3/day)
based on the strength of the primary tank effluent, with
hydraulic loadings ranging from 1O to 3O mgad (23O to 69O
gal/ft2/day).
The Accelo-Filter (5574), which is relatively deep, has direct
recirculation of unsettled filter effluent back to the inlet
of the distributor. Organic loadings, based on the organic
content of the primary tank effluent, are in the same range
as the Biofilter, of the order of 2,54O-3,OOO pounds of BOD
per acre-foot per day (55 to 69 Ib. BOD/1,OOO ft3/day, or
1.5 to 1.9 pounds per cubic yard per day). It has been de-
scribed (1563) as accelerated biological treatment by direct
recirculation of large quantities of material to a rotary
distributor on the biological filter. Gillard (1474) reported
that this system may be operated as either a high-rate or a
conventional percolating filter.
55
-------
Much of the basic work published by Halvorson et al. (1665)
led to the development of the Aero-Filter, which has a rela-
tively deep media bed of six to nine feet. This system (5574)
utilizes a low momentary rate of sewage application by means
of a special distribution apparatus designed for "rain drop"
application over a maximum area of the filter at one time.
Recommended organic loadings are from 3,OOO to 3,2OO pounds
of BOD per acre-foot, or 1.9 to 2.O pounds per cubic yard per
day (69 to 73.6 Ib BOD/1,OOO ft3/day), and the hydraulic load-
ing rate (1563) is more than 10 mgad (23O gal./ft2/day) with
recirculation if necessary to maintain this rate. Flowthrough
diagrams are shown in Figure 4, page 90, illustrating the
versatility of the recirculation systems.
The development efforts by Hamlin (1697) culminated in the
super-rate filter (see Table I and section on "Roughing Fil-
ters") . This high-rate filter has a very shallow bed (one
foot) and is dosed at hydraulic loading rate as high as 455
mgad (1O,465 gal./ft2/day). Hamlin says, "With a BOD removal
of 9,OOO Ib in the preliminary units and the super-rate filter
the power requirements were some 8O hp and a BOD removal of
6.25 Ib per kilowatt hour with a 6O percent removal of overall
BOD." With the advent of the plastic era, lighter prefabricated
materials were substituted for the heavier inorganic materials
used as media in pilot and sewage treatment plants' (488, 538,
678, 1083, 1106, 112O, 2955, 3329). With this change, high-rate
filters were successfully operated at hydraulic loadings from
62.5 to 375 mgad, (1437 to 8625 gal./ftz/day) and organic load-
ings from 5,OOO to 64,OOO Ib of BOD per acre-foot per day (115
to 1472 Ib BOD/1,OOO ft3/day). The advantages of plastic media
are removal of high BOD at high hydraulic loads; light weight,
requiring less support structure and thus less cost; and free-
dom from corrosion.
In establishing the design criteria for the hydraulic loading
rate, Eliassen (1150, 1154) developed charts to facilitate
the determination of the filter capacity based on gallons of
sewage per acre-foot per day or pounds of BOD applied per
cubic yard per day. As one of many investigators, Imhoff
(2211) was concerned with the difference in loading between
low-rate and high-rate percolating filters, and suggested that
to ensure the necessary cleansing action the surface load
should not be less than 27.4 ft3/hr. The depths of the German
filters were reported (2211) to be not greater than 1O to 13
feet. Schulz (391O) operated an experimental tower percolating
filter, 26 feet deep, which had forced ventilation, at a
hydraulic load of 15 cubic meters per cubic meter of media per
day (4.9 mgad or 113 gal./ft2/day).
Bachmann (133), in his review of the design and operation of
the Biofiltration system, noted that if the hydraulic rate ex-
ceeded 1OO to 125 mgad (2,3OO to 2,58O gal./ftz/day) flooding
would occur. According to Bogren (383), cotton finishing waste
56
-------
liquors, which have been treated previously by a standard
trickling filter, have been studied in a pilot plant with a
high-rate percolating filter. Greater reductions in BOD were
realized when operating at 10 mgad (230 gal./ft2/day) than with
a low-rate filter.
Organic loads in excess of 1,OOO Ib BOD/acre-foot/day (23 Ib
BOD/1,OOO ft3/day) were used at ninety military 'installations
and as many municipalities in 1948 (1567) . An upper limit on
the applied organic loading was calculated by Greeley et al.
(1566) to be 10,OOO Ib/BOD/acre-foot/day (23O Ib BOD/1,OOO ft3/
day) . The load-efficiency relationship is affected by local
factors such as treatability of the sewage, temperature, etc.
German experience (376, 2185, 3785), over a period of more
than 24 years, showed successful use of the high-rate filter,
even though the waste was more concentrated than U.S. sewage.
Rankin (3502), in his analysis of operating data, found the
effect of organic load on the efficiency of Biofilter
plants is primarily dependent on the ratio of recirculation.
Dosing rate, loading of filter, or depth of filter have no
significant effect in the range covered by the data. Homack,
in the discussion of the paper by Rankin (35O2), and Pearson
(3328) were among many of the investigators who stated that
the waste character was a critical factor in the design of
high-rate trickling filters. High-ifate trickling filters
have been used for many industrial waste waters with high
organic concentrations, for example, by Reimers et al. (3551)
for fine chemical wastes, by Mercer et al. (2972) for liquid
canning wastes, by Burns and Eckenfelder (538) for pulp wastes,
and by Pearson (3328) for a variety of wastes.
Critique
After the first use of the trickling filter at the Lawrence
Experiment Station of the Massachusetts Board of Health in
1889, studies at the pilot plant stage and of operational
data showed the way for necessary improvements in design,
materials of construction, and operation. The advantages of
a high-rate trickling filter, along with the governmental
sanitary requirements, provided the necessary stimuli for the
adoption of the system for urban and industrial sewage.
Foreign work was often reported dealing with high-rate or
high-capacity systems, and, after conversion to units familiar
to the U. S. figures, the loadings were not always in the high-
rate range. This simple fact may well have been, and still is,
a problem in the communications of performance factors. Thus,
it has been stated that the super-rate filters being used to-
day in many applications have hydraulic loadings of 1 to 6,
whereas low-rate filters had loadings of 1 to 4. The dif-
ference is, of course, the units. The super-rate filter is
expressed in gallons per minute per square foot cross-sectional
area and the low rate filter is expressed in millions of gallons
per acre of area per day. To be consistent, the filters should
57
-------
be expressed in the same units, e.g. 62 to 455 mgad for the
high-rate and 1 to 4 mgad for the low-rate loadings.
Low-Rate Filters
The low-rate or standard trickling filter was the backbone of
the secondary biological treatment process for nearly 5O years
and many are still in operation today (25O4, 4358). As late
as 195O, the standard rate trickling filter represented more
than 5O% of all secondary treatment works (4358). The hy-
draulic and organic loadings of the low-rate filter are indi-
cated in the ranges as shown on Table I. The low-rate trick-
ling filter was described as being from 4.5 to 6 feet deep
(4596, 5574). The rock media varied from 1.5 to 2.5 inches
nominal diameter and was dosed with fixed nozzle distributors
and, later, rotary distributors (658, 3327, 5574). Low-rate
filters commonly had a dosing period of 3 minutes and a rest
period of 6 minutes (658), which was a topic of considerable
controversy (4187). The underdrains were sized to flow half-
full at the designed flow for adequate ventilation (35O1).
A primary difference in the low-rate and high-rate filter
system was the degree of nitrification achieved (1O29). The
low-rate filter, having a greater detention time and lower
hydraulic and organic loading, produced a more highly nitri-
fied effluent than the high-rate trickling filter (1O29, 115O) .
Most of the recent published work has been on the high-rate
rather than low-rate trickling filter system (1542, 1559).
Grantham (1542) reported on the operation of standard rate
design filters that were operating at an overloaded condition.
These filters were described as intermediately loaded systems.
The hydraulic loading of between 5 and 20 mgad (115 to 460
gal./ft2/day) on 6-ft deep filters places this operation in
a gray area between low-rate and high-rate filters, as indi-
cated in Table 1. It was concluded that organic loadings of
over 2,5OO Ib BOD/acre-foot/day (575 Ib BOD/1,OOO ft3/day)
and hydraulic loadings greater than 1O to 12.5 mgad (23O to
288 gal./f12/clay) should be avoided. However, Grantham (1542) ,
based on studies by others, stated that standard rate fil-
ters operated in excess of their design loading without re-
circulation may be expected to reduce BOD with little or no
decrease in efficiency and without operational difficulties.
The nitrification of the filter effluent was reported to
decrease as the organic and hydraulic loading increased.
Lanz (2629) tested the hydraulic loading incrementally up to
the rate of 3O.5 m3/m2/day (748 gal./ftV^ay) in a 3-meter
(10-feet) deep filter (approximately 10 cubic meters sewage
per cubic meter media, 3.3 mgad). He found that the effi-
ciency of the filter decreased with increasing load, but even
at the highest rate tested the effluent was not putrescible.
Aeration at a rate of 3O.8 yd3/hr (23.5 m3/hr) was suffi-
cient to maintain the biological purifying power at maximum
load studied.
58
-------
Typical of the transformation which took place in trickling
filter design was the conversion of the Black River Sewage
Treatment Works, Baltimore, from low-rate filters to high-
rate filters. Keefer and Meisel (24O4), in rechecking results
obtained in 1938, verified the earlier work. By increasing
the hydraulic load from 2.68 to 16.39 mgad (62 to 377 gal/ft2/
day), the resulting BOD removals were in the 90$ range at the
low rate (2.68), in the 8O$ range up to 13.53 mgad (312 gal/
ft2/day), and about 70$ at the 16.39 mgad rate.
Critique
In the gradual development of the low-rate trickling filter
into a high-rate trickling filter, there was a good under-
standing of the many factors which influenced the operation
and efficiency. However, one point which cannot be over-
looked is the degree of nitrification. It was thought for
many years that the early treatment plants were operating most
efficiently when the degree of nitrification was the highest.
In England, nitrate measurements were made originally to
assess the degree of stability of the effluent. It has been
pointed out by many investigators that if the dilution factor
of the receiving body of water is high enough, a highly nitri-
fied effluent was not required. Cases have been reported
where a highly nitrified effluent was instrumental in the
development of a littoral aquatic growth. It would seem now
in light of present understanding that the low-rate trickling
filter would still have an application at a remote installa-
tion where the receiving stream was either dry or did not have
sufficient dilution to establish a proper nitrogen balance.
As will be discussed later, low-rate trickling filters with
slight modifications and in series operation with other units
have found limited application. Walton's paper (4596) is very
indicative of the slowness of adopting new technology of a
proven process to develop more efficient waste treatment sys-
tems. It must be remembered that this conservative approach,
or absence of innovation, which has been demonstrated time
and again throughout the history of trickling filter waste
treatment, was predicated upon the heavy responsibility
placed upon those controlling the change. This responsibility
to the public, and the operation of the sewage plant with
limited funds and manpower, nurtured the replication of known
and proven technology. For changes to take place on a wide-
spread basis, exhaustive experimental data were required as
a safeguard to the populace. Today, this responsibility
still exists, but changes in technology are more easily ac-
ceptable.
Single-stage Filters
The single-stage trickling filter was the unit most commonly
employed (with low-rate designs) during the early development
of the biological filtration technique (4358). It has been
described (5574) as being 6 to 1O feet deep, filled with
59
-------
various media, usually a crushed rock or slag, and followed
by a clarifier (4358). Single-stage filtration appeared
inferior (2118, 3551, 48O4) when its performance was compared
with two-stage and alternating double filtration techniques.
In many of the comparisons (35, 326O), two-stage and double
filtration operated with less difficulty and gave a better
quality effluent. Hansen (1719) concluded that single-stage
trickling filter treatment has been found unsatisfactory for
strong sewages.
Jenks (23O5) discussed the advantages and costs for a single-
stage, as well as for a two-stage, trickling filter plant by
operating with much higher rates of recirculation, the tech-
nique of which was developed into the commercial Biofilter
process. The hydraulic and organic loadings on the single-
stage filter were determined usually by local factors and
the effluent requirements (3161, 4538). Meltzer (2965)
stated that there is no evidence that two-stage processes are
any more efficient than single-stage filtration in deep bed
filters. Escritt (12O4) commented that filters operating by
single filtration should be as deep as practicable. The
American Society of Civil Engineers (4912), in a review of
sewage developments during 1944 and 1945, and based on a
review by Chase (658) , stated that the medium for a single-
stage filtration is usually 6 to 1O feet deep, whereas the
high-rate filter may be 5 feet for a single stage and 3 feet
for two stages.
Design charts were prepared by Baker and Graves (15O) for
single-stage filters using several design formulas. Data
were reviewed and interpreted by several investigators (85,
352, 1233), in which the performance of single-stage operation
was studied in detail, and it was shown that organic and
hydraulic loadings, recirculation, and temperature have non-
linear effects on performance. In addition, the cross-
product terms in their analysis indicated that these variables
are not independent in their effects (352). When perform-
ance was evaluated as a percentage of BOD removal, the ef-
fect of the strength of sewage applied to the filter was
negligible (1233). Archer and Robinson (85) related the ef-
ficiency of a single-stage filter with media volume and re-
circulation ratio based on performance formulas of the Na-
tional Research Council and compared them with field per-
formances. Predictions of efficiencies using charts and
formulas were found to agree closely with field performances.
Recirculation improved the overall efficiency of the plant.
Montgomery (3O74) discussed the results obtained at various
plants where the single-stage Aerofilter has been used, and
Fischer (1285) discussed the Biofilter principle from the
standpoint of the effectiveness of a single-stage applica-
tion. Both reported that between 7O and 8O$ BOD removal
was expected when the recirculation ratio is 1.5:1, although
as high as 87$ BOD removal has been obtained.
Typical of the many applications of single-stage filtration
was that reported by Imhoff and Dahlem (22O4) for a small town
60
-------
waste treatment system, but they cautioned that the low rate
of flow created problems due to odors and flies. The nuisance
problems are reduced by filtration at a high rate with re-
circulation of the effluent. At the recently constructed
sewage treatment plant at Aylesbury, England, (4940), the
filters were flexibly designed to operate by single-stage
filtration among other schemes. Water quality was good enough
that the final effluent could be discharged directly into the
River Thames or used for irrigation. Other single-stage ap-
plications were, typically, treatment of hospital wastes
(3912), rubber processing waste waters (3O55), and the combined
waste waters from a paint works, creamery, foundry, laundry,
malt processing, and domestic sewage (4582).
Critique
The literature has indicated an awareness by the investigators
that, although the single-stage filter is adequate and satis-
factory, increased detention time or two-stage filtration,
which achieves essentially the same thing, was required. Limita-
tions of the increased depth were that the media may not be
able to support a very deep bed and required a significant
foundation structure. Modifications of existing installations
of single-stage units were often forced by economics to expand
the plant by adding a second-stage filtration system to com-
pensate for the added depth required for sufficient BOD re-
moval. Media other than rock could have been used, but
acceptance of these other media was slow. With the development
of media which were lighter weight and possessed proper venti-
lation characteristics, single-stage installations using deep
beds were again considered. Montgomery (3O74) noted that,
although it was possible to improve single-stage filtration
by parallel recirculation, the process was not economical in
this situation as it was cheaper to use two-stage filtration
when a high degree of purification was required. In all
probability, this statement is still true today, with the
exception that the two stages need not both be trickling
filters. The "acid test" would appear to be the projected
costs of removal of a pound of BOD per day over a twenty-year
period for both systems.
Two-Stage Filters
Two-stage or double f i 11 ra t ion me ans two biolo g i cal
trickling filters in series with or without intermediate clari-
fiers, followed by final clarification. This technique was a
natural development for overload plants, as suggested by Nelson
(3159). The increased popularity of high-rate trickling fil-
ters since 1944, with two-stage operation to maintain effluent
quality, developed different design criteria (1915). Different
international trends were established. The U. S. favored two-
stage treatment in shallow high-rate percolating filters with
elaborate provisions for recirculation. Great Britain favored
61
-------
two-stage treatment and relatively deep percolating filters at
more moderate rates with the added condition of being able to
interchange the sequence of filtering units (1719).
i
Sorrels and Zeller (4111) investigated the design factors in-
volved in the operation of two-stage trickling filters with
and without intermediate sedimentation. Archer and Robinson
(85) and Germain (1460) concurred with Sorrels and Zeller
that removal of the intermediate clarifier did not harm the
treatment plant efficiency. Meltzer (2965) suggested the so-
called two-stage operation is, in actuality, a single-stage
deep bed from the efficiency viewpoint. Brown (492) stated
that the cause of an odor problem was due to an inadequate
temperature differential between the waste waters and the
atmosphere, which inhibited natural ventilation. Introduction
of forced ventilation solved the odor problem and improved the
operating efficiency of the plant. Whitehead (4711) suggested
for design purposes that, when the influent BOD does not rapidly
increase, trickling filters alone would probably give adequate
treatment.
Montgomery (3O74), in discussing the results obtained on an
Aero-Filter at Austin, Minnesota, calculated the removal of
BOD using the same volume of medium and concluded that shal-
low filters give inferior results when recirculation is not
used, but the increased amount of recirculation required to
maintain the minimum flow per unit area on the shallow bed
compensates for its lower efficiency. Recirculation ratios
from 1.5:1 up to 6:1, in the two-stage process, have been used
by Fischer (1285) and by Jenks (23O5), depending on the local
conditions and operating techniques required to render the ef-
fluent acceptable. Archer and Robinson (85) reported that the
slight differences in medium volume between first-stage and
second-stage had a very limited effect on the overall effi-
ciency, to have the volumes equally divided between the two
filters and the recirculation equal for both stages. Smith
and Wittenmyer(4O86) reported on the use of the plastic
media, Surfpac®5? as the first stage in a two-stage filter
system to take advantage of existing percolating filters to
increase treatment capacity.
Remote locations, such as turnpike rest areas, have been serv-
iced by two-stage biological filtration (977). Bowes (426),
forty years ago, noted that small installations were equipped
with two-stage filters and identified the auxiliary equipment
and manpower required. Package treatment systems using the
two-stage approach were available in eight sizes and operable
with little or no attention (5118).
Municipal applications of two-stage trickling filters number
in the low thousands (4916). Typical of these are installa-
tions at Oklahoma City, Oklahoma (93O), Suffern, New York (54U),
A registered trademark of The Dow Chemical Company.
62
-------
Newbury (5226) and Aylesbury (494O) in England, and Walvis
Bay (37OO) in Southwest Africa. Many other two-stage bio-
logical systems have been installed and operated and will be
discussed later under sections emphasizing other factors.
Industrial waste applications of two-stage trickling filters
are numerous. For instances, the treatment of chemical waste
(3551), combined sewage and milk waste (48O4), meat packing
house waste, in combination with other processes (1959), as
well as many more, are described and referenced under their
specific wastes in Part IV.
Of the many mathematically oriented papers (most of which are
discussed elsewhere) dealing with two-stage trickling filters,
design is discussed by many, including Archer and Robinson
(85) , Wishart (4793) , and Baker and Graves (15O) . Wishart ex-
pressed the opinion that there were discrepancies between
actual BOD effluent data and BOD values calculated from the
National Research Council (NRC) formulas. Archer and Robinson,
through additional studies, found that the National Research
Council formula predicted quite closely field performances
which were obtained in a newly constructed two-stage biological
filtration system. Later, Baker and Graves suggested modifica-
tions of other formulas for second-stage filters to closely
follow the National Research Council formula and presented de-
sign charts for each, with a computer program to optimize the
two-stage filter volume.
Critique
The literature on two-stage trickling filter applications
showed that this development grew out of necessity from over-
loaded conditions at treatment plants. As noted previously
in the discussion on single-stage filtration, an awareness of
the significance of depth and contact time was demonstrated by
the investigators who were active in the pre-World War II
period. Once the stresses of war were removed and the data re-
sulting from the operation of the military two-stage trickling
filter plants could be observed, more reliable predictions were
incorporated into the design of new systems. Many times the
development of a two-stage plant was governed by factors other
than the optimum hydraulic or biologic information. For in-
stance, if the existing single-stage trickling filter unit was
equipped with media of sufficient volume and the waste was of
low BOD strength, this trickling filter could be used as the
first stage in a two-stage system. The second stage could be
another trickling filter or, if the regulatory agency required
a very high effluent quality, the activated-sludge process.
Another example would be an existing single-stage biological
filter with a relatively small medium volume and an influent
waste of significant BOD which was operating under an
63
-------
overloaded condition. The solution to this problem could
have been the installation of a two-stage trickling filter
plant. With the development of media other than rock, in the
last 1O to 20 years, the use of a roughing filter ahead of the
existing trickling filter has been a popular practical solution.
If the waste was fairly uniform and a high enough BOD, an
activated-sludge unit may have been installed ahead of the
trickling filter, converting the existing trickling filter
into a polishing filter. The design engineer had at his dis-
posal several options which were dictated by local conditions
and regulations at the time of design. It is anticipated that
overloaded conditions will also be rectified in the future by
two-stage trickling filter operation.
Three-Stage Filters
In present day terminology, three-stage or tertiary treatment
usually connotes activities concerned with pollution control
of nutrients, primarily phosphorus and nitrogen. However,
three-stage biological treatment was practiced more than 38
years ago by Rhode (3582) who fitted brown coal slag filters
with perforated clay pipes for better aeration to treat waste
waters from a dye works. Dibdin (946) concluded in 191O that
tertiary waste treatment may be necessary in exceptional cases
and recommended sand filters, land irrigation, fine contact
beds, and percolating beds. Packing house wastes were treated
by Howson (2O71, 2O72) in a fashion similar to that described
by Rhode (3582), but the primary trickling filter was equipped
with a compressed air supply and wash water and was operated
at a hydraulic load of about 6 mgad (138 gal./ft2/day). The
two final filters were (2O71) loaded at about 1.4 mgad (32
gal./ft2/day), handling a waste which had 1,OOO mg/1 BOD and
after primary filtration was reduced to 25O-35O mg/1 BOD.
The effluent resulting from treatment of packing house waste
with a BOD of 1,5OO mg/1 in a three-stage process could be
reduced by 98 to 99$ (2O72) .
Pharmaceutical and other fermentation wastes have been treated
with a three-stage treatment system (952, 3O58). Quite often
with these systems it was desirable to have an activated-
sludge unit included in the process (952, 956). Patented
equipment, such as that proposed by Dekema (914), often is
composed of multi-stage filtering units with various types
of enclosures and aerated sections to minimize the nuisance
problems and yield an efficient process. Meltzer (2963)
stated that tertiary filtration may be 8O to 1OO# superior
to double filtration, based on hydraulic advantages.
Critique
The treatment plant involved in a three-stage trickling
filter process requires high capital investment, so that
64
-------
either biological or nonbiological waste treatment could be
used. For economical trickling filter use, local topography
should provide sufficient head (6 to 10 feet per filter) or
pumping costs must be included. A factor in favor of the
three-stage system is the development, under heterogeneous
microorganism population conditions, of selected strains of
microorganisms in each filter stratum. The microorganisms
can cope with the influent environment of a complex waste
stream and survive the shock of a strong sewage. With rock
medium, a three-stage trickling filter system would have
sufficient medium depth and thereby provide a long detention
time of the waste water on the biological film. With the
development of media other than rock (e.g. plastic), deeper
beds have been constructed with the resulting longer deten-
tion time. One should consider, however, that once the
biological film has sloughed from the media, biological
activity does not stop. The activity of the microorganism
population, while contained in a clarification tank, has been
demonstrated many times, and even the casual observer has seen
the rising solids in the secondary clarifiers. Yet, Sorrels
and Zeller (4111) and Archer and Robinson (85) reported that
intermediate clarification had negligible effect on efficiency.
Therefore, in a three-stage biological filtration treatment
system, intermediate clarification may not be necessary.
Alternating Double Filtration
Alternating double filtration has been described by Hurley
(2110), Rumpf (3782), Calvert (593), Whitehead (4712), Escritt
(1194), Daviss (891), and others, as a two-stage biological
trickling filter operation which is arranged in such a manner
that the order of the primary and secondary filter may be
interchanged periodically. In a discussion of the paper by
Whitehead (4712), D. M. Watson stated that surface clogging
of the trickling filters was avoided by this process. In
reviewing the development of recirculation, deep aerated
covered filters, and alternating double filtration systems,
W. Watson (464O) suggested that, when alternating double
filtration was employed, dissolved oxygen in the primary
filter effluent caused sloughing.
At the same time that the high-rate biofiltration type
processes were being developed in the United States alter-
nating double filtration systems were being employed. Hurley
(2118) reviewed O'Shaughnessy1s observation (3253) that a
clogged filter could be restored by dosing it with partially
purified sewage or filter effluent, i.e., alternate the first
and second stage recycle. For maximum efficiency (2118) it
was recommended that each part of a filter should always carry
out the same type of work, which does not occur in the process
65
-------
of alternating double filtration. He suggested that greater
nitrification would be obtained if other means were used for
keeping the surface clean, e.g., by using coarser medium in
the surface layers, or by covering the filter so that scav-
enging organisms would come to the surface and destroy the
slime. With alternating double filtration, satisfactory
results have been obtained with doses of 150 to 2OO gal. per
cu yd per day (2121) . Filters operated by alternating
double filtration do not require as deep a bed as single-
stage filtration beds, according to Escritt (12O4). With the
high-rate and alternating double filtration systems available,
Herriot (1899) decided the selection of the processes for the
wastewater treatment from an economic standpoint. Discrep-
ancies in using the National Research Council formula for
calculating effluent BOD values on alternating double filtra-
tion systems were pointed out by Wishart (4793).
The alternating double filtration system was applied mainly
in countries other than the United States. Belgium (4352),
England (891, 494O) and other European countries have found
the system satisfactory. In recently designed plants (179,
891, 4352, 494O), alternating double filtration and recircu-
lation of effluent have provided flexibility in operation.
Waste stream from tanneries, gas works, dairies and other
industries have been treated satisfactorily by using alternat-
ing double filtration (35, 3121, 48O4, 5446, 5522). In treat-
ing the tannery waste (35), alternative systems were tested,
and the results of the systems were judged by the degree of
nitrification achieved with alternating double filtration
being the best, contrary to Hurley's doubts (2118). The
effluent from the alternating double filtration plant at a
dairy in South Africa was used as cooling and boiler feed
water (5446). Alternating double filtration in the United
States, such as at Marysville, Ohio, (4804), has occasionally
treated combined sewage and dairy waste with satisfactory
resulcs.
Critique
Alternating double filtration was quite common in Great
Britain and other parts of Europe, while the application in
this country has been limited. This limited application
may be due to economics, since many of the two-stage biologi-
cal filtration plants were built to operate by gravity flow
and a major modification would be necessary to install suf-
ficient pump capacity to reverse the order of the filters.
At the same time, the development of the high-rate (biofiltra-
tion type) trickling filter process approximately paralleled
the development of alternating double filtration. The high-
rate filtration process was an American innovation, while
alternating double filtration was chiefly promoted by the
British. For conditions which may exist in a waste stream
which cause troublesome biological growth (e.g. high
carbohydrate waste), alternating double filtration has an
application.
66
-------
In essence, in this process the filter is being rested and
washed by running the secondary effluent over the primary
filter and vice versa. The operational technique of resting
and rinsing has been used in England, this country, and others
for scores of years. In some of the early work reported by
Jenks (2305) and Halvorson (1667), plugging problems were
overcome by an increased rate of recirculation. This re-
circulation, which lowered the organic load and raised the
hydraulic load, may be thought of as a resting and rinsing
procedure similar to alternating double filtration.
Roughing Filters
Roughing filters have been described in the book, "Sewage
Treatment Plant Design," (4916) as systems used to reduce
the organic load in which subsequent treatment may be applied
to the effluent or where intermediate treatment is required.
The design of the roughing filter (4916) differs from other
filters principally because the determining factor is the
volume of the waste (Table 1) , as well as the high BOD of
certain waste which is to be handled. Limitations in the
medium available for use in roughing filters were overcome by
considerations outlined by Pearson (3328). Roughing filter
design required (3328) an open flow pattern which allowed
the unhindered flow of the microorganism-waste stream down
the medium column and encouraged uniform distribution of
this flow. Organic loadings on these filters have been re-
ported (1O29) to vary from 5,OOO to 2O,OOO Ib BOD/ac-f/day
(115 to 46O Ib BOD/1,OOO ft3/day). Recirculation was normally
practiced to maintain efficiency and keep the biological
film in a wetted condition (681, 1O29). Criticism was made
of the low efficiency of the roughing filter, based on the
difference between the influent-effluent BOD, until effi-
ciency was viewed as the total pounds of BOD removed per
volume of medium, a quantity almost twice that of a normal
high-rate filter (1029).
Roughing filters have been used for many years and their per-
formance was published, for instance at Atlanta, Georgia, (5261),
in which 12 inches of stone media with a possible rate of
175 mgad (4O25 gal./ft2/day) were used on Imhoff tank effluent.
Quite often in the conversion to a two-stage trickling filter
plant, the primary trickling filter was operated as a roughing
filter (547O). An advantage of the roughing filter is its
utility on waste streams of fluctuating strength and flow
characteristics. The waste stream was saturated with oxygen
in the roughing filter, but usually needed further treatment
by other secondary processes (3474). A typical use of the
primary filter in a two-stage operation as a roughing filter
was published by Jenks (23OO) , during his discussion on the
operation of Biofiltration systems. The super-rate filter
was a type of roughing filter as defined by hydraulic and
organic loadings.
67
-------
In treating the waste waters from abattoirs, various forms of
pretreatment were used as well as roughing filters, followed
by percolating filters (3381). Hansen (1719) reported in
1944 that a routine preliminary treatment using a roughing
filter could be used to decrease the organic load to sub-
sequent processes by about 50$.
Several fabricated media, such as tiles, perforated blocks,
wood lath, asbestos-cement sheets, ceramic random packings
such as Raschig rings and Berl saddles, and more recently
plastics, were used (3194, 5196) in roughing filters. The
development of plastic media provided (681) the design engi-
neer with an economic solution to treatment of high organic
influent waste streams. Commercially available plastic media
were discussed by Noble (3194) and others (98, 1O83, 5196) .
A previous limitation of some of the fabricated media was
that they were not economically competitive with conventional
random packed media. However, Chipperfield (681) discussed
the economics of plastic media for roughing filters and
concluded that substantial (25 to 55$) savings in capital
cost over conventional filters and activated-sludge plants
are possible.
Critique
Most of the primary filters with plastic packing in a two-
stage system may be operated at high hydraulic and organic
loadings, as indicated under super-rate filters in Table 1.
Many of the applications of two-stage trickling filter
system, which have been reported in previous sections, in-
volved the operation of the primary filter as a roughing fil-
ter. A description of this operation by Eckenfelder (1077)
and Busch (549) was discussed in the theory section. They
indicated the advantages of using roughing filters for in-
dustrial or municipal waste treatment where occassional high
organic shock loads occur. Generally, it has been recommended
that a roughing filter should not be used as the total treat-
ment for an installation. It is used advantageously ahead of
activated sludge and other secondary processes. The roughing
filters were characterized as deep beds with media which had
large void spaces and were dosed continuously, the effluent
being recirculated as the operation required. The highest
hydraulic loading found in the literature approached five
hundred million gallons per acre per day (11,5OO gal./ft2/
day). However, this high hydraulic loading rate was applied
to a shallow bed filter. The depth of the roughing filter
has been quite dependent upon the type of medium employed.
Polishing Filter
Polishing filters have been used in multi-stage treatment
facilities to decrease the remaining soluble organic and
suspended pollutants in the effluent of a conventional
68
-------
waste treatment plant (5574); for example, the slow sand filter
(698) and intermittent sand filter (2984). However, sand fil-
tration, which incorporated some biochemical degradation and
mechanical filtration, required extensive land area per unit
volume of sewage treated and was, therefore, replaced by other
systems, as described by Metcalf and Eddy in their book on
"Disposal of Sewage" (2984). The influent waste from a popula-
tion of 2,50O, or a flow of 25O,OOO gpd, is considered the
maximum to be handled by intermittent sand filtration (4916).
For polishing filters, a crude type of trickling filter was rec-
ommended by Dibdin (946) and trickling filters, following
partial treatment by aeration, were reported by Thomson
(4380) to handle waste flows in Great Britain. In the two-
stage operation, as previously discussed (1719), if the
primary filter is considered as a roughing filter, then the
secondary filter is a polishing filter. This application
has more significance in design consideration than semantics,
as indicated by Bryan (516), in using polish filters on in-
dustrial waste streams. In designing a polish filter to
produce a high quality final effluent, the retention time
must be increased on the surface of the media. Deep bed,
high specific surface area media have been used to attain
effluent polishing (3328). The organic load applied to pol-
ishing filters has been reported as between 50 and 2OO mg/1
BOD (3328) or 900 pounds BOD per ac-f/day (2O.7 Ib BOD/1,OOO
ft3/day) . The hydraulic load may be similar to that applied
to low-rate trickling filters (Table 1), between 1.1 and
4.4 mgad (25.3 to 1O1.2 gal./ft2/day).
Critique
The design of polishing filters has quite often been governed
by necessity rather than by choice. When an existing low-
rate trickling filter became overloaded, quite often the
procedure has been to insert ahead of it a high-rate trickling
filter or activated-sludge system. With most of the organic
load being removed and the flow being buffered, the low-rate
trickling filter would remove a major portion of the remain-
ing BOD and result in a stable effluent. This was shown to
be a very economical solution, since the existing physical
plant was used.
Contact Aeration
Contact aeration is an aerobic process whereby the supporting
medium is covered with the biological film under submerged
conditions and aerated by compressed air (p. 193 of 4916).
This system was first suggested by Buswell, according to
Metcalf and Eddy (2984), to satisfy a need in waste treat-
ment somewhere between trickling filters and activated-
sludge plants. The use of this technique was common in the
late 192O's and early 193O's in this country and abroad
(315O, 4O37). A typical contact aerator (2984, 5364) is a
69
-------
tank with a false wood bottom, consisting of 2" x 4" planks
set on edge with 0.75-inch slots between them to allow the
air to rise through a coke filling. The coke was kept in
place by a galvanized steel wire mesh one inch square, and a
layer of broken limestone 2 inches thick over the mesh pre-
vents the coke from floating. Other systems were available
(3864) which used movable wooden bodies of small volume but
large surface. Aeration systems were described, typically,
designed by Passavant (33O7), with primary emphasis on the
distribution systems. Hydraulic loadings were reported
(2984) to vary from one-half to two million gallons a day
with contact periods of from 1O to 9O minutes. During World
War II, the contact aerator treatment system was installed
at many of the military installations, but the results were
not altogether satisfactory. The American Society of Civil
Engineers Manual of Engineering Practice No. 36 (4916) stated
in 1959 that ... "The contact aeration process has not been
widely accepted and there are few published operation data."
Critique
The application of the contact aerator has been sporadic.
Goldthorpe, in discussing Chipperfield's paper (674), wondered
if plastic media had been considered in the application of a
tauchkbrper, which is a contact aeration system. The system
may have an application in present day usage. However, one
should be aware in operating a type of activated-sludge plant
without a high degree of control of the biological solids
that operating problems would arise. The design of this
system without more available data would require experimental
investigations. The advantages of the contact aerator proc-
ess are: (a) its ability to hold a quantity of microorganisms
even through a shock load of toxic or high organic waste
(detrimental to part of the biopopulation) , (b) minimum prob-
lems with distribution, (c) little freezing, and (d) control
of odors and other nuisances. Cleaning the media would be a
disadvantage in maintenance. Based on experimental investi-
gations, however, it is entirely conceivable that this hybrid
between the trickling filter and activated-sludge system
could be designed to optimize the strengths of each system
and minimize the weaknesses. Qualified operators would have
to be available.
MULTI-UNIT SYSTEMS
The review and discussion of multi-unit systems provide
additional information on the applicability of trickling
filters. Trickling filters were designed to operate in
parallel, in series with activated-sludge units, in series
with lagoons, in series with septic tanks, and in series
with Imhoff tanks and other units.
70
-------
Trickling Filters In Parallel
A procedure for handling an overloaded biological trickling
filtration plant either in a single stage or two stages
is parallel expansion (1788, 338O). Hopper (2O29) reported
on the design of a waste treatment plant which was comprised
of three sections of trickling filters, all operating in
parallel, handling high and low level sanitary waste as well
as industrial wastes. Flexible designs enabled many bio-
logical trickling filtration plants to be operated in parallel
or in series, with and without circulation (1341, 30O8, 5538).
Operations of parallel trickling filters in colder climates
have indicated high-rate filters are affected by temperature
(2871). Froman (1362) described the design of two parallel
plants to provide a means of testing and comparing new
chemical treatments. The effluents of parallel operating
portions of the biological filtration plant are mixed and
further treated before discharge (338O).
Critique
Trickling filters in parallel were a design concept which was
used (4223) to handle plant expansion, overloaded conditions,
specialized waste problems, and other requirements which would
be satisfied by having an additional plant operating alongside
an existing facility.
Further comments are found in preceding critiques on single-
stage, two-stage, high-rate, and low-rate trickling filters.
Trickling Filters and Activated-Sludge Process
Trickling filters and activated-sludge units have been used in
sewage operations to overcome the limitations of each process.
Pruss and Blunk (3474) indicated that the main disadvantage
of the activated-sludge process is its sensitivity toward
the variations in the sewage and this sensitivity is over-
come by pretreatment on trickling filters. According to Ball
(159), when trickling filters and activated-sludge units are
used in series the plant capacity is increased. Also, where
there were large seasonal variations in the population served,
the activated-sludge process alone or in combination with
percolating filters is most economical. Installations receiv-
ing peak loads of brewery (4583), textile (2037), and paper
wastes (1449) have successfully used a combination of trick-
ling filter and activated sludge. When the activated-sludge
process preceded biological filtration, Sickert (4O13) found
that not only was clogging prevented, but also the efficiency
of the plant was measurably increased. In the discussion of
Whitehead's (4711) paper, Watson (4712) was a proponent of
the use of preliminary treatment by activated sludge to avoid
clogging of the surface of trickling filters. However,
71
-------
Whitehead (4711) stated that bioflocculation followed by
trickling filter treatment was more economical and effective
than either activated sludge or trickling filter alone.
Erode (1168) cites the advantage, where high purification is
required, of using series operated percolating filters and
activated sludge. Ball (159) and Fischer and Thompson (1279)
indicate that when there are large variations in the flow, or
when an activated-sludge plant is overloaded an economical
approach was combined trickling filter - activated-sludge
operation. This approach in handling an overloaded situation
was supported by Hurley (21O8), Garner (1416), and Allen (35).
Various wastes were treated effectively with combined acti-
vated sludge - trickling filter operations in Germany (4O13),
England (4583), Belgium (640), Holland (2027), Poland (1405),
and in this country (956, 1449) . Typical of the wastes that
have been treated are the effluents from municipalities (64O,
1308, 1979, 2027, 4O13), chemical industry (335, 956), pulp
and paper (14O5, 1449, 3976), malting and brewing (1773,
4583), textiles (4O1), Pharmaceuticals (2036), and food proc-
essing (533O). Loading and design parameters have been re-
ported by Cauterman (64O) for a population variation of 8,OOO
to 50,OOO people, which developed a fluctuation in flow of a
maximum 4,50O cubic meters per day (1.19 mgd) versus 800
cubic meters per day (O.21 mgd) minimum. The filters were
designed (640) to treat the low discharge rate, with the
added load beyond the capacity of the filters to be handled
by a preceding activated-sludge unit in series. By using the
combined system at Volksdorf, Germany, an effluent of 1O-2O
mg/1 BOD5 was established with waste flows up to 4,5OO cubic
meters per day (1.19 mgd) (4013). Hurley (2108) reported
that 93$ removal of the BOD was obtained in a combined waste
treatment plant which had an influent composed of 4O# trade
waste plus sewage from 120,OOO people (2108).
Bilfilters loaded with normal sewage (2OO-25O mg/1 BOD) gave
70 to 8O# removal with an effluent suitable for activated-
sludge treatment, according to Fischer and Thompson (1279).
Twenty-five hundred cubic meters of sewage per day (656,000
gal./day) were treated in the plant with the filters loaded
at 1.2 cubic meters sewage per cubic meter filter media
(9O gal./ft3) and with the aeration time in the activated
sludge tank of 1.3 hours (O.5 m3 air/m3 of sewage, or 1 ft3
of air/15 gal. sewage) developed sufficient treatment. By
adding settled filter effluent to primary settling tank
effluent, the quantity of these effluents treatable in the
aeration tanks was increased 30 to 4O$ (2441) with no
decrease in efficiency.
Hall (1655) described the Mineola, New York, sewage plant
which has five aeration tanks with a total capacity of
22O,OOO gallons providing five hours' retention time for a
flow of 1 mgad (23 gal./ft2/day), plus eight percolating
filters, each HO by 125 feet.
72
-------
The concentration of sludge in the aeration tanks is main-
tained at 18 to 24$ in the summer and raised to 28 to 36$
in the winter. This combined waste treatment plant was
treating one million gallons a day satisfactorily. Popel
and Daser (3427) described a trickling filter - activated-
sludge plant which treated the waste from a population of
70,OOO in which the filters were loaded at a surface flow
of eight meters per hour (26.2 ft/hr) to assure sludge wash-
out. The aeration tanks were designed to operate under
stage loadings and step aeration.
Studies by Gellman (1449) indicated that plastic media trick-
ling filters, loaded at 1OO-25O Ib BOD per 1OO cubic feet of
media (1,OOO to 2,5OO Ib BOD/1,OOO ft3/day), would produce
5O$ removal with the advantage of cooling, which suggested
the use of combined trickling filter activated-sludge treat-
ment for handling the pulp and paper mill waste. Along with
Gellman, Popel and Daser (3427) and Allen (35) used trickling
filters followed by the activated-sludge process without
intermediate sedimentation. Pruss and Blunk (3474) and
Kershaw and Finch (2441) indicated efficient treatment with
trickling filtration and activated sludge with intermediate
settling. Recirculation of waste,activated sludge, clarified
effluent, and mixtures thereof have been reported by Whitehead
(4712), Allen (35), and Schreiber (3893) to be effective in
maintaining the flora on the biological trickling filter and
aiding subsequent treatment in the activated-sludge system.
Critique
The combined trickling filter - activated-sludge system is not
new, but has gained some popularity recently due to higher
effluent standards and other requirements. Many engineers
take one side or the other, and either use trickling filters
in their design or the activated-sludge process exclusively.
Many studies based on the comparison of the two processes
illustrate this, as may be noted in Part IV. The literature
reflects the efforts of a few workers who have used the
advantages of each process to solve difficult problems. It
is obvious that there are applications for each process and
for the combined system. A roughing tower has simplified the
operation of activated-sludge plants. Fluctuations in organic
load, toxicity and temperature are just a few of the factors
which may be dampened by the trickling filter treatment.
However, the effluent from this process will not be acceptable
for discharge and many, in fact, have several milligrams per
liter BOD in the final effluent. This type of effluent is
ideal for further treatment by activated sludge.
Actual design criteria were not clearly defined, but trends
of past experience were indicative of the size and loading
of the systems. Investigators have not reported the data
73
-------
completely to exploit the usefulness of this concept. New
construction materials, better qualified operators, more
automatic control, and higher effluent standards all point
to the combined trickling filter-activated sludge waste
treatment plant as a reasonable solution to complex pollution
problems.
Trickling Filters and Lagoons
During his thesis study on removal of low level radioisotopes,
Lawrence (2642) investigated the application of biological
filtration and lagoons (large shallow ponds designed for days
of detention) for the removal of phosphorus-32, iodine-131,
potassium-42, cerium-141 and -144, and strontium-89 among
others. The results showed that primary sedimentation is
relatively ineffective in removing colloidal and dissolved
radionuclides from sewage, and the effectiveness of biological
filtration is limited by the ability of the filter to remove
the isotopes from the liquid phase and concentrate them in
the solids, leaving the filter in a settleable form. The re-
moval of radionuclides within the filter was found to be due
to concentration of the nuclides by the zoogloeal slimes.
Another combined waste which has been treated by the use of
trickling filters and lagoons has been domestic as well as
from a milk processing and poultry processing, recorded by
Grewis and Burkett (158O). This plant used nutrient addition
and a roughing filter and produced an effluent with 1O to 36
mg/1 of BOD. Anderson et al. (73) described the biological
filtration plant at Lansdale, Pennsylvania, where the effluent
was treated by mechanical filtration and lagooning, and was
subsequently used as cooling water for the municipal power
station. Kempton and Roskopf (2428) added low-cost sedimen-
tation lagoons after the percolating filters to give 96$ BOD
removal, which also raised the capability of shock load
handling for the plant. A typical satisfactory operation of
biological filtration, followed by lagoon treatment before
discharge into a small creek, was described for a small
English village (5555). However, as a step further, the
installation of an Air-Aqua aeration system in the lagoons,
which were used as final treatment after biological filtra-
tion in Regina, Saskatchewan, doubled the sewage capacity,
as reported by Goodnow (1517).
An interesting problem in treating distillery waste in
Cuba was noted by Maiz (2841), where the effluent from
lagoons was further treated by two-stage biological filtra-
tion to achieve a high degree of nitrification. For the
treatment of cannery waste Warrick et al. (4621) gave the
details of design and operation of lagoons and the operation
of lagoons in conjunction with high-rate Aero-Filters. The
waste water was recirculated through the filter at a constant
rate of two million gallons per day. In case of high organic
loads, they recommended that the lagoons should be designed
74
-------
to retain 25% of the waste discharged to them. The treat-
ment of sugar beet factory waste waters was investigated by
Carruthers et al. (629). Sedimentation in lagoons alone
was not a very satisfactory method, but after dilution of
the waste with treated sewage, percolating filters were
used with sedimentation in lagoons, both before and after
filtration. The treatment of potato wastes was the topic
of a symposium in 1965 (533O), and the processing waste
waters were amenable to treatment by trickling filters and
sewage lagoons. Treatment of phenolic waste waters was ex-
plored by Biczysko (32O), using percolating filters, 14 to
16 yards deep, followed by oxidation channels (aerated
lagoons), which were able to reduce phenol concentrations
of the waters from 5OO to 95O mg/1 down to 1 mg/1.
Critique
Lagoons and trickling filters have been used for many years.
Unfortunately, lagoons are very seldom properly designed and
problems with odors and other nuisances are common. By stor-
ing the trickling filter effluent in lagoons, some of the
nuisance conditions may be avoided. Many of the existing
lagoon facilities were built as "stop-gap" methods of treat-
ment, especially for difficult trade waste waters. Lagoons
ahead of trickling filters may be used as a type of balanc-
ing tank to allow a more uniform waste to be applied to the
filter. The large area requirements of lagoons are not
compatible with the recent trends of less land area being
used by deep bed trickling filters. As a step to decrease
the land problem, aerated lagoons have been used. The
aerated lagoon following the filters provides a semi-activated
sludge type treatment without the advantage of controlled
suspended solids or the disadvantage of sludge recycle pumping
costs.
Trickling Filters And Septic Tanks
Investigators, such as Meckel (295O), reported that for farm
homes septic systems, followed by either trickling filters
or sub-surface disposal, were preferred early in the 19OO's.
Hewitt (1926) discussed the difficulties encountered in
obtaining a high quality effluent at a reasonable cost. For
small communities in Australia, he recommended primary treat-
ment by septic tanks, followed by secondary treatment, using
percolating filters with or without subsequent land treatment.
Abbott (2) described isolated hospital sewage treatment for
7,OOO people, using septic tanks with six-hour detention time
and trickling filters of about one-half acre. Short detention
periods in the septic tanks were reported by Ken worthy (2437),
together with high dosage rate on the filter, to be responsible
for the overloaded filters in Perth, Australia.
75
-------
German construction of septic tank - trickling filter plants
was common as early as 1898, as reported by Mohle (3035).
As these plants became overloaded, a large storm water tank
and sewage retention tanks were installed to improve the
effluent. Wood shaving trickling filters following septic
tanks were suggested by Girard (1483) in removing colloidal
material, such as iron sulfide, from rural district effluent.
Septic tanks and percolating filters were stated by Davies
(881) to be cheap, since the low initial costs, the low
manpower requirements for maintenance and operation, and an
effluent of consistent quality were definite advantages.
Great Britain developed standard designs (2811) consisting
of septic tank and percolating filters for rural districts.
The sewerage systems for air training schools in Canada
during World War II took advantage of the simple operation
of septic tanks followed by biological filtration, according
to Langman (262O). Specialized factors in design were dis-
cussed by Subrahmanyan (4262) for percolating filters suitable
for treating effluent from septic tanks in the rural districts
of India.
In a discussion (5518) of a paper by Faulkner, regular main-
tenance of small sewage treatment plants, each consisting of
a septic tank and percolating filter, but of different designs,
was emphasized. Faulkner (1244) stated that treatment in
septic tanks, followed by biological filtration, is suitable
in most cases for sewage disposal in rural villages. Lynas
(2798) reported, in 1951, that small sewage treatment plants
comprised of septic tanks and percolating filters were
appropriate for conditions in Northern Ireland.
After an examination of some of the small sewage works,
Escritt (12O6) reported that the choked condition of the fil-
ter was usually due to leaving the septic tank unsludged for
too long a period of time and that the size of the trickling
filter media should be at least two inches in diameter.
Rolston (3683) presented a paper at the Engineering Society
in British Columbia where he states that an adequate primary
treatment for small communities was by septic tanks and
secondary treatment by biological filtration.
Studies were carried out on the BOD loadings of percolating
filters in Japan (2O32) for effluents which had been given
primary treatment in septic tanks. The conclusion was that
adequate capacity of filters could be determined for treat-
ing domestic sewage and sewage from shops, offices and
factories. Developments in the design of septic tanks and
effluent treatment by biological purification through filtra-
tion in Austria were published by Lengyel (2683). He sug-
gested passing the settled sewage through a percolating fil-
ter consisting of fibrous materials on perforated trays and
separated by air spaces. Other modifications resulted in
patents, e.g., Bomhoff (398). Liepolt (2718) disclosed
similar work of septic tank effluent treated by fiber
76
-------
percolating filters. Limitations were proposed on the septic
tanks by Peel (3331), in 1966, by noting that inadequate
detention time, lack of servicing and maintenance, and in-
correct estimate of required capacity were instrumental in
yielding poor quality effluents.
Critique
Septic tanks are still popular for waste treatment at rural
locations. Prior to 193O, large systems servicing cities
of several thousand used septic tanks. The effluent from
the septic tanks used today in modern suburban living is
passed through field lines of gravel, broken rock, etc.,
under aerobic conditions. Basically, the system is good, but
it does require proper design and regular maintenance, as
many of the workers have cautioned. There are certain package
plants in operation today that operate quite satisfactorily
as combined septic tank - trickling filter plants.
Trickling Filters And Imhoff Tanks
Imhoff tanks, which are rectangular with an inverted pyramidal
bottom and a chambering arrangement, have been used in sewage
treatment in conjunction with biological trickling filters
since the inception and development by Karl Imhoff, Germany.
Papers by Hommon (2OO1), Perry (3334), Radebaugh (3499) and
Imhoff (22O4) are only a few examples of the extensive liter-
ature on the utilization of Imhoff tanks with trickling
filters. Applications in New Jersey (5362) and in Kansas
(397O), as well as in foreign countries, for example, in
Brazil (1463) and Australia (52O8), are also numerous. The
Imhoff tank operating in series with percolating filters
and sedimentation tanks has several advantages, such as
economics, according to Timmerman (4394), at locations with
small populations [Carpenter (625), Doubleday (1O08), Greeley
(1549), etc.], or at remote locations [Dye and Mundt (1053),
Sharp (3970)].
Hommon (1999) reported that for a waste from 7O,000 people
with an oxygen consumption value of 86 and suspended solids
of 238 mg/1 a recommended design would be grit chambers,
Imhoff tanks, sprinkling filters, and settling tanks. A
bacterial removal efficiency of 82$ and a reduction in
suspended material from 85 to 3 mg/1 were obtained on an
Imhoff tank - sprinkling filter plant in California (5076)
early in the 190O's. Bacterial investigations by Hotchkiss
(2O45) indicated chemical and biological activities caused
by the film around the stones of a sprinkling filter are
similar to those in an Imhoff tank.
Bass (211) reported the rejection of plans to build a direct
oxidation plant for Austin, Minnesota, in favor of an Imhoff
77
-------
tank - percolating filter plant to treat the 1.33 mgd at a
cost of about $22O,OOO. A description of the Akron, Ohio,
treatment plant by Backherms (138) indicated that the Imhoff
tanks were designed for a settling time of 2.26 hours at
33 mgd with sludge compartments providing 2.7 ft3/head and
14 acres of trickling filters (operated at 54 gal./ft2/day) .
In evaluating new designs for the expansion of the Akron
facility, Kemmler (2426) stated that the combined cost per
million gallons of sewage for operation, construction, and
interest at 4$ over a period of 3O years was $21-$23 for
Imhoff tanks - trickling filters, $15-$38 for chemical precipi-
tation, $2O-$37 for the activated-sludge process, and $3O-
$46 for intermittent sand filtration, and concluded that the
last two are too expensive for consideration for Akron. Data
on dissolved oxygen and biochemical oxygen demand of the final
effluent of the combined system were reported by Tatlock
(43O6). Replacement of seepage pits and septic tanks by
Imhoff tanks - percolating filter system in Argentina increased
the treatment capacity of the daily volume of sewage by 50$
(628) .
The sewage system to treat a highly acid tomato cannery waste
(six-weeks peak load) was described by Hicks (1934) as con-
taining Imhoff tanks and percolating filters (high rate).
The BOD loading was estimated to be 51O Ib/day, 35$ of which
was removed by the primary treatment and 55$ of the remain-
ing BOD was removed by the percolating filter. An Imhoff
tank - biological filtration plant was found to be sufficient
to serve a population of 80O at Bristol, South Dakaota (4904),
and an additional creamery waste of 5,OOO gallons or 68 Ib
BOD/day, where the total load was not expected to exceed 136
Ib BOD/day. Pieczonka (3379) reported in 1956 that an
Imhoff tank - percolating filter plant was operated in such a
manner that the secondary treatment by biological filtration
can be stopped during the winter months. No adverse effects
were observed, providing the operation was gradually reduced
in late October and early November. The system became
operational within three weeks early in May of the following
year.
An Imhoff tank - high rate percolating filter plant operating
with recirculation was described by Kimball (2481) to handle
a designed BOD load of 144O Ib/day, of which no more than
444 Ib/day should be contributed by waste waters from the
canning of pickled cucumbers. However, these waste waters
were found to contain a BOD of 1,277 Ib/day, which caused a
severe overload; the treatment plant recovered after the
passing of these wastes and satisfactory results were obtained.
The treatment of a sugar refinery effluent was published by
Shukla et al. (4O06), using an Imhoff tank - biological fil-
tration plant plus nutrient supplements. A reduction in
pollution of the wastes of 7O$ was obtained, which increased
to 92-95$ after stabilization of the treated effluent in a
tank for 72 hours.
78
-------
Eddy and others (1O93) reported favorable applications of
Imhoff tank - trickling filter plants, but could not recommend
their use in Chicago, in the 192O's, because of objectionable
odors, and substantially the same cost would be involved for
an activated-sludge system. Other Imhoff tank - trickling
filter plants (5364) which were at the limit of their capacity
were modified by building a new Imhoff tank with contact
aerators or other aeration devices (4440, 5365). A comparison
of the cost of an activated sludge versus on Imhoff tank -
sprinkling filter plant resulted in the construction of the
activated-sludge plant at Lawton, Oklahoma (5367) in 1929.
Hudson (2O82) compared the cost of five sewage treatment
plants as $28 to $87 per million gallons and $O.83 to $1.7O
per head for Imhoff tank - trickling filter plants (Illinois
and Ohio), and from $26 to $43 per million gallons and $1.5O
to $2.OO per head for activated-sludge plants on an annual
basis. After a seriously overloaded condition was reached
on the Imhoff tank - trickling filter plant at Rolla, Missouri,
a larger trickling filter sewage treatment plant was con-
structed (4889). Larson (263O) described the modernization
of Detroit Lakes, Minnesota, Imhoff tank - trickling filter
plant to be: a high capacity Aero-Filter with tile medium?
the old Imhoff tank remodeled into a sedimentation tank; the
original standard percolating filter, and a final sedimentation
tank. Typical of many Imhoff tank - percolating filter plants
was that at Marion, Ohio (5OO). This plant gradually lost its
efficiency because of lack of maintenance and normal wear of
equipment. The remodeled plant (originally built in 1924) is
a combined activated-sludge process - trickling filter plant.
Other examples of modernization of Imhoff tank - trickling
filter plants were reported by Hagglund and Omachi (1646),
Schaetzle (3857), and Cosulich (796).
Critique
Most of the literature dealing with trickling filter-Imhoff
tank applications were favorable for the continued use of
this means of sewage treatment. The overall plant efficiency
may be lower than that of a separate sludge digestion plant,
but the low manpower requirements may offset that for certain
locations. As noted by several investigators, the economics
of the application appear to be the limitation compared to
other more rapid systems.
The number of Imhoff tank - trickling filter plants recently
constructed is very small compared to those in the 193O's.
The system should have an application, with more and more
attention being directed toward industrial wastes. The
process is claimed to be an adequate industrial waste system
requiring minimum operation and equipment. However, the
trend was demonstrated that many of the Imhoff tank - trick-
ling filter plants are being converted to more reliable,
efficient processes, particularly at municipal treatment
plants.
79
-------
Trickling Filters With Other Methods
Brauss (448) compared the effect of treatment in irrigation
fields and in percolating filters. Irrigation fields removed
turbidity more completely and reduced bacterial number and
coli titer to a greater degree than percolating filters.
Marcotte (2860) also compared the treatment of sewage by
irrigation, biological filters and activated-sludge process
in France. A review by Heidusenka (1856) contains the prog-
ress in sewage purification up to 1928, and the comparison
of the methods. The Emscher filter was mentioned for the
treatment of phenol-containing coke plant waste, provided the
effluents are diluted with sufficient domestic sewage. Reich
(3532) also reported the use of biological Emscher filters
for effluents from coke plants. Kres (2568) stated that
phenols from industrial wastes may be destroyed quite readily
by biological treatment in Emscher filters or by activated
sludge.
The Emscher tank is another concept which has been applied
to sewage treatment according to Datesman (874), Schmidt
(3875), and Csepai (836). Maier (2835) evaluated several of
these combined treatment tanks in 1914, and evaluated their
cost versus chemical treatment. Emscher tanks were reported
by Datesman (874) to give the best results of clarification
and sludge digestion. Scheffel (3859) used Emscher tanks
in conjunction with percolating filters. They had ISO cubic
meters' (39,60O gal.) settling space with 195 cubic meters'
(51,5OO gal.) digestion space to treat a sewage flow of 1,5OO
cubic meters per day (396,OOO gal./day). The use of a two-
stage Emscher tank in conjunction with a closed aerated
trickling filter was reported in 1938 by Se'e (3955) . Emscher
tanks were used at Atlanta, Georgia, (526O), and retained
nearly all suspended matter (213 to 3O3 mg/1) . The effluents
from the tanks contained 6O to 88 mg/1 suspended solids which
provided the sprinkling filters in the next stage with excel-
lent operating conditions (526O). Instructions were issued
by government agencies, for example the Belgian Ministry of
Public and Family Health (5567), on the design and use of
septic tanks and small Emscher tanks in conjunction with
percolating filters. In 1966, Menkens (297O) discussed the
existing theoretical and practical experiences regarding the
efficiency of Emscher tanks and percolating filters. Perco-
lating filters were considered superior to activated-sludge
plants due to their efficiency and adaptability, and espe-
cially their economy. However, large scale sewage works for
3O,OOO inhabitants and above were more efficient when based
on the activated-sludge principle (297O).
An anaerobic process similar to the Emscher tank, which used
an anaerobic contact tank, a percolating filter and a small
lagoon, was investigated by Fall and Kraus (1237). The proc-
ess removed 77$ of the suspended solids and 34% of the BOD, but
the performance of the percolating filter was poor. However,
80
-------
a lagoon was designed which considerably reduced the BOD in
the final effluent from this combination plant, which could
be safely discharged into the receiving body of water. The
"Bio-reduction" process (5318) operated similarly to a pre-
aerated settling system followed by percolating filter or
other secondary process with charred "bio-loam" being added
to the preaeration device.
Critique
Additional multi-unit systems have been used with as many
arrangements as there were investigators. Most of these
applications were a one-time use for a special problem and
would not generally satisfy the requirements of plant design
under other conditions. Emscher, Imhoff, Kremer, Stiag and
Newstadt tanks have been used with various degrees of success.
Most systems that were efficient at capturing solids or could
be used for ultimate sludge or effluent disposal have had
some application with trickling filters. The usual limita-
tion has been economics and total process efficiency.
MEDIUM SELECTION AND BED DEPTH
Pertinent design factors for the medium of biological trickling
filters were given by Bachmann (L36) and by "Sewage Treatment
Plant Design, No. 36," (4916) for the materials used, size and
shape, durability, and placement. The depth of the filter bed
medium has received considerable attention and it will be dealt
with separately in following paragraphs. Standards, such as the
Ten States Standards (5O11) and the British Standards (4951),
have been established, which outlined requirements for an
acceptable medium, such as: the durability of the medium
against spalling and flaking; very limited solubility, freedom
from iron, structural stability and chemical and biological
inertness. Specific sizes and gradings are given, restricting
the amount of fine and oversize material. Standards for the
use of fabricated media stated that the media be evaluated
based on experience with similar waste and loadings before
installation. Procedures were specified for delivering and
handling the material to guarantee that the underdrains were
not damaged and the medium was put in place as designed (5O11).
Medium Selection
Requirements of an ideal medium were reported by Pearson
(3328) and by Chipperfield (678) as: (a) must offer a surface
over which the biological film can grow successfully; (b) the
liquid to be purified must flow over the biological film
with maximum thickness to assure atmospheric oxygen transfer
to the organisms; (c) void space must be adequate for
sufficient ventilation to assure aerobic condition; (d) the
void space must provide free passage for all organic solids
81
-------
shed (sloughed) from the biological film and be rapidly
carried to the next stage of the process; (e) biological
inertness must be demonstrated by not being subject to
degradation nor inhibiting biological growth; (f) chemical
stability must be demonstrated in the presence of dilute
quantities of solvents and organic chemicals, etc.; (g)
mechanical stability must be demonstrated for long-term com-
pressive stress; and (h) materials should be competitive
with any conventional system. A wide variety of media has
been proposed for various waste streams, starting convention-
ally with rock, gravel, coke, peat, then more ingenious uses
of coral, empty condensed milk cans, scrap metal, refuse,
brush branches, wooden blocks, wooden laths, and, later
fabricated elements made of ceramics and clay, cement-fiber,
and most recently plastics. These materials of construction
are covered in Part II.
In a discussion of the effect of temperature and ventilation
on trickling filter operation, Brown (492) emphasized that
no general design criteria were fixed for percolating filter
media to treat trade waste waters, but that each waste must
be considered individually. Smith and Leibee (4074) were of
the opinion that the design of a proper medium was important
for similar reasons, and stated that it was commercially
feasible to produce a better prefabricated medium. Of primary
importance from the cost standpoint was the quantity of medium
required which was discussed by Temple (4335). Mathematical
relations between quantity of medium and efficiency in re-
duction of BOD have been presented, for example, by Archer
and Robinson (85), Germain (146O), Balakrishnan et al. (154),
and Roesler and Smith (3658).
Bayley and his coworkers (227) stated that there was some
evidence that the size of the filter medium has an effect on
the settling properties of the humus generated, such as the
larger size medium produced a suspension which was more
difficult to settle. The size of the stone was attributed
by Montgomery (3O76) to be one of the causes of inferior
results from high capacity trickling filters. The size of
the stone should be from 5 to 7 millimeters (0.2O to O.27 in.)
as reported by Imhoff (2185).
Included in his discussion of the design of percolating fil-
ters, Fox (1328) noted the size of the medium and compared
his recommended results with those of the Water Pollution
Research Laboratory. Watson (463O) reported early in the
19OO's that the size of the medium should depend on the
character of the waste to be treated and that his experience
indicated that the best size was from 0.75 to 1.25 inches
in diameter. Easdale (1O62) expressed the view that the
size of the material for percolating filters should be
82
-------
sufficiently coarse that the humus would be automatically
washed out of the filter. The secondary filters used in
England (426) contained fine graded medium from 1 to O.25
inch. The New Jersey State Health Department in 1935 speci-
fied that the medium be 1 to 2 inches in size and free from
fines (5O95). Jenks (23OO) stated that the medium size
suitable for the biofiltration process should be 1.5 to
2.5 inches for the primary filter and O.75 to 1.5 inches
for the secondary filter. Schreiber stated that a filter
bed material 1O to 15 millimeters (O.39 to O.59 in.) in
diameter could be removed and cleaned to eliminate possible
clogging problems (3893). The medium proposed by Schreiber
(3893) to be used in the "SU" trickling filters was 5 to
10 millimeters (O.2O to O.39 in.) in diameter compared with
4 to 8 centimeters (1.56 to 3.12 in.) in diameter for
ordinary trickling filters.
Meltzer (2965) indicated that maximum efficiency occurred when
the optimum rate of flow is used for a specific size and con-
figuration of the medium. A properly shaped medium, which
had a large surface area, was stated by Smith and Ellison
(4072) to be self-cleaning. However, Schroepfer (39O2) con-
tended that the shape has very little effect on filter ef-
ficiency. An article entitled, "The Significance of Particle
Shape in Relation to Percolating Filter Media," by A. M.
Bruce was published in 1968 by the Water Pollution Research
Laboratory in the Journal of British Granite and Whinstone
Federation (5O8) . Past work on the effect of particle shape
was noted, along with information dealing with structural
integrity, and it was concluded that elongated particles
would provide more void space than equidimensional particles.
Depth of Medium
In the design of biological trickling filters, the volume of
material used to support the biological growth was determined
by the cross-sectional area and the depth of medium required
to produce a desired effluent quality. Easdale (1O62) stated
that, "within certain limits it is generally found that the
same quantity of sewage can be satisfactorily treated per
cubic yard of material whether it is in the form of a shallow
or deep filter." This is in contrast to Escritt (12O4) who
said that filters operating by single filtration should be
as deep as practicable. A typical example of the procedures
used to increase treatment plant efficiency was cited by
Molesworth (3055), in which the next phase of experimentation
on filters treating effluents from rubber production would
be to investigate the efficiency of deeper filters. Another
typical example of using deeper filters was given by Fletcher
and Anderson (1295), where the existing sewage works at
Cumbernauld New Town, England, had its capacity increased
by doubling the depth of the trickling filters. Schreiber
83
-------
(3892) proposed to solve the problems from plugged 6 m (~ 19.7
ft) medium beds by a system in which the medium could be
recirculated for cleaning. Several media were developed to
achieve a deeper, nonplugging, well ventilated biological
filter. Liepolt (2718), Eden et al. (11O6), and Chipperfield
(681) reported on these fabricated media in which the emphasis
was to provide, among other characteristics, a maximum sur-
face area and a maximum void space to facilitate operation
of a deeper bed.
All were not convinced, however, that deep bed filters were
required, as Reimers et al. (3551) showed that a filter
4 feet deep performed as well as one 6 feet deep. Rudolfs
and Peterson (3721) stated in 1926 that filter beds must be
at least 6 feet deep to handle 75 mg/1 suspended solids.
Under the conditions studied, Smith and Rodeberg (4O82) con-
cluded that the maximum practical depth of medium was 1O
feet, depending on the flow conditions. In a discussion of
the differences between low-rate and high-rate percolating
filters, Imhoff (2211) described experiments with a filter
8 meters (~ 26 ft) high and one meter (39 in.) in diameter
which produced a 9O$ BOD reduction. However, he advised
that it was difficult to construct and that depths of from
3 to 4 m (~ 1O to ~ 13 ft) should not be exceeded, but that
instead the filter could be duplicated or effluent recircu-
lated. Keefer and Kratz (2396), in discussing experiments
with high-rate trickling filters at Baltimore, found that at
a flow rate of 1O mgad (23O gal./ft2/day) 47.5$ BOD reduction
occurred through the top 2 feet of the filter, 68$ through
the top 4 feet, and 8O$ through the whole depth of 8.5 feet.
At a flow rate of 15 mgad (345 gal./ft2/3ay), the BOD removal
vas considerably less. This work was followed much later by
Meltzer (2963), who showed that filters 9 and 12 feet deep
have little, if any, advantage over 6-foot deep filters.
Much of the confusing design criteria were strongly criticized
by Montgomery (3O76), in the second of a series of papers
clarifying high-capacity and high-rate trickling filters (as
previously noted in this review). Part of the cause of
inferior results was attributed to injudicious depth of the
filter bed. A typical high-capacity filter was that described
by Halvorson (1667) to be 8 feet deep and 8 feet in diameter.
The shallow, high-rate filter often reported as the Biofiltra-
tion process (133, 1286, 1897, 23OO, 2303, 23O5, 3159) was
designed to operate with 3 feet of medium and recirculation.
Fischer noted (1284) that increasing the effective depth of
the filter above 3 feet had little effect on high organic
strength distillery waste. Depth and recirculation were
correlated by Rigbi et al. (3612) to show that the size
of the filter medium had no effect on efficiency in a re-
circulating filter 9 feet deep or more. The trend of the
depth of medium is shown in Table 2.
84
-------
Table 2
Depth of Medium Trends
Year Depth, feet Reference
1914 8.33 4498
1915 4.6 1O89
1919 5.5 3652
1928 7.5 3054
1932 6-6.5 1958
1932 8.5 2564
1934 6-8 3775
1936 7.5 5485
1936 3 2300
1937 7.5 2108
1945 6.2 5652
1956 30.5 3912
1960 Approx. 30 3329, 681, 3328
1106, 1083
The many applications of trickling filters have dictated
varying depths based on design criteria and local conditions.
For instance, Furman (1384) reported that percolating filters
at a depth of 4 feet were sufficient for the Florida climate
instead of the standard 6 feet recommended for colder climates.
The New Jersey Health Department established a standard in
1935 which adopted the regulation that filter depths should
not be less than 6 feet or greater than 9 feet (5O95). Stan-
dards have been defined, for example, "Guides for Sewage Works
Design," by the New England Interstate Water Pollution Control
Commission (5331). The most prominent of the standards often
referred to is Ten States Standards (5011), which specify
"The filter media shall have a minimum depth of 5 feet above
the underdrain and should not exceed 7 feet in depth except
where special construction is justified by studies."
The theory behind the relationship of the depth of the filter
and the contact time has been discussed previously (Background
and Theory of Process). The mathematical descriptions of these
relations have appeared frequently in the literature, e.g.,
Caller and Gotaas (1395), and Robertson et al. (3647). The
depth relationship developed by Velz (4535) was later modified
by Howland (2064) and Schulze (3915) to expressions relating
contact or retention time in the trickling filter to the
85
-------
efficiency of treatment. Levine et al. (2702) reported that
the BOD load removal of packing house wastes per unit volume
of filter varied directly with the concentration and rate
of application of the waste and inversely with the depth
of the filter. In studies of air flow and economic design of
trickling filters, Ponninger (3410) in 1938 and, much later,
Petru (3342)suggested factors such as ratio of the diameter of
the filter to the filter depth were significant. Using Velz's
(4535) depth formula, Walton (4596) developed a nomograph to
aid in the design of a percolating filter which would discharge
a final effluent with a BOD of 40 mg/1. A dimensionless factor
which related the hydraulic dosing rate, the depth of the
filter, and the BOD reaction rate constant was proposed by
Gerber (1456). Ingram (2222) suggested that a filter loading
ratio, which was the sewage strength expressed in terms of
depth of filter, area of filter, volume of sewage, and weight
of BOD, would provide a simple basis for calculating filter
size. Russian interest in developing a mathematical formula
relating the various factors involved in biological filtration
and size of filter was expressed by Yakovlev and Galanin
(4840) to aid in the design of high-rate percolating filters.
Other Slovak efforts in mathematical relations between factors
affecting performance of percolating filters have been pub-
lished by Tucek and Chudoba (4474), in which depth, hydraulic
loading, and temperature were related as a primary factor in
the design of filters. Eckenfelder (1081) and with Cardenas
(1084) proposed relationships based on published and experi-
mental data in terms of hydraulic loading and depth for any
specific biological trickling filter medium. Caller and Gotaas
(1395, 1397) initially determined depth as an independent
variable, and later, during optimization studies, found that
depth was dependent on the BOD of the filter influent and the
desired BOD reduction, but it was independent of the volumetric
rate of the influent.
Critique
After reviewing the literature published on medium selection
and bed depth, the extensive information fostered an over-
whelming appreciation for the amount of work which the many
investigators have performed. From a rather humble beginning
of essentially using natural material with a rapid percolation
rate placed at a convenient depth to the design and fabrica-
tion of specialized media and mathematical expressions relat-
ing the depth required to achieve desired removals, the path
has been long and tedious. The literature reflected much
apparent disparity at first glance. However, it must be
borne in mind that a more efficient and economical waste treat-
ment process under specific localized conditions was sought.
It is not surprising that at the time Jenks was reporting
that 3 feet of rather small rock were required, others were
reporting that 6 to 10 feet of a coarse medium were required.
86
-------
The significant point dealing with medium selection and depth
appears to be a complex factor involving an interfacial
reaction of the contact time, aerobic biological film, and
the waste stream. Whether the contact time was increased by a
shallow bed of small material with a controlled recirculation
rate, or a deep bed of coarse material with little or no
recirculation,was dictated by the economics, the available
head, and the existing facilities. The requirements for the
medium have been met by several different materials offered
over the years. The indications are that with more under-
standing of the variables involved media will continue to be
developed. The mathematical relations dealing with the design
of biological trickling filters will be discussed in a later
section. However, it has been amply demonstrated in the
literature that previous investigators were well aware of the
effect of depth and many of the other factors of the medium
as they developed their equations and designs.
The economics of a particular situation may advantageously
dictate a certain type of material to be used as the medium
for biological trickling filtration. As long as the require-
ments for an ideal medium are satisfied and sufficient experi-
mental evidence of satisfactory operation is available, there
is little support to suggest the advantage of one medium over
another. Differences in medium will be discussed under the
Construction Materials section.
SOLIDS-LIQUID SEPARATION AND RECIRCULATION
Solids-liquid separation and recirculation are both involved
in secondary clarification. The solids must be removed from
the effluent of the trickling filter. Recirculation of the
clarified effluent, the settled sludge plus effluent, and
effluent prior to clarification has been practiced in a
variety of arrangements. Recirculation and solids-liquid
separation are not independent of one another. However,
solids may be separated without recirculation, such as
primary treatment, and recirculation of nonclarified effluent
is an example of recirculation without solids removal.
Separation
Solids-liquid separation was of significance prior to and
after biological filtration, as published by Dibdin (946) more
than 60 years ago. Primary treatment has been responsible
for the removal of the major portion of grit and settleable
solids (2984). Dibdin (946) stated in 1910 that the aerobic
treatment process on percolating filters provided "the case
where sedimentation or filtration is required to collect the
humus discharged from the percolating beds." This was
secondary clarification. Dahlem (853)concluded much later
87
-------
that high-capacity trickling filter units required good
subsequent clarification of the sewage. Hurley (2121)
cautioned that the selection and design of humus tanks
following percolating filters and sludge removal were im-
portant. Typical of the many systems proposed was that of
Hall (1652) who described a tangential inlet sedimentation
tank.
Factors affecting the sedimentation of humus and effluents
from percolating filters were investigated by Bayley and
coworkers (227) . They determined that it was possible to
predict the effects on sedimentation due to the initial
concentration of suspended matter, the depth of the sus-
pension, temperature, and the degree of purification achieved
by the trickling filter. Unfortunately, no accurate method
was found for predicting the performance of humus tanks from
their design and operating conditions. For the purposes of
design, under quiescent settling conditions, no significant
improvement was obtained by introducing transverse baffles
at the inlet, by reducing the width of the tank, or by
installing longitudinal baffles in very wide tanks. The
fundamental factors were considered by Eliassen (115O) to be
the detention period in the settling tank and the rate of
overflow of the clarified waste at the weirs. Whitehead
(4713), in a written discussion of the paper by Escritt,
stated that 200 gal./ft2/day of dry weather flow should be
provided for both upward-flow and horizontal-flow tanks.
The approach of Berry (296) to achieve better solids capture
was to use upward-flow clarification in a Dortmund-type tank
and automatic sludge removal. Upward flow sedimentation has
been used ahead of trickling filter waste treatment by
Sviridov (4271), as well as by Watson (4630) in post-trickling
filter applications. Watson reported in the early 1900's
that the best sedimentation was obtained with an upward flow
of about 21 ft/hour. A sludge-blanket principle was success-
fully used by Barnard (180).
Mechanical filtration by various sand filters to achieve solids
capture has been described in several treatises (2984, 4916).
Detailed design and operations information for the mechanical
filtration of sewage and sewage sludge with several types of
filter medium was dealt with by Dickey (957) in his book.
Typical of the many patents issued on various mechanical
filtration schemes was that to von Roll (3678) and others
(4098), covering various processes of cleaning the medium.
Centrifugation has been considered off and on since the 1920's,
as noted by Fuller (1376). In attacking the conditions of
many wasted acres of sewage farms around Berlin, Bottcher
(418) suggested that centrifugation should be considered
for sludge dewatering and drying. He also mentioned that
88
-------
the agricultural value of sewage is greater after mechanical
treatment than after biological treatment, which favored
centrifugation. More recently, Ingram and Edwards (2225)
successfully applied centrifugation prior to the waste being
applied to air-gapped trickling filter medium. Centrifuging
of the settled sewage before the filter reduced the BOD to
about 67% of the total BODf and the filter then removed 63%
compared with 53% for mixed settled sewage. The overall
removal of BOD from settled sewage by centrifuging and filtra-
tion was 75%. Few suspended solids remained after filtration.
An objection was raised by Garrison and Geppart (1417) who
used several different processes to remove suspended material,
such as greases from packing house waste. Their results
indicated that the high capital investment and operating costs
ruled out: centrifugation.
Solids-liquid separation has been reported so extensively
that a comprehensive coverage of the subject is beyond the
scope of this review. Of the several unit operations used
in waste water treatment, sedimentation, mechanical filtration
and centrif'i.gation, as well as flotation, are among those which
are best understood. L. G. Rich (3586) and Babbitt and Bauman
(112) have written textbooks relative to sewage sedimentation
with references to the original sources.
Recirculation
Recirculation is a common technique to control the efficiency
of biological trickling filters. Recirculation on standard
and high-rate trickling filters was shown by Hansen (1719)
to have certain advantages, but economic questions have been
raised. By the end of World War II, according to Hurley
(2121) , recirculation of the effluent from percolating fil-
ters was widely used in this country. Various schemes of
recirculation on single and two-stage trickling filters are
illustrated in Figures 4, 5, and 6. The flexible operation of
the recirculating system on trickling filters to obtain uniform
quality of effluent established its wide-spread use, as
described by Eliassen (1151). A system could then accommodate
the varying requirements of Army camps in the United States.
Fowler (1326) discussed the revival of the percolating filter
concept based on recirculation.
In the operation of a Biofiltration plant, Verna (4538) pointed
out the significance of the ratio of the recirculated effluent
to the settled sewage influent. Imhoff (2205) reported on
investigations which demonstrated the advantage of a constant
rate of flow of sewage, which could be kept by recirculating
part of the filter effluent. Recirculating trickling filters
were simpler to operate than enclosed aerated filters.
Similar comments were made by Escritt (1199) on the design
89
-------
RECIRCULATION
TAKE-OFF
LAUNDER
(A)
QR,(INCL.HUMUS SLUDGE)
(B)
1C)
Q+QR
(D)
(E)
(F)
Fig. 4 - Biofiltration Recirculation Diagrams
Typical Biofiltration Flowsheets. (A) and (B) are single stage;
(C) is two-stage with progressive feed; (D) and (E) are two-
stage, and (F) is single stage, intermediate. Sludge lines are
not shown except as noted in (B). In all flowsheets humus
sludge from the final clarifiers and intermediate clarifier is
returned to the influent line of the primary clarification tank.
[by Permission of Public Works Journal, "Handbook of Trickling
Filter Design" (5574, p. 19)]
9O
-------
R(+SL)
hpO1
SL*SLUDGE RETURN
PRIMARY TANK
FILTER
SECONDARY TANK
Fig. 5 - Flow Diagrams of Single Stage High-Rate Filtration
[by Permission of American Society of Civil Engineers
(4916, p. 157)1
Flow diagram (A) - the secondary sludge is returned to the
primary tank and, other than (G), all systems required separate
pumping arrangements for secondary sludge removal. Flow dia-
grams (A), (B), (C), (E), and (G) dampened fluctuations in the
organic loading applied to the filter. Flow diagram (D) re-
cycled filter sloughings to the filter. All flow diagrams,
except (D) and (F), required consideration of the recirculated
flow from which (E) was usually considered uneconomical. Flow
diagrams (F) and (G) were used for partial treatment. Flow
diagram (C') included the recirculation features of (A) and (B),
91
-------
»^ T *>
SL = SLUDGE RETURN
PRIMARY TANK
FILTERS
INTERMEDIATE TANK
SECONDARY OR
FINAL TANK
Fig. 6 - Plow Diagrams of Two-Stage High-Rate
Filtration [by Permission of American Society
of Civil Engineers (4916, p. 158)]
Flow diagrams (H), (I) , (J), and (M) have been used most
frequently. Flow diagrams (H), (L), and (M) could be
utilized as low-rate filters for the second-stage unit.
Flow diagrams (I), (J), and (K), in addition to eliminating
the intermediate settling tank, were attempts to improve
treatment by developing greater biological activity on the
second-stage filter. In flow diagram (J) a part of the
primary settling-tank effluent was by-passed directly to
the second-stage filter; in flow diagram (I) a part of the
settleable sloughings from the first filter effluent was
applied directly to the second filter and the recirculated
flow was not settled.
92
-------
of recirculating trickling filters in England. Recirculation
relative to the efficiency of -percolating filters was dis-
cussed by Lumb (2781) in a symposium on the treatment of
waste waters. Typical of recirculation ratios used by
Fischer (1285) in the biofiltration system were:
(a) 2:1, producing a reduction in BOD of from 50 to 60%.
(b) 1.5:1, producing a BOD reduction of 75 to 85%.
(c) 1:1, producing a BOD reduction of 90 to 95%.
Meltzer (2964) stated that a recirculation ratio of 1:1 will
give an increase in detention period of 19%, while 2:1 recir-
culation will increase the detention period by 30%. Archer
and Robinson (85) found that recirculation improved overall
efficiency of the plant and that a recirculation ratio of
4:1 can reduce the required volume of medium. Dreier (1029)
in 1947 maintained the flow rate to the filters at a 1:1
ratio with recirculation ratio at average flow.
Meltzer (2965) pointed out, however, that effluent recircula-
tion appeared to have no theoretical advantage over filtration
in deep filters from the efficiency standpoint. Imhoff (2185)
found that stronger sewages required either deeper filters
or recirculation of the effluent to achieve a desired removal.
It was noted by Rigbi et al. (3612) that, when the rate
of recirculation produced the effect of a filter 9 feet deep
or more, the size of the filter medium had no effect on the
efficiency. Escritt (1204) used recirculation instead of a
deeper filter, and suggested that the depth of the filter
multiplied by the recirculation ratio plus one should equal
the depth of a single filter.
Imhoff and Dahlem (2204) suggested that single filtration
operating at a low rate of flow is the simplest form of
filtration, but has disadvantages of odors and nuisance from
flies; however, with filtration at a high rate and recircula-
tion of the effluent, these were reduced. The obvious disad-
vantage of recirculation is the cost of pumping the sewage;
however, the filter area is less and the recirculation can be
controlled automatically. Recirculation is one of four ideas
to reduce blockage caused by excess amounts of organic matter,
as cited by Watson (4640) .
Typical of reports published on the operation of percolating
filters with and without recirculation were those by Yakovlev
and Galanin (4840) , Ziegler (4863) , Benzie and coworkers
(273), and Greeley (1564), and Greeley et al. (1566).
Additional papers by Hurley (2118) and others (4940) compared
recirculation with the alternating double filtration technique.
Recirculation has been used to increase the efficiency of
plants in England (2879), Israel (3612), Germany (3785), and
93
-------
India (3055). Recirculation on many industrial wastes has
been reported by Hatfield (1792)- and by Kirchoffer (2494)
to such an extent that they will be reviewed in the sections
of the specific industrial waste streams in Part IV.
Recirculation involves the collection of the waste water at
some point in the process and transporting it to another
point. Examples of different schemes of recirculation are
illustrated in Figures 4, 5, and 6. Recirculation of the
trickling filter effluent directly without any clarification
has often been used by Bachmann(133) in the Biofiltration
system. Direct recirculation of strong, fine-chemical plant
waste was part of a system reported (3551) to produce 80$
BOD removal. Typical of many of the patented systems was
that issued to Gillard (1474), in which the operation of
the Accelo-Filter depended on the direct recirculation of
unsettled filter effluent to the stream of settled sewage
being applied to the filter. With the development of the
package plant concept, as exemplified by Knapp et al. (2514) ,
recirculation played a vital role. At Suffern, New York,
recirculation of primary filter effluent was an integral
part of plant expansion in 1960 (5411) .
Bartlett (2O5) in 1954 stated that it was desirable to design
trickling filter plants with recirculation of filter and/or
humus tank effluent. Operational experiences reported as
far back as 1938 by Herrick (1897) suggested the use of re-
circulation of effluent from other unit operations, such as
Imhoff tanks, to stabilize the biological growth in per-
colating filters. A practice in the redesign of percolating
filter plants in England was to operate with recirculation
of the effluent of all units of the system up to and including
the trickling filters (3383). In the design of secondary
treatment plants, the practice was to provide recirculation
of the secondary sludge to the primary sedimentation tanks
(4905) .
Another scheme of recirculation used by Sickert (4O13) was
the two-stage operation of a trickling filter activated -
sludge plant in which one bay of the activated-sludge plant
was used for reactivation of recirculated sludge. Other
designs which used direct recirculation of percolating filter
effluent were reported by Theinpont (4352) and Borden (4O9).
Monson (3068) recirculated a milk trade and domestic waste
influent, and Bergman (277) described a recent plant expan-
sion using direct recirculation for three townships in Orange
Free State, South Africa. Ouellette (326O) at Elizabethtown,
Kentucky, and others (5O46) in France, and Greaves (5533)
and Calvert (593) in England used direct recirculation.
Herriot (1899) outlined some engineering points dealing
with recirculation and alternating double filtration including
94
-------
economics. Clarified effluent was frequently recirculated,
as was mentioned by Guerree (1614), Lugt (2777), Griffin
(1593), Fletcher and Anderson (1295), Weller (4673), and
Banister and Ellison (162).
Several design factors of recirculation in conjunction with
trickling filter operations were reported in the literature.
Climatic effects and loading factors were discussed by Benzie
and coworkers (273), where recirculation through the filter
had a marked cooling effect during the winter and lowered the
efficiency of summer operations by 21$. Smith (406O) reported
recirculation rates in terms of pounds of BOD/ac-f/day. A
single stage filter operated with recirculation of 1,000 to
2,000 pounds on an influent of 35O to 4OO pounds of BOD/ac-f/
day (8 to 9 Ib of BOD/1,OOO ft3/day).
Recirculation of sewage gave a washing effect, as noted by
Imhoff (2182), and had to be done at a velocity of sewage
sufficient to keep the filter from, plugging. Reimers et al.
(3551) experienced the tendency of ponding, which was mini-
mized by increasing the recirculation rate, supporting the
concept of Imhoff (2182) some 17 years earlier. Several
papers, e.g., Sharp (397O), Eliassen (1154), and Smith
(406O) , have been published specifying factors of design for
recirculation. Recirculation was used as an acclimatization
procedure by Oeming (3216) and Tedeschi and Lucas (433O),
who reported that stage operation at varying recirculation
rates resulted in a stable effluent.
Much of the data gathered and compiled by the Great Lakes-
Upper Mississippi River Boards of State Sanitary Engineers
was statistically correlated by Fairall (1233). The per-
formance of trickling filters with and without recirculation
was compared to that predicted from the Ten State curve. The
best correlation was obtained for first-stage filters with
recirculation. Caller and Gotaas (1395, 1397) considered
recirculation as one of five independent variables in their
data analysis. Increasing the recirculation ratio was
advantageous until the ratio reaches four volumes recirculated
liquid to one of the plant influent. The Biofiltration process
which incorporated a high degree of recirculation was reported
by Jenks (23O5) to have its capacity decreased once the ratio
of recirculation to rate of flow exceeded six. The practical
upper limit for domestic sewage was found to be 12O mgad
(2,76O gal./ft2/day). Fischer and Thompson (1279) and
Fischer (1284) described the effectiveness of the Biofiltra-
tion process of single-stage and two-stage at low recircu-
lation ratios and noted that the dosing rate on the filter
should not fall below 4 to 8 mgad (92 to 184 gal./ft2/day)
or should not exceed 10O to 125 mgad (2,3OO to 2,875
gal./ft2/day).
High recirculation processes require land area slightly greater
than activated sludge, but less than one-quarter that for a
trickling filter, according to Jenks (2305).
95
-------
Montgomery (3074), in a discussion of poor operation of
Aerofilters, stated that it could be due to inadequate
recirculation, among other things. Economic preference
was given to two-stage filtration over single-stage fil-
tration with parallel recirculation. Caller and Gotaas
(1397) later related the cost of recirculation rate, filter
size, and filter diameter.
Critique
It would appear conservative to say that the solids-liquid
design technology has been worked on to the extent that mean-
ingful criteria are established. Techniques such as poly-
electrolyte treatment to aid the separation are not covered.
However, for design purposes, chemical addition would only
change the detention time required for a given clarity or
effluent quality for the same detention time. The design
must still have a facility for collection, separation and
transport to the next process step, as was reported by most
workers. Until power costs become extremely reasonable, or
the price of land too high, centrifugation and other energy-
consuming systems may be reserved for specialized applications.
The literature on recirculation was typical of that encountered
in other areas of this review. Most of the designs for new
plants were reported after the fact, and each installation was
improved over the last ones, producing a confusing picture.
This confusion was noted by authorities, such as the American
Society of Civil Engineers, which only have examples of the
practiced art and not the correct designs, as shown in
Figures 5 and 6. Work has been done, as reviewed in the
theory section, but much of its practical application is
for future work. Generally, it can be said that recircula-
tion does help plant operation, but not necessarily effi-
ciency, and that an upper limit is four to six volumes of
recycle to one volume of influent. Those in the field, not
mathematically inclined, should be enlightened by some of
the reasoning used in recirculation design. Furthermore,
a set of operation criteria could be established based on
past experience and problems. These criteria, if properly
prepared, could be used by operators and designers to handle
specific waste treatment problems. With knowledge of pre-
vious experience available, some of the unsupported design
and poor operation may be lessened.
EFFECT OF PRETREATMENT
"Pretreatment" describes those unit operations beginning at
the influent pipe of the plant up to and including primary
sedimentation. Economic factors have caused investigators
and operators to study the effect of these pretreatment units
on subsequent treatment processes.
96
-------
Reid (3548), a Scottish investigator; has outlined designs
for typical sewage works for small villages, such as per-
colating filter plants built without screen or detritus
channels, to lower the cost of maintenance labor for isolated
installations which have little chance for inspection. Rumpf
(3782) reviewed the work of Tomlinson and Hall (4425) on
the operation of percolating filters with and without primary
sedimentation during the development of the alternating fil-
tration system. From some of Hall's work (1653), it was
concluded .that the degree of purification in the trickling
filter does not bear a direct linear relation to the strength
of the tank effluent applied to it. Hall did note that im-
proved sedimentation was capable of giving higher filter
loadings by 15 to
A patent was issued to Jung (2354) for a process in which raw
sewage .without sedimentation was applied directly to a per-
colating filter. Osborn (3251) found that the absence of
primary humus tanks resulted in an increase in the suspended
solids in the secondary filter effluent, but did not reduce
the quality of the final effluent during double filtration
operations. These results were in fair agreement with those
of the Water Pollution Research Board (5193), which noted
the presence of the clarifier was not proven essential. The
Water Pollution Research Board was primarily interested in
the effect of secondary treatment with little or no pre-
treatment based on two developments (5193) . These develop-
ments were: " first, the commercial introduction of small
maceraters, and secondly the development of plastic filter
media which were not so readily blocked as conventional
media."
Most of the designs were to condition wastes for subsequent
treatment, as Lawton (2648) reported. Where the suspended
solids were reduced to less than 1OO mg/1, as would be the
case with chemical precipitation, then a fine filtering
medium 0.125 to 0.375-inch diameter was preferable.
The detention period in primary sedimentation in the early
1900 's was recommended by Ogden to be about two hours (322O) .
A common pretreatment technique in the early 19OO's was the
use of a septic tank dosing chamber/ in which an average
detention period of 4.5 hours removed 72$ of the suspended
organic matter, according to Doten (1OO5) . Prior to 194O,
pre- and final sedimentation of 1.5 to 2 hours was specified
by Imhoff (2185) to be necessary for proper waste treatment
operation. Rank in and coworkers (35O1) described in con-
siderable detail the type of treatment required before bio-
logical filtration.
Typical designs of the 193O's cited by Hughes (2084) in-
cluded coarse screens, detritus tanks, fine screens, and
continuous flow sedimentation tanks. A plant flow sheet in
1931 (2O69) included a screening process, an automatically
97
-------
cleaned screen system, a grit scrubber, and mechanically
cleaned settling tanks, which was not too different from
the type of plant described in 1956 by Stuewer (4261) .
In addition to the preliminary facilities noted, chlorina-
tion was used (4673) in handling septic waste. Another
pretreatment approach described by Barron and Lawrence (199)
was raw sewage chlorination for odor and nuisance control.
The application of chemicals to aid sedimentation at a
recent sewage plant development was noted by Greeley (1557) .
Preflocculation without chemicals to aid solids capture was
discussed by Fischer and Hillman (1282).
Preaeration to raise the dissolved oxygen in a septic or
very strong waste was reported by Collom (763), Holland
(1979), Vogler (4545), Anderson (77), and Thomson (4380).
Dreir (1O29) noted that preaeration, either mechanical or
by diffused air, for 3O to 40 minutes, at a maximum hourly
flow, satisfactorily removed grease and increased primary
sedimentation efficiency as measured by primary BOD removal.
He also suggested that by increasing the primary sedimenta-
tion efficiency operational advantages could be gained.
Higher dissolved oxygen in the waste, toleration of a much
higher organic load, and,withstanding the introduction of
the supernatant into the flow were among the benefits of
this technique. Nauta (3150) stated that with continuous
short aeration periods 7O$ purification could be obtained,
and the aerated sewage could be treated on percolating fil-
ters at increased rates. Ingram (2224) agreed that with
suitable preliminary treatment of the waste water, which
involved pH manipulation and suspended solids removal, a
higher rate of controlled filtration was feasible.
Pretreatment in industrial applications was common in proc-
essing potato waste waters, such as screening, disintegra-
tion, sedimentation, and flotation (5330). Vogler et al.
Hendee (4545) stated that strong pharmaceutical and organic
chemical wastes can be treated by biological filtration but
preliminary treatment, similar to that outlined by Reimers et
al. (3551), is often necessary. Genetelli and coworkers
(1452) indicated that minimal (7O$ BOD removal) preliminary
treatment was necessary before discharging pharmaceutical
waste into a domestic sewer system. Pretreatment of this
type would require a more sophisticated plant than was sug-
gested by Snyder (4O94) which consisted of an equalization
pond and a final sedimentation tank to handle a variable
flow textile waste. Other industrial wastes, such as milk
processing waste as recorded by Kirchoffer (2494) and Salvato
and Bogedain (3814), required preliminary treatment to handle
high organic and suspended solids loads. For turnpike sewage
treatment plants, Dixon and Kaufman (977) recommended anaero-
bic treatment in advance of biological filtration to solubi-
lize greases and colloidal suspensions from detergent emulsi-
fication. Antipina (83) recommended that the waste waters
98
-------
from ethylene oxide production should be pretreated by a
4:1 dilution with domestic sewage. The dilution of indus-
trial sewage with domestic sewage is common practice.
Critique
Pretreatment of the influent by mechanical and chemical means
has improved the efficiency/ operation and maintenance of
trickling filters. Studies were also made on the partially
successful operation of trickling filters with no pretreat-
ment, other than a bar rack. However, to deviate from es-
tablished practice, special design must be considered or new
operational problems must be expected. An example would be
a trickling filter plant designed without a primary clarifier.
If the medium was coarse enough and the distributor and
underdrains sized properly, this system should perform ade-
quately. On the other hand, if a low-rate plant is operated
without the primary sedimentation tank, there should be no
surprise when operational problems arise.
Chemicals for pretreatment have been applied for many years
and will probably continue for many more. Nuisance control,
solids capture, pH adjustment, and nutrient supplements are
just a few of the methods of pretreatment of sewage.
Additional information of unit operations and their relative
performance with pretreatment would be of value using newly
developed instrumentation. Use of chlorine for pretreatment
is covered more deeply under maintenance of trickling filters.
EFFECT OF POST-TREATMENT
The necessity for post-treatment of effluent from biological
filtration was recognized as necessary very early and became
commonplace. Lawton (2648) observed that to obtain a good
effluent it is essential to remove suspended solids which are
carried from the filter. In the early 19OO's, Ogden (322O)
stated that a two-hour retention period was used for humus
tanks (clarifiers) to retain the particles in suspension,
thereby clarifying the effluent. Information dealing with
the design, construction, and operation has been published in
a series of papers in the Public Works Magazine (3501, 5378,
5574), which illustrate the importance of the post-treatment
of filter effluent. Typical of many design papers was that
of Borden (4O9), which indicated that the biological filtra-
tion process is integrally related to the final sedimentation
process. Hottelet (2048) stressed the importance of secondary
sedimentation when using biological filtration.
99
-------
During the development and testing of the Biofiltration
process, Fischer (1286) and Fischer and Thompson (1279)
stressed the significance of post-treatment clarification
with careful monitoring of suspended solids. Installations
at Sullivan County, New York, were reported by Borden and
Rodie (41O), Rochdale, Lancaster, England (5218), Hayton-
with-Roby in England by Pilkington (3383) and on the New
York Thruway by Sander (3818) to be typical in using post-
treatment procedures on percolating filter effluents. Data
gathered by Rigbi and coworkers (3612) under field condi-
tions indicated the difference in efficiency with and with-
out post-treatment sedimentation and secondary sedimentation
was favored due to the 1O$ increase in BOD removed. The
significance of final removal of suspended solids and its
effect on the treatment of trade waste waters were discussed
by Escritt (1205) . Reimers et al. (3551) noted that sec-
ondary or final clarification was necessary for the discharg-
ing effluent and not for the recirculated portion of a
pharmaceutical waste.
Sand filters, described by Streander (4244) as being com-
posed of 28 inches of sand and 4 inches of gravel, were
used for post-treatment for many years. Microstrainers
as a tertiary treatment were studied by Fish (129O), Keefer
(24O6), Truesdale et al. (4468), Roberts and Lawson (3637)
and Colthrop Board and Paper Mills (4963), and incorporated
for successful solids removal. Disinfection with chlorine
was stated by Newton (3179), Brown (49O), Lafontaine and
Armstrong (2607) and Rofman (366O) as a common practice of
post-treatment.
Critique
Pretreatment and post-treatment of biological filtration
were reviewed to provide the identification of distinct
areas apart from the trickling filter which have very
strong bearing on its performance. It is well agreed that
solids separation after filtration is important. However,
the investment for secondary clarifiers may not be justified,
If only 1O$ increase in BOD removal is obtained by an ex-
pensive appurtenance such as a sludge scraping settling
tank, then, perhaps another method should be sought or the
money not spent. The weakness here is using the BOD test to
measure effluent quality and an implied BOD removal due to
solids capture. The BOD test was designed for a specific
purpose, to assess the stability of an effluent in a
receiving stream, and not to measure the effectiveness of
secondary clarifiers. The polluting powers of the
suspended materials would have to be evaluated in a manner
other than the dilution BOD test.
1OO
-------
Disinfection of the final effluent before release from the
sewage plant is a common practice and would justify a
separate review. Microscreens for filtering solids have
been used, but screen plugging on secondary sludge is to
be expected.
SOLIDS DISPOSAL
A function of the biological trickling filter plant is the
coagulation of colloidal particles and removal of soluble
organics which generate a sludge disposal problem. Various
designs have been proposed by Dreier (1O28) and Eckenfelder
and O'Connor (1O80) for sludge disposal. A comprehensive
review of sludge handling techniques has been prepared by
Burd (533), and detailed discussions of these methods are
beyond the scope of this review. Selected techniques are
briefly mentioned to demonstrate an awareness of the sludge
disposal problem.
Imhoff (2197) stated, in 1941, that fresh, fully uigested and
dried sludges are odorless. Very offensive odors are produced
in the intermediate stages; therefore, nothing but digested
sludge should reach the drying beds. If it is necessary to
dry fresh sludge, more rapid methods of dewatering were recom-
mended. Geographical location was stressed by Furman (1384)
as significant for sludge drying, i.e., southern climates
require less area. Sludge drying beds were used and the
dried sludge stored or delivered to farmers, according to
Burger (535). Occasionally, unique geographical situations
were used, as noted by Collom (763) in Auckland, New Zealand,
where a volcano crater was used as a sludge lagoon follow-
ing anaerobic digestion.
Easdale (1064) described a technique which was a forerunner
of the Imhoff tank, except that decomposed solids were in-
tentionally allowed to pass out with the effluent to primary
contact beds or percolating filters. The importance of
sludge recirculation was often noted by investigators such
as Fowler (1326) in 1938 and Condon (766) in 1966. Wide-
spread use of Imhoff tanks in conjunction with trickling
filters to handle sludge digestion was described by various
investigators, e.g., Burger (535), Kuhn (2586), and Richey
(3592) .
A very popular method of reducing sludge volume and putres-
cibility during the 1930's, '4O's, and part of the '5O's wa
sludge digestion under aerobic conditions, as described by
was
101
-------
Allen (35) . Greaves (5533) reported sludge from primary
sedimentation tanks was concentrated, digested, and dried
on beds, and the sludge from the final sedimentation tanks
was recirculated to the head of the plant. Heated and un-
heated sludge digestion was cited by Mclntosh (2914),
Storrie (4237), Phillips (3369), Ellsworth (1162), Beaton
(1846), and Veatch (4524), with the digested sludge being
discharged to open or glass-covered drying beds.
Sludge presses were used (1846) to dry the sludge more
rapidly. Sludge which was treated with lime and pressed
was commonly used as fertilizer by farmers in Great Britain.
Watson (4712), as a proponent for sludge disposal on land,
emphasized that sludge from sewage works preserved most of
the mineral constituents of sewage in a form more readily
available as fertilizer than when the sewage was applied
directly to the land. Primary sludge has often been
deposited in lagoons, while secondary sludge was spread over
land (5464) . Reimers et al. said that sludge from a pharma-
ceutical waste treatment plant could be satisfactorily
spread on land for disposal (3551) . A common technique used
today is chemical dewatering on vacuum filters and incinera-
tion of the dried sludge (766).
Critique
Other than to note that the investigators and workers in the
field of trickling filter design were cognizant of the prob-
lems, discussion of sludge disposal will be minimized.
Efforts were made wherever possible to eliminate the sludge
from the receiving body of water. The methods of sludge
disposal were so varied that it took a sizable volume to
contain a recent review of them. Simply stated, it was
noted by Sir Lawrence Chadwick of the British Royal Sewage
Commission, prior to the turn of the century, that the
disposal of solids to the land and sewage to the river should
provide the solution for waste disposal. However, he was of
the opinion that "the solution to pollution is dilution."
This opinion has since been proven slightly shortsighted, as
shown by the rise of secondary waste treatment facilities,
such as percolating filters. However, his comments dealing
with the disposal of solids on land still have current prac-
tical and economic significance.
EFFLUENT DISPOSAL
In complete waste water treatment, the job is not finished
until the effluent from the sewage treatment plant has been
released to the receiving body of water and is harmless to
the environment. The state and condition of the effluent
are quite important to its receiving body of water, as ex-
pressed by many investigators. Concern was expressed by
102
-------
Shaw (3975) over the effluent standards that have been leg-
islated and observed early in the literature by many inves-
tigators, such as Donaldson et al. (995) and Rohde (3666).
Puller (1369) cited six types of conditions of stream flow
which should dictate the type of treatment required. He
strongly emphasized that a uniform degree of purification
in all situations was not practical. According to a survey
by Thompson (4371) covering 25 years, the developments in
methods of sewage treatment since 1910 can be attributed to
the work of the Royal Commission on Sewage Disposal, and
later of others, such as Mulvany (3121). These investigations
evaluated efficient methods of sewage disposal to reduce the
pollution of streams, as well as set up monitoring procedures
for various pollution control agencies. A classic example
of this type of effort was the investigations by Biery (325)
on the Ohio River system which were carried on for some 25
years to develop local, regional, and federal legislation
to establish effluent disposal requirements. Typical of
state activities was that described by Poole (3424), in which
the Indiana Stream Pollution Control Board developed criteria
which would avoid gross pollution of streams during periods
of low flow.
During the design of a waste treatment facility and before
selecting a specific process, engineers, such as McKinney
(2925) and Smith and Leibee (4O75), have often stressed the
importance of evaluating the characteristics of a waste
water and the effluent standards in the local area, and the
need for preliminary treatment. It was pointed out by Phelps
(336O) that early investigations to demonstrate the effective-
ness of waste treatment were done by the Pasteur Institute
to establish effluent disposal which would reduce gross
pollution of the streams in northern France. Weber stated
(4654) that the effluent quality of the sewage treatment plant
was established after an examination of the receiving stream,
the Enz in Germany, which dictated the degree of purification
(8O-85$ BOD removal) the plant must achieve at Pforzheim.
Typical procedures involving effluent disposal design for
the process of upgrading waste water treatment have been de-
scribed. Irwin and French (2235) stated that primary treat-
ment alone resulted in producing an effluent which affected
the receiving stream by the development of a heavy growth of
fungi and a reduction in the dissolved oxygen content as well
as an increase in the BOD of the river water. Ammon (69)
determined the initial condition of streams receiving milk
processing waste waters and the corrective treatment was spec-
ified. Mills and Wheatland (3O17) discussed the treatment of
high salt concentrations which make the effluent unsuitable
for irrigation or for a public water supply. Boone (4O3),
Tiedeman (4392), and Rowe (37OO) modified the typical over-
loaded plant to chlorinate the effluent, which was finally
discharged to a small creek to prevent disease or nuisance
problems.
103
-------
Effluent quality becomes very important in areas of scarce
potable water, Which has developed a growing demand for ef-
fluent reuse. In 1931, German investigators (5614) noted
that biological treatment is necessary when a stream affords
less than a 50:1 to 1OO:1 dilution of the waste and is used
for bathing or washing beyond the point of discharge. Ef-
fluent must be of high quality when it is discharged to a
body of water that is used as a potable water supply, as
reported by Majewski (2842), Saunders (3831), and McPhee
and Smith (2941). In Pennsylvania, the city of Bethlehem
was very much interested in the treatment which Allentown
provided to their sewage because it was discharged to the
Lehigh River, and four miles downstream it was the source
of Bethlehem's water supply, according to Krum (2575). Very
high degrees of treatment have been accomplished over the
years for municipal waste (Yonkers and Fairfax) as well as
industrial waste in areas where the water was to be used
as a potable water supply (3831, 5138). It was pointed
out by Shastin (3973) that effluents of a rosin plant in
Russia have been purified to a high degree (90-95$) and were
discharged, without problems, to a small river. Ground water
recharge was practiced on Long Island as cited by Baffa and
Bartillucci (431). Another use of the effluent, depending on
its quality, was studied by Minett et al. (3O27), who supplied
this effluent to cattle. The symptoms shown by cattle in
cases of alleged sewage poisoning were attributed to other
causes.
The very old procedure of irrigation as a final step in
wastewater treatment was described by Fidler (126O). The
rates of application were around 8,OOO gal./acre/day (O.184
gal./ft2/day) on a 495-acre farm, while even a low-rate
standard design trickling filter was designed to operate at
3,6OO,OOO gal./acre/day (82.8 gal./ft2/day), according to
Galligan (1399). In studies in Germany by Brauss (448) on the
use of sewage effluent as a water supply, it was noted that
the bacterial numbers introduced into a stream by filter ef-
fluent are about the same as those introduced from sewage ir-
rigation fields. Popkin and Bendixen (3430) studied the re-
lationship of proper irrigation rates for various soils. Girard
(1483) described the disposal of the final effluent from septic
tanks by irrigation in some rural or isolated areas, and Guiver
and Hardy (162O) stated that the final treatment of sewage
should be irrigation over grassland. Even though fertilizing
benefits have been cited by Machenko (28O6) for sewage ef-
fluent irrigation of potatoes, objections to disease transfer
have been made by Mundel et al. (3122) . Roberts and Jones
(3639) reported that, when complaints were received on dis-
charging sewage plant effluent into public areas, a community
had purchased farm land for disposal by irrigation, and the
land was leased for farming.
104
-------
Treated effluents with the required quality have been reused
by industry, e.g., 1.5 million gallons a day sold to an
English steel works, reported by Barker (178); effluent from
Noose Jaw, Saskatchewan, used by a Canadian oil industry for
cooling purposed, described by Haack (1636); and the Corpus
Christi, Texas, effluent (98$ removal) sold to a refinery
company, as stated by Allison (57) . Other uses for effluent
of the proper quality were published: Dye (1O55) noted the
city of Tucson received more than $7O,OOO a year from the
sale of effluent for irrigation and sludge as a soil condi-
tioner; for hydraulic power generation (5137); boiler feed
water (5088); or cooling water for a power station, (4988).
Another possibility for disposal of waste water effluents
from paper mills was reported by Southgate (4118) . He
concluded that to reduce pollution the various influent
streams can be segregated, partially treated, and the efflu-
ent from these recirculated back into the industrial complex
for use as washing water, in beaters, and for diluting stock.
Critique
Effluent disposal is one of the last steps in the operation
of biological trickling filters. Systems to make economical
use of the treated water have been designed. As the cost of
pollution control goes up, more use will be made of recycle
of effluents.
The biological trickling filter is only part of a total
system to condition effluent for disppsal. However, complex
waste treatment problems have been solved simply and economi-
cally through use of the trickling filter, and in all proba-
bility its use will be continued. Limitations in the quality
of effluent from trickling filters will require their use in
combination with other systems. Designers should be encour-
aged to direct effluent disposal to other than water courses,
especially if the treated .effluent has valuable characteris-
tics, e.g., an effluent high in phosphorus and nitrates
being used for land irrigation.
Only recently has the literature indicated that control of
effluent quality can be economically beneficial. More
experience illustrating the economic advantages of effluent
use and elimination of stream pollution should be reported,
along with efforts to educate the public to realize the value
of effluent reuse. The new more rigid stream standards will
be a motivating force to develop better sewage treatment
plants, in which the trickling filter may serve a useful
function.
VENTILATION AND UNDERDRAINS FOR TRICKLING FILTERS
The design of the underdrain and ventilation system for the
trickling filter has been the subject of many investigations
105
-------
and a large number of papers have been published on this
topic. Much of this information resulted in the development
of design guidelines, such as "Guides for Sewage Works
Design" (5331). The Ten States Standards (5011) specify that,
"the underdrain system, effluent channels, and effluent pipes
should be designed to permit passage of air. The size of the
drains, channels, and pipes should be such that not more than
5O percent of their cross-sectional area will be submerged
under the designed hydraulic loading. Consideration should
be given in the design of the effluent channels to the possi-
bility of increased hydraulic loadings." The underdrain
system was so interesting technically and economically that
the Trickling Filter Floor Institute was formed to provide
design information to assure proper performance of these
units (5574) .
Several investigators, such as Escritt (1191, 12O1), published
information on the specifications of underdrain systems and
their construction. Lawton (2648) indicated that underdrain
systems were well established in the early 190O's. Imhoff
(2189, 2211) provided examples and design information on the
size of a false bottom and to obtain sufficient aeration.
Baltimore (5137) had a novel underdrain system, and the 18-
foot drop of the effluent to the river generated power. As
higher rate trickling filters were developed and the Bio-
filtration system was popularized. Nelson (3159), in 1943,
stressed that ordinary percolating filters could not usually
be adapted for Biofiltration, since the underdrain systems
could not handle the increased flow of sewage.
Early work developed criteria on ventilation and the effect
of porous walls, but Watson (463O) concluded that there was
no benefit from open sides if sufficient air was circulating
freely under the medium. In disagreement with Watson,
Voorhies (4556) stated that the most economical trickling
filter had open side walls to improve aeration. Furphy (1385)
recommended that the air circulation should be sufficient to
supply enough oxygen and carry off the carbon dioxide fast
enough to prevent its concentration reaching more than O.2$.
Halvorson et al. (1665) disagreed with Weidlich (4662) on the
function of air flow through a filter being dependent on the
cooling or heating of the air by the liquid, change in density
of the air due to change in humidity, and heating due to
microbial activity. For dosages up to 2O million gallons the
rate of flow of air is independent of the dosage and is a
straight-line function of the difference between the temper-
ature of the air and the sewage (1665). The aerodynamics
of trickling filters was explained in considerable detail
by Piret et al. (3389), who gave reasons for, and rates of,
air movement under varying conditions. They, along with
Benzie et al. (273), observed that in summer the sewage is
colder than the air and the air moves down through the filter,
106
-------
whereas in winter the sewage is warmer and the air moves
upwards; when the temperatures are about equal, very little
air movement occurs. Studies in seventeen sewage plants
in Michigan (273) showed that recirculation of effluent
through a well ventilated filter in winter had a marked
cooling effect and this reduced the efficiency of the filters
by about 21$. The actual oxygen consumption was measured by
Pbnninger (3413) to be 6.5 to 7.5 grams per day per liter of
biological film. Dahlem (853) stated that natural ventilation
is sufficient if the filter is not too high, and these views
were supported by Imhoff (2185). Petru (3342) supported the
findings of Benzie et al. (273), and spoke of the cooling
taking place in the effluent from the trickling filters, but
thought it was caused not so much by external temperature as
by the ratio of the filter diameter to the depth. Using Kr85
(as a tracer gas), Mitchell and Eden (3O3O) investigated
ventilation rates in trickling filters and found that the
controlling factors were the difference between ambient
temperature and the temperature of the filter, the direction
and velocity of the wind, and the elevation of the filter
above ground level. Other investigators such as Chase (658)
and Kolobanov (2543) emphasized the importance of the volume
of air feed and system designed for adequate ventilation
throughout the trickling filter.
When the temperature of the atmosphere and of the filter are
the same, and when the strength of the waste is high, deep
trickling filters with forced ventilation have been used to
provide aeration by Caller and Gotaas (1397), Schreiber (3898),
Halvorson (1667), and Imhoff (2182). In the following section,
a detailed review of the literature is made on forced and
natural ventilation for enclosed trickling filters. The
quantity of air flow required under forced ventilation was
studied by Lanz (2629) and Pohninger (3411). No advantage
was shown by increasing the volume of air to more than 5O
times the volume of sewage . Natural aeration (3411)
could not supply the oxygen necessary for a unit treating
a waste influent heavily loaded with concentrated sewage.
Comparative studies were made of percolating filters under
natural and forced ventilation conditions with and without
effluent recirculation by Ziegler in Germany (4863). Deep-
bed trickling filters arranged in layered section were
reported by Ingram for controlled aeration (2222) and Schulz
for forced aeration (391O). Controlled amounts of air were
introduced (2222) to each section of the filter and provided
the degree of treatment required. Schulz (391O) recommended
a ratio of the filter diameter to bed depth of 1:6 with
forced ventilation, while the data obtained by Ingram sug-
gested at least a nine-foot depth should be provided for
stabilization purposes. The use of forced ventilation by
107
-------
Brown (492) adequately solved odor problems and increased
efficiency, which was also verified by Galler and Gotaas
(1397). Deep filters with forced ventilation were more
economical than shallow filters, but only for high reductions
in BOD, and they lost this advantage if pumping was required
(1397) . Many patents have been granted in the area of
forced ventilation, e.g., Schreiber (3893, 3898), and Girard
(1482). A brief review of artificially aerated trickling
filters was published by Heilman (I860).
ENCLOSURES ON TRICKLING FILTERS
Shortly after the turn of the century, the Chicago Sewage
Disposal Experiment Station (53O8) found that trickling
filters did not need a roof. Work at Pennsylvania State
College over a fourteen year period (4578) clearly demon-
strated that closed trickling filters were superior to open
ones in efficiency, better oxidation, less odors, freedom
from flies, algae, snow and ice, and from nozzle freezing.
Ormsby (3248) showed that a closed filter gave greater
bacterial reduction and an effluent with lower oxygen demand.
Damm and Bock (859) reported in 1935 that closed trickling
filters must have compressed air from above to treat dairy
wastes; open filters have fly and odor nuisance. Enclosed
filters were examined in South Africa by Hamlin (1690)
using a Pruss-type enclosure, operating under the same con-
ditions as an open filter.
In 1937, Bottcher investigated several methods to reduce the
odor nuisance from sewage treatment processes in Berlin (418).
He concluded that the least effective of the artificial bio-
logical processes, from a point of view of odor control, was
the open trickling filter. If odor control was essential
a plant with a closed settling tank followed by a closed
trickling filter with sludge digestion and sludge drying in
an enclosed centrifuge was considered the most desirable.
Performance Differences - Open and Enclosed Filters
Closed and open filters were compared in South Africa by
Hamlin (1691) with only a slight difference in performance,
contrary to the results obtained by Walker in the fourteen
year study at Pennsylvania State College (4578). Hurley and
Windridge (2112) described six months' experimental operation
of two enclosed aerated filters. The purification obtained
included BOD reduction of 95$ and a significant degree of
nitrification. For comparison, open filters treating the
same sewage at half the hydraulic load produced a better
effluent. However, if the open filters were required to
mature at the same rate as the enclosed filters, the enclosed
filters produced a better quality effluent. Mechanical parts
might corrode more in an enclosed filter (4123).
108
-------
Hamlin and Wilson (1693) stated that summer temperatures and
dilution with storm water had considerably less effect on the
enclosed filter than on the open one. In the discussion of
this paper (1693), Scouller remarked that the percent of
surface area left open was significant for operational
efficiency.
Hurley (2110) found that operation of the enclosed filters
without forced aeration lowered the effluent quality. Dahlem
(853) pointed out in 1938 that complete enclosure and arti-
ficial aeration were not essential for the operation of
trickling filters at high rates of flow. He gave importance
to the size of the filter medium, the depth of filter suitable
for natural aeration, free access of air to the bottom of the
filter, good distribution of influent, and sufficient flow for
rinsing action. Enclosing only the sides of the filter pre-
vents the escape of flies and ensures good distribution of
liquid over the surface. Complete enclosure is not necessary
to control odor and overcome problems associated with cold
weather. For the same population, an open filter two meters
(78 in.) high and an enclosed filter four meters (156 in.)
high would cost about the same to construct. Operational
costs of the open filter are lower and the space required is
greater.
In 1938, Blunk (373) was concerned that chemicals/such as
chlorine for controlling nuisances on the trickling filter,
were upsetting to the ecological balance. He thought that
the fly larvae and other insects were part of the required
fauna on the trickling filter and should not be destroyed.
Loss of sunlight from enclosing the trickling filter is no
cause for concern, since sunlight might lead to the develop-
ment of organisms that are harmful to the biological processes,
With the exception of the few light-requiring organisms, the
ecology of the open and closed filters should be the same.
Hurley stated (2111) that enclosed filters appeared to mature
more rapidly than open filters. It was expressed that there
is the possibility of severe corrosion resulting from the
enclosure of mechanical plant parts.
In 1939, comparative tests of open and enclosed filters were
conducted in Germany (5323). Wilson and Hamlin in South
Africa (4748) pointed out that enclosed ventilated filters
and an open filter at the Klipsperuit Works in Johannesburg,
South Africa, could deal satisfactorily with nearly equal
loads. However, in cold weather the enclosed filter could
treat nearly four times as much settled sewage per cubic
yard of medium per day as the open filter to give the same
percentage purification. Husmann (2145) gave the various
physical, chemical, and biological processes in conventional
and high-rate filters. Enclosing the sides of a trickling
filter allows ventilation to occur only vertically and
accumulation of carbon dioxide in concentrations harmful
109
-------
to bacteria may occur. In completely enclosed filters, this
carbon dioxide problem appeared to be serious enough to
necessitate artificial aeration.
The main features of open and enclosed percolating filters
used in South America were described by Barbeito and coworkers
(189) and the advantages of the enclosed type discussed. Deep,
enclosed, artificially ventilated filter beds in South Africa
were compared with conventional open filter beds by Dekema
and Murray (915) . Over a three-year test period, the quality
of the final effluent from the enclosed filter was slightly
better than that of the effluent from the open filter. How-
ever, the enclosed filter was treating slightly more than
two times the amount of sewage per cubic yard of medium as
the open filter to about the same degree of purity. The
enclosed filter was less affected by variations in atmospheric
temperature and in strength of applied sewage than was the
open filter. The operation of uncovered filters in North
Dakota in cold weather was feasible, as reported by Stewart
(4216) . In 1948, Arnold (92) reported a comparison of the
performance of two 6-foot filters at the sewage works of
Grafton, North Dakota, which were identical, except that
one was open and the other enclosed. Although temperature
has some effect, the efficiency of open filters, as measured
by the BOD5 test, is slightly inferior to that of enclosed
filters, even under severe winter conditions, and open filters
are more effective during warm weather.
Hunter (2O99) described an experiment on filtration of settled
sewage through an enclosed percolating filter at Glasgow,
Scotland. After the experiment had been in operation for two
months, the filter was treating a hydraulic load equivalent
of 388 gallons per cubic yard per day with purification equal
to that obtained with open filters. However, ponding did
occur in this enclosed filter in winter. In a discussion of
the work by Hunter and Cockburn (21OO), Lovett (21O1) stated
that the fungal growth in the enclosed filter could be due to
the large volume of sewage applied per unit area. A paper in
195O by Cameron and Jamieson (6O4) on the operation of the
enclosed filter at Glasgow which developed fungal growth
proved that good quality effluent could be produced at load-
ing rates of 459 gallons per cubic yard. Aeration of the
filter prevented the egress of flies and controlled slight
odors. The superior performance of the enclosed filter was
attributed to the grading of medium and the retention of
heat.
Imhoff (22O9) discussed the operational results of high-rate
percolating filters in various countries, both open with
natural aeration and closed with artificial aeration. Leibee
(2671) pointed out the following advantages of enclosing
percolating filters: (a) they will function better during
110
-------
cold weather, (b) ventilation can be improved, (c) odor and
flies can be controlled, and (d) the contact of insects with
the flora of the filter can be prevented. In well-designed
open and covered filters in Minnesota, natural draft due to
wind and differences in air density were observed to be
sufficient to provide the air requirements, according to
Johnson (2329) in 1952.
It was concluded from work done in South Africa (5358) that
open filters with larger media showed less ponding and
quicker recovery after reduction of load than did enclosed
filters with smaller media. Nellist (3158) cited in 1965
the plugging of the filter medium of enclosed percolating
filters treating waste waters from a coke works. Waste
from other coke works has been treated by mixing 2% of the
waste water with sewage prior to treatment on the percolating
filters.
Tedeschi and Lucas in 1955 (433O) compared enclosed percolating
filters with normal open filters, and noted from examination
of the flora and fauna of the enclosed filter that numerous
worms and larvae were present, algal growths were completely
absent, and a large population of flies inhabited the air
space under the cover. Ponninger described (3422), in 1957,
modern developments and the design of enclosed tower trick-
ling filters. In 1956, during parallel operation of uncovered
and covered filters in Midland, Michigan, the covered filters
were not less efficient than the uncovered (5645). Since
maintenance problems of freezing and a considerable dis-
charge of fog over an adjacent highway were experienced
with the uncovered filters, the remaining filters were
covered. Water, containing sulfides, was reported by Stone
(4227) in 1962 to be made potable by means of biological
filtration through a covered, coarse rock filter, followed
by chlorination.
Critique
The above section may be summarized by noting that, after many
comparisons of the operating data and results of open and en-
closed percolating filters, no equivocal statement can be
made. Covering of trickling filters has been shown to be
an advantage under conditions of high organic loading, septic
sewage influent to the plant, extremely low or extremely high
atmospheric temperatures, location of the plant in a resi-
dential area and easier maintenance. Disadvantages of cover-
ing the percolating filters are the increased cost, specific
wastes requiring artificial ventilation, difficulty in making
large scale modifications to the distribution device or trick-
ling filter and insulation from the summer heat which is re-
ported to be helpful for trickling filter operations.
Ill
-------
Temperature Factors of Enclosed Filters
Ingram (2226), Wilson and Hamlin (4745), and others (5637)
quoted advantages of enclosure of the trickling filter for
maintaining suitable temperatures. Wilson and Hamlin (4745)
stated that the upper surface of the filter must be kept
warm in winter and that the enclosure conserves this heat.
Ponninger (3407) remarked that so long as the filter was kept
free from sludge the results obtained were as good in cold
as warm weather and that heating the air blown in was un-
necessary. The heat created by the process of oxidation in
the filter raised the temperature of the sewage by 1-2°C.
However, to raise the temperature in winter, the upper sludge-
collecting part of the filter should have air blown in from
below, countercurrent to the sewage. Reddie and Griffin
(3523) agreed that maintaining a higher temperature and
aeration aided the ecology of the filter. Coste (795) ob-
served the enclosure overcomes the cooling effect due to
evaporation caused by free access of air through the filter.
However, high summer temperature cannot materially assist
the enclosed filter.
Wilson and Hamlin (4748) reported experiments showing 9O° to
95°F. temperature to be optimum in obtaining good nitrifica-
tion and effluent quality. Arnold (92) agreed with the
observations of European investigators of temperature effect
on open and enclosed filters. The open filter had slightly
better removal during warmer months, but during winter months
the open filter was inferior. The influence of temperature
was discussed by Christ (683), who used an enclosed filter to
treat paraffin oxidation plant waste. Bayley and Downing
(226) studied the heat balance in percolating filters and
concluded that the most important factor is the temperature
of the applied sewage. The exchange of heat from the sewage
with air, especially near the surface, and the production of
heat by biological oxidation are also significant. Enclosed
filters reduce the risk of freezing; however, the average
increase in temperature would not be more than 3-4°C.
Critique
The temperature relationship of closed vs. open trickling
filters has been explored, but more quantitative information
would be desirable. Extrapolation of data from the various
treatment of wastes in many different locations has been the
basis of design for many years. Protection of the filter sur-
face from the wind and cold helps keep the filter surface
warm. The fast moving waste water being loaded onto the media
112
-------
has sufficient energy to keep from freezing, provided the
added heat sink of wind velocity is controlled. The waste
water temperature controls performance, not the temperature
of the air.
Ventilation Factors of Enclosed Filters
Ponninger (3417) studied field operations and concluded that
the use of aeration of high-rate trickling filters depends
on the concentration of suspended matter in the sewage. High
solids concentration will accumulate sludge on the filter,
and artificial aeration on high-rate filters would be required
to satisfy the oxygen demand. Blunk (372) stressed that down-
ward air distribution through the filter rather than upward
was desirable for the control of odor and filter flies. The
rate of air supply to an enclosed filter was reported by Dekema
and Murray (915) to be varied within the range of 0-1.96 ft3/
min./yd3 of filter medium. They also noted that there did not
appear to be any direct relation between the quantity of air
supplied and the performance of the filter; however, there
was a limit below which the efficiency of the filter deteri-
orated. Ponninger (3421) cited some advantage in long rest
periods for percolating filters which were covered and aerated.
He previously did experimental work (34O7) with the direction
of flow of air opposite to the flow of sewage, and reversing
in winter to warm the air, in the operation of an enclosed,
.nonplugging biological filter. The volume of air was main-
tained continuously at 30-4Otimes the volume of sewage over
a 24-hour period. The capability of reversing the flow of
the air was useful.
Concern was expressed by Bach (123) that carbon dioxide build-
up in the air space of an enclosed trickling filter would limit
the normal biological growth. Hurley and Lovett (2113) studied
enclosed trickling filters with and without aeration and con-
cluded that the conservation of heat obtained by enclosing the
filter more than counterbalances the cooling effect of the air
blown into the bed. It was further suggested that a shallow
enclosed filter 6 feet deep could be used for partial treat-
ment with about 50$ purification. Hurley and Windridge (2112)
described experiments based on the operation of two enclosed
filters without forced aeration. The quality of effluent
decreased appreciably in the absence of aeration. The
treatment of dairy waste was reported by Trebler et al. (4444),
using large enclosed filters and open filters with artificial
aeration. A rather unique medium for their filter was empty
condensed milk cans, which allowed easier movement of the
air. Hamlin and Wilson (1693, 4748), in a series of papers
on percolating filters, noted that artificial aeration pro-
duced more stable operation. Comparative costs for different
113
-------
types of covered percolating filters were given and discussed.
Artificially aerated filters were suggested by Wuhrmann (4831)
to be used in series operation with highly loaded activated-
sludge plants, taking advantage of the existing compressed
air capability. No relation was found between the efficiency
of the filter and the intensity of aeration. He concluded
that with a deep filter neither artificial aeration nor
covering was necessary.
Critique
Forced air ventilation may be required for strong waste waters
or adverse conditions. For municipal and low industrial waste
waters, natural ventilation should be adequate. Much research
and development have not firmly established the exact quantity
of air flow required for optimum operation. The direction of
air flow is preferably downward in the direction of liquid
flow. For deep beds, it is understandable that covers and
forced ventilation are not required. However, caution is
suggested in using uncovered filters in severe weather con-
ditions, because the distribution system may still freeze.
Natural Ventilation in Enclosed Filters
References to the literature were given by Imhoff (22O9) deal-
ing with natural aeration on open filters. Johnson (2329)
reported in 1952 on studies of natural flow of air through
percolating filters in Minnesota. Comparatively low rates of
flow of air were adequate to supply sufficient oxygen to re-
duce the BOD by 85 to 9O percent. Whitehead (4745), in dis-
cussing the work of Wilson and Hamlin (4748), was of the
opinion that there was more scope for improved operation in
using naturally ventilated filters in alternating series
than normal operation with forced ventilation. An enclosed
percolating filter known as the "Natural Draught, " as dis-
cussed by Tedeschi and Lucas (433O, 5637), performed well in
treating industrial as well as domestic waste waters in the
neighborhood of 5OO gal./yd3/day. Some molasses slop wastes,
reported by Krige (2571) to have been treated by enclosed
percolating filters under natural aeration conditions, were
inadequately treated due to the high organic suspended solids
loading. Aeration of the slop in admixture with domestic
sewage was also not satisfactory.
Critique
Natural ventilation is adequate under lightly loaded
conditions such as a conventionally loaded municipal waste
treatment trickling filter. Increases from the normal organic
and hydraulic loadings may require forced ventilation or
aeration to achieve the desired BOD level.
114
-------
Forced Ventilation with Covered Filters
Several technical papers, e.g., those of Hurley (2110, 2115),
Husmann (2136), Guilbert (1618), Imhoff (2209), and Mohlman
(3O51), have been published on the various aspects of using
forced ventilation with covered low-rate and high-rate per-
colating filters. The papers, in general, cite the advantages
of this concept, but are quick to point out that specific
cases of low solids concentration or dilute waste water made
enclosure or forced ventilation of the filter necessary.
Several patents on methods of forced aeration and enclosure
equipment have been issued. Typical is that of Girard (1482)
for forced aeration. A patent to Brintzinger (463) used
forced aeration through granular burnt pyrites media. Griffin
(1598) was issued a patent for uniform aeration as a means of
preventing clogging. Griffin (1597) was also granted a patent
in which air is passed through successive chambers in the same
order as the sewage, and is not liberated until the liquid has
reached its maximum degree of purification. Dekema (914) de-
veloped a patented process for passing the sewage or other im-
pure liquids downward through a series of enclosed filters
into which air was injected at a controlled pressure. This
method provides for changing the order of the filters to
permit cleaning the medium. Blunk (369) obtained a patent
which revealed a sewage filter with an impervious wall and
roof so that air can be drawn or forced uniformly upward
through the bed and discharged through ventilation shafting.
Pruss and Blunk (3477) developed two patents on the purifica-
tion of trade wastes and sewage. The liquids are forced into
a closed space above a filter bed and distributed over the
bed. Air is also forced into the closed space under pressure,
and liquid and gases separated below the bed, paying particu-
lar attention to the hydrogen sulfide balance.
In the period of the late 1930's and the early 194O's, there
was considerable activity in the development of forced venti-
lation of an enclosed trickling filter. Papers on the Pruss-
type enclosed trickling filter, e.g. Antill (82), and general
surveys of the type of equipment used, e.g. Se*e (3955) ,
evaluated the effectiveness of enclosed trickling filters with
forced ventilation. Hurley and Windridge (2111, 2112) pre-
sented a paper at a meeting of the Institute of Sewage
Purification in 1938, which generated much discussion and
indicated the advantages of enclosed filters, e.g., maturing
more rapidly than open filters, and requiring forced aeration
to maintain a good quality of effluent. Ponninger (3414) was
concerned with the amount of sludge being added to the en-
closed percolating filters and recommended that a medium of
suitable size be used in conjunction with periods of aeration
without the addition of sewage. Ponninger also emphasized
(34O7) that the direction of the flow of air under the forced
aeration conditions was important and that by forcing the air
115
-------
to go in the opposite direction to the waste some operational
advantages may be gained, especially in winter. Hurley and
Lovett (2113) worked with deep, shallow, enclosed filters and
found that deep covered filters may: (a) treat four to five
times as much sewage as open filters which were brought into
operation at the same time and were not yet matured, and (b)
treat about three times the flow treated on mature open fil-
ters. Blunk's findings (373) that the forced aeration should
be from the top down to handle the odor and fly problems
agreed with those of Ponninger (3407). The carbon dioxide
concentration was acknowledged as a problem by Husmann (2145),
who asserted that sufficient air should be supplied to trick-
ling filters to maintain the activity of the aerobic bacteria
and prevent harmful accumulation of carbon dioxide. Concern
expressed by some investigators regarding the carbon dioxide
concentration was minimized by Hamlin and Wilson (1693) in
studies which determined that carbon dioxide was not injurious
and the necessity for adequate ventilation was not related to
the carbon dioxide removal.
Operations in South Africa by Dekema and Murray (916) indi-
cated the effectiveness of the alternating double filtration
process using enclosed forced ventilated filters. Upward
ventilation gave slightly better results than downward ventila-
tion, but more power was required and the nuisance factors were
greater. The same authors (916) concluded later that this al-
ternating double-filtration process was capable of, as he de-
fined, a "purification capacity" of 3.2 compared with 2.3 for
a single enclosed filter,with the reference.of l.O for a sin-
gle, six-foot open filter. Factors affecting the efficiency of
covered percolating filters used in the treatment of waste
waters and sewage were reported by Yonner (4842), with particu-
lar reference to the von Roll system of forced aeration.
Many plants were built using the enclosed forced ventilation
trickling filter. Examples of these were the treatment of
dairy waste as recorded by Damin and Bock (859), of municipal
waste in South Africa [Hamlin (169O)] and in Glasgow [Cameron
and Jamieson (604) ], of waste water from abattoirs in France
[Planchon (3393)], of municipal-industrial waste in Germany
(5250), at a waste treatment plant located near a residential
section in Minnesota [Bardwell (172)], or recent plant expan-
sions in Transvaal [Barnard (179)], and treatment in western
France and Normandy [Guilbert (1617)], to mention just a few.
Enclosed aerated trickling filters have been used instead of
the activated-sludge process for secondary treatment of sewage
on the upper Ruhr, according to Priiss (3475) . Imhoff tank and
septic tank effluents have often been treated by enclosed
aerated trickling filters, e.g. Rohde (3668) and Davies (88O) .
Critique
Forced ventilation has been successful in many industrial ap-
plications of enclosed trickling filters. Evidence has pointed
116
-------
to the organic strength, suspended solids and hydraulic load-
ing as factors to be considered when questioning the desira-
bility of forced ventilation. Most municipal wastes being
treated on conventional depth filters (6 to 8 feet deep) re-
quire little or no forced ventilation. Forced ventilated
filters may be necessary during peak flows and more controllable
at stable atmospheric conditions. Air requirement of the acti-
vated-sludge process is in general more than that of aerated
trickling filters. The importance of good quality effluent
governs the economics of aerated filters.
ECONOMICS IN DESIGN
Cost reporting was practiced early in waste treatment history,
e.g. by Clark (690), and was the basis for economical designs.
Many investigators such as Smith and Ellison (4O72), Anderson
(74), and Sullivan and Wiley (4267) emphasized the signifi-
cance of an economical design and some of the factors involved.
Metcalf and Eddy (2983, 2984) reported in their texts, "Sewer-
age and Sewage Disposal" and "Disposal of Sewage," the princi-
ples of methods of financing, construction, and operation of
sewage works, along with some process cost data, to provide a
basis to the young design engineer. Later, texts by Babbit
and Bauman, "Sewerage and Sewage Treatment, " Q.I2 ) , and by
Besselievre, "The Treatment of Industrial Waste," (3O8) con-
tinued this effort to modern times.
Calvert (595) reported on the wide variation in unit costs
of sewage treatment works, depending upon the cost of construc-
tion of the plant, the site, equipment required and other
facilities of a variable nature. Thoman and Train (4359) re-
viewed previous studies on the estimation of costs for municipal
sewage treatment and gave an illustrated description of the use
of generalized cost curves in planning a pollution control pro-
gram. They also illustrated how this information had been
gathered previously and evaluated, based on limited information.
Reference was made to previous work by Velz (4534) as well as
other investigators, and to the comprehensive U. S. Public
Health Service compilation (5595), in which generalized cost
curves of trickling filters with separate sludge digestion
(Figure 7) and imhoff-type plants (Figure 8) were developed,
showing the cost per capita versus the population served. Cost
data on industrial waste treatment were at best meager (2145),
and, therefore, generalized planning was difficult since there
was no basis for extrapolation. The use of Engineering News
Record's Cost Index, based on 1913 dollars (4359), was recom-
mended in determining estimated costs for planning purposes.
In an effort to bridge the lack of information on industrial
waste disposal. Stack (2224) introduced modifications to the
equation previously developed to express the performance of
biological filtration and to illustrate the relation of filter
design to the economics of the process. Economics of design
117
-------
$1,000 p
$100:
COST PER CAPITA |
1,000 10,000
POPULATION SERVED
(DESIGN CAPACITY)
100,000
Fig. 7 - Construction Cost per Capita for Trickling Filters
[by Permission of U.S. Public Health Service
(5595, p. 16)]
118
-------
$1,000 p
$100 i
COST PER CAPITA |
1,000 10,000
POPULATION SERVED
(DESIGN CAPACITY)
100,000
Fig. 8 - Construction Costs per Capita_for Imhoff Sludge
Digestion Plants [by Permission of U. S. Public
Health Service (5595, p. 17)1
119
-------
were considered by Rankin et al. (3501), Biddle and Hoffman
(324), and the Trickling Filter Floor Institute (5574).
Besselievre's textbook, "The Treatment of industrial Waste,"
(308) stresses the importance of economics in operation and
capital expenditure for industrial waste treatment design. In
explaining the duties of the sanitary engineer, Southgate
(4128) specified that economical methods for the disposal of
domestic and trade waste waters must be incorporated into the
design of the sewage plant. However, Schelle (3860) suggested
that effluent standards created problems in the design of such
plants and problems in financing with which the engineer must
cope. The usual economic condition sought in waste treatment
plant design is to minimize operational expense, such as
pumping costs or artificial aeration, as described by Imhoff
(2205) . A common and preferred practice was to have high
operating costs rather than high initial installation cost,
according to Hansen (1718) in 1942, for the treatment of
sewage from military camps during World War II.
The decision to go one way or another on a design for large
installations in Chicago has often been based on economics as
reported by Eddy and coworkers (1093). The activated-sludge
process was used here rather than the Irahoff-tank trickling
filter system. Pbnninger (3410) in 1938 evaluated trickling
filter designs from the viewpoint of the greatest economy in
building costs. Process design and plant modifications with
economics in mind have been published for many years; e.g.,
Voorhies (4556) noted that the most economical layout of the
plant was to place the trickling filter entirely above ground;
Gascoigne and associates (1427) concluded that a new type of
biochemical process would cost less to build than activated
sludge or trickling filter and operational cost would be
between them; Jackson, in a chapter in the book, "Waste Treat-
ment, " edited by Isaac (2238), found it economical to salvage
an animal food from distillery waste, and Tiedeman (4392) and
Collom (763) evaluated the cost and effectiveness of chlorina-
tion in conjunction with trickling filters. Many other ex-
amples can be found in the literature.
Many of the early reports in the literature simply stated the
cost of construction or operation of a waste treatment plant
and the population it was serving, e.g., Doten stated that a
septic tank trickling filter plant for a population of 2,OOO
cost $8,OOO in 1919 (1O05); Hartley in 1914 published the cost
of an Imhoff-tank sprinkling filter plant as in excess of
$18,OOO for a population of from 3OO to 1,200 people (1762) ;
operational cost of a 3-million gallon a day plant was slightly
over $5,OOO in 1913 and almost $6,000 in 1914, based on an ini-
tial investment of $189,OOO for a population of over 20,OOO,
according to Eddy and Hammer (1O89) . Schreiber in 1942 (3893)
described the cost of construction of his "SU" trickling fil-
ters to be 4,OOO German Reich marks per thousand cubic meters
of sewage treated daily.
120
-------
Relative economic information was detailed by Wilson and
Hamlin (4746). Construction of an enclosed filter is about
eight times more expensive than an open filter, and the medium
is about three times more expensive than for an open one. Rela-
tive economics were reported by Jenks (2305), e.g., construction
costs for a Biofilter were slightly more than for an activated-
sludge plant and one-half to two-thirds those of a trickling
filter plant, while power costs were generally double those for
a trickling filter and one-half those of an activated-sludge
plant. Chipperfield (679, 681), while developing a case for
the use of plastic medium trickling filters, reported (a) 48$
savings in capital cost compared with conventional medium fil-
ters and 24$ compared to activated sludge for a strong carbo-
hydrate waste, (b) 28$ savings compared to conventional fil-
ters on a distilling waste where activated sludge proved to
be impractical, (c) 4O$ savings compared with conventional
filters on a brewing waste where activated sludge, also, was
impractical, and (d) 55$ savings compared with a conventional
filter on a food processing waste. Several investigators,
such as Chalmers (647) in 1967, developed various formulas
for calculating cost of treatment of waste waters from food
manufacture and coffee processing.
A separate sewage system for the treatment of hospital waste
would reduce the operating cost of a municipal treatment plant.
Poole (3423) described methods for reducing cost; e.g., changes
in plant process, reuse of waste water, inexpensive types of
treatment, and reducing the amount of mechanical equipment
required. Economics for minimum capital expense were estab-
lished by Smith and Wittenmyer (4O86), Bischofsberger and
Wurm (335), and Herriot (1899) during their design of bio-
logical treatment plants in this country and Europe. Several
modifications, such as the Emscher tank in combination with
percolating filters, have been reported by Bohnke (388) and
Kehr (2413) to reduce the cost of the waste treatment plant
and simplify operations. Many of the published papers contain
detailed design, construction costs, and tables of operating
costs, e.g., Weller (4673), Br'uhne (5O9) , Anderson (78),
Borrie (411), Fernie and Utting (1256), and Moss and Mengel
(3101) .
In several publications, such as that by Hepburn (1889), the
trickling filter was compared economically with other proc-
esses. He concluded that activated-sludge plants occupy
smaller areas than trickling filters, cost less to install,
but cost more to operate, and required skilled supervision.
Goudey (1525) after surveying the added benefits of sprin-
kling filter plants, concluded that the improvement in sus-
pended solids and BOD removal was not sufficient to justify
121
-------
the increased cost of trickling filters. The cost of in-
stallation and operation was estimated to be $10O,OOO mgd
capacity and $10/mg for the sprinkling filter plant versus
$25,OOO installation and $5 operation for separate sludge
digestion, and $6O,OOO installation and $15 operation for
the activated sludge process. Operational costs of 42 south-
eastern trickling filter plants were calculated by Franzmathes
(1344) and related to expanding flow capacities. German
investigators, such as Popel and Daser (3427) and Schreiber
(3894), published the comparative cost of construction and
operation on a basis similar to investigators in the United
States, such as Ellsworth (1162) and Montgomery (3O74). As
additional economic factors, several investigators, e.g.,
Imhoff (2212), Hipwood (1958), Pruss and Blunk (3474), and
Calvert (591) noted the land area and location of the waste
treatment plant relative to population centers.
Critique
It would seem that the area of economics in design would be
extremely clear-cut with answers prepared for most circum-
stances. This would appear especially true since much of
the capital spent for waste treatment was public funds, all
of which should be well justified. Unfortunately, waste
treatment, like city planning, crept upon this complex
society and only a few of the forward thinkers were able to
build logical, economic, well-designed plants and systems.
The results of the work of various agencies in this country
and abroad were made available so that the cost of treatment
of a million gallons of waste could be evaluated for the
next design problem. One difficulty in economic comparisons
is finding a common base. The workers have dutifully reported
their installation operational costs. However, the inves-
tigation to compare one city with another or, even worse,
one industry with another, to determine an economical design,
has only been initiated.
Papers on units such as the cost of a pound of BOD removal
or the cost of a foot of hydraulic head have been submitted,
but not widely used to relate economics to design. The day
is past when the engineer's design is beyond reproach. Com-
petition is keen and only the lean, efficient, reliable design
is sought. With computer inventory techniques and heavy
Federal financing, it should not be surprising that bases
for economical design will be established in an effort to
choose the best and most economical.
122
-------
SECTION IX
CONSTRUCTION TRENDS
Selected papers are noted below, indicating the type of treat-
ment plants which were constructed throughout the various
decades. Information on other construction aspects is in-
cluded, such as reports on prefabricated or package plants,
construction materials, and other considerations. Several
general reviews of construction were available in the litera-
ture, e. g. Khox (2532), Pearse (3326), Hyde (2157), Escritt
(119O), Daviss (892), and Friel (1355), with some evaluation
by the authors. At a general meeting, sanitary engineers
(2449) discussed the trends in the methods of waste treat-
ment in Great Britain and the influence of these trends in
equipment and instrumentation development.
DECADES OF TRICKLING FILTER CONSTRUCTION
In the decade prior to the turn of the century, complete dis-
posal of sewage, apart from screenings and inorganic matter,
was thought to occur by using contact beds or by the aid of
septic tanks. As more was learned about the bacterial treat-
ment of sewage, it was acknowledged that these processes
were not optimum. In the decade between 1900 and 1910, more
modern forms of tanks were developed and experience with
trickling filters was gained (5O59).
In 1892, experiments by Dibdin at Northern Outfall Works at
Barking Creek, Great Britain, were begun, resulting in the
development of fine contact beds (946) . Fine contact beds
were modified into coarse contact beds at Button which were
later used as trickling filters at Exeter (946).
1910 - 1920
For the period 191O-192O, the treatment plants in England were
characterized by detritus tanks, settling tanks, trickling
filters, and humus tanks, as reviewed by Marsland (287O). The
Kelowna, British Columbia, treatment plant consisting of
Imhoff tanks, percolating filters, and sedimentation tanks was
described by Meckling (2951). Sewage plants using covered
septic tanks with effluent discharge onto intermittent sprin-
kling filters (5441), or mechanical screens, Dortmund tanks
and storm water tanks (5058) were also constructed during this
decade. Hansell (1712) described the construction of a 16 mgad
(368 gal./ft2/day) waste treatment plant at Atlanta, Georgia,
which included coarse screens, grit chambers, Imhoff tanks,
fine screens at one of the plants, and roughing filters at the
other, dosing tanks and ventilated percolating filters. An
interesting plant (5459) was built at Prescot, England, on the
basis of the Travis theory of the dissolution of sewage solids
and consisted of a series of sand pits, screens, detritus
tanks, a "hydrolytic tank," primary and secondary filters of
123
-------
clinker and final settling tanks. A special feature of the
hydrolytic tank (5459) was the self-cleansing colloidors,
the function of which was to remove the finely divided
suspended solids and colloidal matter by surface contact.
McDonnell (29O8) supplied an engineering description of the
plant at Sedalia, Missouri, which operated on a combined
sewer system and handled the sewage of 2O,OOO people. The
plant was comprised of grit chambers, Imhoff tanks, dosing
tanks and sprinkling filters with a surface area of O.6 acre
and 5.5 feet deep. The 1914 sewage treatment plant for
Madison, Wisconsin (5O52), was designed for a population of
25,500 and comprised of settling tanks, a sludge digestion
tank with a contact dosing chamber and sprinkling filters,
followed by shallow secondary settling basins. Buchler (519)
described the construction in 1913 of the sewage system for
Singapore to treat the waste of 300,OOO people. The plant
consisted of screens and detritus tanks, 32 Imhoff tanks with
two- to three-hour retention, 55 coal filter beds, each 1OO
feet in diameter and 6 feet deep, and humus tanks with about
two hours retention. Patents were issued to Girerd and
Drapier (1484) and Girerd (1485) in 19O6-19O7 on the trick-
ling filter process.
1920 - 1930
During the period of 1920 to 193O, Russell (3793) described
the new southwest sewage treatment plant at Springfield,
Missouri, which was designed to serve 47/50O people and con-
sisted of screens, grit chamber, settling tanks equipped
with revolving sludge scrapers, sprinkling filters and sludge
digestion tanks, the effluent being discharged into Wilson
Creek. Similar plants were built at East Rochester, New
York, according to Skinner (4O44), Baltimore, Maryland, de-
scribed by Keefer (2388), and Rothwell Yorks, Great Britain,
(5464), with some modification of settling equipment. Hart
(1754) reported on the sewage plant at Leeds, where chemical
precipitation was used preceding trickling filters. The
waste from the population of 650,OOO with 598 mg/1 suspended
solids was treated by this method. The chemical precipita-
tion plant at Teddington was modified into percolating filter
treatment with the construction of five new filter beds. The
reason for greater adoption of the activated-sludge process
in America than in England was suggested by Morling (5463)
to be due to the average strength of the English sewage being
greater than that in America. Chemical treatment of the
sludge from a standard percolating filter plant (5263) ren-
dered it acceptable for a fertilizer. In the Berlin area
(5238), where local regulations permitted free irrigation,
this method was satisfactory. Storm water tanks in conven-
tional percolating filter plants were used routinely, e. g.,
by Silcock (4O25). Construction of trickling filters with
and without enclosures was reported by Walker et al. (4579).
124
-------
Imhoff tank-trickling filter plants were described by
Montgomery (307O) to provide adequate treatment for Wichita
Falls, Texas; a similar plant served Cleveland, Ohio (5O99).
However, the Imhoff tanks put into operation at San Bernardino,
California, were eventually converted to two-stage digestion
tanks due to an increased flow and required plant expansion,
according to Livingstone (2727). Imhoff (219O) discussed the
construction and operation of the trickling filter plant in
Moskau-Koschukodo. These plants were designed to operate with
artificial aeration through boiler clinker media 2.5 to 5 cm
(1 to 2 in.) in diameter and the surface load of O.95 meter/
hour, following a short aeration period by activated sludge
ahead of the trickling filter unit. Ma thews (2884) in 1928
stated that there were three kinds of filters, i. e., sand
filters, contact beds and percolating filters, and with all
three the requirements for satisfactory operation were sound
filter media, uniform dosage, sufficient aeration and proper
underdrainings.
1930 - 1940
As indicated by the literature, the decade between 193O and
194O was a time of high productivity and construction. Bowe
(424) discussed the reconstruction and enlargement of the
New Canaan sewage works, which involved the addition of
auxiliary equipment for aeration, solids-liquid separation
and sludge dewatering. Typical of the many plants built was
that at Mason, Michigan, described by Shephard (3988), which
was comprised of a grit chamber and bar screens, two primary
settling tanks with a detention period of 196 minutes, two
trickling filters with ventilating fans to improve aeration,
final settling tanks, digestion tank and glass covered sludge
drying beds. Ellipitical concrete domes were used to cover
the sewage filters at Hibbing, Minnesota (3O56) . At Marcy
State Hospital, New York, separate sludge digestion (to avoid
grease scumming) along with trickling filters, secondary
settling tanks, and chlorination equipment were constructed
to handle the flow for a population of 5,OOO according to
Ryon (38O5). WPA labor was used, according to Tatlock (43O1),
to construct the Dayton, Ohio, trickling filter plant. In
1935, out of 85 sewage treatment plants either completed or
in the course of construction, 41 were with sedimentation
and separate sludge digestion, 26 were modified Imhoff plants,
11 were activated-sludge plants, 4 used sedimentation and
trickling filters, 2 plants used plain sedimentation and one
was a septic tank plant (5319). A national census of sewage
treatment projects (5322) indicated that in 1939 a total of
848 plants were built which represented an increase of 18$
on the total of 4,7OO plants in service at the beginning of
1939. The trends during this era in sewage treatment were
for separate sludge digestion, the activated-sludge process,
and trickling filters (high-rate and Biofilters).
Dunbar filters were used by Thackwell (4349) at Tyler, Texas.
The plant, which was designed for an average flow of O.6 mgd
125
-------
(13.8 gal./ft2/day) and a maximum flow of 1 mgd (23 gal./ft2/
day), was comprised of a grit chamber, bar screen, hydraulic
pump, vacuum chlorinated effluent, multiple hopper bottom
settling tanks (1.5-hr detention), heated sludge digestion
tanks", glass covered drying beds, primary and secondary
Dunbar filters, and a step aerator.
Imhoff tank-trickling filter plants were still being built in
1938-1939, such as that at Hoopeston, Illinois (1947), to
replace septic tank-land irrigation plants. The waste treat-
ment plant built at Hamilton, New York, prior to 1931 con-
sisted of a pumping station, preliminary settling tanks,
sludge digestion tank with floating cover gas collecting dome,
a glass-covered sprinkling filter, an open final settling
tank, and sludge drying beds plus a short submerged outfall
sewer.
Construction in England was represented by circular trickling
filters with rotary distributors, described by Hurley and
Chibbett (21O9), and efforts to use settled effluent on
percolating filters without chemical treatment (3O95) were
similar to efforts in this country. Legal pressure was brought
to bear and resulted in the construction of additional filters
at Peterborough, England, to treat sugar beet factory waste
(4953). The sewage works at Truro were reported by Barnes
(181) to consist of a grit and screen chamber,pumping station,
settling tanks, trickling filters, humus tanks, storm water
tanks and sludge beds, which were designed to treat six times
the daily dry weather flow from a population of 12,OOO.
Oliver (3231) reported on the construction and operation of
the Buxton sewage disposal works. The use of roughing filters
and chemical addition satisfactorily treated the waste. Un-
derstanding of the construction of trickling filters and
activated sludge systems was reported by Reichle (3536), as
the advantages and disadvantages of the two processes were
noted and the combination of the two processes was recommended
for some wastes. Construction in England was reported by
Honey (20O4) to involve the use of screens, detritus channels,
hopper bottom upward-flow settling tanks, storm water tanks,
sludge dewatering tanks, uncovered digestion tanks, and sludge
drying beds. Prior to World War II, satisfactory results
were obtained from several small treatment plants which were
soon overloaded. Conversion of the smaller plants was de-
scribed by Mills et al. (3O16). Areas such as Romford and
Hornchurch, according to Snook (4O92), and Sanlesbury kept
up with their increasing population and sewage load by ad-
dition of more percolating filters, Dortmund humus tanks,
and storm water handling capabilities.
Construction in Germany (4257) , (3397), Argentina (24),
and Canada (8O1) was indicative of the world-wide activities
to treat various waste waters. These activities involved en-
closed trickling filters, activated sludge in series and
parallel, replacement of outdated septic tanks and irrigation
126
-------
systems, and repairing and upgrading existing facilities to
provide the degree of treatment necessary in each locale.
Activity in Rio de Janeiro was slightly different than that
experienced in many parts of the world (5O89). The construc-
tion and operation were by a private company which owned the
sewage system and used chemical-biological treatment as well
as no treatment at all to handle the waste from 92,000 homes.
1940 - 1950
In the period of time between 194O and 195O many of the re-
ports were involved with the modernization of waste treatment
facilities, e.g., the paper by Bragstad (440). Allen (51) pub-
lished the flow diagram and description of the sewage plant
at Regina, Saskatchewan, which treated the combined trade
waste waters and municipal sewage from a population of 55,000
people by sedimentation, biological filtration and activated
sludge. New plants were constructed at Lebanon, Tennessee
(5435), and plant extensions were made at many cities, such
as Stockton, California (2651), and Auckland (5215) , to up-
grade the degree of treatment. Biofilters were used often,
e.g., at Minneota, Minnesota (1963), and Santa Maria, Cali-
fornia (2892), in combination with normal grit removal and
primary sedimentation. The plant at Santa Maria also used a
Vacuator in an effort to remove the suspended solids by flo-
tation. Construction in Brazil involved the use of Imhoff
tanks, chemical treatment, treatment by the activated-sludge
process, percolating filters or a combination of the processes
(1463). Industrial wastes were treated using recirculating
high rate filters, according to Eldridge (1141) and Kirchoffer
(2494). The general developments in industrial waste treat-
ment, along with their legislation, were reviewed by Russell
(3790) in light of the progress made in the design, construc-
tion and operation of sewage treatment plants.
In this period, the literature indicated projects were con-
structed for different reasons. The motivation, such as that
at Neosho, Missouri, described by Smelser (4056) , was due to
the increase in population by the establishment of a military
camp. Sewage works at Broken Hill, Great Britain, were under
construction (consisting of coarse screens, primary settling
tanks, two secondary settling tanks, four trickling filters
with humus tanks, and four digestion tanks), but was stopped
August, 1942, with the understanding that it would be com-
pleted after the war (5O12). Other construction specifically
aimed at supplying services to the war effort was reported
(51O7) , such as the construction of the waste treatment plant
to handle sewage at the War Office Building in Arlington,
Virginia. This plant was comprised of aeration tanks, high-
rate trickling filters, final settling tanks, heated sludge
digestion tanks, and glass covered sludge drying beds.
Workers in the field adopted the attitude, as discussed by
127
-------
Siebert and Milligan (4O15), that war time pollution prob-
lems were to be solved with the minimum use of strategic
materials, e.g., the percolating filters were made without
sides or with sides of any material available such as rubble
or unreinforced concrete.
1950 - 1960
Construction activity, from 195O to I960, was extremely vigor-
ous as war time restrictions were relaxed and the Robert A
Taft Sanitary Engineering Center, of the United States Public
Health Service, was opened as a research facility. Plants
were generally constructed on conservative lines emphasizing
secondary treatment, disinfection of effluent, and high-rate
percolating filters. Also, the growth of the activated-
sludge process was encouraged and industry was taking an
active part in developing methods for the treatment of trade
waste water (4913). A common situation was that expressed
by Uzzle (45O4) for Hickory, North Carolina. After the war,
the waste treatment plant was overloaded and reconstruction
increased its capacity. The Atomic Energy Commission (327)
built the first full-scale plant for handling radioactive
wastes which was composed of mixing, primary sedimentation,
biological filtration with recirculation, secondary sedimen-
tation, chlorination and sub-surface disposal.
In answer to the demands of the housing boom, construction of
waste treatment plants for subdivisions was reported by
Greenleaf (1575, 1576), and Rader and Associates (5418), to
consist of primary sedimentation, biological filtration,
final sedimentation, effluent chlorination with Imhoff tanks
and sand filters or separate sludge digestion, depending on
the quality of effluent which was required under local con-
ditions. Voters backed the construction of new waste treat-
ment plants in Perryville, Missouri (48O7), and of adequate
facilities in Charlottesvilie, Virginia (17O8). The sewage
works often became a showplace of the communities, such as
at Long Beach, New York, (366O) and^ Easton, Pennsylvania
(189O). However, these waste treatment plants were built
along rather conventional lines, e.g., at Charlottesville,
Virginia, the plant, which served a population of approxi-
mately 18,OOO, was comprised of a barminutor, detritus and
primary sedimentation tanks, percolating filters, chlorin-
ation and final sedimentation tanks, primary and secondary
sludge digestion and sludge drying beds. New developments
were incorporated into designs to satisfy specialized con-
ditions, such as high oxygen demand, as reported by Lindsey
and Smith (2721). Sludge washing and thickening at Caldwell,
Idaho, were noted by Reynolds (3564), along with comminutors
and centrifuges, which were indicative of the popularity
gained by mechanical equipment. However, low cost operation
was still a desired construction goal, as exemplified by
Noble (3192) by the plant built at Vernal, Utah, which was.
128
-------
essentially, a conventional percolating filter plant.
Continued activity in the treatment of industrial waste in
the form of plant construction was reported by Ackerman (9)
for a specialty paper mill. The literature (277, 833, 1837,
1884, 2926, 448O, 4838, 4849) is extremely replete with
articles dealing with expansion, reconstruction, moderniza-
tion, and updating of existing waste treatment plant to meet
initial legal guides established. There was no particular
problem, as reported by Klassen and coworkers (25O3), in
normal waste treatment plants handling ground garbage dis-
charging into the sewers, which became popular equipment.
Combined industrial and domestic plants were fairly common,
e.g., Charlotte, North Carolina (134O). City-industrial
cooperation to develop new waste treatment facilities which
provide the higher degree of treatment necessary was cited
by Thomas (4363) . Construction of a new treatment plant at
Dexter, Missouri, was reported by Ross (3694), and the down-
stream farmers no longer complained about pollution of the
creek receiving the effluent and benefited by using the dry
sludge as fertilizer.
Efforts similar to those experienced in the United States
were reported overseas, such as the publication of the British
Standard Code of Practice which was used as a construction
guide for waste treatment plants (4952). Plants were expanded
and modified extensively in many locations in England, such as
Calne (1839) and others. These waste treatment plants were
equipped with sedimentation tanks, chemical treatment tanks,
percolating filters, humus tanks, sludge removal by monorail
at Bisley (5O25), alternating double filtration at Liverpool
(5219), modification of sand filters at Wisley (4289),
abandoning of cramped facilities and reconstruction at
Taunton (4494), incorporation of storm water treatment in-
cluding some trade waste waters at Bromwich (5532), and
sludge from the alternating double filtration plant at
Lundwood was used to produce gas for generating electricity
(5270).
Elsewhere in Europe, activities were similar in the expansion
and renovation of existing waste treatment plants and new
construction for overloaded areas, such as that reported by
Haury (1811) for several German communities. The Heilbronn
waste treatment facility, described by Pbpel and Baser (3426),
added biological filters, and an aeration tank with the
necessary appurtenances was modified. The construction in
Germany in the post-World War II period indicated the trend
to more secondary treatment and beneficial uses of digested
sludges for power, fertilizer, or fill material (67, 47O,
3426). New construction of sewage disposal plants in South
Africa was also continued and, typically, was reported by
Hamlin to be (1696) treatment by primary sedimentation,
129
-------
biological filtration, final sedimentation, and filtration
of the effluent on reversible sand filters. Percolating
filters were used in Switzerland along with activated-sludge
processes for handling populations from 1,OOO to 16,OOO in-
habitants plus industrial waste waters (4927). Construction
of sewage plants at Bad Voslau, according to Kreuth (2569),
was comprised of screens, two Emscher tanks, two covered
percolating filters, and secondary sedimentation tanks to
handle the waste water from 6,OOO inhabitants. A new sewage
treatment plant built for Oshawa, Ontario, was described by
Reilly (355O) as a standard biological filtration plant with
the effluent discharged to a lagoon prior to discharge to
Lake Ontario.
I960 - Present
The literature from I960 on represents activities in legisla-
tion and in further construction and modifications along
similar lines as that of the 195O's, with a few modified
processes being built. Expenditures for construction of waste
treatment plants were reported by many, e.g., the staff of
Water and Waste Engineering (5647) in 1966. Laws similar to
those of counties in Florida provided motivation for the con-
struction of secondary treatment plants, some of which are
high-rate filters and vacuum filtration of digested sludge.
Torgerson (4434) described the facilities built to keep up
with population growth of Brigham City, Utah. With future
expansion in mind, the Charlotte, North Carolina, plant was
built in two stages: first stage had biological filtration
and the construction connections provided for the second
stage (4295). Efforts by the Government to provide new waste
treatment facilities for new installations and to update
existing installations were published, e.g., Whitesands
Missile Range (5627) and Pearl Harbor (4O79). Complete waste
treatment efforts were intensified as reflected in the litera-
ture by Linke (2722) and others (5644) to meet more rigid ef-
fluent standards. Increasing waste treatment capacity was
reported by Smith and Wittenmyer (4O86) as being aggressively
sought, using established and new techniques such as plastic
medium oxidation towers. Several areas in this country, such
as Rockingham, North Carolina (47O7), and overseas at Hudders-
field and Winsford, England (1512, 5551), found it economical
to construct regional waste treatment facilities using con-
servative designs of standard percolating filter waste treat-
ment plants. Construction activities in Johannesburg, South
Africa (4991), Sweden (5612), India (331O), Canada (642),
France (4352), and Great Britain (5286) were indicative of
the intensified world-wide interest in pollution control.
Many of the waste treatment facilities that were constructed
overseas were built along the lines of extending primary
plants to secondary plants and correcting an inefficient
process which had been operating for several years (5277) .
130
-------
Critique
Papers on the construction and description of new or modified
waste treatment plants were published extensively, and it
appears that there were three to five papers written for each
facility built. These pages were informative, but the value
of their volume in the literature was questionable. It would
appear that authors should use more discretion in titling and
summarizing their work. Several times titles and abstracts
gave the impression that the "ultimate in waste treatment"
had just been constructed. Upon review of the paper it was
noted that it was a typical construction-operation report.
Quite often construction elements were reported as being used
with little or no justification. The cost of construction
was included less frequently in the current literature. Con-
struction trends in the decades demonstrated continued in-
terest in biological trickling filters, with intermittent
challenges from other processes.
PACKAGE PLANTS AND PREFABRICATED INSTALLATIONS
Early in the twentieth century, Meckel (2950) and Gerhard
(1458) reported the need for wastewater treatment facilities
which could handle waste flows from small or remote communi-
ties efficiently and without much operational attention.
Several papers were issued on the problem of sewage treatment
for individual houses (5614), and waste treatment systems for
rural areas and institutions were described by Hopkins (2O23)
and by Berry (291). Standards were developed by the British
Standards Institution (4952) for 3OO people and by Germany
as reported by Antze (84). The principles of design and
construction were outlined for small domestic installation
treatment facilities by the Ministry of Housing and Local
Government (5189). Pettet and Jones (335O) found that by
using some of the recommended standard procedures, under cer-
tain circumstances, additional treatment schemes were required,
e.g., percolating filters following septic tank installations.
Faulkner (1246), Sharp (3970), Rogers (3662), and Stephenson
(42O9) reported on various aspects of design and construction
of small waste treatment facilities (populations of 3OO or
less or remote locations) and described the installations.
These were composed of pumping and screening, sedimentation,
biological filtration and/or extended aeration processes,
sludge handling devices, and the use of package plants.
Package plants have been prepared and installed throughout
the world to satisfy various conditions, some of which are
described below. Gibson (1467) discussed a package plant
comprising a septic tank which encircled a percolating filter
and was designed to handle the household wastes of five adults.
LaFontaine (26O7) described a typical package plant as con-
taining primary sedimentation, aeration, final sedimentation
131
-------
and chlorination of the effluent for the treatment of sewage
of small municipalities at low cost. In the development of
the Haifa waste treatment plant, several small package plants
were constructed in outlying areas of the city in addition
to the two-stage, high-rate biological filtration plant located
centrally, according to Goldstein (1496). In the development
of their waste treatment system, the Auckland Metropolitan
Drainage Board constructed and evaluated package plants to
serve small communities (40O4, 4924). Illustrated descrip-
tions were given of various package sewage treatment plants
which were available in capacities from 1,OOO to 125,OOO
gallons per day by Public Works Magazine (5413). The design
and operation of the two-stage biological filtration unit
packages to serve populations from 5O to 5OO, have been de-
scribed (5118, 5414, 5641). The construction of the package
plant has been quite often around some central element, such
as that reported by Crawley and Brouillette (815), in which
a plastic media trickling filter was used. Package waste
treatment plants have also been constructed to serve as tem-
porary treatment until permanent works could be built (5113).
They have also been used (4514) as the next step in the
development on to extremely simple treatment plants, such as
lagoons or seepage pits. During the construction of small
communities in Florida (2475), package plants were used, and
the sewage from a population of 4OO was handled quickly and
efficiently on an estate in the Netherlands (3688) by this
method. Using the package plant concept, Ingram and co-
workers (2228) investigated several conventional systems to
handle the waste treatment problems in space capsules, but
found the results were not as satisfactory as distillation.
Many patents have been issued on package plant designs, e.g.,
to Gilbert (147O), Dannebaum (866), L. von Roll A. G. (3678),
Knapp et al. (2514), Mead Corporation (2944), and Schreiber
(3897), to name just a few.
Critique
Package plants have been used with various degrees of success.
Unfortunate experiences and optimistic advertisements have
damaged the credibility of this concept. For the solution of
a waste treatment problem at a remote location handling a
relatively small flow, package plants are extremely valuable.
Temporary treatment systems during construction, expansion or
seasonal operations have also used package plants. The
various plants available were designed for specific purposes
and were constructed to operate with little maintenance.
However, to intentionally design a system and not specify
procedures for some form of maintenance is a practice that
may be more oriented to marketing than results. Papers on
the construction of package plants were complete and usually
based upon justifiable claims.
132
-------
MATERIALS OF CONSTRUCTION FOR TRICKLING FILTERS
Most of the reports on trickling filters in the literature
have dealt with the process modification and operation. How-
ever, of equal significance are the materials of construction,
which have also generated a considerable quantity of publica-
tions. Calvert (593) described the advantages of some of
the electrically driven equipment and construction materials,
such as precast concrete, to be used for waste treatment ap-
plications. Corrosion of distributors and other com-
ponents was a problem, as noted by Schaetzle (3853), and an
alternative was offered by Lux and Brady (2797) by a greater
use of plastic materials in equipment for the treatment and
disposal of corrosive waste waters. Quite often the mate-
rials used in sanitary works were described in detail, e.g.,
the book by Blake (348). Escritt published several papers
(12O1, 12O6, 12O9) dealing with various aspects of the con-
struction of biological trickling filters, e.g., the type of
medium, the construction of underdrains and walls, the con-
struction of fixed and rotating distributors, and the shape
and methods of drainage of trickling filters. In an effort
to discourage corrosion and deterioration of waste treatment
plants, Browne (5O4) studied the effectiveness of various
paints to inhibit deterioration. Experiments have been car-
ried out in Great Britain by Lea and Bessey (2656) to deter-
mine the extent of cement deterioration at waste plants,
treating dairy waste waters, and it was found that deteriora-
tion of Portland cement mortars was influenced by the tempera-
ture of the effluent. The most severe cases occurred (2656)
in storage and holding tanks and partial protection was ob-
tained by using tar or bitumen based paints. The only com-
pletely satisfactory results were obtained from untreated,
high alumina cement. The structural details of walls, floors,
drainage channels and ventilation systems were described by
Rankin et al. (35O1) and the "Handbook of Trickling Filter
Design" (5574).
Distribution
The significance of distribution of waste water over the
trickling filter was recognized by Simpson (4O35), Eliassen
(1152), Chase (656), and others. For proper distribution and
intensity of the waste water falling on the medium, flows
were adjusted (4035) to avoid stripping the biofilm so that
some larvae remain in the filter to aid in checking excessive
growth of the film and control filter fly breeding. Milk
wastes, when treated on biological trickling filters packed
with small grain material, did not respond to any particular
method of distribution (5373). Hewitt (1927) reported that
rotary distributors constructed with a main arm and an auxil-
iary arm operated much better with the auxiliary arm con-
tinually working. Spray jet distributors were reported by
133
-------
Merciot (2974) to increase the aeration of sewage, but their
installation and upkeep were costly. Maier and Sohler (2830)
in 1929 used several types of distributors, such as travel-
ing, scattering and rotary sprinklers, with respective cost
information. Several distributor arrangements were noted by
Baxter (224), e.g., the Salmesbury distributor consisting of
corrugated metal sheets with staggered outlets.
While using spray distribution, various workers, such as
Watson (4633) and O'Shaughnessy (3257), modified nozzle design.
Spray nozzles which delivered sewage a distance of 9 feet
raised the dissolved oxygen 5O$ or to about 6 ppm, according
to Anderson (8O). Watson (4628) disagreed and stated that
better distribution caused by movable distributors is not
worth the increased cost over that of fixed sprays and that
the difference in efficiency was hardly noticeable. Watson
(463O) recommended that fixed sprinklers be used working under
a head of from 5 to 6 feet. Wilson (4762) cited the advan-
tages of aluminum and its corrosion resistant alloys for use
in sewage works construction. Tests illustrated that aluminum
and its alloys were almost unaffected by the exposure to
average domestic or combined domestic and industrial sewage
and were a desirable material for trickling filter distributors,
as well as other appurtenances. Information dealing with non-
corrodible material for sewage construction was published,
and it suggested that a wider use of steel substitutes such
as Armco and Tonco iron, monel metal, copper bearing steel,
tin, nickel, brass, noncorrodible irons and steels and alu-
minum should be used in wastewater treatment plants. It was
stated that aluminum alloys had been able to withstand im-
mersion in sewage for 360 days at Cleveland, Ohio, without
detectable deterioration (537O).
Several devices were constructed to distribute the sewage,
such as the Fiddian rotary machines (958, 5458). Other
distributor arrangements involved the flow of sewage downward
from pipes above the beds in vertical jets which impinged
upon cup-shaped splash plates (513O), channeling in the medium
(5125), water wheels (2844), automatically controlled travel-
ing distributors on rectangular beds (5O51), and injecting
controlled quantities of sewage at various levels into an air
gap medium (2222) . Sufficient disagreement on the optimum
distributor was generated that Saylor (3841) evaluated the
relative efficiency of several of the sprinkling devices
and found nozzles to be very satisfactory. Devices described
by Ippolito (2231) which used spraying paddles on rotating
distributors were reported to perform better than rotating
distributors with simply bored arms. The Candy-Whittaker
patent (1173) for a rotary sprinkler device was one of the
earliest patents for distributors. Patents were issued on
many of the distributing devices, for example, Bolton (396),
Bohman and Nettermann (386), Geiger (1446), and Hoffmann
(1976). The quantity of patents on distribution systems has
necessitated a section on patents, which is given later in
this review.
134
-------
Critique
Several materials were used in many designs for distribution
of waste waters on biological trickling filters. The results
of several investigations indicated that aluminum rotary dis-
tributors were well suited to handle waste waters for many
years without corrosion problems. The volume of literature
was indicative of the many problems which have arisen in the
development of distribution devices, and the economic poten-
tiality of providing this equipment. Problems with corrosion
were common in the plants built in the 1920's and 1930's.
New alloys became available after World War II, and design
and construction trends incorporated them into existing and
new installations. Sufficient work was reported on effects
of distribution/and manufacturers' claims have been adequately
evaluated by the plant operators.
Medium Materials
A wide variety of materials was used for trickling filter
media; for review purposes, these materials have been divided
generally into the four categories: (1) man-made inorganic,
(2) natural inorganic, (3) man-made organic, and (4) natural
organic. A critique of the literature on materials of con-
struction for media follows the media evaluation section.
The quantity of literature available on the various media
types, tests, and operational advantages and disadvantages
requires review of selected articles with heavy referral to
additional references. A high percentage of the literature
was concerned with the comparison of one medium versus an-
other for a given waste and set of circumstances (which is
summarized at the end of this section), information on speci-
fic media (which follows), and development of procedures for
medium durability testing, for instance, the British Standards
Institute (4951), Payrow (3314), and Kreige (2566).
Man-Made Inorganic Media
Under the category of man-made inorganic media, materials such
as asbestos, brick, clinker, metal, lime-treated peat, vari-
ous slag materials, and various tile or ceramic forms are
included. Asbestos material was used by Eckenfelder (1O82)
and Zigerli (4865) as a medium for biological trickling fil-
tration. Metal was used as trickling filter media in the
form of copper-mine waste rock by Horasawa (2O3O), empty
condensed milk cans by Trebler et al. (4444), rotating metal
disks by Hartmann (1772) and plastic impregnated metal or
paper by Sullins and Self (4265).
A tremendous amount of literature (more than 1OO reports) is
available on the use of clinker, i.e., the solidified products
from a combustion process, as a medium for biological trick-
ling filtration. The literature indicates (1241, 5421) that
135
-------
clinkers for trickling filter media have been used since the
early 19OO's and are being used contemporarily (4428, 4463)
to evaluate other types of medium, waste to be treated, and
efficiency of removal of waste and bacteria.
Cotton finishing waste (384, 5616), paper waste (1394), and
chemical waste (1573) are just a few of the many examples
which are reported as being treated using clinker-filled
biological trickling filters. Patents, such as that assigned
to Arnatt and Hopper (9O), using clinkers in various configu-
rations for biological trickling filtration were issued. It
is interesting to note that research was done at the Hamburg
Hygienic Institute on clinker-filled trickling filters to
prove that they were biological reactors, thus disproving the
so-called Travis theory that the purification process is not
the result of biological activity (5124).
Another source of man-made inorganic medium has been referred
to as boiler slag (3664) , gas plant slag (2998) , and blast
furnace slag (4854) . The wide acceptance and usage of slag
as a trickling filter medium were based on soundness tests
such as those reported by Kreige (2566) , and Payrow (3314) to
demonstrate the durability of the material. Hornmon (1995)
recommended that the specifications for slag to be used in
trickling filters are: slag should pass the sodium sulfate
durability test at 20 cycles, metallic iron should be re-
moved, the gradation should be such that 95^ of the material
is between 1.25 to 2.5 inches or 2.5 to 3.5 inches as re-
quired, and placement of the material should avoid breakage
and segregation. Organizations/ such as the British Slag
Federation (4950) and the National Slag Association (1993) ,
as well as the British Coke Research Association (4947) , in-
vestigated and promoted the use of slag for various purposes,
one of which was its use as a biological trickling filter
medium. A patent was issued to Thomas (4362) in which sew-
age was passed upward through a loose bed of granular mate-
rial, such as cellular slag, for colloidal and organic re-
moval, and Rudolfs (3738) described a typical slag-filled
filtration operation . Solin and Erlebach (41O3) demonstrated
that slag had a sorptive property for the removal of mono-
and divalent phenols. Erlebach et al. (1188) reported that
removal of fatty acids from waste waters by slag was analogous
to the phenol removal. Sorption of cresols on slag was re-
ported by the same group (41O6), with the process of sorption
and resting described to occur on a seven- and two-day basis.
Madera and coworkers (2816) reported on the successful use
of slag, taking advantage of sorptive capacities as well as
biological growth. A variety of aqueous solutions was used,
e.g., monovalent and polyvalent phenols, cyanides, trinitro-
toluene and synthetic detergents or industrial waste con-
taining some of these substances. A waste slag obtained from
the manufacture of phosphorus was reported by Fromke (1363)
to contain less solids, had a higher pH value than boiler
136
-------
slag, and would accumulate less sludge. The Water Pollution
Research Laboratory (5193) has, for many years, investigated
various aspects of the use of slag as a biological trickling
filter medium. Waste waters from the fermentation process
were reported by Pitts (339O) to have caused some disintegra-
tion of blast furnace slag used as a filter medium. Sima
(4030) described the operation of slag filters in Prague and
stated that slag medium was better than limestone for the
given hydraulic and organic loadings. Size effects of slag
were reported by Oliver and Hall (3241); slag in sizes smaller
than two inches produced effluents of higher quality, but
were subject to considerable amount of ponding. Information
may be found in the many additional references dealing with
various industrial wastes having been treated on slag as a
medium for trickling filters, i.e., biological functions
occurring during filtration, various loading rates, waste
treatment at various geographic locations, specific oper-
ational difficulties, ventilation and temperature effects,
and miscellaneous facts.
Ceramic materials such as vitrified clay or brick are another
type of inorganic man-made trickling filter media. Goldthorpe
and Nixon (15O8) have proposed different modifications of tile
media, such as the "Huddersfield" tiles, which were described
as a tray supported on three legs. As the liquid overflowed
from the tray, it was aerated as it was exposed as a thin
film on the curved underside of the tile before forming drops
which fell into pools on the trays below. Leibee (2673) and
Smith and Leibee (4O75) reported on the advantages of glazed,
vitrified tile for use as a medium. Food processing wastes,
high in organic material, were treated satisfactorily on a
tile medium which had one-inch diameter holes extending from
top to bottom according to King (2492). Raschig rings
(designed by H. R. Straight) were compared by Levine et al.
to crushed quartzite for the treatment of milk and packing
house wastes, with the Raschig rings being superior (2701).
Studies by Edwards and Adams (1116) at the Lawrence Experiment
Station were made on tiles with one-inch diameter holes
and crushed stone. It was observed that the crushed stone
had a tendency to clog at high rates but the perforated tile
filter did not. Several investigators, e.g., McCulloch
(2903), Lubbers (2774), and Hamlin (1697), reported apparent
advantages of concrete or unglazed earthenware pipe and tile
as a biological filtration medium. The development of many
of the ceramic materials resulted in patents being issued,
such as to Halvorson (1675), Lewis (2711), and Page (3268).
Ceramic materials and clays for trickling filter material
medium were used to such an extent that tentative standards
for strength determinations were issued in 1945 by the
American Society for Testing Materials (4917) and again
in 1948 (4918) . Many additional references dealing with
the various forms of tile media in biological trickling
137
-------
filters for the treatment of domestic and industrial waste
under various geographic conditions have been frequently
reported.
Natural Inorganic Media
Other media which were used in trickling filters may be
classed in a group called natural inorganic materials, which
refer to crushed stone which may be basalt, granite, lime-
stone, sandstone, slate, vermiculite or other materials
which would be fractured, and various forms of gravel. The
popular use of crushed stone for trickling filter media
stimulated the report (5436), in 1944, of tentative standards
for high rate percolating filters. The recommended size and
type of material were stated to be greater than two inches
but less than four inches, and could be crushed rock, gravel,
or its equivalent, placed in layers to a depth of not less
than 6 feet. The use of several mineral aggregates was
reported by Kriege (257O) and specifications dealt with the
various properties of the media, such as soundness and frac-
ture faces. Stanley (4183), as well as Levine (2706) and
Dahlem (854), stressed the importance of the use of a good
hard limestone or other material. Small particles and dust
should be avoided (4183) and the stone should be rough
enough to afford a surface for bacterial film development,
but not so rough that the film will not slough off at least
twice during the season. Crushed stone has been used by
Eaves (1O68) in conjunction with other media such as clinker
in stratified layers. The stone size ranged from O.75 inch
(1116) to 3 inches (1O68). A variety of industrial wastes
has been treated by percolating filters using crushed stone
medium, e.g., Munteanu (3124) described the treatment of
waste waters from cellulose and viscose plants using this
medium.
The Duribar filter, used in the sewage treatment plant at
Tempe, Arizona, was a modified trickling filter which con-
tained 2O inches of three- to six-inch stone, eight inches
of 1.5-inch stone and six inches of 0.25-inch gravel with
sixteen inches of O.25- to 0.125-inch sand which provided
(858) a mechanical-biological treatment. Many other
crushed stone references are recorded in the literature.
Volcanic materials have been used as media very early in
sewage treatment history and their popularity was attributed
to the satisfactory results for industrial and domestic
waste treatment (5458). Wormser (4822) studied the decon-
tamination of radioactive liquids by percolating filters
filled with pozzolana and observed that the removal depended
largely on sorption of isotopes by the pozzolana bed. Brebion
and Huriet (452) reported on physico-chemical and biological
purification of waste waters from a pulping process and com-
pared a pozzolana percolating filter to a partition apparatus,
138
-------
favoring the percolating filter. Whinstone (another form of
volcanic material) was reported by Hunter and Cockburn (21OO)
to be a satisfactory medium. An 18-foot deep bed was designed
using a gradation of from 0.75-inch to 3-inch material. Suc-
cessful uses of various basaltic forms were reported (630,
387O, 4163, 4285, 4537, 5127, 5458); the results show that
these forms are a desirable medium capable of handling strong
European domestic waste, industrial waste, waste in remote
locations, and to minimize nuisance problems during operation.
Demoll et al. (928) investigated the use of lava filter beds
and concluded that they have the ability to retain waste for
extended periods, thereby providing the best structure for
rapid and complete colonization of purifying organisms. Blunk
(367) investigated detention times and growth of organisms
on lava slack media and determined the relative detention time
and the biofilm distribution to be quite promising. Munteanu
and coworkers (3125) studied percolating filters containing
basalt and crushed stone. The basalt of the same particle
size as the crushed stone and under identical operating con-
ditions had an oxidation capacity of 5 to 6 times greater than
the crushed stone (4OOO grams of oxygen per cubic meter per
day in the summer as compared to 75O for the stone). Since
there were abundant deposits of volcanic basalt slag in
Romania, Murgoci et al, (3132) used high-rate tower percolat-
ing filters with basalt medium (3O to 5O millimeters nominal
was more efficient than 2O to 3O millimeters) and stated that
the filters were easily adaptable to the several wastes treated.
Typical of the patents issued was that -to Blunk (37O), where
a filter medium of granulated lava was described.
Granite is another form of crushed material which has been
used quite extensively for a medium, according to Halvcrson
(1667) . In 1928, an examination of trickling plants (4525)
showed that about one-third of the plants suffered from
disintegration of trickling filter medium and clogging; however,
granite, trap rock, quartzite, and gravel have proved satis-
factory, if the material had a minimum size of not less than
one inch. During the reconstruction of several plants (429,
353O), the existing medium was replaced by granite or quartzite
medium. Fractured face material, such as broken granite, 2 to
3.5 inches in size, was used by White (47O5) to provide a large
area of contact and ample space for ventilation. In a plant
which was engineered for flexible operation, 8-foot deep
quartzite medium (graded from 2.5 to 3 inches) was used with
and without forced ventilation to satisfy the quality require-
ments at Webster City, Iowa (5O34). Industrial wastes, such
as the creamery, packing house, tannery, and rubber wastes
reported by Levine (2699), Veeraraghavan and Hariharan (4528),
and Molesworth (3O55), were treated on granite media in a
trickling filter with only minimal difficulties. Many examples
of the application of crushed granite as a medium have been
published.
Limestone is another example of a crushed inorganic material
which was successfully used as a trickling filter medium.
139
-------
A motivating factor in the use of limestone as a medium was
the local availability of material which would satisfy the
hardness, toughness, and wear required for long-term opera-
tion (3686, 54O7). The physical strength of the limestone
created much discussion and resulted in the development of
testing procedures to determine its resistance to abrasion,
freeze-thaw, and other mechanical failures, as described by
Schaetzle (3849), Lamar (26O9), and Wukasch and Bloodgood
(4834) . Limestone has been preferred over other materials
by Harrison (1751) and Maier (2832), and examples of operation
under ice conditions (167) have indicated its desirability.
Mohlman (3O45) used dolomite limestone (2 to 3.5 inches
diameter) as an experimental high-rate trickling filter for
the sanitary district of Chicago during developmental inves-
tigations in 1935. Limestone trickling filters have been
reported by Logan (2753), Meyeren et al. (2989), and Wisner
and Pearse (4795) to handle satisfactorily a variety of
difficult wastes, such as dairy food processing and meat
packing wastes. In treating waste from malting and brewing
installations, Oliver and Walker (3239) used local limestone
as the filter medium to help stabilize the possibility of
acidic waste waters, but it was found that, as erosion occurred,
the pieces of limestone tended to knit together, thereby reduc-
ing the available surface area.
The use of sandstone has been reported by Humphrey (2O87),
Mars (2865), and Grace (1529) to have some advantages at
small installations. Others (244, 1529, 4O5O, 4979) have
used sandstone medium in conjunction with other trickling fil-
ter media where the sandstone chips were usually of a coarser
size (1 to 4 inches) for adequate ventilation (1529). Slate
was used in biological filtration processes by Clark and Gage
(693) and others (5O74) for contact beds. Slate beds were
more often used in sludge handling (5454) and on the Dibdin
slate beds (1323, 322O, 4771) which were used to trap colloi-
dal material (5454) . other references to slate medium provid-
ing satisfactory treatment were reported by Dibdin (944, 945).
Coral was reported by Molesworth in the book, "Waste Treat-
ment, " (2238) to have been used for trickling filter operations.
In 1929, eight coral trickling filters handled half of the
secondary sewage treatment from the city of Singapore (5O87).
Similarly, in 1932 Buchler (519) stated that Singapore's sewage
treatment system consisted of 55 coral filter beds, each 1OO
feet in diameter, and satisfactory effluent was being obtained.
Vermiculite was a suitable filtering medium for biological
treatment of cyanide waste waters from automotive manufacturing
installations (462) .
A great deal of literature has been published on the use of
gravel as a filter medium, beginning with that by the Lawrence
140
-------
Experiment Station in 1908 (691) in which biological treat-
ment was investigated. Gravel (1.5 to 2.5 inches) was reported
by Steel and Zeller (4198) to produce an effluent equal in
quality to that produced by crushed stone of the same' size.
In a later study by Steel and Zeller (4197), comparing crushed
limestone and gravel, the limestone deteriorated after three
cycles of the sodium sulfate test, whereas the gravel was
able to withstand more than 20 cycles. Thompson (4374) re-
ported that percolating filters filled with water-worn gravel
(1 to 3 inches in size) would have a smooth surface and would
result in a shorter period of contact than rougher media, but
suspended material would pass through more easily, minimizing
ponding. The properties of chert gravel were investigated
by Lamar (261O). The results were not optimistic because
the gravel fractured and undesirable algae growths developed.
Furman (1384) stated that local limestone and river gravel
were both satisfactory for Florida operations. In investigat-
ing the ecology of gravel trickling filters, Weninger" (4677)
observed that the biological population fluctuated through-
out the gravel material as the seasons changed. Hydraulic!
investigations of gravel percolating filters were made by
Thompson (4373) on high-rate dosing applications. Gravel
was used in combination with other materials (165, 1O31,
5468) in mechanical-biological filtration systems. Objections
to the use of gravel were made by Herring (1898), who showed
that 1.5-inch size gravel had only about one-half the speci-
fic surface area possessed by an equal volume of slag. Clinker
was reported to give better results than crushed gravel fdr
the treatment of combined trade waste and domestic sewage -in
Great Britain (1573). The many combined uses of gravel and
its common availability throughout the world generated a large
quantity of reports on its successful application and problems
in usage.
Natural Organic Media
Another broad category of material which has been used for
trickling filter media may be considered as naturally occur-
ring organic material. This type of media includes wood in
various forms, organic deposits such as peat, soft coal, hard
coal, and other ingenious uses of available materials. Wooden
lath was used (5113) as a medium for trickling filters; an
advantage was the light weight, which was appreciated for port-
ability. Kimball and Wadhams (2482), while investigating
methods of treating industrial waste, determined that a trick-
ling filter composed of layers of lath lattice work would
handle the wastes which were discharged from distilleries>
breweries, and dairies to produce a 9O$ purification without
clogging. Levine and coworkers at Iowa State College (2691,
2692, 2693, 2696, 5231) used a lath filter for a series of
investigations on the biological purification of a d^iry
waste and were able to obtain satisfactory effluent quality.
141
-------
Steel and Zeller (4196, 4858) have used a lath filter for the
treatment of dairy waste with similar successful results.
Kennedy (243O) used a lath filter following an aeration unit
to handle a creamery waste with domestic sewage. Food process-
ing waste was reported by Kimberley (2484) , in 1927, to be
treated on lath at the rate of 2 million gallons/acre/day
(46 gal./ft2/day). Middlebrooks (2991) used a random wood
stack in handling high temperature waste from paper process-
ing as an effective complementary treatment to plastic media
trickling filters. Distillery wastes were treated according
to Hoover and Burr (2O21) to 9C$ removal on lath trickling
filters when a load of 2,OOO mg/1 BOD was applied at a rate
not exceeding 1 mgad (23 gal./ft2/day). Lath was used by
Sander (382O) to treat compressed yeast factory waste waters
and, with proper dilution of 1:3 to stabilize the pH, good
results were obtained. Wooden slats were also used in com-
bination with other materials such as coke, as revealed by
Biczysko (321), for treating phenolic waste waters in a high
temperature biological oxidation system.
Other examples of naturally occurring organic materials used
for trickling filter medium were brush branches, twigs and
straw. Imhoff (2168) and Thackwell (4347) reported the eco-
nomical use of brush wood as a medium for contact aerators
for certain conditions such as partial purification. Prior
to 193O, brush wood was often used (3148, 5366) instead of
stone for percolating filters at loading rates of 2.75 mgad
(62 gal./ft2/day) and the filters were still operating after
three years. Successful operation of brush wooc* percolating
filters was reported by Lloyd-Davies in South Africa (2745) ,
because the flow could be three times that of a broken stone
filter, and the stone filters would cost almost four times
as much as the brush wood filters. Hamlin described the ap-
plication of the brush wood filters in South Africa (1687) ,
but Francis (1338) critically remarked that brush wood filters
were not as satisfactory as stone filters. Redwood bark was
proposed by Carpenter (622) to be used as a medium and it
was suggested that the specific surface area was as much as
one hundred times greater the specific surface area of con-
ventional media. A residual charge on the redwood fiber
was claimed to enhance better colloidal entrapment and devel-
opment of microbial colonies. Burton (543) proposed redwood
bark as a suitable medium for use in percolating filters which
could subsequently be used as a soil conditioner and fertil-
izer. A patent was issued to Girard (1482) in which wood
shavings were provided between the dosing siphon and the
bacteria bed to provide a higher degree of treatment. Straw
filters were used by Richards and Weekes (3587), and the
straw with the retained organic solids was an excellent soil
builder. The practical application of this process (3587)
was limited to sewage works having access to large quantities
of straw and where the demand for soil conditioner exceeded
the supply.
142
-------
Peat, either directly or with treatment with lime, is another
example of naturally occurring organic material used as a
medium. Schlick (3869) measured the relative efficiencies
of Iowa peat, corn cobs and other inorganic materials for
small plants. Daire (856) and Acklin and Camenisch (11) ob-
served operations using peat for mechanical filtration, as
well as biological filtration medium. Industrial wastes
were treated satisfactorily by filtration through peat, e.g.,
dye works (388O), glycerol wastes (472O), and decontamination
of cyanide waste (526). An objection to the use of peat was
raised by Guth and Keim (1628) because the mechanical stability
of peat blocks failed after four months' operation. Slag was
much to be preferred, unless the price of peat was exception-
ally low. Buttner (57O, 571) described a process which in-
volved the use of peat which was treated with powdered lime-
stone. Successful operation (571) which produced an adequate-
ly treated effluent plus a source of fertilizer was obtained
from a trickling filter with peat media. The upper layer
of peat was removed weekly, mixed with fresh sludge and al-
lowed to rot in piles for two to three months. Patents were
issued to Maier (2836), Meyer (2986), Acklin (12), and Dick-
mann (965) on various arrangements using peat as a trickling
filter.
Anthracite coal is also an example of material in the category
of a naturally occurring organic medium. Turner (4482) stated
that the Lawrence Experiment Station was one of the first
installations to use anthracite, in 19O1, in a filter 12 inches
deep. Anthracite filters were used to handle high solids
waste water or effluent, (3349, 4482), but Kountz (2556) also used
coal as a trickling filter medium. Loadings of 2O mgad (46O
gal./ft2/day) gave a BOD reduction of 4O$ without recirculation
for treating slaughter house waste waters. Mechanical fil-
tration of effluents (3346, 3348, 4243, 5488) utilized an
anthracite medium, the loading rates of which were 2OO to 30O
gal./ft2/hr. Clifford (717) indicated that liquid contact
time of coal versus gravel medium trickling filters favored
the longer detention time of the coal medium, but little
significant removal difference was demonstrated. The sewage
experience of the city of Bradford, Great Britain, which was
reported over many years (244, 4978, 4979, 4980, 4981, 5300)
indicated that the coal medium trickling filters, which had
been built and operated for many years, had developed serious
operational problems due to plugging and that stone medium,
graded gravel and other inorganics would replace the coal.
Man-Made Organic liedia
Plastics for trickling filter media are another area where
materials can be classified into a fourth category, that of
organic material developed by man. Plastics such as poly-
vinyl chloride (98, 643, 678, 815, 1876, 223O, 3O25, 4965,
4970), polystyrene (513, 514, 1772, 497O, 5184, 5193, 5324),
143
-------
and polypropylene (497O) , as well as others, have been pro-
posed and used in the treatment of industrial and domestic
waste waters.
British investigators, e.g., Isaac (224O), Pearson (3328),
and Chipperfield (677), have reported on the advantages of
using plastic media biological trickling filters relative
to other media in waste treatment processes. Smith and Leibee
(4O74) suggested the use of plastic media as a superior al-
ternative to commercial grade rock to provide more specific
surface area, yet still allow adequate ventilation void space.
Some of the plastic media were designed to allow the waste
water to fall in a thin film (5O3O), while others have re-
ported designs which create droplets in more turbulent flow
(154). Plastic media trickling filters as combination cool-
ing towers and biological units were investigated by Burns
and Eckenfelder (538) and found to be only slightly desirable.
While handling wastes which were characterized as having the
tendency to plug normal biological filters, Brown (492) sug-
gested the possibility of improving the ventilation of the
filters by using a plastic medium. Egan and Sandlin (1120)
loaded plastic filter media up to 75O pounds of BOD per 1,OOO
cubic feet per day and removed up to 35O pounds per 1,OOO
cubic feet per day by this treatment. Berridge and Brendish
(286) examined the successful use of plastic media for han-
dling digester supernatant.
The development of new plastic media for trickling filters
generated additional information on design (2955). The in-
creasing popularity of the use of plastic media for indus-
trial waste applications was indicated by recent publications
of an annual review (4948) and book (3O8), and Sak (3811) re-
ported further experimentation to be underway. Biological
industrial waste treatment using plastic media trickling fil-
ters was reported by Lux and Brady (2797) and Stack et al.
(4158) to be an intelligent use of the material where a waste
provides a difficult sloughing biological film or where waste
waters are particularly corrosive.
Industrial waste applications of plastic media have been
heavily reported, e.g., pulp and paper mill effluent treat-
ment (488, 643, 1009, 1O82, 1O83, 1449, 3795, 3812, 4574,
497O), effluent from coke ovens, steel mills, and metal
processing (1936, 3437, 3489), for handling detergent wastes
(815), for handling textile mill effluent (326, 678, 681,
3124, 3808, 3812, 4O94), for the treatment of oil and petro-
chemical plant effluent (516, 1331, 1332, 1740, 2941, 3968),
effluent treatment from food processing plants (643, 983,
2972, 3812, 5324), treatment of meat processing plant effluent
(1417), brewing and distillery effluent treatment (678,
681), and treatment of dairy effluents (1157). The various
loading conditions and advantages and disadvantages of these
media for the industrial waste applications will be reviewed
144
-------
separately under the specific waste in question. The great
world-wide activity in the development of the plastic bio-
logical filtration media was indicated by the patents which
were issued, e.g., Japan (4342), Belgium (2216), Austria
(3481), and Britain (2215). U. S. patents were issued to
several individuals and companies, such as Porter and Kohl
(3438), Hall (1661), Quinn and Franzoso (3495), Sullins and
Self (4265), and Fry (1364).
Another form of organic material which has been developed by
man to be used as a medium for biological trickling filtra-
tion is coke. Coke was used by Matthes (2887), Clark and
Gage (693), and Pollock (34O3) as a percolating filter medi-
um, early in the 1900's and before, and as a mechanical-
biological filter. Loading rates were reported at 1O mgad
(23O gal./ft2/day) on a 24-inch deep filter (34O3), and 5O
to 7O gallons per cubic yard per day (2666, 5476) for deeper
coke medium filters. Biological filtration using coke filters
was described by McCracken (2902) for the development of a
waste treatment facility involved in the water supply at Ames-
bury, Massachusetts. Coke was used in novel applications
such as the sectioned filter described by DeMoll and Liebmann
(927). Ten-centimeter deep coke granules (3 cm in diameter)
were spaced apart by 1O centimeters and produced an acceptable
degree of treatment. Coke used in combination with other
materials, such as clinker, was used by Snook (4O92) to aid
in the development of a nonplugging percolating filter.
Goldthorpe and Nixon (15O8) compared coke to a newly developed
tile medium, but the effluent from the tile, even though it
had a longer detention time, was still inferior to the coke
effluent.
Many industrial wastes have been treated using coke as a
trickling filter medium, e.g., agricultural industry effluents
(971), dairy and milk product factories (5168), phenolic and
high BOD waste from chemical plants (1393, 2583), textile
and tannery waste disposal (21O5, 5565), ammunition plant
waste (473O), and brewery and distillery waste (3O89, 3693,
5O57). As with other media developed by man's ingenuity,
there were several patents issued, such as to Mars (2864,
2865, 2866) and to Duval (1O52).
Additional materials and different shapes of previously
reviewed materials have been used for trickling filter media.
Paper, waterproofed by various methods, was used (1611, 356O,
3561, 4265) for biological trickling filters. Phosphate-
impregnated asbestos papers for biological trickling fil-
tration were patented by Quinn and Franzoso (3495) to provide
added nutrient benefit during the growth of biological orga-
nism. Different shapes and configurations of the medium, such
as disks, were used in laboratory and field operations (355,
926, 99O, 1621, 177O, 1772, 235O, 2464). A further extension
of the work using disks was accomplished by using spheres or
145
-------
inclined planes, either plastic or ceramic or other materials,
which were reported to give varying degrees of operating ef-
ficiency (31, 10O, 154, 364, 781, 2065, 3481, 4O42, 5186).
Free fall media (clear drop flow pattern) were reported lay
Demoll and Liebmann (926) to have some advantage. Free fall
(926) generally was avoided because contact time between
waste and microorganisms was minimum. Honeycomb structures,
arranged to allow the waste to trickle downward (513, 3268,
3495, 356O, 3561), had structural stability and adequate
ventilating properties. Air gap media were reported as sec-
tional devices whereby the media were suspended in layers
one above the other. Multiple layer media were reported by
various authors (522, 526, 1O31, 3348, 4824) to have advan-
tages on nonplugging and mechanical filtering ability.
Screens of various materials and sizes have been used to
support biological growth in the laboratory and in the field,
e.g., polymer foam with screens by Imperial Chemical Indus-
tries (2216), vertically suspended wire screens by Schulze
(3914) and by Green and coworkers (1571), and Reinsch-Wurl
screens with a mat of anthracite (510O). Quite often the
biological trickling filter media were fabricated in the
form of sheets (98, 25O, 513, 1364, 1661, 3155, 3481, 389O,
4096).
Media were devised, as previously noted, to be constructed
in various layers of either different materials, different
sizes, or structural configurations.
Media Evaluations
By way of summary, it was of interest to note the investiga-
tions which were involved with the comparison of one medium
versus another. This type of study was reported extensively
in the literature with various conclusions being drawn.
Investigators, such as Farmer (1239) and Grant, Fulton and
Litton (1534), demonstrated early in this century the aware-
ness of alternative medium testing to provide the most reli-
able system for biological treatment. Materials such as
clinker, gravel, broken pottery, quartzite, granite, and
limestone were tested.
The tempo of investigations on media increased during the
193O's and investigators such as Hurley (5210) found that
the difference between porous and nonporous materials in
filters was evident only for a period of time and then
tended to disappear. Lloyd and Mountfort (521O) only par-
tially agreed with Hurley's comments. They considered that
good clinker gave better results than coke medium, but that
coke had the advantage of uniformity and durability. Gravel
was suggested by Miller (3OO4) as the most economical medium;
however, experiments had shown it to be unsuitable for
conditions similar to those investigated by the previous
three authors.
146
-------
While investigating the Aero-Filtration system, Halvorson
(1667) compared the use of gravel, trap rock, and tile and was
able to demonstrate that only with high rates of recirculation
could media other than tile be used. Work by Rudolfs et al.
(3739) on crushed stone, slag and gravel in various sizes
indicated that: (a) the slag retained more solids than the
other media, (b) the degree of purification achieved by the
two media with similar sizes was about equal, (c) there was
a marked difference between the effect of coarse and fine
materials (an increase in the BOD load caused a poorer ef-
fluent but the percentage reduction was greater), and (d)
an increase in temperature appeared to have a similar effect
as an increase in surface area. Based on considerable in-
vestigation and observation, Stanley (4185) concluded that
excessive clogging was involved in the use of clinker, coke
and coal. The more durable materials such as crushed trap
rock, quartzite and certain granites, as well as crushed slag,
when available, and limestone, if it passed the hardness test,
were preferred.
Filters constructed from sectional lath, cinder or clinker,
quartzite, gravel, ceramic rings, broken tile, and corncobs
were studied by Levine and coworkers (27OO) for the purifica-
tion of milk wastes. They observed that: (a) the cinders
proved very efficient but clogged after five months' opera-
tion; (b) a sectional filter with air gap layers was less ef-
fective than a continuous solid filter; (c) the lath filter
was inferior to the cinder filter; (d) the ceramic rings
produced results inferior to those obtained with gravel; (e)
the tile filter was very efficient for reducing BOD, but
clogged badly; and (f) the corncob filter gave high removals,
but the bed depth shrank 35$. Rudolfs and Chamberlin (3728)
investigated crushed stone, slag, gravel, and one-half inch
galvanized wire mesh laid at one-half foot intervals. The
effectiveness of the medium was measured by the degree of nitri-
fication and the wire filter was about three-quarters as ef-
fective as the others. Schaetzle (3855) reported that lime-
stone, slag, and trap rock were equally efficient as trick-
ling filter media, if the size of the pieces of each medium
was not less than one inch nominal.
In Great Britain, Hawkes and Jenkins (182O) , Wilkinson (4733),
and Truesdale et al. (4463) measured the effectiveness of
crushed granite, crushed gravel, and river gravel in various
sizes, as well as slag and clinker in large scale installations,
Data were reported by Truesdale et al. (4463) on different
sizes and textures of media with specific surface areas of
28, 32 and 35 ft2/ft3. After 12 months' operation no sig-
nificant difference in BOD removal was observed. The effec-
tiveness of a polystyrene plastic medium was reported for
application as a high-rate roughing filter (4463) and was
compared with 2.5-inch rounded gravel. The effluent from
the plastic medium was about 25$ higher than the effluent
147
-------
from the gravel medium, but after six months' operation, the
two effluents approached the same strength. The film growth
on the surface of the plastic media (4463) never approached
the high development exhibited on the rock media.
Investigations by Kalabina et al. (2365) were conducted on
quartzite, gravel, limestone, coke, anthracite, and boiler
slag. it was shown that the oxidative process took place on
all filters at approximately the same rate. There was con-
siderable loss of the biofilm from the limestone, quartzite,
and coke filters, which was interpreted as an advantage since
plugging would be minimized.
Plastic media were the subject of several investigations,
such as by Egan and Sandlin (112O) , the Water Pollution Re-
search Laboratory, Great Britain (5193), Chipperfield (678),
and Germain (146O). The normal observation was that compared
to inorganic media higher hydraulic loads were used (4841)
and higher organic loads were effectively removed (112O).
Industrial waste applications for plastic media were strongly
recommended (678), especially as roughing filters, for which
mosi^ plastic media have been designed. The Water Pollution
Research Laboratory (5196) reported on the various character-
istics such as bulk density, specific surface and proportion
of voids for commercially available plastic media, as well
as conventional two-inch rock (Table 3). it was strongly
emphasized (5196) that plastic packings were designed pri-
marily for partial treatment of wastes. High hydraulic and
organic loadings could be used on plastics which were not
particularly suitable for the treatment of effluents to a
high standard. The more conventional random pack media
were , superior (5196) for establishing high effluent quality.
Critique
Af1;er many pages of referenced comments dealing with construc-
tion materials used as media for biological trickling filters
rand a documented summary-evaluation, there is little left to
say. Generally speaking, the quality of papers was high and,
with the exception of a few, common conclusions may be ex-
trapolated.
It would appear that detention time, i.e., specific surface
area, has been given a primary role in determining the ef-
ficfency of a biological trickling filter. The significance
of the micro-surface characteristic is, at best, only minimal.
Media which possess suitable physico-chemical properties, in
addition to providing a biosurface, are helpful for wastes
which are difficult to degrade. The properties of strength,
durability, and chemical inertness are commonly accepted and
required by all investigators, except those wishing to use
the medium for subsequent purposes.
148
-------
Table 3
Characteristics of Types of Filter Media (5196)
Specific
Composition of Bulk Density Surface Proportion
Medium kg/itf* Ib/ft^ mVm-* ftVftJ of Voids, c,
Polystyrene
sheets 64 4 82 25 94
Polystyrene
sheets (close-
packed) 48 3 187 56 94
Polyvinyl
chloride
sheets 37 2-3 85 26 98
Polyvinyl
chloride
tubes 8O 5 2 2O 66 94
Rock, 2 in.
diameter 13 5O 85 1O5 31.5 ca 5O
Based on this review, it seems evident that a wide variety
of material can be used as the support and have satisfactory
performance, provided the guides for design of an ideal
medium are sought. The basis of choice of a medium should
be economics and an evaluation of the properties of the types
available.
Materials for Underdrains
Information dealing with the design, construction, and mate-
rials for the underdrainage of biological trickling filters
has been reported in a series of articles by the Public Works
Journal (35O1, 5378, 5382), by the Trickling Filter Floor
Institute (5574), by Mahlie (2824), Escritt (1191), and others
(12O1). Underdrainage systems (5382) must support the medium
and the drains be designed so that there is an uninterrupted
straight line following the floor slope to a central drain.
Dickinson (958) and Watson (463O) recommended the use of tile
underdrains laid on an impervious floor (concrete) with open
joints between the tile. Mieder (2994) reported on a situation
where a filter with no underdrainage system, other than large
stones, had crumbled; the floor levels were very uneven,
149
-------
creating an undesirable drainage-ventilation situation. A
large number of reports were published on the use of spe-
cially designed tiles for underdrain systems, which were
carefully placed to provide the desired drainage (426, 1529,
38O4, 4O15, 4628, 5458) . Inverted channels in the concrete
floor were reported (5125) to be arranged in parallel rows
connected with larger channels at right angles to provide
adequate drainage for large trickling filter beds. Rein-
forced concrete was used by Walker et al. (4579) to provide
underdrains and a floor for an experimental trickling filter
in Pennsylvania. Underdrainage systems of vitrified paving
brick were constructed by Wagner (4568) and were quite
similar to previous references of various forms of tile
underdrains. Another example of tile usage was that noted
by Rhode (3582) in which the filter was underdrained with
perforated clay pipes. Small purification plants for house-
hold drainage were described by Hepburn and Drosten (1888) .
They suggested that the filters be constructed of timber and
inserted in excavations of porous dune limestone. Chipper-
field (678) noted that with the development of lightweight
plastic media heavy supporting underdrain structures would
not be required. The Handbook of Trickling Filter Design,
sponsored by the Trickling Filter Floor Institute in co-
operation with Public Works Magazine, illustrated the use of
various types and the specifications of tile underdrain
systems, such as that outlined by the American Society of
Testing Materials, Section C, No. 159 (5574). Typical of
the many patents issued were those to Whitacre (47O2),
Harbour (169, 17O), and Levine (2694) for modified underdrain
systems.
Critique
Supporting structures made of strong durable materials were
reported. Little conflicting evidence was offered on mate-
rials of construction. Designs included the concept of rapid
drainage, high ventilation, media placement on the drains,
and other factors of construction. Only slight mention was
noted of underdrain systems and materials being designed with
cleaning ease in mind. The development of lightweight media
has placed materials such as noncorrodible metal gratings and
others into an area previously occupied by inorganic clay
types. It would not be surprising to see, in the future,
other lightweight, economical materials used to provide the
support, effluent transport, and ventilation required for
the operation of biological trickling filters.
Materials Used For Construction of Enclosing Walls and
Architecture
The design and material selection for the construction of
walls of percolating filters were described by Stockley (4221)
as a matter of some concern. Reinforced concrete was common-
ly used with the walls being designed as ring tension units.
150
-------
Frequent failure of these walls was explained by Bergman
(278) as being due to the daily fluctuation in temperature
which caused expansion during the day and contraction at
night. Investigators such as Modersohn (3O33) have been
proponents of perforated walls surrounding the medium to
provide better aeration and removal of carbon dioxide. Per-
colating filters constructed after 1960 for the City of
Bradford, Great Britain, took advantage of thin, impervious
plastic membrane placed on earth-fill excavation with the
stone trickling filter medium placed over the membrane.
Kratz (2564) is one of several investigators who reported
on the use of rubble masonry as a perforated wall surrounding
the trickling filter medium. Reinforced concrete can be
used (5129) economically for the walls for trickling filters
because other unit operations in the plant require this
design. Steel has been used (302O) as a containment mate-
rial for biological trickling filters and advantages in econ-
omy and speed of construction were discussed. Jenks (2312)
used corrugated metal walls instead of the more usual con-
crete walls on the Biofilters constructed at San Mateo and
other locations in California. Prefabricated storage steel
walls were used (5642) for constructing containments for
percolating filters with advantages of economy and speed
of construction.
Architectural considerations were discussed by Mueller-Neuhaus
during the design and construction of trickling filters (3110) .
As stated previously under the section on enclosures, the
architecture was considered (5645) in developing the total
trickling filter system. In preparing design standards. Mad-
docks (2811) suggested in 194O the importance of architectural
considerations in the design and construction of waste treat-
ment facilities.
Critique
The determining factor in selecting a material for wall con-
tainment is mainly economic. Walls were used to provide a
tank for the filter which could be flooded for the control
of filter flies (Psychoda). However, media have been developed
upon which flies are controlled by hydraulic loading and
other factors. This control removes the limitations of mate-
rials for construction, such as metal, plastic, wood and
others, from which to build walls. Under the sections on
theory and design, the reasons for using walled enclosures
were reviewed.
Architecture is an area which obviously can be improved in
the design of waste treatment plants in the United states.
Plans of plants built in Switzerland usually have definite
architectural influence and the total plant appears as one
modular unit. Interesting design problems arose when
these facilities were expanded while maintaining the
original theme. Unfortunately, architectural considerations
151
-------
were not stressed in most of the design or construction
reports. With more public attention being directed at
waste treatment and its facilities, architectural consid-
erations will undoubtedly be more important.
Materials for Enclosures
Steps to be taken for odor and nuisance control were published
in the "Recommended Standards for Sewage Works"(5O11). Scharfe
(3858) reported on the construction and efficiency of high
rate enclosed percolating^filters. Descriptions by Lundie
(2794) of the use of a Priiss-type filter in South Africa
indicated that by forcing air into the top and through the
filter impurities in the vapor phase were oxidized in the
same manner as those in the liquid phase. Antill (82) de-
scribed the Priiss-type enclosed trickling filter as being
a conical roof with a small fan at the apex for ventilation.
Reference was made in this review to the use of totally
enclosed sewage works, which indicated that the concept of
covering percolating filters was fairly common. Bardwell
(172) described an attractive sewage treatment plant at
Excelsior, Minnesota, which was totally enclosed to control
odors.
Domes have been constructed in Sarasota, Florida,with aluminum
panels hinged to aluminum pipes, in the geodesic design, to
allow dimensional changes under conditions of loading stress
or irregular expansion. The aluminum dome solved the air
pollution problem and eliminated the need for preliminary
chlorination of the sewage (1478, 5115, 5116). In the early
1930's, papers were published (423, 4316) describing the
construction and operation of percolating filters with glass
covers. Eight-ounce translucent fiberglas reinforced panels
were used for covering tanks at the Whittiers Narrows water
reclamation project at Los Angeles (5656). Concrete-covered
domes of spirally generated Styrofoam~* over 8O ft in diameter
were built in 1966 (5645) to reduce fog discharge problems
from filters. Similar 136-ft diameter domes protected the
trickling filters at Elmira, New York, as reported by Dunbar
in 1969 (1044) . Gilde (1472) patented the use of a flexible
cover over a trickling filter which was kept at a positive
pressure to contain the objectionable odors until the sewage
is filtered out by absorption and direct metabolism.
Molke (3O56) and Foster (1312) utilized elliptical concrete
domes (3.5 to 6 inches thick) to protect the trickling fil-
ters from the adverse weather in Hibbing, Minnesota. The
literature (37O4, 4885) in the late '3O's to the mid '40's
indicated that the use of enclosed trickling filters was a
common feature of sewage plants. Quite often the development
of a specialized medium (463, 235O) or issued patents (1212,
*Registered trademark of The Dow Chemical Company,
Midland, Michigan.
152
-------
4O97) involved the use of various materials specifically for
trickling filter covers.
Critique
Covering systems have presented other problems, such as cor-
rosion. New materials should be tried by design engineers
in an attempt to solve these problems. A lower cost cover,
such as the flexible type, may, however, produce higher
operational and maintenance expenses.
153
-------
SECTION X
OPERATION OF TRICKLING FILTERS
A general description on the operation of waste treatment
plants involving biological trickling filters was published
in 1942 (5389), and other publications were so similar that
a detailed review and listing of these reports would be
duplication. Various aspects of operation and operator
activities for trickling filters are reviewed in the fol-
lowing subsections.
OPERATION — KEY TO EFFLUENT QUALITY
Agar has reported (16) the need for proper supervision and
operation of sewage works including daily inspection, a
record of conditions, immediate remedies for breakdowns,
and familiarity of plant personnel with alternative operat-
ing procedures. Wisely (4783) charged that it was the
foremost responsibility of the plant operator to maintain
an efficient and economical operation of the waste treat-
ment plant. Escritt (1194, 1199) stressed the importance
of the plant design which would allow flexible operation
and intelligent management, e.g., series operated trick-
ling filters which could be operated by alternating double
filtration or parallel operation. The operation of small
sewage works was considered by Stanbridge (4165) in which
he suggested short term, manual labor, low expense items
of operation, e.g., control of ponding by forking the sur-
face of percolating filters and by prechlorination of
sewage before filtration. Operation of newly designed
plastic roughing tower waste treatment plant was discussed
by Chipperfield (677).
Operational parameters available for the control of waste
treatment plants were emphasized by Malinowski (285O) as dis-
solved oxygen, settleable and suspended solids, biochemical
oxygen demand, degree of nitrification, pH value, and rela-
tive stability among others. The influence of laboratory
determinations as a means of measuring the effectiveness of
the operation of a waste treatment plant was stated to be
necessary (4859). Lumb (278O) discussed several analytical
techniques useful to the efficient operation of sewage treat-
ment plants. Imhoff (2199) objected to the use of nitrate
as the parameter which would indicate a satisfactory effluent
and suggested operating trickling filters at a high rate to
wash out the accumulated sludge before it has been decomposed.
Biological examination was considered by Kalibina (2367) to
be a rapid means of judging the quality of operation of a
sewage plant. Sima (4O29) found that a count of psychro-
philic bacteria was one of the most suitable criteria indi-
cating the failure of biological filtration plants.
Typical operational problems such as the accumulation of a fun-
gus on an enclosed percolating filter were discussed by Hunter
155
-------
and Cockburn (21O1). Control of the fungus was attempted by
inoculating the filter with Achorutes with promising results.
Siphons to dose percolating filters to ensure that the dis-
tributor arms are not plugged were extensively used through-
out the world. Their operation and problems were outlined
by Grubb (16O7), who noted that overloading of a filter may
be due not only to the small size of the filter but also
unsatisfactory operation of dosing siphons. A severe chal-
lenge for sewage works operation was stated by Killam (2478)
to be the combined effect of variations in the rate of flow
and the concentration of sewage arriving at different hours
of the day. A conservative approach to the design (yet
always satisfactory) was to use the maximum concentration
as the controlling factor to determine capacities, and during
periods of light load the system would be operated as if in
a resting phase. A great deal of literature was published
on the operational techniques to control filter flies, e.g.,
Murray (3138), such as diluting the influent, controlling
temperature, dosing with chloride of lime, dosing with free
chlorine, flooding the filter, enclosing trickling filters,
applying salt to the surface of the filter, painting filter
walls with oil, flaming the filter stones, and interrupting
the flow for seven days. In the discussion of Murray's
paper, the degrees of success, advantages and disadvantage
of the methods were emphasized.
Many articles were published on the operation of various
units, such as contact filters, by Clark and Adams of the
Lawrence Experiment Station in Massachusetts (697) in 1914.
Operation of Imhoff tanks and contact beds was reported by
Fish (1289), indicating the success of the operation on
domestic sewage. Operation of dipping contact filters
while treating mixtures of sewage in trade waste waters
was related by Hartmann (1773). Reports (5648) on the
operation of trickling filters by alternating double fil-
tration indicated less trouble with plugging and filter
flies. Reichle (3536) compared the operation of two-stage
activated sludge and anaerobic treatment with two-stage
trickling filter and anaerobic treatment with respect to
operation and purification obtained. The popularity of the
Bio filter was mentioned in reports on its operation by
Lawrence and Eichenauer (2644) and Divet (975). Problems
of icing were handled (2644) by recirculating sewage to
the secondary filter and operating the primary filter as a
low-rate system. Typical of various locations which had
operating problems and which were solved were Pontiac,
Michigan (34O4), a girls' school in Indiana (95), District
No. 2 of Hamilton County, Ohio (719), trickling filter plants
throughout Illinois (2O81), and operation of waste treatment
plants at hospitals (13O6).
Critique
In most cases, the problems which had developed after a
period of operation were corrected by operator ingenuity and
156
-------
resourcefulness with help from other interested individuals
such as consulting engineers, regulatory agencies, and
others. The operation of today's major metropolitan sewage
treatment plant is a large step for operators in a little
over fifty years. Fountain (1317) in 1915 reported that
the treatment plant serving Calvert, Texas, produced ade-
quate effluent, had no odor 2O feet away, cost $5,OOO to
build, and required one man 45 minutes a day to operate it.
Whereas in 1966, the chemical industry alone has more than
1,7OO people working full time and is spending $40 million
a year to produce satisfactory effluents, according to Sadow
(3809) .
OPERATIONAL PROBLEMS WITH START-UP
Operational problems with the start-up of percolating filter
systems after reconstruction, flow interruption, or new
construction have been reported (355O, 5182, 5422). Beedham
(244) outlined the start-up problems generated when several
trickling filters were taken out of operation during times
of minimum flow and were later restarted which caused the
effluent quality to suffer. The solution to this difficulty
was recirculation of final effluent, but the cost of pumping
would be high. Kleeck (2505) recommended procedures for
start-up of new filters. However, biological investigations
(4396) indicated that when the operation of a filter was
interrupted for seven days the top layer microorganisms
were killed, but the deeper layer microorganisms retained
their viability. Upon restarting, the efficiency of the fil-
ter was expectedly lost, but was reestablished in 1O to 15
days. An operational problem which required considerable
skill was acclimating biological processes to treat trade
wastes, such as reported by Orford (3247) , which enabled a
percolating filter to handle the waste waters from a natural
gum processing plant. Problems encountered during the initial
operation of a plant for the biological treatment of waste
waters from the manufacture of Pharmaceuticals were published
by Pitts (3390) , who showed that continuous return of the
sludge from final sedimentation tanks to primary aeration
tanks aided the operation.
A good example of the type of work on acclimatization was
given by Tucker (4477) who investigated two different methods
of starting up percolating filters. He concluded that it ap-
peared that a preliminary rapid rate of flow enables filters
to reach their maximum efficiency more rapidly than the gradual
increase of the rate of flow, and more satisfactory results
were obtained if a new filter was seeded with partly purified
effluent than if it received settled sewage only. Simons
(4033) described start-up difficulties and their elimination
in the start-up and initial operation of a percolating filter
plant in Orlando, Florida. Treatment of domestic sewage plus
trade wastes from a cotton mill generated odors, which were
masked by treatment of the incoming sewage with Orthosolv,
157
-------
an emulsifiable chlorobenzene (2688). Freeman (1348) pub-
lished, as have many authors, the start-up problems which
were eventually overcome, in this case at Fayetteville,
North Carolina. The acid conditions were prevented by the
addition of lime, the pumping equipment and valves control-
ling the pressure of waste gas were adjusted, and grease
balls formed in the crude distribution box were broken up
with water.
Critique
Start-up problems have been adequately reported and tech-
niques to solve them evolved due to operator experience and
participation in forums and discussions, and most situations
are now well understood. Operation data did imply the lack
of design and construction considerations for certain common
problems in waste treatment plants. However, with improved
communication and more qualified operators being available,
corrections of these design difficulties are being incorpo-
rated into new plant designs.
OPERATOR TRAINING AND RESPONSIBILITIES
Several workers in the sewage treatment field expressed the
necessity for formal training for sewage plant operators
(1122, 45O5). The importance of operator training was rec-
ognized (946) early in the history of waste treatment, as
W. J. Dibdin stated in 1897 in his book, "The Purification
of Sewage and Water."
"if the work is to be trusted to men ignorant of
the first principles of the duty entrusted to
them, failure will be directly invited. Hitherto
the one great difficulty in the way of success is
the persistence of authorities in placing in charge
men who, however intelligent, willing, and honest
they may be, have not had the training required
for that systematic supervision so essential in all
operations conducted upon scientific principles.
Let those in authority ask themselves, 'What would
become of a brewery if entrusted to the management
of a man who knew nothing of brewing? or of a
foundry, or, in fact, of any ordinary business if
placed under the direction of one who knew nothing
of the scheme underlying his work?' The remedy is
obvious."
In an effort to supply some of the technical information,
agencies such as the Iowa State Board of Health (523O) and
the New York State Board of Health (33O2) published opera-
tional rules and guidelines for operators to establish ef-
ficient waste treatment plants. As part of the formal
training/there has been strong interest (4728) in estab-
lishing a suitable licensing law for water and sewage plant
158
-------
operators. The law would provide for the appointment of a
competent board of men with recognized experience in the
field who would issue licenses to qualified personnel. The
board would have the power to force compliance with the law
which would establish protection for the public, municipal
officials, and competent plant operators.
As an aid to the training of plant operators, manuals were
prepared (537, 5337) which dealt with routine instructions
for operating sewage works of various types and a variety of
information on operation, data gathering, process modifica-
tion, and safety, along with other assorted information which
should be available to a competent operator. The U.S. Corps
of Engineers used a model of a bio filtration plant (5109)
to train operators for Army sewage treatment plants. Con-
tributions to the field have been made by several workers,
such as Cosens (793), to provide basic information and data
for operators of sewage works. Methods of measuring flow,
concentration and temperature of the sewage, taking samples
for analysis and making laboratory determinations were dis-
cussed and a description of the sewage plant, including per-
colating filters, was given. Tomlinson (4412) published a
paper dealing with the description of the animal life in
percolating filters which was written intentionally to enable
the nonspecialist to identify the various forms.
The result of much of this training has been enthusiastic
response by the operators, as illustrated by operator forums
(5149) in which various operating problems and the solutions
of these problems were discussed, providing practical opera-
tional methods for the operators. Typical of the operation
papers were those submitted by Brooks and Ross (479), and by
Booth (4O4). Organizations such as the Federation of Sewage
Works Association and the American Society of Civil En-
gineers (2392, 5148) have published papers dealing with
operating and cost records, annual report writing, routine
analysis, supervision levels, operator qualifications, and
analytical determinations necessary for controlled operation
of waste treatment facilities.
In an effort to continue the education of waste treatment
plant operators, an active program of short courses has been
offered by several academic institutions. Examples of the
academic institutions which had active programs of opera-
tor training courses were Louisiana State University (1616,
3632, 4957), the Texas Water and Sewage Works Short School
(dating back to 1918) (2893, 2943, 2969, 4347, 4499, 4858,
5568, 5569, 557O), the University of Wisconsin Sewage Works
Operators School which began in 1945 (561O, 5611), the North
Dakota Water and Sewage Works Conference (2636), the Michigan
Sewage Works Operators Association (1142), the Sewage Plant
Operators of Pennsylvania Conference (2946), and other or-
ganizations overseas, such as the Association of Managers of
Sewage Disposal Works in Great Britain (5463).
159
-------
Critique
The quality of effluent is directly proportional to the
quality of operation and the quality of operation reflects
the operator training. The trend in favor of more formal
operator training has been well established and documented
by descriptive papers. The participation of operators in
publications and other communications indicates their desire
to improve their profession and to establish the waste treat-
ment facility as a point of community or industry pride.
OPERATIONAL COST FACTORS
The review of the literature indicated considerable concern
over operational costs of waste treatment facilities. Gen-
erally, the initial capital investment for trickling filters
has been higher than the initial capital investment for al-
ternative processes* such as activated sludge. However, the
operational expenses for trickling filter plants have been
much lower than for activated-sludge plants. Several in-
vestigators have quoted dollar figures as the cost for
treating a million gallons of waste water and brief reference
was made to selected publications to give the reader an in-
dication of the magnitude of the cost involved. It should,
of course, be kept in mind that for the purpose of comparing
actual cost in terms of present-day dollars and under indi-
vidual circumstances the economics would extend this com-
plex problem beyond the scope of this review.
Examples of operating costs were reported throughout the
literature, e.g., that reported in 191O at Manchester, Eng-
land (4993) . It cost $6O,OOO per year to operate 2O biologi-
cal filters and 4 storm water filters plus $42,OOO a year for
sludge removal. Hoover (2O16) stated that the cost for the
operation of a conventional treatment plant in 1917 was $1.84
per million gallons. Alford and Burdick (62) calculated the
annual operating cost for the Lincoln, Nebraska, waste treat-
ment facility, in 1921, was $12,5OO, treating 4.5 million
gallons per day ($7.6O/miHion gallons of sewage) . Streeter
(4246) arrived at the operation cost of the sewage plant at
Sheffield, England, with a dry weather flow of 18 mgd which
was about one-third industrial wastes and a wet weather flow
of 65 mgd, to be 18 shillings and 2 pence ($2.1O present day)
per million gallons for handling the dry weather flow.
In 1933, Cohn (747) published the operating costs for
the Schenectady Imhoff-trickling filter plant handling 8.93
mgd as $8.33/million gallons of sewage, which was 20$ less
than in 1932. Operational costs on chemical processes were
stated in 1934 by Blew (356) to be from $8 to $2O/million
gallons, exclusive of fixed charges. Rudolfs et al. (3743)
remarked on the cost of chemical treatment in percolating
filter plants and the cost of chemicals for plain sedimenta-
160
-------
tion (at an experimental waste treatment plant in New York
in 1935). The cost for chemicals alone for sedimentation
with filtering was $4.3O to $6.7O more than with the filter
alone for equal removal of non-settleable solids, from $1.30
to $6.6O more for equal removal of total solids, and from
$1.20 to $2.40 more for equal reduction of BOD. The dif-
ferences in cost increased with the degree of purification
obtained.
Also in 1935, Belbin (247) reported operational costs in
Scotland to be 177 pounds, 13 shillings and 10 pence (~$4,25O)
per year to handle the waste from 4,7OO people. Kemmler
(2426) in 1936 calculated the combined cost for operation,
construction, and amortization (4%) over a period of 30 years
for four types of plants: (1) Imhoff - trickling filter plant—
$21 to $23 per million gallon of sewage per day; (2) activated
sludge plant - $2O to $37; (3) chemical precipitation plant—
$15 to $38; and (4) intermittent sand filtration plant— $30
to $46 per million gallons per day. Hosegood (2O40) also
discussed the operating cost of a trickling filter plant in
San Bernardino, California, in which chlorination for nuisance
control was practiced. For a flow of about 2.9 mgd, the
average cost was $10.49 per million gallons of sewage.
Operational costs for a trickling filter plant were stated by
Humphrey (2O87) for waste treatment at a New York State can-
nery to be $O.5O per thousand gallons when 5O million gallons
of waste waters were treated during 1939 ($5OO/million gallons)
In 1940, Reece (3529) treated dairy wastes on trickling fil-
ters with chlorination and lime treatment with the cost of
chemicals being $3.7O per million gallons. Babbitt and Bau-
man (112) published the data obtained by BesselieVre in 1945
which indicated the operational costs listed in Table 4.
Table k
9,
Treatment Operational Costs of Sewage Plants
Cost of
Operation
Plant Location Type of Plant $/million gal.
Rockville Center, N.Y. Activated sludge 59.68
Gary, Ind. Activated sludge 8.79
Galesburg, 111. Trickling filter 15-9^
Elgin, 111. Sedimentation and separate sludge 13.79
digestion with trickling filters
Waukegan, 111. Chemical precipitation 19.22
Fort Dodge, Iowa Sedimentation and separate sludge 27.9^
digestion with trickling filters
Urbana-Champaign, 111. Imhoff tank-trickling filter 16.11
Compiled in 191*5.
161
-------
The cost of industrial waste treatment was reported by Hub-
bard (2075) in England for chemical plant waste waters to be
3 shillings ($O.36) per thousand gallons of crude waste on a
treatment system of 100,OOO gallons per day. In treating
the effluent from pesticide manufacture on an alternating
double filtration plant with pretreatment, Sharp and Lambden
(3971) estimated the cost at one British pound ($2.40) per
thousand gallons including labor.
Another method of reporting costs was used by Smith (4084) .
The capital cost per pound BOD removed per day, using chemi-
cals and plastic medium treatment, was estimated from $15 to
$11.2O for a BOD removal of 2,5OO pounds to 2O,OOO pounds
per day, and the power costs (at Iff KWhr) for the various
units were estimated at 5.2^/lOO Ib BOD removed per day for
a sewage flow from O.42 to 5.O3 mgd. Kempton and Roskopf
(2428) tabulated the costs for a small capital investment
which reduced operational costs by taking advantage of local
conditions and the requirements of the waste stream to be
treated. Relative savings in operation and capital expendi-
ture were calculated by Chipperfield (678), in 1967, for
several industrial wastes, but because of commercial interest
absolute figures were not published. Several investigators
have proposed (5643) various formulas to establish charges
based on sewer usage characteristics.
Critique
In review of operational costs, the literature reflected an
interesting trend. As the price of waste treatment increased,
due to rising costs of labor and material, figures on
the cost of treatment per million gallon were not as readily
available in the 196O's as they were in the 193O's. Texts,
such as those by Babbitt and Baumann (112) and Metcalf and
Eddy (2984), made an effort to include cost and economics
along with other design criteria. Operational costs are
dependent on so many variables that a common basis of com-
parison is difficult. However, figures such as the cost of
the removal of a pound of BOD or of some other pollutional
characteristics certainly appear to be a step in the right
direction. The descriptive terminology required in develop-
ing effluent standards should be made considering the re-
quirements for operational cost accounting. For example,
the use of a percent removal figure by a regulatory agency
really does not guarantee the receiving body of water suf-
ficient pollution protection. Similarly, the cost per mil-
lion gallons is on a basis which would allow less expenditure
for less treatment. A vote of confidence should be given
to the operators of the many waste treatment plants who set
the example of getting the greatest degree of treatment for
the least cost.
162
-------
SECTION XI
MAINTENANCE OF BIOLOGICAL TRICKLING FILTERS
Papers such as that by Hamlin (1695) on factors of mainte-
nance of waste treatment plants to achieve efficient operation
have been frequently published. During operation orientation,
maintenance of equipment is an item that is stressed, as re-
ported in the published literature (5389). The importance of
proper maintenance in small sewage works was emphasized by
Escritt (12OO). Lack of maintenance and the normal wear on
equipment have been the cause of reduced efficiency in waste
treatment plants (5OO).
CORROSION PROBLEMS
Concern over corrosion of metal, wood, and concrete structures
(5O4) and serious damage to concrete and pumps (2598) has been
expressed by waste treatment plant operators for several years.
Much of the corrosion was attributed to hydrogen sulfide
(4143), and it was noted that the presence of zinc in copper
alloys tended to reduce the corrosion by hydrogen sulfide.
The literature (4291) reflected the concern of workers on the
operation and maintenance of pumps, electrical equipment, the
percolating filters, and other potential trouble spots in the
waste treatment facility. Alternative materials such as
plastics have been suggested by Lux and Brady (2797) for
handling particularly corrosive trade waste waters. Hazlip
(1843) described a classic case of corrosion, in Louisiana,
where the poor condition of the sewer pipes caused delay in
delivering the sewage to the treatment plant, and,consequently,
the sewage was frequently septic when it reached the plant,
thereby generating nuisance problems and increasing mainte-
nance cost. During the operation of the waste treatment
facility, modifications have been incorporated into the main-
tenance program such as changing the direction of the jets on
distributor arms (467, 7O4) . Selection of the trickling
filter distributor should be based on low maintenance require-
ments, as indicated by Baxter (224) . Because of the lower
required maintenance, traveling distributors were more common
in England, in 1924, than fixed nozzles (18O7). Quite often
the operator was faced with the maintenance of aging equip-
ment, as was the case at the Epsom and Ewell sewage works
(417O),but the plant must be operated with the equipment
at hand. In rural areas where several small waste treatment
facilities were located, Shimnin, in a discussion of a paper
by Easdale (1O61), advocated the use of roving crews which
would have the mobility to maintain several facilities in an
effective way. Many examples are available of waste treatment
plants which have been operated without proper maintenance
(3455, 5O77), and then deteriorated in efficiency, effluent
quality, and subsequently required additional attention and
expenditures.
163
-------
SYSTEM CLEANING
The cleanliness of the plant has been reported (38OO, 5447)
to be a primary factor in the reduction of general plant
nuisances and operational problems. Procedures, such as out-
lined by Chase (655), which were no more than simply clean-
ing tanks every six weeks to four months, are prevalent
throughout the literature. Examples of cleaning procedures
for the maintenance of biological trickling filter plants
are the cleaning of screens (5315), maintaining and clean-
ing siphon chambers and filter media and distributors (25O5,
4951), daily cleaning of multiple nozzles (1781) and main-
taining sprinkler filter and feed pipes (3337), as well as
many others. Chlorine chemicals, to open clogged pipelines
for flow, have been used by Rideal (36O2). During operator
discussions (2636), methods of cleaning included wire brush-
ing, sand blasting, and sanding. However, it may be cheaper
to replace metal parts rather than to keep corrosion under
control in the waste treatment atmosphere. Occasional cases
were pointed out (539O) where waste treatment unit processes
were completely out of balance and it was necessary to flush
out the plant before normal operation could be resumed. One
method of cleaning a filter medium such as coke, patented
by Mars (2866), forced hot gases through the filter, leaving
only a coating of fine carbon on the coarse coke. Several
papers (1948, 566O) dealt with the removal and cleaning of
trickling filter media by screening and washing. Depending
on the type of medium and strength of waste, cleaning could
be required from every five months for slag (5373) to every
five or six years for stone or cinders (2O78, 2678).
Biological cleaning methods have been used by Carlson and
Cornell (618), which involved an enzyme preparation to clean
the nozzles of the filter clogged with grease and to remove
grease and scum from walls and other surfaces. Bell (253)
recommended the colonization of Anchorutes viaticus to keep
trickling filters from building thick film layers. Others
(1113) have reported on the importance of Enchytraeid worms
for keeping trickling filters clean during operation.
Several physical methods of relieving clogged sprinkling
filters were recorded (5134) as: (a) resting the beds and
washing out the dry accumulated matter by normal distribu-
tion, (b) washing the filter with water from a fire hose,
(c) applying bleaching powder on the surface of the filter,
and (d) application of a strong disinfectant or bleach during
normal distribution. Many investigators (596, 597, 1949,
2706, 3896, 5131} have used backwash devices for unclogging
the trickling filter medium, with and without compressed air,
to aid the hydraulic action and with and without mechanical
handling of the medium.
164
-------
IDENTIFICATION OF MAINTENANCE PROBLEMS
In addition to the above information on maintenance, much
work was done on identifying maintenance problems at perco-
lating filter plants, such as the effect of synthetic deter-
gents producing foaming and ponding as well as blockage of
distributors (783, 4566). Grease from domestic as well as
waste waters caused the plugging of the filter medium,
blockage of distributors, and reducing the biochemical ef-
fectiveness (1887). Procedures which have been used to
correct grease build-up have been published, e.g., use of
alternating double filtration, high-rate recirculation,
inoculation of the medium with an active worm population
(1887), cultivating the top six to eight inches of the fil-
ter medium (33O, 467), the addition of enzymes (789), and
other techniques (2825, 4887). Maintenance problems were
caused by severe weather conditions (52, 711), and freezing
was minimized or eliminated by arrangements of recirculation,
raking the surface of the bed, chlorination, and enclosure
of the percolating filter.
Percolating filter ponding was discussed many times and was
identified, typically, by Lanphear (2628) and Wilkinson (473O).
Methods of prevention and cure are well established (5377).
Foaming (357, 2386, 2437) produced objectionable conditions
and safety hazards, and required additional treatment, e.g.,
the use of chlorine, antifoam baffles, or design modifications.
s
A major maintenance item, as indicated by the literature,
was the control of filter flies (Psychoda alternata). Kemper
(2427) stated that chemicals such as borax, lime, copper
sulfate, sulfate of iron, pyrethrum, nicotine, chlorine,
and carbon disulfide are useless to kill the larvae and
pupae of the fly, since these substances also damage the
organisms necessary for the biological treatment of sewage.
Other methods were recommended, such as flooding the filter
with water, surrounding it with wicker work or green plants,
colonizing the filter with Achorutes viaticus, spraying the
surface with a mixture of petroleum and pyrethrum, treating
the sewage with an emulsion of ortho-dichlorobenzene or of
creosote and crude paraffin oil, but each of these has certain
disadvantages and is not totally effective. Dekema (914) noted
in a patent that nuisance from flies or odors was prevented by
adding insecticides, disinfectants, or deodorants with the air
stream. To kill the fly in the larva stage was part of the
disadvantages of chemical addition noted by Riedmuller (36O9)
because the Psychoda are necessary in lightly loaded filters
for the destruction of solid matter. Gammexane (benzene hexa-
chloride) has been applied (4421) to control the breeding of
the filter fly, but it was cautioned that drastic treatment
could cause ponding and decrease the efficiency of the filter.
A discussion among plant operators (5229) indicated no particular
165
-------
advantage to any one of the suggested procedures for fly con-
trol. Hypochlorite was used (4884) to control filter flies.
In North Carolina, flies were controlled (2853) by spraying
the filters with Malrin and the few adult flies surviving were
controlled by a portable fogging machine. Sedimentation tanks
were covered with a film of fuel oil to prevent larvae from
hatching. A severe problem of active breeding of stable flies
(Stomoxys calcitrux L^.) arose in Greenville, South Carolina
(317O), due to a meat packing waste in high concentration in
the influent sewage. Satisfactory control of the fly breeding
was established by submerging the filters for 12 hours twice a
week. Waller and Ingols (4587) stated that filter flies were
controlled in Atlanta, Georgia, by flooding the filters at
weekly intervals; however, a serious odor problem was created
and chemical addition was required for odor control.
Ponding of biofilters caused by a profuse growth of the sul-
fur bacterium, Beggiatoa alba, was reported by Shepherd
(3989). Flooding the filter with chlorinated water for 48
hours killed the growth. However, flushing out the dead
growth was difficult. When the detention time of the
flooded filter was long enough to allow anaerobic digestion
to occur, the filters were flushed repeatedly, solving the
dead growth problem and resuming normal filter operation.
Flooding of filters for 24 hours every four days was reported
by Rhame (3577) to completely eliminate flies, but seriously
impaired removal of BOD. Later reduction of the flooding
time to 2-3 hours every four days considerably improved the
BOD removal, but did not eliminate the flies. Fuller (1375)
completely eliminated filter flies and ponding by flooding
the filter 24 hours once a month, and Browne (501) at Marion,
Ohio, controlled filter flies by flooding the trickling
filters.
Cohn (738) used chlorine to aid in the maintenance of trick-
ling filter plants, e.g., reduce the odors, the biological
growth in nozzles and distributing pipes, and the number of
Psychoda flies, and increase the amount of suspended and
colloidal solids in the effluent. In addition to the above
uses, Chamberlin (648) indicated that chlorine was effective
in the removal of grease, reduction of BOD, improvement of
clarification, foaming, and the control of ponding on per-
colating filters. Lohmeyer (2755) utilized chlorine to
eliminate distributor stoppage due to snails (Physa) which
were collecting in the nozzles. Chlorine was very effective
in correcting trickling filter ponding, according to Agar
(18). Prechlorination of an odoriferous waste (19, 1352,
1896, 43O4, 4667, 5O6O) controlled the release of odors from
the trickling filter and the flies, and retarded ponding
tendencies. Cohn (743) stated that chlorination did reduce
ponding in trickling filters, but was not a very effective
agent in controlling filter flies.
166
-------
NUISANCE CONTROL BY ENCLOSURES
Several general papers (2671, 5497) outlined the advantages
of enclosing nuisance-causing unit operations in waste treat-
ment plants. Blunk (372), in reply to criticism of another
of his papers, described the effectiveness of nuisance control
by enclosure in the late '3O's. Ponninger (34O7, 3417, 3421)
published much work dealing with control of the Psychgda fly
and odors on enclosed trickling filters in Germany. Ponninger
worked with sewages of various concentrations and septicities
and developed experience in ventilating enclosed trickling
filters. Imhoff (2197) expressed the opinion that, if,through
faults in design or operation of the plant,odors cannot be
completely controlled by chemical means, then the whole or
at least part of the plant may be covered and ventilated, with
the air treated with activated carbon or ozone. Not only was
there concern for nuisances, but also for safety and the quality
of the local waste treatment plant environment. Muller (3119)
briefly summarized the toxicity, odor generation and corrosive
effect of hydrogen sulfide in sewage treatment works.
The nuisance conditions experienced around treatment plant
works may be similar to that published by Planchon (3393)
in treating the waste from abattoirs in France. The heavy
organic load and the small plant usually available dictated
that it was best treated on a closed aerated percolating
filter to diminish odor and fly nuisance. Likewise, others
have reported (1618, 17OO, 525O) from the early 19OO's to
the present that enclosure of the percolating filters and
forced ventilation were a reasonable approach. The covering
of waste treatment facilities minimized nuisance problems in
golf club (2341) and housing areas (172). Institutions such
as hospitals have employed covered sprinkling filters (2619)
to solve a nuisance and hygiene problem.
Covered sprinkling filters have been built during the period
191O to 192O to control nuisances (65O, 1437, 155O, 5138).
Again, in the late '3O's (3397), industries such as those based
on milk, found the use of covered trickling filters desir-
able. It was frequently published (1258, 2569, 3912) in the
'5O's that enclosed percolating filters were being used in
Germany, France, and Austria. Quite often the population
figure generating the sewage treated by the percolating fil-
ter plant was in the neighborhood of 6,OOO. This type of
loading usually required two covered percolating filters,
e.g., at Bad VSslau (2569). In an effort to combat the nui-
sances of odor and insects, other processes have been tried,
e.g., the anaerobic system described by Dewes (942) . Equip-
ment was also developed and patented such as that disclosed
by Blunk (369).
167
-------
Fly Control
Conflicting recommendations and operational experiences were
published by Ponninger (3414) and Dahlem (853) . Ponninger,
with enclosed aerated trickling filters, found that within
four to five weeks the filters were free from sludge and the
upper layers had a large population of larvae and pupae of
Psvchoda. Dahlem stated that it was not necessary to enclose
the filters completely to prevent odors, the troubles in cold
weather, or the egress of flies in summer. The escape of
flies can be prevented by enclosing the sides of the filter
and by ensuring good distribution of liquid over the surface.
Husmann (2145) agreed with Po'nninger that the problem of the
control of filter flies could be best solved by the complete
enclosure of the filters. The flies can thus be confined
within the filter without causing a nuisance. The larvae
may be present in sufficient numbers to prevent accumulation
of thick biofilm on the medium. Blunk (375) also supported
the practice of enclosed filters to avoid trouble from flies.
The flies which emerge from the top are trapped by the cover
and destroyed by spiders. In the late 195O's, Kleeck (25O6)
discussed various methods of control of the filter flies.
One of the most positive methods was to cover the percolating
filter to prevent the laying of eggs.
Odor Control
Around the turn of the century, many communities, such as Mt.
Vernon, New York; College Hill, Ohio; Chicago, Illinois; and
many more, were aware of the problems of odor control (1721,
5130, 53O8). The trickling filters were covered to prevent
odors reaching nearby residences. Whether a filter was covered
or not was often based on availability of funds and economics.
In 1935, Greeley (1557) described the progress of odor con-
trol by covered structures. In this country, as well as
overseas, the control of odor by various means including en-
closure was practiced. A limitation of enclosures was ob-
served by Le Lan (2678) . In his opinion, the media of all
trickling filters should be renewed every five to six years,
a difficult operation with enclosed filters. Several tech-
niques of odor control were available, according to Blunk
(373), but methods other than enclosure were expensive.
Large metropolitan areas, such as Berlin (418) , were con-
cerned with methods of preventing odor and also the danger
of corrosion caused by formation of sulfuric acid in covered
plants already in use. Mount Vernon, New York, used covered
bed and ventilation systems passing the foul odor through
iron oxide for treatment (513O).
About 194O, Hurley (2115) reviewed the recommendations of
the Royal Commission on sewage disposal (England), and con-
sidered the problems and methods of odor control, which in-
168
-------
eluded covering sedimentation tanks and using enclosed
aerated filters. Husmann (2145) suggested that the cause
of odor production in high rate filters was due to the sep-
ticity of the sewage being applied. By passing the air
through the filter in the same direction as the sewage, the
odor nuisance from decomposing organisms and sludge was pre-
vented. Economics became even more important during the
'50's, and low cost, geodesic shaped covers (1478, 5115) were
used to trap the objectionable odors generated by waste treat-
ment plants. A small single-stage enclosed trickling filter
plant (4073) was built at a cost of $3O.6O per capita com-
pared to $25 per capita for open-type plants.
The tempo of odor control programs for waste treatment plants
continued at a high rate during the '6O's. Odor control units
(34O2) were installed in Port Washington, New York, an ozone
generator in Coral Gables, Florida (4395), and digester gas was
biologically treated by passing through the trickling filter
(34O5). Cheap enclosures such as fiberglas panels (5656)
and reinforced concrete (4O8O) were shown to be quite ef-
fective.
Many industrial wastes have characteristic odors. Treatment
plants handling canning wastes have been described by King
(2492) to operate quite well using enclosed filters. Slop
from the acid fermentation of molasses produces an odor nui-
sance and enclosed percolating filters have found some suc-
cess, according to Krige (2571). Work on the treatment of
phenolic effluent was done in England, and Hall (1651) stated
that covered percolating filters in combination with acti-
vated sludge worked well. The successful operation of
covered percolating filters for the sewage from a paraffin
oxidation plant waste was reported by Christ (683) in 1958.
Several systems were developed and patents issued dealing
with odor control. Yonner (4842) effectively applied forced
aeration in covered filters. Odor control patents were is-
sued to several people, e.g., Blunk (37O), Reddie and Griffin
(3523), PrQss and Blunk (3477), Gilde (1472), and Griffin
(1597). These patents dealt with various means of enclosure,
air recirculation, gas treating towers, and other process
equipment.
Icing
Eddy and Vrooman (1O88) in 19O9 concluded that covered fil-
ters were required for complete success in winter operation.
Ogden (322O) remarked that there was no advantage with porous
walls (no side walls) and that the widely used revolving dis-
tributors were subjected to ice interference. Pritzkow (3465)
stated that in an extremely large waste treatment plant icing
was not a problem on the sprinkling filters even with tempera-
tures as low as -16° in which the effluent temperature reaches
169
-------
2°*. This statement was corroborated by Bantrell (167); he
reported ice sometimes four feet thick on sewage filters at
Rochester, New York. However, Walker (4579) observed in
193O that enclosing the percolating filter gave slightly
better oxidation, more efficient removal of bacteria, and
less icing problems than open filters.
The advantages of enclosing the filters for winter operation
were described by Cleland (711) in 1933. He stated that
troubles caused by freezing of pipes and nozzles, especially
during low night flow, could be eliminated by covering the
trickling filters, resulting in a general increase in effi-
ciency and less winter maintenance problems. Very poor re-
sults were obtained in winter operation at Cedar Rapids when
ice reduced the ventilation of the filter beds (2918) . The
construction of elliptical concrete domes to cover the trick-
ling filters, as reported by Molke (3O56), protected them
against the severe weather in Hibbing, Minnesota. During a
comparison test of open versus enclosed filters in South
Africa, Vosloo and O'Reilly (456O) observed that when the
filters are overloaded, especially in cold weather, the open
filters showed signs of ponding while the enclosed filters
did not pond; however, the quality of the effluent deteriorated.
Improved results found by other workers with enclosed aerated
filters could be due to the greater depth of medium employed
by them. Hurley (2125) in 1945 attributed the reason for the
decrease in efficiency of percolating filters in cold weather
compared to warm weather to the loss of scouring organisms
for the surface layers. A paper was given by Taylor (4324)
on the operation of percolating filters in Canada under
severe weather conditions, with atmospheric temperatures
3O° to 5O°F below zero being recorded. He concluded that
open percolating filters were unsuitable for districts where
the winters were cold, but covered filters dosed either in-
termittently or continuously operated satisfactorily.
In locations where the average diurnal temperature is below
freezing, the air in an enclosed percolating filter is main-
tained at 35°F (45O9) . Locations where the flow may be in-
terrupted for periods of weeks or months, such as schools,
have been operated quite successfully as closed filters
(3591) . Operation in Saskatchewan of open and closed per-
colating filters was described by Allen (52) , who noted
that it was almost impossible to operate the open filter in
January and February. A concrete and fiberglas enclosure
(2383) prevented icing of the filter even during periods of
sub-zero temperature in Iowa.
Critique
Maintenance of biological trickling filter plants is required
just as it would be on any mechanical or chemical process
*This is a 1911 reference which does not specify whether the
temperature is in degrees Centigrade or Fahrenheit. It ap-
pears likely that it is in degrees Centigrade.
170
-------
equipment. The literature reflects the awareness of workers
in the field that maintenance is an important aspect. Fur-
ther, it was observed that the usual techniques are not
adequate for the development of new material and new oper-
ating procedures. Lack of awareness by taxpayers of the
significance of maintenance of a waste treatment facility
is apparent.
The isolated one-unit waste treatment plant has been replaced
by a multi-unit, large capital investment facility in which
the taxpayers or the corporation have a prime interest. It
is obvious, from this review of the literature, that operator
training will increase to a higher level and that money will
have to be made available for maintenance of these large
capital investment facilities. Careful attention should be
paid by the designing engineer to the new materials which
have been developed and new processes which have been tried
with the objective of a minimum maintenance system. However,
the capital expenditure with the newer materials and processes
should not exceed the present operating processes by any
great extent, or make the new process uneconomical.
The nuisance control-aesthetic point of view has been a very
strong consideration in covering waste treatment facilities,
especially trickling filters. As the need for a more con-
trolled environment increases with the demands of urban
development, the time-tested approach will be used at a greater
rate. The capital expenditure for the structure versus opera-
tional costs of other methods is very suitable for bond issue -
tax-based applications. The little or no maintenance re-
quired by covers provides the treatment plant operator with
additional time to concentrate on other solids-liquid treat-
ment problems.
The filter fly, being a fragile organism, does not normally
stray great distances from its breeding ground. However,
concentrated swarms are produced which may be dispersed by
wind currents. Enclosing the filter controls the dispersal
of the filter fly. High organic wastes produce dense popula-
tions of flies and sludge growths. The beneficial effect of
the fly larvae in loosening the sludge is maintained by using
enclosures.
The use of covers to prevent icing has been shown to be of
advantage. Plants may be operated under iced conditions with
a great deal of effort and a loss of efficiency. The geo-
graphical location, i.e., temperature extremes, is a major
factor in determining the necessity of covering the waste
treatment system. Heating the ventilating air must be estab-
lished for specific locations where low temperatures are
common, but generally is not required.
171
-------
The odor removal is partially solved by enclosure. Forced
ventilation helps if it is in the direction of the liquid
flow. Positive treatment techniques of sorption and/or oxi-
dation of the odoriferous gas have been promising. The
capital investment is higher for a covered system, but the
net effect is a more complete waste treatment system. Re-
moving the pollutant from the water and dispersing it into
the air is only an intermediate solution and is not accept-
able in the true environment management scheme of waste
treatment.
172
-------
SECTION XII
PERFORMANCE SUMMARY OF TRICKLING FILTERS
Performance of a sewage plant has been of major importance in
the literature reviewed. For example, the British Royal Com-
mission on Sewage Disposal, which was appointed in 1898, pub-
lished conclusions and recommendations involving all phases
of waste treatment (3983). The Commission reported in 19O4
on the efficiency of the percolating filter relative to depth
of bed and evaluated the performance of the filter versus a
contact bed. Early efforts in this country (4326) and abroad
(336O) were directed at evaluating the performance of various
modified trickling filters and these efforts have been con-
tinued. Harris (1747), in 1925, compared treatment on per-
colating filters with that by activated sludge. During the
construction of new plants, such as at Milwaukee (444O),
alternative processes were pilot tested and the most reliable
one chosen.
In the 192O's and 193O's, performance evaluations were direct-
ed towards the treatment of industrial waste effluent (665,
3466, 5231) and into some of the mechanics of biological puri-
fication (564, 3411). In an effort to evaluate the overall
waste treatment picture, compilations were made listing the
number of communities and the type of plants which were con-
structed or being planned (51O4).
Throughout the 194O's many facets of industrial waste research,
development, operation and management were studied (27O4, 3344,
3764). Also, experimental plants (3689) were operated to
evaluate the performance of new systems (3159) and the ef-
fect of process modifications (4645). Operators continued
to have forums (5149) to discuss the evaluation of the per-
formance of their plants.
In the 195O's work was being continued in evaluating the per-
formance of trickling filters in handling industrial waste
(3773, 4646). New developments in media for percolating fil-
ters were included (5568), along with evaluations of newly
constructed waste treatment plants (1788, 4654). Process
modifications, such as recirculation, were evaluated (267)
and the literature was reviewed (363).
The 196O's were typified by further exploration of industrial
waste treatment (318, 11O2), process modification (1237, 1773),
and additional descriptive information for a mathematical
model of trickling filter theory (3917). More than 65 papers
dealing with various aspects of mathematics applied to trick-
ling filters appeared in the 1960's, 46 papers in the 1950's,
11 in the 194O's, 8 in the 193O's, 2 in the 192O's, 1 in the
19lO's, and 4 in the 19OO's.
173
-------
MATHEMATICS USED FOR DESIGN AND PERFORMANCE EVALUATION
Design equations used to predict the performance of biologi-
cal trickling filters began humbly with simple calculations,
e.g., calculation of sprinkling distribution coefficients
(139O, 2689, 3359, 477O) and calculation of mean contact time
(4297). Mathematical deductions were made by Tatham (4296)
in 1921 on the percentage purification obtained during a mean
contact time which involved an additional variable called the
"avidity constant." This constant, a numerical measure of
the activity of the biological oxidation, was dependent on
the type of waste being treated and the conditions for oxi-
dation. A review of biological treatment dating back to 1671,
which was written by Dibdin (946) in 191O, did not indicate
any particular formalized mathematical expression for the
prediction of percolating filter efficiencies. Investigators
such as Imhoff (2184), Ponninger (3412), Escritt (1191), and
Gaultier (1432) applied mathematical procedures for deter-
mining medium size, depth of bed, hydraulic load and cross-
sectional area of filter beds, oxygen consumption values, con-
centration of sewage expressed by McGowan'j§ formula, and air
requirements. A booklet was prepared by Popel (22O2) which
outlined operations and calculations in connection with trick-
ling filters.
Velz (4535) proposed that the performance of the percolating
-pt —kD
filter can be expressed by the equation: _D = 10 , where
L
L represented the total removable fraction of BOD, D was depth,
and L_ represented the quantity of removable BOD remaining
at depth D. The velocity constant, k, was at temperature t°.
Stack (4157) derived equations for the performance of percolat-
ing filters operated with and without recirculation. His
equations were based on the assumption that a percolating fil-
ter was a self-regenerating absorption tower or that each
unit depth of the filter will remove a constant fraction of
the BOD applied to that unit depth. Rowland (2O63) proposed
a modified equation, along these lines, which involved a tem-
perature coefficient which was recommended to take into account
the fundamental temperature effect on the biological process.
Sorrels and Zeller (4111) found a hyperbolic relation between
the applied BOD and that removed by primary filters in a two-
stage operation.
Using the concept of the straight line formula (y = mx + b) ,
Gerber (1456) interpreted y to equal the designed BOD divided
by the applied BOD, m to be a coefficient equalling O.0052,
x to be the hydraulic dosing rate in feet per day divided by
the product of the depth of the filter in feet and a reaction
rate constant in units of reciprocal day. Using b to equal
O.O7, Gerber stated that the curve was good up to a value of
x = 5O. Ingram (2222) suggested that a filter loading ratio
(R_), defined as the strength of sewage expressed in terms
174
-------
of depth of filter, area of filter, volume of sewage, and
weight of BOD, provided a simple basis for calculating fil-
ter size. A similar concept was used by Gaillard (1391) who
compared the performance of percolating filters in Johannes-
burg, South Africa, on the basis of a "load unit." The "load
unit" was the product of the sewage strength (ppm oxygen ab-
sorbed from permanganate) and the rate of flow in gallons per
cubic yard of medium per day.
Modified equations were proposed by Yakovlev and Galanin
(4840), Meyer and Kirsten (2988), Bushee (552), and Tomlin-
son and Hall (4423), which involved trickling filters operat-
ing with and without recirculation, the quantity of flow and
concentration of the influent, deviations from nonstandard
conditions, and two-stage monomolecular reactions. Schulze
developed (3915, 3916, 3917) formulations to describe the
filtration process which equated biological filtration to an
adsorption process in which the hydraulic load determined
the period of contact and, therefore, determined the effi-
ciency of treatment. Zeller (486O) proposed a formula which
included nitrogen loading. The BOD loading is a major fac-
tor in BOD removal by a trickling filter. The hydraulic
rate is a minor factor and operates inversely as 0.5 power.
McCabe (2895) proposed and defended a two-phase mathematical
formulation which fits the laboratory and field data con-
cerning the biooxidation of waste materials. The BOD removal
is attributed to bioadsorption. The sludge growth during
BOD removal is divided into the constant growth phase, the
declining growth phase, and an auto-oxidation phase. The
formula given for the relation between biooxidation and
sludge growth rate is a two-phase discontinuous function.
Hartmann (1770) proposed design equations to be used for evalu-
ating the efficiency of the revived dipping contact filter
concept. Melzer (2966) developed an equation which was com-
pared with the law proposed by Velz and showed that data
obtained agree more closely with the values calculated by
the newly proposed formula than Velz's formula.
Recent foreign interest in performance evaluation has been
expressed in the development of design and performance for-
mulas by Kolobanov (2543), Tucek (4474), Kucharski (2583) ,
and Ganczarczyk and Suschka (14O4, 14O5). These formulas
involved variables similar to those used in this country, e.g.
hydraulic load, height of tower biofilters or depth of bed,
volume of air feed, temperature, biological activity, indus-
trial waste fractions, aerating ability, and retention times.
Germain (146O) developed an equation for predicting perfor-
mance of a trickling filter containing plastic media for
domestic sewage containing a small amount of industrial
waste. Mathematical models were reported by Swilley and
Atkinson (4276) and Eckenfelder and Cardenas (1O84) which
used laboratory operation to develop necessary data. Further,
Eckenfelder (1O81) developed a relationship using average
175
-------
coefficients based on a statistical analysis of published
data. This relationship was proposed to be used for the
design of trickling filters treating domestic sewage and a
hypothesis was proposed to explain the effects of oxygen
transfer and temperature on the performance of the filters.
Correlation between specific surface area, applied BOD con-
centration, and a removal rate coefficient was defined by
Balakrishnan and coworkers (154). Roesler and Smith (3658)
in 1969 proposed a mathematical model for trickling filters
which involved capital cost, operation cost, and recirculation
along with normal operating variables. Computer testing of
field data agreed with the model.
Nomographs and Charts
Graphical expressions have been proposed in the form of
charts and monographs for estimating the performance of stan-
dard percolating filters by Bernhart (283), and high-rate
percolating filters by Newberry and coworkers (3174). Charts
and graphs were used by Eliassen (1154), e.g., to determine
optimum loading conditions and dimensions of filters by
Sorrels and Zeller (4110), to determine the effect of recircu-
lation by Eliassen (1150), to provide a rapid means of calcu-
lating filter capacity based on gallons of sewage per acre
foot per day or on pounds of BOD applied per cubic yard per
day, and to evaluate many combinations of hydraulic load,
recycle, BOD loading and tower size by Mehta et al. (2955).
Walton (4596) prepared nomographs for determining the depth of
filter required to produce an effluent which would impart a
final river BOD of 40 mg/1 under various conditions. Nomo-
graphs were used to quickly solve design equations, such as
that of the Upper Mississippi Board of Public Health Engineers
(28O1) and the Trickling Filter Floor Institute (5574). A
nomograph developed by Davis (885) shows the relation between
BOD of settled filter effluent and that of crude settled
sewage to BOD of filter effluent and the ratio of recirculated
flow to flow of crude sewage. Rincke (3621) prepared a nomo-
graph to be used for evaluating the performance of trickling
filters which involved influent BOD, surface loading, depth
of bed, load capacity, temperature, percent decomposition,
and effluent BOD.
Statistics
Statistical considerations for performance evaluation of
trickling filter are important throughout the literature;
e.g., Childs and Schroepfer (67O) reported statistics during
analysis of data from trickling filter plants in 1931. Mohlman
(3O52) evaluated the data of the National Research Council
gathered at military installations to determine the precision
expected in the performance of trickling filters and other
waste treatment plants. Statistical analysis and correlation
176
-------
of data obtained from 44 plants in the Upper Mississippi
Valley States was undertaken by Fairall (1232, 1233) for the
purpose of compiling general requirements for trickling filter
design as reported in the Ten States Standards (5011). Pilot
plant data were analyzed statistically (352, 1400) to determine
the performance of single- and two-stage operations, the effect
of BOD loading, temperature and recirculation. The deviation
in measured parameters between several replicate pilot fil-
ters was, for example, the mean standard deviations of BOD =
6*7 ± 1.6; permanganate value = 11.5 ±1.2, ammonia nitrogen =
5.9 ± 2.8; oxidized nitrogen = 32.6 ± 2.5,as reported by
Gameson et al. (14OO). Statistical measurements were used
by Oldaker (3227) to evaluate chemical and biological studies
on flax waste stabilization, temperature effects on trickling
filters by Schroepfer et al. (39O3) , detention times using
radioisotope techniques by Eden and Briggs (11O7), comparison
of double filtration and alternating double filtration while
treating industrial waste by Hawkes and Jenkins (1827), and
proof of the advantages of plastic biological filtering media
in handling specific industrial wastes by Chipperfield (678) .
Formula Evaluations
The papers which referred to the evaluation of proposed for-
mulations, such as noted by Rankin et al. (35O1), contributed
materially to the discussion of the performance mathematics
of biological trickling filters. Hardenbergh and Rodie (1731)
used the design equations developed by the Subcommittee on
Military Sewage of the National Research Council (559O) to
evaluate the performance of single-stage and two-stage fil-
ters under varying load and recirculating conditions. The
National Research Council formula (559O) for efficiency was
proposed to involve the weight of BOD in raw settled sewage
(w) applied to a filter volume (V) for the number of passes
(F) as follows:
-------
Homack (35O2) criticized the conclusion given by Rankin
(35O2) on the use of three methods of evaluation of the Bio-
filtration process where the Velz formula checked most ac-
curately with actual data. Homack questioned if the domi-
nant factor in the performance of a trickling plant is the
ratio of circulation. However, Nelson and Lanouette (3161)
successfully applied the National Research Council formulas
to high-rate biological filters. Climatic factors were also
evaluated by Benzie and coworkers (273) based on the per-
formance relationships of the National Research Council and
Great Lakes-Upper Mississippi River Board. It was found
that there was a significant difference in filter efficien-
cies between summer and winter.
Behn (246) discussed mathematical formulations on the process
with and without recirculation as proposed by Phelps, Velz,
Schulze, Rowland, and Stack. Special emphasis was placed upon
reaction kinetics and the role of recirculation. Further
work was suggested to determine the validity of first order
relationships, the time versus depth as a basis for BOD ex-
traction, the concept of filter saturation, and the role of
filter medium in filter saturation and time of contact. A
formula was proposed by Sima (4O31) and compared with the
formulas of the National Research Council, from which it was
concluded that the Council's formulas gave the best results,
but involved much more work than the one proposed.
Caller and Gotaas (1395, 1396, 1397) presented an analysis
of biological filter variables which eventually led to the
optimization of these factors to be used in design formula-
tions. Archer and Robinson (85) investigated single- and two-
stage biological filtration along with medium volume and re-
circulation ratio and found that predictions of efficiencies
could be adequately expressed by the National Research Coun-
cil performance formulas. Baker and Graves (150) compared
the trickling filter efficiency formulas proposed by the
National Research Council, Eckenfelder, and Galler and Gotaas.
Based upon a computer optimizing program, they determined
that minimum volumes of media would occur when the volumes
in each stage of a two-stage process were equal when the NRC
formula and a modified Eckenfelder formula were used. A
ratio volume of the first to the second stage of 1:2 occurred
if the Galler-Gotaas formula was used. A statistical analysis
of trickling filter performance formulas was reported by
Robertson et al. (3647) to determine, among other things, the
accuracy of the existing design equations in predicting the
performance of the filters and the possibility of improving
predictions by the adjustment of certain parameters using
regression analysis. Based on about 4,OOO single observation
datum points, the formulas of Velz, Schulze, Stack, Rankin,
National Research Council, Farrell, Germain, and Galler and
Gotaas (all put in common nomenclature) were subjected to rig-
orous analysis (3647). An over-simplification of the conclu-
sions reached was that the Galler and Gotaas equation for domestic
178
-------
waste water could be used to fit the predictions of trickling
filter treatment of paper mill waste better than the others
considered.
Defined Functional Relationships
Many additional mathematical studies were published which
dealt with determining the mean residence time in a trickling
filter (99, 364, 1449, 2064, 2O65, 29O4, 2964, 4042, 42O7).
An in-depth review of these investigations was presented by
Sheikh (3981) , who found that the mean retention period used
by many workers was inadequate to define the retention char-
acteristics of a filter. Sheikh used the modal time and the
standard deviation of the log-normal distribution to derive
and compare expressions for filter performance. Additional
mathematical evaluations have been reported by Ames, Behn and
Collings (68) involving the transient operation of trickling
filters, investigations of turbulent flow on vertical walls
by Belkin et al. (25O), and the effect of particle shape on
porosity and surface area of media by Schroepfer (39O2). and
Ganga (2377) on the performance of the filters. Mathematics
involving growth kinetics of the biological film (3824), the
role of diffusion in the rate of biological uptake of organics
(1621), oxygen demand of microorganisms (3428), and mass
transfer of nutrients (2838) are typical of the large number
of mathematically oriented papers which have been published
in the last decade and a half. Mathematical models have
also been proposed for laboratory conditions (548), as well
as on the basis of published literature for the biological
oxidation of industrial wastes (1O77) to provide new tools
(1O85) for performance evaluation by designing engineers.
Critique
Sufficient efforts have been reported to define the primary
variables and identify functional models which may be used
for design and description of filter performance. Solutions
to previous questions, such as the order of mathematical
relationships, the effect of recirculation on efficiency,
and the relation of contact time or medium depth to perform-
ance, have been investigated. Unfortunately, the answers
have not been reported in as straightforward a manner as
those of the National Research Council. It appears that the
advent of the computer has brought into the field investi-
gators who may be more interested in the functions of the
mathematics than in applying the mathematical functions to
practical performance evaluation and prediction. It would
be highly desirable to extend the work of Robertson et al.
(3647) to other data and report the results in a usable
manner. It is agreed and understood that the functions of
the filter are complex, but approximations can be made and
179
-------
a rationale developed. This rationale must satisfy theory,
but be applicable to real problems based upon measurable
parameters.
TREATMENT EFFICIENCY
The importance of determining the efficiency of treatment
plants has been reported by Chase (658) from the standpoint
of design and operation. Studies were made by Knechtges (2516)
and Fairall (1233) on the hydraulic and organic loadings
and of biological trickling filters and their modifications.
Childs and Schroepfer (67O) studied the effectiveness of the
usual parameters, e.g., five-day biochemical oxygen demand, and
suspended solids removal taken on 24-hour composite samples,
for determining treatment plant efficiencies. The large
quantity of information developed on the nitrogen balance
in sewage during biological filtration made it possible to
develop a performance formula which included the nitrogen
loading (486O).
Hydraulic and Organic Loading
The literature has a large number of reports dealing with
hydraulic and organic loading of the standard rate of low
rate single-stage biological trickling filter, such as those
by Schulze (3916), Grantham (1542), and Beatty (232). Ef-
ficiency is affected by organic and hydraulic load with re-
circulation in two primary ways: an increase in the organic
load affects the effluent quality more than an increase in
the hydraulic load, and,as either load is increased, the
total amount of BOD removed increases, but the percentage
removal decreases, giving an effluent of lower quality.
Overall reductions in suspended solids (94$) and BOD5 (93$)
were reported by Lose (2762) on the llO-foot diameter trick-
ling filters of Chatham, New York, which were designed to
treat a maximum daily flow of 5OO,OOO gal. A waste treatment
plant at Albemarle, North Carolina, designed to treat 1.75
mgd (3O81) performed similarly,with the average removal being
90$ for suspended solids and 8O to 95$ for BOD5 (with post-
treatment) . Performance of the Princeton, New Jersey, waste
treatment plant was affected by flooding the filter, which
then discharged an effluent which was four times the normal
15 to 2O mg/1 BOD5. Even though slightly overloaded, Has-
further (1788) reported that the new biological filter at
Fairfield, Illinois, reduced the BOD by 92.5$ and the final
effluent contained 8 to 18 mg/1 nitrates. Digester super-
natant liquor (3767) at loadings up to 4OO mg/1 BOD was ef-
fectively reduced by 8O to 87$ by a trickling filter loaded
at the rate of 1 mgad (23 g'al./ft2/day) .
Some of the industrial wastes which have been treated by
biological trickling filters are as follows:
180
-------
1. Gas Liquors - loaded at 70 gal. /yd3 /day (about 250
Ib of BOD/1, OOO cu ft/day) produced 6O to 7O$ phenol
content reduction when the gas liquor was kept between
O.5 and 1$ of the influent (1413).
2. Refinery Wastes - loaded at 9 to 29 mgad (2O7 to
667 gal./ft2/day) with recirculations of 1.5 up to
5 and BOD5 loadings of 4OO Ib/ac-f/day (92 Ib of
BOD/1,000 ft3/day) produced satisfactory effluents
(2891) .
3. Flax Wastes - 1$ concentrations were treated at a BOD5
loading of 5OO Ib/ac-f/day (11.5 Ib of BOD/1, OOO
ft3/day) and produced 57 to 72$ removal (3226) .
4. Canning Wastes - were treated, after screening, to
8O$ - 95$ BOD 5 removal at loadings of 1/2 mgad (11.5
gal./ft2/day) and loadings of 2 mgad (46 gal./ft2/day)
dropped the efficiency to 5O$ - 70$ (2486) .
5. Tomato Canning Wastes - with BOD of 548 mg/1 and
275 mg/1 suspended solids were loaded at a rate of
O.724 Ib BOD5/yd3/day (26.8 Ib BOD5/1,OOO ft3/day)
and resulted in 7O$ BOD5 removal, producing an ef-
fluent with a BOD of 161 mg/1 (190O) .
6. Packing House Wastes - of 1,OOO to 2, OOO mg/1 BOD5
and 2OO to 2, OOO mg/1 suspended solids were treated
to 7O$ BOD removal with a recirculation of 3:1 on
a 4 mgad (92 gal./ft2/day) treatment plant (1716).
7. Laundry Wastes - were treated at O.5 mgad (11.5 gal./
ft2/day) to give 85$ BOD5 reduction, 1 mgad (23 gal./
ft2/day) to give 76$ reduction, and 2 mgad (46 gal./
ft2/day) to give 62$ reduction on a waste stream which
had been subjected to chemical pretreatment (433) .
8. Kraft Pulp Mill Wastes - were treated at a BOD5
loading of 1,42O g/mVday (87 Ib of BOD/1, OOO ft3/
day) and a hydraulic loading of 1.53 m3/m2/day
(37.5 gal./ft2/day) to produce BOD5 reductions of
8O$ (3207) .
9. Chipboard Wastes - were treated to 85$ BOD5 removal
with organic loadings of 15 Ib of BOD/1, OOO ft3/day
and hydraulic load less than 1 mgad (23 gal./ft2/^ay) ,
but with a recirculation ratio of 1:1 BODs removals
were 93$ at the same organic loading and a total
hydraulic loading less than 2 mgad (46 gal . /f t 2/day)
(3593).
A two-stage biological trickling filter was studied by Sorrels
and Zeller at loadings from 32 to 69 Ib BOD5/1,OOO ft3 /day.
181
-------
An increase in BOD5 and suspended solids removal in both the
primary and secondary filters was observed with increased
load. Ammonia nitrogen oxidation was also reduced, which
suggested that for an equal volume of sewage series filtration
was better for treatment than parallel filtration. Two-
stage filter operations loaded organically at 1,OOO, 2,000,
3,OOO, and 4,OOO Ib BOD/ac-f/day (23, 46, 69 and 92 Ib BOD/
1,OOO ft3/day) and hydraulically at 4.46, 8.92, 13.38 and
17.83 mgad (1O2, 2O5, 3O7 and 41O gal./ft2/day) indicated
that the total reduction in BOD in the primary filters in-
creased with increased loading up to 3,OOO Ib/ac-f/day, (69
Ib BOD/1,OOO ft3/day), but fell *t the loading of 4,OOO Ib
(92 Ib of BOD/1,000 ft3/day). Effluent BOD of 3,69O mg/1
from the Porteous process (a heat sludge-conditioning process)
was loaded at 5O gal./yd3/day on a two-stage filter and pro-
duced effluent BOD's of 5OO and 86.7 mg/1, and the settled
effluent was 38.3 mg/1 (2783). After further investigation
(2783), it was determined that the primary filter could be
loaded at rates of 156 gal./yd3/day and the settled effluent
could be applied at 72 gal./yd3/day to the secondary filter
to yield substantial BOD removal.
Industrial wastes such as that of compressed yeast manufacture
were treated on two-stage operations with no recirculation
at loadings of 300 to 2,8OO Ib BOD/ac-f/day (6.9 to 64 Ib
BOD/1,OOO ft3/day) and produced reductions of 5O$ BOD (3769).
A marked decrease in efficiency was observed (3769) at load-
ings in excess of 1,2OO Ib BOD/ac-f/day, (27.6 Ib BOD/1,OOO
ft3/day), but performance was not significantly affected by
flow rates between O.19 and 1.73 mgad (4.4 and 4O gal./ft2/
day). Pharmaceutical wastes were treated (1941) by two-
stage trickling filters loaded at 2.43 mgad (55.5 gal./ft2/
day) and an average BOD of 13,OOO mg/1 which produced 96$ re-
moval. Milk wastes were treated by two-stage filtration (48O2),
in which the first-stage filter removed an average of 1,35O
Ib BOD/ac-f/day (31 Ib BOD/1,OOO ft3/day). The second-stage
filtration increased the total reduction to 9O.4$ BOD5 re-
moval,with effluent having the desirable properties of dis-
solved oxygen, nitrates and alkalinity.
Converted high-rate trickling filters in Baltimore, Maryland,
were operated from 6.5 to 26 mgad (15O to 6OO gal./ft2/day)
during which the highest loading rate still produced 5O$ BOD
removal. It was noted by Keefer and Meisel (24O4) and Keefer
and Kratz (2396) that by loading between 2.68 and 16.39 mgad
(62 to 377 gal./ft2/day) 77.3 to 91$ removal was achieved,
while with loadings of 2.68 to 13.53 mgad (62 to 311 gal./ft2/
day) 8O$ BOD removal was experienced. Experiments in Chicago
(3O45) were conducted with a hydraulic loading of 20.2 to
26.6 mgad (465 to 612 gal./ft2/day) in which the raw sewage
containing 97 mg/1 BOD and 122 mg/1 suspended solids was
reduced to 12.3 and 44 mg/1 BOD, depending on the season.
The Upper Mississippi River Basin Sanitation personnel (4759)
182
-------
investigated high rate percolating filters treating domestic
sewage in conjunction with trade wastes. Plants using two-
stage high rate treatment had overall efficiencies of 83.6
to 95.1$ BOD removal, but single-stage treatment resulted in
57.5 to 91.4$ removal. Suspended solids removal in the two-
stage operation was 88.6 to 94.3$ with organic nitrogen being
removed in the two-stage process by 28 to 64$. Provided the
organic load was less than 1,000 Ib of BOD/1,OOO ft3/day, when
the hydraulic load of the filters was between 7 and 30 mgad
(26 to 69O gal./ftVday), the reduction in BOD was proportional
to the load (4759).
Roughing filter applications with 13 to 15 meter (43 to 49
ft) deep beds (2988) handled phenolic and fatty acid wastes
at organic loads of 1,OOO mg/1 of BOD which reduced the phenol
concentrations of 5OO to 95O mg/1 down to 1 mg/1. Plastic
media were used (513) to treat phenol waste to 82$ removal at
hydraulic loadings of 11 mgad (253 gal./ft2/day) and 85 Ib
BOD/1,OOO ft3/day.
Food processing and milk waste waters (1157, 5325) were
treated effectively on roughing filters. Seventy-five per-
cent reduction in chemical oxygen demand was experienced
for the milk waste (1157) that was loaded at approximately
71 mgad (1,633 gal./ft2/day) and 8.8 Ib of COD/1,OOO ft3/day.
Trickling filter towers loaded at 3,1OO to 3,9OO grams of BOD
per cubic meter per day (O.23 to O.3O Ib BOD/1,OOO ft3/day)
produced a reduction of 71$ in BOD. " The two-stage operation
handled a load of 3,OOO to 3,5OO grams of BOD per cubic meter
of medium (O.23 to O.27 Ib of BOD/1,000 ft3/day) and produced
91$ reduction in BOD with 83$ of the organic nitrogen being
removed in the last stage (391O). Optimal surface loading
of high rate filters was found to be O.8 to l.O m3/m2/hr (475
to 59O gal./ft2/day) and the optimal organic loading was 0.8
to 1.1 kg BOD/m3/day (5O to 69 Ib of BOD/1,OOO ft3/day) (403O).
Greater efficiency was experienced in BOD reduction when operat-
ing at the rate of 1O mgad (23O gal./ft2/clay) than at an aver-
age rate of 1.1 mgad (25 gal./ftv^ay) (383), when treating
industrial wastes from cotton finishing. A cotton mill waste
(35O to 6OO mg/1 BOD) was treated adequately on a high-rate
biological filter (2688) at 28 mgad (645 gal./ft2/day). Milk
and packing house wastes were treated on high-rate filters
of various media (27O1) to 15, 43.4, 48.8, and 49.4$ BOD re-
duction and were loaded at 16 mgad (368 gal./ft2/day). After
sedimentation, BOD reductions were 8O to 9O$ with significant
increase in nitrification (27O1). The BOD removals per acre
foot of filter were 35O to 41O pounds (8 to 9.4 Ib BOD/1,OOO
ft3/day) at 2 mgad (46 gal./ft2/day) and 2,2OO to 2,60O pounds
at 16 mgad (5O to 6O Ib of BOD/1,OOO ft3/day at 368 gal./ft2/
day).
183
-------
Gas plant, milk, sugar refinery, and tannery wastes were
treated by biological filtration (1138). Phenolic wastes
were loaded at 1 mgad (23 gal./ft2/day) and the greatest
removals attained were 4.72 Ib of BOD/1,OOO ft3/day of medium
with concentrations of 1OO to 115 mg/1 phenol. When loaded
at 16 mgad (368 gal./ft2/day) with the milk processing waste
the BOD was reduced 90.6$, with the influent BOD of 537 mg/1
and the effluent BOD of 46 mg/1.
High-rate trickling filters have treated chemical wastes
containing organic oils and acids and up to 5,OOO mg/1 of
formaldehyde (3OO to 1O,OOO mg/1 BOD5) that were diluted and
applied at 5O mgad (1,15O gal./ft2/day). The formaldehyde
content was reduced to 46 mg/1 with an 82$ BOD reduction.
By treating the composite waste waters of the chemical plant
(948), 94.5$ of the formaldehyde was removed. High-rate or
tower percolating filters provided adequate treatment loaded
at 24 mVmVday (59O gal./ft2/day) for a waste (1OO mg/1 of
BOD) which was subjected to an oxygen capacity of 75O g/m3/
day (46.8 lb/l,OOO ft3/day) (3125).
A modification of the trickling filter, the Pruss enclosed
filter (1691), treated 3.5 to 24.2 gal./yd3 of media per day
(13O to 892 gal./I,OOO ft3/day) of domestic sewage and pro-
duced an effluent with a BOD of 16 to 8O mg/1. The Aero-
filter was loaded (1667) with a waste (4,OOO to 8,OOO mg/1
BOD) at 25 mgad (575 gal./ft2/day) and produced a 95 to 97$
BOD reduction on a diluted influent. BOD loadings of 18 to
20 Ib/yd3/day (666 to 742 Ib BOD/1,OOO ft3/day) were applied
to a Biofilter to produce 9O$ BOD reduction of a distillery
waste (20,OOO mg/1 BOD). It was further noted (1284) that
the hydraulic loading should not be less than 48 mgad (1,1O4
gal./ft2/day) or greater than 1OO to 125 mgad (2,3OO to 2,875
gal./ft2/day). The Accelo-Filter was reported by Gillard
(1475) to treat normal domestic sewage with 100 to 20O$ of
the average flow recirculated, and generally it was expected
that a BOD reduction of 8O to 95$ was achieved.
Use of trickling filters and activated sludge was described
by Oldaker (3227) for the treatment of flax waste. The acti-
vated sludge reduced the flax waste BOD by 15 to 18$ after 24
hours and by 53 to 59$ after 54 hours. The percolating fil-
ters took a recirculation ratio of 3:1 with a hydraulic load-
ing of 3.5 mgad (8O.6 gal./ft2/day) and 649 pounds BOD/ac-f/
day (14.7 Ib of BOD/1,OOO ft3/day) to produce 88$ removal or
1,432 pounds BOD/ac-f/day (33 Ib of BOD/1,OOO ft3/day) to pro-
duce 71$ removal. Treatment at Greenville; South Carolina,
on combined textile finishing waste indicated that to produce
a BOD reduction of 7O to 8O$ a hydraulic loading of 4 mgad
(92 gal./ft2/day) was used, but the BOD loading must be less
than 3,OOO pounds/ac-f/day (69 Ib of BOD/1,OOO ft3/day).
184
-------
This treatment was followed by an activated-sludge process
which would produce a final effluent of less than 30 Tig/1
BOD (1464). A septic tank followed by a contact bed treat-
ment plant was replaced (1952) by settling tanks and trickling
filters to handle food processing waste (influent BOD 220 to
1,23O mg/1) and were loaded at 1.5 mgad (34.5 gal./ft2/day)
to produce 95$ BOD removal.
Imhoff tank-trickling filter plants were reported by Drury
(1O37) to be aided by 15 to 3O minutes' aeration which re-
duced the BOD by 2O$ and allowed increased trickling filter
loadings of 4OO to 5OO pounds BOD/ac-f/day (9.2 to 11.5 Ib
of BOD/1,OOO ft3/day) for dosing of 16 mgad (368 gal./ft2/
day). Wastes from rayon manufacturing were reported by
Roetman (3659) to be treated on percolating filters at the
rate of 1 mgad (23 gal./ft2/day) and 3OO pounds BOD/ac-f/day
(6.9 Ib of BOD/1,OOO ft3/day) to 9O# removal, and by using
the effluent from Imhoff tank at a rate of 1,O5O pounds BOD/
ac-f/day (24.1 Ib of BOD/1,OOO ft3/day) to complete removal,
as measured by the BOD test. Microstrainers and sand filters
were used (3187, 5184) as treatment after the filters and
decreased the suspended solids in the effluent from 3O to
11O mg/1 to less than 2O mg/1, with 75$ as an average reduc-
tion.
Critique
The efficiency of trickling filters has been reported in
numerous domestic and industrial applications. However, the
usual percent removal and the occasional hydraulic and organic
loading information may not be adequate criteria to determine
performance. Specific data dealing with enzymatic rates of
biological activity and general information on nuisances and
operational problems have been published in an effort to
supplement the usual performance efficiency. It is apparent
from the reviewed literature that the performances of biologi-
cal trickling filters were variable from the standpoint of
background, design, construction, operation, and maintenance.
Both hydraulic and organic loadings of the filter have been
described as being the dominant factor in treatment performance.
To summarize expected performance of biological filtration is
extremely difficult and, perhaps, misleading. However, some
performance information should be provided in every paper for
comparative purposes. The following comments in Table 5 are
submitted for the novice, not the expert or practitioner, and
it should be fully understood that exceptions may be raised
with valid reason.
It is to be carefully noted that the information in Table 5
may change if the waste water fluctuates in strength and flow,
toxic elements are available, and limitations are imposed
upon the physical plant. The value of observing these data
is that these figures have been reported on a variety of
waste water and conditions of treatment. One may view this
185
-------
information in light of its source, i.e., the performances
of operating plants have been described and it should not
be difficult to design other plants to duplicate the per-
formance of the trickling filters.
Table 5
Summary of Filter Criteria
Filtration Filter Usual Equipment
Rate Stage Modification
Low
Low
High
High
Super
Single Clarifier
!Iwo Clarifier
Recirculation
Single Clarifier
TVo Clarifier
Recirculation
Single Deep bed
Super or Single Followed by
High activated
sludge
Expected
Efficiency,
60-70
90+
40-50
85
50
95+
Possible
Improvements
Recirculation
Additional stage
Nutrient removal
Alternate system
Additional stage
Recirculation
Alternate system
Nutrient removal
Clarifier
Recirculation
Sludge handling
Recirculation
Alternate system
Bacterial and Viral Removal
The performance of biological filters in reducing the bacteria
and virus particles has been the subject of numerous papers.
General bacteriological studies by Malguth (2849) indicated that
there was a continual decrease in the number of bacteria in
each unit of the waste treatment plant and the average final
percentage reduction was 95$ at 37.5°C (99.5°F) and 94.3$ at
2O°C (68°F). Coffin and Hale (731) discussed in 1916 the
effects of treatment units on bacteria reduction. The septic
tank reduced the total number of bacteria by 72$ and Bacterium
coli by 4O$. Primary contact beds reduced the total number
of bacteria by 16$ and B. coli by 15$. Settling basins re-
moved an additional increment (731), e.g., total number of
bacteria by 2$ and B. coli by 13$, and sand filtration removed
91$ of B. coli. Effluent chlorination destroyed a 99.9$ of
total bacteria and 99.997$ of B. coli. Similar data were pub-
lished by Allen and Smith (48), in which they showed that
percolating filters and humus tanks removed 97.2$ of B. coli
and 98.3$ of Streptococcus faecalis from settled sewage,
which gave a total treatment plant reduction in these bacteria
186
-------
of 98.68$ and 99.4O$, respectively. Septic tanks followed
by trickling filters removed B. coli by 89 to 97$.
Endamoeba histolvtica cysts were reduced by 5O$ in primary
sedimentation and 9O$ due to trickling filter with secondary
clarification, according to Kott and Kott (2553). However,
cysts of Endamoeba histolvtica were reported in sewage in-
fluent in Israel (2552), but after biological filtration
cyst counts were very low or no cysts were found. Average
bacterial reductions reported of ten waste treatment plants
(4245) indicated that total bacteria count was reduced 95.9
without chlorination and 99.8$ with chlorination, while the
B. coli were reduced 99.6$ without chlorination and 99.98$
with chlorination.
The relationship between the efficiency of the filter opera-
tion and the number or organisms present was discussed by
Nakajima (3144) and it was recommended that the optimum BOD
loading was 900 ± 1OO g/m3/day (56.1 ± 6.2 Ib BOD/1,OOO ft3/
day) . While checking for tubercle bacilli, Mu'ller (3115)
found that in the sludge from lightly loaded percolating
filters there was a positive indication in 9O$ of the samples,
while 29$ to 5O$ of the samples of sludge from high-rate
filters and activated sludge were positive. The bacterial
activity of the bio film was reported by Pohl (34O1) to be so
great that a depth of one-half meter (19 inches) was suffi-
cient for purification if the waste was applied intermittent-
ly and frequently in small amounts well distributed over the
whole surface. Rudolfs et al. (3739),also reported that B.
coli reduction was the greatest in the upper half of the
filtering bed (2 feet) where the protozoa were the most nu-
merous. Research by Knop and Husmann (2524) has shown that
two-stage trickling filter operations produced a better re-
moval of Bacterium coli than single-stage, but the advantages
did not offset the installation or operational cost. Allen
et al. (4O) supported the conclusion reached by Knop and
Husmann that two-stage filters provided better bacterial
removal.
Investigations showed that sewage in the Birmingham Tame
district (4933) with a high percentage of trade wastes con-
tained less bacteria than a weak domestic sewage. Studies
at Pennsylvania State College (3248, 4579) indicated that
closed percolating filters gave slightly better removal of
bacteria than open trickling filters. Effluent examinations
by Brauss ( 448) revealed that irrigation fields removed more
total bacteria and B. coli than percolating filters. Low-
rate percolating filters were preferred, however, by Neumeyer
(3172) to that of high-rate filters because of the better
bacterial count reduction. While comparing various units of
an Imhoff-trickling filter plant, Hotchkiss (2O44) noted that
the effluent from the sprinkling filter contained the fewest
bacteria compared to other systems studied. Biofilter opera-
tions in New Zealand were reported by Rowntree (37O7) to
provide a high degree of treatment in which B. coli was
absent from the final effluent.
187
-------
In an attempt to aid bacterial removal, aeration was pro-
posed as an auxiliary to trickling filter treatment (36O5).
It was found that the value of the aerator was mainly in the
reduction of biochemical oxygen demand and in flocculation
of part of the colloidal matter, but that there was no re-
duction in B. coli. During a discussion of bacteria and ef-
fluent quality of percolating filters, iron bacteria and
sulphur bacteria were reported (2252) to be found in the ef-
fluent from series operated percolating filters, but not in
the effluent from single filters. Interest was expressed
by Hurley (212O) and Grigorieva (16O3) on the effect of bio-
logical trickling filtration on viruses and various other
forms of microorganisms.
Kabler (2364) reviewed work on the effect of biological fil-
tration on pathogenic bacteria, fungi and viruses and in-
dicated that more efficient methods were needed for enumerat-
ing these life forms in sewage. Aerobic and anaerobic treat-
ment were investigated on bacteriophages (985) and indicated
the resistance of this life form to normal biological treat-
ment. Human enteric viruses in sewage were studied by
Clarke and Kabler (7O2) and they concluded that primary
treatment has little or no effect on the virus; however, bio-
logical filtration reduces viral numbers by about 4O$, while
activated sludge reduces them by 9O to 98$. Similar work
reported by Shuval et al. (4O11) indicated that only about
one-third of the phages present were removed during treat-
ment which indicated that biological filtration plants are
ineffective in removing viruses.
Coxsackie A-13 virus was observed (2243) to pass through
both activated sludge and trickling filters and, was con-
trolled only by chlorine dosage of O.5 ppm with 8 hours'con-
tact. In continuing studies on the efficiencies of mechani-
cal-biological treatment of sewage, Schutt (3920) added
known numbers of Staphylococcus phage 3C. with radioactive
sodium used as a tracer. He concluded that there was simi-
larity in the removal of radioactivity and the inoculated
phage, and that inadequate purification efficiency of per-
colating filters was demonstrated. Mundel et al. (3122)
expressed concern on the viral carry-over from humus tanks
following trickling filters and concluded that vegetables
grown on land irrigated by this liquid should be cooked be-
fore consumption by human beings.
Bacterial removal efficiencies reported by Rudolfs et al.
(3731) indicated that up to 9O$ of B. coli was removed in
the upper 2.5 feet, where most of the bacteria-feeding
protozoa were to be found on the trickling filter.
Helminthic ova were reduced by 18-26$ by biological filters,
whereas Imhoff tanks removed 97$ and, when followed by sec-
ondary sedimentation tanks, removed 87$ of that remaining
(4125). Helminthic ova were in the crude sewage (316) in
India, and after two-hour sedimentation 5O to 7O$ was re-
moved, whereas treatment with percolating filters or acti-
vated sludge reduced them by 9O to 1OO$.
188
-------
Allen et al. (46) concluded that percolating filters removed
8O to 9O$ of the bacteria and that further treatment on sand
filters would improve the bacterial effluent quality but
could not be relied upon. A high-rate biological filter was
reported by Rabinowitz (3496) to reduce the coliform count
of 16 x 1O organisms per milliliter in crude sewage by 7O
to 9O$. Using a membrane filter method of testing the ef-
ficiency of biological sewage treatment plants, Beling et al.
(249) showed that the percolating filter is important in the
removal of intestinal bacteria. The complete sewage treat-
ment resulted in an average reduction of 94.5$ in total
bacteria and 93.7$ in B. coli. High-rate filters were re-
ported (5181) to remove 75$ of the bacteria from a spent coke
oven liquor and produced an effluent of acceptable quality.
Effect of medium and seasonal changes was noted by Tomlinson
et al. (4428); e.g., the number of coliform bacteria in the
settled sewage were higher during the warmer months, but the
total bacteria count showed no regular seasonal variation.
Filters containing four different kinds of media produced
a 95$ reduction regardless of the medium. Typical of in-
dustrial waste investigations was that reported by Dickinson
(96O) who observed that the bacterial reduction during treat-
ment of a cannery waste on a percolating filter was 37 to 99$
at 37°C (97°F). Heukelekian (1917) reported that the coliform
bacteria were reduced to a greater percentage during the sum-
mer than in the winter on the Biofilter investigated, with the
sedimentation tank considered as part of the unit. The per-
centage reduction of the total bacteria bore no relation to
the percentage reduction of the coliform, but generally
90$ reduction was achieved.
Chlorination, of course, has been the workhouse for disin-
fection and bacteria kill for many years. Mebus (2947) stated
in the 192O's that the B. coli were reduced 99.992$ during
sewage operations, which included chlorination. Chlorine
was used by Shuval et al. (4O12) effectively for virus con-
trol and they obtained 99.9$ reduction in virus concentrations
and coliform counts with MPN (most probable number) less than
1OO per 1OO ml. Reference to the effect of chlorination on
Coxsackie virus A-13 has already been made. Chlorination
following percolating filter treatment was reported by Goodwin
(1518) to reduce the bacteria content by more than 99$. After
chlorination of trickling filter effluent, bacterial contents
were 144/cc at 37° (2574) and 3 to 38/cc (4212), with B. coli
reductions down to 4.2/cc (2574), and no counts/cc (2852,
4212). In 1928, a waste treatment facility at a recreational
area, consisting of a septic tank and trickling filters with
post-chlorination, was able to remove 99.99$ of B. coli (5O88) ,
189
-------
Critique
Biological trickling filters have demonstrated effective
and economical removal of bacteria during the treatment of
sewage. An awareness of the importance of bacterial removal
was basic in the literature. Some divergent opinions ex-
pressed over a period of several years probably indicate the
changes in the identification and isolation techniques more
than changes in treatment plant efficiences. Soluble organic
removal and bacterial removal may be generally considered as
responding similarly in a treatment system; e.g., a system
reducing 9O$ of the BOD will provide high bacterial removal,
but a low BOD will indicate low bacterial removal. There are
conditions, such as roughing filters, where a sizable BOD
removal may be achieved, but without proper post-treatment,
the bacterial removal may be low. There was considerable
evidence that trickling filters do not provide adequate treat-
ment for virus removal. However, the testing procedure for
enumerating viruses leaves much to be desired and additional
work is underway to improve this situation.
Disease Prevention
Disease transfer was recognized as the major problem to be
faced in waste treatment. Diseases transmitted by sewage
were reported by Wilson (4753), e.g., cholera, typhoid, para-
typhoid, dysentery, Malta fever, Weil's disease, tuberculosis,
anthrax, and poliomyelitis. Requirement standards have been
established since 191O (3261) that the effluent from a facil-
ity must be unobjectionable and practically free from typhoid
and other pathogenic germs. Under normal conditions, the
objective of the sewage works treatment was to eliminate the
risk from infection from the effluent (4753) ; however, the
effluent should not be used for irrigating crops which are
eaten raw, and during epidemics the effluent should be dis-
infected. Hardy (1735) considered that the risk of infection
transported by birds and insects to reservoirs or water
supplies had been over-emphasized.
Work dealing with the control of disease transfer was re-
ported in California (496O), where a three-year survey of
direct utilization of waste waters was successfully conducted.
Ingram (222O) compiled a selected bibliography dealing with
organisms associated with processes of sewage treatment, in-
cluding Sphaerotilus, zoogloeal bacteria and parasitic orga-
nisms causing disease.
Waste waters from a tuberculosis sanitarium, reported by
Rutz and Weigmann (3794), were treated by biological fil-
tration with chlorine post-treatment. Kabler's (2363) re-
view of several investigators' work on pathogenic organisms
indicated that enteric bacteria, viruses and tubercle bacteria
were reduced considerably by biological filtration and other
aerobic processes. However, Heukelekian and Albanese (1925)
190
-------
investigated sedimentation, septic tanks, biological filtra-
tion, and activated sludge, and indicated that all were in-
effective in removing Mycobacterium tuberculosis from sewage.
This work supported the earlier work by Jensen (2317), who
showed that sedimentation tanks, Imhoff tanks, or percolating
filters handling waste from tuberculosis sanitaria or slaughter
houses in which tubercular animals were killed all had T.
bacilli in the effluent (control was effected only by chlori-
nation). It was demonstrated by Jusatz (2358) prior to the
above two papers that treatment of sewage on a trickling fil-
ter does not prevent the appearance of Myco. tuberculosis and
that they may be killed by adding sufficient chlorine to kill
B. coli and reduce the bacterial count by 99 per cent. Studies
carried out by Bhaskaran et al. (317) in Calcutta on waste
waters from sanitarium and slaughter houses indicated that
viable Myco. tuberculosis found in 95% of the sanitarium
sewage samples and that biological filtration was not very
effective in removing these organisms. No tubercle bacteria
were found in the 33 samples of slaughter house waste waters
which were taken at different times. A special plant designed
for the waste waters from Gauting Sanitarium (near Munich)
(4471) handled the waste of a thousand patients and a staff
of 50O by biological filtration with post-treatment and under-
ground effluent disposal.
Experiments documenting the longevity of Escherichia coli in
comparison to Salmonella typhosus were established by Beard
(23O). He showed that 96$ reduction of the typhoid organism
occurred after six hours' aeration in activated sludge and up
to 99$ reduction was obtained when treating the sewage on
trickling filters at rates up to 6.6 mgad (152 gal./ft2/day) .
At least partial treatment was recommended by Wilson (4747)
in South Africa to avoid infection by typhoid or hookworm
when using sewage as irrigation for areas where body contact
is a possibility. In 191O, strict requirements were developed
for the effluent of a sewage plant for Fort Benjamin Harris.
It was required that the effluent not be putrescible, must
be clear, colorless and practically free of typhoid and other
pathogenic germs (3261) .
The uses of hypochlorite (1O7), "oxychloride" (36O2), chlorine
gas (2554), and high temperature incineration (1458) were among
many of the methods used for disinfection purposes. It was
noted that an outbreak of poliomyelitis resulted in the rec-
ommendation that digested sludge not be used as a fertil-
izer, even though no direct relation was proved between the
methods of sewage disposal and the spread of this disease
(3886) . Kelly (2425) found that primary and secondary treat-
ment for biological filtration without chlorination did not
reduce the frequency of virus isolation. Virucidal effects of
chlorine in waste water indicated (539) that less chlorine
was required to inactivate phage in settled sewage than in
filter effluent, but testing techniques left much to be
191
-------
desired. It was reported (3146) that the activated-sludge
plant aerators and the trickling filters were the major
source of emission of coliform organisms. However, 5O% of
the particulate matter emitted was above 5 micron diameter
(3146), would be filtered by nasal passages, and would not
reach the lungs. In the Biofilter (1158), the high protozoan
activity controlled the undesirable bacteria by their activity
increasing with increasing load.
Critique
The effectiveness and the limitations of the biological trick-
ling filter as a disease controller are demonstrated in the
literature. The aerobic secondary treatment has been shown
to be quite useful for the control of enteric, anaerobic,
and pathogenic bacteria for many years. Trickling filtration
is but one method of providing the aeration. However,
indications are that the sorptive capacity of the biofilm
for colloidal particles, e.g., pathogens, etc., possesses
additional treatment capabilities. For positive control
of virus and disease, chlorine or other disinfectants must
be used.
It is of interest to note that litigation is presently under-
way dealing with organic pollution abatement and other viola-
tions of waterways. Accusations question the inadequacy of
plant designs to provide the required treatment. Additional
Government standards and guidelines are introduced dealing
with the nutrient content of waste treatment plant effluents.
This activity has been intensified considerably from previous
years, when storm water sewers were causing the problems in
receiving bodies and simple primary treatment was required.
In addition to the stream receiving the effluent being con-
taminated, disease transfer, as exemplified by the epidemics
which spread throughout Europe and this country, was a dis-
tinct possibility. The purposes of waste treatment should be
kept in mind when pointing out the inadequacies of an old
plant for nutrient control, but which is still keeping disease
from running rampant.
192
-------
PART III
TRICKLING FILTER RESEARCH AND DEVELOPMENT
APPROACHES, ECOLOGY, AND PATENTS
This section first supplies information pertaining to the
type of approaches used to develop the current technology
of biological trickling filters. The approaches to research
and development were divided into areas which were commonly
reported in the literature, i.e., laboratory-scale, pilot-
scale, and field- or full-scale. Secondly, due to the im-
portance of the biological reactions and relationships, the
ecology of the trickling filter is reviewed in some detail.
Thirdly, a section on patents is included, as patents rep-
resent the end-product of research and development. No cri-
tique or comment is made since many of the patents lapsed,
and the detail which would be required to review them ade-
quately is beyond the scope of this review.
193
-------
SECTION XIII
LABORATORY-SCALE INVESTIGATIONS
In this section, laboratory-scale investigations may be de-
fined as bench-top measurements, usually under room tempera-
ture and/or controlled conditions, with proper and sufficient
analytical equipment readily available. Unless defined
otherwise, all experiments described in this section are on
the laboratory scale.
Typical of academic studies were those from the University
of Newcastle upon Tyne which used laboratory, as well
as larger size, biological filters (5608) to determine the
ability of isolated biofilm bacteria to metabolize various
synthetic sewages. It was concluded that no specific strain
of organisms possesses greater purification capacity than
that of a heterogeneous population. Brebion (45O) described
an apparatus which was designed to simulate conventional
treatment of sewage by coagulation, mixing, primary sedimenta-
tion, activated sludge, biological filtration, secondary
sedimentation, and modifications thereof which were used to
study the treatment of sulfite waste waters. Concern was
expressed by Gameson et al. (14OO) on the validity of data
obtained on small trickling filters, and therefore twelve
replicate systems were maintained simultaneously to deter-
mine reliability of the data. Stopler et al. (4236) reported
on the design and operation of a glass-column, laboratory
filter. Later, Eckenfelder and Cardenas (1O84) discussed the
selection and operation of laboratory activated-sludge units
and percolating filters in conjunction with techniques to be
used to establish full-scale design criteria.
Laboratory studies were conveniently used by Nemerow and
Armstrong (3165) on the activated-sludge process prior to
pilot-scale operations using activated sludge and trickling
filters. Brebion and Huriet (452) suggested that the selec-
tion of a treatment system was most effective and economical
based on results from laboratory and pilot plant conditions.
This concept was practiced by Dickerson (956) to determine
optimum conditions of treatment for an existing two-stage
high-rate biological filtration followed by activated sludge.
Alternating double filtration to handle waste treatment in
Ventura, California, was shown not to be beneficial, contrary
to reported experience, based on laboratory investigations
(4266). Studies which would be very difficult on larger
scale, such as measurement of data on sub-surface disposal
of trickling filter effluents, were made on laboratory soil
filters which were dosed under intermittent and continuous
conditions (343O).
Detailed functional and practical data were gathered on lab-
oratory filters (423.6). Oxygen saturated dilutions of a nutrient
source circulated through tubes containing biofilm increased
195
-------
the flow and caused increased turbulence (1775). From this
simulation, it was concluded that more turbulence increased
the supply of food and oxygen, and, therefore, accelerated
the metabolic rate of the total system (1775) . Laboratory
models of biofilters 165 cm (65 in.) high and 1O.5 cm (4.17
in.) diameter were used (2365) to evaluate oxidative proces-
ses on biological filters composed of quartzite, gravel,
limestone, coke, anthracite, and boiler slag. Experiments
with variations of periodicity of dosing in laboratory per-
colating filters were made (4989) and it was concluded (573)
that the maximum interval was eight minutes. Ganczarczyk
and Suschka (14O4) used laboratory columns 12O cm (47.2 in.)
long and 2O cm (7.88 in.) in diameter of zinc plated steel
pipe as biolfilters to determine kinetics of aeration by
measuring a decrease of sulfite concentration in a synthetic
sewage.
Results from experimental studies indicated (37O2) that media
depth had little effect on the efficiency per cubic yard of
medium based on settled sewage flow rate of 22O gal./yd3/day
(8.14 gal./ft3/day) applied to filters 3, 4, 5, 7-1/2, and
9 feet deep (2.71, 2.O4, 1.63, 1.O9 and O.91 gal./ft2/day,
respectively). Balakrishnan (153) found a good correlation
between specific surface area of the six-foot deep media and
nitrification rates. Stone (4227) experimented with ground
water supply, which led to the construction of a potable
water treatment plant incorporating a biological trickling
filter to remove sulfides. Analytical techniques such as
photographic methods were used to study the behavior of
water flowing freely down vertical surfaces at Reynolds
numbers between 2OO and 3O,OOO. The thickness of the liquid
layers under varying dosing conditions (25O) was thus deter-
mined. Neutron scattering techniques were described by Eden
et al. (11O4) to demonstrate the relationship between weight
of film, retention period, and performance.
Rotary tubes were used as laboratory models of biological
trickling filters (14O2, 1489) in which detailed information
on oxidative capacity, radioactive sorption, ecology, and
other bio-control factors were obtained. Inclined rotating
tubes which develop a biofilm were used by Tomlinson and
Snaddon (4429) to determine the critical depth related to
nitrification as 0.2 mm (O.O8 in.) and evaluate diffusion
coefficients. Kornegay and Andrews used a rotating annular
ring reactor to develop a mathematical model for biological
filtration. The biofilm reached a constant thickness of
approximately 2OO microns, at which time the shear on the
film surface was sufficient to wash the newly formed organ-
isms into the liquid phase (2547).
Gulevich (1621) applied a rotating disc for a laboratory
simulation of a trickling filter. The disk presents, math-
196
-------
ematically, the simplest surface with regard to mass transfer.
This device was well suited for evaluating the relative im-
portance of diffusion using glucose as a nutrient, and it
might be of value on more complex compounds. Laboratory and
pilot plant experiments to treat phenolic wastes from air-
craft maintenance were evaluated by Reid et al. (3547) on
rotating drums, biological filtration, and activated sludge.
They concluded from these studies that the rotating drum type
plant was the best for very high concentrations of phenolic
waste.
Mathematical relationships were developed by Rowland (2O64)
using laboratory models of biological trickling filters.
Maier (2837) used a flat inclined surface covered with a bio-
film to simulate and study the purification of sewage in
percolating filters. Mass transfer and growth rate were re-
lated to the flow pattern and other biotic factors. Swilley
and Atkinson (4276), Kehrberger and Busch (2419), and Busch
(548) used laboratory models composed of an inclined plane
which was used to describe and analyze trickling filter sys-
tems. Filtration simulation by an inclined plane was also
reported by Atkinson et al. (1OO) in the study of the phases
of growth of the biofilm .related to hydraulic and organic
factors. Mathematical studies of reactions in biological
filters were proposed (1OO, 548, 2419). Schulze (3916) and
Green et al. (1571) used a vertical laboratory screen simu-
lated filter to develop a relationship between hydraulic and
organic load and efficiency, as well as to measure growth
rates under controlled temperature conditions.
Contact time on trickling filters was determined by Bloodgood
et al. (364) by a laboratory simulation, such as a string of
spheres hanging vertically and an inclined plane. The con-
tact time was determined (364) to be inversely proportional
to the two-thirds power of the liquid application rate.
Sinkoff et al. (4O42) also studied the mean retention time
on trickling filters and used a simulation consisting of
columns packed with spherical media, such as glass and
porcelain, of varying sizes.
A thin film aeration reactor was used (45O7) to determine that
oxygen saturation was reached within two feet of filter depth,
using a loading rate equivalent to that of normal practice in
biological filtration. Forced ventilation in these experi-
ments did not increase the oxygen uptake. Oxygen consumption
was measured and found to be between 6O mg/1 and 160 mg/1 for
sewage supplied by normal atmospheric ventilation. Atkinson
and Swilley (1O1) used laboratory percolating filters to show
that suspended solids and dissolved salts reduced the effi-
ciency, and addition of ferric chloride to the influent sharp-
ly reduced the removal efficiency. However, the addition of
salt to the influent to twelve laboratory percolating filters
was observed by Mills and Wheatland (3O17) to have little
effect upon the total number of scouring organisms present.
197
-------
The Water Pollution Research Laboratory applied biological
filtration and activated-sludge units to determine the treat-
ability of a specific industrial waste and to develop process
modifications on existing waste treatment plants (5186).
Porter and Dutch (3437) described a laboratory apparatus used
to obtain data on plastic biological oxidation media to han-
dle industrial wastes such as phenol or cyanides.
Successful laboratory operation of biological filtration and
activated-sludge processes was followed by pilot-scale sys-
tems operated in series to handle pulp and paper mill wastes
(2661). Brebion (449) in 1958 used 36 porous columns seeded
with various fungal cultures which were evaluated for treat-
ing dilute sulfite waste waters and the encouraging results
suggested that 8O$ to 9O$ reduction in BOD could be obtained.
Biological filtration followed by activated-sludge treatment
was determined successful (683) and the relationship of tem-
perature and content of hydrogen sulfide on the efficiency
of treatment was defined based on experiments on the treat-
ment of fatty acids by covered percolating filters. Labora-
tory and pilot plant experiments by Christ (684) supported
these findings and also found that waste waters from the manu-
facture of fatty acids could be treated to 9O$ organic removal.
Kucharski and Krygielowa (2582) investigated the treatment of
phenol-formaldehyde plant waste on laboratory filters and the
successful operation provided input for larger scale testing
to confirm that the phenolic concentrations of the waste
waters were reduced from 2OO mg/1 to 1-2 mg/1. Laboratory
filters were used by Zdybiewska (4852, 4854) in Poland, to
determine the feasibility of biological purification of
phenol wastes, and to define the procedure of microorganism
acclimation. Similarly, laboratory and full-scale percolat-
ing filters were constructed at Knostrop Sewage Works by the
Joint Research Committee of the Gas Council of the University
of Leeds to study the biological treatment of mixtures of
sewage and trade waste fractions of problem effluents (56O6).
Laboratory-scale studies on oil refinery waste waters were
reported by Luchter (2775) who evaluated slag, coke and
calcium carbonate as filter media, finding the last one to
be unsatisfactory. The Magdeburg Process of direct biologi-
cal treatment of spent gas liquor without sewage was studied
under laboratory conditions (3193). Many laboratory treat-
ability studies, e.g., Michail and Popescu (2998), developed
methods for the treatment of waste waters containing tar and
phenols discharged from coal gasification plants.
Patil et al. (331O) stated that it was originally proposed to
carry out pilot-scale studies on the treatment of synthetic
drug wastes by biological filtration, but the cost of pre-
paring a synthetic waste to simulate the composition of the
actual waste effluent required preliminary experiments. The
198
-------
results of these studies indicated that biological filtration
had little effect in reducing the oxygen demand and consider-
able savings were made. Both trickling filters and activated
sludge were effective in reducing the BOD of sugar refinery
wastes (32O3), as was demonstrated, initially, by laboratory
and, later, by pilot-scale operation.
Industrial wastes such as that from the tanning industry were
evaluated (2O89) using laboratory and pilot-scale studies,
from which it was determined that a rate of 7O gal./yd3/day
of sewage containing 3O mg/1 trivalent chromium could be
applied to the filters. Approximate toxic levels of metal
finishing wastes were established, initially, by laboratory
operation (3979), and later expanded to pilot-scale opera-
tions . Safe limits of treatment of metal plating wastes
were reported by Kittrell (25O2) based on experimentation
with metal plating wastes, such as cyanide. During addition-
al work, a pair of insulated laboratory-scale filters equipped
with sampling ports of various depths was used by Ware (46O6)
to determine the effect of temperature on the biological
destruction of cyanide waste.
Critique
A significant portion of the literature contained references
dealing specifically with the use of laboratory-scale opera-
tions simulating various facets of the biological trickling
filter processes. Many of the studies were undertaken using
laboratory-scale operations to facilitate control of vari-
ables, such as the influent stream, temperature, humidity,
invading organisms, and other factors which normally affect
the process. As more data were made available and the ten-
dency was established to develop rational formula for design
purposes, several investigators found it convenient, economi-
cal and highly desirable to use laboratory-scale filtration
systems.
From the viewpoint of research, the more well-defined the
system the more accurate and applicable are the data generated.
From the practical standpoint, the use of a smaller system to
investigate specific problems has been shown to be efficient,
economical, and, with proper interpretations, reliable. The
proper interpretation is a point not to be minimized, since
the temptation is great to immediately extrapolate data
gathered under well-defined and controlled conditions at the
liters per day scale to predict responses for large waste
treatment facilities handling millions of gallons of waste
per day. Techniques have been evolved, however, in which
information generated from this scale operation may be. used
to make these full-scale predictions. If time is the limiting
factor for the development of a well-founded waste treatment
system, laboratory-scale investigations, if properly designed.
199
-------
operated, and interpreted, may produce data from which recom-
mendations may be made with more justification than simply
opinion.
It appears from the review of this literature that the most
common procedure has been to use laboratory-scale simulation
to develop ranges of performance and design criteria from
which pilot-scale operations may be developed. For a large
installation involving significant capital outlay, this
procedure obviously has merit. However, as larger quantities
of data become available and further basic understanding of
the processes is developed and disseminated, it is not beyond
comprehension to consider that laboratory-scale simulation
(not necessarily in the form of miniaturized plants) could be
used extensively to measure specific reactions and phenomena
required for design at various points in the waste treatment
plant. These laboratory data would be applied directly to
establish design relationships for full-scale sizing and
operation. Furthermore, by computerizing the pertinent
available data from the literature, the probability of elim-
inating the laboratory or even pilot plant step in the devel-
opment of a process for the purification of waste waters
could become a reality.
2OO
-------
SECTION XIV
PILOT-SCALE INVESTIGATIONS
Pilot-scale investigations may be considered as studies at a
large enough scale to be representative of phenomena experi-
enced by the prototype, yet be small enough to have some con-
trol and allow changes to be made economically. Stenburg
et al. (42O6), in a general review of wastewater treatment
approaches, outlined several pilot plant and full-scale
studies, and considered pilot plants as the modern method to
wastewater treatment. Procedures for selecting industrial
waste treatment processes were recommended by Stack et al.
(4158), involving local requirements, economic considera-
tions, plausible alternative treatment systems, with the final
decisions based on pilot scale operational data. As a con-
sulting engineer, Higgins stressed (1935) the value of pilot
plant data used in the design of waste treatment plants,
specifically those employing plastic media trickling filters.
Burgess et al. (536) determined the reliability of measurements
for testing treatment efficiency, and loading was evaluated,
based on pilot deep-bed percolating filters. The results
substantiated the Velz basic law. Statistical methods used
to design experiments and analyze data obtained from pilot
plants were emphasized by Blarigan and Lamb (352), and ob-
servations were made at as many combinations of the different
levels of each variable as possible, .rather than the usual
method of altering only one variable at a time. Typical of
pilot plant studies was that of Ingram and Edwards (2227),
where an 18-foot deep experimental filter was divided into
four sections and the ecological population and stratifica-
tion in each section of the filter bed determined.
Sponsoring agencies, such as the New Brunswick Research and
Productivity Council (533O), were active in the development
of pilot- and full-scale trials employing various methods of
treatment, such as percolating filters with plastic media to
treat trade wastes. With the development of new biological
trickling filter media, pilot plants became valuable in deter-
mining design criteria, such as contact time and efficiency
in removing specific organic pollutants (154). Hawkes and
Jenkins (182O) studied the comparative behavior of various
media trickling filters and the accuracy of the data gathered
from these studies, e.g., Bruce et al. (5O7). Studies on
specific media to justify their use were frequently performed,
such as the one by Althaus (6O), who used the bullrush Scirpus
lacustris in a pilot plant contact aeration device and demon-
strated a 95.7$ reduction in BOD. Another example was that
reported by Truesdale et al. (4463), in which eight rectan-
gular filters containing various inorganic media were studied
to determine the effect of the media on the behavior of the
percolating filter.
Operational changes have been demonstrated by pilot-scale
testing of existing waste treatment systems. For example,
201
-------
Luinb (553O) reported that the capacity of an existing waste
treatment plant was increased by a 1:1 dilution of the sedi-
mentation tank effluent with that from aeration tanks before
biological filtration. In this way, the load on the filter
could be at least doubled, and a superior effluent produced.
During pilot plant studies. Earth et al. (2O4) found that
the inability to control process variables effectively in
trickling filter treatment made more efficient nitrogen re-
moval by modifications of existing structures more unlikely
and favored the use of activated sludge. Pilot plant studies
were used to determine required facilities for plant expansion
and to define criteria from which a full-scale unit was de-
signed, e.g., plastic media pilot plant used to remove a mini-
mum of 5O$ BOD applied at 35O lb/l,OOO ft3/day (4O86). Oliver
and Hall (3241) and Culver (839) also used pilot plants to
develop data to determine the modifications required to ex-
isting facilities to improve the effluent to conform with
standards. Additional techniques, such as bacteriological
analyses, have been used by Sima (4O29) to evaluate the per-
formance of pilot scale biological filtration plants. A
neutron scattering technique to determine the amount of bio-
logical film in a pilot scale percolating filter was described
(1782) and the results from this method were reported and
discussed. Tertiary treatment of biological filtration ef-
fluent, such as deep well injection with various forms of
chemical and mechanical treatment, has been studied by
Hennessy et al. (1881) in pilot scale operations.
Petru (3342) and Bayley and Downing (226) determined the ef-
fect of temperature on the flow of air through percolating
filters based on pilot-scale filters operated over a signifi-
cant period of time. They concluded that cooling was affected
not so much by the external temperature as by factors affecting
aeration and by the characteristics of the individual filter,
such as the ratio of diameter to depth (3342) . The most im-
portant factor is the temperature of the applied sewage, but
the exchange of heat with air, especially near the surface,
and the production of heat by biological oxidation are also
significant (226).
Specific problems, such as to determine that recirculation of
humus to the influent reduced the amount of inorganic coagu-
lants required for efficient treatment of aircraft cleaning
wastes, have been attacked (3O) on pilot plant systems. Pilot-
scale determinations to verify laboratory results of retention
time tests were reported by Sheikh (3981) to be necessary in
the logical progression of the application of the developed
information to full-scale field design.
The effect of digester supernatant was investigated "by Sorrels
and Zeller (4113) on pilot plant experiments with percolating
filters. They observed that the primary filter appeared to
act mainly by flocculation and the secondary filter by oxida-
202
-------
tion of carbon and nitrogen. These pilot plant operations
demonstrated safely and economically that digester supernatant
liquor was amenable to biological filtration at loadings up
to 23 Ib of BOD/1,OOO ft3/day on primary filters and up to
92 Ib of BOD/1,OOO ft3/day on secondary filters. The effect
of hydraulic loading and depth on percent BOD reduction was
measured (4) on an 11-foot high pilot filter, 5-foot diameter,
and it was concluded that the highest BOD removal per 1,OOO
ft3 of filter media occurred at the 3-foot depth with a
hydraulic loading of 3O mgad (69O gal./ft2/day).
Pilot plants have been frequently used to obtain design data
for specific industrial wastes. Experiments were used by
McGaughey (2909) to determine the biological treatability
of "hard" and "soft" detergents and the data compared closely
with that from full-scale operating waste treatment plants
of similar types. Husmann et al. (2153) also evaluated, as
did many others, the biodegradability of soft detergents in
laboratory-, pilot-, and full-scale activated sludge and
percolating filter plants, which indicated that 8O$ of the
detergent could be removed by either process. Lawrence (2642)
determined the effectiveness of biological filtration in the
removal of low level radioisotopes from waste waters on a
pilot plant scale.
According to the results of pilot plant operations, activated
sludge was more effective than biological filtration in
treating tannery waste waters (3923). For textile wastes,
Snyder (4O94) found the effectiveness of plastic media trick-
ling filters sufficient to design a full-scale treatment
plant and assure a BOD reduction of 6O to 7O$. The effect of
pH on trickling filter performance was determined by Walter
(4592) by pilot plant studies on combined sewage and textile
waste. He found that pH values above 11 or 12 reduced BOD
removal rates. Three pilot plants were used simultaneously
by Munteanu (3124) to determine design parameters for full-
scale plants to treat waste waters from the manufacture of
cellulose and viscose artifical fibers, thereby reducing ob-
servation time and increasing flexibility. A pilot plant
used in Russia to determine the operating techniques to treat
waste from an artificial fiber plant was also designed to
operate under extremely flexible conditions (1393).
Noble et al. (3193) in 1964 summarized the activities of the
Gas Council since 1951, which sponsored pilot-scale experi-
mentation on activated-sludge and biological filtration proc-
esses to treat gas liquors (2296). Oil refinery (3205)
phenolic wastes were tested under pilot plant conditions to
demonstrate the applicability of biological filtration and
activated sludge processes. It was concluded that the shock-
resistant biological filtration process was preferred, and a
full-scale plant was constructed. Laboratory experiments
(3O18) indicated that certain pesticides were amenable to
203
-------
biological treatment, and pilot plants, with plastic media
filters and activated-sludge systems, were operated to deter-
mine design criteria. It was determined (3O18) that a full-
scale activated-sludge plant would reduce the phenol con-
centration by 99$ and the BOD by 9O to 95$. Black and
Fairall (342) reviewed in 1963 the pilot plants for develop-
ing operating data for the biological treatment or organic
waste waters by activated-sludge and percolating filter
plants. Pilot plant operations determined that the.activated-
sludge process was more suitable (5621) than percolating fil-
ters to treat chemical plant effluents.
Ratliff (35O8) described the advantages of pilot plant studies
on a paper mill effluent to consider the activated-sludge
process and biological filtration. Neither method was satis-
factory, and the effluent from a circulating lagoon was
treated by biological filtration to produce a satisfactory
treatment. A pilot plant designed to treat 10O m3/day (26, 4OO
gal./day) of pulp mill waste was operated by Ganczarczyk
(14O3); however, Schmidt (3873) found that dual treatment by
activated sludge and biological filtration provided a flex-
ible waste treatment facility, reducing the BOD by 8O$.
Similarly Nowaclci and Pilotek (32O6, 32O7) used a pilot
plant to treat the waste waters from the manufacture of kraft
pulp which was designed to include physico-chemical treatment
followed by biological treatment. After a twelve-month in-
vestigation on pilot plant-scale trickling filters, Richey
(3593) concluded that chipboard wastes were removed effec-
tively on a low rate trickling filter with the effluent
satisfactory for discharge. It was concluded {2993) that at
all hydraulic loadings from 1 gal./min (1,440 gal./day) to
7 gal./min (1O,OOO gal./day) the BOD removals exceeded 5O$
on a plastic medium in a pilot trickling filter with paper
mill effluent, and that supplemental nutrient such as anhy-
drous ammonia would be the most economical. Minch et al.
(3O24) and Ruus (3795) used pilot plant scale plastic media
trickling filters for the treatment of paper mill effluent
and indicated that the percent BOD and COD removal decreased
significantly with increased hydraulic loads.
Norgaard et al. (32OO) described pilot plant studies to es-
tablish design parameters for canning wastes which included
(a) anaerobic digestion followed by either activated-sludge
process or high rate biological filtration, (b) primary sedi-
mentation followed by either single- or two-stage high rate
biological filtration, and (c) the activated-sludge process
both with and without primary sedimentation. Based on
pilot studies, cost estimates for alternative plans and
expected removals were developed and presented. Food
processing wastes, such as that described by Leete and co-
workers (2669), have often been characterized by their
performance on pilot-scale operations used to develop de-
sign criteria, e.g., a recirculation ratio of about 4.5:1 on
204
-------
a pea cannery waste was required to produce an overall BOD
reduction of 78.8 percent. Chalmers (647) found from pilot
plant studies of activated sludge and trickling filters
treating the effluent from a complex food manufacturing and
coffee processing plant that the waste was more economically
treated by physical means. Sugar refinery effluent was
treated (4OO6) in a pilot plant by a high-rate percolating
filter to give a 7O$ reduction, which increased to 92 to 95$
after stabilization of the treated effluent in a 72-hour de-
tention holding tank. Pilot-scale evaluations were made by
Grau (1544), in 1961, on the effect of screening fruit and
vegetable wastes prior to biological filtration with and
without recirculation. He concluded that best results were
obtained without recirculation and a pH value of 6.6 to 6.9.
Adverse effects of recirculation are attributed to the lack
of nutrients. The temperature of air or waste water had
little effect.
Hirlinger and Gross (1959) found from pilot plant studies on
biological filtration that waste waters from a packing house,
after pretreatment, were treated satisfactorily if care was
taken to ensure a constant flow over the filter. Tests on
pilot-scale anaerobic digestion plants for treating slaughter-
house wastes were used by Silvester (4O27) as input so that a
full-scale plant was constructed with the effluent discharged
on existing percolating filters with no expected problems.
Pilot plant studies on the treatment of dairy waste waters
were carried out by Svoboda and coworkers (4272), who deter-
mined that the optimal BOD loading was 6OO to 8OO g/m3/day
(37-5O Ib BOD/1,OOO ft3/day). Other investigators, such as
Tidswell (4389), found from pilot scale treatment studies that
brewery wastes could be treated at 75 gal./yd3/day (2,775 gal.
of sewage/1,OOO ft3/day) to produce satisfactory effluents.
Critique
Pilot-scale operations were applied to several areas of re-
search and development for biological trickling filters. The
use of pilot-scale operations is a logical step between
laboratory and plant-scale operations, and investigators were
able to evaluate laboratory results and theories without in-
vesting large sums of money and capital equipment. Pilot
plants have the flexibility that, once constructed, changes
could be made without causing major design problems.
It should be emphasized that a major factor in the credi-
bility of the results from pilot plant operation is the
condition under which it operates. The example which was
most common was that of pilot plant operations to develop
design criteria for a specific trade waste. Pilot plants
of this type were established, usually, in the geographical
area where a prototype plant would be built and were sub-
jected to the waste in question for a significant period
of time. Once the investigator was assured that the pilot
205
-------
system had experienced the perturbations that the prototype
system would experience, the next task was that of correlat-
ing this information to be used for prototype applications.
On occasion, it was reported that the operation of a specific
pilot plant was a failure and that the waste was not removed
by the filter system. This, by definition of the use of
pilot plants, was not a failure because it saved the owner
considerable expense by not constructing the prototype plant
based on the evidence at hand, but only the costs for the
construction of the pilot plant.
Several companies have found it a necessary step in the de-
velopment of a market to provide their particular component
for pilot plant testing under the conditions which would be
expected after construction of the full-scale plant. An un-
fortunate by-product of pending legislation and the conserva-
tive approach of using pilot plants for an extended period
of time has been, in essence, a stalling action for pollution
abatement. Considerable data are available on the operation
and maintenance of biological trickling filters under many
conditions treating many wastes. These data, when supple-
mented with limited data of laboratory and pilot-scale
origin, would appear to provide sufficient information so
that full-scale design and construction would be more readily
pursued.
It is unfortunate, but understandable, that pilot plant
operations have not been used extensively for the development
of rational formulation describing the performance of the
biological filtration process. It would seem logical, follow-
ing the conservative lines used for construction of new
facilities, that mathematical models developed in the labora-
tory should be exhaustively evaluated at pilot plant-scale
and modifications of the model be made to adjust any dis-
agreements. It should be noted, however, that the use of
pilot plant-scale investigations to develop design criteria
does not need to be perpetuated, if basic understanding of
the processes involved is established and agreed upon at any
scale of testing.
206
-------
SECTION XV
FIELD- AND FULL-SCALE INVESTIGATIONS
Studies accomplished in the field and at prototype scale have
been regarded by many as the real "acid test" of a concept,
e.g., investigators such as Hall (1654) developed information
at pilot scale, but stress that proof must be given at full-
scale conditions. A good example of field studies was re-
ported by Benzie et al. (273), who evaluated climatic condi-
tions and loading factors on the performance of percolating
filters. The operational results of seventeen full-scale
installations in Michigan were compared and factors of filter
efficiency, BOD loading, recirculation, and ventilation rates
were evaluated.
Observations of field-operated systems have been used by Otto
(3259) to predict expected results of process modifications
that often result in considerable savings of time, money, and
damage to the environment. Field evaluations of the perform-
ance of existing plants were noted (4959) as useful to deter-
mine when reconstruction was necessary and to indicate the
type of plant which should be built. Ecological investigations
were conducted under field conditions by Chaudhuri (663) to
establish microbial population identity and occurrences in the
several unit processes of the treatment plant. However, most
industrial waste surveys involved, in addition to published
information and laboratory investigations, data gathered from
the field to assess the volume and strength of the waste
(3436) . A field survey of municipal waste treatment plants
was made (203) to determine the effect of metals on sewage
treatment, which confirmed pilot-plant studies that heavy
metals in concentrations of 1 to 9 mg/1 caused no serious re-
duction in efficiency of either aerobic or anaerobic waste
treatment processes. Barth et al. (204) studied five biologi-
cal treatment plants in Ohio to determine the efficiency of
nitrogen removal by municipal wastewater treatment plants.
From these field studies they concluded that where nitrifica-
tion occurred denitrification followed, suggesting an anaero-
bic process was possible for the removal of nitrogen from the
effluent.
Field comparisons on activated-sludge and biological filtration
systems for the removal of various pollutants, such as viruses
(2243), gave valuable information for system modifications. In
this case, chlorination of effluent to inactivate the virus
would be acceptable, for design. Valuable comparison data of
cost and effectiveness of activated-sludge and trickling
filter plants operated in parallel were reported by Dye (1O55),
based on a million gallons of sewage per day operation. Field
evaluations have often been used for full-scale installations
of package plants (5118) where units were presized, shipped
to the job site, and evaluated. Kehr and Mohle (2417) con-
ducted systematic experiments at various plants treating
2O7
-------
highly polluted waste waters to assess the efficiency of two-
stage biological treatment, namely, the activated-sludge proc-
ess followed by percolating filters. This procedure was typi-
cal of field-gathered information which was used for future
design and from which actual operating procedures were defined.
Typical procedures required for a large capital investment ex-
isting facility to be updated were related by Oliver and Hall
(3241). They obtained information on grading and depth of
medium satisfactory for a full-scale filter from a pilot-
scale operation in conjunction with existing high-rate filtra-
tion with recirculatio.i. Expected performance of a modified
plant was verified by Wilson and Harrison (4757) using labora-
tory- and full-scale studies, and they concluded that the ef-
ficiency of percolating filters was improved by recirculation
of the effluent. The improvement was explained to be due to
the effects of recirculation of organic matter which more
than compensated for the reduction of retention periods.
Laboratory- and full-scale experiments were used by Seidel
and Bauman (3957) to determine whether preliminary aeration
had a beneficial effect on biological filtration. The pre-
liminary treatment was advantageous when plain sedimentation
was the least effective and it was demonstrated on a plant-
scale that aeration for 45 minutes, followed by sedimentation
for one hour, gave consistently better results than plain
sedimentation alone for two hours. Based on the results of
prolonged full-scale tests of percolating filters, the design
of a system to handle 9O mgd (2O7 gal./ft2/day) and 1,OOO
pounds of BOD/acre/day (23 Ib of BOD/1,OOO ft Vday) , as well
as several other modifications to the existing Dallas, Texas,
waste treatment plants, was described by Graeser (153O).
Kastler (2383) reported on field investigations of enclosing
a trickling filter in a concrete and fiberglas structure
which resulted in efficient treatment and minimum maintenance
with no icing occurring during periods of sub-zero tempera-
tures at Northwood, Iowa. The alternating double filtration
technique was developed under full-scale operating conditions,
but it was based on the results of pilot-scale experiences
(3251) prior to full-scale construction.
A mathematical formulation of the biological oxidation proc-
ess had been proposed by McCabe (2895), which fitted the
laboratory and field data. Sheikh (3981) tied together
several studies on retention time using laboratory-, pilot-,
and full-scale experimentation with radioactive tracers.
He stressed that the performance data obtained under pilot
plant conditions were applicable to field-scale installa-
tions only if the two filters were operating under the same
conditions (depth, medium and hydraulic loading) . Field
installations, operated in conjunction with laboratory and
pilot plant studies to verify operational variables and aid
in this statistical interpretation, as in the case with
single- and two-stage trickling filter plants, were studied
in detail by Blarigan and Lamb (352).
208
-------
Pilot plant and field data were used by Germain (146O) to
develop an equation for predicting the performance of per-
colating filters containing plastic media for treating
domestic sewage containing only a small amount of industrial
wastes, from which further evaluations were to be made. The
National Research Council (559O) used the field observations
from several hundred sewage treatment plants to develop the
NRC formulation. Stanbridge (4175) carried out large-scale
experiments on sewage works to determine the effect of period-
icity of dosing on the performance of the percolating filters.
He concluded that controlling the accumulation of film in
low frequency dosing enabled high-quality effluents to be
produced throughout the year, while the winter performance
of filter treated differently deteriorated because of ponding.
General papers, such as that outlined by Albertson (27) , pro-
vided information on handling industrial waste based on field
investigations. Often the problem had been previously solved
and field studies were used to test their effectiveness so
that designed modifications and operations in the future would
correct the errors in past designs. Quite often field investi-
gations were in the form of plant trials such as reported by
Petru. (3341), who evaluated a biological filtration plant for
treating flax retting waste. An assessment of the pollution
load is a necessary requirement for a trade waste design,
as illustrated by Wheatland and Borne (4697) while investigat-
ing waste disposal at dairy farms. Occasionally, reports were
published which used laboratory-scale devices in conjunction
with field testing to determine the most effective treatment
system to be designed, as was the case of treatment of waste
from a hydrolysis plant carried out on an industrial filter
and on a glass model 1.5 meters (59 in.) high (1034). Con-
clusions were reached based on the glass model and correlated
with the performance of the industrial filter such that the
minimum height and contact time were established. Gerlich
(1459) reported that the operation of full-scale plastic
media filters for Cedar Rapids, Iowa, for treating sanitary
and meat packing wastes was better than expected from the
pilot model.
Pilot- and full-scale operation of plastic media biological
filtration towers was reported by Minch et al. (3024). The
operation of the treatment plant showed that the trickling
filters are capable of withstanding shock loads and severe hy-
draulic overloads while treating pulp and paper mill waste.
Design and operational requirements were developed to justify
an installation of this type and a full-scale plant was con-
structed. Pilot studies verified by field studies were used
by Chipperfield (681 ) to indicate the usefulness of plastic
medium trickling filters for single- and multi-stage plants
treating various industrial wastes.
2O9
-------
Large scale experiments have been frequently made by British
investigators to determine the treatability of various trade
wastes, such as reported by Blackburn and Tomlinson (345) in
1955 for the treatment of gas works liquor with biological
trickling filters. Field tests were performed (4285) to
establish an economical process by which combined domestic
and phenolic waste waters could be treated using treatment
plant elements such as settling tanks, rotating distributor,
percolating filters, and recirculation. However, laboratory-,
pilot-, and full-scale studies were used for the treatment
of gas works liquor by Drabek (1O24), from which it was con-
cluded that with careful operation the percolating filters
could achieve complete removal of the phenols from waste
waters.
Following pilot plant studies, field studies were used to
considerable advantage during the detergent degradability
studies of the early sixties, whereby effluent samples from
several existing trickling filter and activated-sludge plants
were taken to determine the degradability of hard and soft
detergents (29O9). It was concluded that linear alkylsulfo-
nate-type of detergent ("soft") is more readily degraded than
the "hard" type and where secondary treatment is provided it
should reduce foaming in the receiving waters. Typical of
this type of work was that by Jendreyko and coworkers (2271,
2272), who made full-scale investigations of detergent bio-
degradability in trickling filters and activated sludge by
adding known quantities of detergent to waste waters and
measuring the percent decomposition under field conditions.
Commercial detergents were tested by Husmann et al. (2153) in
laboratory-, pilot-, and full-scale biological filtration and
activated-sludge plant which indicated that 8O$ or more of
the soft detergents was decomposed and high concentrations
did not interfere with the performance of treatment plants.
Kumke and Renn (259O) specifically used a low-rate percolat-
ing filter which served an institutional home to determine
the biodegradability of linear alkyl sulfonate (LAS).detergents
and concluded from these field studies that the detergent was re-
moved as efficiently as naturally occurring organic materials.
Critique
The major contribution of the field and full-scale testing
program, as derived from the literature, was the high level
of credibility given the data generated. These data, which
were obtained under actual operating conditions, reflected
all of the variables of the environment plus sampling and
interpretation. When a clear relationship was discovered
among the variables, the information became extremely valu-
able. Theories which were developed under laboratory and
pilot plant conditions and verified by full-scale testing
have been well accepted as representing the phenomena.
210
-------
It should be noted that the expenses involved in developing
full-scale data are much higher than those of other scale
operations. Occasionally, data have been shown to be value-
less due to improper sample handling and delayed analytical
testing. Because of the many coexisting variables, basic
understanding and definition of the process mechanism are ex-
tremely difficult under full-scale testing conditions.
Full-scale testing and measurement have been used to advantage
in the development of modifications to existing waste treat-
ment facilities. Industrial waste surveys before and after
the construction of a waste treatment plant have been used
for many years to develop criteria for full-scale design and
operation. The conservative approaches of limited laboratory
investigations, extensive pilot-scale investigations, followed
by full-scale construction and operation are extremely valu-
able for the development of complex waste-treatment facilities
As may be inferred from the bibliography, full-scale testing
occupied a significant portion of the literature. The refer-
ences reviewed here are indicative of only a small fraction
of this material, but cover, in general, the performance
evaluation, ecological investigation, process modification,
and similar type studies.
211
-------
SECTION XVI
ECOLOGY
The trickling filter presents three elements which determine
its function and its population:
1) A trickling filter is designed to contain a ventilated
surface to which is applied a polluted substrate containing
required minerals and materials which microorganisms can
metabolize with net production of energy for growth or suf-
ficient energy TCO maintain the existing population.
2) The filter is always fed from the top and designed to
collect effluent at the bottom. The feeding may be continu-
ous or intermittent and the feed may include recycled liauid
and solids which have previously been passed over the fil-
ter.
3) In general a trickling filter is followed by a secondary
settling device which removes material converted to settle-
able solids by the combined forces of synthesis and/or bio-
coagulation in the filter.
The trickling filter is always used as a device for biologi-
cal oxidation, but a substantial portion of the purification
made possible by the filter is usually accomplished in the
sedimentation system which follows the filter.
Perhaps the most comprehensive descriptions of the popula-
tions found in trickling filters was presented by Cooke
(779). Cooke's listing of organisms reported to be found in
trickling filter populations is presented as Table VI, and
is given on the following pages.
Cooke (775) analyzed the interactions of the complex popula-
tions found in trickling filters in some detail and comment-
ed on their relationship to percolating filter beds, e.g.,
"binding" organisms which formed the biofilm, "free living"
organisms which fed on the bacteria, and "scouring organisms"
which fed directly on the biofilm and sewage solids and gave
additional information on technicmes and growth characteris-
tics.
Algae growing on the surface of trickling filters are usual-
ly an inconsecruential element of the population, limited to
illuminated surfaces, although the organisms have been
charged with responsibility for bed clogging (1826). Cov-
ered filters and filters with very limited illuminated sur-
face are at least as effective for treatment as are open
illuminated filters. Algal organisms on filters are clearly
tolerant of organic material and high levels of carbon diox-
ide (779) .
213
-------
Table 6
Composite List of Trickling Filter Organisms - Cooke
(Genera in major groups listed alphabetically
regardless of importance or systematics)
(779)
Schizomycetes (bacteria)
Sphaerotilus natans
Zoogloea ramigera
Many unidentified species
Cyanophyceae (Blue Green Algae)
Amphi thr ix j anth ina
Anacystis montana
Oscillatoria limosa
Phormidium uncinatum
Pleurocapsa sp.
Eumycophyta (Fungi)
Absidia corymbifera
cylindrospora
Allescheria boydii
Alternaria tenuis
Ascodesmis microscopicum
Aspergillus spp.
candidus
clavatus
flavipes
flavus
fumigatus
niger
ochraceous
sydowii
terreus
ustus
versicolor
wentii
Cephalosporium spp.
Chaetomium globosum
Cladosporium cladosporioides
Coniothyr ium spp.
fuckelii
Dematiaceae spp.
Epicoccum nigrum
Fusarium spp.
aquaeductuum
oxysporum
roseum
solani
Geotrichum candidum
Gliocladium roseum
Gliomastix convoluta
Leptomitus lacteus
Margarinomyces heteromorphum
Memnoniella echinata
Monilia sitophila
Moniliaceae spp.
Mucor sp.
fragilis
plumbeus
racemosus
saturninus
Myrothecium verrucaria
Paecilomyces varioti
Penicilliun spp.
brevi-compactum
chrysogenum
clavigerum
cyclopium
digitatum
expansum
funiculosum
herquei
implicatum
j an thine Hum
lilacinum
? luteum
martensii
nigricans
ochro-chloron
oxalicum
palitans
piscarium
purpureogenum
roquefortii
stipitatum
variabile
velutinum
vermiculatum
Phoma spp.
Pullularia pullulans
214
-------
Table 6 (Cont'd)
Penicillium spp. (Cont'd)
Rhizopus spp.
arrhizus
nigricans
rhizopodiform is
Rhodotorula spp.
Sphaeronema spinella
Sporotrichum pruinosum
Stachybotrys atra
Stemphyllium consortiale
Subbaromyces splendens
Syncephalastrum racemosum
Trichoderma viride
Trichothecium roseum
Verticillium spp.
lateritium
White yeasts
Chrysophyceae (Diatoms)
Nitzschia palea
Chlorophyceae (Green Algae)
Aeronema polymorphism
Characium sp.
Chlorella vulgaris
Chlorococcum humicola
Microthamnion kuetzingianum
Richteriella sp.
StigeocIonium nanum
Ulothrix tenuissima
Bryophyta—Musci (Mosses)
Leptodictyum riparium
Protozoa
Sarcodina (Amoeboid organisms)
Actinophrys sol
Actinosphaerium eichhornii
Amoeba proteus
radiosa
verrucosa
villosa
Arcella dentata
vulgaris
Centropyxis aculeata
Chlamydophrys stercorea
Cochliopodium bilimbosa
Difflugia pyriformis
Distigamoeba gruberi
Euglypha alveolata
Hartmanella hyalina
Protozoa (Cont'd)
Pomphagus nutabilis
Protamoeba primitiva
Trinoma line are
Vahlkampfia albida
guttula
limax
Mastigophora (flagellates)
Anthophysa vegetans
Bodo caudatus
lens
mutabilis
Cercobodo sp.
longicaudata
ovatis
Distigma proteus
Euglena sp.
Heteronema sp.
Mastigamoeba sp.
Monas sp.
Notosolenus sp.
Oicomonas socialis
Peranema trichophora
Tetramitus sp.
Trepmonas agilis
"Many small flagellates"
Infusoria (Ciliates)
Amphileptus sp.
Amphisia sp.
Aspidisca sp.
Blepharisma sp.
Carchesium sp.
Chilodon sp.
cucullus
Cinecochilum margaritaceum
Colpidium sp.
striatum
Colpoda sp.
inflata
Cyclidium sp.
glaucoma
Euplotes sp.
sharon
Glaucoma sp.
scintilans
Halteria sp.
Holophrya sp.
Lembus sp.
Lionotopsis sp.
Lionotus sp.
fascicola
215
-------
Table 6 (Cont'd)
Infusoria (Ciliates (Cont'd)
Loxodes sp.
Loxophyllum sp.
Macrymaria sp.
Metopus sp.
Microthorax sp.
Nassula sp.
Opercularia sp.
Oxytrichia sp.
pellionella
Paramoecium sp.
caudatum
Pleuronema sp.
chrysalis
Pleurotricha lanceolata
Podophrya sp.
Prorodon sp.
teres
Spirostonum sp.
Stentor sp.
Stylonichia sp.
pustulata
Uroleptus sp.
Uronema sp.
marina
Vorticella sp.
Metazoa
Turbellaria (Flat worms)
Stenostomum leucopus
Rotatoria (Rotifers)
Euchlanis dilatata
Philodina roseola
Rotifer vulgaris
Nematoda (Nematode worms)
Diplogaster strictus
Diploscapter coronata
Dorylaimus saprophilus
Rhabdites sp.
Gastrotricha (Gastrotrich worms)
Chaetonotus sp.
Oligochaeta (Round worms)
Aeolosoma hemprichi
Dero limosa
Limnodrilus sp.
Lumbricillus lineatus
Lumbricus rubellus
Pristina sp.
longiseta
Tubifex sp.
Metazoa (Cont'd)
Polychaeta (Polychaete
worms)
Polychaeta sp.
Crustacea
Cyclops sp.
Mollusca
Lymnaea humilis modicella
Physa anatina
cubensis
halei
Integra
Tartigrada
Macrobiotus sp.
Arachnida
Torrenticula anomala
Insecta
Cercyon fimbriatus
Diploneura cornuta
Leptocera fontinalis
Limonia simulans concinna
Psychoda alternata
Scatella paludum
stagnalis
216
-------
As a short term retention device the trickling filter would
not be expected to be am effective reduction device for
pathogenic microorganisms. This is indeed the case. Thus,
there are reports of limited reduction in the numbers of
Sal, typhosa, Sal, paratyphi, Mvco. tuberculosis (43, 23O,
2317, 2358, 3847). This is true also of pathogenic proto-
zoa such as Entamoeba histolvtica (812, 2552, 4829).
When loaded heavily with carbonaceous material, nitrifica-
tion in a trickling filter may be absent or nominal. How-
ever, the trickling filter when loaded lightly usually does
some nitrification. Applied to secondary effluent, the
trickling filter has been found to nitrify ammonia effec-
tively in studies now pending at The Dow Chemical Company.
Much of the early study of trickling filter populations was
aimed at the control of filter flies. These nuisance
organisms were controlled by flooding, chlorination, vari-
ous pesticides, etc. (62O, 1228, 1723, 1816, 1823, 1824,
1825, 1829,^2731, 274O, 3931, 441O). The filter fly per-
sists as a nuisance where low rate filters are used. How-
ever, high rate filters tend to have little fly associated
difficulty, except where feed application is very irregu-
lar.
The "ponded" filter occurs when strong wastes are applied
at low hydraulic rates (in terms of surface loading).
These difficulties are related to excessive amounts of
growth which clog rock filters and may be controlled by
higher flow rates to keep surfaces flushed.
The "sloughing" of the low rate trickling filter is a key
to its continuing effectiveness. The continuing presence
of an effective grazing population is believed to be needed
in order to prevent excessive build-up of attached slimes.
The use of insecticides, for instance, has been reported to
have been followed by "ponding" on occasion (1817, 1821).
Relatively large losses or releases of attached slimes may
occur during periods of high activity by worms or insect
larvae so that there may be either build-up or unloading of
slimes and retained sludges by trickling filters.
Since filter slimes can receive oxygen only at their free
surface, a thick slime may have a definitely anaerobic
interior and sulfur reducing bacteria have been isolated
from trickling filter slimes (3665). Under such circum-
stances filters can develop odors, and conceivably slough-
ing can occur as the result of gas generation in the interi-
or of the slimes. Sloughing occurs very frequently after
temperature changes in the spring and fall and has been
attributed to increased activity by worms and/or insect
larvae.
217
-------
The similarities and differences of the trickling filters
and activated sludges perhaps are worth some comment.
1) Because of the short retention time, trickling filters
show a somewhat greater resistance to various types of in-
sult than do activated sludge systems. This desirable fac-
et may also be related to lower adsorption capacity.
2) The trickling filter population in large part remains
attached to the filter. Accordingly, the age of the total
biota is older than that of activated sludge, even though
surface slime may have a very low age in high rate filters.
3) Whereas thready fungal or bacterial growth may greatly
impair activated sludge by causing reduced separation effi-
ciency in secondary settlers, such growth is ouite accept-
able in trickling filters.
4) The demands on the operator of the trickling filter are
low, and keeping mechanical devices used for distribution
working well is the larger part of continuing maintenance.
Activated sludge plants present more need for continuing
skilled operation and decisions.
5) Hold-up time in the trickier is low compared to the
aeration times used in the activated sludge process and
overall treatment efficiencies, in terms of "BOD" removal
will tend to be lower.
6) The slime or biological population of a trickling fil-
ter at any given time consists largely of attached material
as compared to the suspended and moving biological floe of
an activated sludge system. Accordingly, the waste moves
over and past the trickling filter slime, whereas the
sludge floe moves with the "mixed liquor." In consequence,
there is more opportunity for qualitative biological dif-
ferences in different portions of the trickling filter as
compared to an activated sludge unit.
Critique
The trickling filter is based on the metabolism of
attached bacteria and fungi and a predation or flushing
system which prevents excessive build-up of the primary
organisms to block the filter. A filter presents opportu-
nity for spatial variation in population vertically. There
is also opportunity in a heavily slimed filter for aerobic
populations at air-waste interfaces with anaerobic and sul-
fur reducing organisms below the surface layers of growth.
218
-------
As a Tcey to items relating to filter ecology, the following
listing of references is presented:
Reviews and Summations: 8, 44, 231, 254, 461, 566, 73O,
775, 779, 809, 987, 1228, 1251, 1253, 1471, 1817, 1821,
1825, 1826, 1828, 1912, 1916, 2O31, 2237, 2238, 2266, 2364,
2367, 2714, 2731, 2732, 2734, 28OO, 3126, 3168, 3432, 3575,
3576, 3699, 3712, 3714, 3715, 3759, 3868, 4O76, 4O83, 4266,
4384, 4403, 4429, 4473, 4579, 4641, 5456.
Growth Distribution: 173, 197, 523, 555, 777, 928, 987,
1612, 1820, 1821, 1825, 1826, 1829, 1914, 1916, 1917, 1919,
2442, 2628, 2713, 2715, 2742, 3144, 3721, 3824, 2903, 4384,
4396, 4404, 4405, 4414, 4429, 4648, 4672, 4677, 4821.
Algae: 2O9, 3O9, 343, 4O7, 653, 778, 779, 1158, 1213, 1735,
1757, 1821, 1826, 1987, 1994, 2O88, 2246, 2266, 3336, 3467,
3480, 3715, 4404, 44O5, 4473, 4717, 4923, 5173, 5577.
Bacteria: 43, 44, 88, 121, 185, 192, 23O, 249, 254, 461,
567, 582, 617, 664, 718, 729, 73O, 777, 779, 86O, 927, 97O,
1251, 1252, 1273, 14O1, 1522, 1545, 1676, 1682, 1683, 1826,
1912, 1987, 2005, 2O74, 222O, 2227, 2237, 2252, 2317, 2358,
2628, 2679, 3248, 3537, 3665, 3714, 3739, 3847, 3978, 3989,
4273, 4314, 4396, 44O5, 4473, 4541, 46O5, 4648, 465O, 4821,
4853, 56O5, 5608.
Fungi: 88, 197, 235, 254, 3O9, 343, 449, 776, 778, 779,
1252, 1253, 1665, 1735, 1821, 1826, 1829, 1831, 1904, 1987,
2100, 2170, 2227, 2237, 2266", 26O3, 3131, 3271, 3274, 3336,
3570, 3645, 3715, 3774, 3799, 4O47, 4274, 428O, 4375, 4396,
4399, 4404, 4405, 4473, 4939.
Protozoa: 88, 174, 175, 176, 177, 461, 779, 812, 832, 85O,
987, 1059, 1251, 1432, 1826, 1987, 2227, 2266, 2348, 2552,
2603, 2713, 2728, 3O17, 3144, 3248, 3384, 34O1, 3716, 3732,
3739, 4047, 4O5O, 44O4, 4544, 4579, 4677, 4821, 4829.
Insects, Worms and Miscellaneous Life Forms: 146, 252, 373,
461, 578, 579, 62O, 774, 779, 926, 987, 1228, 1365, 1475,
1480, 1492, 1723, 1815, 1816, 1817, 1819, 1821, 1823, 1824,
1825, 1829, 1869, 1914, 21O1, 2145, 2221, 25O6, 273O, 2731,
2732, 2733, 2734, 2735, 2738, 2740, 2754, 2755, 2894, 3131,
3143, 3170, 3258, 3338, 3425, 35O7, 3567, 3568, 357O, 3931,
4101, 4185, 4343, 441O, 4411, 4415, 4677, 4737, 539O.
Pure Culture: 88, 332, 45O, 567, 582, 1216, 1251, 1273,
1522, 1683, 1826, 3O74, 2262, 2611, 284O, 3158, 3764, 4O83,
4399, 4428, 4605, 46O6, 4852, 5124, 5166, 5456, 56O5, 56O8.
219
-------
SECTION XVII
PATENTS
An awareness of the significance of patenting various proc-
esses and modifications thereof dealing with biological trick-
ling filters was indicated (877, 4181) by discussions of the
significance of the developed technology. Throughout this
review, brief mention has been made of typical patents issued
on material pertinent to the section under consideration. It
is beyond the scope of this review to consider an in-depth
patent search, but patents may be regarded as an end-product
of research. Therefore, it was deemed a section prepared as
follows would be an efficient method of acquainting the
novice with past issued patents and provide the contemporary
practitioner with a reference source. By referring to the
bibliography, one may observe those references which were
noted to be specifically oriented towards patented informa-
tion. No comments or critiques were considered necessary
due to the specific nature of patent information which would
require the interested reader to delve into the patent more
deeply than may be done here.
For the purposes of organization and rapid review, the
following categories have been included.
Patents for the general sewage treatment by trickling
filtration are listed as follows:
338, 395, 542, 917, 1O04, 1156, 1165, 1509, 1672, 1675,
1772, 1966, 1968, 2OO9, 2173, 2376, 2424, 2618, 272O,
3268, 3452, 3481, 3561, 368O, 40O3, 4O21, 4455, 4456,
4602, 4819, 4867, 4868.
Items of interest for the pretreatment of sewage for
biological trickling filtration are found in the follow-
ing issued patents:
1677, 2164, 2354, 3366, 3678, 3679.
There were several patents issued on various procedures
for distributing the waste over the trickling filter
such as follows:
386, 387, 396, 496, 524, 78O, 1171, 1446, 1484, 1759,
1760, 1761, 1763, 1764, 1765, 1767, 1967, 2339, 2349,
2529, 253O, 2637, 2854, 349O, 3597, 3685, 3898, 434O,
4836.
The media used for biological trickling filters have
been covered in considerable detail under construction
materials, and patents were issued, typically, as
follows:
221
-------
90, 312, 463, 917, 94O, 965, 1O52, 1364, 1482, 1661,
1705, 1976, 2215, 2216, 223O, 235O, 2588, 2706, 2836,
2865, 2866, 2986, 31O7, 3438, 3495, 3895, 4265, 4361,
4362, 4454, 4603, 4763.
Problems arising from odor were very closely related to
ventilation and underdrains which gave rise to several
patents being issued, typically, as follows:
12, 169, 170, 369, 37O, 489, 914, 1212, 1472, 1595,
1597, 1598, 1599, 2172, 2325, 2588, 2694, 281O, 3476,
3523, 3822, 4O02, 4O97, 5637.
Treatment was not considered complete in most cases
without some form of post-treatment on biological
trickling filters which involved various forms of
clarifiers and sludge disposal techniques. In this
category, the following patents were issued:
1002, 1428, 1435, 272O, 3677, 3679, 3897, 4367.
Several investigators were unsatisfied with the per-
formance of biological trickling filters and suggested
process modifications using trickling filters as the
basis in combination with other unit operations or
changes in the standard design of trickling filters.
The issued patents are given in the following list:
141, 368, 544, 68O, 757, 781, 782, 866, 966, 1OO1,
1003, 1164, 1307, 1435, 147O, 1481, 1485, 16OO,
1622, 1841, 1854, 2155, 2267, 2514, 2633, 27O7,
2944, 2954, 2982, 3169, 3469, 3495, 389O, 3965,
4096, 4211, 4310, 4581.
Cleaning and maintenance of biological trickling filters
and their appurtenances required techniques of sufficient
uniqueness that patents have been issued, typically, as
follows:
398, 2632, 2694, 27O6, 2864, 3155, 4O98.
The above references were offered in this manner in the text
of the review because of the strong relation between research
and development and patents. It is suggested that with these
references and "chose provided in the discussion sections the
reader should ha able to know the patents in the field. Due
to the activity of investigators applying for patents, it is
to be noted that this partial list has been gathered only as
a basis and is by no means considered comprehensive.
222
-------
PART IV
APPLICATION OF TRICKLING FILTERS
TO SPECIFIC INDUSTRIAL WASTES
This section includes significant information on the per-
formance of biological trickling filters on several specific
industrial wastes which appeared frequently in the litera-
ture. The categorization of the wastes was established by
similarity of characteristics, trends noted in texts, and
groupings in the literature. Within each section, after
a discussion of the characteristics and occurrence of the
waste, there are the following subcategories:
1. Pretreatment required.
2. Efficiency of trickling filter application.
3. Comparison to other methods of treatment.
4. Post-treatment and effluent quality.
5. Special operational problems.
Each waste is separately critiqued, noting information on
past, present and future applications of the process. The
category of institutional and military waste was included
in this section because of its unique characteristics. A
few of the subcategories were condensed due to the redundancy
of reports, such as use of humus tanks. However, post-
treatment is reported with every application.
223
-------
SECTION XVIII
BREWERY AND DISTILLERY WASTES
Trickling filters have been used extensively for over fifty
years as a satisfactory method for treating brewery wastes.
In 1916, Emmerling (117O) recommended trickling filters be-
cause the management is simple and requires little attention
after the process is operational. A survey by Faber (1223)
of brewery wastes in Chicago showed BOD in the 12OO mg/1
range, suspended solids levels at 6OO mg/1, and nitrogen
averaging 5O mg/1. Efforts to coagulate this waste chemi-
cally were unsuccessful because of the high degree of soluble
BOD in relation to the concentration of suspended solids.
In most cases, brewery waste was blended with domestic sew-
age prior to treatment. A typical treatment plant included
grit removal, primary clarification, two trickling filters
in series with sedimentation between each filter, and final
clarification. Variations in treatment consisted of pre-
and post-chlorination, and substitution of an activated-sludge
unit for the second filter. Due to the easily oxidative
nature of the waste (4782), retention times in settling tanks
were held up to 1.5 hours (51O5) . Hydraulic loading rates
for domestic waste - brewery waste blends were reported to
run from 75 gal./yd3/day (1,35O gal./ft2/day), (Tidswell, 4389
4390) to 100 gal./yd3/day (1,8OO gal./ft2/day), (Moncur,
5265). Recently, Chipperfield (681) described three
years' experience of brewery waste treatment on both single
and multistage trickling filters constructed of plastic media.
The main waste byproduct from distilling operations was re-
ported as thin stillage. This material is acidic (pH 3.5 to
4.5) , had a BOD5 of 25,OOO to 45,OOO mg/1 and contained 3 to
5% solids (414, 4O71). Boruff and Elaine (415) claimed that
85$ of the stillage solids in the U.S.A. were recovered
as animal feed. Analysis of waste water discharged from an
Indiana distillery (3642) demonstrated that more than one-
half of the BOD resulted from process wastes. About 77$ of
this waste was due to the condensate from the quadruple-
effect evaporators used to concentrate the stillage. This
condensate had a pH of 3.8, a BOD of 685 mg/1, less than 1
mg/1 of nitrogen, no phosphorus, and 1 to 1.5 mg/1 of copper
from copper tubes and piping of the evaporators. Biological
filtration was a satisfactory method of treating the resulting
waste from stillage (415). Single stage filtration had been
employed, and Laurent (2635) documented satisfactory per-
formance utilizing double filtration. Stillage produced
from the distillation of wood waste contains unfermented
sugars, furfural, and organic acids and has a BOD of 6,OOO
mg/1 according to Bollen (392).
225
-------
PRETREATMENT REQUIRED
Various investigators have studied pretreatment schemes in
an effort to facilitate the treatment of brewery wastes
across trickling filters. Wisely (4782) related that the
wastes at the sewage plant in Highland, Illinois, were
aerated in a large hydraulic dosing chamber to keep them
fresh and alkaline. Similarly, according to Rowe (37O1),
lime was added to the wastes to maintain alkalinity, but
large quantities of sludge were produced. Raux and Maringe
(3511) and Moreau and Aubertin (3O89) reported on ferric
sulfate and lime being used for an identical purpose in
France, while the sewage works in East Ham, London, (5214)
used lime and chlorinated copperas. Oliver and Walker (3239)
tried to avoid adding lime and instead applied the waste to
filters with limestone media, but this proved unsatisfactory
because erosion occurred and the stone tended to knit to-
gether, thus reducing the available surface area.
Chlorine has been added to the primary filter influent for
the control of filter flies (3179). Faber (1223) found that
the intermittent addition of 4O mg/1 chlorine was adequate
to accomplish this control.
Trickling filters have also served as a useful pretreatment
device for other forms of treatment on brewery wastes, par-
ticularly the activated-sludge process (2O35, 4583). At
Chilton, Wisconsin, (489O) blended waste was screened,
settled, passed through high-rate filters, settled, aerated
for 4 to 6 hours, and again settled. Sudden changes in the
waste water caused by the brewery wastes were balanced by
passage through the filters, and the operation of the ac-
tivated-sludge plant was not affected.
Waste waters from distilleries have a high content of or-
ganic matter and are difficult to treat by an aerobic proc-
ess of sewage treatment owing to the high oxygen demand of
the liquors. Bloodgood (361) obtained satisfactory results
by anaerobic digestion of the waste, followed by treatment
on trickling filters. Roberts and Hardwick (3642) found it
necessary to add supplementary nitrogen and phosphorus to
the evaporator condensate prior to the application on the
filters. Maiz (2841) and Cosculluela (792, 339) found that
it was necessary to dilute distillery slops with three to
four times their volume of water, followed by sedimentation
and inoculation with ammonifying bacteria. The resulting
fermentation eliminated the putrescible fraction of the ef-
fluent and was amenable to biological filtration where the
ammonia would be oxidized to nitrite and nitrate in two-
stage percolating filter treatment. Buswell (564) used a
similar scheme except that the slop was fed to two digester
tanks in series and the digested effluent was diluted with
final effluent before being passed to a single trickling fil-
ter. Under some conditions lime was added to the acidic
waste before filtration (3693) .
226
-------
Distillery wastes from the manufacture of wine and brandy
pose slightly different pretreatment problems. Research by
Hodgson and Johnston (1969) indicated the addition of lime
and sedimentation. Sulfuric acid was added to adjust the pH
back to between 7 and 8 before the waste was two-stage fil-
tered. This effluent was then sent to a domestic waste treat-
ment plant. Soeiro and Lima (4O99) recommended mixing wine
and brandy waste with cooling water, adding lime, and then
settling. Hydrochloric acid neutralized the effluent prior
to processing by trickling filters equipped with recirculation
of effluent. At Sonoma, California, Vaughn et al. (4517)
noted that the brandy stillage was treated to remove tartrate,
the pH was adjusted upward, and then the waste was diluted
with condenser water and waste wash water prior to sewer dis-
charge. Tartrate removal was accomplished by the "hot throw"
method using calcium hydroxide and calcium chloride.
In the treatment of wood distillation wastes, experimentation
by Riley et al. (3616) proved that mild autooxidation was a
necessary preliminary treatment to reduce the toxicity before
biological filtration could function. Coagulation with lime
removed suspended solids satisfactorily.
EFFICIENCY OF TRICKLING FILTER APPLICATION
Good removal rates of soluble BOD, suspended solids, and
dissolved nutrients contributed by brewery wastes have been
attainable with trickling filters (1963, 4325). wisely
(4782) found that the hydraulic fluctuations caused by the
irregular dosing of the filter from a large dosing tank af-
fected the final efficiency- Investigations by Bushee (551)
proved that a brewery waste containing 37O mg/1 BOD5 and 340
mg/1 of suspended solids gave better results when treated
with two filters operated in series than when run in parallel.
Measurements of BOD showed that more of the impurity was re-
moved in the first stage than in the second. Krum (2575)
stated that complete treatment of a combined domestic brewery
waste in Allentown reduced organic nitrogen from 17.05 to 3.66
mg/1, ammonia nitrogen from 12.23 to 1.O5 mg/1, oxygen demand
from 129 to 31 mg/1, suspended solids from 186 to 35 mg/1,
and BOD from 213 to 29 mg/1. A Houston brewery (51O5) was
able to achieve 9O$ BOD and 75$ suspended solids reduction
through a two-stage trickling filter process. Serving as
pretreatment devices or roughing filters prior to activated-
sludge units, trickling filters assisted in attaining total
treatment efficiencies of 95 to 99$ of the BOD and of 75 to
9O$ of the suspended solids (2522, 489O).
The efficiency of trickling filters for the treatment of
distillery wastes was quite good, without preceding anaerobic
digestion, biological filtration was capable of achieving a
77$ reduction in BOD according to Roberts and Hardwick (3642).
227
-------
Initial removals were only in the range of 33%, however,
until the filter microorganisms became acclimatized to the
copper present from the evaporators and 3O mgad (690 gal./ft2/
day) were loaded at 2 Ib BOD/yd3/day (74 Ib BOD/1,000 ft3/day)
after some pretreatment. Davidson (878, 879) reported on a
Maryland distillery which produced a waste having a BOD of
7OO mg/1. Recovery of cattle feed after distillation and
passing the remaining waste material through high-rate trick-
ling filtration made BOD removals of 99.9$ possible. Sawyer
and Anderson (3839) ran 1$ rum waste through a two-stage trick-
ling filter plant equipped for a recirculation ratio of three
volumes of effluent to one volume of sewage. When the primary
filter was loaded at a rate of 2,59O Ib of BOD/ac-f/day (58.6
Ib of BOD/1, OOO ft3/day) , the average BOD was reduced from 485
to 19.7 mg/1. The overall BOD reduction of a wine and brandy
stillage-sewage mixture, as reported by Vaughn (4517), aver-
aged at 88$ for a three-year period.
Plastic media have also been investigated for processing dis-
tillery waste. Askew (98) treated waste from a distillery
containing 922 mg/1 BOD on two-stage plastic media packed
towers with the average BOD loads and removals at stages one
and two being 3.5 and 1.1 Ib/yd3/day (129 and 39.6 Ib of BOD/
1OOO ft3/day) and 68.6 and 78.4$, respectively. Treatment
of wood distillation waste on trickling filters at a
rate of 1 gal./ft3/day with aeration and recirculation ap-
peared to be the most desirable process with 5O$ BOD removal
(392).
COMPARISON TO OTHER METHODS OF TREATMENT
Several investigators (43 9O, 5214) have compared the relative
effectiveness of trickling filters for brewery waste treat-
ment with other methods. Both trickling filters and activat-
ed sludge have been used alone or together to process brewery
waste or combined brewery-domestic waste. Allen (51) report-
ed that Regina, Saskatchewan, Canada, used both processes,
with 6O$ of the flow treated by trickling filters and the
remainder treated by the activated-sludge process. Bushee
(551) asserted that undiluted brewery wastes were adequately
processed by series filter operation, whereas the contact
aeration process yielded unsatisfactory BOD reductions. Ad-
vantages of biological filtration over the activated-sludge
process, according to Jackson (2251), included more econom-
ical operation, less sensitivity to shock loading, and reduc-
tion of the cost of treatment by the recovery and sale of
animal feed. However, Newton et al. (3180) described
the conversion ot an existing trickling filter plant at
Frankenmuth, Michigan, to the activated-sludge process to
achieve a higher degree of treatment efficiency. The recent
advent of plastic media for biological filter packing has
stimulated evaluation studies by Askew (98) and Chipperfield
(681). Advantages cited are savings in capital costs over
conventional filters in activated-sludge plants in the order
of magnitude of 25 to 55$.
228
-------
Trickling filters were generally accepted in the treatment
of distillery waste when compared to other processes. David-
son (879) pointed out that the activated-sludge process was
not acceptable for treatment of these waste waters since
they contain some copper from the copper distillation equip-
ment and this copper accumulated in the sludge and had a
toxic effect on the microorganisms. According to Boruff
(414), treatment by the activated-sludge process was not
successful unless the wastes are first digested, then mixed
with sewage. Smith and Fargey (4O71), however, reported that
the total solids and the chemical oxygen demand could be
reduced by both activated sludge and biological filtration,
and still greater reductions were achieved operating the two
processes in series, e.g., COD reduced by 95$. Sawyer and
Anderson (3839) studied rum distillery wastes and discussed
the difficulties in treating such wastes by standard trickling
filters or by the activated-sludge process. One source (4925)
stated that, although an activated-sludge unit is being in-
stalled at a new distillery, biological filtration was the
predominant form of treatment. Advantages of plastic (PVC)
media for handling distillery wastes in comparison to con-
ventional filter media were also described (4965).
POST-TREATMENT AND EFFLUENT QUALITY
Trickling filters are rarely used as a post-treatment type
facility. Of the plants utilizing this concept. Artist (96)
was able to verify that the trickling filters at Wakefield,
England, which were preceded by the activated-sludge process,
did function well and with no filter flies. It was speculated
that most of the food normally available to these flies was
removed in the aeration tank.
Normally, ordinary sedimentation followed the trickling filter
operation in a system designed for handling distillery wastes.
Stevens (4211) described the separation of liquid from dis-
tillery slop by either settling or centrifugation and the
treatment of the liquid on a trickling filter before being
mixed with a reagent, such as calcium carbonate, to precipi-
tate an insoluble lactate which was filtered off. Part of
this filtrate or the original trickling filter filtrate, or
a mixture of both, may be concentrated and re-fermented.
Another form of post-treatment was reported by Schieck (3863).
He used a circular aeration tank for further treatment of a
tar distillery waste water after biological filtration.
SPECIAL OPERATIONAL PROBLEMS
Past experience demonstrated that brewery wastes when mixed
with sewage were satisfactorily treated with two-stage series
biological filtration (551, 2575, 5105). Moncur (5265) dis-
cussed the pilot plant operation of six filters in four ways:
229
-------
single filtration, single filtration with recirculation of
effluent, series (double) filtration, and alternating dou-
ble filtration. Tidswell (4389) followed up this research
with his own and came to the conclusion that double fil-
tration should be practiced in preference to alternating
double filtration. It was interesting to note, however,
that seven years later he reversed his position after ob-
taining operating data from the full-size treatment plant
(4391) . He also recommended that the order of the filters
be reversed every two weeks (439O) . Within the past ten
years, most trickling filter plants were designed to function
either as single, double, or alternating double filtration
units (4582, 5274, 556O). One source (5228) did mention
that alternating double filtration was used more frequently
if the proportion of brewery waste to sewage increased sig-
nificantly. Severe corrosion in a sewerage system resulting
from the presence of brewery waste rich in organic sulfur
compounds, which evolved as hydrogen sulfide from bacterial
action, prompted the use of plastic material for the con-
struction of pipes, channels, and sprinkler equipment at a
new biological filter plant (2264).
The treatment of distillery waste has not been without prob-
lems (539O). Blohm (357) recorded the difficulty encountered
in a Maryland plant when distillery waste water amounting to
1$ of the total domestic sewage flow was introduced into a
conventional single-stage trickling filter plant. Exces-
sive septicity occurred in a primary settling tank which
was caused by foaming in the digestion tank and overflowing,
followed by frothing over of solids onto the trickling fil-
ter. Normal operation was not resumed until the trade waste
was diverted from the treatment plant. Due to the high oxy-
gen demand of these wastes. Sawyer and Anderson (3839) have
advocated the installation of a separate piping system to
transport rum distillery waste directly to a municipal plant.
The mixing of this particular waste with sewage created ob-
jectionable anaerobic conditions in the sewer system.
Critique
Brewery and distillery wastes can be adjusted to be treated
adequately by trickling filters. Advantages of using trick-
ling filters, and specifically filters with plastic media,
were adequately reported. It appears that the technology
for treating this waste is well developed. Now, develop-
ments for automatic control, recoverable byproduct, and
other aspects are being studied.
The high oxygen demand of these wastes caused concern to
some investigators because low-rate filters developed nui-
sance conditions. A satisfactory approach was high- or
super-rate filtration in series or in series with other
systems. Anaerobic treatment ahead of trickling filters
found some application, but the attendant nuisance odors
230
-------
must be considered in the design. The reports reviewed
indicated the necessary pretreatment and post-treatment
required to make these wastes biodegradable. Acclimati-
zation was required, and some problems with toxic metal
contents were to be expected, but the wastes were generally
treatable by biological filtration to conform with the
required standards.
231
-------
SECTION XIX
ORGANIC AND INORGANIC CHEMICAL WASTES
Chemical wastes as used here are defined as the effluents
from installations in the chemical industry, and it is not
meant to cover the chemical characteristics of all wastes.
It was estimated in 1966 by Sadow (38O9) that $4O million
annually were spent in operating costs and 17OO full-time
personnel were employed by the chemical industry to fight
pollution. The types of wastes, their concentration and
volume vary widely from plant to plant. A large number of
the chemical plants separated their wastes, allowing un-
contaminated storm and cooling water to be discharged direct-
ly to the river, while process waste waters were treated by
various methods. The waste waters from petroleum production
were often odoriferous, highly colored, and had high organic
content (BOD). Waste waters from the production of dimethyl
terephthalate, in Russia, were turbid, odorous, and intense
yellow-orange, contained considerable amounts of aliphatic
acids, and had a high BOD (4OO1).
Pettet suggested (3355) that,by modifying manufacturing proc-
esses, it was possible to reduce the volume and strength of
the waste water, and savings in raw materials and water were
realized. He recommended that waste waters containing
organic substances be treated separately from those contain-
ing organic substances. In general, the inorganic wastes
are treated by chemical methods, and the recovery of these
materials may be of importance in the manufacturing processes,
while the organic wastes may be biologically treated. Many
chemical plants are so large and the waste so complex that
it is virtually impossible to separate the organic and in-
organic wastes. Harlow and Powers (1743) reported in 1947
that wastes from the production of 4OO chemicals in 5OO build-
ings totaled 2OO mgd which required disposal. About 125 mgd
of clear, warm condenser waters was discharged directly to
the river, but phenolic wastes, varying from weak (1.5 ppm)
to strong (7OO ppm), were separated and treated by separate
means.
Experiments were carried out by Brink and Thayer (462) on the
treatment of cyanide waste waters by biological filtration.
Preliminary results indicated that concentrations of free
cyanide as high as 15O ppm were removed and complex cyanides
containing metals such as copper were the most difficult to
treat. Liquid organic wastes containing formaldehyde and
methanol were successfully treated when combined with domestic
sewage and cooling water (5O47). The treated water was not
toxic to fish when discharged to the river.
The manufacture of synthetic fibers produced new waste treat-
ment problems. Sadow (38O8) described the treatment of wastes
containing acrylonitrile and zinc from the manufacture of
233
-------
acrylic fiber. Dryden et al. (1O39) treated the wastes from
a pharmaceutical plant manufacturing antibiotics, synthetic
vitamins, and cortisone by biological filtration. The treat-
ment of effluents from pesticide manufacturing was described
by Sharp and Lambden (3971). This plant manufactured dinitro-
orthocresol insecticides, copper oxychloride fungicides,
DDT insecticides, and a number of other weed killers and in-
secticides incorporating oils, solvents, and surface active
agents. These wastes, typical of many chemical plant ef-
fluents, were strongly bactericidal and were not amenable
to biological treatment in their raw state.
PRETREATMENT REQUIRED
Treatment of organic waste waters (4545) , such as from the
manufacture of pharmaceutical-type products and organic chemi-
cals, by biological filtration was feasible, but preliminary
treatment and pilot plant studies were often necessary.
Many chemical wastes required neutralization and equalization.
This procedure was often combined with primary sedimentation,
as reported by Sadow (38O8) for the treatment of a synthetic
fiber waste. The zinc from the fiber waste was precipitated
by coagulation with caustic soda and ferrous sulfate in an
upward flow tank, and the effluent mixed with sanitary sew-
age before it was pumped to plastic media trickling filters.
Dickerson discussed (951) the need for neutralization and
settling before biological treatment of the waste liquors
from the manufacture of a smokeless powder, which contained
decomposition products of nitrocellulose, alcohol, acetone,
ether, nitric oxides, nitric and sulfuric acids, waste from
dehydration, and some soluble organic materials. Powers (3451)
and Harlow and Powers (1743) stated that lagoons were used
for storage and blending of strong phenol wastes. Lagoons or
ponds were also used for cooling.
Loewnstein observed (2751) that an alcohol still residue and
spent gas liquor from a South African coal gasifier had a
temperature of 14O°F and contained up to 3O ppm of phenol.
This waste was passed through spray ponds to reduce the tem-
perature to 95°F before biological filtration. More recently,
Loewnstein and Waal (2752) reported that the same waste now
receives additional pretreatment by chemical coagulation using
ferrous sulfate and hydrated lime prior to biological filtra-
tion. Since the wastes from pesticide manufacture are highly
toxic. Sharp and Lambden (3971) described the preliminary
treatment as adsorption on activated charcoal, followed by
addition of hydrated lime, settling, and neutralization with
sulfuric acid. They further stated that wastes of this type
were mixed with settled sewage or a commercial biological
inoculum before treatment on percolating filters.
234
-------
EFFICIENCY OF TRICKLING FILTER APPLICATION
One of the more common wastes from the chemical industry is
phenol. Powers (3451) and Harlow and Powers (1743) discussed
the application of trickling filters to remove phenol dosed
at a rate of 1O.2 mgad (235 gal./ft2/day). The concentration
of the phenol applied to the filters was normally 3O to 5O
ppm and 4.29 to 9.O pounds of phenol was removed per day/
1, OOO ft3 of filter media when the temperatures of the waste
water were 56° and 83°F, respectively. At the same temperatures
the reduction in BOD was 15.8 and 27 lb/l,OOO ft /day, respec-
tively.
Sadow (38O9) reported that phenol removals of 99.9$ and COD
reduction of 8O to 85$ were obtainable by roughing filters
and aeration devices. The wastes from a synthetic resin plant
were applied to trickling filters at the rate of O.5 Ib BOD/
yd3/day (18.5 Ib BOD/1,OOO ft3/day), which allowed a margin
for the possible inhibiting effect of toxic substances (947).
Diammonium phosphate had previously been added to the waste
water to provide nutrients for the filter organisms and the
reduction in BOD varied between 79 and 95$. The content of
phenols in the crude waste waters reached 5O ppm and the ef-
ficiency of their removal varied from 9O to
The River Spree near Berlin, Germany, was protected from the
discharge of phenol from a tar distillery and chemical factory
by a two-stage biological treatment plant (4547). The plant
was equipped with a tower percolating filter and an activated-
sludge unit. The overall reduction in the concentration of
phenol was 95$, and the BOD5 was reduced by 67$. The percolat-
ing filter removed 1O$ of the phenol and 32$ of the BOD. Mills
(5354) discussed the operation of a pilot plant utilizing a
plastic media percolating filter and an activated-sludge tank
to treat phenolic and 2,4-D wastes. Effluent from the acti-
vated-sludge tank was recirculated to the percolating filter
and sludge was passed to a sedimentation tank and then re-
turned to the system. Although 99 to 1OO$ of the phenol was
removed, the 2,4-D waste waters (main constituent dichloro-
phenol) was more difficult to treat and 5:1 dilution was
necessary.
In 1959, Erlebach et al. (1188) showed that removal of fatty
acids from waste waters by slag filters was analogous to the
removal of phenols. The retention of fatty acids exceeded
the adsorption capacity of the slag, although the efficiency
of the filters did not decrease. Dickerson (948) reported
that an exceptional high-rate trickling filter was dosed at
5O mgad (1,15O gal./ft2/<3ay) with a recycle ratio of 4O:lwith
an influent containing phenol, formaldehyde, and resin oil
waste.
235
-------
A plastic media percolating filter was used by Harlow et al.
(1740) and Shannon (3968) to remove 9O$ of the BOD, 60$ of
the COD, and 95$ of the phenol from a petrochemical waste.
For loadings of formaldehyde up to 26O mg/1, Dickerson (948)
removed 2.59 Ib of formaldehyde per cubic yard of stone (96
lb/l,OOO ft3/day), and the overall efficiency of removal by
the filter was 23 to 40$. He recommended (948, 949) that
the pH be controlled between the 4.5 and 8.5 range. Organic
wastes containing formaldehyde (573) were passed through
percolating filters after mixing with process water and
domestic sewage.
Paleni described (3276) the treatment of waste waters from a
vegetable oil refinery and chemical plant which had a BOD of
5,OOO to 19,OOO and an average pH value of 1.5. The Ingram
controlled biological filtration process in packed towers was
used for these waste waters to produce an effluent BOD of
5O and a pH value of 7.O. Strong waste liquors (3,OOO to
1OO,OOO mg/1 BOD) from the manufacture of antibiotics, vitamins,
and sulfa drugs were evaporated and incinerated. However,
Reimers et al. (3551) reported that these wastes could be
combined with weaker wastes and adequately treated by high-
rate trickling filters. In the pilot plant, four- and six-
foot deep filters were used, with clarification necessary only
for the final effluent. BOD5 removal (8O$) was obtained in a
single-stage operation, and designed organic loadings of full-
scale units were reported as 1.9 to 2.O Ib/yd3/day (70 to
74 lb/1,000 ft3/day).
The waste waters from rosin extracting processes containing
resinic and fatty acids, polymerized terpene hydrocarbons,
and other organic substances had a BOD5 up to 13,OOO mg/1
(3663). These wastes were diluted with river water or domestic
sewage to give a BOD of about 4OO mg/1 influent for treatment.
In a typical experiment with the percolating filter, the BOD
was reduced to 46.8 mg/1.
To remove the color and BOD from wastes from the production
of pigmented lacquer materials, three slag filters in series
were used by Griinwald et al. (1613) . When the concentration
of organic matter in the influent changed, there was a direct
corresponding change in efficiency of the filters and in the
amount of organic matter retained. The slag filters lowered
the pH value of alkaline waste waters and raised the pH value
of acid waste waters, with the overall efficiency of the
three filters averaging ~32.7$. The waste waters from the
manufacture of synthetic rubber in Romania were reported by
Munteanu et al. (3126) to be treated in high-rate aerated
percolating filters after pretreatment. The filters operated
with hydraulic loads of about O.5 to l.O m3/m2/hour (295 to
59O gal./ft2/day) and five-day BOD loads of about O.3 to O.5
kg/mVday (18.7 to 31 lb/l,OOO ft3/day) . The addition of
domestic sewage was unnecessary, provided that phosphorus
was supplied in doses of 3 to 5 mg/1.
236
-------
Blarigan and Lamb (352) reported the results of a statisti-
cal interpretation of pilot plant studies which involved
single- and two-stage trickling filters. They found that
temperature had a marked effect on BOD removals at all load-
ings and recirculation ratios, and the effect was more pro-
nounced at higher recirculation ratios. The effect of BOD
loading was small in comparison to the effects of temper-
ature and recirculation. However, the effect of recircu-
lation was more pronounced with single filtration than with
two-stage filtration.
COMPARISON TO OTHER METHODS OF TREATMENT
Trickling filters have often been compared with activated-
sludge units, although the two processes may be combined.
Shannon described (3968) an installation which treated mixed
liquor from the activated-sludge unit. The filter gave con-
siderable oxidation and made it unnecessary to waste excess
sludge. However, in periods of shock loadings or excess
demand on the treatment plant, the filter was operated in
parallel with a second filter which preceded the activated-
sludge unit.
During experiments on TNT wastes, Schott et al. (3885) found
that difficulties in operation would occur in activated-
sludge plants where 5% or more TNT was in the waste waters.
However, a trickling filter, when dosed at the rate of 1
mgad (23 gal./ft2/clay) was capable of treating sewage con-
taining 1O$ TNT waste waters without appreciable decrease
in the percent reduction of BOD. When dosed at the same
rate with sewage containing 25$ TNT waste waters, the per-
centage reduction of BOD was reported to be considerably
decreased. Phenol-containing wastes were treated (4547)
by combined trickling filter and activated-sludge process
and 85$ of the phenol and 35$ BOD were removed by the latter
unit.
Reimers (3551) reported that strong waste liquors (3,OOO to
1OO,OOO mg/1 BOD) from the manufacture of antibiotics, vi-
tamins, and sulfa drugs were successfully treated by high-
rate trickling filters. Previously, the wastes had been
treated by other methods such as evaporation and incineration,
but the biological treatment improved the effluent emptying
into the river, and was more economical.
POST-TREATMENT AND EFFLUENT QUALITY
Trickling filters as roughing filters, followed by activated
sludge, are frequently used for treatment of chemical wastes,
and have been covered extensively in a previous section of
this survey. Several authors (952, 956, 1743, 3451, 38O9,
4547) have discussed the use of this approach of dual treat-
ment. A petrochemical plant in Texas (4971) reused the ef-
fluent from the dual system after sedimentation for cooling
purposes.
237
-------
Dickerson (948) noted that lagoons could be used if further
purification after the trickling filter was required. Liontas
(2723) pointed out that a pharmaceutical plant manufacturing
antibiotics and synthetic chemicals used intermittent sand
filtration after two-stage biological oxidation in high-rate
percolating filters and secondary sedimentation. Chlorination
was used following the sand filtration, and 98$ average over-
all reduction in BOD was noted.
SPECIAL OPERATIONAL PROBLEMS
The effect of pH on biological systems was well documented,
as confirmed by Walter (4592) during pilot plant studies on
the combined treatment of sewage and textile waste waters. A
roughing percolating filter removed about 60$ of the applied
BOD with loadings of 3,OOO to 6,OOO Ib/ac-f/day (69 to 138
Ib BOD/1, OOO ft3/day). The greatest removals occurred with
influent pH values of 8 to 9 and the reduction in BOD decreased
as the pH value increased to 11 or 12. Biological filtration
reduced the pH value of the waste when the initial value was
8 to 11, but above and below this range the filter had little
effect on the pH value of the waste. Little additional re-
duction in BOD was achieved by settling the primary filter
effluent and, since a second stage of biological treatment
was necessary for this waste, it was concluded that intermedi-
ate sedimentation was not justified.
Bucksteeg (522) observed that cyanide waste was only reduced
about 3O$ in percolating filters and this was due mainly to
aeration. When the filter medium was replaced with low tem-
perature coke and the concentration of cyanide was increased
from 1OO to 2OO rag/1, it was found that the effluent contained
only about O.I mg/1 cyanide. Further experiments indicated
that the cyanide adsorbed on the coke was destroyed within a
few days, which left the adsorption surfaces free to adsorb
more cyanide. He suggested that a process similar to this
with a two-layer percolating filter be used to treat municipal
sewage containing cyanide waste waters. The two-layer per-
colating filter is composed of a top layer filled with coke
and ceramic material to destroy the cyanide and a bottom layer
filled with a conventional medium for biological oxidation.
Dickerson (951) noted that, when the hydraulic rate of a
trickling filter was increased from 16 mgad to 39 mgad (368
to 897 aal./ft2/dav)fPsvchoda flies were no longer a nuisance.
Loadings at 8 mgad (184 gal./ft2/day) minimized ponding prob-
lems (951). An ecological study was made by Reynoldson (3575)
of a sewage containing a high proportion of chemical waste.
He found that L. lineatus was scarce and E. aldidus was more
abundant, but the total worm population was numerically much
less than found in percolating filters of other works. The
worms were fewest at the surface and the population was
relatively high in the area surrounding the cental distributor,
which could explain ponding problems.
238
-------
Hess (19O3) discussed the problem of acid waste waters and
oil in waste waters at a treatment plant employing perco-
lating filters. The waste was comprised of effluent from
food canning, processing of milk, manufacturing of cheese,
plating and pickling of metals, manufacture of dry batteries
and coolants, and oil-bearing degreasing wastes. The ef-
fects of oil were overcome by treating the filters with
emulsified ortho-dichlorobenzene and breaking up the surface
of the medium.
Ingols (2219) discussed the difficulties in sewage treatment
at a municipal plant caused by the presence of synthetic
detergents in the sewage. He found that the presence of
detergents interfered with sedimentation and resulted in over-
loading of the percolating filters. A pilot plant study was
made by Wilson (4756) for a chemical plant in North Wales,
and the breakdown of pure chemicals in experimental perco-
lating filters was.examined. Although acetic acid depressed
the rate of breakdown of phenols in the filters operating on
mixed chemical waste waters, the acetic acid was fairly
readily destroyed. A study of Jenkins (2281) on the bio-
logical oxidation of stearic acid in percolating filters
showed that sodium stearate was oxidized if nitrogen, phos-
phorus, and potassium were present. The oxidation occurred
more readily in the presence of domestic sewage, but at cer-
tain concentrations of stearate the filters might become
choked. However, these concentrations were greater than
that likely to be found in most waste streams. Munteanu et
al. (3126) noted that copper accumulated in the biological
film and caused difficulties in subsequent sludge digestion
during the treatment of waste waters. In the treatment of
effluent from pesticide manufacturing. Sharp (3971) estimated
the cost at|Ll.OO/1,OOO gallons ($2.50/1,OOO gal.) including
labor, where the treatment consisted of passing through
charcoal towers, treatment with hydrated lime, settling,
neutralization with sulfuric acid, and treatment on perco-
lating filters before final sedimentation.
Sadow (38O9) reported that the chemical waste treatment plant
at Texas City, Texas, had an annual operating budget of $75O,OOO
per year including all fixed charges, taxes, insurance, and
depreciation. Roughing filters and aeration devices were used
for biological treatment, with 99.9$ of the phenol and 8O to
85$ of the COD being removed.
Critique
Papers on the treatment of chemical wastes with biological
trickling filters were published frequently, and the investi-
gators demonstrated knowledge of physical, chemical and bio-
logical treatment techniques. The inorganic wastes were
usually treated chemically. However, only occasionally re-
ported, but of rather common knowledge, was the heavy reli-
ance on dilution to handle inorganic chemical wastes and
thermal pollution.
239
-------
Biological trickling filters have been of value for high
organic strength waste of varying character and flow. The
examples of chemical wastes reviewed using biological trick-
ling filters were a small part of the total literature. The
pretreatment required was frequently in the form of equali-
zation and suspended solids removal. Post-treatment was
usually clarification, additional oxidation and disinfection.
The toxicity of some chemical waste required deep-well dis-
posal of the treated effluent.
The chemical industry, as well as others, has apparently
spent much effort in time and money to render their waste
waters less objectionable to the environment. Unfortunately,
systems such as the biological trickling filter, even oper-
ating at high efficiency, are very expensive and one must
reflect if there is a better way and if the technology for
treatment of these wastes is really developed. If a plant
removes 9O$ of the BOD of an influent which has 5OO,OOO mg/1
BOD, an effluent of 5O,OOO mg/1 BOD would still be discharged
to the receiving waters. If one were to consider the total
pounds of BOD discharged per day to the receiving body rather
than concentrations, the performance of the sewage plants
becomes more alarming. The recent literature reflects these
concepts and engineers and scientists are cognizant of these
weaknesses and are endeavoring to rectify them.
24O
-------
SECTION XX
GAS AND COKE PLANT WASTES
The processes described herein and their associated wastes
deal with the production of coke by carbonization of coal
and the conversion of solid or liquid fuels to gaseous
products that are suitable either as a source of energy
or as a feed material for chemical synthesis. In the United
States, natural gas has largely supplanted manufactured gas
as a fuel source. In other countries, however, manufactured
gas is still the principal source of gaseous fuels. The manu-
factured gas industry can be subdivided into two distinct and
different industries: the production of fuel gases, and the
production of feedstocks for chemical synthesis. The gas
industry is closely related to the coke industry, since coke
is also a major raw material for gas production.
Fuel gas is produced by heating soft coal or oil with air
and steam, oxygen and steam, or hydrogen. The principal
products of this gasification are CO, H2 and CO2 in decomposed
steam; there are also small to moderate amounts of CH4, N2,
and traces of sulfur-bearing gases. These gases subsequently
are used directly for fuel purposes or as feedstocks for the
production of ammonia, methanol, or a variety of liquid fuels.
The production of coke, which is closely related, consists of
the carbonization of coal upon heating in the absence of air
at an elevated temperature, and driving off volatile products.
In all of these processes, the primary waste products result
from gas scrubbing and solid quenching operations. Consider-
able quantities of ammonia,phenol,cresols,sulfides,cyanides,
thiocyanates, various tars and other residues are discharged
in liquid wastes from vapor condensates from gas manufacture
and from coke quenching operations. These compounds have a
high oxygen demand, are toxic, and impart taste and odors to
the receiving waters. The removal of phenolic compounds from
these wastes has been of prime concern because of their low
detectable concentrations. Methods of treatment of these
wastes have included biological oxidation, through the use of
trickling filters or activated-sludge units, or various chemi-
cal methods, primarily oxidation, filtration, precipitation,
and extraction.
Consideration of the biological and ecological conditions
arising upon discharge of gas and coke plant waste into re-
ceiving waters was reflected in the literature. Meissner
(2961) made special reference to the removal of phenol and
ammonia from coal-associated wastes and gave account of the
biological conditions of a number of rivers affected by such
wastes. Bandt (161) discussed the origin and harmfulness and
the treatment of phenolic waste waters. A summary of the
241
-------
literature dealing with the volumes, composition, and treat-
ment of gas works liquor and effluents has been compiled
(4992) . Bondy and Priestley (400) reviewed legislation con-
cerning river pollution with special reference to the waste
waters from gas and coke works.
The general treatment of phenol-containing effluents was
considered in two early papers. Pruss (3470) and Lesperance
(2686) reviewed the biological treatment of phenolic waste
waters and made special reference to the need for fundamental
design criteria. Trickling filters were recommended for con-
sideration where treatment was required for large volumes
of warm waste waters containing low concentrations of phenol.
Drabek (1O24) investigated the use of slag filters for the
treatment of phenolic waste waters. Solin (41O5) purified
generator effluents on slag filters after preliminary de-
tarring. Zdybiewska and Bursztynowicz (4854) reduced the
phenolic content of the waste water using acclimated organ-
isms on blast furnace slag. Chipperfield (675) reviewed
the basic requirements for plastic filter media treating
these wastes. Noble (3194) presented graphical results on
the treatment of spent gas liquor on plastic filters. Three
waste treatment systems for waste containing phenolics from
chemical manufacturers, including biological treatment, were
described by Smith (4O83).
PRETREATMENT REQUIRED
Pretreatment of coke and gas plant wastes was primarily
dilution with domestic sewage or recirculated effluent.
Zdybiewska (4852) preconditioned blast furnace slag with
municipal sewage before introducing phenolic waste waters
from coking plants for treatment. Muns and Thompson (3123)
diluted coke plant waste with domestic sewage followed by
treatment at the municipal sewage works. Taubert (43O8)
stated that gas works waste waters were diluted with well
water or by recirculation with about 80$ of the filter
effluent. Lesperance (2686) recommended dilution of phenolic
waste waters to reduce the toxic effects upon microorganisms.
Slater (4O48) obtained satisfactory treatment of ammoniacal
liquor from gas works upon diltuion with sewage, clarification
on sand filters, and subsequent biological oxidation on trick-
ling filters. Hurley (21O5) diluted crude gas liquor seven
times with volumes of water. Oliver (3233) filtered crude
ammoniacal gas liquor wastes through clinkers after diluting
the waste with domestic sewage. Kabakova and Vil'gel'mov
(2361) stated that biological purification of phenolic waters
from coke ovens and gas plants in combination with domestic
sewage was advantageous, provided the relative amounts of
phenolic waters were small and the recovery of phenol was
uneconomical. Bach (118) purified phenol waste biologically
following dilution of the waste with final effluent.
242
-------
The removal of tars prior to biological treatment was a pre-
requisite, as suggested by Mihail and Popescu (2998) where
preliminary acidification followed by sedimentation and
filtration through scrap iron or coke was used. Phenols
were removed on a slag trickling filter. Drabek (1023) re-
moved tar in a sedimentation tank, followed by filtration
through a coke filter, and the subsequent spraying of the
waste over the surface of a slag filter. Abson and Todhunter
(7) described a three-stage process for treatment of coke
oven effluents as (a) phenols and higher tars were removed
by an activated-sludge process, (b) thiocyanate, thiosulfate
and cyanide were subsequently removed by biological fil-
tration, and (c) ammonia was then removed by nitrifying
bacteria in sludge tanks. Solin (4105) obtained improved
filtration of gas generator effluents on slag filters after
preliminary detarring of the waste water.
The preconditioning or acclimatizing of the biofilm on trick-
ling filter media with municipal sewage prior to the intro-
duction of phenolic wastes was recommended by Zdybiewska
(4852). Preconditioned blast furnace slag was used with
municipal sewage before introducing phenolic waste waters
from coking plants for treatment. Zdybiewska and Burszty-
nowicz (4854) conditioned blast furnace slag with municipal
sewage for ten days and gradually increased the amounts of
phenolic wastes. Nellist (3158) noted that coke works
liquor could be treated on trickling filters if combined
with domestic sewage. Additional filter capacity was not
required (3158), with only slight deterioration of effluent
quality. Norwood and Schaeffer (3205) recommended passage
of phenolic waste waters through retention ponds before fil-
tration. Barritt et al. (198) described the pre-use of
effluents from condensate plants in a coal washery and froth
flotation plant. Final reduction of the phenol was obtained
by biological treatment in an activated sludge plant, and
trickling filters were cited as a possibility. Hsu et al.
(2074) obtained partial removal of phenols upon pretreatment
of the waste through storage and aeration, but most of
the removal was obtained by subsequent trickling filtration.
Coke and gas plant wastes have also been preconditioned for
biological treatment by chemical extraction of the phenolic
compounds. Johnson (2325) purified a phenolic waste by first
extracting the phenol, aerating the liquid, and finally sub-
jecting the waste to biological treatment on a trickling
filter. Pruss (3472) removed most of the phenol from a coke
plant waste by extraction. The unextracted phenol was
destroyed biologically by acclimatized filter beds or activated
sludge. Key and Etheridge (2464), however, found no advan-
tage in the pretreatment removal of phenol from spent and
ammoniacal liquors before treatment with municipal sewage
on trickling filters or activated sludge. Putilina (3487)
removed phenol, sulfur, and cyanide compounds from undiluted
243
-------
waste water by means of special bacterial cultures followed
by biological filtration. Reich (3532) used pretreatment
which consisted of neutralization, sedimentation, and phenol
extraction, with the phenolic effluent combined with domestic
sewage and treated with a biological filter.
EFFICIENCY OF TRICKLING FILTER APPLICATIONS
Biczysko (321) reduced the phenol content after first-stage
filtration by 56-70$. Bryan (516) also reported a 50-75$
reduction in a first-stage treatment and an overall effi-
ciency of 95$ in combination with activated sludge. Noble
(3194) obtained an efficiency of 93-95$ from a liquor con-
taining 3,OOO-5,OOO mg/1 total phenols in a two-stage
treatment. Jenkins et al. (2296) obtained a higher treat-
ment efficiency for domestic sewage combined with coal
carbonization plant effluents than with domestic sewage alone.
Phenol was reduced from 1,19O-3,15O mg/1 to 0.2-5 mg/1 in a
two-stage biological process. Porter and Dutch (3437) ob-
tained 95$ reduction in phenol and 9O$ reduction in cyanide
from coke oven waste using a plastic trickling filter medium.
The efficiency of treatment as a function of the initial
concentrations of components was observed by Zdybiewska
and Bursztynowicz (4854) where a reduction in phenolic
content of 98.6$ was obtained on blast slag. This waste
contained 1OO mg/1 of total phenols in combination with
municipal sewage. Drabek (1O22) treated waste waters from
gas works containing 843-951 mg/1 of monohydric phenols
and 353-392 mg/1 dihydric phenols and found no phenols in
the filtered water during the first six months of operation.
Eventually, 11.7-172 mg/1 of total phenols were found in
the effluent in the following nine months of operation.
Specific active strains of bacteria removed 96$ of the
phenolic compounds from a gas condensate waste containing
2,OOO mg/1 of phenolic compounds, according to Husmann and
Malz (2154).
Some authors had reported effective treatment on undiluted
gas and coke plant wastes, but most agreed that some pre-
liminary dilution was required for effective treatment.
Gwilliam et al. (1635) found that concentrations of 2O-4O
mg/1 phenol were completely destroyed during biological
filtration, but an upper limit of phenol concentration existed
which could be treated satisfactorily. The actual limit was
not determined, but a filter charged with 2OO mg/1 phenol
became choked with black septic matter and could not oxidize
the phenol completely. Biczysko (32O) conducted experiments
to verify the efficiency of deep bed filters and the oxidation
available for the treatment of phenolic waste waters.
244
-------
COMPARISON OF TRICKLING FILTERS WITH OTHER METHODS
Sherwood (3994) reviewed and compared the treatment of phenolic
waste waters from gas producing plants by the activated-sludge
process and by the trickling filtration process. Bischofs-
berger and Wurm (335) described the treatment of phenolic
waste waters by biological processes using activated sludge,
trickling filtration, and both processes in series. Bandt
(161) described the treatment of phenolic waste waters by
both trickling filtration and by activated sludge. Velek
(4532) described various methods of removing phenols from
waste waters, e.g., biological filtration, ion exchange, and
adsorption. Knop (2525) stated that residual phenol in coke
works waste waters may be destroyed by treatment in either
trickling filters or activated sludge. The advantages of
plastic trickling filtration media for treating phenolic
wastes over other materials (crushed stone, clay blocks, or
Raschig rings) were discussed by Bryan (514). Husmann and
Malz (2154) found that phenol-containing sewage was success-
fully treated by both trickling filtration and activated-
sludge processes. The best treatment was accomplished with
the aid of very active strains of bacteria together with long
aeration.
There was disagreement in the literature as to whether the
trickling filter or activated-sludge process was preferred
for coke and gas plant waste treatment. Noble et al. (3193)
found that the power requirements for the activated-sludge
treatment of gas liquor wastes, as in municipal works, were
more than for packed towers. A longer period of acclimation,
less space utilized, and less sensitivity to change were
attributed to the packed towers (3193). Norwood and Schaeffer
(32O5) agreed that the activated-sludge process was more
sensitive to shock loadings and variations in feed concen-
tration, whereas trickling filters were considered (4948)
much more resilient to sudden changes in loading. An
effluent containing 2O$ of the"devil" liquor from a coke oven
operation was treated satisfactorily with trickling filters,
while only 2% was treatable using an activated-sludge process.
Investigations (4947) have shown that percolating filters
could be used instead of aeration tanks and were more suit-
able for certain of these trade effluents.
In some instances, the activated-sludge process has been pre-
ferred to trickling filtration as a result of comparisons
at various flow rates of coal carbonization waste waters
(3622). In another instance, effluent from a coke works was
successfully treated by the activated-sludge process but an
enclosed trickling filter was not successful (5O33). Nellist
(3158) reviewed previous studies and stated that the acti-
vated-sludge process was successful in treating coke works
wastes but that percolating filters treating the same
wastes were observed to age. This aging was apparently a
245
-------
result of choking of the filter medium with sludge and not
due to a filter for growth development. This author reported
that trickling filters were successful in treating up to 2%
coke waste liquor in combination with sewage. Prliss (3470),
in an earlier paper, described the treatment of phenol-con-
taining effluents by sprinkling filters and by activated
sludge. Costs for the sprinkling filters at that time were
considered too great. The activated-sludge process was
reported to be better suited for the varying composition
of the phenolic waste.
Chemical methods for the destruction of toxic components in
coke and gas plant effluents have also been described by
Abson and Todhunter (6). These included oxidation by ozone,
nitrous or nitric acids, chlorine, treatment with formalde-
hyde, lime, or sulfuric acid, and ion exchange. Micro-
biological processes,including trickling filtration and
activated sludge,were compared with chemical methods.
Savage (3832) described studies on the treatment of ammonia-
cal waste waters by biological filtration and by oxidation
with chlorine dioxide. Neither method was entirely satis-
factory and a countercurrent extraction of phenols was
employed, followed by distillation of ammonia before dis-
charge.
Bulicek (528) reviewed methods of removing phenol from waste
waters which included the preparation of fertilizers, com-
bustion, evaporation, use in coke-quenching and coal-
washing, sedimentation, filtration, chlorination, adsorp-
tion on activated carbon, steam distillation, extraction
with hydrocarbons, oxidation, and precipitation. Bulicek
reported that the most common methods for treatment were
oxidation, filtration, precipitation, and extraction.
Sherwood (3994) also reviewed existing methods for the
treatment of phenolic waste waters from gas producing
plants. These (3994) included air flotation, evaporation,
and catalytic oxidation. The intense taste and odor pro-
duced by phenolic compounds were recognized (3994) as a se-
vere problem. Abson and Todhunter (6) described the re-
covery of phenols from carbonization effluents by three main
methods: distillation and absorption or adsorption of phe-
nolic vapors, direct adsorption, and solvent extraction.
POST-TREATMENT
Post-treatment of coke and gas plant wastes initially treated
by trickling filtration has usually consisted of a second-
stage biological treatment and sedimentation to. remove
sludge. Bryan (516) described two-stage biological treat-
ment of phenolic waste: a first-stage trickling filtration
was followed by a second-stage activated-sludge process.
Meissner (2961) also used staged treatment of coal waste
containing phenol and ammonia. The final stage was bio-
logical treatment by trickling filtration and activated
246
-------
sludge. Malz (2855) recommended a two-stage trickling fil-
tration or activated sludge process in which the stages
were operated at different temperatures (18° and 55° C) con-
ducive to the growth of mesophilic and thermophilic organisms
Studies were carried out on the removal of the "intractable
residue" which remained in coke and gas effluents after
biological treatment. This residue was removed by passage
of the effluent from biological treatment through a bed of
suitable ion exchange resin (5607). In the treatment of
coke oven effluents described by Abson and Todhunter (7),
phenol and tars were removed in an activated-sludge process,
thiocyanate, thiosulfate and cyanide by trickling filtra-
tion, and the ammonia was subsequently removed by nitrifying
bacteria in sludge tanks. Suszka (4270) investigated the
purification of coke plant wastes containing phenolic com-
pounds by primary sedimentation and trickling filtration
followed by pond oxidation, secondary sedimentation, and
sludge recirculation.
SPECIAL OPERATIONAL PROBLEMS
Blackburn and Kershaw (347) described the treatment of gas
works and coke oven liquors in admixture with sewage on
trickling filters. Considerations of the volume and height
of the packed filter to reduce the "end" and "wall" effects
were discussed by Noble (3194). DeLaporte (921) emphasized
the segregation and separation of phenolic waste for treat-
ment to prevent pollution by the waste water.
Noble et al. (3193) observed some difficulty in settling
biological solids in the effluents of gas liquor wastes
treated with trickling filters. This problem was overcome
by adding chlorinated copperas or by returning settled
sludge to the aeration tanks. Stewart (4217) stated that
the presence of phenol reduced the proportion of BOD removed
in settling tanks. Phenol remained in solution but was de-
composed under aerobic conditions in a trickling filter and
no phenol reached the sludge digestion tanks.
The concentration of pollutants and the degree of dilution
required as well as other operational parameters were
considered by Kucharski (2583). Porter and Dutch (3437)
reported that sulfides in concentrations up to 50 mg/1 had
a negligible effect upon the operation of plastic trickling
filters used in treating coke oven wastes. Hsu et al. (2074)
emphasized the necessity of supplying inorganic nutrients
as well as maintaining optimum oxygen content to promote the
vigorous growth of organisms that degraded phenolic wastes.
Biczysko (321) noted that the efficiency of treatment of
phenolic waste waters on trickling filters decreased with
increased temperature up to 43-46°C, then increased to a
247
-------
maximum at 55-6O°C, and decreased again at higher tempera-
tures. Taubert (43O8), however, recommended an optimum
temperature for treatment of gas works effluent in the
range of 20-25°C. The effects of temperature, loading rate,
biological growth, and nitrogen balance on the rate of phenol
removal were observed by Malz (2855) . Husmann and Malz
(2151) discussed the importance of the growth of mesophilic
and thermophilic organisms for biological filtration of
phenolic wastes.
The use of slag filters stimulated considerations of adsorp-
tion and biological oxidation. Solin et al. (41O6) observed
that some cresols were adsorbed on slag in a manner similar
to the adsorption of phenol and pyrocatechol. These ad-
sorbed phenolic compounds were then subsequently oxidized
into substances which condensed to form humic acids. Solin
and Erlebach (41O3) stated that the adsorptivity of phenolic
compounds onto slag depended upon the load and period of use
of the filters. Solin et al. (41O2) also described the satis-
factory performance of slag filters even without sufficient
aeration to oxidize phenols to form humic substances. Kres
(2568) indicated that activated carbon was a suitable material
for purification of phenolic wastes for special cases such as
drinking water supplies.
The character of the microbial populations growing upon trick-
ling filters has been described in other sections of this
review. Zdybiewska (4853) observed that the population of
living organisms was greater in the upper layers of trick-
ling filters and increased with increasing concentrations
of phenols in the waste treatment. Palaty (3274) inoculated
a trickling filter with the fungus Oospora and used this unit
to treat generator waste waters containing phenols. Jenkins
et al. (2296) observed some adverse effects upon nitrifica-
tion and the responsible organisms following treatment of
coke carbonization plant effluents by biological filtration.
Critique
General observations on the operation of trickling filters
for the treatment of coke and gas plant wastes have been
primarily concerned with the nature and the concentrations
of the components of the waste treated, the temperature of
operation, and the effect of trickling filtration upon
subsequent treatment processes, such as adsorption and
nitrification.
Satisfactory treatment by trickling filter and activated
sludge individually and together was reported to the extent
that a clear choice is not obvious. A major difficulty is
that most of the literature dealt with "phenolic wastes"—
248
-------
a term which is not really very descriptive. Toxicity
factors were present in some waste streams and absent in
others. A comprehensive piece of work outlining all the
ramifications of gas and coke plant wastes has on occasion
been undertaken, unfortunately, for a specific plant, location,
or agency. It would appear from this review that one may use
trickling filters, after some pretreatment, to purify these
wastes. If a high quality effluent is required, staged oper-
ation should be used whether it may be another filter or
activated sludge. The economics of a particular situation
may aid in making the decision. Systems other than biological
treatment have been fully explored and evidently found economi-
cally inferior.
249
-------
SECTION XXI
FOOD PROCESSING WASTES
Food processing wastes as used in this section are defined
as wastes from edible food processing other than meat wastes.
A history of the use of biological trickling filters and
other methods for the treatment of canning waste was published
by Murphy et al. (3133) in 1928. Food industry wastes were
classified by McLachlan (293O), in which it was generalized
that these wastes were treated by screening, sedimentation,
chemical treatment, and biological fermentation. Typically,
the reviews and symposia which have been held to deal with
the problem of food processing waste waters (4956, 5589)
involved various conventional and novel methods of biological
treatment. Hemens (1875) and Loesecke (275O) have character-
ized food processing wastes and reported treatment methods,
including biological filtration.
Alternative methods available to the designer and contractor
of waste treatment plants handling food processing wastes
were published in 1931 by the New York State Department of
Health. Heider (1855) reviewed the canning industry in the
state of Indiana and found that there were 235 installations
in 1945 which produced waste water having an organic content
with the equivalent of ten times that of domestic sewage.
The seasonal operation of the canning industry has been a
continuing waste treatment problem (1855, 4332). Trade groups
such as The National Canners Association outlined the proper-
ties of waste waters from canners and reviewed earlier methods
of treating these waste waters (4619).
The facilities for the treatment of corn and soybean wastes
of the Decatur plant of A. E. Staley Company, which was
reviewed by Willenbrink (4734), had a population equivalent
of 22O,OOO to 23O,OOO per day. Imhoff tanks, percolating
filters and lagoons plus contact-stabilization plants were
used for treatment. Pearse and Greeley (3319) used an
experimental sprinkling filter plant to handle corn products
and gas works waste. A typical problem at many food pro-
cessing effluent treatment plants was maintaining adequate
flow during off-seasons, while at the same time having
sufficient capacity for the canning season (531, 5646). The
high-rate biological filtration plant in North Davis County,
Utah, described by Templeton (4338), maintained satisfactory
performance in spite of the variety of cannery wastes being
treated. The wastes from cucumber (2481) or pea and bean
(162) canning and dehydration of various foods (924) were
all biologically treatable, usually on recirculating high-
rate filters.
Literature in I960 by the U. S. Public Health Service (5594)
characterized several food processing wastes, such as potato
251
-------
chip wastes. Characteristics dealing with pH, population
equivalent, solids and permanganate values were reported by
DeMartini et al. (924) for several vegetables and milk product
materials. Buzzell et al. (572) characterized potato starch
waste. Typical food and fermentation waste strengths were
tabulated by Dlouhy and Dahlstrom (983) and advantages of
primary, secondary, and tertiary treatment were considered.
General aspects of industrial waste treatment and uses of
trickling filters for treating waste from canneries were out-
lined by Eldridge (1141), where recirculating trickling fil-
ters, loaded with over 1 Ib BOD/yd3 of medium per day (37 Ib
of BOD/1,OOO ft3/day) at the rate of 15 to 2O mgad (345 to
46O gal./ft2/day), achieved the desired BOD removals.
Loesecke et al. (2749) reported on experimental treatment of
citrus cannery effluent, where the influent BOD varied from
10O to 1,4OO mg/1 and contained inorganic salts, organic
acids, pectin, and sugar with very low nitrogen content.
Domestic sewage flow of 75O,OOO gal./day and about 3OO,OOO
gal./day of waste waters from processing fruit and vegetables
were reported to have a BOD varying from 22O to 1,23O mg/1
and to contain 3O to 363 mg/1 suspended solids (1952). The
BOD of a fruit processing plant making apple butter and
preserves was reported by Hommon (1991) to be ^OO mg/1.
Tomato canning waste was described by Hert (19OO) to have a
BOD of 548 mg/1, 1,341 mg/1 total solids, 275 mg/1 suspended
solids, and 1,064 mg/1 dissolved solids, while pumpkin
canning waste had an average BOD of 4,6O6 mg/1 to contain
1O,6O9 mg/1 total solids, and 1,1O3 mg/1 suspended solids
(19O1). The characteristics from a vegetable and fruit can-
ning factory were reported by Miller and Clark (3OO9) to be
5O,OOO gal./3O tons of vegetables/day, to have a high pH,
and to have a tendency to pond trickling filters. The wash
waters of the first washing from production of dried vege-
tables were large in volume, but not very polluting, and
after sedimentation the BOD usually ranged between 1O and
6 mg/1. Waste waters from the peeling process had a BOD of
3OO to 2,OOO mg/1, and from the blanching process had a BOD
of from 5,OOO to 15,OOO mg/1, with their volume usually less
than that of waste waters from preliminary washing or from
the peeling process, according to Jones (2337).
Successful use of biological filtration after screening and
sedimentation was reported by Pettet (3347) on grain process-
ing waste waters which were characterized to have a maximum
of 4,47O mg/1 BOD and 2,190 mg/1 COD. The usual waste
volumes were from 47.5 to 175 gallons per quarter of wheat
washed with clarified BOD of O.39 to O.48 pound per quarter
of wheat (3347) . An edible oil refinery effluent was char-
acterized by Cunningham (843) to be approximately 2OO gal./
min, a BOD of approximately 1,OOO mg/1 and a high fat
content. The BOD of corn canning waste waters varied from
1,2OO to 4,OOO mg/1 and total solids from 1,5OO to 8,OOO mg/1
252
-------
with total volatile solids from 1,4OO to 7,OOO mg/1, and
suspended solids from 30O to 4,OOO mg/1, as reported by
Halvorson et al. (1673). Horton et al. (2O39) reported on
the waste water characteristics in coffee production and
stated that the waste volume varied from 10O to 26O gal./
1OO Ib of coffee beans produced and the fermentation waste
waters had a BOD of 1,700 mg/1 and 90O mg/1 suspended solids
with 2,1OO mg/1 total solids.
PRETREATMENT REQUIRED
Eldridge (1144) surveyed the application of chemical processes
in industrial waste treatment, specifically noting the utility
of these processes in the seasonal discharge of cannery waste
and the use of chemicals for pretreatment such as lime, fer-
rous sulfate, alum, ferric chloride, ferric sulfate, or
zinc chloride. Slow settling of the solids was an objection
(2O39) to some chemical coagulants. The Ohio Canners Associ-
ation (2484) reported in 1926 that biological filtration was
an effective method for treating vegetable processing waste
waters at the rate of 2 mgad (46 gal./ft2/day) and suggested
pretreatment of these wastes to include sedimentation using
lime or sodium aluminate, pH adjustment and chlorination.
From pilot plant studies to determine the loading rates and
media requirements for the treatment of several food process-
ing wastes, for the Indiana Canners Association, Miller (30O7)
concluded that tomato wastes could not be clarified without
the addition of chemicals such as lime and ferrous sulfate.
Investigations by Warrick (4614) on pea cannery wastes indi-
cated that 75$ removal could be achieved by chemical treatment
and, therefore, 33 Wisconsin canneries installed this process,
as of 1934. However, treatment of beet and carrot cannery
waste on trickling filters at municipal sewage works had been
proven satisfactory (4614).
Humphrey (2087) reported, in 1940, that canning wastes often
needed separate treatment if the cannery was not near a
sewage works of sufficient size which could handle the ef-
fluent. Waste waters from the production of sauerkraut were
reported by Tanzler (4290) and Haseltine (1786) to contain
lactic acid, sodium chloride, and protein compounds with a
high sulfur content and which was treated by hot lime to
neutralize the pH (6.6 required) prior to discharge to sewers
and biological treatment. Wastes from tomato canning caused
difficulties in a septic tank trickling filter, plant, were
treated with lime at 1 pound per 5OO gallons waste (2O87), and
then successfully treated by a settling tank trickling filter
(1O65). Ventz and Zaenger (4537) treated fish processing
waste water with ferric chloride to break the fat-protein
emulsion prior to biological filtration.
Satisfactory results were obtained by Hatfield et al. (1793)
for the treatment of corn wastes when the pH value of the
253
-------
effluent was between 6.0 and 8.5, the ratio of the recircu-
lated filter effluent was constant, and the ratios of BOD
to nitrogen of 20:1 and of BOD to phosphorus of 75:1 were
maintained. Pearse (4214) outlined the characteristics of
corn product waste and general method of treatment and noted
that if the wastes were diluted four times with condensed
water they could be treated on trickling filters at O.7 mgad
(16.1 gal./ft2/day) .
One sugar factory (3284) treated their waste by aeration and
biological filtration followed by a 1 to 1O dilution (Shukla
Process). McLachlan (293O) stated that balancing tanks should
be used ahead of the trickling filters to maintain a constant
load on the filters. Acid fermentation might occur in the
waste before it reached the filters and was counteracted by
the addition of lime and chlorine (293O) or recirculation of
treated effluent (963). A segregated high strength citrus
waste was reported by Kimball and Furman (2483) to be pumped di-
rectly to the municipal waste treatment plant and a combination
of domestic sewage and citrus waste mixed in a ratio of 1:1
was successfully treated biologically.
Pretreatment by roughing filter or preaeration was advisable
in treating waste from a margarine factory with high-rate
percolating filters in combination with activated sludge and
also by alternating double filtration (2415). A high-rate
filter installed ahead of the standard rate filter to remove
a large portion of the BOD was a typical plant modification,
where the standard rate trickling filter plant was overloaded,
according to Moyer (3108). Hatfield (1796) reported, in 1928,
that activated sludge operated between an Imhoff tank and
trickling filter relieved the overloaded conditions from a
corn products factory to the point that the filter capacity
was increased three or four times. The "bottled-up" system
of corn processing,which employed reuse of waste water/
drastically reduced the pollution load (1574).
Anaerobic fermentation was used successfully at a Florida
installation on citrus fruit canning waste prior to biological
filtration to remove compound toxic to fish and eliminate odor
problems (5153). Methane produced in the process was used
as a source of fuel to aid in lowering the cost of treatment.
Both biological filtration and anaerobic digestion consider-
ably reduced the pollution load of starch wastes, but were
not truly satisfactory for the conditions required, according
to Stander and Gien (4177).
EFFICIENCY OF TRICKLING FILTER APPLICATION
The typical effect that a canning waste would have on a
municipal waste treatment plant indicated (3O87) that when
sewage alone was treated the BOD was reduced from 32O to
254
-------
21 mg/1 and when cannery wastes were treated in combination
the BOD was reduced from 1,566 to 466 mg/1 and the suspended
solids were comparably affected. Treatment of the pea can-
ning waste on trickling filters, after screening, removed
8O$ to 95$ of the BOD when the filters were operated at
O.5 mgad (11.5 gal./ft2/day), which was sufficient if the
final dilution was not less than 5:1 (2486). By quadrupling
the hydraulic load, the percent BOD removed dropped to 50
to l&fo, and required a final dilution of at least 2O:1.
Dickinson (964) concluded that food processing wastes were
amenable to treatment in percolating filters when an in-
fluent BOD of 1,140 mg/1 and a 4-hour oxygen demand of 472
mg/1 were reduced by filtration to a BOD of 3.5 mg/1 and
an oxygen demand of 3.6 mg/1.
Treatment of malting and cannery wastes by dipping contact
filters was reported by Hartmann (1773) to produce 6O$ BOD
removal on the filter with the remaining being removed in a
post-treatment aeration tank with a retention period of
2O minutes. Food processing wastes were treated by a 1.5
mgd (39.5 gal./ft*/day) trickling filter plant to produce
80 to 85$ removal of settleable solids and 95$ removal of
BOD, according to Hill (1952). Crawley and Brouillette
(815) reported BOD and average suspended solids removal of
about 53$ and 59$, respectively, on PVC roughing filters
from the frozen food processing industry wastes. Several
reports (2O39, 2749, 3284) indicated that food processing
wastes could be treated on biological filtration systems,
with and without recirculation,to attain reductions in BOD
in the high ninety percent range.
According to a paper by Young (4844), a sewage flow of
2OO,OOO gal./day increased to 38O,OOO gal./day during the
cherry canning season, when the influent BOD was 84O mg/1
and final effluent was 350 mg/1. Chemical treatment with
lime and copperas reduced the BOD to 382 mg/1 and coagulation
of cherry or tomato canning wastes with alum could reduce
BOD by as much as 5O$. Kruse (2578) removed 80$ of the
polluting load, as measured by the permanganate and BOD
values, using a high rate percolating filter to treat fruit
packing waste at 1,OOO to 2,5OO grams of BOD/m3/day (6.2 to
15.5 Ib of BOD/1,000 ft3/day).
Sedimentation reduced (2337) the suspended solids from
dried vegetable production by 76$. However, the BOD of 310
mg/1 from the preparation of potatoes was 33O mg/1 after
sedimentation. When carrots were being prepared, the BOD
change due to settling was reduced from 1,220 to 86O mg/1.
The oxygen absorbed permanganate value in 4 hours was 144
mg/1 before sedimentation and 76 mg/1 after sedimentation
for potatoes, and 94O mg/1 before sedimentation and 81O mg/1
after sedimentation for carrots (2337). In experiments by
the Wisconsin State Board of Health (5O9O), pretreatment
involved chemical precipitation with lime and ferrous
255
-------
sulfate which reduced a composite beet canning waste by 96$
suspended solids and 5O$ BOD, and a carrot canning waste by
98$ suspended solids and 67$ BOD. These wastes (5O9O) were
discharged to the municipal trickling filter plant which
produced an effluent containing 82 mg/1 total solids and
57 mg/1 suspended solids and 88 mg/1 BOD.
Waste liquor from a tomato canning plant which had primary
treatment was combined with domestic sewage by Hert (190O)
to remove 11.9$ of the BOD, 41$ of the suspended solids and,
after the tomato waste was applied to percolating filter at
0.724 Ib BOD/yd3/day (26.8 Ib BOD/1,OOO ft3/day), 7O$ BOD
was removed, producing an effluent with 161 mg/1. A high-
rate percolating filter added to an overloaded tomato cannery
waste treatment plant was loaded (1934) at about 51O Ib/day
and 55$ BOD was removed by the percolating filter after
sedimentation. A recirculation ratio of 4.5:1 was used in
the treatment of blancher waste water from a pea cannery
(2669) to produce an overall BOD reduction of 78.8$. Hert
(19O1) stated that 25$ of the BOD and 58$ of the suspended
solids were removed in primary treatment from pumpkin can-
ning waste water admixed with domestic sewage, and with
biological filtration 74$ of the BOD was removed, producing
an effluent of 965 mg/1 when loaded at 1.232 Ib/ydVday (46
Ib BOD/1,000 ft3/day).
Fatty waste waters from a vegetable oil refinery and a chemi-
cal plant were treated on packed towers, where the BOD was
reduced from 5,OOO to 5O mg/1 and the pH increased from 1.5
to 7 (3276). Eighty-three percent of the organic contaminants
were reported by Ventz and Zaenger (4537) to be removed from a
fish processing waste water, where ferric chloride was added
to break up and precipitate fat protein emulsions and the
effluent was treated by biological filtration and lagooning.
Pilot plant studies were made by Haseltine (1786) on the
treatment of waste waters from kraut and pickle manufacture
by biological filtration and the maximum efficiency was esti-
mated when the BOD of the recirculated effluent was less
than 4O$ of the total BOD applied to the filter, i.e., a
recirculation ratio of not more than 7. BOD removals of
85$ were obtained from the kraut and pickle waste liquors
on a pilot plant filter containing blast slag with no corre-
lation between recycle ratio and BOD reduction observed.
One sugar refinery increased removal of organic matters from
its wastes from 7O$ to more than 90$ by the addition of a
percolating filter and final clarifier (13O8). Another plant
(2149) experienced difficulties because of rapid fungi growth
on the filters, although 75 to 95$ purification of the sugar
waste was obtained. Brandon (444) reported on methods of
treatment of waste waters from processing coffee, which
256
-------
were equivalent in polluting character to about 6O,OOO
gallons of domestic sewage for each ton of clean coffee
produced. Biological filtration was satisfactory, contrary
to the experience of Chalmers (647) , to treat the mixed waste
waters (BOD of 1,100 to 1,8OO mg/1) which were diluted four
times by filter effluent and applied to the filter at the
rate of 1OO gai./yd3/day (11.1 Ib BOD/1,OOO ft3/day) from
which the settled effluent had a final BOD of 17 to 1O4 mg/1.
Plastic media biological filtration systems treating various
food processing wastes were reported (98, 643, 678, 2972,
4O84, 5196, 5325, 533O) to be economically successful and
highly efficient. These systems treated a variety of fruits
and vegetable wastes without the use of polyelectrolyte
flocculants and satisfactory operation was achieved as an
economic solution to the high BOD waste. A plastic media
trickling filter, which also doubled as a cooling tower for
a corn steep process and a vegetable oil refinery waste, was
described in a paper by Bryan (516). About the same time,
Hamlin (1697) described a super-rate biological filter with
a cellular clay pipe medium to successfully treat fruit and
vegetable cannery wastes. The filter had good resistance
to shock loads and the performance has been excellent. Rock
trickling filters were included in a discussion of food pro-
cessing waste treatment methods by Rudolfs and Heukelekian
(3772), and later Chipperfield (674) compared synthetic filter
media with conventional media.
COMPARISON TO OTHER METHODS OF TREATMENT
Siebert and Allison (4O16) discussed several examples of
typical plants used to treat waste waters from canneries.
Chemical addition with lagoons, screening with sand fil-
tration and with chemical addition, percolating filters
followed necessarily by large lagoons, storage, and again
large lagoons was discussed (4016) with regard to inherent
limitations and advantages of these installations. Mercer
and Rose (2973) investigated high-rate trickling filters,
air flotation, circular vibrating screens, and chemical
addition to treat food processing wastes. Alternative
methods were discussed by Wakefield et al. (4570) for the
disposal of citrus waste and the recovery of byproducts,
for example, lagoons, discharge to municipal sewers, and the
activated-sludge process were commonly utilized. Several
methods of cannery waste treatments, such as lagooning, bio-
logical filtration, coagulation and disposal by spray ir-
rigation, were compared by Anderson (74) in 1967.
Chalmers (647) described the treatment of waste waters from
coffee manufacturing which had a BOD of from 6,OOO to 8,OOO
mg/1 and alternative methods of treatment, such as activated
sludge and trickling filters as well as physical removal,
were evaluated. Physical removal involving screening,
257
-------
centrifuging, filtration, evaporation of selected strong
liquors and incineration was determined to be the most
economical.
Comparison of chemical treatment with ferrous sulfate and
lime with biological trickling filters was made by Warrick
(4615). The chemical treatment reduced the oxygen demand
by 5O to 75$ and the trickling filter reduced the oxygen
demand by 76 to 94$. The BOD was reduced by 70 to 92$ when
sweet pea canning wastes were chemically treated and applied
to the filter at rates below 3 mgad (69 gal./ft2/day), but
the efficiency was slightly reduced when screened waste was
applied at rates of 1 to 2.1 mgad (23 to 48 gal./ft2/day).
Jones (2338) experimented with treatment of cider waste
admixed with domestic sewage on single stage filtration,
with and without recirculation, and with alternating double
filtration. The most satisfactory method was by single
filtration with recirculation of the effluent and special
attention was paid to the nitrogen deficient nature of the
waste. Recirculation of the effluent was employed and a
diammonium phosphate nutrient was added to a plastic media
filter for which BOD and COD removal was three times greater
than that reported for conventional filters (5324).
Standard and high-rate percolating filters were compared by
Buzzell et al. (572) with activated-sludge units in the
treatment of synthetic protein water and that from a potato
starch factory. Activated sludge removed 95$ of the BOD with
loadings up to 8O pounds of BOD per 1,OOO Ib of mixed liquor-
suspended solids per hour of aeration. Standard rate fil-
ters loaded up to 1,3OO Ib of BOD/ac-f/day (29.9 Ib of BOD/
1,OOO ft3/day) removed 9O$,,as did high-rate filters loaded
up to 3OO Ib/ac-f/day (69 lb/l,OOO ft3/day). Wakefield
(4571) reported in 1954 that citrus fruit waste water could
be processed satisfactorily by high-rate biological filtra-
tion and by the activated-sludge process, but both required
nutrient supplement. As alternatives, if sufficient land is
available, spray irrigation was satisfactory, but chemical
coagulation with air flotation was not satisfactory.
However, it was suggested by Kammann and Herb (2375) in
193O that activated-sludge treatment was replacing biologi-
cal filtration to the point that it was not to be considered
as an economical means of treatment. Likewise, an Imhoff-
trickling filter plant was modified in favor of activated
sludge in 1928 for treating corn product wastes at Decatur,
Illinois, due to the added expense and equal operating ef-
ficiencies (5365). In contrast, Pearse (4214) considered
both systems for comparing the treatment of corn product
waste and activated sludge as not being successful. Norgaard
et al. (32OO) reported, in I960, on the experience of a
pilot plant composed of activated sludge and high-rate bio-
logical filtration. From the results obtained by the pro-
totype, the projected area required for handling the cannery
waste waters was excessive and disqualified the use of high-
rate filters.
258
-------
POST-TREATMENT AND EFFLUENT QUALITY
Typical of examples of post-treatment was a California
sewage treatment plant which installed high-rate biological
filtration followed by extensive post-treatment to cope with
a 76$ increase in food processing wastes over a period of
five years (645). Warrick et al. (4621) described their
studies on the disposal of cannery waste waters by discharge
to lagoons, and for many wastes lagoons are a partial solu-
tion to cannery waste disposal problems with the limitation
being control of odors and the large area required. Waste
disposal methods of solid and liquid waste from food canner-
ies were discussed by Wisniewski (4797) with strong emphasis
on the economic nature of the problem. The seasonal factor
of the canning industry gave rise to the use of land disposal
screening, biological filtration, chemical precipitation and
lagooning, and discharge to municipal treatment (3816, 4797).
Henry et al. (1883) reported in 1954 on the disposal of bio-
logical filtration effluent on crop irrigation systems and
it was found that Reed canary grass, which was tolerant to
wet conditions, would accept 4O inches or more of water
applied during the growing season (May to October). Almost
all of the nitrogen, phosphorus, and potassium were removed
from the effluent by the crops in the soil. There was no
evidence of bacterial pollution of a nearby creek as a re-
sult of drainage from the area, but concentrations of sodium
and chloride in the creek water were increased.
SPECIAL OPERATIONAL PROBLEMS
Fifteen pounds of chlorine were applied daily to a flow of a
little over a million gallons per day consisting of domestic
sewage and fruit processing wastes to combat problems arising
from odors during the summer and autumn (19O6). Problems
with foaming and flies on a trickling filter handling corn
packing plant wastes were solved by the addition of hypo-
chlorite to the filters, and chlorination of the effluent
allowed it to be discharged into a small stream and lake
(4884). Corn sugar refinery steep water was shown (3869)
to be a possible source of nitrogen for the fermentation of
alcohols from waste sugars which left almost a stable final
waste. A possible use for waste molasses and yeast was
composting vegetable matter to give a higher nitrogen con-
tent than with molasses alone, which would be an attractive
proposition for cane growing countries with surplus molasses
and a limited demand for alcohol (1955).
Typically overloaded conditions were reported (564O) in 1965
for Cambridge, where the biological filtration plant was
inadequate to handle the increased domestic sewage waste
waters plus waste discharged from a jam and other cannery
factories. To alleviate this condition, modifications were
259
-------
made which included additional percolating filters with re-
circulation and other appurtenances. Mangold (2856) described
experiments in which wheat starch wastes were treated on an
enclosed percolating filter with artificial aeration. The
treatment was satisfactory, but the cost of treatment would
be high.
In 1968, Koehler (2538) described possible treatment methods
for the problem of seasonal wastes from sugar, potato, and
sauerkraut processing plants. Shallow ponds were considered
along with their disadvantages, as well as successful appli-
cation to trickling filters. Segregation of the waste waters
from domestic sewage and cannery waters was practiced in
Australia, according to Furphy and Ley (1386), to avoid treat-
ment plant operating problems due to alkaline peach or acid
pear canning wastes.
Infiltration of oil and washing waters from hangars and
restaurants was reported by Kelez (2422) to be a chief
operating difficulty for trickling filter operation at an
airport location. The primary problem in wastes occurring
from the preparation of foods at restaurant facilities
along turnpike installations (54O3) was grease, which, once
removed by means of a preliminary digester, allowed the
percolating filters to operate with recirculation of chlori-
nated effluent.
Gulp (838) reported in 1963 on laboratory and pilot plant
studies treating mixtures of domestic sewage and cannery
wastes on a system which involved chemical coagulation,
primary adsorption on alum, rapid filtration, followed, when
necessary, by secondary adsorption on activated carbon.
Biological trickling filtration was not used in this process;
however, the principal disadvantage of the process appeared
to be high operating costs due to the cost of chemicals and
skilled operation. Thus, because of economics the problem
of treating waste is forced back to conventional methods.
Critique
The large quantity of literature available on biological fil-
tration of food processing wastes is indicative of the suc-
cessful use of this method of treatment. Since most food
waste waters contain organic matter in varying degrees of
concentration, with at best intermittent flow and, more
importantly, increased seasonal discharges, elevated tem-
peratures, alkaline and acid conditions, as well as other
factors, pretreatment of the waste waters was determined
to be of considerable importance in this application. The
sources of these waste waters came from washing, cutting,
peeling, juicing, blanching, pasteurizing, maintenance of
equipment, and other water-using procedures. Information
260
-------
regarding the characteristics, sources and treatment of food
processing wastes has been reported in considerable depth in
several industrial waste treatment texts, e.g., "The Treatment
of Industrial Waste," by Besselievre (3O8).
Much of the literature showed that on similar studies similar
conclusions were reached at essentially the same time. This
observation indicated poor communication, or awareness on the
part of the investigators, or doubt of the validity of the
previous work. Most of the literature appeared to have been
developed from a "make do" motivation for situations requir-
ing attention. Much of the published information informed
the public that a waste treatment plant was constructed and
was operating properly except for a few minor problems. Some
design information was available and, as more reports were
made on operational experience, these data were also used for
design. It must be pointed out that discrepancies in results
reported in this review on apparently identical waste streams
may, in actuality, support rather than refute the conclusions
of the investigators. The reason for this conflict is that
one tomato canning waste or ooffee processing plant effluent
may, quite conceivably, be significantly different from
another similar waste. This statement is not meant to imply
that generalizations are impossible or to be discouraged, but
that conflicting conclusions should be appreciated.
In summary, biological filtration has been shown to be
partially effective for the treatment of food processing
waste. Considerable pretreatment was reported to aid mate-
rially in producing an efficient waste treatment system. In
comparing biological filtrations with other methods, certain
waste and conditions arose which favored the other methods.
However, quite often biological filtration was an economical
means of waste treatment. As an aid to provide this economic
incentive, low cost, high performance synthetic filtration
media were introduced for the treatment of high BOD waste,
such as that found in food processing effluent. Post-
treatment of trickling filter effluent,other than the more
conventional secondary settling tank and chlorination,has
been extensive use of lagooning. Most operational problems
have evolved around nuisance conditions common to percolating
filters, such as odors, ponding, and fly production. Each
of these problems has had many answers proposed and applied
to make the process or biological filtration acceptable for
the treatment of food processing waste.
261
-------
SECTION XXII
INSTITUTIONAL AND MILITARY WASTES
The waste waters from institutional and military sources are
unique because they closely approach the characteristics of
domestic sewage, with possibilities of having high grease
and oil content and producing flows of extreme variability.
It was estimated in 1955 by Thompson et al. (4379) that 78$
of the air bases in the United States treat their own sewage,
while 20$ of them discharge to nearby municipal installations
and 2$ discharge untreated sewage to receiving bodies of water
(4379). Much of the design and operational experience in 1941
for the operation of treatment plants for camps under war-
time conditions was based on criteria obtained from camp
sewage plants during the First World War in 1917 and 1918
(5384).
Problems of population expansion in specified locations were
related, in 1942, due to the establishment of military camps
such as that at Neosho, Missouri, which required that a new
treatment plant be constructed which consisted of a grit
chamber, a comminutor, sedimentation tanks, percolating fil-
ter, and humus tanks (4056). Sullivan and Wiley (4267) re-
ported on the activities in southeastern United States deal-
ing with efforts to combat pollution from military installa-
tions, and the severity of the problem prompted the issuance
by the U. S. Public Health Service of a pamphlet entitled,
"Notes on Basic Design Data for Emergency Water and Sewage
Treatment Plants in Areas Affected by National Defense Pro-
gram. " Greeley (1561) stressed that the design of treatment
plants handling military installation or related wastes in
1942 should provide sufficient treatment to prevent nuisance
and to protect health, but that complete treatment in war
time was unnecessary and expensive and some pollution was
allowable.
Manning (2858) was concerned that data on 206 sewage treat-
ment plants built during World War II at military installations
and related activities which were analyzed should be applied
to future design situations. The sub-committee on sewage
treatment of the Committee of Sanitary Engineering, National
Research Council (559O), presented data on the operation of
several hundred waste treatment plants constructed at mili-
tary installations in the United States. A critical study
of the design and operation was conducted on these facilities.
Results on characteristics of military sewage, experience
with various unit operations and processes, and data repre-
sentative of these treatment plant operations were described.
Mohlman (3O53) discussed the reports by the Commissioners of
the Upper Mississippi River Basin and the National Research
Council on the results of high-rate percolating filter oper-
ation. Little nitrification occurred on high-rate filters
loaded with more than 2,OOO Ib BOD/acre/day (46 Ib of BOD/
263
-------
l.OOO ftVday) . Although high-rate filters were satisfactory
for treating sewage from military installations, this did not
mean that they were necessarily suitable for permarsnt use
at municipal installations. Greeley (1567) reported in 1948
on the extensive experience gained on the use of high-rate
trickling filters at 9O military installations and discussed
design relationships with regard to performance and flexible
operation.
Military procedures for waste treatment in other countries
were also reported. Early British procedures, according to
Balfour (157), considered each military waste plant individ-
ually and the degree of treatment depended on local conditions.
Hodgson (197O) summarized Australian experience on waste
treatment from military installations, and trickling filter
applications were popular. French military installations
(5155) generally follow separate sewage systems and when
treatment was provided it was by screening, sedimentation,
biological filtration with separate sludge digestion. A
trickling filter which was used to treat waste waters from
a National Guard Camp was the subject of a study in Japan
(2246) dealing with the ecology of the trickling filter.
The unique characteristics of military installation waste
waters were outlined by Hansen (1718) as being entirely do-
mestic and free from industrial waste and contained less
water and more grease. Other characteristics were the sew-
age flows experienced and loading rates applied. Military
installations prefer higher operating costs to high instal-
lation costs. Waste water characteristics from the U. S.
Army Ordnance Depot, which included waste from the vehicle
repair shops, had appreciable amounts of free and emulsified
oils, large quantities of suspended and dissolved solids, a
pH of 7.5 to more than 1O, temperatures from 5O°F to 9O°F,
and a BOD as high as 18O mg/1 (4191) . Daigh (855) described
the characteristics of waste from an Air Force installation
which contained free and emulsified oil, chromium salts,
solvents, grease, acids, alkalies, phenols, cyanides and
heavy metals. A waste treatment plant built for the Pentagon
in 1942 was designed to handle 1.2 mgd for a population of
4O.OOO which was treated at a rate of 3.2 mgd for 9 hours
each day, with the sewage estimated to have a BOD of 4OO
mg/1 and 57O mg/1 suspended solids.
Thompson et al. (4379) reported that the average flow of
sewage from U. S. Air Force installations was estimated at
1OO gallons/capita/day for resident personnel and 3O gallons/
capita/day for nonresident personnel, and the most common
method of treatment was by biological filtration. The char-
acteristics of military installation waste water were esti-
mated by Reinke and Pratt (3554) to have an average flow of
7O gallons/capita/day with a maximum flow of three times
this rate, a BOD of O.17 Ib/capita/day, suspended solids of
264
-------
O.27 Ib/capita/day and grease content of O.088 Ib/capita/day.
During World War II, unique situations were developed at camps
that were established under National Defense projects (5384)
which had from 1,5OO to 66,OOO people and the average flow
was estimated at 7O gallons/capita/day, most of which was dis-
charged during 16 hours. The flow of waste waters from air
training schools in Canada was estimated to be 60 gallons/
capita/day; the original figure was modified to 1OO gallons/
capita/day due to an increase in consumption of water (2931).
Hospital waste water characteristics were described by Megay
(2953) to have a flow greater than the amount per capita for
ordinary domestic sewage (in Germany) and to be generally
more dilute, higher in temperature, and contained more putre-
fiable organic matter. Institutional plants were reported
(2772, 2948) to handle 3O gallons/capita/day (2772) up to a
maximum of 1O,OOO gallons/140 persons/day (2948), in which
treatment was provided by primary sedimentation, Imhoff tanks,
sprinkling filters, secondary settling basins, and sand fil-
ters. An allowance of 35O liters (92.5 gal.)/head/day was
recommended by Fontaine (13O6) when designing treatment plants
for hospitals due to the bacterial contamination and divergent
nature of the waste, and the cost of treatment could be reduced
by use of a separate sewer system for the hospital. The waste
treatment works for Marcy State Hospital, New York, which were
designed to treat sewage from a population of 1,OOO at an
average flow of 1OO gallons/capita/day, was found (38O5) to
be overloaded, as many have been (1O53) , and a new plant was
built to handle 1 mgd. These overloaded conditions resulted
from underestimating the sewage flow per person. Typically,
hospital water consumption, in 1911, was assumed at 26 gallons/
capita/day and the primary sprinkling filter was dosed at the
rate of O.54 mgad (12.5 gal./ft2/
-------
tration plant, algal reduction tank for final clarification,
and an infiltration basin for underground disposal of the
effluent. The New York State Department of Health (5655)
was typical of concerned agencies involved in improving state
institutions by installing new equipment such as rotating
distributors on trickling filters and enclosing the waste
treatment plants.
PRETREATMENT REQUIRED
Due to the natural biodegradable characteristics of the wastes
generated by institutional and military installations, a great
deal of pretreatment was not warranted. The usual techniques
of clarification and some prechlorination were common. How-
ever, the photographic laboratory wastes at Norton Air Force
Base were given pretreatment (1882) prior to discharge to
municipal sewers for subsequent treatment on standard rate
percolating filters or activated sludge, but more concentrated
wastes were trucked to an ocean outfall disposal site. Pre-
treatment in the form of a 3.5-hour aerator in an institutional
plant was reported by Redenour (3524) to produce a partial re-
duction in the BOD and a flocculation of part of the colloidal
material. The net effect was the reduction of the required
area of the sprinkling filter.
Wastes from the airfield at Pendleton, Oregon, were treated
(5386) by a high capacity filter with provisions for chlorin-
ation prior to and after biological filtration. Daigh (855)
combined Air Force metal solution wastes with sanitary sewage
which were treated by biological filtration with certain of
the effluent streams receiving pretreatment to condition them
for biological treatment in the form of chemical addition, ion
exchange, and pressure sand filtration (855). A high degree
of treatment was accomplished (4191) using combined chemical
and biological techniques on ordnance repair shop wastes
plus post-treatment.
EFFICIENCY OF TRICKLING FILTER APPLICATION
A summary of operating results from biofiltration plants
handling military wastes noted (5388) that the average con-
sumption of water was 97 gallons/capita/day. The average
flow of sewage including laundry waste was 1O4 gallons/
capita/day. The influent waste had 193 to 34O mg/1 BOD and
was reduced to 9 to 43 mg/1 BOD, depending on the operation
of the individual installation. Hardenbergh(1729) summarized
data on military installations and emphasized that 7O$ removal
efficiencies on high-rate filters was experienced and two-
stage filtration systems loaded at 85OO Ib BOD/ac-f/day
(195.5 Ib BOD/1,OOO ft3/day) produced effluents of 66 and 53
mg/1 BOD. Waste treatment facilities for military instal-
lations were required (51O8) to operate for population
fluctuations between 3,5OO and 35,OOO people. The best
266
-------
operating results on five army Biofilter plants were obtained
(5388) from a plant comprised of comminutor, grease trap,
primary clarifier, Biofilter, humus tank, and separate sludge
digestion. In 1942, Fischer (1286) stated that ISO biofil-
tration plants were in operation or under construction in
the United States, with 62$ being at military installations.
It was observed that 1.21 to 5.O1 Ib BOD were removed per
cubic yard of filter medium (44.6 to 185 Ib of BOD/1,OOO ft3)
and power requirements (pounds of BOD removed per kilowatt
hour) compared to activated-sludge systems were slightly high-
er. Conventional treatment employing Biofilters was described
(51O8) using a flow estimated at 85 gallons/capita/day with
a maximum of 21O gallons/capita/day; the expected BOD removal
produced an effluent containing 30 mg/1 BOD and 35 mg/1 sus-
pended solids. Nelson (3159) reported on the experience of
a biofiltration plant treating military waste over a period
of 8 months which gave a total BOD reduction of 6O to 95$,
depending on the stages of treatment employed. The efficiency
of the waste treatment plants at the military installations
was reported by Kessler and Norgaard (246O) to provide 85$
BOD reduction by biofiltration, 6O to 83$ BOD reduction in
Aero-filters, and 80 to 90$ BOD reduction in two Biofilters.
A mechanical-biological system, which produced a 7O$ removal,
was proposed by Wegenstein (4658) to be used for a 5OO-person
military camp.
White (47O5) reported on the overload conditions in Spartan-
burg, South Carolina, due to the construction of a military
installation which required that a larger waste treatment
plant be built. The new Aero-filter installed to increase
the capacity of the sewage plant produced 9O$ reduction in
BOD and 92$ reduction in suspended solids. Another recon-
struction due to overload was the Neosho plant, where a re-
duction of 87.5$ BOD was obtained by the expanded plant (4O56).
Ellsworth (1162) stated that waste treatment plants composed
of high-rate trickling filters serving 3O,OOO to 40,OOO troops
provided removal of suspended solids and BOD of from 7O to
9O$ during the winter and summer months, respectively. Rown-
tree (3707) used biofiltration at New Zealand military in-
stallations and loads of 2.2 Ib/yd3/day (81 Ib of BOD/1,OOO
ft3/day) produced efficient treatment, reducing suspended
solids an average of 98$ and BOD 93$.
Schultz (3909) reported that effluent from a tuberculosis
sanitorium treated by artifically ventilated percolating
filters was 12 mg/1. Garbage grinding affected the waste
water stream by increasing the BOD 2OO to 290$ above normal,
and short periods of flow exceeded the daily average by 27O$
at a state school in Illinois (25O3). The increased loading,
however, did not appear to have any adverse effect on the per-
colating filters which still removed 9O$ of the BOD and sus-
pended solids. A BOD of less than 2O mg/1 was produced in a
two-stage filtration sewage plant at the State Hospital at
267
-------
Anna, Illinois, which was originally a standard rate per-
colating filter plant converted to high-rate operation (3627).
COMPARISON TO OTHER METHODS OF TREATMENT
Reid (3544) used biological filtration and other methods of
treatment to handle aircraft cleaning wastes at a military
installation. Later, efforts (3541) were made to combine
the merits of Biofilters and Aero-filters to treat sewage at
a Canadian Air Force installation, and forced ventilation
could be applied when necessary to the filter. An evaluation
of knowledge and experience was made in 1947 by Dreier (1O29)
who noted that the U. S. Army basis of design of 6OO Ib BOD/
ac-f/day (13.8 Ib of BOD/1,OOO ft3/day) was not excessive
when other requirements for treatment were respected. Dreier
also suggested that considerable experience indicated no
danger in applying intermediate loadings between 6OO Ib BOD/
ac-f/day (13.8 Ib BOD/1,OOO ft3/day) of standard filters and
3,000 Ibs BOD/ac-f/day (69 Ib BOD/1,OOO ft3/day) on high-rate
filters, particularly if recirculation was provided. Gradual
improvement in efficiency with high-rate filters,at military
installations in New England due to operational changes in
the primary tank allowed loads at 7,OOO to 9,OOO Ib BOD/ac-f/
day (161 to 2O7 Ib BOD/1,OOO ft3/day) with acceptable ef-
fluents. Brown (493) noted the use of high-rate percolating
filters with humus recirculation and final effluent recircu-
lation for the treatment of military installation waste in
South Carolina, where dosages of 35 mgad (8O5 gal./ft2/day)
and loadings of 3.6 Ib BOD/yd3/day (126 Ib BOD/1,OOO ft3/day)
were adequately treated at this installation. Other systems
were used (559O) with some success, but required skilled op-
eration and additional capital e.'ipenditure.
Stowell (4239) reported on several methods of sewage treat-
ment at mental hospitals, prisons, parks, and installations
where percolating filters treated saline sewage. Berry (29O),
while questioning the blind ..adoption of the activated-sludge
process, observed that installations such as hospitals pos-
sess a sewage flow which is too variable in quantity and
composition, and personnel are not sufficiently skilled to
give satisfactory results from activated-sludge operation.
Typical of problem conditions was one noted by Fowler (1320),
who related the experiences of C. C. James at a leper colony
near Bombay. Experiments were made with septic tanks and
various types of filters, including contact beds and trick-
ling filters, with which very good results were obtained.
Trickling filters have been long preferred because of their
simple operation and less sensitivity to variations in qual-
ity and quantity of sewage (13O6). An economic alternative
was a portable waste treatment plant (5113) which contained
a wood block media biological filter which could handle sew-
age from 4OO persons at the rate of 12 gallons/capita/day
with a BOD of 2OO mg/1 in the effluent. This system was used to
268
-------
handle waste treatment problems at new installations, such
as schools, pending construction of permanent facilities.
POST-TREATMENT AND EFFLUENT QUALITY
Franks and Obma (1342) discussed in general the procedures
and methods of treatment used at United States military in-
stallations which included Imhoff tanks, activated-sludge
process, Hays Process, high-rate and standard rate percolating
filters, and they observed that recirculation of filter ef-
fluents improved the performance of these plants. Sludge
and oil from all repair shop processes on an Army base (4191)
were stored in lagoons, and other treatment processes, includ-
ing flotation, were considered. However, very little attention
was paid to post-treatment devices, other than clarification
and, if required, disinfection.
Megay (2953) strongly recommended that infectious sewage from
hospitals be disinfected in addition to normal treatment, or
be so diluted that the content of pathogenic bacteria was not
greater than that occurring in domestic sewage. Many examples,
such as Langlois (2619) or Bergsman and Vahlne (279), dem-
onstrated that post-treatment chlorination of biological fil-
tration effluent from a tuberculosis hospital required a con-
tact period of 3O minutes and a concentration of 15 mg/1 chlorine,
under normal operating conditions, but under overloaded con-
ditions 25 mg/1 chlorine were necessary. However, Dixon
(982) stated that calcium hypochlorite in the amount of 2-3
ppm of available chlorine was used to sterilize biological
filtration effluents from a Pennsylvania hospital.
Bertelmann (3OO) reported that the sewage from a skin dis-
ease hospital was used for subsoil irrigation after biologi-
cal filtration with no particular problems. A one mgd sewage
disposal plant employing the biofiltration process satisfac-
torily treated waste from a state hospital and the effluent
was used (23O3) for irrigation. Ponninger (3419) described
the waste water characteristics from three sanatoria and noted
that lightly loaded percolating filters at least 4 meters
(4.4 yards) deep were operated with sewage diluted 1:1, fol-
lowed by sedimentation tanks and chlorine contact tanks with
detention times of 1 to 2 hours.
SPECIAL OPERATIONAL PROBLEMS
A general review of military installation waste treatment
plants was reported by Pedersen (333O) for installations in
the 9th Service Command and problems, such as overloaded con-
ditions and ponding of small trickling filter medium as well
as unskilled operators, were observed. Concern was expressed
by Owen in 191O over the construction and operation of waste
treatment plants to handle military waste for the control of
disease, especially typhoid and other pathogenic organisms
(3261) .
269
-------
Problems with military waste treatment were reported by
Shepherd (3989) when a profuse growth of sulfur bacterium
developed on Biofilters,and a combination of flooding and
chlorination corrected the problem. Kessler and Norgaard
(246O) remarked that problems with disintegration of rock
and trickling filters had occurred during the operation of
military installation waste treatment plants. Smith (4O59)
in 1915 reported on a problem which was of importance (but
of doubtful significance today) at military installations,
to handle waste from an army corps of 42,OOO men with 1O,OOO
horses which required extended sedimentation and trickling
filter treatment. The effect of the construction of base
housing at Chase Field was to overload (4653) the existing
waste treatment facility. The design and construction of
an additional plant rather than expand the existing one was
undertaken, based on local conditions requiring a high quality
effluent discharge.
Hood (2OO7) reported on efforts to automate military instal-
lation waste treatment by the use of oxidation-reduction
potential to simplify the degree of skill required for oper-
ation. Problems which developed from underloaded waste
treatment plants due to transfer of personnel at military
installations included sedimentation tanks being too large,
high rates of recirculation required for biological filtration,
and septic conditions in influent sewage (2577). Manning con-
cluded (2858) that standardized designs may have been justi-
fied under wartime conditions,but that they served little or
no purpose under normal conditions.
Problems in the design and operation of biological trickling
filters at a tuberculosis hospital were reported by Brossman
(485), who indicated that the filter took six.to seven weeks
to mature and was infected with Psychoda. Problems in esti-
mating the volume and characteristics of waste flows were
reported by Schulz (3912) for design of hospitals and sanatoria.
The peak flow was recommended to be estimated as 1/5 instead
of 1/1O the average daily flow. The treatment of waste from
a tuberculosis sanatorium for which it was not possible to dis-
charge to a municipal sewer nor an available stream required
extensive treatment involving both chlorination and dechlo-
rination (3794). Antibiotics, disinfectants and other mate-
rials were discussed (2953) relative to their interference on
biological treatment processes,and antibiotics were cited
particularly for their inherent ability to form scum. Pratt
(3455) reported a typical problem in maintenance which oc-
curred at institutional waste treatment plants where, after
excellent results were obtained the first year of operation,
18 months later the settling tanks decreased in capacity and
the filters clogged because of neglect. To alleviate the in-
herent problem of institutional waste treatment by lack of
maintenance and operation, hospital waste treatment facil-
ilities were being designed and constructed to operate auto-
matically (5469) .
270
-------
Critique
The literature reflected a high incidence of use of biologi-
cal filtration for the control of pollution from institutional
and military facilities. These wastes, generally, were simi-
lar in character because of the absence of much of the trade
wastes commonly found in a city sewer system. The investi-
gators indicated concern and had some difficulty in establish-
ing design flows for the facilities. Many design criteria
were evaluated after several years and some agreement was
shown.
By discussing military wastes in a separate section, the
literature was more clearly examined. It was observed that
there were definite indications during and after World War
II that responsible individuals were, in essence, actually
recommending limited pollution. This criticism of distin-
guished engineers such as Greeley is made in light of the
circumstances of the times in which this nation was under
the economic pressure of war. At that time, in the '40's,
it was practical, expedient, and evidently safe to build
waste treatment facilities with limited capability. However,
for future purposes, the results of these endeavors should
be noted. Many trickling filter waste treatment plants were
built in which the media deteriorated or other parts of the
waste treatment plant failed shortly after or during the war
period. In several instances, the plants were designed so
that further expansion was difficult if not impossible.
Some reports indicated that biological filtration would
have been phased out altogether if it were not for the re-
maining plants left over from abandoned military facilities.
The u. s. Public Health Service had issued information deal-
ing with procedures and design of plants which could be in-
stalled or constructed to partially abate pollution. The
military operations have favored higher operational costs
and lower capital investment due to the temporary nature of
the facilities built. It is hoped that the experience of
the past and the total economic implications of a military
facility are recognized so that if the installation is aban-
doned the community remaining would still have adequate, well-
designed, expandable waste treatment facilities.
Institutional waste has, it would appear, been relegated to
disposal by rural methods. Much concern dealing with path-
ogenic transfer from hospital-type installations was ex-
pressed in the literature along with workable solutions. Dif-
ficulty in defining the per capita flow at the various insti-
tutions over the years reviewed does not indicate incompetence
of the investigators, but changing water use habits of the
institutions. There is literature evidence that institu-
tional and military waste treatment facilities are ideal
applications for semi- or fully automatic waste treatment
271
-------
plants. The desire for automated waste treatment plants
has been expressed in several areas. With the Government
emphasizing that the waste waters from its various agencies
be treated properly to conform with the issued regulations,
research and development to design and evaluate an economical
and reliable automated plant would be profitable.
272
-------
SECTION XXIII
LAUNDRY AND CLEANING WASTES
Waste water generated from laundry and cleaning activities
can possess an extremely variable character. These wastes
can be organic, inorganic, acid, alkaline, or even radio-
active. Obviously, no one single method of waste treatment
can treat all of these wastes satisfactorily, although trick-
ling filters have been accepted widely in this application.
Because of the variations in flow, strength, and characteris-
tics of laundry waste waters, chemical and biological proc-
esses at small sewage works may be seriously affected.
Gehm (144O) maintained that sewage works with a total flow
of more than 5 mgd usually experience little difficulty
treating laundry waste waters. Boyer (433) found laundry
waste waters to be alkaline, turbid, and to contain large
quantities of soap, soda ash, grease, dirt, dyes, and cloth
scourings. According to Spade (4129), high-rate trickling
filters were most suitable for efficiently lowering the con-
centrations of suspended solids and BOD from this type of
waste. In a survey of four laundries in Dayton, Ohio, Horation
(1991) found that BOD concentrations ranged from 91 to 333
mg/1, while the suspended solids averaged 113 to 473 mg/1.
McCarthy (2896, 2897) did extensive work at the Lawrence
Experimental Station on laundry wastes with a pH of 11,
caustic alkalinity as high as 1,OOO mg/1, and a grease
content of 267 mg/1. The laundry waste was satisfactorily
treated on a rock trickling filter, 8 feet deep, at an
application rate of 5 ragad (115 gal./ft2/day). Jenks (2303)
used a 2.5 mgad (57.7 gal./ft2/
-------
Waste water from laundries in which clothing was contaminated
with explosives was treated (473O) by trickling filters after
mixing with domestic sewage in a ratio of at least 4 volumes
of sewage to 1 volume of laundry waste. Reid and Janson
(3544) and Reid and Libby (3545) confirmed that waste water
from cleaning operations involving aircraft maintenance was
adeauately treated with trickling filters. The most dif-
ficult portion of the waste emanates from the decarbonization
of aircraft engine parts. Troublesome contaminants include
cresylic acid, a mixture of ortho-, meta-, and paracresols.
PRETREATMENT REQUIRED
Pretreatment of cleaning wastes in certain instances has
enhanced the efficiency of trickling filter operations.
With respect to laundry waste water, Ens low (1178) advo-
cated the precipitation of alkaline laundry wastes by the
addition of acid prior to further treatment. Milk of lime
added to a sewage-laundry waste mixture was found by
Drummond (1O35) to improve operation and reduce odors. At
another plant, lime and chlorinated copperas as coagulants
were added to the waste water prior to settling and filtra-
tion (5214). In the treatment of radioactive laundry waste
water, Wiederhold (4715) recommended adding ammonium nitrate
as a nutrient prior to biological filtration. In his studies,
Newell (3177) found it necessary to add phosphate in addi-
tion to ammonia. Wilkinson (4730) recommended that separate
settling of laundry and domestic wastes be employed before
mixing of the two together and loaded the experimental
trickling filter at a rate of 63 gal./yd3/day.
Waste water from aircraft cleaning in tank washing operations
invariably required (3544) pretreatment before biological
filtration. The same authors applied batch pretreatment
techniques, such as skimming to remove oil, neutralization
with lime, and the addition of diammonium phosphate, on
engine cleaning wastes to break an emulsion of cresol-type
compounds. Reid (3545) noted that waste phosphoric and
nitric acids would supply necessary nutrients to phenolic
waste water which was in the alkaline state. Gutzeit and
Enyart (1631, 1632, 1633) treated waste waters from the
cleaning of tank cars and described a pretreatment system of
adsorption on coal, coagulation with ferrous sulfate and
lime, settling, storage in a lagoon for five days mixed with
domestic sewage, followed by a trickling filter. Safronova
and Yaroshevskaya (381O) cautioned that filters handling
laundry waste mixed with sewage must be gradually matured.
Hartmann (1776) stressed that treatment of "new" detergents
in 1967 by biological systems which were properly acclimatized
was still difficult due to the poor nutrient quality of the
detergents. Corder (789) reported on the use of Bionetic
at an airport installation to reduce the formation of grease
and scum in the trickling filter plant.
274
-------
EFFICIENCY OF TRICKLING FILTER APPLICATION
Trickling filters have been reasonably efficient in removing
contaminants from cleaning waste waters. Studies of laundry
waste treatment at military installations revealed BOD and
suspended solids reductions in the range of 81 to 89$ when
double filtration was employed (3O12, 5388). Newberry et al.
(3174) blended raw waste with sewage and used two different
recirculation ratios on one filter. They obtained effi-
ciences of 91$ BOD and 89$ suspended solids removals at a
2:1 recirculation rate, and 88$ BOD and 91$ suspended solids
removal at a 1:1 rate. Without any chemical pretreatment,
Boyer (433) tested a single stage limestone filter with ap-
plication rates of 2, 1, and O.5 mgad (46, 23 and 11.5 gal./
ft2/day) of raw waste. The BOD removals for each rate were
62$, 76$, and 85$, respectively. Crawley and Brouillette
(815) studied the performance of laundry wastes on plastic
media and obtained removals of 73$ of BOD, 4O$ of alkylbenzene-
sulfonate, and 72$ of the suspended solids.
Dobbins (984) claimed that 9O$ of mixed fission products
from radioactive laundry waste water was removed when the
waste waters contained detergents plus citric and nitric
acids. Using a similar waste, Newell et al. (3177) found
that biological treatment on a single stage filter was
effective only when settled effluent was recirculated at
a ratio of more than 8:1. Operation of two filters in
series with recirculation ratio of 6:1 or more also pro-
duced satisfactory results with BOD removals of 9O$ or more.
Cleaning wastes from aircraft maintenance activities were
investigated by Reid and Janson (3544) and it was found
necessary to dilute the waste with sanitary sewage before
reductions of 80$ phenol and 75$ BOD were realized with
biological filtration. Phenol present in waste waters
from tank car cleaning operations was reduced from initial-
ly 2O8 to 864 mg/1 to finally less than 1 mg/1 when trick-
ling filters were used in conjunction with adsorption on
coal and chemical coagulation (1631, 1632).
COMPARISON TO OTHER METHODS OF TREATMENT
The trickling filter process does not have a monopoly in
the treatment of cleaning waste waters (4876) . Moore
(3O8O) noted that laundry wastes were treated on trickling
filters or by chemical coagulation. Boyer (433) outlined
a coagulation process involving pH adjustment (6.4 to 6.6)
with sulfuric acid, followed either by 24O mg/1 of ferric
sulfate, 16O mg/1 ferric chloride, or 2OO mg/1 aluminum
phosphate, obtaining BOD reductions of 85 to 9O$. Ferrous
sulfate and sodium aluminate caused no precipitation, but
alum dosed at 23 grains per gallon of waste formed a rapid
settling floe, according to McCarthy (2896). Gehm (144O)
stated that the addition of sulfuric acid with alum or
275
-------
ferric sulfate lowered the amount of these chemicals needed
for coagulation. Carbon dioxide proved to be less effective
than sulfuric acid when examined for decreasing the amount
of ferric chloride required for coagulation. The same re-
searcher found that 5OO mg/1 of magnesium sulfate coupled
with 8OO mg/1 lime achieved high BOD removals.
Trickling filters were judged preferable to chemical pre-
cipitation since they required little operating skill, cost
of chemicals was avoided, they were more adaptable to fluc-
tuations and could be used for partial treatment. Spade
(4129) found that the high-rate trickling filter operation
efficiently reduced suspended solids and BOD from laundry
wastes. He cited comparative costs and operating problems
of sand clogging and odor nuisance for open sand filters.
Gehm (144O) maintained that sewage containing up to 20$
laundry waste could be treated by the activated-sludge
process with normal periods of aeration. Snook (4O91)
observed that it was necessary to reduce the flow of
laundry waste to an aeration tank to avoid bulking. Laun-
dry waste applied to plastic media trickling filters com-
pared favorably with the activated-sludge process (2944)
as well as with package treatment plants (643).
Wiederhold (4715) and Pazdernik (3317) reviewed the treat-
ment of these wastes by both the activated-sludge process
and biological filtration. Newell et al. (3177) compared
the coagulation of radioactive laundry waste by ferric
chloride and lime, followed by biological filtration and
then slow sand filtration. Calcium chloride, sodium hydrox-
ide, activated silica, and iron chloride were used for radio-
active waste prior to an activated-sludge process but ex-
cessive foaming was observed by Ruf (3776) and by Newell
et al. (3177). Chemical treatment produced 25 to 30 times
more sludge than that produced by biological filtration.
Dobbins (984) noted that the radioactivity removed by two-
stage filtration was about the same as that removed by
single-stage filtration when the total volume of filter
medium was the same.
The treatment of aircraft cleaning waste waters high in
phenol content was compared by Reid et al. (3545, 3547)
on rotating drums, by biological filtration, and by the
activated-sludge process. Wastes high in phenol content
were best handled on rotating drums, although trickling
filters were able to handle concentrations below 1OO mg/1.
The activated-sludge process was capable of treating phenol
up to 50O mg/1 with a BOD of 1,OOO mg/1. When the waste
contained cresylic acid and various cresols, an Aero-
accelator was able to achieve substantial BOD and phenol
reductions if suitable pretreatment was employed (2355,
3544). One airport (5258) did report trouble with the
activated-sludge process when oils, phenols, and metal
plating wastes were encountered, and a high-rate biological
276
-------
filter with recirculation was constructed as the substitute
treatment process.
POST-TREATMENT AND EFFLUENT QUALITY
Additional treatment on cleaning wastes following biological
filtration has been employed in some instances. In most
cases, sand filtration was the most popular choice of post-
treatment. At Liberty, New York, settled two-stage filter
effluent from a combined domestic sewage-laundry waste mix-
ture (53.8J.) was fed to a magnetite filter at a filtra-
tion rate of 1.5 gal./ft2/min (216O gal./ft2/day). In Covina,
California, ferric chloride was mixed with effluent from a
trickling filter humus tank and passed through a rapid sand
filter (2536). The filtrate was then chlorinated and fed
to underground percolating wells. Gutzeit (1633) mentioned
that tank car cleaning effluent from a filter was disin-
fected with chlorine dioxide before discharge. With regard
to radioactive waste water, Wiederhold (4715) recommended
either two-stage trickling filter operation or final coag-
ulation of the effluent to remove residual radioactivity.
Newell et al. (3177) found rapid sand filters satisfactory
for removing suspended matter from a radioactive trickling
filter effluent.
SPECIAL OPERATIONAL PROBLEMS
The treatment of cleaning wastes has caused some abnormal
operating problems. Spiess (4146) reported that a sewage-
laundry waste mixture encouraged the abnormal growth of
filter flies at one particular plant. The most effective
fly control method employed the application of trichloro-
benzene to the filter which was flooded and allowed to
stand overnight. In the treatment of radioactive laundry
wastes, difficulty in chemical coagulation prior to fil-
tration was encountered,probably due to the presence of
complexing agents (3177). Gehm (1440) observed that the
irregular flow pattern of laundry waste was felt most on
Mondays and Tuesdays. The grease and suspended solids
levels on these two days were 14% higher than the Monday
through Friday average, while the BOD was 12% greater.
Critique
Cleaning wastes were shown to be adequately treated by
biological filtration after proper pretreatment. It
should be noted that ABS type detergent wastes were never
treated adequately, but the impurities carried in the
waste streams were handled effectively. With the removal
of hard detergents from the general market, additional
BOD was loaded to the system, but total removal was
greater. Chemical pretreatment was shown to be required
277
-------
for deeming wastes to remove high suspended solids and toxic
elements. The trickling filter system was shown to be of
value here, as in other areas, where heavy metals or toxi-
cants were discharged. Also, the intermittent flow of
laundry wastes placed the recirculating trickling filter
process in a desirable position. Sufficient experience was
reported in the literature to properly design and operate
biological trickling filter plants to handle laundry and
cleaning wastes.
278
-------
SECTION XXIV
MEAT AND POULTRY WASTES
Meat packing wastes were listed (3O8) as having been success-
fully treated on biological trickling filters and, therefore,
the quantity of available literature is large. A general
discussion of the literature and practice of the disposal of
meat packing wastes was published in 1966 by Steffen (4203),
which outlined several alternative methods of treatment, as
well as the waste water characteristics, and the advantages
of producing salable by-products from the waste treatment
processes.
The growth of the meat packing industry in the United States
affected the waste water characteristics, and Hill (1949)
stated that the volume of waste water per 1,OOO pounds of
animals killed usually ranged between 835 and 4,16O gallons.
The seasonal variation in the waste characteristics indicated
that 2O$ above the average were killed in the winter, while
1O$ below the average were killed in the summer. The BOD
and the suspended solids per 1,OOO pounds of animals killed
varied from 5.2 to 18.9 pounds and from 2.9 to 22.0 pounds,
respectively. During the weekdays, according to Cropsey
(826), more than 9O$ of the sewage flow to the South St. Paul
sewage plant consisted of waste waters from the meat packing
industry. On slaughtering days, the sewage characteristics
were 28,OOO mg/1 total solids, 781 mg/1 suspended solids, a
BOD of 1,O09 mg/1, and by rather conventional treatment the
BOD was reduced to 236.8 mg/1, suspended solids to 51 mg/1,
and the total solids to 1,751 mg/1. A paper published in
Waste Engineering (5619) in 1956 stated that the most satis-
factory method for treating meat waste was in admixture with
dome stic s ewage.
Usual treatment processes which have been used (5619) in-
cluded fine screening, sedimentation, chemical precipitation,
biological filtration and the activated-sludge treatment as
well as various chemical precipitants. Nichols and Mackin
(3184) reviewed work of British investigators handling meat
packing wastes and listed, typically, that a treatment plant
included a grit chamber, grease extractor, primary settling
tank, a sludge digestion chamber, a secondary settling tank,
a mixing tank where domestic sewage was added, a third settling
tank, a cascade aerator, and an intermittent sprinkling filter
followed by humus tank. Dual treatment by trickling filter-
activated sludge systems, as well as mechanically backwashed
trickling filters, was reported by Hill (1949). Steffens
(42O4) described waste treatment from small abattoirs in
which the waste was characterized as resembling domestic
sewage, but was much stronger and required grease traps,
screens, and a settling tank with a capacity for one day's
flow. If the effluent was to be discharged to a stream, then
trickling filter treatment was recommended.
-------
Another plant was described by Howson (2O71) as containing
fine screens, grit chamber, grease filtration, flocculation,
primary sedimentation, primary trickling filters equipped
with air and wash-water operated at 6 mgad (138 gal./ft2/day),
intermediate clarifiers, with one hour detention time, and
either two- or three-stage biofiltration, which operated in
parallel or series as required. The filters were normally
operated in parallel at the rate of 1.4 mgad (32.2 gal./ft2/
day) .
Meat packing wastes have been problems in several countries,
e.g., New Zealand (37O5), and the data were similar with
those of typical plants in the United States. Meat packing
waste water characteristics were described by Aikins (25)
as quite similar to domestic sewage, and combined municipal-
trade waste waters were satisfactorily treated by percolating
filters or by activated-sludge plants. According to Just
(2359), once the fat and blood were removed, the wastewater
characteristics were quite similar to those of domestic
sewage, and admixtures of up to 5O$ could be treated satis-
factorily by biological filtration. Chief characteristics
of waste waters from packing houses are the high temperature,
and the high content of sodium chloride, fats, organic nitrogen,
nitrates, and total solids (27O8). Most of the effluent from
a poultry packing station was discharged during the short
periods of wash-down operations (5575) . Rowntree (37O5)
suggested conventional treatment of waste waters of meat
packing and slaughter houses by biological filtration or
activated sludge, while Heukelekian et al. (1922) used high-
rate filtration.
Russell and Axon (3788) described a municipal treatment plant
in Tennessee which utilizes trickling filters as a roughing
device for domestic sewage and meat packing wastes. The
operation of the Cedar Rapids, Iowa, waste treatment facility,
which was receiving intermittent discharge of a packing house
waste, was detailed in a series of articles by Mclntyre
(2916, 2917, 2918). Melzer (2966) studied the applicability
of treatment of meat packing waste with an average BOD of
2,14O mg/1 over a tower trickling filter filled with blast
furnace slag which had provision for intermediate layers of
air. Deyl (943) tested Biofilters 5 and 6 m high (16.4 and
19.7 ft) to treat slaughter house wastes. The optimum load-
ings were 2,OOO g BOD/m3/day (125 Ib BOD/1,OOO ft3/day) with
recirculation, 8, OOO g (511 Ib BOD/1, OOO ft3/day) for
continuous operation without recirculation, and 5,OOO g (311
Ib BOD/1, OOO ft3/day) for shift operations without recircu-
lation.
Heukelekian (1922) characterized the waste waters from four
poultry processing plants where the flow of waste water per
1,OOO chickens varied from 2,62O to 7,OOO gallons. The
highest figure was due to the large volume of water used
280
-------
during killing and plucking, but the BOD and suspended
solids resulted primarily from wash water of batteries or
coops. Wash water figures at 5,460 gallons per 1,OOO turkeys
and 2,2OO gallons per 1,OOO chickens with BOD contributions
of 80 pounds ^er 1,OOO turkeys, and 31 pounds per 1,OOO
chickens were also recorded (5575). An average suspended
solids load was 51 pounds per 1,OOO turkeys and 11 pounds
per 1,OOO chickens. Data collected by Kountz (2557) showed
that the BOD load was 25 pounds per 1,OOO chickens. For
plants using the " flow-away" system for processing feathers
and offal, Bolton (397) calculated that 6,OOO gallons of
water per 1,OOO chickens processed was reasonable if water
conservation measures were taken. In a plant in which manual
removal of feathers and offal was employed, the volume of
waste water was 2,600 gallons per 1,OOO chickens processed.
When reasonably efficient blood recovery is practiced, the
waste waters had a BOD of 25 pounds and contained suspended
solids of 12 pounds per 1,OOO chickens processed. Sharpley
(3972) described a trickling filtration plant which treated
domestic, milk processing, and poultry processing wastes
at a rate of application of 42O gal./yd*/day. Whiten (4714)
recommended filtration through a roughing filter and a
standard rate filter for a sewage plant treating predominantly
wastes from the poultry processing industry.
PRETREATMENT REQUIRED
Plants utilizing trickling filters were recommended by Bolton
(397) to employ preliminary aeration as a pretreatment step.
He also stressed that for economic reasons, as well as for
ease of treatment, it was important to provide an efficient
recovery of blood, adequate screening for removal of feathers
and offal, and removal of gizzard contents at the processing
plant. As a pretreatment measure, Roberts (3640) observed
that all poultry processing plants were required to install
screens to reduce the amount of feathers and grit reaching
the treatment plant.
Bryan (518) discussed pretreatment at a Wisconsin meat pack-
ing plant which treated its waste waters by biological fil-
tration prior to discharging the effluent to the municipal
sewage treatment plant, where it is further treated by
activated sludge and biological filtration. Combinations of
trickling filters and activated sludge were utilized (516)
to treat meat packing wastes prior to discharge to the
municipal wastewater treatment plant. Martin (2876) gave
the requirements for construction of slaughter houses in
France. Where the effluent could not be discharged into
sewers or a river, the treatment included settling, either
mechanical or with addition of chemicals, and trickling
filters or broad irrigation. Anaerobic digestion was used
by Hicks (1933) to reduce the BOD from 2,OOO mg/1 to 2OO -
4OO mg/1,which was amenable to further treatment on per-
colating filters or other aerobic systems.
281
-------
Packing house wastes were reported by Garrison and Geppert
(1417), in I960, to receive preliminary treatment by sedi-
mentation and grease recovery prior to discharge to munici-
pal sewers. With regard to rendering plant waste waters,
Granstrom (1533) claimed that a recirculation ratio of at
least 2:1 must be maintained to dilute the waste and equalize
the strength. At loadings greater than 5 pounds of BOD/yd3/
day (175 Ib of BOD/1, OOO ft3/day), the mixed waste water and
the recirculated effluent should be pretreated by aeration
prior to filtration. Rohde (3669) added iron sulfate and
lime to sewage entering the sedimentation tank and the sew-
age was also aerated in a preliminary section of the tank.
The percolating filters were completely enclosed and aerated
from the top.
Anaerobic digestion of meat packing waste waters was reported
by Steffen (42O1) as a pretreatment to biological filtration,
followed by sedimentation and chlorination. Silvester (4O27)
reported, in 1962, that anaerobic digestion at 33°C was the
preliminary treatment prior to biological filtration to
produce a settled effluent acceptable for discharge.
Primary sedimentation was used in chicken packing plants
(1922) to reduce the content of suspended solids in mixed
waste waters by 8O$ and the BOD by 4O$. Blocks of alumino-
ferric were placed (4068) in the waste channels prior to
primary settling to improve sedimentation of effluent from
poultry dressing plant since most of the impurity in the
waste water was colloidal. Grit removal, aeration and co-
agulation with ferrous sulfate and chlorine were used in
France (5O46) prior to high-rate percolating filters. When
a flotation unit preceded biological filtration, the addi-
tion of 2O mg/1 of alum to the waste water did not affect
the performance of a filter. Crist (819) reported that a
plant designed to treat domestic and industrial waste in-
cluding poultry processing waste waters used pretreatment
with lime and chlorinated copperas to adjust the pH and to
improve sedimentation for removal of 5O^ BOD and suspended
solids.
When blood pretreatment removal techniques were employed
(5575), the BOD of poultry processing wastes was reduced by
41$. It was suggested by Just (2359) in 1939 that it was
economical to recover the fat and blood from the waste waters
of abattoirs. Acid precipitation was used (4979) to remove
as much grease and organic matter as possible prior to
trickling filters in the treatment of waste waters containing
a high grease content from wool scouring.
EFFICIENCY OF TRICKLING FILTER APPLICATION
Typical of the relationship of meat packing waste to a munici-
pal population was that reported by Hansen and Hill (1716),
282
-------
where waste waters of a population of some 18,OOO were com-
bined with the effluent of a packing house which had an
equivalent of 195,000 people. Pretreatment was heavy chlorin-
ation, followed by sedimentation which would reduce the
BOD's from 2,OOO to 600 - 8OO mg/1, and the suspended solids
from 1,9OO to 200 mg/1 before discharge to the sewers. They
stated that 7O$ BOD reduction could be obtained by recircula-
tion ratios of 3:1 through biological filters. Bragstad
(439) reported that after chemical-biological treatment,
the final effluent contained 9.9 mg/1 suspended solids and
had a BOD of 1O.6 mg/1 based on an influent of 2.6 mgd of
domestic sewage and 1.7 mgd of meat packing waste water
which contained 4O3.1 to 720 mg/1 suspended solids and
BOD's of 324.7 to 954.7, respectively. Seven years later
modifications to this plant were described by Hill (1948),
which involved changes in the settling tank weir system and
increased the BOD removal from 4O to 55$, and suspended
solids from 60 to 65 - 7O$. Once the modifications were in-
stalled, 9O$ BOD reduction was achieved across the filters.
Levine (2699) reported, in 1935, that the BOD of settled
packing house waste could be reduced by 97$ and the organic
nitrogen content by 95$ using an application rate of 7 mgad
(161 gal./ft2/day) on an aerated granite filter followed by
percolating the settled effluent through a cinder or gravel
filter at the rate of 3 mgad (69 gal./ft2/day).
Overall BOD removal through the chemical-biological filtra-
tion plant treating poultry waste waters was 93$ at loading
rates of 1.35 Ib of BOD/yd^ of medium/day (5O Ib of BOD/
1,OOO ft3/day), according to Crist (819). British instal-
lations were noted to produce 9O$ removal. Meat packers
have on occasion built and operated adequate waste treatment
facilities (4837) which produced acceptable effluents. One
million gallons a day of screened packing house waste were
treated (4487) to 92.7$ BOD removal by biological trickling
filtration on a quartzite filter, 1.3 acres in area and 7
to 8 feet deep, with fixed sprays. Eldridge (1143, 1148) re-
ported on the troublesome characteristics of a meat packing
waste in Michigan. He indicated that a percolating filter
plant with recirculation removed 99.2 to 96.6$ BOD from a
septic tank effluent and, after dilution with cooling water
from the refrigeration plant, the final effluent contained
5 to 12 mg/1 of BOD. Levine (27O8) observed that the over-
all reduction of BOD by a standard filter treating packing
house wastes was 4OO Ib/ac-f of medium (9.2 Ib BOD/1, OOO
ft3/day), as compared to the reduction of 1,OOO Ib/ac-f of
medium (23 Ib BOD/1, OOO ft3/day) by the primary filter of
a two-stage filtration plant. Investigations (1417) on the
use of plastic media filters indicated that 7O$ of the BOD
could be removed at high rates of loading of packing house
waste waters. A waste treatment plant handling meat packing
wastes in North Dakota was reported by Howson (2071) to handle
283
-------
7OO,OOO gpd and reduced the BOD by 95$. After primary fil-
tration, the BOD which had an influent of 1,OOO mg/1 had
been reduced to 25O to 3OO mg/1.
When packing house wastes at Cedar Rapids, Iowa, with a daily
volume of 1.2 mg and a BOD of 2,OOO mg/1, and sometimes 4,OOO
mg/1, were discharged to the Cedar Rapids waste treatment
facility, the influent BOD was satisfactorily reduced by 40
to 5O$ after an elaborate sequence of clarification and
filtration aided by a roughing filter (5O97). A roughing
filter used in Oklahoma City was reported by Cunningham (841)
to reduce the BOD to 6O mg/1 on an influent of combined
domestic sewage and packing house waste. A pilot plant bio-
logical filter, studied by Hirlinger and Gross (1959),
operated as a high-rate unit with a rate of flow of 2O mgad
(46O gal./ft2/day) and a recirculation rate of 50$, pro-
ducing a reduction in BOD of 5O$ with BOD loadings as high
as 4.97 lb/yd3 (184 Ib BOD/1,OOO ft3/day). During the non-
killing hours, when the BOD load on this filter was about
2.O6 lb/yd3 (76 Ib BOD/1, OOO ft2/day),the average reduction
in BOD was 61.2$.
COMPARISON TO OTHER METHODS OF TREATMENT
Steffen (42O2) discussed and compared various methods of
treating slaughter house waste waters including that of
biological filtration. Biological filtration and the acti-
vated-sludge process have not been widely used (3436) in
poultry plants because of the high capital cost and the
care required in operation, but an extended aeration modi-
fication of the activated-sludge process was used at one
plant. Meat processing plant waste waters were treated by
the extended aeration process, which was explained by
Willoughby and Patton (4739) to have certain advantages over
conventional activated sludge and over two-stage biological
filtration. Kountz (2556) determined that neither the acti-
vated-sludge process nor biological filtration was suitable
for treating waste water from small slaughter houses because
the BOD was too high for biological filtration, except with
several stage operation or with high recirculation rate.
However, he showed that a coal filter operating at 2O mgad
(46O gal./ft2/day) without recirculation reduced the BOD
by 4O$, indicating that this method could be used as a pos-
sible preliminary treatment.
Mohlmann (3O41, 3O54) reported, in 1928, that activated
sludge was chosen for the treatment of packing house wastes
in Chicago rather than trickling filters due to the equal
performance and more economical installation cost. Waste
waters from packing house installations in Oklahoma City
were studied by Benham (212), and the advantages of two-
stage percolating filters, two-stage biofiltration, and the
284
-------
activated-sludge process were observed, with the two-stage
percolating filter method of treatment adopted. Final re-
sults of studies by Deyl (943) showed the advantages of
tower percolating filters over the activated-sludge process
for treating slaughter house waste waters. He concluded that
there was no difference in the results achieved by batch or
in continuous operation which permitted treatment on a more
economical shift basis, an operating procedure not possible
with the activated-sludge process. Levine et al. (27O2)
reported that the poor efficiency of an activated-sludge
plant (54.8$ removal BOD) on a packing house waste did not
refer to the unsuitability of the activated-sludge process
to handle meat packing wastes, but to the unusually high
salt content. The percolating filters removed 1,173 to
4,343 Ib of BOD/ac-ft(27 to 100 Ib BOD/1,OOO ft3/day) by
primary filtration, and 246 to 1,362 lb/ac-ft(5.7 to 31.6 Ib
BOD/1, OOO ft3/day) by secondary filtration.
Schroepfer (39O4) compared the anaerobic process for treating
waste waters from meat packing houses where, for removal of
BOD and suspended solids of 95$, a loading of 0.2 Ib of BOD/
ft3 of digestion tank per day was required. To obtain the
same degree of purification by the activated-sludge process
or percolating filters, loadings of only O.O2 to O.O5 Ib of
BOD/ft3/day (20 to 5O Ib BOD/1, OOO ft3/day) were possible.
Schroepfer et al. (39O5) reported the cost of treatment by
anaerobic digestion when compared to biological filtration
was approximately one-half to effect an equivalent degree
of purification. However, the costs of operation and mainte-
nance were slightly greater to produce a 95$ BOD and 9O$
suspended solids removal. Syme (4281) stated that conventional
treatment by biological filtration or activated sludge was not
entirely satisfactory for treating waste waters from meat
processing plants in New Zealand and that good results were
obtained at some plants using anaerobic digestion.
High-rate filtration, which could be backwashed with air and
effluent, was often used for the treatment of packing house
waste in 194O, according to Fischer (1281). Quite often,
small slaughter houses have required extensive treatment and
it was indicated by Stiemke (4218) that treatment by bio-
logical filtration of settled waste waters with recirculation
was poor. Satisfactory results were obtained on sand filtra-
tion where waste waters with a BOD of 8OO mg/1 with little
or no settleable solids were applied at rates up to 15O,OOO
gal./ac/day (3.45 gal./ft2/day). Chemical coagulants, such
as alum and calcium hypochlorite or chlorinated lime, produced
an average reduction of BOD of 95$ and a clear colorless
effluent. Kountz (2557) compared the cost for percolating
filters vs. spray irrigation and subsurface tile fields to
be in favor of the latter, but percolating filters were made
more competitive by adopting recirculation.
285
-------
POST-TREATMENT AND EFFLUENT QUALITY
Rather than building an expensive post-treatment device in
1937, Jacoby (2259) used abandoned strip mines to serve as
oxidation ponds and final sedimentation units. The BOD
was reduced from 1OO rag/1 coming from the filter to below
2O mg/1. Addition of sodium nitrate, however, was required
to assure a high quality effluent. Lagoons were used (3436)
successfully for treatment of poultry wastes or in admixture
with domestic sewage. Disposal by irrigation was satisfactory,
but required a large amount of land. As a form of final
treatment, the use of grassland irrigation for the trickling
filter effluent was encouraged (5575) .
Smith (4O68) found that it was preferable to recirculate
unsettled trickling filter effluent rather than final set-
tling tank effluent to establish high quality effluent. In
Gainesville, Georgia, Mower (31O6) described the provisions
made to apply chlorine to the recircul ating effluent from
trickling filters used to process meat wastes. French ex-
perience (5O46) demonstrated that an effluent BOD of 1O mg/1
and suspended solids of 2O rag/1 could be achieved in a chemi-
cal precipitation - high-rate biological filtration plant
treating slaughter house waste waters.
SPECIAL OPERATIONAL PROBLEMS
Typical operational problems that have occurred on waste
treatment plants handling meat packing wastes were reported
by Lovell (2764) in 1941. The surface of the trickling
filters at Fort Dodge, Iowa, was continually clogged and
had to be cleaned mechanically and by hosing. Operations
were improved using chlorination and post-primary sedimenta-
tion. Increased sloughing, reduced clogging and improved
efficiency were observed, and the overall performance of
the plant was 91$ reduction in BOD and 9O.4$ reduction in
suspended solids. Shepherd (3989) had problems with the
medium in the percolating filters and it was removed, screened,
washed, replaced. Older percolating filters were equipped
with new rotary distributors.
Ingols (2219) has found that large quantities of detergents,
when mixed with chicken processing wastes, interfered with
sedimentation and thus resulted in overloading of the trick-
ling filters. Biological purification of meat packing wastes
was the recommended treatment in France (665, 3393) where
concentrations of organic matter in the heavily polluted
water may be more than 6,OOO mg/1 with 2O to 30 times as
many pathogenic bacteria as in domestic sewage. It was also
recommended that a closed aerated percolating filter be used
at least on small plants.
Banister and Cloud (164) described the expansion of an existing
biological filtration plant to include a conventional activated-
286
-------
sludge unit to follow the trickling filter as second stage
treatment. Difficulties realized during operation have been
due to the presence of feathers and turkey lungs. Light
perchlorination was used when excessively large quantities
of blood were discharged at a Department of Agriculture
Research Center (5096) and, with a combined chemical treat-
ment - trickling filter plant, BOD reductions in excess of
have been achieved.
The method of charging industries to treat their waste waters
encouraged them to effectively reduce the volume and strength
of waste entering the treatment plant by 50 to 56% (Reaves,
3521) .
Critique
The heavy coverage in the literature on biological trickling
filters was indicative of the successful use and some problems
which were reported during the treatment of meat packing, '
slaughter house, and poultry processing waste. The wastes
were adequately categorized and characterized as to organic
content, suspended solids, inorganic content, pH, and other
operational factors. Grease was identified as a primary
problem and several authors stressed that pretreatment was
of importance for the removal of grease and blood prior to
biological treatment using trickling filters or activated-
sludge processes.
The value of salable byproducts from waste treatment and
recovery of materials was recognized early by several workers
and reemphasized recently. It was noted that poultry process-
ing wastes have changed slightly over the years due to in-
creased mechanization and that quantities of wash water
required have increased, but systems have been developed to
recycle wash water which reduced the pollutional effects.
The removal of feathers and other parts of the poultry was
important, since these materials interfere with the unit
operations of waste treatment facilities.
It would appear that sufficient evidence has been provided
by investigators over the years to demonstrate that the
waste waters discharged from meat processing and poultry
processing plants may be rendered biodegradable. Available
evidence indicates that, because of the severe conditions
imposed on a waste treatment plant by these wastes, pretreat-
ment is required and techniques have therefore been developed.
Most operational difficulties have been overcome with the
result than an economical waste treatment facility may be
operated with a minimum of trouble. It should be considered,
however, that only through proper maintenance and continually
modified operation will these wastes be adequately treated.
If a waste treatment plant is allowed to deteriorate, such
287
-------
as biological trickling filters developing plugging condi-
tions or dense fly populations, intolerable effluent will
be discharged to the receiving body of water, as well as
poor conditions for the local treatment plant environment.
288
-------
SECTION XXV
METAL WORKING WASTES
The sources, composition and treatment of electroplating
wastes were described in a comprehensive review by Pettet
(3352). The safest method of disposal of plating wastes
was reported to be by discharge into a public sewer, with
permission from the local authorities, although pretreat-
ment of some sort was indicated. Cyanide wastes were treat-
ed (3352) on a biological trickling filter with or without
an admixture of domestic sewage, and a review of the treat-
ment of cyanide waste waters was published by Brink and
Thayer (462). Mohlman (3O51) reviewed the various develop-
ments in percolating filters and activated sludge for the
treatment of wastes from the metal and other industries in
1946. A critical review of the literature published in 1952
compared various methods of treatment of industrial and munic-
ipal wastes in which biological filtration was compared to
activated sludge and mechanical filtration (5582) . The com-
position of waste from U. S. Air Force installations was
described by Thompson et al. (4379) as a variable flow do-
mestic sewage which contained plating and metal wastes, plus
waste waters from cleaning of engines and fuselages and paint-
ing.
PRETREATMENT REQUIRED
Most investigators found that some type of pretreatment, such
as screening, pH adjustment, or sedimentation, was required
for the treatment of metal wastes by biological filtration.
Eldridge (1139) discussed the effects of various industrial
wastes, including metal plating and acid mine drainage, on
municipal waste treatment plants. Many acid metal wastes
required pretreatment, such as neutralization and removal
of certain metal ions, to prevent overloading or destruction
of biological processes. When the proportion of industrial
wastes to domestic wastes was high, pretreatment definitely
was advised. Rummel (3781) treated brown coal mine wastes
on a pilot plant scale using lime water, aeration, and rapid
filtration. Pretreatment by flocculation of the iron was
used and pH adjustments in the range of 6.8 to 7.4 were sat-
isfactory. A rather unique method of treating plating wastes
on percolating beds, which also doubled as sludge drying beds,
was discussed by Rincke (362O) and involved pretreatment to
oxidize cyanide and reduce chromium. Pretreatment of plating
wastes, such as alkaline chlorination of cyanide, was also
recommended by Kittrell (25O2). Pretreatment by screening,
neutralization, coagulation and sedimentation of alkaline
wastes from roller bearing manufacture was also used before
the secondary treatment of screening, comminution, primary
sedimentation, two-stage biological filtration, and final
sedimentation, as described by Geyer (1462). Stolbov suggest-
ed (4225) that certain trade wastes, such as metal and phenolic
289
-------
wastes, could be mixed with domestic wastes which would make
them amenable to biological treatment without the use of a
Biofilter. Chlorination was used for preliminary treatment
and was combined with electrolysis for the recovery of metals.
EFFICIENCY OF TRICKLING FILTER APPLICATION
Working with plastic trickling filter media on a laboratory-
and pilot plant-scale basis, Porter and Dutch (3437) demon-
strated that consistent removals of 95$ phenol- and 9O$ cyanide
were readily attainable for various coke and steel wastes.
These high removals were accomplished on synthetic feed solu-
tions of phenol and cyanide. The effect of sulfides in con-
centrations up to 15 mg/1 was negligible. The treatment of
gas plant wastes and other ferrous metal industry wastes by
alternating doable filtration reduced the BOD from 15O to 11
mg/1 (3O14) . In 1928, Drury (1O36) described the installation
of sprinkling filters in combination with Imhoff tanks to
alleviate the severe pollutional load from an automotive
assembly plant. Concentrations of free cyanide as high as 150
mg/1 could be removed by biological filtration; however, cyano-
copper complexes were much more difficult to treat (462). A
filter designed to treat waste pickle liquor, spent gas liquor,
and other trade waste waters was described by Hunter and Cock-
burn (21OO). The biota and efficiency of these filters,
operated both closed and open, were compared. Koch (2533)
reported successful treatment of phenolic metal wastes on
biological trickling filters.
Experiments by Green et al. (1571) with domestic sewage con-
taining a high fraction of metal finishing wastes indicated
that the greatest growth occurs on Biofilter media when the
biofilm was at 2O°C, with BOD removal of 95$. At 3O°C> per-
formance was much the same, while at 5°C BOD removal was less
than half. Satisfactory treatment of synthetic rubber wastes
containing copper and chromium was obtained on high-rate
aerated percolating filters, according to Munteanu et al.
(3126). Hydraulic loadings were from O.5 to l.O m3/m2/hr
(295 to 590 gal./ft2/
-------
COMPARISON TO OTHER METHODS OF TREATMENT
Knie (2517) described various methods for treating cyanide-
containing waste waters, such as chlorination, acidification
and aeration, and ferrous sulfate addition, in addition to
percolating filters. Experimenting with lead wastes, Stones
(4233) demonstrated that biological filtration would remove
30$ of the lead from the settled waste, while activated sludge
would remove 9O$. In other experiments with copper and zinc
wastes, he found that Biofilters removed 2O$ and 3O$, respec-
tively, as compared to removals of 8O$ and 6O$ for activated
sludge (4231, 4232).
The sewage treatment plant built at Pforzheim, Germany, de-
scribed by Weber (4654), had percolating filters installed
for secondary treatment because of their simpler operation,
lower energy requirements and greater shock-load resistance
as compared to the activated-sludge process. Percolating
filters were considered by Kittrell (25O2) as more adaptable
than activated sludge for treating varying concentrations of
plating wastes. Recirculating filters were more efficient
than standard filters.
Knop (2527) found that the activated-sludge process was more
suitable than biological filtration for the treatment of mine
drainage wastes which were ultimately discharged into streams
in the Overhausen area of Germany. Tarvin (4294), from lab-
oratory scale studies with plating wastes, compared the effi-
ciencies of activated sludge and percolating filters. While
low concentrations of metal and cyanide did not greatly affect
the efficiency of the two processes, he noted that, if failure
occurred, activated sludge had an advantage over percolating
filters of being restarted much more easily. Walton and
Smith (4597) stated, in 1958, that no satisfactory method
had yet been developed for the treatment of mine drainage
wastes.
POST-TREATMENT AND EFFLUENT QUALITY
Post-treatment was reported (3352, 5582) to be routinely
practiced in the form of solids-liquid separation with
emphasis on lagooning troublesome wastes.
SPECIAL OPERATIONAL PROBLEMS
Oeming (3215) and Bloodgood (359) observed that metal plating
wastes killed organisms and reduced efficiencies in activated
sludge and percolating filters. Barth et al. (2O3) , in 1965,
stressed that heavy metals in concentrations of 1 to 9 mg/1
posed no serious detriment to high-rate biological filtration,
as well as other aerobic or anaerobic treatment processes.
Malaney et al. (2847) developed methods using laboratory
techniques, such as the Warburg respirometer, to predict the
effects of given metal ions on biological filtration and
activated sludge.
291
-------
Examination of the effects of copper on various biological
treatment processes revealed that trickling filters were not
seriously affected by small quantities of copper (2O-25 mg/1)
(4O23) . Horasawa (2O3O) experimented with copper mine waste
rock as a trickling filter medium and found no appreciable
effect on Aspidisca, Vorticella, Trochilia. Lionotus, Amoebae,
or zoocrloea. In 1939, Spencer (4133) stated that copper and
nickel plating wastes had no deleterious effects on percolat-
ing filters, but that biological purification ceased when
chromium was present. However, amounts of copper larger than
25 mg/1 reduced biological respiration, retarded nitrification,
and released significant amounts of free acids from the hy-
drolysis of the copper salt (4O23).
Jenkins and Hewitt (2285) experimented on a laboratory scale
with chromium and its effect on bacterial filters. He found
that the effects on the biofilm at 1 mg/1 were slight, with
a good quality effluent; at 1O mg/1, chromium retarded nitri-
fication and organic removal and moderately reduced the ef-
fluent quality; and at 1OO mg/1, chromium reduced nitrificatior
by 66$ and an effluent of poor quality was produced. The
toxic effects of finishing and plating wastes were also exam-
ined by Tarvin (4294) on a standard percolating filter type
pilot plant. Chromium was tolerated up to 4 mg/1 with no
ill effects. However, at 5.5 mg/1 some loss of biological
growth was visible. Concentrations of 1 mg/1 chromate in
the sewage were observed to occasionally affect nitrification
(4134). The effect of heavy metals, especially chromium,
was considered (5592) in conjunction with various percolating
filter surfaces. Sheets (3979) examined the toxicological
effects of various metals on biological treatment processes,
and concluded that a shock load of 4 mg/1 chromium partially
damaged the biofilm, and activated sludge would withstand
shock loads of 1O mg/1 for 12 hours without detrimental ef-
fects. Barium, resulting from the treatment of chromium
waste with excess barium chloride, appeared (4134) not to
affect the percolating filter when a large amount of bicar-
bonate was present. Pettet (3353) also found that certain
metals, sulfates, and cyanides in sufficient concentration
had deleterious effects on the trickling filtration and acti-
vated sludge process. Choking of trickling filters by iron
waste precipitate was reported by Cameron (598), and a com-
parison of trickling filters and activated sludge in the
treatment of acid iron waste was discussed. The use of lath
filters in combination with septic tanks for small instal-
lations was also described by Allton (58) . Operational dif-
ficulties, caused by covering of the media of percolating
filters by acid waste waters and oil, were alleviated by
treating the filters with emulsified ortho-dichlorobenzene
and breaking up the surface of the medium (19O3).
292
-------
Critique
Metal working industries produce wastes which may be highly
acid or alkaline and toxic, plus having other non-biodegrad-
able characteristics. Most metal finishing operations,
whether it be painting, electroplating, or anodizing, usually
require an alkaline treatment of some sort. Where pickling
operations are used, wastes are highly acidic. When each of
these two corrosive solutions is treated separately, the
extreme pH's of these solutions create problems, especially
where biological processes are involved. Frequently, an
industry generates both wastes and by combining the two a
waste of fairly neutral pH is obtained. The main problems
in the electroplating industry are the toxic wastes, partic-.
ularly from chromium and cyanide. Government regulations
are very stringent in the area of toxic discharge. However,
vague government regulations on pollutant concentrations
have sometimes allowed discharge of a waste which was not
always prudent. Fairly conventional methods have been used
in the treatment of metal processing wastes, as detailed in
the book, "The Treatment of Industrial Waste, " by Besselievre
(308) . The only real advances in treating these wastes in-
volve the economics of chemical handling. The most frequent
treatment methods are by trickling filters and activated
sludge, following some form of pretreatment.
293
-------
SECTION XXVI
MILK WASTES
Milk wastes or the waste waters from the processing and
production of milk products have generated serious pollution
problems. It was remarked by Guth (1625) in 1912 that waste
waters from creameries quickly underwent acid fermentation
and putrefaction and were considered difficult to purify.
A general review of investigations concerning treatment of
milk processing waste suggested (2485) that chemical treat-
ment was useful as a partial treatment device, but for com-
plete treatment sand filters and trickling filters were pop-
ularly used in 1931. Levine and Watkins (2696) did consid-
erable work on the treatment of milk processing waste by
biological filtration in which bacteriological considerations
were applied to explain the treatment efficiency and the
functions throughout the filter handling these milk wastes.
Several papers by Levine and his associates (2691, 2693,
2698, 27OO) assembled much of the technology of milk waste
treatment by biological filtration.
In 1954, Wisniewski (4796) reviewed various treatment process-
es available for treating milk processing wastes. Zack (4851)
outlined a waste treatment plant that handled 175,OOO gallons/
day of milk processing waste waters by primary sedimentation,
two-stage high-rate biological filtration, humus tank and
sludge digestion, which provided effluent of sufficient
quality to be discharged. The U. S. Public Health Service
(5593) published a guide on types of treatment plants use-
ful for the treatment of these wastes.
Kountz and Forges (2558) outlined government research carried
out to determine the characteristics of dairy waste and ex-
plained the basic theory of the biological oxidation of dairy
wastes on filter beds and in aeration tanks. The Water Pol-
lution Research Board (Great Britain) (3287) also was active
in the development of data for the treatment of milk process-
ing waste waters for many years. The cost of construction
and operation of a biological filtration system to handle
waste waters from various milk processing installations in
France was described by Fertin (1259).
Several specific processes were used to handle this trouble-
some waste. The treatment of milk processing waste was gen-
erally reviewed in 1942 by Hyde and Rawn (2161) , who noted
that the Mallory Process had been used. Milk processing
wastes were treatable (1139) by biological filtration or by
activated sludge, if the latter process was operated under
carefully controlled conditions. Grewis and Burkett (1580)
used a roughing filter to handle milk waste, which was com-
bined with a treatment of poultry processing waste waters,
to provide satisfactory treatment. Ellis (1157) and others
(1294, 158O) used the roughing filter concept to advantage.
295
-------
Levine (2691) in 1925 characterized creamery wastes as having
5O$ more carbohydrate lactose present than in ordinary sew-
age, and said that anaerobic conditions rapidly produced a
drop in pH from the acid formed by the bacteria.
Lawton (2649) reported that high saline whey in milk process-
ing wastes did not affect the development of the biofilm,
providing the BOD loading was not excessive. These waste
waters with a BOD in excess of 1,5OO mg/1 did not readily
develop normal biofilm either in the presence or absence of
sodium chloride. A small waste treatment plant in South
Dakota was reported (4904) to treat 5,OOO gallons/day of
waste waters from a creamery which contributed 68 Ib BOD/day.
Treatment consisted of sedimentation, biological filtration
and final sedimentation.
PRETREATMENT REQUIRED
The U. S. Public Health Service issued a bulletin (5593)
dealing with methods of reducing milk losses and, therefore,
polluting material, as well as information on effluent treat-
ment by irrigation, percolating filters, and activated sludge.
Chemical precipitation as preliminary treatment followed by
biological filtration was recommended by Kershaw (2438) and
Guth (1625). Pretreatment using base exchange clays and
other inorganics was studied by Robertson (3649) in 1938 in
treating milk processing effluents by biological filtration.
Banister and Ellison (162) gave an account of the treatment
of two trade wastes, one a canning, and the other milk pro-
cessing, and noted that an equalization tank and recirculation
of clarified effluent could maintain a constant rate of flow
so that a high capacity filter would provide the degree of
treatment required. Preliminary settling was found to be
unnecessary by Kennedy (243O) for the successful use of com-
bined activated-sludge trickling filter process to handle
milk processing waste.
A typical situation was a small town (population of 76O) with
a waste treatment plant which was required to handle waste
waters from a dairy processing 2OO,OOO pounds of milk per
day (3581). The domestic sewage was screened and treated
in an Imhoff tank and an Accelo-filter. The dairy waste
waters enter the plant by a separate sewer system and mixed
with filter effluent before passing through the percolating
filter. Kimberley stressed (2485) that complete information
regarding quantities, factory processes involved, and product
decomposition was necessary before the design of an efficient
milk waste treatment plant for a particular set of conditions
could be made. For economic treatment of waste waters from
the processing of milk, the loss of milk should be reduced
to a total of not more than 2% of the milk received, accord-
ing to Davy and Noth (895). They gave a layout of a plant
296
-------
suitable for complete treatment on a percolating filter
combined with domestic sewage. Provision was made for recir-
culation of the filter effluent, and secondary tank effluent
to the filter effluent.
EFFICIENCY OF TRICKLING FILTER APPLICATION
Eldridge (1129) reviewed in considerable detail various sys-
tems which were used in 193O to treat milk processing wastes,
such as (a) sand filters were desirable in areas where land
was plentiful and sand was available at.low cost; (b) gravel
filters were advantageous for operation under nonfreezing
conditions and could handle one mgad (23 gal./ft2/day) with
93$ reduction from an initial 1,6OO mg/1 BOD; (c) brush fil-
ters were found unsuitable due to deterioration and attack
by acid forming bacteria; and (d) cinder filters produced
89$ removal of wastes up to 1,740 mg/1 BOD with the use of
chlorination to control ponding. McKee (2919) gave an account
of experiments on aeration of dairy waste waters. Aerations
of milk processing waste water reduced the BOD from 260 pounds/
day to 81 pounds/day. Primary sedimentation reduced the BOD
by 15$ and, when followed by aeration, raised this to 7O$.
Septic tank treatment increased the BOD removal further.
Based on these results, a waste treatment plant was construct-
ed where the waste waters were aerated, settled, and sludge
separately digested, and the cost of installing and operating
the aeration plant was about half of that providing treatment
by biological filtration. Ellis (1157) investigated the per-
formance of plastic media biological filters which were shown
to reduce the chemical oxygen demand by 75$ when operating at
rates approaching 71 mgad (1,633 gal./ft2/day) and 88 Ib COD/
1,OOO ft13 of media/day.
Pretreatment with high-rate percolating filters reduced (896)
the BOD of waste waters from 826 to 396 mg/1 before milk
wastes of 13,333 gal./day were discharged to the sewage treat-
ment plant. A total combined flow of 74,90O gal./day was
treated at the sewage works to reduce the resulting 323 mg/1
BOD of the mixed influent to 65 mg/1. Chemical-biological
filtration of milk processing wastes and other trade wastes
indicated, according to Bruhne (5O9), that the chemical
treatment reduced the permanganate demand by 79$ and the
BOD by 93$, with further improvements affected by biological
treatment. Wisely (4782) reported on a combined trade waste
comprised of 46$ milk waste and 9$ brewery waste which over-
loaded the treatment facility, but operational changes to
incorporate chlorination and recirculation resulted in effi-
ciencies as measured by suspended solids reduction of 350 to
45 mg/1 and BOD reduction of 491 to 54 mg/1.
Municipal sewage treated in combination with milk processing
waste was (3529) chlorinated, comminuted, the pH raised by
297
-------
lime addition, followed by primary sedimentation, trickling
filtration, and final sedimentation. This plant performed
with an average BOD removal of 92.2$. It was noted (5593)
that percolating filters would handle diluted milk process-
ing waters at O.5 to 3.O Ib BOD/yd3/day (18.5 to 111 Ib
BOD/1,OOO ft3/day) at an efficiency of 7O to 9O$ BOD re-
moval. When compared to well operated activated-sludge sys-
tems, which were capable of BOD removal as high as 99$,
other considerations for the use of biological filtration
must be used, such as economics and less supervision.
Wieselberger (4721) briefly summarized the literature of the
use of alternating double filtration on milk processing wastes
and stated that a high rate percolating filter in Bavaria
produced 9O$ reduction of BOD and with lower loads an effluent
of 2O mg/1 BOD or less was obtained. As was typical of many,
Rumpf (3782) used alternating double filtration in 1953 to
treat milk waste which had a BOD of 45O to 1,28O mg/1 and
produced an effluent BOD of 3 to 1O mg/1.
COMPARISON OF OTHER METHODS OF TREATMENT
Bettels (311) reported in 1928 that milk processing waste
waters could be successfully treated in percolating filters
or discharged directly into any stream affording 1:1OO or
1:15O dilution. The characteristics of dairy waste waters in
Germany in 1933 implied mechanical and biological methods for
treatment, with the choice governed by local conditions
(1O57). A submerged trickling filter, as designed by Dewes
(942), to handle milk processing effluent to adequately
reduce the pollution did not develop metabolic products
which were toxic to aerobic organisms in normal secondary
operations, such as percolating filters or activated sludge.
Based on a review of recommendations by the Royal Commission
(Great Britain) of the late 18OO' s, and the activities of the Water
Pollution Research Board, Calvert (586) described the treat-
ment of milk processing wastes by alternating double filtra-
tion. Later, Jenkins (2282) detailed successful treatment
of milk processing waste by alternating double filtration.
Percolating filters and activated sludge were discussed as
treatment systems by Barritt (194) for the treatment of milk
processing wastes. Either process was adaptable after pre-
liminary treatment and followed by humus tanks, but slight
advantage was given to activated sludge due to its inherent
ability to take advantage of higher temperature waste.
Cantinieaux (613) also investigated the use of two-stage
biological filtration and activated-sludge processes to
treat milk processing waste and concluded that both methods
were satisfactory.
298
-------
POST-TREATMENT AND EFFLUENT QUALITY
Falkenhain (1235) reported in 1947 after reviewing fermen-
tation, biological filtration, surface irrigation, spray
irrigation, and fish ponds to treat milk processing waste
waters that, from the point of view of purification and
utilization, the best methods were to distribute the waste
waters on land through spray irrigation during ten months of
the year and by surface irrigation during the winter. Imhoff
(2195) discussed the advantages of recirculation of effluent
on trickling filters which were handling milk processing
wastes. Post-treatment by slow sand filtration was popular
in 1938 as indicated by Robertson et al. (3649).
SPECIAL OPERATIONAL PROBLEMS
Occasional chlorination (1129) was required to control pond-.
ing. Rapid acid fermentation caused problems in long sewer
lines (2691). Combined domestic and milk processing wastes
required special attention (3529), such as chemical treatment
by chlorination and lime for control of pH and odor.
Critique
Milk wastes, generally characterized as high in dissolved
organic matter, exerting relatively high BOD's, and ini-
tially neutral in pH, but readily susceptible to acid fer-
mentation, have been found to be extremely amenable to treat-
ment by biological filtration. This success and the numerous
plants in operation, ranging in size from very small to large
industrial complexes, have been reported in considerable
detail.
A very small fraction of the literature available was reviewed
here. Sufficient evidence was presented to indicate that bio-
logical filtration following pretreatment in the form of pH
neutralization, solids capture, and partial disinfection was
satisfactory. Trends were noted in the literature, e.g.,
prior to 193O treatment of milk wastes was difficult and when
admixed with domestic sewage undesirable results of plugging
and nuisance conditions developed. With increases in devel-
opment effort, the operational experience for treatment of
these wastes was successful.
As biological filtration was demonstrated effective, other
methods such as activated sludge were attempted and on occa-
sion were quite successful. Due to the readily biodegradable
nature of the waste, preliminary aeration has been valuable
in removing initial BOD. Biological filtration was operated
under considerably higher than normal saline conditions which
were developed at certain milk processing plants producing
cheese.
299
-------
The literature showed some overlap of investigators' efforts,
which may be viewed to substantiate further the efforts of
each of the investigations. The conclusions reached by the
investigators indicating that only activated sludge or bio-
logical filtration or some other method should be used for
the treatment of milk processing waste may not be well found-
ed. It is suggested that the decision for the use of one or
more of these processes would be more rationally dictated, as
some investigators have noted, based upon waste characteristics.
Most importantly, economic factors, such as availability of
land and operational personnel qualifications, must enter into
the decision. It may generally be concluded, however, that
biological filtration with limited pretreatment or little
flow time between the source and the treatment plant will
provide adequate treatment for most milk waste installations.
3OO
-------
SECTION XXVII
PHARMACEUTICAL AND FERMENTATION WASTES
The importance of pharmaceutical and fermentation wastes has
been recognized by groups such as The American Chemical
Society, which sponsored symposia on various aspects of water
pollution control and specifically on fermentation waste
disposal (2059). Genetelli et al. (1452) outlined an ap-
proach to selecting an industrial waste treatment system to
treat waste waters from the manufacture of pharmaceutical
chemicals which involved a survey of the receiving body of
water, examination of individual waste streams within the
plant, experimentation on the possibility of biological
treatment of one or more of the waste streams along with
other procedures to determine the design criteria for a
plant.
Pharmaceutical industrial wastes were treated (2815) by a
local sewage works and investigations of these wastes indi-
cated that the greatest effect on the works was the waste
from the manufacture of tetracycline. Molof (3058) reviewed
pharmaceutical waste treatment by activated sludge or bio-
logical filtration, and Hurwitz et al. (2134) described the
treatment of penicillin and streptomycin waste waters by bio-
logical filtration. Chipperfield (681) summarized waste
treatment developments, with emphasis on plastic media for
biological trickling filters, which were proven economically
sound for the treatment of fermentation and yeast manufactur-
ing waste with savings on capital expenditures of 25 to 55%.
Rudolfs commented (3764) on the operational factors of char-
acteristically high organic wastes such as may be found in
fermentation industries. These wastes usually contained
less suspended matter but more soluble organic matter than
sewage. Physical or chemical treatment was not as success-
ful as biological treatment, and many of the wastes required
additional nitrogen source. Penicillin wastes were char-
acterized in 1949 by Heukelekian (192O) to have a pH valve
between 2 and 2.5, a BOD of 2,150 to 10,000 mg/1 for the
spent broth, wash waters having a BOD of from 210 to 13,800
mg/1, with various organic and inorganic compounds used in
the process as well as some suspended material. Howe and
Coates (2056) characterized waste waters from the manufacture
of antitoxins, anti-sera and vaccines as containing total
solids concentrations of 4,000 to 8,500 mg/1, volatile solids
of 3,000 to 7,500 mg/1, BOD 1,000 to 1,700 mg/1, and alkalin-
ity 600 to 800 mg/1. The waste waters, 15,000 gallons/day,
including blood from test animals and wash water as well as
waste chemicals and animal excreta, were generally red.
A French publication (4955) reported the characteristics of
waste waters from the manufacture of penicillin, streptomycin,
aureomycin, which agreed well with the comments by Heukelekian
301
-------
(192O) . Muss (3142) and Paradise and Howe (3283) character-
ized the combined waste waters from the manufacture of peni-
cillin, primarily from spent broth, wash water, and total
solids of I2.5OO mg/1. Aureomycin waste waters, 63,OOO gal./
day (1593), were combined with sanitary sewage and storm water
totaling about 2.5 mgd, and the resultant solution had a BOD
of 46 to ISO mg/1, total solids of 1,5OO mg/1, suspended
solids 2OO mg/1, and a pH of from 2.5 to 8, and was highly
colored.
Dlouhy and Dahlstrom (983) tabulated typical fermentation
waste strengths which ranged in BOD's from 2OO to 5O,OOO
mg/1 with suspended solids from 3O to 5,OOO mg/1. The quan-
tities of yeast waste were reported by Kiby (2471) as being
great,e.g., a small factory using 5,OOO to 6,OOO kg (5.5 to
6.6 tons) of molasses per day produced more than 1OO m3 (13O
yd3) of waste per day. The wastes were characterized as
having variable quantities of alcohol, suspended solids, and
salts, and were treated experimentally by anaerobic fermenta-
tion followed by several stage settling and biological fil-
tration.
PRETREATMENT REQUIRED
Process changes, such as that noted by Kiby (2471), involved
in the manufacture of yeast, altered the effluent so that
previous treatment methods were modified or discontinued.
Preliminary treatment at the source was necessary, e.g.,
neutralization and distillation of the solvent used to ex-
tract the pencillin (1O98). Due to intermittent discharge
of the waste, a schedule of discharging was coordinated with
recirculation at the waste treatment facility to assure ade-
quate degradation of the spent metabolite from the penicil-
lin manufactured.
Wittmann (48OO) reported that waste waters from the produc-
tion of antibiotics were treated partially by evaporation and
by activated sludge, and later by high rate biological fil-
tration, to produce a satisfactory effluent. Griffin (1593)
reported on the treatment of aureomycin waste to which lime
was added to give a pH of 10 prior to discharge to a lagoon.
The lagoon was later replaced by a plant consisting of a
holding and aeration tank, primary and secondary sedimenta-
tion tanks, and recirculating trickling filter.
Combined treatment with activated sludge and percolating
filters was used successfully by Black and Fairall (342) for
the treatment of pharmaceutical wastes with primary emphasis
on biological filtration. This system was also described by
Hbrne and Rinaca (2O36) and others (3O58) to provide
adequate treatment for combined pharmaceutical waste.
3O2
-------
Because of overloaded conditions of biological filtration
plants, Edmondson (1115) indicated that fermentation wastes
from the manufacture of antibiotics be pretreated by triple-
effect evaporation at the rate of 2,50O gal./hr, with the
condensate then being treated by aeration and biological
filtration. Harvey (1783) noted that septic tanks were
used as preliminary treatment ahead of trickling filters on
wastes from yeast plants to provide successful treatment
in Scandinavia. Anaerobic pretreatment was reported by
Rudolphs (3764), Krige (2571), and Wittmann (4799) to reduce
the BOD significantly, with the effluent amenable to further
secondary treatment. Lagoons were occasionally used for
pretreatment, and their successful operation overshadowed
the effectiveness of percolating filters and activated
sludge, which were then discontinued (4433).
EFFICIENCY OF TRICKLING FILTER APPLICATION
Two-stage biological filtration was reported by Tompkins
(443O) to reduce pharmaceutical and antibiotic waste by
95 to 98$ BOD. Three-stage biological treatment, outlined
by Howe (2O59) to treat antibiotic production wastes, re-
duced the BOD 9O to 95$ and removed antibiotic activity
completely.
Heukelekian (192O) investigated chemical treatment, anaerobic
digestion, activated sludge, and biological filtration. He
concluded that digestion, followed by biological filtration
through coarse medium then fine medium, produced a final
effluent with a BOD of 35 to 4O mg/1 from an influent of
2,15O to 13,8OO mg/1. Antitoxin and vaccine wastes were
treated by an Imhoff tank and a high-rate percolating filter
(2O56). The effluent characteristics were 5O8 mg/1 suspended
solids, 12.3 mg/1 BOD, an absence of color and odor and an
alkalinity below 318 mg/1 from an influent having 4,OOO to
8,5OO mg/1 total solids, and 1,OOO to 1,7OO mg/1 BOD. Accord-
ing to Reimers et al. (3551) , wastes from the manufacture
of antibiotics, sulfa drugs and vitamins were usually treated
by evaporation and incineration. Weaker wastes were combined
with sewage and adequately treated by high-rate filtration
with 8O$ BOD removal on a single stage operation loaded at
approximately 2 Ib/yd3/day (74 Ib BOD/1,OOO ft3/day). A
penicillin waste with a BOD of 4, 9OO mg/1, a permanganate
oxygen demand of 2,93O mg/1, a pH of between 6 and 7 and
containing 26,8OO mg/1 total solids was treated to 99$ re-
moval of the BOD, producing an effluent of 3O mg/1 BOD, if
the loading were diluted to 0.5 Ib BOD/yd3 (18.5 Ib BOD/
1,OOO ft3/day). Preliminary treatment of a pharmaceutical
plant manufacturing antibiotics by lime and dilution 1:1
with river water plus primary sedimentation allowed high-
rate biological filtration with secondary sedimentation,
followed by chlorination, to produce 7O$ removal of suspended
solids and 35 to 50$ BOD removal (1464).
303
-------
The treatment of compressed yeast wastes was studied by
Rudolfs and Trubnick (3769) over five years on trickling
filters and, with loadings of 3OO to 2,800 Ib BOD/ac-f/day
(276 Ib BOD/1,OOO ft3/day), there was a marked decrease in
removal efficiency. They (3767) also treated waste waters
from the production of yeast combined with sewage at 1 mgad
(23 gal./ft*/day) on a biological filter. Concentrations
of 4OO mg/1 BOD were reduced by 8O to 87$; with higher con-
centrations efficiency was sharply decreased. Active nitri-
fication occurred with all concentrations and the rate in-
creased on passage through the filter.
Liontas (2723) described an industrial waste treatment unit
also treating sanitary sewage. The waste waters contained
inorganic acid, fermentation waste, salts, acetic solvents,
and other aliphatic and aromatic compounds, and had, after
preliminary treatment, a BOD of 1O,OOO mg/1. The overall re-
duction in BOD averaged 98$. A typical yeast plant effluent
was treated by Rudolfs (3764) anaerobically in three stages
over 3.25 days on an influent BOD of 5,ISO mg/1, producing
an effluent with 77.5$ reduction. Another system involved
lagoon storage, followed by treatment in covered percolating
filters at the rate of 24 mgad (552 gal./ft2/clay) to reduce
an influent BOD of 2OO-25O mg/1 by 7O$.
COMPARISON TO OTHER METHODS OF TREATMENT
An acid fermentation waste was described by Krige (2571) as
molasses slop which had an average BOD of 1,OOO mg/1 and
total solids of 8,OOO mg/1. Possible methods of treatment
investigated were evaporation, anaerobic digestion, and
passage through percolating filters. Previous experience
indicated chemical coagulation was of little or no value.
Penicillin waste waters with a BOD of 12,OOO mg/1, a pH of
2 to 3, and suspended solids of 12,OOO mg/1 were treated by
Wittmann (4799) with anaerobic digestion and biological
filtration, and it was concluded that the latter was the
better method.
Waste waters from a yeast factory with a BOD of 5,OOO to
9,OOO mg/1 were digested anaerobically, and experiments on
aerobic systems by Merkel (2976) indicated that two-stage
treatment was successfuL but single-stage was susceptible
to a strong growth of fungus. Jensen (2316) described
operations in the production of yeast similar to that of
Kiby (2471). He observed that preliminary treatment, using
anaerobic digestion followed by lagooning, prepared a waste
which, when treated on percolating filters, produced little
suspended matter in the effluent. Humus tanks were un-
necessary. The BOD was reduced from 8,6OO mg/1 to 1OO mg/1
with a pH going from 4 to 7.8 as a result of this treatment.
3O4
-------
Tooloose et al. (4433) reported on a treatment plant develop-
ment in which it was shown that a properly operated lagoon
would handle fermentation waste waters, provided that sudden
changes in loading were avoided and seasonal efficiency could
be tolerated. This method was just the opposite of that
selected by Vogler (4545), in which a fermentation waste
was originally clarified, treated on percolating filters,
followed by an aeration unit and final sedimentation tanks
with some reliance on chlorination and sludge lagoons.
Jackson (2251) preferred biological filtration to activated
sludge for the treatment of fermentation waste due to its
more economical operation and lower sensitivity to shock
loads from the fermentation industry. Bopardikar (4O7) ex-
perimented with the use of algae as an alternative method
of treatment to conventional biological filtration and other
processes for antibiotic production effluent, and found the
method to be less expensive than conventional ones. Deep
well injection has been practiced by Cushman and Hayes (849)
and by Gurnham (1624) to handle the disposal of pharmaceuti-
cal wastes as an alternative to biological methods. However,
Cushman and Hayes (849) also investigated various modified
biological treatment systems in combination with chemical
wastes and domestic sewage. Based on pilot studies it ap-
peared that all the waste waters were amenable to treatment
by biological filtration.
POST-TREATMENT AND EFFLUENT QUALITY
Sisson (4O43) reported that treatment by two-stage biological
filtration, following*preliminary neutralization, floccula-
tion, and sedimentation, conditioned pharmaceutical waste
waters so that they could be used for ground water replenish-
ment by pond percolation. Tompkins (443O) found the BOD of
the pharmaceutical waste effluents from two-stage biological
filtration with recirculation was reduced by 95 to 98$.
Jensen (2316) stated that percolating filters following
anaerobic digestion removed solids to the extent that a
post-treatment humus tank was not required.
The effluent from the percolating filter for yeast waste
treatment showed (2471) a reduced ammonia content, a high
nitrate and nitrite, dissolved oxygen, and did not decolor-
ize methylene blue at 2O°C, indicating stability. The
effluent was discharged after dilution with untreated
cooling water.
Experiments (2571) on the use of two-stage enclosed percolat-
ing filters treating a diluted fermentation waste with a re-
circulating ratio of 1O:1 indicated that for the effluent to
to be treated to the degree necessary 3O,OOO yd3 of filtering
medium would be required. A greater volume of water or sewage
than was available would be needed for dilution.
305
-------
SPECIAL OPERATIONAL PROBLEMS
Overload of the plant during operation was experienced
occasionally, e.g., Hamlin (1691) stated that 1O,OOO gallons
of yeast factory effluent was dumped into a sewage system
treating approximately 1OO, OOO gallons per week which severely
overloaded the trickling filters. The waste stream had to
be diverted to another sewer and the filters quickly re-
covered. Wett and Hartshorne (4696) noted that an over-
loaded condition of a pharmaceutical plant was relieved by
the installation of plastic media in the trickling filter
oxidation tower. The capacity of treatment was doubled
and a 99$ reduction in BOD was maintained,
Patil et al. (331O) had difficulty using biological filtra-
tion for the treatment of complex waste waters resulting
from the manufacture of synthetic drugs, even when procedures
of dilution and some acclimation were employed. Kiby (2471)
reported on process modifications which were made in 1934
for the manufacture of yeast which drastically changed the
waste water quality and it was noted that sedimentation fol-
lowed by discharge into lagoons or field irrigation was un-
satisfactory as the method of treatment of these waters.
Pitts (339O) reported that loadings of 5,OOO Ib BOD/ac-f/day
(115 Ib BOD/1,OOO ft3/
-------
Based on this review, biological filtration should be used
in conjunction with other processes for the treatment of
fermentation and pharmaceutical wastes. Problems arising
with antibiotics and other drugs in the biological systems
have been met rather routinely. There was no concerned in-
dication in the literature of any relationship between viral
production or transfer in treating wastes from vaccine pro-
duction facilities. It may appear at first glance that the
injection of effluent from pharmaceutical waste treatment
facilities into ground water basins is a bad practice. How-
ever, the limited travel distance of most life forms in
alluvial material, clays, or other subterranean aquiferous
materials must be considered. Biological filtration was
shown to be effective for the treatment of these high organic
content wastes. The requirement for extensive land area
for biological filtration to achieve the required quality
of treatment may, in part, be eliminated by fabricated medi-
um and other process modifications.
307
-------
SECTION XXVIII
PULP AND PAPER WASTES
The use of biological trickling filters on pulp and paper
mill waste was reported often since the early 19OO's (2O79).
To evaluate the effectiveness of the treatment and character-
ize the waste, McGowan (2911) in 1921 devised an indexing
system based on dissolved oxygen absorption. Paper mill
wastes had an index of 22, distilleries 33, tanneries 22,
and printing and dyeing 25. Sciver (3926) criticized
McGowan's formula and suggested that trade wastes,such as
from the pulp and paper industry, increased the suspended
solids content of the sewage and this characteristic made
sedimentation more difficult. However, waste waters from a
paper factory in Europe (4654) contained little fibrous mate-
rial, but were heavily colored.
Southgate reported (4118) the characteristics of a paper mill
waste to be: (a) 18,OOO gal./day of spent liquors and wash-
ings from digestions of rags and cold washings from straw
and wood with a BOD5 of 4,6OO mg/1 and 1$ alkali (as Na2CC>3) ;
(b) 1OO,OOO gal./day waste water from potchers and concen-
trators with a BOD of 25O mg/1 and neutral; and (c) an average
of 12,OOO gal./hr of waste water from paper making machines
with an average BOD of 11O mg/1 and neutral or slightly
alkaline. The mixed waste waters after sedimentation had
an average BOD of 23O mg/1, with liquors containing spent
lye washings from the digested material being the most dif-
ficult to treat. About 540,OOO gallons of waste water were
produced from a process which required 3O,OOO gal./ton of raw
materials per day (4118) . Waste water from the cardboard or
hard board manufacture was found by Schmidt (3873) to have a
BOD5 of 338 mg/1, a permanganate demand of 1, 58O mg/1 and a
phenol content of 3 to 4 mg/1 with the pH being 3.6 to 6.
Lime coagulation, superphosphate addition,and recirculation
were necessary to alter the characteristics of the cardboard
waste waters for efficient biological treatment. Mueller and
Schulz (31O9) reported on the characteristics of the waste
water resulting from the bisulfite process to have a pH of
3.8 to 6.7, with a permanganate value of 79O to 3,475 mg/1,
the BOD 5 of 65 to 368 mg/1, and with nutrient addition re-
quired for trickling filter or other biological treatment.
Eden et al. (1O99) discussed the three types of waste waters
from a paper mill: (a) black liquor at 13,2OO gal./day being
comprised of spent lye, first wash water, and flow washings
from the digester house; (b) strong washings amounting to
55,2OO gal./day composed of latter stage wash water and
bleaching of digested hemp; and (c) 1.65 mgd composed of
washing, bleaching and beating of materials other than hemp,
and discharges from paper machines, lime sludge from the water
softening plant, and some storm water. It was reported by
Holderby (1977) that the waste waters from the manufacture
of cellulose by the sulfite process contained lignin, sugars,
sulfur dioxide, and organic acids. Krussman (2579) outlined
3O9
-------
pilot plant studies for treatment of waste waters from a
paper mill. Waste water from this plant included rag cook-
ing liquor, wash water, white liquor and power plant dis-
charge.
Gehm (1442) concluded that de-inking waste waters were not
treated satisfactorily by trickling filtration because of
their high alkalinity. He also concluded that percolating
filters were not suitable for treating waste waters from
kraft mills due to nonbiodegradable materials.
While the characteristics of the waste were investigated,
Tyler et al. (4491) studied the types of organisms most
suitable for seeding filters to treat waste water from the
sulfite process. By seeding with various organisms and
molds, and then removing the sulfur dioxide, they success-
fully treated waste liquor on percolating filters. The
American Society of Civil Engineers formed a task force to
investigate the biodegradability of alkaline waste waters
(4915) and it was concluded that alkaline waste could be
treated biologically without neutralization although less
oxidation would occur at high pH values. Because some paper
mill wastes contain various turpentine mixtures, polyvinyl
chloride^ plastic medium (497O) was used in biological
treatment of kraft pulping waste waters as well as in other
applications (643). Horlock et al. (2O34) and Eden et al.
(1O99) agreed that a supply of nitrogen, such as sewage or
ammonium sulfate, was necessary to maintain a stable flora
population in trickling filters. Paper mill effluents be-
came a source for the recovery of materials of economic value
in India (1615). However, the effluent still had 2OO mg/1
BOD and 3OO mg/1 suspended solids and, from processes other
than the sulfite process, required further treatment consist-
ing of sedimentation, with or without coagulation, and some
type of biological filtration. Waste waters, from the pre-
liminary treatment of flax fibers for the preparation of
specialty papers, consisting of spent cooking liquor and
other cleaning waters, were discharged at the rate of 3O,OOO
gal./day, of which 13,OOO gal./day was spent black liquor
with an overall BOD5 of 3O,OOO to 4O,OOO mg/1 and 13O,OOO
to 14O,OOO mg/1 suspended solids with O.3 to 0.5$ caustic
soda (3226). A specialty paper manufacturer produced 1.3
mgd waste water which had a BOD of 25O to 300 mg/1 and con-
tained 15 to 2O Ib suspended solids (9).
The waste from the steam distillation of wood hydrolysis
plant was described by Black and Minch (34O) to vary great-
ly in volume and in BOD and to contain soil, wood fibers,
tannin, rosin, oil, solvents, and furfural. Anaerobic di-
gestion and percolating filters were used by Roznoy (37O8)
to treat cellulose acetate wastes, which were composed of
acetic acid, acetate, cellulose acetate fines, and sugars
from hydrolysis. The waste waters were characterized by
31O
-------
BOD5 of 2,7OO to 3,90O mg/1, pH of 3.5 to 4.3, and total
solids of 9,7OO to 13,2OO mg/1. After biological filtration,
the waste water from a wood hydrolysis plant was shown by
Tkachenko (4396) to contain appreciable amounts of wood and
gum sugars, acetic acid, and traces of formic and levulinic
acid. As a follow-up to this work, he (4397) also studied
the effect of temperature on the microorganisms in percolat-
ing filters treating waste waters from hydrolysis plants.
He showed that exposure to low temperature for several days
had no effect on the viability of the organisms.
Successful removal of 27 to 7OO mg/1 of furfural in a spent
alkaline fermentation liquor with a BOD5 of 40O to 5OO mg/1
was reported on a wood hydrolysis plant waste by Drublyanets
and Ivanova (1O33) . They also offered an explanation of the
removal kinetics during biological oxidation of hydrolysis
plant effluent (1O34). Detailed investigations of the waste
water from wood hydrolysis plants by Tkachenko and Yudina
(4398) indicated that Escherichia coli did not survive the
strong disinfecting action of these waters, A large number
of yeasts have been isolated by Tkachenko (4399) from per-
colating filters treating waste waters from wood hydrolysis
plants.
PRETREATMENT REQUIRED
Because of the general characteristics of pulp and paper waste,
usually some effort was required to make the waste amenable
to biological treatment. Ackerman (9) stated that sedimenta-
tion and equalization in a lagoon prior to biological filtra-
tion were desirable based on results of pilot plant studies.
Preliminary treatment of the waste in a save-all yielded an
effluent which could be treated satisfactorily by biological
filtration according to Richey (3593). A Spaulding Precipi-
tator was used by Stahl (4162) to remove 44 to 68^ of the
suspended solids in paper mill waste water effluent prior
to filtration. While sand filtration proved unsatisfactory
because the solids rapidly clogged the bed, quiescent settling
and withdrawal of the supernatant liquid through fine screens
was reported by Geer (1439) to be adequate in a tank on the
fill and draw method operated with a detention time of 1.5
hours. Burton (544) was issued a patent on a system which
involved entrapment of colloidal particles by prewetted
natural bark fibers. The effluent of this patented process
was treated in special percolating filters. An inclined
screen was used by Howard (2052) with the filtrate being
stored in tanks so that uniform liquor concentration could
be obtained for the paper mill waste stream.
The possibilities of reusing the soda in washing waters and
the black water in beaters as well as treating the waste
311
-------
waters from paper mills by sedimentation, with and without
coagulants, were discussed by Southgate (4118). Hodge (1965)
supported the concept of reuse of the water and recovery of
byproducts from the pulp and paper waste streams, as well as
treatment in lagoons, chemical coagulation and sedimentation.
Eldridge (114O) stated that nine tons of material per day
were recovered from a flow of 15,20O gph of waste when 95
mg/1 alum were added. Seven tons of usable materials were
recovered from a dry paper machine with 60O pounds of alum
added per day to a 85,6OO gph flow of waste.
Dilution with fresh water or effluent and chemical addition
was a common practice (1O34, 3226). The additions of
ammonium sulfate and superphosphate as nutrient sources were
also common, as for example, Drublyanets and Ivanova (1O34),
as well as others (1O99, 3873). Coagulation of mill waste
waters with alum, calcium chloride, or ferric chloride was
proven impractical by Oldaker (3226), as large amounts of
sludge were produced and the maximum reduction of BOD was
only 5O$. Eden et al. (1O99) used neutralization and dilu-
tion as well as dechlorination and addition of nutrient salts
to handle waste waters from a mill manufacturing special
grades of paper from hemp, linen, cotton, and other textiles.
A two-stage biological treatment process in which activated
sludge was followed by biological filtration provided 8O$
reduction in BOD; yet Schmidt (3873) stated that coagulation
with lime reduced the BOD by 4O$ and permanganate value by
2O$. One type of pretreatment frequently practiced was de-
scribed by Black and Minch (34O) in which the waste was
segregated into four distinct elements, where controlled
filtration, sedimentation, coagulation, screening, as well
as biological filtration, could be effectively used to
provide the treatment required.
Studies by Riiffer (3777) on coagulation of the wood process
waste waters with lime indicated that a pH value of about 11
was required to create the large floes in the hard fiberboard
waste stream necessary for good removal. Gascoigne (1422)
discussed the utilization of chlorine for the treatment of
white water from paper mills. Hommon (2OO1) commented in
1916 that coagulation of straw board waste with aluminum
sulfate, followed by rapid filtration, was not very success-
ful. In 1910, Clark (692) decolored paper mill waste by
intermittent sand filtration for a period of time, but best
results were obtained by adding calcium hydroxide and ferric
chloride in an amount not over one ton per million gallons
treated. However, Merryfield (2979) later reported that
waste from the flax industry was precipitated with O.5$
lime solution, and was followed by biological filtration.
Rohde (3669) described the successful handling of an in-
dustrial waste problem which involved two paper factories
by using iron sulfate and lime added to the sewage entering
the sedimentation tank and allowing the sewage to be aerated
in a preliminary section of this tank. This effluent was
312
-------
thereby rendered amenable to biological filtration on en-
closed filters containing a media bed of lava slag four
meters (4.38 yd) deep.
Anaerobic digestion before discharge to a sewage treatment
plant was used by Rohde (3671) on tannery and straw board
mill waste in Germany. As a preliminary or partial treat-
ment of straw board waste water, digestion of a mixture of
waste waters and digested sludge from domestic sewage produced
an anaerobic flocculant sludge which settled easily and had
strong adsorptive properties (2353).
Hydrolysis plant effluents were diluted with fresh water or
low BOD effluent, and nutrient added to aid biological treat-
ment (3166). In treating the waste waters from thermal treat-
ment of wood, aluminum sulfate was the best coagulant tested
by Busnita (553) and with the addition of nutrients gave suf-
ficient pretreatment so that admixtures of the waste water
with sewage at a ratio of 1:3 could be handled by biological
filtration.
EFFICIENCY OF TRICKLING FILTER APPLICATION
The efficiency of the biological trickling filter process was
measured over the years by several different techniques, for
instance, McGowan (2911) and Mueller and Schulz (31O9). Dur-
ing his pure culture studies, Tkachenko (4399) found that the
yeast Candida tropicalis was able to reduce the BOD of wood
hydrolysis plant waste water by 61 to 64$. Investigations
with acid sulfite liquor by Holderby (1977) determined that
BOD5 was reduced from 6O to 65$ with loadings of up to 3.O3
Ib BOD/yd3/day (112 Ib BOD/1,OOO ft3/day). He concluded that
the percolating filter had the ability to reduce the BOD of
sulfite liquor mixtures by more than 75$. Mueller and Schulz
(31O9) reduced by 7O to 75$ the BOD of the waste water from
the cellulose industry's sulfite process by the trickling
filter.
Laboratory scale apparatus was applied to treating cellulose
acetate waste. Roznoy (37O8) found that, by double filtra-
tion with an applied load of 2 Ib BOD/yd3/day (74 Ib BOD/
1,OOO ft3/day) and the recirculation ratio of 3:1, the BOD
decreased from 1,170 mg/1 to 174 mg/1 in the first filter.
This BOD was further reduced to 82 mg/1 in the second filter
under the optimum loading limit of 1.6 Ib BOD/yd3 (59 Ib BOD/
1,OOO ft3/day). Based upon a general review of methods in
treating paper mill effluent, Horlock et al. (2034) reported
the improved efficiency obtained by using alternating
double filtration and recirculation of effluent as well as
supplying nitrogen by sewage or ammonium sulfate addition.
During experimental studies by Tyler et al. (4491), the fil-
ters were loaded at 25 Ib BOD/ydVday (925 Ib BOD/1, OOO ft3/
day) , and it was shown that approximately 77ft of the sugars
313
-------
and 49$ of the BOD were removed from the sulfite process
waste water. When the load was dropped down to 10 lb/yd3/
day (37O Ib BOD/1,OOO ft3/day), the corresponding reductions
were 8O$ of the sugars and 76$ of the BOD. Further studies
done at elevated temperatures and higher concentrations of
sulfite waste liquors led the authors to the conclusion
that operation above 25°C was not economical and that as
high as 8O$ sulfite waste liquor could be treated on the
filters without adversely affecting the efficiency of treat-
ment .
Eighty-five percent removal was experienced on chip board
waste on a low rate trickling filter by Richey (3593) with
an organic loading of 15 Ib BOD/1,OOO ft3/day and a hydraulic
load of not more than 1 mgad (23 gal./ft2/day) . By using a
1:1 recirculation ratio of effluent at the same organic load-
ing and a total hydraulic loading including recirculation of
2 mgad (46 gal./ft2/day) , he achieved 93$ removal of the BOD.
In pilot plant investigations on kraft pulp mill effluent in
Poland, Nowacki and Pilotek (32O7) obtained optimal results
with a BOD loading of 1,42O g/m3/day (89 Ib BOD/1,OOO ft3/
day). A hydraulic loading of 1.53 m3/m2/day (37.5 gal./ft2/
day) was used to provide reductions in BOD and permanganate
demand of 80$ and 17.5$, respectively. The authors noted,
however, that the removal of color was poor and that re-
circulation of effluent did not improve treatment efficiency.
Voight (4546) described a two-stage biological process treat-
ing straw board waste waters which reduced a BOD of 2,5OO to
less than 6O mg/1 within the effluent suspended solids of
less than 5O mg/1. The results of a study by Brebion (449)
indicated that ion exchange, filtration, and anaerobic diges-
tion processes were inapplicable to industrial scale situa-
tions. Activated sludge and biological filtration appeared
feasible enough to treat waste from industrial processes
permitting 8O to 9O$ reduction in BOD in 36 to 48 hours.
Several packing arrangements of the media in biological trick-
ling filters have been used on pulp and paper mill effluent,
e.g., Renzi (356O) , Minch et al. (3O24), Crawley and Brouillette
(815), Follett (13O5), Egan and Sandlin (1119), Eckenfelder
and Barnhart (1O82), and Middlebrooks and Coogan (2993).
Results from these studies using fabricated media indicated
BOD loadings of 3.5 times that of conventional rock could
be used (356O) . BOD removal was 9O$ for fine paper
effluent (13O5). Loadings of 750 Ib BOD/1,OOO ft3/day
were applied without impairing the efficiency of the opera-
tion and gave removals as high as 37O Ib BOD/1, OOO ft3/day
(1119). The tower oxidation approach was reported by West
(468O) and Ganczarczyk et al. (14O5) as desirable for aera-
tion and for the reduction of BOD and concentration of
oxidizable substances in the mixture of pulp wash water and
bleaching effluents. Influent hydraulic loads of 1 gpm
314
-------
and recirculation of 7 gpm were reported by Middlebrooks
and Coogan (2993) to effect 5O$ BOD removal on the plastic
filter media. When toxic shocks did not occur, an average
of more than 6O to 7O$ of the BOD was removed at surface
loadings of 5.7 gpm/ft2 or 37O mgad (8,600 gal./ft2/day).
They favored fabricated media for roughing filters as
efficient treatment devices capable of removing 6O^ BOD at
3.7 gpm/ft2 (5,328 gal./ft2/day) and a recirculation rate
of 2 gpm/ft2 (2,88O gal./ft2/day).
Many of the laboratory and pilot investigations on the bio-
logical treatment of paper mill wastes have been expressed
in terms of mathematical relations, e.g., West (4680). Sev-
eral of the mathematical models dealing with paper mill ef-
fluent data were evaluated statistically by Robertson (3647).
Additional mathematical discussion was previously covered,
but it is of interest to note that the relationship described
by Caller and Gotaas (1397) was found to best fit the pre-
diction of the trickling filter performance while operating
on paper mill waste.
COMPARISON TO OTHER METHODS OF TREATMENT
The troublesome characteristics of paper mill and wood hy-
drolysis waste were indicated by the pretreatment required and
the efficiency of the trickling filters. Many methods were
investigated, and Grant (1538) discussed the mechanical,
chemical, biological, and preventive methods for reducing
the strength of the waste before it reaches the receiving
water. Biological filtration was generally preferred to
activated-sludge treatment in spite of the large plant re-
quired, but a closed circuit system of counter washing was
strongly recommended. Blosser (366), under the auspices of
the National Council for Stream Improvements, worked with
aerated lagoons and found that four days' aeration was re-
juired for an 8O$ BOD reduction on de-inking waste and that
the sludge produced and the quality of the effluent were
similar to those from the activated-sludge process.
In a literature survey, Rennerfelt (3558) reviewed the com-
position of waste water from forest product industries, bio-
logical treatment, aerated and nonaerated stabilization la-
goons, percolating filters and the activated-sludge process.
A review by Gleeson (1486) included methods of treatment,
such as the activated-sludge process, biological filtration,
aeration utilizing the tendency of liquor to foam and ac-
celerate oxidation, fermentation with production of fodder
yeasts or alcohol, evaporation with or without burning,
disposal in lagoons, the Howard process of coagulation with
lime, hydrolysis at high temperatures and pressures, and
ion exchange. In 195O, Holderby and Wiley (1978) reviewed
the result of the work of the Sulfite Pulp Manufactuers
Research League, Inc., as: (a) the activated-sludge process
315
-------
would reduce the BOD of the waste sulfite liquor by 95$;
(b) percolating filters would reduce it by 65$, and under
favorable conditions 75$; (c) contact aeration would reduce
it by 7O to 75$; (d) aerobic methane fermentation was not
satisfactory; and (e) promising results were obtained from
the cultivation of yeast. Each of these processes had its
limitation under a given set of circumstances and must be
evaluated with this consideration.
Studies by Brebion (449) in France showed that ion exchange,
filtration and anaerobic digestion did not have the advantages
of biological filtration. A porous partition-type apparatus
was compared by Brebion and Huriet (452) to conventional per-
colating filters, which was an advantage to the trickling fil-
ter process due to lower construction cost. In contrast to
other work (449), an anaerobic filter, in which the flow of
waste water was upward through the filtering material, was
successfully used by Dewes (942). A phosphate-impregnated,
honeycombed asbestos paper medium was used by Quinn and
Fronzoso (3495) to reduce the BOD by 61$ at the greatest
flow and 86$ at the lowest flow, which was ten times greater
per unit than the flow generally applied to trickling filters.
Tabb (4288) has developed a process for treating dilute
waste waters from mechanico-chemical processes of digestion
by passing carbon dioxide through the waste waters to con-
centrate the organic matter.
Autooxidation used by Riley et al. (3616) on wood hydrolysis
waste showed promise for complete treatment of the waste
water. Experiments using only mild autooxidation still re-
quired the waste to be diluted 1:9O before biological treat-
ment. Temperature and the partial pressure of oxygen con-
trolled the reaction rates. High pH wastes were biologically
treatable (4915) although less oxidation occurs at these pH's.
Davis (884) found an experimental trickling filter unsatis-
factory for the treatment of a sulfite mall waste and it was
abandoned. However, an experimental bio filter reduced the
BOD of the impounded waste by 3O$, removing about 2.1 Ib BOD/
yd3/<*ay (7.8 Ib BOD/1, OOO ft Vday) • Lebo and Hassler (2661)
indicated that laboratory investigations of integrated pulp
and paper mill wastes were amenable to both activated sludge
and biological filtration. Subsequently, pilot scale, two-
stage trickling filters were shown to give a total reduction
of more than 65$. Irwin and French (2235) reported experi-
ments on felt paper mill effluent which indicated that treat-
ment in lagoons would not be practical and chemical coagula-
tion of effluent would be expensive and ineffective. Based
on pilot percolating filter plant data, a full-scale filtra-
tion plant with recirculation was built; this treatment re-
duced the effluent BOD to less than 5O mg/1.
Contrary to the above results (2235), Gehm (1443) in 1957
stated that percolating filters have not yet been proved to
316
-------
be effective in the removal of BOD, but modifications of the
activated-sludge process have shown promise. Ten years ear-
lier, Holderby (1977) was so convinced of the effectiveness
of biological filtration for the treatment of sulfite waste
liquors that he concluded it had greater possibilities of
development than treatment by the activated-sludge process.
Recently, Wheatland (4698) reviewed both the biological fil-
tration and activated-sludge process as they were applied to
the biological treatment of waste waters from paper and board
mills. Two-stage treatment by an activated-sludge unit and
biological filtration was used by Ganczarczyk (14O3), Gellman
(1449), and Nowacki and Pilotek (32O6). After investigating
several^methods including chemical coagulation, biological
filtration, and activated-sludge processes, lagooning and fil-
tration through fly ash, the activated-sludge process was
chosen by Palladino (3277) for reasons of economics, efficiency,
and ease of operation. Ratliff (35O8) found that neither acti-
vated sludge nor biological filtration was satisfactory for
the treatment of wastes from the manufacturing of jute liners,
chip boards and corrugating materials from waste paper. He
concluded that a circulating lagoon was required for the
treatment of this particular paper waste to reduce the BOD.
However, the effluent from the lagoon required further treat-
ment which was obtained by biological filtration. After
studying the effect of activated-sludge process and biologi-
cal trickling filters on wood fiber waste, it was noted (2541)
that the best results were obtained with the "P-method of
Magdeburg." In the presence of iron salts this method gave
a reduction of greater than 7O$ BOD5 and 5O$ permanganate
value after six hours of aeration with a sludge concentration
of 2.2 kg/m3 (6.3 lb/yd3).
Typical of many of the papers presented on the effectiveness
of activated sludge or trickling filters was that of Morgan
(3O93) in which he reported that an influent BOD of 414
mg/1 and suspended solids of 1,6O2 mg/1 were reduced to: (a)
169 and 1, 6OO mg/1 after six hours' aeration which had 12.5
mg/1 sodium nitrate added; and (b) 358 mg/1 BOD with no
nitrate added and 295 mg/1 BOD where nitrate was added after
passage through a six-foot deep percolating filter. The ef-
fects of thermal loading were studied on activated sludge and
synthetic medium trickling filter-cooling tower combination
units and evaluated by Burns and Eckenfelder (538). These
tower units were a feasible addition to existing waste treat-
ment facilities, but due to the inhibitory factors of paper
mill wastes there was only slight improvement in the final
effluent BOD.
POST-TREATMENT AND EFFLUENT QUALITY
Post-treatment required by a trickling filter operation for
the treatment of paper mill effluent was investigated by Gehm
317
-------
and sponsored by the National Council for Stream Improvement,
Inc. (1441), where by alternative processes the uses of ef-
fluent were explored. Grant (1537) reported that waste
waters from paper mills which do not make their own pulp can
be treated satisfactorily by flotation, settling in open
beds, or filtration. Sheen (3976) emphasized that the
presence of color, toxic compounds, and substances which
would cause taste in a municipal water supply must be dealt
with using mechanical, biological, and/or chemical methods.
Chemical clarification plus organic removal by percolating
filters alone and in conjunction with the activated-sludge
process have been used.
Typical of the many patents on treating paper mill wastes
was that issued to Metallgesellschaft (2982). Ferrous or
aluminum sulfate in the presence of activated sludge or a
mixture of similar organisms was used as a post-treatment
to remove the colloidal organic material. Gehm and Gellman
(1444) observed successful operation of paper mill effluent
treatment by biological filtration followed by the activated-
sludge process, with the activated sludge returned to the
influent of the filter. Some paper mills found it economical
to transport their cooking liquor waste to other locations to
be treated, as noted by Krussman (2579). The treatment con-
sisted, typically of alum and Separan® 2O* coagulation, sedi-
mentation, biological filtration, and chlorination of the
effluent before discharge to the receiving creek. Burns and
Eckenfelder (538) used two fabricated media trickling filters
ahead of an existing activated-sludge plant to take advantage
of the rapid BOD reduction and cooling effect. Biological
trickling filtration as post-treatment of anaerobic digestion
was not efficient for single filtration; however, double
filtration on a laboratory scale proved desirable, according
to the results obtained by Roznoy (37O8). Clarification was
a common post-treatment procedure (3O93).
SPECIAL OPERATIONAL PROBLEMS
In determining the efficiency of biological processes treating
paper mill effluent, Kalbe (2366) found that the 5-day BOD
test (as determined by the dilution method) could give mis-
leading results; but the Warburg method gave an accurate
picture of the course of decomposition of the organic com-
pounds. Horlock (2O33) discussed the application of the BOD
test to pulp and paper manufacturing wastes and stressed the
need for a standard synthetic dilution water to measure the
effects of treatment by percolating filters and by the
activated-sludge process and anaerobic fermentation. The
Department of Scientific and Industrial Research (England)
has been active in developing tests to evaluate the perform-
ance of various unit operations (5175).
*Registered trademark of The Dow Chemical Company.
318
-------
Sciver (3926) observed that suspended solids in paper mill
effluent usually made sedimentation more difficult; however,
cases have been experienced where paper solids acted as
coagulants. A breakdown of treatment costs using the activated-
sludge process, trickling filters, lagooning, and land dis-
posal for the treatment of paper mill effluents was prepared
by Caron (621). Research on the development of treatment
systems to handle waste from the sulfite process was financed
and supervised by the National Council for Stream Improvement
(5391) to provide operational and design data on trickling
filters and stream aeration. This sponsored work was based
on laboratory experience of the Wisconsin Sulfite Pulp
Manufacturers' Committee on Waste Disposal (5662).
A patent was issued to Neil (3155) on a specially operated
filter which had vertical surfaces and retainers which accu-
mulated solid material such as from pulp effluents. A tem-
porary operation was used during the period of reconstruction
in Germany after World War II where paper factory waste waters
were discharged directly to the sewerage system after coag-
ulation (1811). Typical of the uses of sludge generated by
the treatment of waste waters from paper factories was that
described by Lesenyei (2685), where the sludge was mixed with
peat and used as compost at the percolating filter treatment
plant. Studies in Japan (2163) indicated that the treatment
of calcium-based acid sulfite process waste water by biologi-
cal filtration and the activated-sludge process was effective
in removing organic matter, but not effective for the removal
of lignin. Investigations on the removal of furfural from
wood hydrolysis plant waste waters demonstrated that biologi-
cal filtration would completely oxidize this material (1O33).
Critique
Considerable work was published on the use of biological
trickling filters for treatment of pulp and paper related
waste waters. The evidence was presented repeatedly that
with certain pretreatment, e.g., pH and suspended solids,
the waste was biologically treatable. Several papers were
prepared on the advisability of using trickling filters
versus other systems such as the activated-sludge process.
Data were not available to document the reasons for the
decision of the selected treatment based on economics. Per-
formance information was available and could be compiled and
reduced for common comparison. Trickling filters used in
combination with the activated-sludge process and other
systems were reported several times with essentially the
same conclusion, that it was a desirable approach.
This industry should be congratulated for its efforts to
provide resources and emphasis on clean water through the
formation of agencies like the National Council for Stream
319
-------
Improvement. Cooperative efforts of these agencies and
other industries to develop treatment systems and measuring
techniques have provided the basis for contemporary efforts.
Methods other than biological treatment have been used, but
the large quantities of water required in this industry
necessitate the most economical form of treatment. Often
this economic criterion determined that biological treatment,
and specifically trickling filters, alone or in combination,
is the pollution abatement answer.
32O
-------
SECTION XXIX
RADIOACTIVE WASTES
General surveys have been sponsored by the Atomic Energy
Commission (5578) to investigate design factors and behavior
of unit operations dealing with the removal of radioactive
isotopes. From this work, information was disseminated on
many aspects of radioactive waste disposal. Pazdernik (3317)
and Ruf (3776) reviewed the general literature available on
the treatment of radioactive wastes, such as that from the
waste waters from laundries, and concluded that both activated
sludge and biological filtration were very effective.
The concentration of radioactivity throughout waste treatment
facilities and the various unit operations were discussed by
Kenny (2434) and it was found that very little adsorption
of radioactivity by sewage solids occurred during primary
sedimentation. Treatment by activated sludge or biological
filtration partially sorbed iodine-131, cobalt-6O, phosphorus-
32, and carbon-14, while sodium-24, potassium-42, bromine-
82, sulfur-35, strontium-89 and 90, and radioactive tritium
passed straight through. Investigators such as Newell (3178)
and Colas (753) in 195O studied radioactive waste pollution
and described methods of treatment, such as ion exchange,
distillation, evaporation, sand filtration, biological fil-
tration, absorption by volcanic ash, adsorption, lagoon
treatment, chemical coagulation, activated sludge, and ulti-
mate solids disposal. Radioactive waste treatment facilities
were also described by Sumiya and Muramatsu (4268) to in-
clude biological filtration with other unit operations.
Pettet (3355) and Brown (499) recognized radioactive waste
disposal to be a significant problem and established special
groups to study and handle the waste and suggested alterna-
tive methods of treatment. Newell et al. (3176, 3177) char-
acterized radioactive contaminated laundry wastes as con-
taining soap, synthetic detergents, citric acid, and lint,
and had a BOD of 2OO to 60O mg/1, and an average alpha count
of 1,OOO up to 2O,OOO/min/1. Pilot plant studies in-
dicated that plutonium can be removed from waste waters by
coagulation with ferric chloride and lime, followed by slow
filtration through sand (3176).
PRETREATMENT REQUIRED
A fundamental paper was presented in 1951 by Belcher (248),
which dealt with the theory of adsorption of radioactivity
by suspended solids and the relationship of sewage treat-
ment processes for the removal of five commonly used radio-
isotopes. Later, Eden et al. (11OO) also described similar
experiments and their results to remove common radioisotopes.
321
-------
The Atomic Energy Commission built its first full-scale plant
for radioactive waste disposal near Idaho Falls, Idaho
(327) . The radioactive waste waters were mixed with domestic
sewage in the ratio of 1:1O. Treatment was by primary sedi-
mentation and biological filtration with continuous recircula-
tion of the effluent in the ratio of 5:1. Secondary sedi-
mentation was also used and the chlorinated effluent dis-
posed underground.
The effect of strontium-89 was measured by Ilyin (2166) who
observed that the maximum concentration of the isotope in
the biofilm was 1 x 1O~5 cAg of raw weight. This amount
of isotope did not affect the biochemical activity of the
microorganisms or increase the weight of film. Similarly
Zhogova (4861), during experimentations with polonium-210
and strontium-9O, did not observe any evidence that the
presence of radioactive contaminants had any harmful effect
on the biofilm or on the purification of sewage.
EFFICIENCY OF TRICKLING FILTER APPLICATION
The U. S. Atomic Energy Commission has published reports
(5577) dealing with the effectiveness of sewage treatment
processes for the removal of radioisotopes. Studies on
activated-sludge process and biological filtration were
described for different isotopes. The removal of fission
products by biological filtration varied from 7O to 85$
and was unaffected by recirculation of the effluent or by
variations in BOD loading from 75O to 3,O9O Ib/ac-f/day
(7.2 to 71 Ib BOD/1,000 ft3/day) and hydraulic loadings of
2.2 up to 12.8 mgad (5O.6 to 294 gal./ft2/day) . Chemical
and biological techniques were investigated by Newell et al.
(3177) and operation of two-stage percolating filters with
a recirculation ratio of 6:1 gave satisfactory results with
a BOD being reduced by 9O$ and low rates of application at
0.3 mgad (6.9 gal./ft5/day) produced the required activity
removal. Dobbins (984) demonstrated that 90$ removal of
mixed fission products from radioactive laundry wastes
contain detergents and organic acids were removed by bio-
logical filtration. Two-stage filtration removed about the
same amount of activity that single stage filtration did
when the total volume of the filter medium was the same. A
percolating filter packed with pozzolana medium was effective
(4822) in the removal of radioactive waste waters. After
passage of these waters through the bed four times, 75$ of
the original radioactivity was fixed on the bed while only
7$ was removed with the sludge.
Iodine-131 was added to settled crude sewage, and, as the
rate of application increased (632) from 2 to 6 mgad (46
to 138 gal./ft2/day), BOD removal decreased from 92$ to
75$; however, at the rate of 2 mgad (46 ga./ft2/day) 85$
of the iodine-131 was removed in 6-foot deep filters.
322
-------
Analysis of the biofilm indicated a concentration of iodine-
131 to give counts of 106/min/g of slime which contained
about 4.5 Mc/g dry weight. Percolating filters were used
at the Brookhaven National Laboratory (5576) to remove
radioactive substances from laboratory waste waters. Removal
efficiencies were high enough so that only 1O$ of the phos-
phorus-32, 12% of iodine-131, 1% of strontium-90 - yttrium-
9O, and 2% of mixed fission products passed through the fil-
ter. This removal was interpreted as an average reduction
of 59$ in the activity of the waste waters from the labora-
tory.
COMPARISON OF OTHER METHODS OF TREATMENT
The efficiencies of biological filtration, activated sludge,
and oxidation ponds for removing radioactivity from waste
waters were compared by Kaufman et al. (2385) in 1956. In
all cases, they observed that the uptake of radioactivity
was inversely proportional to the concentration of the iso-
topic carrier. The activated-sludge and biological filtra-
tion processes were both effective in the removal or radio-
active materials from contaminated laundry waste (4715).
Liebmann (2715) used low and high rate percolating filters
and activated-sludge plants to remove radioactive substances
and indicated that the pH values, periods of aeration, and
other biotic factors were important in providing conditions
for sorption and subsequent removal.
Eden et al. (11OO) determined the efficiency of biota to
remove radioisotopes from water and specifically commented
on the performance of slow sand filtration and the activated-
sludge process. They concluded that for the elements in-
vestigated, which were io
-------
radionuclei discharge to the sea and receiving bodies of
water was discussed in considerable detail by Kenny (2434). He
reported in June, 1957, on the properties of radioactive
materials and the effects of alpha, beta, gamma, and X-rays
on human beings.
It was recommended by Wiederhold (4715) that two-stage opera-
tion or final coagulation of the effluent should be employed
to remove residual activity even though biological filtra-
tion removed most of the radioactivity. Procedures such as
this have been required, according to Newell (3177), to
reduce the plutonium activity in the discharged waste to
be less that 70 counts/min/1 which was accomplished by
chemical-biological treatment. Radioactive tritium passed
straight through the treatment process (2434), as did
cesium-134 and rubidium-lO6 (1489). These failures of bio-
logical systems required additional post-treatment.
SPECIAL OPERATIONAL PROBLEMS
Fowler et al. (1319) determined the conditions of adsorption
of rubidium-86 on biological percolating filters and they
observed that the adsorbed radioactivity was associated with
the biofilm growth. More activity was adsorbed at the one-
foot depth than at the surface of the filter. It was con-
cluded that the biofilm was important in the adsorption of
radioactive isotopes which, in the case of rubidium-86,
was not irreversible. Versene®*, the disodium salt of
ethylenediaminetetraacetic acid, was found to have an
adverse effect on the removal of fission products by both
activated sludge and biological filtration due to its che-
lating or complexing abilities (5577) . The relationship
of pH and period of aeration were considered to be a problem
(3776), and alternative methods and treatment experience were
related. After initial adsorption of the radioactive material,
the adsorption rate dropped off (1489), and cesium-134 and
rubidium-lO6 were not handled adequately by biological treat-
ment.
Critique
The literature indicated some advantage in using biological
filtration for the removal of radioactive wastes. The treat-
ment of these wastes, which is a relatively new industry, was
reported to involve chemical, physical, and biological waste
treatment methods. Evidence was given of the adsorptive
capacity of the biofilm for several radioisotopes. Problems
developed from relying on this adsorptive capacity for certain
elements which were not permanently removed from solution.
'Registered trademark of The Dow Chemical Company.
324
-------
Literature other than that reviewed here has shown considerable
interest in the removal of radioactive materials from waste
effluent. Investigators such as C. P. Straub and R. Eliassen
have reported extensively on the treatment of low and high
level radioactive waste. This literature review serves the
purpose of demonstrating investigators' awareness of the
adsorptive capabilities of biological filtration. Generally,
it has been shown that biological filtration requires pre-
treatment for most radioactive wastes. There are many radio-
active wastes at such high level discharge that methods other
than biological filtration must be used.
325
-------
SECTION XXX
TANNERY WASTES
The significance of tannery waste disposal problems was noted
by the formation of committees (5565) charged with the re-
sponsibility for several investigations dealing with storage,
sediment, filtration, chemical treatment (if necessary), and
secondary sedimentation. This committee recommended in 1931
that final filtration through six feet of coke should be
included after pretreatment by chemical precipitation. Cau-
tion was expressed that logwood should be used as the filter-
ing medium because acid in the waste water reacted with the
iron in the coke to color the solution blue-black. Wastes,
which were categorized by Russel (3787) as tannery wastes,
were formerly discharged untreated to receiving bodies of
water, but, according to Taylor (4328), were also combined
with the municipal waste water and treated by biological
filtration.
Reuning (3562) reported that tannery waste waters contained
lime, hair, and hide particles, and were very caustic and highly
colored. Treatment was generally based on results of experi-
ments in Pennsylvania between 1924 and 1930 (5565) and con-
sisted of coagulation, sedimentation, and biological filtration.
Under difficult situations, the colored water was evaporated
or treated in lagoons. A combined tannery and municipal
waste was characterized by Eddy and Vrooman (1088) to be
warmer than domestic sewage and to possess more organic
material. Kalibina (2367) reported in 1930 on experimental
work dealing with the treatment of tannery wastes by the
activated-sludge process, aerated contact filter, and per-
colating filter. He concluded that biological examination
is a rapid means of judging the quality of a purifying plant
and that the concentration of tannery waste water appeared
to have little influence on the purifying effect.
During a forum on industrial waste and related problems in
195O, McKee and Camp (292O) outlined particulate material,
high alkalinity and hardness, hydrogen sulfide formation,
and discoloration of receiving streams as being primary prob-
lems of tannery wastes which required treatment. In an
early effort to evaluate or characterize the relative
strength of tannery waste, McGowan (2911) developed an in-
dexing system based on oxygen consumption which was related
to the operation of biological filtration.
s
Tannery wastes were characterized by Dienert (972) to be
778 m3 (21O,OOO gal.) of washing water, 816 m3 (22O,OOO gal.)
of waste water from the "limes," 531 m3 (14O,OOO gal.) of
tanning water, making a total volume of 2,325 m3 (614,OOO
gal.) of waste water a day for 82,000 kg (9O tons) of skins
treated in a month. Rosenthal (3692) and Power (345O) dis-
cussed problems of tannery waste treatment and listed the
characteristics of the waste as having a pH of 11.8 and a
total alkalinity of 1,1OO mg/1 as calcium carbonate.
327
-------
Parker and Watlcins (3299) described the waste treatment plant
at Middlesborough, Kentucky, as a combined municipal and tan-
nery waste-treatment facility and characterized the tannery
waste as containing 8,500 mg/1 total solids, 1,7OO mg/1 sus-
pended solids, and 3,3OO mg/1 volatile solids with a BOD of
83O mg/1 and an alkalinity of 1,155 mg/1. They calculated
that the population equivalent of these wastes, based on sus-
pended solids, was 20,700, and based on BOD was 15,5OO, but
the actual population of the community was less than 12,OOO.
One mgd tannery effluent was handled in conjunction with 3.5
mgd of combined domestic sewage and milk processing waste in
a plant consisting of grit chambers, coagulation and sedimenta-
tion processes, standard and high-rate biological filtration
followed by final sedimentation (1768).
PRETREATMENT REQUIRED
Allen (35) reported pretreatment devices, such as fine mesh
screens, were installed at tanneries to remove hair and solid
materials from the influent, in a combination plant, the
tannery wastes were added to the sewage in a mixing chamber,
the pH was adjusted, and further treatment was accomplished
by a conventional trickling filter plant (3299). Morrison
(3O99) identified in 1911 that lime and tan liquors were
mutually precipitable and that these combined solutions could
be used to clarify other suspended matter in tannery effluents
prior to biological filtration. Two chemical purification
processes to handle tannery waste effluents were used by
Genin (1453), but a physical process, such as decantation of
filtration, for removing the precipitate was required. Based
on laboratory studies, pretreatment using lime or alumina
ferric dosed at 1 to 4 pounds/1, OOO gallons removed most of
the 25,96O mg/1 of leather dressing colloidal material (13O2).
An arrangement was reported by Hughes (2O84) whereby tannery
effluents were discharged to the municipal treatment plant
after midnight and were subsequently treated in a separate
settling tank prior to biological filtration. Aue (1O3)
described a separate waste treatment facility which was used
to treat tannery waste waters which involved flow equaliza-
tion, chemical precipitation, and reduction of pH, after which
the waste water was discharged to the municipal system where
sedimentation and high-rate biological filtration were used.
EFFICIENCY OF TRICKLING FILTER APPLICATION
Eddy and Vrooman (1O88) concluded that biological processes
could treat tannery wastes and that covered percolating fil-
ters could be dosed at 1 mgad (23 gal./ft2/day) following pre-
treatraent at the tannery site to maintain 9O$ removal of sus-
pended solids. Color was removed largely through the sprin-
kling filter and completely through the sand filter, but some
inoffensive odor was retained. Clark (692) stated in 191O
328
-------
that pretreatment with calcium hydroxide, followed by sand
filtration, could remove 6O$ of the organic matter from tan-
nery waste. Nemerow and Armstrong (3165) revealed the results
of an industrial waste survey in New York where 22 tanneries
had effluents contributing to the polluti - of the receiving
waters. The survey indicated (3165) that secondary treatment
of the combined industrial and domestic waste was necessary
to produce 65 to 97$ reduction in BOD. Humphreys and Bailey
(2089) studied the treatment of tannery waste alone and in
combination with domestic sewage, and it was determined that
biological filtration at a rate of 7O gal./yd3/day could be
sustained with trivalent chromium up to 3O mg/1 without suf-
fering a decrease in either BOD or permanganate demand effi-
ciency.
A two-stage biological treatment plant was found (2978) to be
particularly suitable for the treatment of tannery waste in
Germany. The activated-sludge process combined with biologi-
cal filtration produced satisfactory results as indicated by
daily tests on BOD, permanganate demand, pH value and turbidity.
The sewage disposal plant at Keighley successfully treated 4
mgd of sewage including a heavy trade waste (5O17) of 9OO,OOO
gal./day of primarily textile and tannery wastes by conventional
percolating filters. The final effluent contained an average
of 35 mg/1 suspended solids and 26 mg/1 BOD.
In 1964, Gorecik (152O) investigated the effect of phenol oc-
curing in tannery waste waters. He determined that for optimum
removal of phenol (86.2 to 94.6$) not more than 2,OOO grams
BOD/m3/day (125 Ib BOD/1,OOO ftVday) could be loaded on the
Biofilter. Experiments on the purification of tannery wastes
by chemical precipitation, primary filtration through cinders
at a rate of 25O,OOO gal./acre/day, and secondary filtration
through sand at a rate of 30O,OOO gal./acre/day were performed
by Hommon (2OO2) in 1917. The suspended solids were reduced
from 1,2OO to 3O mg/1 and the nitrogen was reduced from 70 to
25 mg/1.
COMPARISON TO OTHER METHODS OF TREATMENT
Processes which were reviewed by Chase and Kahn (662) and by
Dienert (972) included straining and settling, followed by the
activated-sludge process, filtration through crushed stones,
filtration through clinker followed by a sand filter and other
clarifiers. In 1960, Paszto (3308) carried out pilot plant
experiments on tannery waste waters using sedimentation,
chemical purification, biological filtration and activated
sludge. He indicated that chemical purification was expensive,
but a perfect solution to the problem. He concluded that bio-
logical filtration could be used only after 9O^ dilution of the
waste water, while the activated-sludge process could be used
after 6O$ dilution of the tannery waste with domestic sewage.
Schwarz (3923) detailed pilot plant studies on combined treat-
329
-------
ment of heavily polluted tannery waste waters to demonstrate
the waste to be more completely treated by the activated-
sludge process than biological filtration, in agreement with
Kubelka (258O). Kabakowa and Kalabina (2362) investigated
the purification of tannery wastes by the activated-sludge
process and slag contact tanks on a laboratory scale. They
concluded that the best treatment was obtained in the activated-
sludge system at 33% sludge and a non-putrefying product was
obtained by oxidation in contact with brown coal after chem-
ical coagulation.
However, biological filtration was noted (972) to give best
removal treatment. Allen (35) constructed and operated a
pilot plant which employed activated sludge and biological
filtration operated in series and parallel plus alternating
double filtration or recirculation of the effluent. He con-
cluded that alternating double filtration was the best
procedure.
Kubelka (258O) discussed the treatment of tannery waste
waters by soil filtration, fish ponding, percolating filters,
and the activated-sludge process. Primary sedimentation was
necessary and treatment by the activated-sludge process was
regarded as preferable to biological filtration. Soil con-
ditions made land disposal impractical and fish ponds
appeared possible, but were not tried. Thabaraj et al.
(4346) made a comparative study of some 5O investigators'
reports on the treatment of tannery effluents by trickling
filter, the activated-sludge process and lagooning. Pre-
liminary treatment, by mixing effluents from individual
tannery processes, resulted in flocculation and particulate
removal. No particular advantage was given to any treatment
system and the advantages and disadvantages of each were
identified. They concluded that the system choice would de-
pend on local, economic, operational and environmental factors.
Gononian (1514) suggested that oxidation, chemical precipita-
tion and filtration were required. Rohde (3671) described the
Niers Process for the treatment of sewage with high percentage
of waste waters from tanneries and textile mills. To purify
tannery waste waters, a filtration gallery-type system, in
which loosely layered granite stone was placed in a shallow
channel, produced results comparable with that obtained in
standard percolating filters, but further investigation was
deemed necessary to determine the effect of various process
factors on this system (4528).
POST-TREATMENT AND EFFLUENT QUALITY
Usual post-treatment was in the form of humus tanks, lagoons
and chlorination. Anthrax was noted as being well under con-
trol. Merz (298O) described methods for treating tannery
waste waters in East Germany which involved the removal of
330
-------
fat, sedimentation in lagoons, and sludge thickening, but tests
were in progress in 1961 to determine the effect of biological
treatment of the effluent in deep-bed percolating filters.
SPECIAL OPERATIONAL PROBLEMS
in 1950, Vrooman and Ehle (4565) discussed problems in treat-
ing the waste at Gloversville, New York, which contained a
high proportion of tannery waste waters. As was later re-
ported by Nemerow and Armstrong (3165), the existing waste
treatment plant was overloaded and a pilot plant was con-
structed to determine the possibilities of digesting sludge
which was obtained from the biological filtration process as
well as the high suspended solids of the tannery waste.
Hofer (1973) reported, in 191O, that tannery waste in sewage
was a detriment to biological purification and separate steps
were required for its treatment. Allen (35) noted that the
resistance to biological treatment of tannery waste water
posed a difficult problem and made the cost of separate
treatment prohibitive. Cohn (737) reported, in 1925, that
filter flies appeared during the treatment of tannery wastes
on biological filters at Norwood, Massachusetts. A dosage
of 1,6OO ppm chlorine was applied before the number of flies
were appreciably reduced, but this also materially reduced
the efficiency of the filters.
Critique
Tannery wastes were adequately reported to have characteris-
tics of significant BOD and excessive-suspended solids load.
Experimentation dealing with several biological, physical,
and chemical processes resulted in the publication of many
reports. Biological filtration was shown to be effective,
but papers show that situations arose where other systems,
such as the activated-sludge process, were more efficient.
The literature reflected good communication among investigators
by the formation of responsible committees to sponsor aggres-
sive research and development and to disseminate the results.
It appears from the literature that a reasonable approach to
the treatment of tannery waste has been adequate pretreatment
at the tannery site with discharge to a larger municipal treat-
ment plant. Pretreatment at the site was the removal of sus-
pended and colloidal solids, pH adjustment, and some roughing
filter-type oxidation. These pretreatment steps have been
shown to be necessary to dampen the extreme characteristics
of tannery wastes arising from the curing, flushing, washing
and soaking, unhairing, lime splitting, bating, pickling,
degreasing, bleaching, fat-liquoring, and dyeing processes.
Biological filtration has not been used as extensively in
this industry as in others reviewed. Inherent difficulties
from the characteristics of the waste have relegated the
treatment to lagoons and other low efficiency operations.
331
-------
SECTION XXXI
TEXTILE WASTES
The problem of waste waters from the textile industry causeu
sufficient interest that associations were formed (4876) to
investigate characteristics of waste waters and explore
methods of treatment. Alkaline wastes (characteristic of
textile wastes) posed a sufficiently difficult problem to
investigators that the American Society of Civil Engineers
sponsored (4915) research projects to investigate biological
oxidation of alkaline wastes. It was concluded that these
wastes could be treated biologically without neutralization
although less oxidation would occur at high pH values
(4915). Eldridge (1139), Pettet (3355), and Masselli et al.
(2881) generally reviewed industrial waste treatment on
municipal sewage treatment facilities and, in categorizing
them, noted that textile wastes required complicated chemical
treatment and their treatment with domestic sewage was not
generally recommended. However, Forges et al. (3434) treated
textile wastes at 24 mgad (552 gal./ft2/day) with recircula-
tion.
Wool scouring wastes were described by Wishart (4789) as
difficult to treat, but with sufficient dilution were amen-
able to biological treatment. An increase in organic matter
was to be expected due to contributions from cotton, flax,
hemp, and jute bleaching wastes as well as caustic alkalinity.
Rapidly putrefying silk wastes were treated adequately by
chemical precipitation. Dyeing waste waters contained sus-
pended material, and recoverable dyes/ such as indigo, were
removed at the factory while other dyes were subjected to
chemical precipitation. Gibson and Wiedeman (1466) stated
that textile wastes included caustic and soda ash, detergents
and soaps, desizing wastes, acids, reducing agents, resins,
and dyes.
Hart (1758) and Souther and Alspaugh (4117) concurred with
other investigators that the characteristics of cotton
bleaching and dyeing wastes were high alkalinity, suspended
matter, color, and BOD. Dickerson (953) characterized
cotton dyeing and finishing plant waste to have from 3,OOO
to 3O,OOO Ib BOD/day with waste flows from 2 mgd to 15
mgd. A major source of the BOD came from the cornstarch
used in processing the cotton which was washed away with
hot water. Textile dye wastes were discussed by Nemerow
(3163), in 1952, and categorized as direct dyes, sulfur
dyes, vat dyes, acetate dyes, and naphthol dyes. The wastes
were characterized as being alkaline with a high oxidative
demand, rich in color, had an elevated temperature, and
sulfur dye wastes were extremely toxic. Gauge (1429), in
1922, described the waste liquors from flax wetting to be
dark yellowish-green liquids with offensive odors and acidic
properties, which had less suspended matter and nitrogen
than raw sewage. These liquors were suitably treated by
333
-------
biological filtration after chemical treatment with calcium
oxide and aluminum sulfate. Sharda and Manivannan (3969)
categorized effluents from a viscose rayon factory as alka-
line, viscose, acid sulfide, and sanitary. Combined waste
treatment might be difficult, but physical and chemical
methods in addition to biological filtration were recommended.
Grey (1581) described a sewage plant to treat textile wastes
and domestic sewage. Treatment consisted of separate sewage
and trade waste influents, screening and comminution, equali-
zation and mixing, coagulation with alum, ferric sulfate,
and lime sedimentation, and biological filtration followed
by final sedimentation with sludge digestion.
Irwin and French (2234) and Lloyd (529O) reported that two-
stage biological filters with recirculation following screen-
ing and sedimentation provided satisfactory treatment for
felt mill and combined textile-sanitary effluents. However,
Gibson (1466) suggested that recirculation or organic load
had little to do with the BOD filter efficiency of trickling
filters and the activated-sludge process. Disposal of silk
mill wastes was described by Geer (1439), based on experi-
mental studies of sand filtration, various forms of settling,
and biological filtration. Later, a biofiltration plant for
silk mill wastes was operated by Jenks (23O8) on the high-
rate filtration principle. The high standards set for the
effluent from cellulose and viscose artificial fiber plants
caused Munteanu (3124) to build pilot plants incorporating
physical, chemical, and biological treatment, both by filtra-
tion and the activated-sludge process, to develop data
necessary for the design of full-scale plants,
Textile wastes were treated adequately on plastic media bio-
logical filtration systems which made activated-sludge tanks
and other conventional methods impractical (98). Biggs (326)
reviewed the literature and the capabilities of biological
treatment and required pretreatment for handling of textile
wastes and commented, specifically, on the application of
plastic media biological trickling filters. A discussion
(5196) on the use of plastic media biological trickling
filters to provide economical and satisfactory treatment of
textile wastes emphasized a roughing filter in congested
areas of,production facilities or waste treatment plants
requiring rapid removal of high oxygen demand.
PRETREATMENT REQUIRED
Pretreatment was necessary, according to Smallhorst (4O55),
for the treatment of cotton finishing plants using ferric sul-
fate and lime to treat strong sulfur dye wastes as well as
copperas and lime. Chemical treatment similar to present
methods was used (5423) in the early 19OO's such as lime and
ferrous sulfate, followed by percolating filters and humus
334
-------
tanks. Flue gas was used (1758) to lower the pH of bleaching
waste water as preliminary treatment. Textile waste waters
at Russian installations were treated (3871) by similar
methods of chemical precipitation followed by biological
filtration. Treatment was commonly reported (953) to be in
the form of chemical precipitation or biological oxidation,
using combined percolating filters and the activated-sludge
process. Textile effluents were reported by Kehren (2420)
to be successfully treated by chemical precipitation using
iron turnings, followed by sedimentation and biological treat-
ment in percolating filters or the activated-sludge process
to provide an acceptable effluent.
The replacement of organic acids with mineral acids in various
dye processes was suggested by Snyder (4O93). Preliminary
treatment suggestions were made by Horton and Baity (2O37)
to reduce the volume of waste and recover by-products at
the waste source. Anaerobic digestion of wool-scouring
wastes was used (2631) prior to discharge of mill effluent
to sewage works, where further biological treatment by
filtration was accomplished. Equalization of synthetic
fiber waste by admixture with domestic sewage and cooling
water, followed by additional biological treatment, was out-
lined by Singleton (4O39). McCarthy (29O1) reported on
methods of preliminary treatment to handle wool-scouring
liquors, such as steam stripping and chemical addition.
By diluting the waste waters from treating flax waste
liquors to 1$ with river water, it was possible to treat
them by biological filtration (3226).
EFFICIENCY OF TRICKLING FILTER APPLICATION
McCarthy (29OO) estimated that the waste waters from wool
dyeing had a BOD of 0.25 to l.O Ib/lOO Ib of wool dyed,
and could be treated by biological filtration after dilu-
tion to produce 9O$ BOD reduction. With recirculation of
the effluent, loads of 5,OOO Ib BOD/ac-f/day (115 Ib BOD/
10OO ft3/day) were reduced by 89$. Waste waters from a
desizing process contained digested sludge residues and
exerted a total BOD of 3,OOO Ib/day in combination with
other plant waste (4O93). Biological filtration was shown
to be effective for 75 to 8O$ removal of this BOD. By
diluting with river water to the point that 1% of the flax
waste existed (3226), treatment was accomplished on accli-
matized biological filters which were loaded at about 50O Ib
BOD/ac-f/day (11.5 Ib BOD/1,OOO ft3/day) and produced 57 to
72$ removal which, after 35 days' operation, was raised
to 70 to 90$ BOD reduction. Chemical pretreatment signif-
icantly removed color from textile dye wastes and, when
followed by biological filtration, the BOD was reduced
by at least 86$ and the color by 83$ (3163).
335
-------
Husmann (2149) found that waste waters from a textile mill,
which included waters from bleaching and dyeing, could be
treated on percolating filters without the addition of sew-
age, but with a recirculation ratio of 3:1 to reduce an in-
fluent BOD of 44O mg/1 to 1O7 mg/1. Seasonal acclimatiza-
tion was demonstrated by Petru (3341) to be effective in
purifying retting waste in which the effluent improved as
the retting season continued. Dahlem (854) treated waste
waters from dyeing and bleaching plants by biological fil-
tration with recirculation even though the pH was at times
12 to 13. Effluent recirculation at 3:1 produced 75% re-
moval of oxygen demand. Biological filtration with recir-
culation at 1:1 of the effluent was used for treating cot-
ton curing wastes, and when operated with alternating double
filtration a BOD reduction of 95.7$ was achieved (5289).
Alternating double filtration (4O39) was effective in im-
proving the quality of effluent from synthetic fiber manu-
facturing waste treatment plants.
With the technique of intermittent filters following per-
colating filtration, Bogren (384) was able to reduce the
BOD by 92.7$ and the suspended solid by 99.6$ in textile
waste waters He also removed 75$ of the BOD from cotton
finishing wastes, in agreement with Dahlem (854). Removal
of 58$ of the BOD and 45$ of the color was achieved by
Souther and Alspaugh (4114) on a high-rate biological trick-
ling filter treating a composite dye house, bleaching, fin-
ishing, mercerizing, and color shop wastes mixed with do-
mestic sewage. Combined treatment using high-rate filtra-
tion and activated sludge (4114) resulted in 93$ BOD removal
and 50$ color removal which was verified later by Souther
and Alspaugh (4115, 4116) . The biological filtration after
sedimentation removed 73$ of the BOD at a pH of 8.5, and
58$ at a pH of 1O.5 (4115).
A treatment plant was described by Brown (491) which removed
82.6$ of the BOD from a mixed waste containing 6O$ textile
waste waters and 4O$ domestic sewage via processes involving
pH adjustment, primary sedimentation, and two-stage biolog-
ical filtration, followed by humus tanks. Walter (4592)
studied the combined treatment of sewage and textile waste
waters and the effect of high pH on the performance of rough-
ing percolating filters. Loadings of 3,OOO to 6.OOO Ib BOD/
ac-f/day (69 to 138 Ib BOD/1,OOO ft3/day) were removed
about 6O$ by biological filtration. The filter reduced the
pH of the waste and, since little reduction in BOD was
achieved by settling plus the necessity for second stage
filtration, an intermediate clarifier was not justified
(4592). Nutritional deficiencies and other biotic factors
were investigated by Oldaker (3227), and it was determined
that the optimum BOD removal occurred when 1 Ib of phos-
phorus was added for every 98 Ib of BOD removed. An 8$ in-
crease in BOD reduction was obtained by reducing the pH
from 11.5 - 12 to 7.5. Kilgore and Sawyer (2477) supported
336
-------
these findings in separate studies of nitrogen and phos-
phorus requirements.
COMPARISON OF OTHER METHODS OF TREATMENT
Milligan (3O1O) discussed the laws of the Pennsylvania Sani-
tary Water Board concerning the effect of textile industry
effluents and treatment on biological filters with recircu-
lation of effluent, and final chlorination was preferred to
chemical coagulation. After comparing performance of chemi-
cal coagulation, biological filtration, and lagooning, Williams
and Hutto (4738) constructed pilot plant aerated lagoons to
treat waste waters from synthetic fiber and other textile
mills which produced reductions in BOD of 75 to 8O$. Souther
and Alspaugh (4116, 4117) recommended treatment similar to
that of other investigators and pointed out advantages and
disadvantages of processes such as lagooning, chemical treat-
ment, coagulation, biological treatment by both percolating
filters and activated sludge. Additional flexibility was
designed into plants by Grey (1581), which allowed chemical
treatment for neutralization and disinfection, as well as
the addition of the influent waste waters at several points
throughout the treatment plant.
Gotaas (1521) discussed textile wastes containing unstable
organic matter and toxic constituents and indicated that the
activated-sludge process removed color better, but was less
resistant to shock loads and to toxic substances than trick-
ling filters. Horton and Baity (2037) used the activated-
sludge process and biological filtration for treatment of
textile waste and sewage mixtures with the usual process
limitations. Nicholas (3183) successfully operated a Kessener
brush aeration unit which provided satisfactory treatment
on a combined waste flow containing waste from the manufacture
of tweed and woolen garments. This waste did not require
preliminary treatment and this aeration unit was chosen in
preference to chemical coagulation and to high-rate biological
filtration.
According to Oldroyd (3229), alternating double filtration
on a laboratory scale was effective on the treatment of tex-
tile wastes when the primary filter was loaded at 40O gal./
yd3/day, while the diluted waste was applied to the secondary
filter operating with recirculation at 500 gal./yd3/day.
This system was compared to two-stage high-rate filtration
without recirculation. However, both systems ponded. He
concluded that the slightly better results of alternating
double filtration were not worth the added capital invest-
ment and operational requirements over two-stage operation
with recirculation at a constant rate.
337
-------
POST-TREATMENT AND EFFLUENT QUALITY
Meyer (2987) recommended the use of spray irrigation of flax
retting waste waters in Germany even though biological fil-
tration had been used successfully for years under lightly
loaded conditions. Fleming (1294) noted that woolen mill
wastes were treated with combined domestic sewage on percolat-
ing filters which produced a sludge used as a fertilizer.
McCarthy (2901) noted common use of backwashing filters and
high dilution of effluent to reach an acceptable performance
level. Reductions in hydroxyl alkalinity, BOD, total settle-
able solids, color, and temperature of the waste were ac-
complished successfully by lagooning the textile waste waters
(4117) .
SPECIAL OPERATIONAL PROBLEMS
A conventional activated-sludge plant treating a combination
of sewage and textile waste waters in North Carolina became
overloaded and was operated only as a chemical precipitation
plant (839) . A volume of 11 mgd, 4O$ of which was textile
waste, was treated. Before designing a new plant, pilot plant
studies were made and it was concluded (839) that an effluent
BOD of 50 mg/1 could be obtained by a two-stage biological
treatment if the pH were controlled to less than 9.5. A
harmful characteristic of textile waste was reported by Ingols
(2219) to be the large quantity of detergents imparted to
the sewers from washings which interfered with sedimentation
and resulted in overloading percolating filters.
Horton and Baity (2O37) discussed problems of treating com-
bined textile and domestic sewage due to the changes in char-
acter of the sewage as it reached the treatment plant. Dif-
ficulties in treatment of waste waters from cotton kiering,
cotton bleaching, and pulping of rag and rope by two-stage
high-rate percolating filters were attributed by Kilgore and
Sawyer (2477) to nutritional deficiencies. The requirements
for nitrogen and phosphorus of the different wastes varied
from 4.7 to 7.4 Ib of nitrogen and from O.36 to 1.25 Ib of
phosphorus per 1OO Ib of BOD removed, with similar needs
for activated sludge and biological filtration. Oldaker
(3227) generally agreed with the nutritional demand.
Souther and Alspaugh (4117) stated that the volume of waste
water which could be treated biologically without preliminary
treatment was limited by the hydroxyl alkalinity rather than
by the pH value. According to Petru (334O), acclimatization
of filters which were to treat concentrated waste waters,
such as that from flax retting, is an operational require-
ment. The maturing of a filter is shown by a steady decrease
of BOD and an increase of pH value.
338
-------
Critique
Waste waters arising from the production of textile goods
have been reported in many texts, e.g., "The Treatment of
Industrial Waste," by Besselxevre (3O8), in conjunction with
biological trickling filters. Textile wastes have been gen-
erally considered to be highly alkaline, contain high dis-
solved organics and suspended solids, be of an elevated tem-
perature and colored in various degrees. Several different
types of chemical, physical, and biological waste treatment
have been employed. Since the development of synthetic
fibers, which have wastes quite similar to those from the
chemical industry, most treatment schemes have been con-
sidered.
Wastes arising from the production of cotton, wool, silk,
synthetic fibers and other materials have been treated by
biological filtration with various degrees of success. Pre-
treatment to establish the proper pH range and to remove
suspended solids was frequently emphasized. Additional
components in the textile wastes, such as detergents and
toxic dyes, complicated the problem of waste treatment.
It would appear, based on this review, that adequate treat-
ment of textile wastes can be achieved by biological fil-
tration in combination with other treatment processes. These
processes would involve pretreatment, not only in the form
of pH control and suspended solids removal, but also in cool-
ing and equalization. Biological filtration operated in
series with other processes, such as the activated-sludge
process, has been shown to be successful for the treatment
of these difficult wastes. Low capital investments were
spent for lagoon waste treatment, which was often used in-
stead of biological filtration.
The literature reflected excellent communication among in-
vestigators. This may have been due to the formation of
trade associations which explored methods of treatment and
sponsored research and development to characterize waste
waters and to propose alternative methods of treatment. Waste
waters from similar production facilities had effluent char-
acteristics that were correspondingly similar. Performance
results agreed well with independent investigations on simi-
lar wastes. Trends similar to those observed for tannery
waste treatment were noted in this area, e.g., for several
years, reports were submitted which dealt with the difficulties
of treating textile waste combined with domestic sewage.
With the construction and operation of newer facilities, data
were made available resulting in further process modifications.
These modified plants were operated under several different
conditions to optimize plant efficiency. Biological fil-
tration was shown to be useful for the partial treatment of
textile wastes, provided sufficient pretreatment was in-
cluded in the design.
339
-------
SECTION XXXII
ACKNOWLEDGMENTS
The following groups and sources aided in the development
and completion of this review.
The staff of The Dow Chemical Company's Technical
Library.
Mrs. Clementine Moore, of the National Environ-
mental Research Center Library, located in the
Robert A. Taft Sanitary Engineering Building,
Cincinnati, Ohio.
Mr. Edwin D. Posey, of the Civil Engineering
Library, Purdue University, Lafayette, Indiana.
Chemical Abstracts and Water Pollution Abstracts.
Messrs. D. G. Parker, K. F. Spear, J. W. Pollack,
E. B. Albright and M. B. Ettinger, and Dr. S. L.
Daniels made contributions to the -bibliography and
to this review under the guidance and leadership of
Dr. C. H. Thompson, who compiled the rough draft.
Dr. R. S. Karpiuk and Mrs. Isabel H. Carr reorga-
nized and edited the rough manuscript to its final
form. Mr. M. E. Arnold gave valuable assistance in
reorganizing and editing the references. The sec-
tions on "Ecology" and "Bacteria}, and Viral Removal1
were technically edited by M. B. Ettinger, J. J.
McDade and J. N. Tadman. Ms. Karen Anderson, Helen
Blank, Loretta Demers, Virgie Fisher, Margo Hardy,
Neva Harrison, Vera Hart, Kim Short, Jean Shuler,
Sharon Tomczak, Ann Wisniewski, and Carolyn Wyse
provided the massive secretarial support under con-
siderable pressure of an approaching deadline.
341
«U.S. GOVERNMENT PRINTING OFFICE: 1972 WU-U87/311 1-3
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
1. Report No.
2.
3. Accession No.
w
4. Title A LITERATURE SEARCH AND CRITICAL ANALYSIS
OF BIOLOGICAL TRICKLING FILTER STUDIES.
VOL. I AND VOL. II
7. Author(s)
9. Organization The. Dow Chemical Company
Functional Products and Systems Department
Midland, Michigan 4864O
12. Sponsoring Organization EPA, Water Quality Office
15. Supplementary Notes
5. Report Date
6.
8. Performing Organization
Report No.
10. Project No.
17O5O DDY
11. Contract/Grant No.
14-12-474
13. Type of Report and
Period Covered
16. Abstract
A compilation, review and critique of the literature on biological trick-
ling filter studies and related pollution abatement processes have been
made. Volume I, the literature review and critical analysis, is divided
into: (a) Introduction, Definitions, History and Background Theory of the
Trickling Filter Process; (b) Plant Design, Materials of Construction, Opera-
tion, Maintenance and Performance? (c) Trickling Filter Research and Develop-
ment Approaches, Ecology, and Patents, and (d) Applications of Trickling Filter
to Specific Industrial Wastes. Volume II is the bibliography of 5, 565 refer-
ences. Based on the review, several general conclusions were drawn. There is no
well-defined theory of design and operation. Much published work was redundant
and European efforts were not readily accepted in the United States, and vice
versa. The literature reflects cycles of interest in trickling filters. The
value of much of the early work was ignored. Solutions to complex pollution
problems will be made by industry strongly supported by local, state, and Fed-
eral governments. The filter will be used in high efficiency, modern waste-
water treatment plants. The process is not applicable to all pollution prob-
lems, but its shock survival capabilities and rapid flow-through time are defi-
nite advantages which cannot be overlooked in any design of a waste treatment
facility.
17a. Descriptors* Re view of trickling filters, ^Bibliography, *Biological filters
*Percolating filters, *Sprinkling filters, *Biofilters, *Theory, *Con-
struction, *Operation and maintenance, ^Performance, ^Research and devel-
opment, *Ecology, *Patents, *Industrial wastes.
776. Identifiers
*Aerobic treatment, *Biological treatment, *Filtering systems, *Pollution
abatement, *Treatment facilities. Application of trickling filters to
brewery and distillery, chemical, gas and coke plant, food processing,
institutional and military, laundry and cleaning, meat and poultry, metal
working and production, milk processing, pharmaceutical and fermentation,
pulp and paper, radioactive, tannery and textile wastes.
17c. TO WRR Field & Group iQp Q 5D
18. Availability
19. Security Class.
(Report)
20. Security Class.
(Page)
21. No. of
Pages
22. Price
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON, D. C. 20240
Abstractor Robert S. Karpiuk \institution The Dow Chemical Company
WRSIC 102 (REV. JUNE 1971)
GPO 9l3.26t
------- |