<
>£ £
||2
111
%£$
Q « OC
United States
Environmental Protection
Agency
Water Engineering
Research Laboratory
Cincinnati OH 45268
EPA/600/9-86/015b
July 1986
Research and Development
r-i
Proceedings
Tenth United States/
Japan Conference on
Sewage Treatment
and
North Atlantic Treaty
Organization/Committee
on the Challenges of
Modern Society
Conference on Sewage
Treatment Technology
Volume I. Part B.
United States Papers
-------
EPA/600/9-86/015b
July 1986
PROCEEDINGS
TENTH UNITED STATES/JAPAN CONFERENCE
ON SEWAGE TREATMENT TECHNOLOGY
v OCTOBER 17-18, 1985
•N/v
^
^ AND
S NORTH ATLANTIC TREATY ORGANIZATION/COMMITTEE ON THE
CHALLENGES OF MODERN SOCIETY (NATO/CCMS) CONFERENCE
ON SEWAGE TREATMENT TECHNOLOGY
r
\J
> OCTOBER 15-16, 1985
^
CINCINNATI, OHIO
VOLUME I.
PART B. UNITED STATES PAPERS
U S Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevacd, 12th Floor
Chicago, IL 60604-3590
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
CINCINNATI, OHIO 45268
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
-------
FOREWORD
The maintenance of clean water supplies and the
management of municipal and industrial wastes are vital
elements in the protection of the environment.
The participants in the Japan-United States-North
Atlantic Treaty Organization/Committee on the Challenges
of Modern Society (NATO/CCMS) Conferences on Sewage Treat-
ment Technology completed their conferences in Cincinnati,
Ohio, in October 1985. Scientists and engineers of the
participating countries were given the opportunity to study
and compare the latest practices and developments in Canada,
Italy, Japan, The Netherlands, Norway, the United Kingdom and
the United States. The proceedings of the conferences comprise
a useful body of knowledge on sewage treatment which will be
available not only to Japan and the NATO/CCMS countries but
also to all nations of the world who desire it.
Lee M. Thomas
Administrator
Washington, D.C.
m
-------
CONTENTS
Foreword . . . .' iii
Japanese Delegation vi
United States Delegation vii
North Atlantic Treatment Organization/Committee on the
Challenges of Modern Society (NATO/CCMS) Delegation viii
Joint Communique 1
Volume I.
Part A. Japanese Papers 3
Volume I.
Part B. United States Papers 367
Volume II.
North Atlantic Treaty Organization/Committee on the
Challenges of Modern Society (NATO/CCMS) Papers 633
-------
JAPANESE DELEGATION
DR. TAKESHI KUBO
Head of Japanese Delegation,
Counselor, Japan Sewage Works Agency
TOKUJI ANNAKA
. Chief, Water Quality Section
Water Quality Control Division
Public Works Research Institute
Ministry of Construction
DR. KEN MURAKAMI
Deputy Director, Research and
Technology Development Division
Japan Sewage Works Agency
DR. KAZUHIRO TANAKA
Chief Researcher, Research and
Technology Development Division
Japan Sewage Works Agency
KENICHI OSAKO
Chief, Eastern Management Office
Sewage Works Bureau
Tokyo Metropolitan Government
SAKUJI YOSHIDA
Chief, Facility Section
Construction Division
Sewage Works Bureau
City of Yokohama
YUKIO HIRAYAMA
Director, Planning Division
Sewage Works Bureau
City of Fukuoka
MASAHIRO TAKAHASHI
Extraordinary Participant,
Researcher, Sewerage Section
Water Quality Control Division
Public Works Research Institute
Ministry of Construction
TAKASHI KIMATA
Extraordinary Participant,
Researcher, Research and Technology
Section, Research and Technology Division
Japan Sewage Works Agency
-------
UNITED STATES DELEGATION
JOHN J. CONVERY
General Chairman of Conference and
Head of Cincinnati U.S. Delegation
Director, Wastewater Research Division
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnai, OH 45268
DOLLOFF F. BISHOP
Co-Chairman of Conference
Chief, Technology Assessment Branch
Wastewater Research Division
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
FRANCIS T. MAYO
Director,
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
LOUIS W. LEFKE
Deputy Director,
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
DR. JAMES A. HEIDMAN
Environmental Engineer
Innovative & Alternative Technology Staff
Systems & Engineering Evaluation Br., WRD
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
HENRY H. TABAK
Research Chemist,
Toxic Research & Analytical Support Staff
Technology Assessment Branch, WRD
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
DR. ALBERT D. VENOSA
Microbiologist, Ultimate Disposal Staff
Systems & Engineering Evaluation Br., WRD
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
ARTHUR H. BENEDICT, Ph.D.
Brown & Caldwell Consulting Engrs
P.O. Box 8045
Walnut Creek, CA 94546-1220
DR. WILLIAM C. BOYLE
Dept. of Civil Engineering
& Environmental Engineering
University of Wisconsin
3230 Engineering Building
Madison, Wisconsin 55706
DR. MICHAEL CARSIOTIS
Dept. of Microbiology
& Molecular Genetics
University of Cincinnati
College of Medicine
231 Bethesda Avenue
Cincinnati, OH 45267
DR. CLEMENT FURLONG
Dept. of Medical Genetics, SK50
University of Washington
Seattle, WA 98195
GILBERT B. MORRILL, P.E.
McCall, Elingson, Morrill, Inc.
Consulting Engineers
1721 High Street
Denver, CO 80218
DR. GEORGE PIERCE
Battelle-Columbus Laboratories
505 King Avenue
Columbus, OH 43201
DR. JOHN N. REEVE
The Ohio State University
Dept. of Microbiology
484 West 12th Avenue
Columbus, OH 43210-1292
DR. H. DAVID STENSEL
Dept. of Civil Engrg, FX-10
University of Washington
Seattle, WA 98195
-------
NORTH ATLANTIC TREATY ORGANIZATION/COMMITTEE ON THE CHALLENGES
OF MODERN SOCIETY (NATO/CCMS) DELEGATION
DR. J. DUANE SALLOUM
Chairman of NATO/CCMS Committee
Director, Technical Services Branch
Environmental Protection Service
- Ottawa, Canada K1A 1C8
DR. B. E. JANK
A/Director,
Wastewater Technology Centre
Canada Centre for Inland Waters
P.O. Box 5050,
Burlington, Ontario L7R 4A6
Canada
DR. ROLF C. CLAYTON
Director,
Process Engineering
Water Research Laboratory
Elder Way, Stevenage, Herts, SGI 1HT,
England
DR. IR. WILHELMUS H. RULKENS
Department of Environmental Technology
Division of Technology for Society
MT/TNO
P.O. Box 342, 7300 AH Apeldoorn
The Netherlands
DR.ING. BJ0RN RUSTEN
Aquateam, Norwegian Water Technology Centre A/S
P.O. Box 6593
Rodeldkka, N-0501 Oslo 5,
Norway
DR. MARIO SANT.ORI
Institute di Ricerca sulle Acque
Consiglio Nazionale delle Ricerche
Rome, Italy 00198
vm
-------
DELEGATES TO THE NATO/CCMS CONFERENCE AND THE TENTH UNITED STATES/
JAPAN CONFERENCE ON SEWAGE TREATMENT TECHNOLOGY
ANDREW W BREIDENBACH ENVIRONMENTAL RESEARCH CENTER, CINCINNATI, OHIO
-------
DR. JOHN H. SKINNER, DIRECTOR, OFFICE OF ENVIRONMENTAL ENGINEERING
AND TECHNOLOGY, MR. FRANCIS T. MAYO, DIRECTOR, WATER ENGINEERING
RESEARCH LABORATORY, U.S. EPA AND DR. TAKESHI KUBO, HEAD OF
JAPANESE DELEGATION'AND COUNSELOR, JAPAN SEWAGE WORKS AGENCY
MR. DOLLOFF F. BISHOP, U.S. DELEGATE AND DR. ROLF C. CLAYTON,
NATO/CCMS DELEGATE FROM THE UNITED KINGDOM
-------
VIEW OF CAPTOR WASTEWATER TREATMENT PROCESS
TEST AND EVALUATION FACILITY, CINCINNATI, OHIO
VISIT TO THE MULTIPLE DIGESTION PROJECT,
TEST AND EVALUATION FACILITY, CINCINNATI, OHIO
XI
-------
JOINT COMMUNIQUE
TENTH UNITED STATES/JAPAN CONFERENCE
ON SEWAGE TREATMENT TECHNOLOGY
Cincinnati, Ohio
October 18, 1985
1. The Tenth United States/Japan Conference on Sewage Treatment
Technology was held in Cincinnati, Ohio, from October 17 to 18, 1985.
2. The Japanese delegation headed by Dr. Takeshi Kubo, Counselor,
Japan Sewage Works Agency,was composed of two representatives from the
Ministry of Construction, three representatives from the Japan Sewage
Works Agency and one each from the local governments of Tokyo, Yokohama
and Fukuoka.
3. Mr. John J. Convery, Director, Wastewater Research Division, Water
Engineering Research Laboratory, U.S. Environmental Protection Agency,
was head of the U.S. delegation,which consisted of seven representatives
of the federal government, five academia representatives and three repre-
sentatives from consulting engineering firms and scientific laboratories.
4. The chairmanship of the Conference was shared by Mr. John J. Convery
and Dr. Takeshi Kubo.
5. During the Conference, papers relating to the joint research projects
on sludge treatment and disposal*including combustion, oxidation and compost-
ing.were presented by both sides. Data and findings on the joint research
projects were mutually useful and provided increasing insights into the
nature of the problems and potential solutions for each country. A decision
was made to expand the scope of the joint research projects to include
anaerobic treatment of wastewater.
6. Principal topics of the Conference were bioengineering applications
in wastewater treatment as well as sludge management and disposal,
aeration practice, wastewater reirse, odor control, small flow sewerage
system, nutrient control and innovative biological treatment processes.
The discussions which followed the presentations were also useful
to both countries.
xn
-------
7. Field visits in Lawrence, Marlborough and Hartford, Connecticut;
Chicago, Illinois; Madison and Milwaukee, Wisconsin; and Sacramento, Cali-
fornia; are planned to inspect wastewater treatment facilities in these
areas.
8. Recent engineer exchanges between the two countries included a two-
w£ek visit in 1985 to Japan by Mr. James F. Kreissl, Wastewater Research
Division, Water Engineering Research Laboratory, U.S. Environmental
Protection Agency,and a fourteen-month visit to the United States by
Dr. Kazuhiro Tanaka, Japan Sewage Works Agency, in 1984 to 1985. Mr.
Takashi Kimata of the Japan Sewage Works Agency is now staying at the
above U.S. EPA Cincinnati Research Laboratory. Both parties agreed to
continue the engineer exchange program.
9. It was proposed by the Japanese side that the Eleventh Conference
be held in Tokyo, Japan, about September 1987, and the future Confer-
ences in the United States would be held in Cincinnati, Ohio and also in
Washington, D.C. as were the past Conferences.
10. A proceedings of the Conference will be printed in English and Japanese.
xiii
-------
UNITED STATES PAPERS
STRUCTURE OF METHANOGEN GENES 369
John N. Reeve, Department of Microbiology, Ohio State
University, Columbus, Ohio
THE START OF NITRIFIER GENETICS 385
Michael Carsiotis and Sunil Khanna, Department of Microbiology
and Molecular Genetics, University of Cincinnati, Cincinnati,
Ohio
REMOVAL OF PHOSPHATE AND OTHER SMALL MOLECULES FROM WASTE STREAMS
WITH BINDING PROTEINS IN CYCLING COLUMN ADSORBERS 397
Clement E. Furlong, Joseph A. Sundstrom, and R.J. Richter,
Departments of Genetics and Medicine, Division of Medical
Genetics, Center for Inherited Diseases, University of
Washington, Seattle, Washington; and John Yin and Harvey
W. Blanch, Department of Chemical Engineering, University of
California, Berkeley, California
DEGRADATION OF CHLORINATED BENZOATES UNDER A VARIETY OF ANAEROBIC
ENRICHMENT CONDITIONS 413
Barbara R. Sharak-Genthner, Jayne B. Robinson, and George E.
Pierce, Batelle--Columbus Division, Columbus, Ohio
ASSESSMENT OF BIOAUGMENTATION TECHNOLOGY AND EVALUATION STUDIES
ON BIOAUGMENTATION PRODUCTS 431
Henry H. Tabak, Wastewater Research Division, Water Engineering
Research Laborabory, Office of Research and Development, U.S.
Environmental Protection Agency, Cincinnati, Ohio
MUNICIPAL SLUDGE OXIDATION WITH THE VERTICAL TUBE REACTOR 501
Gilbert B. Merrill, McCall-Ellingson & Merrill, Inc.,
Denver, Colorado
MUNICIPAL WASTEWATER TREATMENT USING THE CAPTOR PROCESS 517
James A. Heidman, Water Engineering Research Laboratory,
U.S. Environmental Protection Agency, Cincinnati, Ohio
BIOLOGICAL AERATED FILTER PERFORMANCE 535
H.D. Stensel, University of Washington, Seattle, Washington;
and R.C. Brenner, U.S. Environmental Protection Agency,
Cincinnati, Ohio
367
-------
FINE PORE AERATION PRACTICE 559
William C. Boyle, Department of Civil and Environmental
Engineering, University of Wisconsin, Madison, Wisconsin
COMPOSTING PRACTICE IN THE UNITED STATES TODAY 603
Arthur H. Benedict, Brown and Caldwell, Pleasant Hill, California
368
-------
STRUCTURE OF METHANOGEN GENES
by
John N. Reeve
Department of Microbiology
Ohio State University
Columbus, Ohio 43210
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
369
-------
STRUCTURE OF METHANOGEN GENES
by: John N. Reeve
Department of Microbiology
Ohio State University
Columbus, OH 43210
ABSTRACT
Methanogens are members of the third biological kingdom known as the
Archaebacteriales. Their sizes and cellular organizations are typicially
prokaryotic whereas biochemical and molecular biological analyses have
demonstrated the presence of eukaryotic properties. Our studies have two
goals; to establish the structure and mechanisms of expression of methanogen
genes and to apply this knowledge to the manipulation of genes which encode
enzymes directly involved in methane biogenesis. We have cloned and sequenced
several methanogen genes which, when expressed in Escherichia coli, complement
auxotrophic mutations of this laboratory prokaryote. Analyses of the DNA
sequences indicate that methanogen genes are organized into multigene
transcriptional units and that translation of mRNAs employs ribosome binding
sequences. These are properties typical of eubacteria; in contrast, there is
no evidence for eubacterial promoter-like sequences and comparisons of
methanogen and other archaebacterial sequences indicate that archaebacterial
promoters resemble the sequence used in both yeast and in the fruit-fly
(Drosophila) to direct the transcription of heat-shock genes. An analysis of
the structure and presumed regulatory signals in sequenced methanogen genes is
presented in this paper.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
370
-------
INTRODUCTION
Methane biogenesis ranks with processes such as photosynthesis and
nitrogen fixation as a major factor in the global cycling of biochemicals.
Methane generating micro-organisms (methanogens) play a vital role in the
world's environment by catalyzing the final stage in the decomposition of
waste biomass to methane. As the opportunities to use methane as a fuel
increase it is to be expected that methanogens may also soon play a much
larger role in the world's economy. The study of methanogens themselves has
been impeded by their extreme sensitivity to oxygen; they require redox
potentials below -330mV for growth and only recently have techniques been
developed by which methanogens can easily be cultivated as pure cultures in
the laboratory (1, 2). The advent of these techniques has facilitated a rapid
expansion in the study of the biology of methanogens and the biochemistry of
methane biogenesis. Results obtained to date indicate that methanogens form
an extremely diverse group (1, 3, 4 are reviews). They span the full range of
prokaryotic morphological types and have DNAs with base contents ranging from
27.5 to 52% G+C. It has, in fact, been proposed that methanogens are not true
prokaryotes but are representatives of a third biological kingdom called the
Archaebacteriales (1, 5). This kingdom, which includes extreme halophilic and
acido-thermophilic micro-organisms in addition to the methanogens, was
proposed because its members have several properties radically different from
both prokaryotic and eukaryotic species. Archaebacteria have unique
structural subunits in their cell envelopes and lipids (6). Their RNA and
protein synthesizing systems are resistant to most of the antibiotics which
inhibit these processes in prokaryotes and eukaryotes (2, 7, 8). Their
DNA-dependent RNA-polymerases are very different from those of classical
prokaryotes (9) and comparative analyses of the sequences of tRNAs and rRNAs
have provided convincing evidence that whereas the different archaebacterial
types are related to each other they are only very distantly related, in
evolutionary terms, to current prokaryotic and eukaryotic species (1). It has
recently been shown that some archaebacterial tRNAs contain intervening
sequences, introns, a property generally associated with eukaryotic but not
prokaryotic species (10). Our studies have focused on cloning and analysis of
polypeptide encoding genes. We have been able to clone fragments of DNA, from
several different methanogens, which function in E. coli to complement
auxotrophic mutations in this eubacterial species~(ll-14). This procedure has
allowed us to identify and subsequently obtain the nucleotide sequences of the
methanogen genes which function in £. coli. The accumulated data has provided
the first description of the organization of protein encoding genes within the
genomes of methanogenic archaebacteria. This information is presented in
detail.
371
-------
RESULTS AND DISCUSSION
Genes cloned from methanogens. Table I contains a list of cloned methanogen
genes which function in E_. coli. Expression of these genes results in
complementation of the listed auxotrophic mutations. Additional genes have
been cloned whose functions are unknown but which can be identified as open
reading frames (ORF's; 11-13). Methanogen genes cloned by other research
groups encode sub-units of the enzymes methyl-coenzyme-M methyl-reductase
and DNA-dependent RNA-polymerase (15, 16) and encode tRNAs and rRNAs (17, 18).
TABLE 1. CLONED METHANOGEN GENES WHICH COMPLEMENT AUXOTROPHIC MUTATIONS IN
E. COLI
Mutation Methanogen Size of encoded Reference
Complemented* DNA Polypeptide (Kd)
hisAT
hisA
argG
argG
argG*
proC
purE]
pur_E2
purE]
PurE2
* Cloned
T>
M
M
M
M
M
M
M
M
M
M
methanogen
. voltae
. vannielii
. voltae
. vannielii
. barkeri
. smithii
. smithii
. smithii
. thermoautotrophicum
. thermoautotrophicum
DNAs complement mutations
26
26
55
51
51
28
37
37
36
36
in the
11.
11.
19.
Unpublished result
14.
12.
12.
12.
13.
13.
listed E. coli genes.
Salmonella typhimurium (11).
* The tl. barkeri DNA also complements mutations in the argA locus of Bacillus
subtilis (14).
Methanogen gene organization. Methanogens are archaebacteria (1) and our
current experiments are aimed at determining whether the structure and
expression of methanogen genes follow prokaryotic or eukaryotic principles or
372
-------
whether archaebacteria have genetic arrangements different from both
eukaryotes and prokaryotes. The following paragraphs address molecular
biological parameters which are either typically prokaryotic or typically
eukaryotic.
a) Genetic code and introns. The demonstration that cloned methanogen genes
can direct the synthesis of functional enzymes in E_. coli (Table 1) strongly
implied that these genes employed the universal genetic code and did not
contain introns. DNA sequencing has confirmed this prediction (11-13). It
should be noted however that there have been many unsuccessful attempts to
clone methanogen genes which would complement mutations in additional E_. coli
genes. These unsuccessful attempts are, of course, not listed in Table 1 and
it should be recognized that the procedure used to obtain the methanogen genes
listed in Table 1, i.e. functional expression in E_. coli, presumably would
select against the isolation of methanogen genes employing an unusual genetic
code or which contained introns. Analysis of DNA sequences surrounding cloned
genes, sequences whose functional expression was not demanded in E_. coli,
might allow the recognition of introns in polypeptide encoding sequences. One
such candidate sequence has already been identified although the evidence for
this sequence being an intron is currently only circumstantial (11).
b) Operons and ribosome binding sites. Prokaryotic genes are frequently
contained in polycistronic transcriptional units or operons. In contrast,
eukaryotic genes are not usually transcribed into polycistronic raRNAs.
Parameters expected of genes in an operon are that mutations in promoter
proximal genes are polar on expression of promoter distil genes, only short
intergenic sequences are present between genes in the same operon and ribosome
binding sequences are positioned immediately preceding the AUG initiation
codon of each polypeptide-encoding gene. These properties have been
demonstrated for two genes on a fragment of DNA cloned from M. smithii which
complements purE mutations in E. coli (12). Only one of the cloned genes is
needed for purE complementation although both genes direct the synthesis of
polypeptides in _E_. coli. Polycistronic raRNAs must contain signals to
correctly position ribosomes for initiation of translation at the start of
genes embedded within the polycistronic mRNA. Ribosome binding sites contain
sequences which are complementary to a sequence found near the 3' terminus of
the 16SrRNA molecule (20). Table 2 shows that ribosome binding sites precede
the sequenced methanogen genes.
The data from DNA sequencing which indicate that raethanogen genes are
organized into prokaryotic-like transcriptional units and encode
prokaryotic-like mRNAs are supported by a study of the structure of M.
vannielii mRNAs (21). .M. vannielii mRNAs are unstable, have only short 3'
polyadenylation tracts and do not contain 5* capped nucleotides; all features
of prokaryotic but not eukaryotic mRNAs..
373
-------
TABLE 2. RIBOSOME BINDING SEQUENCES PRECEDING METHANOGEN GENES
Methanogen
Methanobrevibacter smlthii
Methanobrevibacter smithii
Methanobrevibacter smithii
Methanobrevibacter smithii
Gene1
Sequence*
** **** *
AGAAGGTATTTTAAAATG
* ***
ATAAGGGATAATTATG
* ** ****
AAGATGTGAAATATATG
**** *
GTGAGGACAAATAATATTTTATG
Methanobacterium thermoautotrophicum ORF-PurEf
Methanobacterium thermoautotrophicum ORF-Bt
Methanobacterium thermoautotrophicum ORF-Ct
TAAAGGTGAATCTCCAGATG
* ******
TGATGGTGATTGAATG
** *****
AGAAGGTGTACTGATG
Methanococcus vannielii
Methanococcus vannielii
ORF-HisA
ORF-76
******
AAAAGGTGAATACAATG
******
TTCTGGTGATTCAATG
Methanococcus voltae
ORF-HisA
* *****
AGATGGTGAAACTGATG
The sequence data and gene designations are taken from publications 11—13.
A sequence exactly complementary to 3' end of the 16SrRNA of M. smithii
and M. thermoautotrophicum would be 5'AGGAGGTGAT. The 16SrRM of
methanococci lacks the 3' terminal U and therefore the complementary
sequence would be 5'GGAGGTGAT (1). The asterisks (*) indicate bases in
the methanogen sequences which are complementary to the 16SrRNA. The
initiation ATG codon of each polypeptide encoding gene is underlined. The
number following the base in the sequence AyGgG^gTgGyAy is the number of
times that base occurs in the location indicated in the 10 sequences
listed in Table 2. We propose this sequence as in the consensus ribosome
binding sequence in methanogens.
374
-------
c) RNY rule. Shepherd (22, 23) has proposed that polypeptlde-encoding ORFs
can be recognized by the use of RNY codons (R=purine, Y=pyrimidine, N=purine
or pyrimidine). RNY codons occur most frequently in the correct reading-frame
whereas non-utilized ORFs do not show preferential RNY codon usage. The only
reported exception to this rule, other than the highly evolved overlapping
genes of bacteriophages such as '5X174, is the archaebacterial gene of R_.
halobium which encodes bacterio-opsin (23). We have now analyzed the
available archaebacterial genes from methanogens, both those known to encode
functional polypeptides and those only identified by computer screening as
ORFs, for RNY codon usages. Table 3 shows that the majority of
methanogen-derived genes follow the RNY rule. There are two exceptions, ORF-D
of M. smithii and ORF-PurEt of M. thermoautotrophicum. ORF-D is encoded on
the DNA strand opposite to that which encodes proC in M. smithii and is
completely contained within the proC sequence. It was not considered a bona
fide gene in a previous publication (12). The RNY rule therefore seems to
provide additional support for the conclusion that ORF-D is not a polypeptide
encoding gene although this could be an erroneous interpretation of the RNY
data if ORF-D is, in fact, an overlapping gene. In contrast, there is
convincing evidence that the ORF-PurEt sequence is a polypeptide encoding
gene. The polypeptide has been identified by its synthesis in minicells of E_.
coli. An in-frame deletion within ORF-PurEt reduces the size of the encoded
polypeptide when synthesized in E_. coll and also inactivates the ability of
the polypeptide to complement mutations in purE of E_. coli (13). The
non-conformity of ORF-Pu£Et with the RNY rule is particularly surprising in
that the purE complementing gene from a different methanogen, M. smithii
(ORF-PurEs, Table 3) does follow the RNY rule and these two methanogen-derived
genes are clearly evolved from a common ancestor. The two DNA sequences are
53% homologous and the encoded polypeptides are 74% homologous if conservative
amino-acid substitutions are considered to maintain polypeptide homology (13).
It appears therefore that during evolution the divergence which has produced a
mesophilic II. smithii having a genome with 30.6% G+C and a thermophilic M.
thermoautotrophicum a genome containing 49.7% G+C has permitted the M. smithii
but not the M. having thermoautotrophicum purE gene to maintain the
preferential usage of RNY codons.
d) Codon usage. It is well established that codon usage is not random; there
is a direct correlation in the choice between synonymous codons and the
availability of isoaccepting tRNAs (24, 25). Table 4 is a comparison of the
codons used by 11. smithii, M. thermoautotrophicum, M. voltae and M. vannielii
with codons used by E. coli and S. cerevisiae.
375
-------
Table 3. FREQUENCY OF OCCURENCE OF RNY CODONS IN METHANOGEN OPEN READING
FRAMES (ORFs).
Methanogen
M.
smithii
M. thermoauto-
trophicum
M.
M.
M.
M.
M.
M.
M.
M.
M.
smithii
smithii
smithii
smithii
smithii
vannielii
vannielii
vannielii
voltae
ORFT
purEs
purEt
BS
IS
proC
D
E
I
3
hisA
hisA
Number of
Codons*
339
334
418
401
251
171
237
502
76
238
242
RNY
120
89
155
98
84
40
77
141
25
81
83
0
Stop
0
0
0
0
0
0
0
0
0
0
0
Frame^
1
RNY Stop
39
39
58
91
37
21
41
85
9
26
25
33
36
33
32
23
13
21
36
5
18
28
Reference
containing
DNA sequence
2
RNY Stop
85
94
95
94
58
56
59
121
23
58
55
37
23
52
47
25
10
31
52
9
27
20
(12)
(13)
(12)
(12)
(12)
(12)
Unpublished
(11)
(11)
(11)
(11)
RNY codons defined by Shepherd (22, 23); R=purine, Y=pyrimidine, N=purine
or pyrimidine.
T ORF designations are given in the cited references. The genetic loci
indicate that mutations in these genes of E_. coli are complemented by the
the cloned methanogen gene.
* Number of amino-acid encoding codons.
£ Frames 0, 1 and 2 begin with the A, U and G of the AUG initiation codon,
respectively. The number of termination codons, UAA, UAG and UGA in each
reading frame is listed under 'Stop1.
376
-------
TABLE 4. CODON USAGE IN E. COLI*, S. CEREVISIAE AND FOUR METHANOGENSf
Residue
and Codon
Ala GCA
GCC
GCG
GCU
* Total
Arg AGA
AGG
CGA
CGC
CGG
CGU
Total
Asn AAC
AAU
Total
Asp GAG
GAU
Total
Cys UGC
UGU
Total
Gin CAA
CAG
Total
Glu GAA
GAG
Total
Gly GGA
GCC
GGG
GGU
Total
E. coll
%
No • syno-
nym
179 22
179 22
236 28
238 28
832 10.5
3 >1
3 >1
14 3
156 33
17 4
280 59
473 6.0
210 75
69 25
279 3.5
259 55
209 45
468 5.9
51 64
29 36
80 1.0
78 23
256 77
334 4.2
454 75
148 25
602 7.6
24 4
243 42
34 6
280 48
581 7.4
S. cere-
visiae
%
No . syno-
nym
13 4
78 23
2 >1
249 73
342 10.9
113 88
4 3
0
1 >1
0
10 8
128 4.1
105 85
18 15
123 3.9
103 58
75 42
178 5.6
1 6
16 94
17 0.5
82 91
8 9
90 2.8
166 94
11 6
177 5.6
3 1
7 3
4 2
238 94
252 8.0
M.
smith 11
%
No . syno-
nym
39 46
10 12
0
36 42
85 5.1
36 68
4 7
2 4
3 6
0
8 15
53 3.2
29 25
89 75
118 7.2
36 37
60 63
96 5.7
12 45
15 55
27 1.6
38 79
10 21
48 3.0
103 89
13 11
116 7.0
59 63
7 8
7 8
20 21
93 5.7
M. thermo-
autotrophicum
%
No . syno-
nym
15 33
21 46
1 2
9 29
46 8.7
10 32
15 48
0
1 3
3 10
2 7
31 5.9
16 84
3 16
19 3.6
11 35
20 65
31 5.9
6 75
2 25
8 1.5
0
4 100
4 0.8
17 43
23 57
40 7.6
11 30
6 16
10 27
10 27
37 7.0
M. voltae
No . syno-
nym
13 54
0
1 4
10 12
24 4.7
10 62
2 13
1 6
0
0
3 19
16 3.1
6 13
40 87
46 9.1
13 32
28 68
41 8.1
2 29
5 71
7 1.3
15 88
2 12
17 3.3
26 72
10 28
36 7.1
7 18
7 18
4 10
21 54
39 7.7
M. van-
nielli
%
No . syno-
nym
30 77
3 8
1 2
5 13
39 5.3
8 35
8 35
0
1 4
1 4
5 22
23 3.1
12 34
23 66
35 4.7
10 22
36 78
46 6.2
4 50
4 50
8 1.1
12 75
4 25
16 2.1
53 96
2 4
55 7.4
32 52
4 6
7 13
18 29
61 8.2
377
-------
TABLE 4. cont
His CAC
CAU
Total
Ilu AUA
AUC
AUU
Total
Leu CUA
cue
CUG
CUU
UUA
UUG
Total
Lys AAA
AAG
Total
Met AUG
Total
Phe UUC
UUU
Total
Pro CCA
CCC
CCG
ecu
Total
Ser AGC
AGU
UCA
UCC
UCG
UCU
Total
E. coll
98 69
45 31
143 1.8
4 >1
321 67
154 32
479 6.1
14 2
67 9
505 67
60 8
40 5
64 9
750 9.5
331 73
123 27
454 5.7
207 100
207 2.6
167 66
84 34
251 3.2
45 14
20 6
69 69
11 11
316 4.0
86 21
27 7
23 6
118 29
43 11
106 26
403 5.1
S. cere-
vis iae
56 75
19 25
75 2.4
3 2
83 47
90 51
176 5.6
18 7
4 1
6 2
6 2
31 13
188 74
253 8.1
62 25
185 75
247 7.9
55 100
55 1.7
80 77
24 23
104 3.3
95 75
4 3
11 8
21 16
131 4.2
2 1
8 3
12 5
96 40
1 >1
119 50
238 7.6
M.
smithii
11 29
27 71
38 2.3
71 40
23 13
84 47
178 10.9
14 10
7 5
13 9
36 26
52 37
17 13
139 8.5
152 94
10 6
162 9.9
27 100
27 1.6
16 29
39 71
55 3.3
24 38
6 10
9 14
24 38
63 3.8
14 15
22 23
28 29
10 10
0
22 23
96 5.9
M. thermo-
autotrophicum
3 37
5 63
8 1.5
37 82
5 11
3 7
45 8.5
3 6
23 50
10 22
9 19
1 2
0
46 8.7
8 28
21 72
29 5.5
20 100
20 3.8
10 67
5 33
15 2.8
3 11
14 52
4 15
6 22
27 5.1
9 21
4 10
14 33
8 19
2 5
5 12
42 8.0
M. voltae
3 60
2 40
5 1.0
27 45
4 7
29 48
60 11.8
4 9
1 2
1 2
4 9
29 64
6 13
45 8.9
47 87
7 13
54 10.7
16 100
16 3.1
2 83
10 17
12 2.4
3 22
4 28
1 7
6 43
14 2.8
2 14
3 20
3 20
0
0
7 46
15 3.0
M. van-
nielii
8 53
7 47
15 2.0
25 34
10 13
40 53
75 10.1
6 8
4 5
1 4
27 37
29 40
6 8
73 9.8
63 87
9 13
72 9.7
19 100
19 2.6
21 100
0
21 2.8
9 41
6 27
0
7 32
22 3.0
2 5
9 23
20 51
1 2.5
1 2.5
6 15
39 5.3
378
-------
TABLE 4. cont.
Thr ACA
ACC
ACG
ACU
Total
Trp UGG
Total
Tyr UAC
UAU
Total
Vol GUA
GUC
GUG
GUU
Total
% G+C
of genome
E. coli
25 6
205 54
44 11
105 28
379 4.8
30 100
30 0.4
119 60
79 40
198 2.5
138 22
86 14
154 24
252 40
630 8.0
51
S. cere-
vislae
14 7
76 41
2 1
95 51
187 5.9
25 100
25 0.8
71 81
17 19
88 2.8
6 2
108 43
9 3
129 54
252 8.0
36
M.
smith ii
40 43
16 17
3 3
33 37
92 5.7
3 100
3 0.2
20 34
38 66
58 3.4
42 42
6 6
2 2
49 50
99 6.1
31
M. therrao-
autotrophicura
3 20
7 47
2 13
3 20
15 2.8
2 100
2 0.4
4 50
4 50
8 1.5
11 20
8 14
16 29
21 37
56 10.6
50
M. voltae
11 50
3 14
1 4
7 32
22 4.3
5 100
5 1.0
7 39
11 61
18 3.5
18 41
2 4
8 18
16 36
44 8.7
31
M . van-
nielii
15 42
3 8
7 19
11 31
36 4.9
6 100
6 0.8
8 33
16 67
24 3.2
14 25
1 2
5 9
35 64
55 7.4
31
Codon usages for approximately 50 E^. coli genes and 15 S. cerevisiae genes,
originally tabulated by Ikemura and Ozeki (25), have been used in
construction of Table 4.
^ The methanogen sequences are given in publications (11-13).
* The total number of codons encoding the same amino-acid. This number is
given as a percentage of all the amino acids under the % synonym heading.
The genome of E_. coli contains 51% G+C, almost the same G+C content as in
the genome of M. thermoautotrophicum. Similarly the genome of S_. cerevisiae
contains 36% G+C which is close to the 31% G+C content of the genomes of Jl.
smithii, M. voltae and M. vannielii. Codon preferences of the methanogens
seem to be dominated by the relative availability of A/T and G/C base pairs.
As examples, the lysine codon AAA and asparagine codon AAU are preferentially
used by M. smithii, M. voltae, and M. vannielii, species in whose genomes G/C
base pairs occur relatively infrequently, whereas the lysine codon AAG and
asparagine codon AAC are preferentially used by M. thermoautotrophicum, a
species in which G/C pairs constitute a much higher percentage of the total
genome. This simple correlation does not hold so consistently for E_. coli
and S. cerevisiae. E. coli (51% G+C) employs AAA more often than AAG whereas
379
-------
j^. cerevisiae (36%) preferentially uses AAG and AAC rather than AAA and AAU.
The somewhat unpredictable codon usages of E_. coli and S. cerevisiae can be
explained once the relative availabilities of different tRNAs in these species
is considered. Codons AAA and AAG are recognized by a single tRNA^ys i-n E.
coli whereas S^. cerevisiae has two lysine accepting tRNAs both of which
recognize AAG but only one of which functions with AAA codons (25).
Unfortunately considerations of this type cannot be applied in evaluating the
codon usages in methanogens as the number and relative amounts of isoaccepting
tRNAS have not been fully determined. It will be interesting, once these
experiments are completed, to see how well the actual amounts of isoaccepting
tRNAs agree with predictions for these values which can be made from Table 4.
In addition to conforming directly with the need to accommodate different G+C
contents, codons preferentially used by methanogens are often codons which are
almost never used by E_. coli e.g. AUA, AGA and AGG. In this respect codon
usages by methanogens are more similar to codons usages of the eucaryote, S.
cerevisiae. Very infrequent usage of codons containing the dinucleotide CG,
which is almost always underpresented in eucaryotic genomes (26), is evident
in the lists of codons used by S_. cerevisiae and by all four of the
methanogens. Selectivity in codons employed is not limited to choices between
different codons designating use of the same amino-acid but also influences
the net usage of different amino-acids. The polypeptide products of the
methanogen genes analyzed in Table 4 contain relatively more isoleucine and
lysine residues and less alanine residues than are found in the E. coli and
^. cerevisiae gene products.
Promoter structure and expression of methanogen genes in E. coli. It was
evident from the first publication describing cloning and expression of
methanogen genes (27) that methanogen derived DNAs must, contain sequences
which can function as promoters in E_. coli. Analysis of the actual DNA
sequences that are now available confirms this expectation (11-13). The
genomes of M. smithii, M. voltae and M_. vannielii are composed of almost 70%
A+T base pairs and the intergenic regions contain sequences which approach 90%
A+T. Sequences which function as promoters in E_. coli are known to be very
A/T rich. Many acceptable versions of the consensus -35(TTGACA) and
-IO(TATAAT) E_. coli promoter squences can be found positioned in locations
which should facilitate transcription of adjacent cloned methanogen genes in
E_. coli. Availability of ribosome binding sequences (Table 2) apparently
ensures that if a methanogen gene is transcribed in E_. coli, the transcripts
can be translated.
The structure of bona fide methanogen promoters remains to be determined.
A comparison of DNA sequences which precede both genes from methanogenic and
halophilic archaebacteria lead to the proposal that 5'GAANTTTCA and
5'TTTTAATATAAA might be consensus archaebacterial promoter sequences (12, 13).
An intriguing correlation is that these sequences are contained within
sequences previously identified as promoters in Drosophila (28). It has also
been shown that the Drosophila promoters are recognized in cells of the yeast
S. cerevisiae. As the DNA-dependent RNA-polymerases of archaebacteria are
apparently structually related to RNA polymerases of yeast (29) the similarity
of archaebacterial and Drosophila promoters may be more than coincidental.
Development and use of in vitro transcription systems using purified
DNA-dependent RNA-polymerases from methanogens is in progress and should help
380
-------
confirm or negate the proposed archaebacterial promoter sequence,
CONCLUSIONS
We now have the complete nucleotide sequence for more than ten different
methanogen genes including genes from four different methanogenic species
(Table 1). These genes direct the synthesis of enzymes when cloned in E_.
coli. The overall structures and organizations of methanogen genes, as
described in this report, are very similar in most respects to the structures
and organizations well established for classical prokaryotes such as E_. coli
and B. subtilis. This result is very encouraging in that it predicts that
sophisticated genetic engineering procedures, developed for use with these
thoroughly studied prokaryotes, ought to be directly applicable to
methanogens. We and other research groups (16) have therefore already begun
cloning genes which encoded enzymes that catalyse biochemical steps in
methanogenesis. Initial results of these studies demonstrate that it is
possible to clone such genes in E_. coli. The resulting availability of
"methane" genes now offers the opportunity of genetically manipulating
methanogenesis using in vitro genetic engineering techniques. Two attractive
goals of such endeavors would be increasing the range of microbial species
which can produce methane and be expanding the range of substrates and
environments which can support methanogensis. The feasibility of such
projects and practical approaches to reaching these goals are the major themes
of experiments we now have in progress.
ACKNOWLEDGEMENTS
This work was supported by contracts CR810340 from the. Environmental
Protection Agency, AC02-81ER10945 from the Department of Energy and
5083-260-0895 from the Gas Research Institute.
381
-------
REFERENCES
1. Balch, W.E., Fox, G.E., Magrum, L.J., Woese, C.R. and Wolfe, R.S.
Methanogens: re-evaluation of a unique biological group. Microbiol.
Rev. 43, 260-296, 1979.
2. Hook, L., Corder, R.E., Hamilton, P.T., Frea, J.I. and Reeve, J.N.
Development of a plating system for genetic exchange studies in
methanogens using a modified ultra-low oxygen chamber. In: Microbial
Cheraoautotrophy, Strohl, W.R. and Tuovinen, O.H. (eds), Ohio State
University Press. Columbus, Ohio, 1984. pp. 275-289.
3. Prevot, A.R., Recherches recentes sur les bacteries raethanogenes. Bull
Insttt. Pasteur 78: 217-265, 1980.
4. Wolfe, R.S. Methanogens: a surprising microbial group. Antonie van
Leeuwenhock. 45: 353-364, 1979.
5. Woese, C.R. and Fox, G.E. Phylogenetic structure of the prokaryotic
domain: the primary kingdoms. Proc. Natl. Acad. Set. U.S.A. 74:
5088-5090, 1977.
6. Kandler, 0. and Konig, H. Chemical composition of the peptidoglycan -
free walls of methanogenic bacteria. Arch. Microbiol. 168: 141-152,
1978.
7. Pecher, T. and Bb'ck, A. In vivo susceptibility of halophilic and
methanogenic organisms to protein synthesis inhibitors. FEMS Microbiol
Lett. 10: 295-297, 1981.
8. Camraarano, P., Teichner, A., Landei, P., Acca, M., Niciolaus, B., Sanz,
J.H. and Amils, R. Insensitivity of archaebacterial ribosomes to protein
synthesis inhibitors. Evolutionary aspects. The EMBO J. 4: 811-816,
1985.
9. Stetter, K.O., Winter, J. and Hartlieb, R. DMA-dependent RNA-polymerase
of the Archaebacterium Methanobacterium thermoautotrophicum. Zbl. Bakt.
Hyg. I. Abt. Orig. Cl. 201-214, 1980.
10. Kaine, B.P., Gupta, R. and Woese, C.R. Putative introns in tRNA genes of
prokaryotes. Proc. Natl. Acad. Sci. U.S.A. 80: 3307-3312, 1983.
11. Cue, D., Beckler, G.S., Reeve, J.N. and Konisky, J. Structure and
sequence divergence of two archaebacterial genes. Proc. Natl. Acad. Sci.
U.S.A. 82, In Press, 1985.
12. Hamilton, P.T. and Reeve, J.N. Structure of genes and an insertion
element in the methane producing archaebacteriura Methanobrevibacter
smithii. Mol. Gen. Genet. In Press, 1985a.
382
-------
13.
14.
15.
16.
Hamilton, P.T. and Reeve, J.N. Sequence divergence of an archaebacterial
gene cloned from a mesophilic and a thermophilic methanogen. Submitted
to J. Mol. Evol. 1985b.
Morris, C.J. and Reeve, J.N. Functional expression of an archaebacterial
gene from the methanogen Methanosarcina barkeri in Escherichia coli and
Bacillus subtilis. In; Microbial Growth on Cl Compounds (R.L. Crawford
American Society for Microbiology, Washington,
1984.
and R.S. Hanson, eds.).
B.C. 20006. pp 205-209
Bollschweiler, C., Kuhn, R. and Klein, A. Non-repetitive AT-rich
sequences are found in intergenic regions of Me thanoco ecu s voltae DNA.
The EMBO J. 4: 805-809, 1985.
Konheiser, U., Pasti, G., Bollschweiler, C. and Klein, A. Physical
mapping of genes coding for two subunits of methyl CoM reductase
component C of Methanococcus voltate. Mol. Gen. Genet. 198: 146-152,
1984.
17. Jarsch, M., Alterbuchner, J. and Bock, A. Physical organization of the
genes for ribosomal RNA in Methanococcus vannielii. Mol. Gen. Genet:
189, 41-47, 1983.
18. Jarsch, M. and Bbck, A. DNA sequence of the 16SrRNA/23SrRNA
intercistronic spacer of two rDNA operons of the archaebacterium
Methanococcus vannielii. Nucl. Acid. Res. 11: 7537-7545, 1983.
19. Wood, A.G., Redborg, A.H., Cue, D.R., Whitman, W.B. and Konisky, J.
Complementation of argG and hisA mutations of Escherichia coli by DNA
cloned from the archaebacterium Methanococcus voltae. J. Bacteriol:
156 19-29, 1983.
20. Shine, J. and Dalgarno, L. The 3'-terminal sequence of Escherichia coli
16S ribosomal RNA: complementarity to nonsense triplets and ribosome
binding sites. Proc. Natl. Acad. Sci. U.S.A. 71: 1342-1346, 1974.
21. Brown, J.W. and Reeve, J.N. Polyadenylated, noncapped RNA from the
archaebacterium Methanococcus vannielii. J. Bacteriol. 162: In Press,
1985.
22. Shepherd, J.C.W. Method to determine the reading frame of a protein from
the purine/pyrimidine genome sequence and its possible evolutionary
justification. Proc. Natl. Acad. Sci. U.S.A. 78: 1596-1600, 1981.
23. Shepherd, J.C.W. From primeval message to present-day gene. Cold Spring
Harbor Symp. Quant. Biol. 47: 1099-1108, 1983.
24. Ikemura, T. Correlation between the abundance of Escherichia coli
transfer RNAs and the occurrence of the respective codons in its protein
genes. J. Mol. Biol. 146: 1-21, 1981.
383
-------
25. Ikemura, T. and Ozeki, H. Codon usage and transfer RNA contents:
organism-specific codon-choice patterns in reference to the isoacceptor
contents. Cold Spring Harbor Symp. Quant. Biol. 47: 1087-1097, 1983.
26. Subak-Sharpe, H., Burk, R.R., Crawford, L.V., Morrison, J.M., Hay, J. and
Keir, M.H. An approach to evolutionary relationships of mammalian DNA
viruses through analysis of the pattern of nearest neighbour base
sequences. Cold Spring Harbor Symp. Quant. Biol. 31: 737-751, 1967.
27. Reeve, J.N., Trun, N.J. and Hamilton, P.T. Beginning genetics with
methanogens. In; Genetic Engineering of Microorganisms for Chemicals.
(A. Hollaender, R.D. DeMoss, S. Kaplan, J. Konisky, D. Savage, and R.S.
Wolfe, eds.). Plenum Publishing, New York, pp 233-244, 1982.
28. Pelham, H.R. A regulatory upstream promoter element in the Drosophila
Hsp70 heat-shock gene. Cell 30: 517-528, 1982.
29. Schnabel, R., Thomm, M., Gerargy-Schahn, R., Zillig, W. Stetter, K.O. and
Huet, J. Structural homology between different archaebacterial
DNA—dependent RNA-polymerases analyzed by immunological comparison of
their components. EMBO J. 2: 751-755, 1983.
384
-------
THE START OF NITRIFIER GENETICS
by
Michael Carsiotis and Sunil Khanna
Department of Microbiology and Molecular Genetics
University of Cincinnati
Cincinnati, Ohio 45267-524
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
385
-------
THE START OF NITRIFIER GENETICS
by: Michael Carsiotis and Sunil Khanna
Department of Microbiology and Molecular Genetics
University of Cincinnati
Cincinnati, OH 45267-524
ABSTRACT
Two bacterial groups, ammonia-oxidizers and nitrite-oxidizers, acting together,
are the primary microbes involved in the conversion of ammonia to nitrate. We have
begun a study of the genetic structure and organization of Nitrobacter hamburgensis
strain XI4, a nitrite-oxidizer. The gene which codes for B-isopropylmalate
dehydrogenase in this organism has been successfully cloned. We are sequencing the
gene and its flanking regions in order to learn the DNA sequences in Nitrobacter
which are used for transcription and translation of genes. This information will be
used to construct plasmid vectors with which to introduce genes into N. hamburgensis
X14.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
386
-------
INTRODUCTION
Nitrification is a desirable process in wastewater treatment and is brought
about largely by the concerted action of two groups of autotrophic bacteria which
together comprise the bacterial family Nitrobacteraceae. One group, designated by
the prefix Nitroso, contains numerous members all of which can oxidize ammonia to
nitrite; they are referred to as ammonia-oxidizers. The second group, designated by
the prefix Nitro, contains various members all of which can oxidize nitrite to nitrate;
they are referred to as nitrite-oxidizers. The term nitrifiers is used when referring
to both groups collectively. Despite their significant contribution to wastewater
treatment and to the global nitrogen cycle, virtually nothing is known of the genetics
of nitrifiers. This laboratory has begun a program designed to discover the genetic
structure and organization of Nitrobacter hamburgensts strain X14 (l), a nitrite-
oxidizing bacterium. The results will provide the type of information eventually
required to apply genetic engineering to nitrifiers in order to produce strains which
will increase the rate of nitrification in wastewater treatment facilities. The
approach which we think will be ultimately most useful in increasing the nitrification
rate is the introduction of genetic information into nitrifiers that will allow them to
grow faster in their unique biological niches such as wastewater treatment facilities.
Our immediate goal is the introduction of genes into _N. hamburgensis strain XI 4
which will increase its growth rate in autotrophic medium. This medium is as
comparably modest in nutritional value as wastewater. Which genetic information do
we think will produce a more rapidly growing N^ hamburgensis strain XI 4?
Specifically, we will introduce DNA derived from other nitrite oxidizing members of
the Nitrobacteraceae family especially those which occupy global ecological niches
other than waste water treatment plants. Another potentially useful DNA donor
would be Rhodopseudomonay palustris (2) which has been shown recently (3) to be
closely related phylogenetically to Nitrobacter.
NL hamburgensisl is a Gram-negative autotrophic bacterium which can obtain
carbon and nitrogen for synthesis of protoplasm from atmospheric CO2 and nitrite.
Its only other requirements are water, oxygen, phosphate and other minerals. From
this simple medium it can synthesize all of the necessary building blocks, e.g., amino
acids, purines,pyrimidines,vitamins, etc. It is assumed that the biosynthesis of these
molecules involves the same pathways used by the familiar heterotroph Escherichia
coli but this point has never been investigated. In fact the actual demonstration of
the existence of any familiar enzymatic reactions in nitrite-oxidizers is largely
confined to those involved in energy yielding processes. Ribulose-l,5-bisphosphate
carboxylase (4) and nitrite oxidoreductase (5) have been demonstrated in cell free
extracts. The former is a key enzyme for the fixation of CO_ in nitrifiers via the
Calvin cycle. The latter is the enzyme complex responsible for nitrite oxidation; it
has been purified and characterized recently (6). We chose to use Nitrobacter
hamburgensis strain XI 4, henceforth simply strain XI 4, in our studies because it grew
relatively rapidly, its nitrite-oxidizing activity was unusually high (1) and it, as well
as the related strain Y (1), were the only nitrite-oxidizers which contained plasmids
(7).
present two strains of N. hamburgensis, designated X14 and Y, have been
identified(l).
387
-------
EXPERIMENTAL
The purpose of our first experiments was to learn if an antibiotic resistance
plasmid could be introduced into strain XI4. We chose to use plasmids called broad
host range plasmids since they have been readily introduced and replicated in
numerous other Gram-negative organisms (8). They contain antibiotic resistance
genes which would allow detection of their entry and replication in strain XI4.
Although there were no reports of the introduction of broad host range plasmids into
nitrifiers the existence of plasmids in N. hamburgensis strains X14 and Y indicated
that-some plasmids can be maintained in nitrite-oxidizers. Introduction of plasmids
into new bacterial hosts can be by means of either conjugation or transformation
(8,9,10). The antibiotic resistance genes are especially useful because if the plasmid
is introduced and maintained the host is then resistant to an antibiotic to which it
was formerly susceptible. Our first experiments were designed to introduce the
broad host range antibiotic resistance plasmid RP1(8) and several derivatives of RP1
(11,12) into strain X14 by means of conjugation and transformation. By either means
success would have been evident from the growth of strain X14 on antibiotic
supplemented agar plates. Despite the use of a variety of plasmids and various
published experimental protocols for conjugation (13) and transformation (14,15) we
did not recover any antibiotic resistant isolates of strain XI4. Two, of several,
possible explanations for our results are seen in Table 1.
TABLE 1. POSSIBLE OUTCOMES OF CONJUGATION AND TRANSFORMATION OF
STRAIN XI* BY BROAD HOST RANGE PLASMIDS
Outcome
Conclusions
A. Isolation of antibiotic
resistant isolates of strain XI4.
B. No antibiotic resistant isolates
of strain X14 are obtained.
A.I. The DNA sequences required for
transcription and translation of
plasmid genes are functional in
strain XI4.
B.I. The DNA sequences required for
transcription and translation
of plasmid genes are not functional
in strain X14.
OR
B.2. The origin of replication of the
plasmid is not functional in
strain XIH.
388
-------
We chose to investigate first the possibility that genes in broad host range
plasmids are not functionally expressed in strain X14. Functional expression of a
plasmid gene, like any bacterial gene, requires the presence of: DNA sequences at
the y end of the gene which permit transcription of the gene into messenger RNA
(mRNA) by the cellular RNA polymerase; Shine-Dalgarno sequences (16) in the mRNA
which allow ribosornes to attach to mRNA; translation termination sequences at the
3' end of the gene and; transcription termination sequences at the 3' end of the gene.
In order to learn what sequences strain X14 uses for transcription and translation we
would have to isolate a gene from strain XI^ and determine its DNA sequence.
The isolation of a gene from strain X14 was achieved as follows. The
chromosomal DNA of strain X14 was digested partially (10) with the restriction
enzyme Sau_3A and the resultant 5 kilobase (kb) fragments were isolated from low
melting temperature agarose (10). The fragments were introduced into the BarnHI
site in the plasmid vector pMK2004 (Figure 1; reference 17); this produced a mixture
of recombinant plasmids.
/?/ Hmdffl
Hoel BamI
HincH "
PstI
HaeH
HaeH
HaeH
Hmdm
Xhol
Smal Aval
Aval
Figure 1. Restriction map of pMK2004. Coordinates are in kilobase pairs. Tet,
tetracycline resistance gene, Kanr, kanamycin resistance gene, Rep, origin of
replication, Ampr, ampicillin resistance gene.
This mixture was used to transform E^ coli C600 (10), a strain which has two
relevant mutations in its chromosome. One is a mutation in the gene leuB which is a
gene that codes for 3-isopropylmalate dehydrogenase, one of three enzymes required
for leucine biosynthesis. Because of this mutation E. coli C600 cannot grow unless it
389
-------
is provided with leucine in the medium and is said to be leucine-dependent. A second
mutation, in the thrB gene of E_. coli C600, causes it to be threonine-dependent as
well as leucine-dependent. After the recombinant plasmids were introduced into E.
coli C600 by the process of transformation, we were able to recover twelve clones of
this strain which were no longer leucine-dependent. In a separate transformation we
recovered three clones of E_. coli C600 which were no longer threonine-dependent.
The successful recovery of both leucine and threonine independent clones indicated
strongly that genes of strain X14 were functionally expressed in E_. coli. Let us
focus only on the recombinant plasmids in the leucine independent transformants. We
used them for a series of experiments designed to obtain the smallest piece of strain
X14 DNA which could express 3-isopropylmalate dehydrogenase. This DNA would
then be sequenced in order to learn which sequences are used for transcription and
translation in strain XI4. We focused on seven of the recombinant plasmids since the
other five were unstable. Three of the seven contained an approximately 12 kb
fragment of strain X14 derived chromosomal DNA; the other four contained an
approximately 6 kb fragment. In all probability the three larger plasmids were
reisolates of the same plasmid; the same is probably true of the four smaller
plasmids. One of the larger plasmids, pNBH6, and one of the smaller plasmids,
pNBH3, were digested with several restriction enzymes. The resultant data were used
to draw the restriction maps of their respective fragments (Figure 2).
pNBH6 fragment
.(12. 5kb)
P c rr
? iY }
,S s
1
SHE
Y ;
A EBB SPE*
I I I I I If
pNBH3 fragment
(6. 7kb)
Figure 2. Partial restriction map of the strain X14-derived fragments within
recombinant plasmids pNBH3 and pNBH6 which confer leucine independence on E^
coli C600. A, Sau3A, B, BamHI, E, EcoRI, H, Hindlll, P, PstI, S, Sail, kb, kilobases.
390
-------
It was evident that the leuB gene of strain X14 must be located in the common
BamHI -EcoRI fragment. Upon subcloning this fragment into pMK2004 the resultant
recombinant plasmid, pNBH601 (Figure 3), was found to confer leucine independence
on E. coli C600.
B S P E
kb 0
2. 4 3.6 3.7
pNBH601 fragment
(3. 7kb)
Figure 3. Partial restriction map of the strain X14-derived fragment in pNBH601
which confers leucine independence on E. coli C600. B, BamHI, E, EcoRI, S, Sail, P,
PstI, kb, kilobases.
The additional subcloning of the BamHI - Sail fragment of pNBH601 into pMK2004
produced the plasmid pNBH602 (Figure 4) which also conferred leucine independence
on E. coli C600.
B S
kb 0
2. 4
pNBH602 fragment
(2. 4 kb)
Figure 4. Partial restriction map of the strain XI^-derived fragment in pNBH602
which confers leucine independence on E. coli C600. B, BamHI, S, Sail, kb, kilobases.
We are now sequencing this fragment in order to learn a number of things. The most
important piece of information we will derive from the DNA sequence is whether
strain XI4- uses the same transcription and translation sequences that E. coli uses.
Once we have determined whether the transcription and translation sequences are the
same or different we will construct a shuttle plasmid vector. The steps involved in
the construction will depend on whether the transcription and translation sequences
of strain X14 and E. coli are identical or different (Table 2).
391
-------
TABLE 2. CONSTRUCTION OF A SHUTTLE PLASMID VECTOR
Transcription and
translation sequences of
E. coli and strain
Steps in the
construction of shuttle
plasmid vector*
A. Identical
B. Different
A.I.
B.I.
§
Clone ori of strain XI 4 chromosome or
plasmid into broad host range plasmid
Clone ori of strain X14 chromosome or
plasmid into broad host range plasmid
B.2. Use in_ vitro recombinant DNA techniques
to position strain X14 transcription and
translation sequences at 5' and 3' end of an
antibiotic resistance gene in the plasmid
*Shuttle plasmid vector can be broadly defined as a plasmid capable of replication
in two microbial hosts which are taxonomically distinct. One example is the
plasmid pHV14 (18) which can replicate in E. coli and Bacillus subtilis.
§
The origin of replication (ori) is the DNA sequence present on all replicons
(chromosome, plasmid, phageTwhich is necessary for the replicon's replication.
At the successful conclusion of the experiments outlined in Table 2 we will have
available a shuttle plasmid vector capable of introducing genes into strain X14. We
will introduce into that plasmid chromosomal DNA from other organisms. The
organisms chosen will be those that we believe contain genes which when expressed in
strain XI^ will increase its growth rate in autotrophic medium. As noted earlier two
sources of chromosomal DNA which seem most appropriate are those of other nitrite-
oxidizers and JR. palustris. The resultant plasmid will be introduced into strain XI4
and standard microbial genetic procedures used to isolate faster growing antibiotic
resistant isolates. The genetic engineering of ammonia-oxidizers would presumably
follow similar lines in order to produce a strain capable of faster growth in
autotrophic medium. The introduction of faster growing nitrifiers into wastewater
treatment facilities would have two beneficial effects. First, the overall rate of
nitrification would be increased. Second, the faster growing strains of nitrifiers
would be able to outgrow any competing nitrifiers which enter the wastewater
treatment facility.
392
-------
RESULTS AND/OR BENEFITS EXPECTED
Nitrification of wastewater is a useful process since it reduces the ultimate
BOD in the effluent to an acceptable level. The slow rate of nitrification adds to the
construction and operating cost of wastewater treatment facilities. The reliability of
nitrification is not consistent because periodic changes (seasonal, diurnal) in
nitrifying activity result in incomplete nitrification. Introduction of a more rapidly
growing strains of nitrifiers would decrease operating costs (less retention time),
decrease construction costs (eliminate the need for long-term aeration) and increase
the reliability of nitrification (more ecologically stable system).
Although nitrification is one limb of the overall nitrogen cycle that occurs
throughout the biosphere, the genetics, microbial physiology and biochemistry of
nitrification are poorly understood. Increased knowledge about nitrification may
permit applications in other waste treatment problems and in areas such as
agriculture, which depend on operation of the cycle and the use of inorganic nitrogen
as fertilizer.
CONCLUSION
A gene has been isolated from strain XI4, a nitrite-oxidizer. This will permit
us to determine the DNA sequences used by strain X14 for transcription and
translation of their genes. This information will be used to construct a shuttle
plasmid which contains an antibiotic resistance gene which should be functionally
expressed in strain XI4. The shuttle plasmid will serve as the vector for introducing
bacterial DNA into strain XI4 in order to increase the growth rate of the organism in
autotrophic medium.
ACKNOWLEDGEMENT
This investigation was supported by a grant from the U.S. Environmental
Protection Agency (Cooperative Agreement No. CR810888-01-0) to M.C.
393
-------
REFERENCES
All references to established background information are drawn from
references 8, 9 and 10. Specific referenced items can be found by consulting their
indices.
1. Bock, E., Sundermeyer-Klinger, H. and Stackebrandt, E. New facultative
lithoautotrophic nitrite-oxdizing bacteria. Arch. Microbiol. 136:281-284, 1983.
2. Pfennig, N. and Truper, H.G. The phototropic bacteria. Iru R.E. Buchanan and
N.E. Gibbons (ed.) Bergey's Manual of Determinative Microbiology, 8th ed.
Williams and Wilkins Co., Baltimore, 1974, p. 29.
3. Seewaldt, E., Schleifer, K-H., Bock, E., and Stackebrandt, E. The close
phylogenetic relationship of Nitrobacter and Rhodopseudomonas palustris.
Arch. Microbiol. 131:287-290, 1982.
4. Shively, J.M., Bock, E., Westphal, K., and Cannon, G.C. Icosahedral inclusions
(carboxysomes) of Nitrobacter agilis. J. Bacteriol. 132:673-675, 1977.
5. Aleem, M.I.H. and Nason, A. Phosphorylation coupled to nitrite oxidation by
particles from the chemoautotroph Nitrobacter agilis. Proc. Natl. Acad. Sci.
U.S.A. 46:763-769, 1960.
6. Sundermeyer-Klinger, H., Weber, W., Warninghoff, F., and Bock, E. Membrane-
bound nitrite oxidoreductase of Nitrobacter; evidence for a nitrate reductase
system. Arch. Microbiol. 140:153-158, 1984.
7. Kraft, I. and Bock, E. Plasmids in Nitrobacter. Arch. Microbiol. 140:79-82,
1984.
8. Hardy, K.H. Bacterial Plasmids. Iru Aspects of Microbiology 4. American
Society for Microbiology, Wash. D.C. 1981.
9. Glover, D.M. Gene cloning: the mechanics of DNA manipulation. Chapman and
Hall, London and New York. 1984.
10. Maniatis, T. Fritsch, E.F., and Sambrook, J. Molecular cloning: a laboratory
manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1982.
11. Weiss, A., and Falkow, S. Transposon insertion and subsequent donor formation
promoted by Tn _5JH in Bordetella pertussis. J. Bacteriol. 153:304-309, 1983.
12. Weiss, A.A., Hewlett, E.L., Myers, G.A., and Falkow, S. Tn5- induced
mutations affecting virulence factors of Bordetella pertussis. Infect. Immun.
42:33-41, 1983.
13. Broda, P. Plasmids. W.H. Freeman and Co., Oxford and San Francisco. 1979.
394
-------
14. Fornari, C.S. and Kaplan, S. Genetic transformation of Rhodopseudomonas
sphaeroides by plasmid DNA. J. Bacteriol. 152:89-97, 1982.
15. Weiss, A.A. and Falkow, S. Plasmid transfer to Bordetella pertussis:
Conjugation and transformation. J. Bacteriol. 152:549-552, 1982.
16. Shine, J. and Dalgarno, L. Determinant of cistron specificity in bacterial
ribosomes. Nature 254:34-38, 1975.
17. Kahn, M.L., Kolter, R., Thomas, C., Figurski, D., Meyer, R., Remaut, E. and
Helinski, D.R. Plasmid cloning vectors derived from the plasmids ColEl, F,
R6K, and RK2. Methods Enzymol. 68:268-280, 1979.
18. Goze, A. and Ehrlich, S.D. Replication of plasmids from Staphylococcus aureus
in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 77:7333-7337, 1980.
395
-------
REMOVAL OF PHOSPHATE AND OTHER SMALL MOLECULES FROM WASTE
STREAMS WITH BINDING PROTEINS IN CYCLING COLUMN ADSORBERS
by
Clement E. Furlong, Joseph A. Sundstrom, and R. J. Richter
Departments of Genetics and Medicine
Division of Medical Genetics
Center for Inherited Diseases
University of Washington
Seattle, Washington 98195
and
John Yin and Harvey W. Blanch
Department of Chemical Engineering
University of California
Berkeley, California 94720
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
397
-------
REMOVAL OF PHOSPHATE AND OTHER SMALL MOLECULES FROM WASTE
STREAMS WITH BINDING PROTEINS IN CYCLING COLUMN ADSORBERS
by: Clement E. Furlong , Joseph A. Sundstrom and R.J. Richter
Departments of Genetics and Medicine, Division of Medical
Genetics, Center for Inherited Diseases, University of
Washington, Seattle, WA 98195
and
John Yin and Harvey W. Blanch, Department of Chemical
Engineering University of California
Berkeley, California 94720
Supported by: U.S.E.P.A. Cooperative Agreement # CR811948
ABSTRACT
This report describes the use of a special class of stable, reversibly
denaturable proteins with high substrate affinity and specificity in cycling
column adsorbers designed for the removal of specific small molecules from
waste streams.
The class of proteins termed "binding proteins" resides in the space
between the inner and outer membranes (periplasmic space) of the gram-negative
bacteria. These proteins are involved in transporting nutrients across the
plasma membrane. They bind their respective nutrients with high specificity
and affinity. In addition, they are resistant to proteases and heat.
Further, these proteins bind their respective substrates through a broad range
of pH values and ionic strengths. These special properties make this class of
proteins especially interesting to consider for use in cycling column ad-
sorbers for removal of specific small molecules from waste streams, or other
aqueous environments.
We have attached the phosphate-binding protein from E. coli to cyanogen
bromide activated Sepharose 4B (Pharmacia). Radiolabeled phosphate (33-Pi) was
used to monitor the performance of the resin-bound protein. We demonstated
that the solid phase phosphate-binding protein efficiently removed phosphate
from a feed stream. The bound phosphate could be released from the binding
protein by reversibly denaturing the phosphate-binding protein. Upon re-
naturing, the resin-bound phosphate-binding protein could be subjected to
continuous cycles of loading, and unloading. Batch experiments were also
carried out with a glutamate-aspartate binding protein. It behaved in a
similar manner.
whom correspondence should be addressed.
398
-------
To facilitate the product
bacterial strain containing a
protein was used for the large
efficiency of the purification
periplasmic proteins directly
production of the phosphate-bi
This paper has been revie
Protection Agency's peer and a
presentation and publication.
The unique properties of
proteins termed "binding prote
to use in cycling column adsor
from waste streams. Proteins
and a high degree of substrate
differentiate chemical isomers
removal of the bound substrate
should be possible to design
different molecules from solut
protein of E. coli as a model
produced in reasonable quantit
that solid phase proteins with
stability can be used in cycli
molecules from solution.
EXP
MATERIALS AND METHODS
Bacterial Strains. E. coli st
supplied by Dr. H. Rosenberg.
Measurement of Phosphate Bindi
of phosphate to the binding pr
or by a simple membrane filter
Purification of Phosphate-Bine
purified from the periplasmic
formed with a plasznid bearing
protein. The overproduction wa
phosphate deficient medium.
described experiments was puri
producer strain by gel filtrat on
Osmotic Shock Procedure. The
10 liter culture by a standard
developed chloroform extractio
on of the phosphate-binding protein, a
lasmid bearing the gene that encodes this
scale purification of protein. To increase the
procedure, a mutant strain that secretes its
nto the medium is being designed for high level
ding protein.
ed in accordance with the U.S. Environmental
ministrative review policies and approved for
INTRODUCTION
he class of bacterial nutrient transport
ns" made them worth considering as candidates
ers for the removal of specific small molecules
ffer the possibility of high affinity binding
specificity. For example, they can easily
If a simple procedure is designed for the
and renaturation of the active protein, it
finity columns for the removal of many
on. We have chosen the phosphate-binding
ystem, since it has been cloned and can be
The experiments described below demonstrate
high affinity, high specificity and high
g column adsorbers to scrub specific small
RIMENTAL PROCEDURES
ain AN1685 transformed with plasmid pAN92 was
g to the Phosphate-binding Protein. The binding
tein was determined by equilibrium dialysis (1)
assay (2).
ng Protein. Phosphate-binding protein (3) was
rotein fraction of an E. coli strain trans-
he gene that encodes the phosphate-binding
achieved by growing this organism on a
he phosphate-binding protein used for the
ied from osmotic shock fluid from the over-
through a Biogel P-100 column.
eriplasmic protein fraction was prepared from a
osmotic shock procedure (4) or a recently
procedure (5).
399
-------
Analytical Gel Electrophoresis Procedures. Standard sodium dodecyl sulfate
polyacrylamide gel electrophoresis (6) was used to analyze the protein
composition for fractions of interest.
Coupling of Protein to the Solid Phase Support — Phosphate-binding protein
was coupled to cyanogen bromide activated (7) solid phase support, Sepharose
4B. The protein loaded gel (2 mg protein/ml resin bed vol.) was packed into a
water jacketed column (1.6 x 5 cm).
Analytical Apparatus. The column feed line was connected to a valve that
switched between the solution containing labeled ligand (0.8 uM,33Pi, 50
uCi/1, in 20 mM Tris-HCl, pH 7.0) and the wash solution (20 mM Tris-HCl, pH
7.0). The column eluate was joined to a stream of scintillation solution and
run through a continuous flow scintillation counter (Radiomatic Flo I, Tampa,
FL). The feed stream flow rate was 1 ml/min and the scintillation flow 10
ml/min. Data were recorded both as linear counts/time from a printer and log
counts/time on a strip chart recorder. A diagram of the apparatus is shown in
Figure 1, and a photograph in Figure 2.
Protein Assays. Protein was assayed by the method of Lowry (8).
IN-LINE SCINTILLATION COUNTER
DIGITAL PRINTOUT OF CONTINUOUS
RADIOACTIVITY FOR PROGRAMMED PIONITORING
TIME INTERVAL OF RADIOACTIVITY
(COUNTS/TIME) (Loc COUNTS/TIME)
FIGURE 1. DIAGRAM OF THE PHOSPHATE-BINDING PROTEIN REACTOR
400
-------
Figure 2.
Photograph of th
detector.
s phosphate adsorber and in-line scintillation
RESULTS
The goal of our initial studies
could be constructed with proteins
possible, the proteins used need
was to determine if a cycling adsorber
on a solid phase support. For this to be
sd to have the following properties:
o High affinity and specif
o Ability to reversibly de
convenient and inexpensi
stream treatment
o Ability to renature afte
cycles of denaturation/r
The phosphate-binding prote
studies, since it could be produ
with a relatively inexpensive anJ
Figure 3 shows the gel filt
protein from the periplasmic pro
electrophoretic analysis of the
protein is shown in Figure 4. T
highly purified phosphate-bindin
produces such a high level of th
The phosphate-binding protein copii
plasmic protein fraction.
Lcity for the solute
lature and release bound solute under
/e conditions that would be useful in waste
r ligand release and undergo many
snaturation
Ln of E. coli was chosen for our initial
ed in quantity, and binding could be monitored
safe isotope, -^-Pi.
ation purification of the phosphate binding
tein fraction of E. coli. A disk gel
eriplasmic protein fraction and purified
ie single step purification produces
g protein. This is possible since the strain
s protein when grown on low phosphate medium.
iprises between 60-70% of the entire peri-
401
-------
0,8
0,6
0,2
O
CQ
2 cf
10 20 30
FRACTION NUMBER
Figure 3. Gel filtration chromatography of osmotic shock fluid.
30,000
22,000
Figure 4. SDS-PAGE analysis of the purification
of the phosphate-binding protein. Lane 1,
molecular weight markers: 94,000, phosphorylase
B; 68,000 bovine serum albumin; 43,000 ovalbumin;
30,000, carbonic anhydrase; 22,000, soybean
trypsin inhibitor. Lane 2, crude shock fluid.
Lane 3, purified phosphate-binding protein.
402
-------
The purified protein was c
at a density of 2 mg protein/ml
packed into a water jacketed co
taining 33_pi was passed throug
monitored with an in-line scint
switching between the feed stre
experiment. A schematic repres
presented in Figure 5.
ATTACH PROTEIN TO INSOLUBLE MATRIX
Figure 5. Diagram of the
the solid phase
upled to the solid support resin Sepharose 4B
resin. The protein conjugated resin was
umn as described above. A feed stream con-
the column and the outflow was continuously
nation counter. A three-way valve allowed
m and the wash buffer as indicated in each
ntation of the solid-phase protein is
PROTEIN BINDS PHOSPHATE
DENATURE PROTEIN TO RELEASE PHOSPHATE
RENATURE PROTEIN FOR NEXT CYCLE
ttachment of the phosphate-binding protein to
resin and its reversible binding of phosphate.
403
-------
A series of experiments were carried out with the phosphate affinity
column. Figure 6 shows a run in which the column was loaded to full capacity,
then rinsed with rinse buffer, and finally stripped by reversibly denaturing
the phosphate-binding protein. From the trace, several things are observed.
The column efficiently removed the labeled phosphate from the stream until the
ligand sites began to saturate. As the binding capacity of the column was
approached, counts appeared in the effluent. If, as in this experiment, the
column was allowed to fully saturate before beginning the wash and elution
steps, there was a long bleed of isotope from the column. During the steady
state rinse, the protein was denatured. It is clear that denaturation of the
protein resulted in the release of its bound solute into the rinse stream.
10
I
CD
X
Q_
6
4
2
C/O
GO
<
CO
I
1
50
100
TIME (MIN)
150
Figure 6. Loading of the phosphate adsorber to full capacity followed by
washing and elution of the bound phosphate.
Figure 7 shows another variation of column loading and elution. in
this experiment, the valve was rerouted to run the feed stream directly
through the scintillation counter. When steady state feed was achieved, the
system was washed with buffer, then the feed stream was switched to the
column. When counts began to appear in the eluate, the valve was switched to
the rinse buffer and the bound phosphate was released by denaturing the
protein.
404
-------
-pa
O
cn
reversible bindi
3
iQ
ri-
ff
hi
O
£
ia
y
a
fD
3
0)
rt
0
h{
01
rt
H-
0
3
X
fD
3
0>
rt
£
hi
01
ft
H'
O
3
•
specificity for
scrub the specif
H- D)
O O
H
cn 0
O rt
H fD
^
rt O
CD Oi
3
Hi
h! cr
O CD
B
01
01 rt
O rt
H fli
0 O
rt ff
H- CD
o a
3
• rt
0
•fl W
0 0
hi H
rt h>-
3- a
CD
hi -a
*• ty1
a>
rt 0)
y CD
CD
01
^3 C
H *a
o >a
rt o
(D hj
H- rt
3 tn
p. Q)
01 3
a
O
0) 0
T3 01
O) CD
cr a
H
CD rt
O
0
Hi
These exper
H-
3
CD
3
rt
01
a
CD
3
0
3
01
rt
hi
0>
rt
CD
ri-
ff
0>
rt
01
V
h!
O
rt
CD
H-
3
s;
H"
rt
y
y
H-
irj
ff
o>
l-h
Hi
H-
3
H-
rt
^
pj
3
Hi ft
CD ff
CD CD
a
H OJ
CD 01
< ff
CD
• 0)
01
W
ft
01
hi
rt
CD
a
0)
01
01
o
o
3
Oi
01
rt
ff
CD
O
O
0
3
rt
01
cr
(Q
Oi
3
rt
0
hi
H-
01
CD
0>
cr
o
o
iQ
h{
O
0
3
a
denaturation , an
appeared to loos
CD a
h4 hi
H- fD
rt 3
rt 0)
H rt
CD 0
hi
0> 0>
O rt
rt H-
H- O
< 3
H- •
rt
^
H9
a ff
0 CD
hi
H- H-
3 B
<£> 3
O
o cr
O H-
3 H
01 H-
CD N
O CD
0 a
rt
H- -O
< ff
CD O
Dl
0 13
^ ff
O 01
H rt
CD CD
01 1
• cr
H-
3
^ a
0 P-
h^ p
tj^
ri-
ff t)
CD hi
0! O
fD rt
CD
hj H*
0 3
3
cn
*•
shows three cons
CD
0
0
rt
H-
CD
O
O
H
fO
01
0
Hi
PJ
O
Oi
a
H-
3
*•
s:
01
ff
H-
3
••
01
0
H
pi
rt
fD
M
CD
H
CD
DJ
01
CD
cr
^
function in subs
CD
0
CD
3
rt
O
n
H
CD
01
O
Hi
cr
H-
3
a
H-
3
iQ
X.
a
3
Oi
rt
^
hj
01
rt
pi
O
3
h!
CD
3
0>
rt
hi
01
rt
H-
0
3
•
ftj
H-
ifl
H
fD
03
ff
CD
CD
X
rt
I
O
tQ
^
CD
Dl
rt
H-
O
3
£
0>
Dl
S
ff
CD
ri-
ff
CD
rt
y
CD
TD
0
rt
fD
H-
3
O
0
H
a
cr
CD
n
fD
3
0>
rt
0
fD
a
0)
3
a
H-
3
rt
O
rt
ff
CD
hi
H-
3
01
CD
cr
0
Hi
Hi
CD
Hi
CD
H
rt
p-
O
3
O
Hi
ri-
ff
fD
cr
O
3
a
^O
ff
0
TO
t3
ff
OJ
rt
CD
Hi
O
3
ri-
ff
fD
Q
O
H
0
B
3
01
01
rt
fD
5
ri-
ff
CD
rt
ff
hi
O
ua
ff
rt
ff
CD
O
0
H
g
3
Hi
O
H
H
O
f.
fD
a
cr
"3
H-
hi
CD
hH
O
0>
H-
3
rt
ff
CD
Hi
CD
CD
a
01
rt
(D
0>
B
a
H-
CD
o
rt
H
rt
O
ri-
ff
CD
a
CD
rt
CD
O
rt
O
CPM x 10"
CO
CD
CD
LOAD DIRECTLY TO DETECTOR
WASH
LOAD TO COLUMN
ELUTE PHOSPHATE
-------
100
150
TIME (MIN)
Figure 8. Continuous cycles of phosphate binding/unloading. L indicates
the beginning of phosphate loading; W indicates the wash step
and E, the elution of the bound phosphate.
The next question is whether large quantities of protein can be produced
economically. Two major considerations come into play in designing automated
procedures for producing large quantities of the desired protein(s). First,
It will be important to engineer strains that produce large quantities of the
desired protein(s). Second, it would be most efficient if the overproduced
protein(s) could be secreted directly into the medium. If these two criteria
can be met, automated, large-scale protein generation procedures can be
developed.
The first problem was approached by using a mutant of E. coli that
secretes its periplasmic proteins into the medium. Figure 9 shows that this
strain appears to also secrete the phosphate-binding protein directly into
the medium. Thus, the first problem of achieving direct secretion of the
phosphate-binding protein into the medium appears to be solved.
406
-------
Molecular
Weight
94,000
68,000
43,000
30,000
22,000
Figure 9. SDS-PAGE analysis of the
production and secretion of the
phosphate-binding protein by E. coli
strain C90. Lane 1, molecular weight
markers (see Figure 4). Lane 2,
Phosphate-binding protein standard
(5 ug). Lane 3, chloroform extracted
periplasmic fraction of strain C90 (50
ug). Lane 4 molecular weight markers.
Lane 5, phosphate-binding protein
standard. Lane 6, concentrated growth
medium from strain C90 (50 ug).
The second problem, that of producing strains that generate and secrete
very high levels of binding protein into the medium, is presently being
pursued. The rationale of the general approach is as follows:
The level of the phosphate-binding protein is controlled by the intra-
cellular level of inorganic phosphate. The phosphate level in turn is
controlled by the level of the phosphate-binding protein dependent transport
system which is composed of the phosphate-binding protain and probably three
membrane-bound proteins (9). The binding protein dependent system is
responsible for controlling the intracellular phosphate level. A mutation in
any of the genes encoding components of the binding protein dependent trans-
port system results in a defective transport system, a drop in intracellular
phosphate, and the derepression of at least 20 different operons involved in
phosphate transport and metabolism (10). Mutations in the genes encoding the
membrane associated protein components of the transport system allow for the
high level of production of the phosphate binding protein in strains bearing
plasmids that express only the phosphate-binding protein (9).
Thus, to achieve a high level of production of phosphate binding protein
in a strain that secretes this protein directly into the medium, the following
are necessary:
o A mutant strain with a defective cell wall that leaks
its periplasmic fraction into the medium
407
-------
o The strain should have a defect in one or more of
the membrane associated proteins of the binding protein dependent
phosphate transport system.
o In addition, this strain should have a high copy number
plasmid that carries the gene that encodes the phosphate
binding protein, but not the gene that encodes the
mutant membrane associated component(s) of the transport
system.
Experiments to generate an overproducer strain that secretes high levels
of phosphate-binding protein into the medium are in progress. An experiment
that illustrates the high levels of phosphate-binding protein that can be
secreted into the periplasm is shown in Figure 10. Growth on low phosphate of
the strain that carries the gene for the binding protein on a plasmid results
in a very high level of phosphate-binding protein in the periplasmic fraction.
Approximately 70% of the entire periplasmic fraction is phosphate-binding
protein.
PBP
Figure 10. SDS-PAGE analysis of the production
of phosphate binding protein by cells grown in
low versus high phosphate. Lane 1 ,
periplasmic fraction (50 ug) of cells grown
on high phosphate. Lane 2, periplasmic
fraction (50 ug) of cells grown on low
phosphate. PBP, Phosphate-binding protein.
The generality of the approach of solute removal by immobilized binding
proteins has also been tested with immobilized glutamate-aspartate binding
protein. Batch binding experiments gave results similar to those obtained
with the phosphate-binding protein studies.
408
-------
DISCUSSION
Proteins are generally thought of as rather fragile, unstable molecules.
The existence of a special class of proteins that are resistant to proteases,
boiling, conditions of very high and very low osmotic strengths and high and
low pH values led us to consider the use of this class of proteins for the
construction of cycling column adsorbers for the removal of specific solutes
from waste steams. The first basic question was whether a regenerable ad-
sorber could be designed that would provide efficient scrubbing of a solute
from a stream. Our preliminary experiments with a small-scale adsorber con-
taining immobilized E. coli phosphate-binding protein indicate that the basic
technology is feasible.
The question of designing production strains that would generate large
quantities of protein and secrete the overproduced protein directly into the
culture medium also appear to be feasible. Our preliminary experiments
indicate that the phosphate-binding protein can be secreted into the medium
(Figure 9). Studies on the regulation of production of the phosphate-binding
protein by others (e.g. 9) and verified in our experiments indicate that it
should be possible to generate strains that secrete very large quantities of
this protein directly into the medium.
The question of the diversity of proteins available for producing
specific reactors is interesting. Table 1 lists the properties of the trans-
port system associated binding proteins from E. coli and S. typhimurium that
have so far been characterized. In addition, two cadmium binding proteins
from E. coli have been described (11). In future experiments, we will
investigate the feasibility of using one of the E. coli cadmium binding
proteins in a cadmium adsorber. We have also described a general procedure
for finding specific ligand binding proteins in virtually any organism (12).
The procedure uses radiolabeled ligand to locate specific binding proteins in
a two-dimensional gel of: whole cell proteins. This procedure should be useful
for isolating proteins with high affinity and specificity for other problem
solutes.
SUMMARY
Proteins with high affinity and specificity for specific small molecules
can be attached to solid phase supports and used to scrub the molecules of
interest from solution. The adsorbed small molecules can be released from the
protein simply by reversible denaturation of the protein. Recombinant DNA
procedures can be used for the large-scale production of the specific
proteins.
409
-------
TABLE I
BINDING PROTEINS FROM Escherichia coli AND Salmonella typhimurium
Binding-
Protein
Organism
Molecular
Weight
KD
(uM)
Amino
Acids
Peptides
Orni thine
Lysine-
Arginine
Ornithine
Cystine
Glutamine
Glutamate-
Aspartate
Histidine
Leucine-
Isoleucine-
Valine-
Threonine
Leucine-
Specific
Oligopeptide
E. coli
E. coli
S . typhimurium
E. coli
E. coli
S . typhimurium
E. coli
S . typhimurium
E. coli
S . typhimurium
E. coli
S . typhimurium
E. coli
S . typhimurium
E. coli
S . typhimurium
28-33K
26-30K
26K
27-28K
23-29K
23K
29-32K
3 OK
25-31K
25K
36. 7K
35-39K
37K
34-38K
52K
52K
Arg .03-.1
Arg 0.15
Lys 3
Orn 5
Cys 0.01
0.15-0.3
0.8-6
Asp 1
0.8
0.15-1.5
0.2-2
Leu 0.43
He 0.15
Val 0.89
0.7
0.54
410
-------
TABLE I (continued)
Binding-
Protein
Organism
Molecular
Weight
(uM)
Sugars
Anions
Vitamins
Arabinose
Galactose-
( Glucose)
Maltose
Ribose
Xylose
Citrate
Phosphate
sn-Glycerol-3-
Phosphate
Sulfate
B12
Thiamine
E. coli
E. coli
S . typhimurium
E. coli
E. coli
S . typhimurium
E. coli
S . typhimurium
E. coli
E. coli
S . typhimurium
E. coli
E. coli
33K 0.2-2
32K 1
33K 0.38
40. 7K 1
29. 5K 0.13
29K 0.33
37K 0.6
28K 1-2.6
34K 0.8
45K 0.2
34. 7K 0.02
22K 0.005
0.03-.1
Data derived from reference 13.
411
-------
REFERENCES
1. Furlong, C.E., Morris, R.G., Kandrach, M. and Rosen, B.P. A multichamber
equilibrium dialysis apparatus. Anal. Biochem. 47: 514, 1972.
2. Furlong, C.E. and Weiner, J.H. Purification of a leucine-specific binding
protein from Escherichia coli. Biochem. Biophys. Res. Commun. 38: 1076,
1970.
3. Medveczky, M. and Rosenberg, H. The phosphate-binding protein of E. coli.
Biochim. Biophys. Acta. 211: 158, 1970.
4. Willis, R.C., Morris, R.C., Cirakoglu, C., Schellenberg, G.D., Gerber, N.
and Furlong, C.E. Preparation of the periplasmic binding proteins from
Salmonella typhimurium and E. coli. Arch. Biochem. Biophys. 161: 64,
1973^
5. Ames, G.F.-L., Prody and Kustu, S. Simple, rapid, and quantative release
of periplasmic proteins by chloroform. J. Bacteriol. 160: 1181, 1984.
6. Laemmli, U.K. Cleavage of structural proteins during the assembly of the
head of bacteriophage T4. Nature (London). 222: 680, 1970.
7. Porath, J. Preparation of cyanogen bromide-activated agarose gels. J.
Chrom. 86: 53, 1973.
8. Lowry, O.K., Rosebrough, N.J., Farr, A.L. and Randall, R.J. Protein
measurement with the Folin phenol reagent. J. Biol. Chem. 193: 265, 1951.
9. Morita, T., Amemura M., Makino, K., Shinagawa, H., Magota, K., Otsuji, N.,
and Nakata, A. Hyperproduction of phosphate-binding protein, phoS, and
pre-phoS proteins in Escherichia coll carrying a cloned phoS gene.
Eur. J. Biochem. 130:427, 1983.
10. Wanner, B.L. Overlapping and separate controls on the phosphate regulon in
Escherichia coli K12. J. Mol. Biol. 166: 283, 1983.
11. Khazaeli, M.B. and Mitra, R.S. Cadmium-binding component in Escherichia
coli during accommodation to low levels of this ion. Appl. Environ.
Microbiol. 41: 46, 1981.
12. Copeland, B.R., Richter R.J. and Furlong, C.E. Renaturation and
identification of periplasmic proteins in two-dimensional gels of
Escherichia coli. J. Biol. Chem. 257: 15065, 1982.
13. Furlong, C.E., in press, Methods in Enzymology.
412
-------
DEGRADATION OF CHLORINATED BENZOATES UNDER A VARIETY
OF ANAEROBIC ENRICHMENT CONDITIONS
Barbara R.
by
Sharak-Genthner.
and George E.
Jayne B. Robinson,
Pierce
Batelle-Columbus Division
Columbus, Ohio 43201
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
413
-------
DEGRADATION OF CHLORINATED BENZOATES UNDER A VARIETY
OF ANAEROBIC ENRICHMENT CONDITIONS
by: Barbara R. Sharak-Genthner, Jayne B. Robinson,
and George E. Pierce
Battelle-Columbus Division
Columbus, Ohio 43201
ABSTRACT
Degradation of chlorinated benzoates was studied under a number of
anaerobic conditions. Enrichments were prepared by the addition of 3- or
4-chlorobenzoate to basal medium and 10% secondary anaerobic digested sludge.
A 3-chlorobenzoate degrading consortium obtained from Dr. J. M. Tiedje was
also studied as a model system for isolating and identifying organisms
capable of dehalogenation.
Degradation of 4-chlorobenzoate was not observed after 29 weeks under
any of the enrichment conditions. Degradation of 3-chlorobenzoate was
observed in one of the sewage enrichments at 10 weeks. It was observed
that the presence of nitrate inhibited growth and dehalogenation and that
excess H£ did not enhance the establishment of a 3-chlorobenzoate degrading
consortium.
We were unable to isolate a novel 3-chlorobenzoate degrader from the
consortium and have recently initiated genetic studies using the 3-chloro-
benzoate degrading pure isolate DCB-1.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
414
-------
INTRODUCTION
Halogenated organic compounds pose a serious environmental problem,
because they enter the environment in substantial amounts, are toxic, and
tend to accumulate In sediments and soils affecting both flora and fauna
(1, 2, 3). Mlcroblal degradation could be employed to remove these
compounds j_n situ or during waste treatment. Aerobic degradation has
been studied, but some of these compounds do not appear to be degradable
aeroblcally (4, 5) and others are degraded in a manner in which highly
toxic Intermediates are formed (4, 6, 7a, 7b, 8, 9, 10, 11). Relatively
little Is known about anaerobic degradation of these compounds. However,
recent investigations indicate that anaerobic degradation of a number of
halogenated organic compounds occurs Including halogenated benzoates (12,
13, 14) halogenated phenols (15, 16, 17, 18) and halogenated short chain
aliphatic compounds (5, 19, 20). These studies have shown that some
compounds which are not degraded aerobically are readily degraded
anaeroblcal ly. In contrast to aerobic degradation, the first step of
anaerobic degradation Is the removal of the halogen (13, 14) leading
Immediately to the formation of a less toxic, more biodegradable
compound. In the present study, we enriched for microblal anaerobic
degradation of chlorinated benzoates under a number of anaerobic
conditions In order to obtain a variety of organisms capable of
dehalogenatlng and degrading chlorinated benzoates. We also obtained the
3-CI-benzoate degrading consortium of Dr. J. M. TIedje (21a) to use as a
model system for Isolating and Identifying the organisms responsible for
dehalogenation and degradation.
MATERIALS AMD METHODS
ANAEROBIC METHODS AND MEDIA
The anaerobic techniques employed for the collection of Inocula;
preparation of media; and handling of enrichments and cultures were
essentially those of Hungate (21b) as modified by Bryant (22) and Balch
and Wolfe (23). The basal medium contained yeast extract (0.15?, w/v),
B-vitamlns, minerals, NaHC03, Na2$ reducing solution, resazurin redox
Indicator, and a 90$ N2-10? C02 gas phase (final pH 7.0). The terminal
electron acceptor was CX>2 for methanogenlc media, while 20mM Na2S04, 15mM
KN03 or 20 mM sodium fumarate was added for sulfate, nitrate or fumarate
enrichments, respectively. Fermentative enrichments were prepared by
adding the carbohydrates used In the Complete Carbohydrate medium (CO of
Leedle and HespelI (24).
415
-------
MOST PROBABLE NUMBERS (MPN) ANALYSIS
Ten-fold serial dilutions were prepared with anaerobic dilution
solution. One ml of each dilution was inoculated Into triplicate tubes
containing nine ml of the desired medium. The MPN was estimated from the
number of tubes In each triplicate set which were positive for a desired
characteristic (25).
SOURCE AND COLLECTION OF INOCULUM
The source of Inoculum was secondary anaerobic digested sludge
(Jackson Pike Plant, Columbus, Ohio). Inoculum was collected In a sterile
2 liter glass carboy containing a stir bar for mixing and sealed with a
black rubber stopper fitted with a one-way gas valve to allow release of
gas pressure.
ENRICHMENTS
Enrichments were prepared by adding sterile anaerobic 3- or
4-ch I orobenzoate to the basal medium and \Q% sewage sludge as Inoculum.
Enrichments were also prepared to determine the effect of hydrogen on the
development of actively degrading consortla by adding \0% ^2 "1"°
enrichments after Inoculation. Enrichments were Incubated at 35°C.
Degradation of the compound was followed over time by HPLC analysis. Once
degradation was observed, the enrichments were transferred to fresh media
and passed several times to stabilize the activity.
HPLC ANALYSIS
E'enzoate, 3-CI-benzoete and 4-CI-benzoate were separated, Identified
and quantified by high pressure liquid chroniatography (HPLC). A reverse
phase C18 Llchroborb column (10 u, 4.6 mm (ID) x 25 cm. All tech
Associates, Inc., Deerfleld, III.) was used v,! th n liquid pluise of
methenolrwaterracetlc acid (6:4:0.5) at a flow rate of 0.7 ml/mln.
Fluid samples for HPLC analysis were collected aseptlcally and
anaeroblcal ly via a 5 ml sterile syringe which had been flushed with
anaerobic gas before sample collection. Samples were prepared for
analysis by filtering through a 0.45 urn Spartan-25 nylon filter
(Schlelcher & Schuell, Inc., Keene, N. J.) Into new dlposable glass test
tubes. Slightly larger than 100 ul of sample was Injected via Hamilton
syringe Into a 100 ul Injection loop (Beckman/Altex; Berkeley, Calif.).
RESULTS AND DISCUSSION
DEGRADATION IN SEWAGE ENRICHMENTS
Degradation of 4-CI-benzoate was not observed after 29 weeks of
Incubation under any of the enrichment conditions used (Table 1).
416
-------
Persistence of 4-CI-benzoate under methanogenlc enrichment conditions had
been prevlosuly reported (12) after 65 weeks of incubation.
3-CI-benzoate was degraded In one of the duplicate nitrate
enrichments without added H2 at 10 weeks (Table 1). It had not shown
degradation at 6 weeks. Degradation was observed at 10 weeks In the
second duplicate as the 3-CI-benzoate concentration had been reduced from
965 uM at 6 weeks to 689 uM at 10 weeks. All 3-CI-benzoate had been
degraded In the second duplicate by 23 weeks. The methanogenlc and
fumarate enrichments, which had not shown degradation at 10 weeks, showed
complete degradation of 3-CI-benzoate at 23 weeks. The sul fate and CCM
enrichments did not show degradation after 23 weeks of incubation.
These results indicate that although anaerobic dehalogenation Is a
reductive process (13, 14), the presence of excess fi? does not enhance
the establishment of a consortium capable of degrading 3-CI-benzoate.
Microscopic examination of the enrichments revealed a mixture of
gram-negative rods of varying lengths.
The nitrate enrichment showing degradation at 10 weeks was refed 800
uM 3-CI-benzoate four times and showed an increased rate of degradation
with each subsequent feeding (Figure 1a). A ten-percent transfer of this
enrichment Into fresh medium under methanogenlc, sulfate-reducing and
nitrate-reducing conditions indicated that nitrate was not required for
degradation, and It was actually inhibitory (Figure 1b). Fifty percent
transfers of the other enrichments into fresh medium without H2 and with
and without the terminal electron acceptor that had been present in the
original enrichment Indicated that neither \\2 n°r the terminal electron
acceptor were required for degradation. After 5.5 weeks of incubation
the transferred fumarate enrichment showed complete degradation of
3-CI-benzoate in the presence or absence of fumarate. At 10.5 weeks the
methanogenlc enrichments showed complete degradation, while the nitrate
enrichments passed to medium containing nitrate showed no degradation.
At 10.5 weeks only one of the duplicate nitrate enrichments passed to
medium lacking nitrate showed degradation. These data support the
conclusion that the presence of nitrate is actually inhibitory to
degradation of 3-CI-benzoate under these growth conditions.
DEVELOPMENT OF A 3-CI-BENZOATE DECHLORINATING ENRICHMENT
A 3-CI-benzoate degrading enrichment (14) was obtained from Dr. J.
M. Tiedje. It was reported that the dechlorlnatlng organism, which had
been Isolated In pure culture (21a), reduced nitrate to nitrite; grew
somewhat poorly with pyruvate; and performed reductive dehalogenation of
3-CI-benzoate. The consortium was Inoculated (10/O Into three media in an
attempt to Isolate the dechlorInating organism which was no longer
available In pure culture. These Included the basal medium plus 15 mM
KNC^; basal medium plus 15 mM KN03 and 50? H2; and basal medium plus 0.3?
pyruvato ane 15 nt' KI"C3. Tlu 3-C!--bcnzGci-te concentration was dnterrnfned
at 0, 3, 16 and 44 deys (Figure 1c). At 44 days the culture containing
417
-------
3-CI-benzoate plus 15mM KN03 showed no detectable 3-CI-benzoate, but a
large peak was observed with a lower retention time (8.4 min versus 14
m!n) corresponding to 694 uM benzoate. The concentration of
3-CI-benzoate In the enrichments containing either H2 or pyruvate did not
decrease during this time.
These data indicate that the 3-CI-benzoate was being dechlorinated,
but not degraded in the nitrate containing enrichment. After transfer
into fresh medium, it was found that 3-CI-benzoate was dechlorinated at a
rate of 3.8 umoles/liter/hour and that as the 3-CI-benzoate concentration
decreased the benzoate concentration increased (Figure 1d).
A gram stain of the culture showed that the dominant eel I type was a
small gram negative coccobaci11 us predominately found in pairs and chains
(Strain BG19, Figure 2a). Some large gram negative rods were also
present. Cel Is with the distinct morphology of Methanosp i r tII urn
h, ungate 1. which were very apparent In the original consortium, were
absent in the dechlorinating consortium. Methane was not detected. As a
result of the degradation data, we can assume that under these conditions
the benzoate degrader(s) and methanogens have been selected against and
the dechlorInating organ!sm(s) is one or both of the cell types present.
CHARACTERIZATION OF THE DECHLORI NAT ING CONSORTIUM
A Most Probable Numbers (MPN) analysis of the dechI or InatIng
consortium was performed (25) In order to determine total cells/ml versus
dech lor Inat!ng organisms per ml. The consortium was diluted from 1 x 10^
to 1 x 10~9 in anaerobic dilution solution and one ml was Inoculated Into
nine ml of basal medium containing 3-CI-benzoate (800 uM) plus 15 mM
KN03. Optical densities were followed for two weeks and revealed a total
of 2.5 x 10« cells/ml. HPLC analysis after 6 weeks indicated the
presence of approximately 1 x 103 dechlorInatIng ceils/ml. Thus, the
dechlorinating organisms are present as a very minor portior of the total
population indicating that direct isolation would be very difficult.
Microscopic examination of the MPN cultures revealed that the small
coccobaci Mus was the predominant cell type at all dilutions, but some-
large gram negative rods were present at the lower dilutions. A few
small gram negative rods were also present at all dilutions. Since these
rods could not be diluted away from the coccobaciI I us, they may be a
morphological variation of the coccobaci I lus which had felled to divide
properly. Their presence could also indicate an obligate syntrophic
association between these two cell types under our growth conditions.
A series of media were Inoculated with the dechlorinating consortium
to determine the physiological capabilities of the organisms present.
These Included the basal medium lacking yeast extract with the following
additions: 5 mM KN03 only; 5mM KN03 + 50* H2; 5mM KM03 +0.1$ yeast
extract; 5mM KN03 + 50$ H2 + Q.1% yeast extract; 5mM KN03 +0.3? lactate-
5mM KN03 + 0.3* lactate + 0.1$ yeast extract; 0.3? lactate + 0.1% yeast
extract; 3-CI-benzoate (800 uM) + 0.1$ yeast extract; 3-CI-benzoate +
418
-------
0.]% yeast extract + 50% H2; 50? H2 + 0.1? yeast extract; 0.1? yeast
extract only ; 3-CI -benzoate + 0.1? yeast extract + 5mM KNC>3;
3-CI-benzoate + 5 mM KM03; and 3-CI-benzoate only. Hydrogen was added as
a potential source of reducing equivalents for reductive dechI or inatlon
and lactate was added as an energy source for nitrate reducers present.
The maximum growth (600 nr.i) In triplicate tubes was followed for two
weeks (Table 2). Growth did not occur in medic lacking yeast extract.
The presence of leictate was found to inhibit growth even in the presence
of yeast extract. The best growth was observed in medie containing H2 In
addition to yeast extract and the presence of nitrate tended tc Inhibit
growth. These data indicate that organisms with a hydrogenase were
present in the consortium and that despite being a nitrate enrichment,
nitrate was not only not required for growth, but was inhibitory.
After 6 weeks HPLC analysis indicated thai dechI or I nation occurred
In all media containing 3-CI-benzoate, except 3-CI-benzoate + KNC>3 and
3-CI-benzoate only. A concomitant increase in the benzoate concentration
was observed in those media showing dechI or I nation. Thus, despite a lack
of significant growth, dech I or I nation occurred In the medium containing
H2 + 3-CI-benzoate, but lacking yeast extract. These data suggest that
reducing equivalents for reductive dehalogenatlon are obtained from the
yeast extract component, and only when reducing equivalents in the form
of H2 are provided will reductive dehalogenation occur without yeast
extract or some other suitable compound. These data also Indicate that
nitrate is inhibitory to dechI or I nation, as well as growth.
Microscopic examination of those cultures containing hydrogen
indicated gram negative rods of varying shape and lengths were
predominant. Other cultures contained the small coccobacl I lus as the
predominant cell type. Few cells were observed in the cultures showing
little or no growth. In the defined H2 + 3-CI-benzoate culture which
dechlorinated 3-CI-benzoate, but showed little growth, a mixture of the
gram negative coccobacil lus and large rod were seen. In this case the
large rod comprised approximately 50? of the population, which Is a much
higher percentage than the original dechlorinating consortium.
ISOLATED STRAINS PRESENT IN THE DECHLORINATING CONSORTIUM
Since dehalogenatlon did not occur in the MPN cultures containing
the coccobaci I I us, but did occur In the lower dilutions containing the
large gram negative rod, it is likely that the rod, and not the
coccobaclI I us, Is the dechlorinating organism. Since a large rod was
present in greater numbers in the defined H2 plus 3-CI-benzoate medium,
in which dehaiogenatlon occurred, we attempted a direct isolation of the
dechlorinating organism from this culture. The culture was streaked onto
3-CI-benzoate plates plus yeast extract and was incubated with and
without H2. After two weeks, 102 colonies were picked Into 3-CI-benzoate
plus yeast extract broth. Cultures were Incubated for one week and
cellular morphology and dehalogenaticn were determined. The rrsjorMy of
cultures contained the coccobacJ11 us previously described, but five other
419
-------
morphological types were isolated. These included two
coccobacillus which differed slightly from the dominant cell
type. Strain BG95 (Figure 9h) was more spherical and strain BG?Q
(Figure 2c) did not form pairs, while chains were only
periodically observed. Three gram negative rods were isolated,
including: strain BG?, a medium rod with pointed ends (Figure 2d);
strain BG49, a medium rod with round ends (Figure 9e); and a
I one, thin rod similar -to ]»ig_th .gnosp 1 r U | urn hun^ate I. The presence
of Hp had little effect on the type of cell isoleted. The
3-CI-benzoete concentration in cultures of each morphological type
was determined weekly for one month, but dechlorIneticn was not
observed. A second attempt at isolation of a dech I or I natIng
organism resulted in the Isolation of two rnoro types of gran
negative rods, including a medium red (strain FG131) which wins
usually found In pairs (Figure 2f) and a long oval rod (strain
BG170) which tended to divide unevenly (Figure 2g). After one week
3-CI -benzoate had not been dech lor inated by either of these
Isolates. It Is possible that dech I or I nation requires the presence
of one or more of these Isolates due to some type of cross-feeding
and may account for the lack of dechlor I nation in pure culture.
This coulc be determined by growing these organisms In various
combinations In 3-CI-benzoate medium.
Since this time, we have received the dechlorinating culture
DCB-1 in pure culture from Dr. J. M. Tiedje (21a). The DCB-1
culture has been shown to also dehalogenatf: 3-lodcbenzoate and is
thought to be related to the genus JD.e^lfpvlbr |p. The pure culture
has been propagated and shown to dehalogenate 3-chlorobenzoate.
Further studies will be directed at elucidating the genetic basis
for the dechlor fnatlng activity. A rapid screening procedure for
the detection of dehalogenatlon and methods for extracting pi asm Id
and chromosomal DNA will be developed.
Extraction of DNA from the DCB-1 strain will be based on
methods used by other investigators working with OesuJfOVlbrJo
strains. These organisms have proven resistant to milder cell
disruption methods Inducing lysosome, SDS and protease treatments.
The method currently producing the best results utilizes treatment
with RNAse A followed by disruption In a French Press and
phenol :chloroform extraction. The extracted chromosomal DNA will
be sheared, size fractionated and then cloned Into an E. col I host
organism in order to establish a clone bank of the entire
chromosome. Any plasmids isolated will be analyzed by treatment
with restriction endonucl eases and cloned into the same £j. £oj_i
host. The vector pDP1 to be used in the cloning experiments Is a
hybrid constructed from pDG5 and the gaptemldeg clindamycln
resistance pi asm Id pCP1. This vector replicates In both E, col I
and Bacjgfjpjjdes fragIIes and contains the RK2 or! T sequence. With
the helper plasmid, pRK231, fragments cloned and banked Into £*.
c_o_l_L can be relntroduced Into the original host or other
appropriate anaerobe hosts, such as gacterpi des to check for
expression of the dehalogenatlon activity.
420
-------
A rapid screening method for the detection of dehalogenation
Is essential for the detection of recomblnant clones expressing the
dehalogenatlng trait. A method Is currently being developed that
is based on the release of Iodide that accompanies dehalogenation
of 3-todobenzoate.
421
-------
LITERATURE CITED
1. Edwards, C. A. 1973. Persistent Pesticides lr the Environment.
CRC Press, Cleveland, Ohio.
2. Chapman, P. 1978. MIcrobial degradation of halogenated compounds.
Blochem. Soc. Trans. 4:463-466.
3. Schneider, M. J. 1979. Persistent Poisons. New York Acad. Scl., New
York.
4. DIGeronimo, M. J., M. NIklado and M. Alexander. 1979. Utilization
of chlorobenzoates by mlcrobleil populations In sewage. Appl.
Environ. MIcroblol. 37:619-625.
5. Bouwer, E. J., B. E. Rittman, and P. L. McCarty. 1981. Anaerobic
degradation of halogenated 1- and 2-carbon organic compounds.
Environ. Scl. TechnoI. 15:596-599.
6. Horvath, R. S. and M. Alexander. 1970. Cometabolism of
jn-chlorobenzoate by an ^JJirc&aclej:. APP'« Wcrobiol. 20:254-258.
7a. Evans, W. C., B. S. W. Smith, H. N. Fern ley end J. I. Davies. 1971.
Bacterial metabolism of 2,A DlcMGrcpherioxycicrlaif!. Biochen. J.
122:543-551.
7b. Evans, W. C., B. S. W. Smith, P. Moss and H. M. Fernley. 1971.
Bacterial metabolism cf 4-ch I orophenoxyacetate. Blochem J.
122:509-517.
8. Gaunt, J. K. and W. C. Evans. 1971. Metabolism of
4-chloro-2-methyl phenoxyaceate by a soil pseudononsd. Bfochem. J.
122:519-526.
9. Ahmed, M. and D. D. Focht. 1973. Degradation of polychlorlnated
biphenyls by two species of Achronobacter. Cand. J. Microbiol.
19:47-52.
10. Hartmann, J., W. Reineke and H.-J. Knackmuss. 1979. Metabolism of
3-chloro-, 4-chloro-, and 3,5-dichlorobenzoate by a pseudomonad.
Appl. Environ. Microbiol. 37:421-428.
11. Reineke, W. and H.-J. Knackmuss. 1980. Hybrid pathway for
chlorobenzoate metabolism in Esauifomojias. sp. B13 derivatives. J.
Bacteriol. 142:467-473.
422
-------
12. Horowitz, A., D. R. Shelton, C. P. Cornell, and J. M. Tledje. 1982.
Anaerobic degradation of aromatic compounds In sediments and
digested sludge. Dev. Ind. MIcrobiol. 23:435-444.
13. Sufllta, J. M., A. Horowitz. D. R. Shelton, and J. M. Tledje. 1982.
DehaiogenaTion: a novel palnway for the anaerobic blodegraoation of
haloaromatlc compounds. Science 216:1115-1116.
14. Horowitz, A., J. M. Suflita, and J. M. Tledje. 1983. Reductive
dehal ogenatlons of halobenzoates by anaerobic lake sediment
microorganisms. Appl. Environ. MIcrobiol. 45:1459-1465.
15. Ide, A., Y. Nik!, F. Sakamoto, I. Watanabe, and H. Watanabe. 1972.
Decomposition of penatach!orophenol in paddy soil. Agrlc. BIol.
Chem. 36:1937-1944.
16. Murthy, N. B. K., D. D. Kaufman, and G. F. Fries. 1979. Degradation
of pentachlorophencl (PCP) In aerobic and anaerobic soil. J.
Environ. Scl. Health Part B Pestic. Food Contam. Agric. Wastes
14:1-14.
17. Boyd, S. A., D. R. Shelton, D. Berry and J. M. Tiedje. 1983.
Anaerobic degradation of phenolic compounds In digested sludge.
Appl. Environ. MIcrobiol. 46:50-54.
18. Boyd, S. A. and D. R. Shelton. 1984. Anaerobic degradation of
chlorophenols In fresh and acclimated sludge. Appl. Environ.
MIcrobiol. 47:272-277.
19. Bouwer, E. J. and P. L. McCarty. 1983a. Transformations of 1- and
2-carbon halogenated aliphatic organic compounds under methanogenlc
conditions. Appl. Environ. MIcrobiol. 45:1266-1294.
20. Bouwer, E. J. and P. L. McCarty. 19835. Transformations of
halogenated organic compounds under denitrIfication conditions.
Appl. Environ. Microbiol. 45:1295-1299.
21a. Shelton, D. R. and J. M. Tledje. 1984. Isolation and partial
characterization of bacteria In an anaerobic consortium that
mineralizes 3-CI-benzo!c acid. Appl. Environ. MIcrobiol.
48:840-848.
21b. Hungate, R. E. 1950. The anaerobic, mesophilic cellulotlc
bacteria. Bacterlol. Rev. 14:1-49.
22. Bryant, M. P. 1972. Commentary on the Hungate technique for the
culture of anaerobic bacteria. Am. J. Clln. Nutr. 25:1324-1328.
23. Batch, W. E. and R. S. Wolfe. 1976. New approach to the cultivation
of methanogenlc bacteria: 2-mercaptoethanesuIfonic acid
(HS-CoM)-dependent growth of Methanobacter1 urn rum Inantlum In a
pressurized atmosphere. Appl. Environ. MIcrobiol. 32:781-791.
423
-------
24. Leedle,J., A. ZJegler and R. B. Hespell. 1980. Differential
carbohydrate media and anaerobic replica plating techniques In
delineating carbohydrate - utilizing sub-groups in rumen bacterial
populations. J. Appli. Environ. Microbiol. 39:709-719.
25. Rodina, A. G. 1972. Methods in Aquatic Microbiology, p. 177-180.
University Park Press, Baltimore, MD.
424
-------
TABLE 1. DEVELOPMENT OF 3-C1-BENZOATE DEGRADING CONSORTIUM
UNDER VARIOUS ENRICHMENT CONDITIONS
Enrichment Type Time (Weeks)1
3-CI-benzoate
methanogenic
sutfate
nitrate
fumarate
CC
4-CI-benzoate
methanogenic
sulfate
nitrate
fumarate
no H2
10-232
>233
6-10
10-23
>23
>29
>29
>29
>29
+H2
10-23
>23
10-23
10-23
>23
>29
>29
>29
>29
1) Weeks of Incubation Before Degradation Observed
2) Degradation not Observed at 10 Weeks, but Apparent at 23 Weeks
3) Degradation not Observed at Last Sampling
425
-------
TABLE 2. GROWTH AND DECHLORINATION OF 3-C1-BENZOATE BY
DECHLORINATING CONSORTIUM UNDER VARIOUS CONDITIONS
Additions1 Growth2 3-CI-benzoate Benzoate
KNO3
H2/KNO3
YE/KNO3
H2/KNO3/YE
Lactate/KNO3
Lactate/KNO3/YE
Lactate/YE
3CB/YE
H2/3CB/YE
H2/YE
YE
3CB/YE/KN03
3CB/KNO3
H2/3CB
3CB
0.05
0.05
0.22
0.32
0.08
0.05
0.07
0.24
0.47
0.42
0.22
0.32
0.05
0.06
0.05
—
—
—
—
—
—
—
147.3
0
—
—
384.4
815.9
0
757.5
—
—
—
—
—
—
—
630.5
752.0
—
—
422.6
65.8
748.0
58.2
1) Basal Medium Without Yeast Extract Plus the Additions Indicated (See Text)
2) Maximum Optical Density (600 nm) After Two Weeks
426
-------
• 1st refeeding
o 4th refeeding
FIGURE l:a. DEGRADATION OF 3-Cl-BENZOATE IN COLUMBUS
SEWAGE BY SGB/NC^j ENRICHMENT.
1000
• Basal
oS04
*NO3
I
I
I
FIGURE l:b.
120 240 360 480 600
Hours
DEGRADATION OF 3-C1-BENZQATE IN BASAL, SULFATE
AND NITRATE MEDIUM INOCUBATED WITH 10 PERCENT
COLUMBUS SEWAGE ENRICHMENT THAT DEGRADES
3-Cl-BENZOATE.
427
-------
3-CI-Benzoate (•) or Benzoate
s
INS
CO
8 I
O 3
*TJ
• to
6
8
3-CI-Benzoate
o
ox*
01
3
en
3<
-------
ro
- i
T ^
• -V
FIGURE 2:a. STRAIN BG19.
FIGURE 2:b. STRAIN BG95.
-------
•£>:£
•TCTSQ
V
\
— I
NIECES -p:2
\
rawns
aaosu
o
CO
•• - • ' I '-"l^
* »i •• • _ . •'*.*
t *- • . <*•"»'.«
•• •»•>*••»>•
* . • •
*•
*• , *
- • * .*,
'' "
( . \
»•
-------
ASSESSMENT OF BIOAUGMENTATION TECHNOLOGY AND EVALUATION
STUDIES ON BiOAUGMENTATION PRODUCTS
by
Henry H. Tabak
Wastewater Research Division
Water Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
431
-------
ASSESSMENT OF BIOAUGMENTATION TECHNOLOGY AND EVALUATION
STUDIES ON BIOAUGMENTATION PRODUCTS
by: Henry H. Tabak
Wastewater Research Division
Water Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
ABSTRACT
An overview of bioaugmentation technology and its application to
biological wastewater treatment through the use of biocatalytic products is
presented. The report defines and characterizes bioaugmentation, discusses
the principles of biological augmentation, describes the methodology for
the production of biocatalytic microbial and enzymatic additives through
selective adaptation and mutation and classifies the varied groups of bio-
augmentation products.
Manufacturers' assessment of bioaugmentation possibilities and the
range of treatment benefits through the use of their biocatalytic products
is counterweighed by a discussion of the shortcomings of bioaugmentation
technology as revealed by laboratory and field scale research studies.
Finally, the report discusses the ongoing evaluation studies funded by
the U.S. Environmental Protection Agency (EPA) using developed label verifi-
cation and performance (efficacy) testing protocols. The project goals are:
to place the use and characterization of these types of products on a
scientific basis; to arrive at consensus standards for label verification/
package contents and performance of these products and to improve the credi-
bility of bioaugmentation technology through the use of standard testing
protocols for evaluation of the biocatalytic products.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved
for presentation and publication.
432
-------
INTRODUCTION
Over the past twenty years there has been an increased interest in the
use of nonstructural approach for improvement of effluent quality in munici-
pal and industrial wastewater treatment without a major upgrade in the
treatment process. Bioaugmentation approaches, such as addition of enzymes
to improve flocculation in a physical/chemical treatment process or bacterial
augmentation of an activated sludge process to decrease the level of pollu-
tants in the treated discharge may provide such a nonstructural approach.
During the same period of time, increased reference has appeared in the
literature on the ability of specialized bacterial cultures to produce
improvements in biological waste treatment. These improvements include:
decreased sludge production; reduced foam in the aeration process; color
reduction; suppressed filamentous organisms in activated sludge; improved
response from a peak load; and an overall improvement in effluent quality.
Various types of enzymes, dried bacterial cultures, by-products of
bacterial fermentation products, biocatalytic additives, and combination of
microbiological cells and their fermentation products have been developed.
Suggested applications of these products include solubilization of grease;
breaking up the scum and dissolving the fat in anaerobic digesters; improv-
ing the performance of septic tanks and leaching fields; increasing the
biochemical oxygen demand (BOD) removal efficiency of primary sedimentation,
activated sludge and trickling filters; controlling foaming in aeration
tanks; unclogging sewers, pipes and trickling filter media; increasing
performance of stabilization ponds; and deodorizing sewers, septic tanks and
small treatment systems. In addition, the added bacterial cultures are said
to improve performance of overloaded plants by increasing kinetic rates.
The beneficial effects on the performance of biological waste treatment
processes associated with the use of such a technique has not always been
clearly demonstrated. Reports on application of bacterial additives have
often overstated the advantages or presented selected examples of success.
In addition, failure to use a full scale control when evaluating these
products makes documentation of their advantages difficult. Under the
evaluation procedures, careful plant operation during the testing periods
could cause the observed improvements in efficiency.
It is the purpose of this report to review the bioaugmentation
industry's applications and uses of bioaugmentation products and their
technological approach in wastewater treatment; to present the manufacturers
assessment of bioaugmentation possibilities and the range of treatment
benefits achieved by the technology; to present pertinent research results
and critiques by critics of the bioaugmentation approach; and finally, to
describe the ongoing evaluation studies on bioaugmentation products funded
by the U.S. EPA at the National Sanitation Foundation.
433
-------
BIOAUGMENTATION TECHNOLOGY AND ITS APPLICATION
As practiced today, bioaugmentation involves the addition of selected
microorganisms to a naturally occuring population in a wastewater treatment
system to increase the biological activity of the system. The bioaugmenta-
tion technologists state that bioaugmentation is very cost competitive with
other upgrading approaches in wastewater treatment, and by actually degrading
pollutants (usually to carbon dioxide and water), solves the pollution
problem instead of transferring pollutants to another location. A variety
of bacterial preparations (including some with added nutrients or enzymes)
are presently available for bioaugmentation in treatment of municipal,
industrial, and agricultural wastes and wastewaters. The addition of
specific cultures of microorganisms to a process, known as bioaugmentation,
is a commonly used procedure in many industries including brewing, cheese-
making, wineries, dairy products, pharmaceutical and fermentation, to
mention a few. The majority of fields using biological processes rely upon
this technique. Bioaugmentation products suppliers indicate that wastewater
treatment constitutes the largest user of biological processes.
THE BIOAUGMENTATION CONCEPT
The concept of bioaugmentation has been variously defined. Flow
Laboratories defines bioaugmentation as "the addition of specific organisms
Into an environment for specific purposes....to assure consistency of quality
by eliminating the opportunity of unfavorable, happen-stance organisms to
cause undesirable effects." This definition is thus applicable to the use of
bacterial culture in the beer, wine, dairy, and pharmaceutical industries as
well as to their use in wastewater treatment. Dominance of the introduced
bacteria is the implied objective. In contrast, Reliance Books defines
biological augmentation as process which "augments the existing bacterial
population with bacteria that are capable of higher rates of organic
oxidation or that are capable of degrading organic compounds that have
previously been nondegradable." This definition is followed by the quali-
fying statement that "the object is not to replace the existing bacteria
but to supplement them for improved efficiency." The differences between
these two definitions is probably more of a semantic than of a scientific
nature as the methods and objectives of these two companies are essentially
identical.
The principle of ubiquity (Figure 1) states "that organisms necessary for
satisfactory wastewater treatment are available in any soil or the wastewater
stream itself, and that the organisms capable of using a specific organic
substrate in the waste, are selected out within a total population through
natural selection processes." In contrast to the ubiquity principle stressing
the natural selection process, proponents of the bioaugmentation approach
stress that organisms are normally not selective toward specific organic
substrates but rather handle an array of various substrates within their
434
-------
THE UBIQUITY PRINCIPLE
"... oil types of bacteria are available
at all times everywhere..."
hence:
Natural population selection mechanism
will always result in the right biological
culture for treatment of a given waste.
Figure 1. The Ubiquity Principle.
metabolic capabilities. The basis of biological supplementation (bioaugmen-
tation) is the need for added bacteria, capable of handling specific problem
organic substrates, to achieve an artificial dominance in the microbial
population that will improve treatment performance.
With programmed bacterial supplementations, sufficient quantities of
selected bacteria are added to the system initially, and on a repetitive
basis, to produce a competitive number of the supplemented organisms. In
this way, the augmentation changes the natural population distribution and
the maintenance dose maintains the necessary numerical advantage of the
desirable organisms. Maintenance doses are used on a variable basis,
depending upon the nature of the system. The proponents claim that bioaug-
mentation thus offers an effective means of controlling the nature of the
biomass.
In the view of bioaugmentation technologists, the various types of
bacteria incorporated in the formulations differ from those which find a
natural dominance through the natural selection processes within a typical
treatment system. One is not simply adding numbers of bacteria but changing
the quality and characteristics of the existing biomass. The effective
formulations are based on selected species and strains which have an
enhanced ability to breakdown problem organic substances compared to
bacteria that achieve natural dominance within a given system. Sustenance
treatments, however, are not normally required in instances where competi-
tive organisms are not routinely entering the system or in chemical waste
disposal processes in which hard to handle chemical wastes are continuously
treated.
435
-------
Bioaugmentation researchers present cases which strongly support the
conclusion that, although microbial ubiquity and natural population selection
and adaptation are effective in problem situations of a continuous nature,
most biological wastewater treatment systems regularly face intermittent
problem situations to which natural adaptive processes are unable to respond
in a timely basis. According to them, blends of specialized preselected,
adapted, mutant microbiota can reduce the response time of such biological
systems and thus improve their performance and stability. They maintain
that a combination of effective microbial applications technology, appro-
priate system design and informed systems operations and maintenance (biomass
engineering) have enormous potential for yielding cost-effective waste
treatment for industry.
METHODOLOGY FOR THE PRODUCTION OF BIOCATALYTIC MICROBIAL ADDITIVES
In general, bioaugmentation organisms are mutated by chemical in situ
mutation, and not by recombinant DNA techniques. With the latter technique,
the organisms are lysed, the genetic feature of interest is isolated, and
then it is reintroduced into living organisms. With chemical in situ
mutation, most of the organisms are likely to be rare mutants. That is, a
capability that was already present in living organisms (often Pseudomonas
aeruginosa) is enhanced by mutation as much as a thousandfold. Unlike the
genetic engineering method which involves gene splicing and plasmid transfer
technology, conventional mutational methods involving strain selection and
chemical mutagenesis are being used in the production of randomly achieved
microorganisms specialized for a particular task. In a treatment situation,
the organisms simply increase the degradation rates in the system.
Although the microbes that the bioaugmentation industries market for
specialized waste treatment are, in a sense, genetically engineered, they
are produced by "natural" methods. In essence, the process by which
evolution devises organisms specially adapted to a given environment is
accelerated in the production of the bioaugmentation additives.
First, product developers collect organisms, often from special sites
where natural selection has already favored microbes adapted to unusual
conditions. The organisms are then grown in the laboratory and are sub-
jected to enrichment culture techniques in a medium containing the pollutant
they are supposed to degrade or under condition that enhanced the desired
treatment effect. This process selects the organisms that accomodate best
to the medium's chemistry and available nutrients, i.e., or to appropriate
environment conditions.
Scientists speed up the organisms' adaptation to the specialized en-
vironment or substrate by increasing their mutation rates using tools such
as radiation and chemical mutagens. The object of the mutagenesis, is not
necessarily to grossly alter the organisms' natural metabolism but to
accelerate the rates of enzymatic activity. The genetic changes induced by
the mutagens result in some microbes that are better at degrading a par-
ticular pollutant or producing a desired treatment effect than their
predecessors.
436
-------
After this trial-and-error process, the best adapted mutants are grown
in large quantities. The companies may market their microbes as dried
powders or liquids. The dried products usually contain mixtures of organisms
adapted to a particular waste treatment task, along with additives such as
wetting agents and emulsifiers to aid in dispersion and nutrients. To
activate the microbes, warm water is added to the contents and the mixture
is stirred. Liquid products are suspension of bacteria, their metabolic
products, enzymes and nutrients.
Selective Adaptation and Mutation
Researchers producing the bioaugmentation additives, maintain that
during the process of selective, adaptation and mutation, these microbes are
"evolving" under conditions other than those found in the "real world".
Perhaps most significantly, they are not subjected during this process to
the strongest competitive selective pressures of the "real world". As a
result, although enhanced abilities to degrade given substrates can be
obtained, this is normally accomplished at the cost of those qualities
which provided the original microbe with "real world" survivability under
such pressures. When the new strains are reintroduced to the competitive
environment of the natural ecosystem (particularly if their favored
substrates are either absent, only intermittently present or present in low
concentration), their survivability may be marginal. Thus their longevity
under such conditions may be significantly reduced, although not terminated.
This means that generally regular reinoculation is required to maintain
them in a given "natural" environment. It also means that they tend to be
"selected out" of the natural ecosystem when conditions favoring their
survival are eliminated (e.g. the removal of a preferred substrate).
CHARACTERIZATION OF BIOAUGMENTATION PRODUCTS
Currently there are approximately between 60 to 70 industrial concerns
and suppliers in the United States, manufacturing and representing the
competitive bioaugmentation products, consisting of either microorganism
formulations, enzyme preparations, or combinations of bacterial cultures
and enzymes. Some of the larger industries include Polybac, a subsidiary
of Cytox Corporation; Sybron Biochemical, a division of Sybron Corporation;
Environmental Cultures Division of Flow Laboratories Inc.; Miles Labora-
tories, Inc.; Bioscience Management, Inc; General Environmental Sciences
Corporation; Industrial Microgenics, Ltd; Reliance Brooks; Jet, Inc.;
Solmar Corporation; Materials Bio-Science Corporation; Worne Biochemicals,
and J.T. Baker (Environmental Protection).
Polybac supplies engineered microbes for commercial waste treatment to
food processors, chemical manufacturers, petroleum refiners, petrochemical
plants and for hazardous waste cleanup. Sybron sells bacterial cultures to
industries ranging from food processing to steel coking, and boasts that one
of its pseudomonads was the first bacterial life form to be patented. The
company maintains that its organisms include some that will degrade Arochlor
1260, one of the most highly chlorinated of the polychlorinated biphenyls
(PCBs), generally considered non-biodegradable. Enviroflow specializes in
municipal water and sewage treatment.
437
-------
General Environmental Sciences Corp. and Jet, Inc. market liquid
suspensions of bacterial cultures. One of these formulations, Liquid Live
Microorganisms (LLMO), is produced by growing bacterial cultures to maturity,
and chemically inducing a dormant state, presumably by addition of a growth
or metabolic inhibitor. These cultures supposedly can be held in this sus-
pended state for an indefinite period of time, and full population recovery
is expected when the culture is diluted and suitable nutrients become
available. Sybron markets both wet and dry cultures whereas the other
companies offer dry cultures only. Dry cultures require strictly dry con-
ditions for storage and cannot be frozen. Maximum shelf life is 1 - 2 years.
Industrial Microgenics offers immobilized cell preparations in which the
microorganisms are supported by semi-solid or granular substrates.
Many industries, such as Reliance Brooks, Sybron, Industrial Micro-
genics, and Polybac offer "mutant" bacterial cultures, but details regarding
the parent strains and the genetic engineering techniques used to induce
and select mutations are apparently of a proprietary nature and are not
made available. However, Industrial Microgenics does present a brief
description of a treatment process involving 5-bromouracil and UV-radiation,
which is used to induce mutations in bacteria isolated and selected for
tolerance of high levels (approx. 1000 ppm) of various compounds (e.g.
phenols). Unlike the above industries, Enviroflow uses organisms selected
from the environment without modifying them via mutation.
In all cases the product literature warns against exposure of bacterial
cultures to toxic levels of heavy metals, especially chromium. Polybac
specifically recommends that the hexavalent chromium concentration not
exceed 2 ppm in the treatment system. Bacterial cultures or similar
products for metal waste treatment or metal recovery evidently have not
been developed or marketed to date. Most presently available formulations
are designed for biological degradation of organic wastes with a few formu-
lations for treatment of cyanide or cyanate waste, and control of T^S
emissions.
Bacterial cultures for bioaugmentation of wastewater treatment are
being used when one or more of the following problems (Figure 2) occur:
foul odors persist; temperature, or hydraulic or chemical loading changes
drastically; grease and oil accumulate, resulting in blockage or formation
of surface scum and reduction of oxygen transfer rates; the waste is a
chemical inhibitor of biological growth or metabolism; the nutrients needed
for bacterial growth are deficient; BOD, COD, suspended solids, color, or
nitrogen levels exceed effluent standards; algae or duckweed growth occurs;
build-up of solids or excessive foaming occurs; corrosion of metals occurs
due to the action of Thiobacillus organisms in the presence of H2S; float-
ing sludge, and high effluent solids.
Only a few companies provide the genera of bacteria present in their
culture. These genera of bacteria used for bioaugmentation are listed in
Table 1. In no instance, are species names or other taxanomic information
provided. In addition to the bacteria listed, several fungi are cited in-
cluding Rhizopus, Aspergillus, Candida, Myrothecium and Tricho-Cellulomonas.
438
-------
COMMON SIGNS OF BIOLOGICAL
SYSTEM STRESS
-FOAMING
-FLOATING SLUDGE
-BOD and COD BREAKTHROUGH
-HIGHLY TURBID EFFLUENT
-LACK OF HIGHER LIFE FORMS
- INCREASED OR DECREASED RESPIRATION RATES
-H&N EFFUIENT SOLIDS
Figure 2. Common signs of biological system stress.
Bacillus is probably the most important bacterium present in these cultures
because of the diversity of substrates it can utilize, and because of its
ability to form heat-resistant spores.
General Environmental Science Corp. provides a number of documents
addressing the potential human health hazards of bioaugmentation. According
to researchers at Case Western Reserve University, experiments failed to show
any toxic effects on mice injected with LLMO. Inspection by the USDA is
required for certification of the absence of Salmonellae and other pathogenic
microorganisms. The USEPA presently does not require registration of LLMO and
for that matter any other bioaugmentation product on the market for wastewater
treatment.
THE CONTROVERSY ON BIOAUGMENTATION APPLICATIONS
I. Manufacturer's Overview of the Treatment Benefits
Achieved by Bioaugmentation Technology
According to L.T. Davis (1) a range of treatment benefits have been
achieved with bioaugmentation technology: generating good bioactivity,
especially in very toxic systems; greatly increasing the rate of degrada-
tion, and sometime precluding the need to build a bigger treatment system;
biodegrading compounds that had previously passed through the system;
maintaining efficient bioactivity at low temperatures to yield significant
savings on heat costs; sustaining efficient bioactivity with lower aeration
costs by means of more efficient oxygen utilization; withstanding variable
waste streams and shock loadings, thereby promoting greater system stability,
recovering quickly from a "kill" situation; reducing foam; and enhancing
sludge settleability.
439
-------
TABLE 1. GENERA OF BACTERIA USED FOR BIOAUGMENTATION
Genus Gram Stain
Act Inomycetes +
Aerobacter
Arthrobacter +
Bacillus +
Cel (ulomonas +
Desulfovlbr lo
MycobacterFum +
Nltrobacter
Nocardla +
Pseudomonas
Rhodopseudomona s *
Oxygen Req u I rement
m I croaeroto 1 erant
facultative anaerobe
strict aerobe
strict aerobe or
facultative anaerobe
strict aerobe
strict anaerobe
strict aerobe
strict aerobe
(ammonia oxldizer)
strict aerobe
strict aerobe
photosythetlc
Spore
Format ion
no
no
no
yes
no
no
no
no
no
no
no
Shape
mycel ia 1
rod
Irregular
rod
irregu ! ar
curved rod
mycel ia 1
pear-shaped
nyce 1 ia 1
straight or
curved rod
rod
* Gram stain Is used to distinguish non-photosynthetic bacteria only.
-------
Davis maintains that bioaugmentation treatment has yielded impressive
results in a number of wastewater applications in chemical and pharmaceuti-
cal industry. Biocatalysts are added to: withstand organic overloading;
maintain efficiency at lower temperatures in winter; and successfully
compete against filamentous growth. Heavy initial dosage is needed to
establish a dominant place in natural population, a maintenance dosage
is also required to compete with naturally occuring organisms and maintain
a population in the system, and to compensate for back mutation, whereby
some of the organisms' potency is gradually lost. Bioaugmentation
researchers have isolated from soils contaminated with Arochlor 1260, and
mutated a group of organisms that exhibit a high degradation rate of this
PCB. Treatment in activated systems augmented with these microbiota
reduced 100 to 400 ppm PCBs to undetected levels. Mutated cultures have
also been produced that successfully degrade the analogs of dioxin
(2,4,7,8 tetrachlorodibenzoparadioxin).
Grubbs (2) describes the use of bioaugmentation products in: waste
treatment plants, enhancing BOD5 removal and COD removal rates; in aerobic
and anaerobic digesters, improving digester performance; pretreatment of
industrial wastes; and hazardous waste applications. According to Grubbs (3)
successful environmental applications of bioaugmentation technology were
made in domestic sewage situations, through enhancement of BOD and COD
removals, control of grease, improved digestion of sludge, reduction of
odors and control of H2S emissions; industrial wastewater pretreatment,
by improving BOD and COD removals; and treatment of petroleum based wastes
and hazardous wastes. Report has also been made by Grubbs (4) on the use
of bioaugmentation products in enhancing BOD5 removal rates in final
effluent for activated sludge wastewater treatment for potato wastes, corn
product wastes, and in increasing % BOD removal for dairy wastes across
aerated lagoon systems.
Bacterial supplementation has been shown (4) to: improve 8005 removals;
increase sludge settleability; lower sludge volumes; eliminate grease mats,
control malodors; reduce H2S corrosion; improve digestion of solids; improve
digester operations; provide much quicker recovery from upsets due to shock
loadings or mechanical failures; prevent malodors from lagoon inversions;
clean grease in collection systems; restore percolation of fields, perco-
lation ponds, etc., which are plugged from organic matter, and give more
predictable results. Factors limiting bioaugmentation can be enumerated as
follows: poor engineering design and operational practices; very low dis-
solved oxygen levels, which will stimulate filamentous growth (Sphaerotilus
species); caustics, chlorine, bactericides; and variations in hydraulic
organic loadings.
Numerous articles (2, 4-16) have described the successful use of cultures
to control grease within sewage collection systems to reduce sewer-line clog-
ging and to improve sewer-line systems treatment. Reductions of odors and
control of hydrogen sulfide emissions (2,4,5,9,10, 11, 17-20) have also been
widely reported. Improvements in BOD5 and COD removals (2,4-7,9,10,21-26)
through the use of bacterial supplements in industrial wastewater treatment
applications have been reported in the literature.
441
-------
Worne (25) reported on the activity of adapted and mutant microorganisms
supplemented in the biological treatment of industrial wastes and on their
use in the degradation of specific organic compounds. The use of adapted
and mutated microbiota offers advantage of immediate activity with pre-
computed levels of biochemical activity for removal of various toxicants in
contrast to the time required for adaptation in nature of wild strains.
The concentration of these biocatalytic additives assures the formation of
significantly higher concentration of enzymes, and establishes higher rates
of efficiency and more rapid velocities of degradation. Biodegradation of
aryl halides, halophenols, aliphatic amines and aryl amines by parent and
adapted/mutant microbiota is shown in Tables 2, 3, 4 and 5, respectively.
Biodegradation of ABS-LAS surfactants by parent and adapted/mutant micro-
biota is expressed graphically in Figures 3, 4 and 5. Table 6 provides
biodegradability data on various inorganic and organic cyanides.
The use of adapted mutant microbiota for the enhancement of refinery
effluent treatment (26-33) is well documented in the literature. The impact
of a mutant bacterial aid on the performance of an activated sludge process
was illustrated in a full scale application conducted on a refinery waste-
water by McDowell and Zitrides (32). Two identical activated sludge
systems operating in parallel provided control and trial units during a
three month investigation. The process receiving the mutant bacteria
provided a 32 percent improvement in effluent quality as measured by total
organic carbon (TOG). In addition, probability distributions of the treated
effluent quality indicated less variability in the trial unit. The authors
suggested that the mutant organisms accelerated the response of the acti-
vated sludge which resulted in a less variable effluent quality. The
mutant organisms increased activated sludge performance, resistance to
upsets and shock resistance, as illustrated in Figures 6,7 and 8,
respectively. A mathematical model using Monod kinetics was proposed to
describe the transient response time of the activated sludge. By varying
the model kinetic constants for maximum specific growth rate, U-max, and
the half velocity coefficient, Ks, the sensitivity of the model to bacterial
modification was illustrated in which a lowering of Ks and an increase in
y-max provided a faster response of the activated sludge to a shock load.
Although the authors proposed that the mutant microorganisms produced such
a result, the parallel activated sludge processes in this study were not
used to test the authors' hypothesis.
Tracy and Zitrides (31) discuss the enhancement of process kinetics
in refinery effluent treatment units containing Phenobac cultures. Whereas
the control units show a decreasing coefficient rate with increasing TOC,
suggesting inhibition of the biomass, the coefficients in the Phenobac
supplemented units, show no inhibitory effects and remain constant with
increasing TOC loadings, suggesting greater resistance of Phenobac supple-
mented cultures to substrate shocks. The comparison of kinetic constants
between treated and control units is illustrated in Table 7.
Enhancement of the treatment of petrochemical wastes by adapted mutant
microbiota has been reported in several articles (34-37). Use of freeze-
dried mutant microorganisms to improve the effluent from a pure oxygen
activated sludge process treating petrochemical wastewater has been studied
442
-------
TABLE 2. DEGRADATION OF ARYL HALIDES BY A MUTANT
PSEUDOMONAS SP. (a) 30°C
f ftiti nfiu nrf
^v VI 1 1 |^wU 1 'VI
Monochlorobenzene
o-Dichlorobenzene
m-Dichlorobenzene
p-Dichlorobenzene
I ,2.3-TrichIorobenzene
I J,4-Trichloroberuene
1 ,3,5-Trtchlorobenzene
1 ,2,3,4-Trichlorobenzene
1,2,4,5-Tetrachlorobenzene
Hexachlorobenzene
CY»nc
^•WIIW
200mg/l
200 mg/1
200 mg/1
200mg/l
200 mg/1
200 mg/1
200 mg/1
200 mg/1
200 mg/1
200 mg/1
Ring Disruption %
Parent
100
100
100
too
87
92
78
33
30
0 •
Mutant
100
100
100
100
100
100
100
74
80
0
Time in Hours
Parent
58
72
96
92
120
120
120
120
120
120
Mutant
14
26
28
25
43
46
' 50
120
120
120
TABLE 3. DEGRADATION OF HALOPHENOLS BY A MUTANT
PSEUDOMONAS (a) 30°C
Compound
Phenol
o-Chlorophenol
m-ChlorophenoI
p-Chlorophenol
2,4-Dichlorophenol
2,5-Dichloropheno
2,3,5-Trichlorophenol
2,4,6-Trichlorophenol
Pentachlorophenol
o-Bromophenol
m-Bromophenol
^-Bromophenol
2,4-DibromophenoI
2,5-Dibromophenol
2,4,6-Tribromopheool
^•___
Cone
500 mg/1
200 mg/1
200 mg/1
200 mg/1
200mg/I
200 mg/1
200 mg/1
200 mg/1
200 mg/1
200 mg/1
200 mg/1
200mg/t
200 mg/1
200mc/l
200 mg/1
Ring Disruption %
Parent
100
100
100
100
100
60
100
100
7
100
51
87
75
58
14
Mutant
100
100
100
100
100
100
100
100
26
100
100
100
100
100
92
Time in Hours
Parent
25
52
72
96
96
120
100
120
120
85
96
84
72
120
120
Mutant
8
26
5!
33
34
38
52
50
120
14
> 25
22
20
35
AA
42
443
-------
TABLE 4. DEGRADATION OF ALIPHATIC AMINES BY A
MUTANT AEROBACTER SP. (a) 30°C
Compound
Triethylamine
N-Propylamine
Di-N-Propylamine
Tri-N-Propylamine
N-Butylmanine
N-Amylamine
N-HexyUunine
N-Dodecylamine
N-Allylamine
Di-N-Allylamine
Tri-N-Allylamine
.
Cone
200mg/l
200mg/l
200mg/l
200mg/l
200mg/l
200mg/l
200mg/l
200mg/l
200mg/l
200mg/I
200mg/l
Degradation %
Parent
100
100
100
100
100
100
100
100
78
62
47
Mutant
100
100
100
- 100 .
100
100
100
100
100
100
100
Time in Hours
Parent
28
31
26
30
22
25
20
18
93
105
120
Mutant
II
9
12
10
7
9
10
5
13
17
22
TABLE 5. DEGRADATION OF ARYL AMINES BY A
MUTANT AEROBACTER SP. (a) 30°C
Compound
Aniline
e-Chioroaniline
m-Chloroaniline
p-Chioroaniline
2,4,6-Trichloroaniline
e-Toluidine
m-Toluidine
p^Totuidine
e-Anisidine
m-Anisidine
p-Anisidine
o-Dianisidine
Cone
500mg/l
SOOmg/l
SOOmg/l
SOOmgll
SOOmg/l
500mg/l
SOOmg/l
500mg/l
SOOmg/l
SOOmg/l
SOOmg/l
SOOmg/l
Ring Disruption %
Parent
100
100
100
100
82
100
100
100
92
80
86
78
Mutant
100
100
100
100
100
100
100
100
100
100
too
100
Time in Hours
Parent
54
60
68
59
120
64
62
' 48
120
120
120
120
Mutant
12
18
16
12
30
6
10
3
16
24
12
36
444
-------
A FMIIMIMIM ••M*»Tf»
O
d
o
o
en H
B S
CO
•H M
d cu
co A
00
r< (S
o o
O -H
!-i rH
O rH
O
H
Time in Days
Figure 3. Growth of adapted and mutated Pseudomonas on ABS-LAS
substrates in Worne media at 25°C.
cu
4-1
•H
iH
O
U
O
•H
Time in Minutes
Figure 4. Oxygen uptake by unadapted and mutant Pseudomonas sp.
in Worne media containing 0.01% solution of Sodium
Alkyl Benzene Sulfonate at 25°C.
445
-------
a,
c
c
o
a
-------
1000
500
•f 200
I
DC'
3
o 100
z
o
tc.
o
_ I I I I I I I I I I I I I I I I
o
50
20
10
I I
• EAST EFFLUENT (DAY 12-58)
• EAST EFFLUENT (DAY 59-102) } PHENOBAC
A WEST AFFLUENT (DAY 59-102)
• INFLUENT (DAY 12-102)
1
WEST EFFLUENT (DA^
(BEFORE PHENOBAC)
i I I * I I I I I I I T I I I I I \\l\\
0.01 0.1 0.5 2 5 10 20" 40 60 BO 90 98
PERCENT OF TIME LESS THAN THE INDICATED VALUE
99.99
Figure 6. Activated Sludge Performance With Additive
-------
280
240
poo
§ 160h
5
DC
O
120
p 80
i i i — I — i — i — i — i — i — i — i — i — i
i — i — r
WEST EFFLUENT
(CONTROL) V. xx
EAST EFFLUENT
(PHENOBAC)
38 , 40 42 44 46
J 1 « 1 ' 1 1
48
DAY
50 52 54 56 58
Figure 7. Additive Increases Resistance To Upsets
448
-------
8
400
360
320 -
280 -
240 -
u 200 -
3
g
j
o
160
120
80
f- EAST EFFLUENT
I .s (PHENOBAC)
^x ^_
\j> —
1 1 L 1 ,
^^£* ^**^*^*«*»^i^^« ^ _
WEST EFFLUENT '^'^
(PHENOBAC)
•WJ 1 1 1 1
^JB
irs>- ^
1 »
97 99 101
103 105
DAY
107 109 111
Figure 8. Increased shock resistance in treated units.
449
-------
TABLE 7. COMPARISON OF KINETIC CONSTANTS
Percentile TOC
10
50
90
Avenge
Treated unit
coefficient
32.6
53.8
42.6
43.0
Control unit
coefficient
12.6
6.5
3.9
7.7
EFFECT ON PROCESS KINETICS
Process kinetics for the treated and untreated cases
were defined to assess additive effects for conditions
other than those which prevailed during the test. In
evaluating the kinetics the following mathematical
model was used:
S>« Sp
I + KXQk
where: S0 and
Sg * influent and effluent substrate
concentration, mg/1
K * rate constant, 1/g—day
Qn «• hydraulic retention time, days.
X • MLSS concentration, g/1
450
-------
by Thibault and Tracy (37). Operation problems cited were solids-liquid
separation in secondary clarification, a deterioration in effluent quality
resulting from shock influent loads, and excessive adsorption of oil on the
biological floe. The test program was conducted in two phases. The initial
phase of the study was a three week trial of a parallel or side-by-side
comparative analysis of the dual train activated sludge process while the
second phase involved two consecutive fifty-day periods. The first fifty-
day period prior to bacterial inoculation was used as a comparison with the
performance during the second fifty-day segment.
Following one week of parallel examination, the trial portion of the
activated sludge process receiving the special bacteria exhibited a differ-
ence in effluent quality. During the subsequent two weeks, it experienced
a 21 percent improvement in effluent total oxygen demand (TOD) and a major
reduction in floating solids. This phase of the study was terminated after
3 weeks because the activated sludge was operating at a temperature of 43°C.
The second phase of bacterial inoculation study was conducted using
"before" and "after" performance data. Activated sludge effluent quality
for TOD, 8005 (both total and soluble) and TSS was monitored during a 50-
day period of bacterial augmentation. This performance period was then
compared with the preceeding 50-day period which served as the control.
Mean influent TOD quality to the activated sludge process during the control
and inoculation periods was approximately the same, however, less influent
variability occurred during the bacterial inoculation period. Comparison
of effluent quality for the two periods indicated the following improvement
during the inoculation period: TOD = 46 percent, 6005 = 73 percent; BOD5
(soluble) = 59 percent, and TSS = 38 percent.
The authors used a mathematical model to estimate the kinetics during
the control and inoculation periods. Their analysis indicated a greater
than 2.5 increase in reaction rate attributed to the use of the mutant
organisms during the inoculation period. Comparison of influent TOD,
effluent TOD, BOD, TSS data before and after Phenobac application is shown
in Figures 9, 10, 11 and 12, respectively. Improvement of treatment effi-
ciency in a side-by-side demonstration, and improved degradaton of tertiary
butanol with the use of Phenobac are illustrated respectively in Figures 13
and 14. Phenobac performance during acclimation period is shown in Figure 15.
Use of bioaugmentation products for improvement of waste removal relia-
bility in wastewater treatment systems through enhancement of population
dynamics and growth rate kinetics (23,32,38,39) has been described in the
literature. It is the contention of the bioaugmentation technologists, that
while it is true that the influent does influence the population dynamics
of a wastewater treatment system by the very nature of its available
nutrients and BOD strength, the selection process for microorganisms does
not necessarily lead to an optimal microflora for the best assimilation
rate of the pollutant loadings being applied. While "naturally" developed
microflora can provide an adequate biological population for many waste
streams, there are some waste streams beset with problems of bulking,
451
-------
-pi
en
10.000
5000
o 2000
UJ
a
fc
o
1000
500
200
100
II I T
l I l I i
r t
50 DAYS
AFTER PHENOBAC
50 DAYS
BEFORE PHENOBAC
PETROCHEMICAL WASTEWATER
UNOX PROCESS
I
I l
J I
I l
J I
J I
0.01 0.1 0.5 1 2 5 10 20 40 60 80 90 95 98 99
PERCENT OF TIME LESS THAN THE INDICATED VALUE
99.9 99.99
Figure 9. Comparison of Influent TOD Data Before and After Phenobac
Application Reveals Similar Distributions
-------
en
CO
1000
500
~ 200
o r
UJ
f. 100
O
50
20
10
TTT
i — i — i — i — i i i i i — i — i — i — r— n — rr
50 DAYS
BEFORE PHENOBAC
50 DAYS
AFTER PHENOBAC
PETROCHEMICAL WASTEWATER
UNOX PROCESS
J—I—LJ—I 1 I
0.01 0.1 0.5 1 2 5 10 20 40 60 80 90 95 98 99 99,9 99.99
PERCENT OF TIME LESS THAN THE INDICATED VALUE
Figure 10. Comparison of Effluent TOD Data Before and After Phenobac
Application Reveals Significant Improvement in Performance
-------
100
s 50
1
I 20
Ul
I
o
UJ
8
m
o
10
i i i i i i — i — i
50 DAYS
BEFORE PHENOBAC
i — i
T—TT
50 DAYS
AFTER PHENOBAC
PETROCHEMICAL WASTEWATER
UNOX PROCESS
J_
J L_J L
J L
J—I L
J L
'
0.01 0.1 0.5 1 2 5 10 20 40 60 80 90 95 98 99
PERCENT OF TIME LESS THAN THE INDICATED VALUE
99.9 99.99
Figure 11. Comparison of Effluent BOD Data Before and After Phenobac
Application Reveals Significant Improvement in Performance
-------
~ — m — n — i
8-aoo
<0 ISO
§
jg M
« TO
I"
S "
40
30
20
i — i
i — i — T-T-T
50 DAYS
BEFORE PHENOBAC
I I I 1 I I I
50 DAYS
AFTER PHENOBAC
PETROCHEMICAL WASTEWATER
UNOX PROCESS
PERCENT OF TIME I ESS THAN THE INDICATED VALUE
MS ft* Ml MM
Figure 12. Comparison of Effluent TSS Data Before and After Phenobac
Application Reveals Significant Improvement in Performance
455
-------
2 MO
o
I.
SOUTH UNIT EFFLUENT
(CONTROL)
I I
I I I I
I I I
NORTH UNIT EFFLUENT
(PHENOBAC)
PETROCHEMICAL WASTEWATEH
UNOX PROCESS
••I •• *l tt •» I I • <• M » «• M • It M MM
PERCENT OF TIME LESS THAN THE INDICATED VALUE
HI m» •• M*
Figure 13. Phenobac Improves Efficiency in Side-by-Side Demonstration
456
-------
'516
TBOH
(mo/I)
en
125
50 •
25
PHENOBAC*
APPLICATION
INITIATED
INFLUENT
PETROCHEMICAL WASTEWATER
UNOX PROCESS
-20 -15 -IO -5 O 5 K> 15 20 25 30 35 40 45 50 55 60
Figure 14. Phenobac Improves Degradation of Tertiary Butanol
-------
350
„ 300
1
3
373 373
I
bl
250
200
I 150
IU
2
> 100
9
>
IL
50
PHENOBAC APPLICATION
INITIATED
\
PETROCHEMICAL WASTEWATER
UNOX PROCESS
-2-1 1 2 3 4 5 67 6 9 10 11 12 13 14 15 16 17 18
DAY
Figure 15. Phenobac Performance During Acclimation Period
458
-------
deficiency in settling solids and an excessive loss of suspended solids in
the discharge. The naturally selected microbiota do not provide a reliable
waste removal capability for the wastewaters. Researchers in biotechnology
maintain that it is possible to influence the population dynamics of various
waste treatment systems (i.e. dairy waste) microflora by altering their
microbial population and improving removal performance through bioaugmen-
tation.
Chambers (23) maintains, that to optimize a biological waste treatment
process, several growth factors must be designed into the system to achieve
the desired water purification. These factors are: to provide and maintain
a physiochemical environment that minimizes suppressive factors on the
waste assimilation process and growth of the biomass; to supply adequate
aeration and mixing to meet the respiratory needs of the biomass microflora
and to assure stabilization of the organic waste constituents; to promote
the establishment of desirable types and quantity of microorganisms to
assimilate the pollutants present in the waste stream; and to control the
balance between the BOD being applied and the quantity of biomass needed to
reduce this loading to a desirable level in the final discharge. However,
the key to the successful treatment of a given waste stream is the capability
of the biomass to assimilate the waste, convert this waste into cells and
suspended solids and form a floe particle that will settle. In accom-
plishing the above, the control of the biomass activity is dependent upon
detention time, control of biomass solids and the maintenance of a desirable
quantity ratio between nutrient loading and biomass solids in the system.
According to Chambers (23) microbial constituents of biomass solids are
critical to the nutrient uptake and new solids yield rate, based on growth
kinetics of biomass. The higher the growth rate the better the efficiency
rate will be for waste removal.
Chambers further indicates that bioaugmentation permits the establish-
ment of selected microbiota which compete more effectively for substrate
and are more tolerant to vascillating growth conditions. Eventually, the
selected bacteria gain population dominance, waste assimilation properties
are improved, and waste removal performance is more reliable. In essence,
the population dynamics is altered through the use of the bioaugmentation
technique.
McDowell and Zitrides (32) have demonstrated Monod's concept of cell
growth response to substrate concentration by using adapted mutant strains
of microbiota specifically selected to assimilate petroleum refinery wastes.
Their observation supported the Monod concept that it is the total amount of
enzyme produced by a given microorganism which dictates the growth response
of the cell. Additionally, the Michaelis-Menton constant, Ks (Monod) for
the limiting substrate can have a direct influence of the microorganism's
ability to compete for available substrate. That is, a microbial cell
which has a lower limiting substrate requirement needed to drive essential
metabolic reaction will have a decided advantage over other competing cells.
This competitive advantage assists the microbial cell in establishing
itself in the waste assimilation environment and certainly influences the
population dynamics.
459
-------
According to McDowell and Zitrides (32) competition for the same food
source always favors the biological cell that has the lower Ks constant.
Decided advantages are a lower limiting substrate requirement, a quicker
growth response and shorter cell generation times. For a nonrestrictive
substrate concentration, all competing microbiota will grow and multiply
but at different growth rates. As the substrate concentration becomes
growth limiting, cells with higher Ks requirements will begin to cease being
competitive. At that point, microbial cells with the lower Ks substrate
requirements will begin to dominate.
The success of the bioaugmentation program lies in the correct selection
of bacteria to assimilate a given waste stream. Those bacteria introduced
into the waste assimilation environment must be able to actively compete for
available food and eventually gain population dominance. Usually this com-
petitiveness is due to the cell's lower Ks (Monod) substrate requirements,
its shorter generation time, and an ability to respond rapidly to a favor-
able growth environment.
According to Chambers (23) and McDowell and Zitrides (32) benefits that
should be realized through the use of boiaugmentation are: decreased sludge
solids yields due to a more efficient destruction of colloidal material;
rapid establishment at start-up or restoration after chlorination of return-
ing sludge of the biological activity in the wastewater treatment process;
improved floe formation and settling; and increased waste assimilation
rates and versatility in substrate uptake.
Enhancement of dairy waste effluent treatment with the use of specific-
ally adapted/mutant microbial supplementation (22,23,40-45) has been reported
in literature. Chambers (23) described the application of bioaugmentation
to an aerated stabilization basin (ASB) system and two activated sludge
processes operating in the dairy industry. The activated sludges exhibited
poor sludge settling characteristics while the ASB process produced a
dispersed biological floe. Poor floe formation associated with dominant
populations of fungal and filamentous bacterial species, resulted in less
than desired levels for BOD and suspended solids being discharged by all
three systems. Accordingly, the "normal" biological mass was supplemented
with specific bacterial cultures. One source of bacteria, culture of
Pseudomonas species was used to routinely inoculate one of the activated
sludge processes. The second activated sludge system and the ASB process
were augmented with a commercially available bacteria. Following the addi-
tion of the special microbiota, the activated sludges exhibited improved
settling, compaction, and a decline in filamentous organisms. This response
allowed a greater concentration of biological solids in the aeration process
reducing the food-to-mass (F/M) ratio and decreasing the sludge production.
Effluent quality for BOD5 and suspended solids also improved following
bioaugmentation. The bioaugmentation of the ASB produced results similar
to the activated sludge inoculation program. A general improvement in
effluent quality was experienced.
460
-------
Reported experience with bacterial augmentation of biotreatment systems
in the pulp and paper industry (46-51) includes trade journal articles,
reports presented at technical meetings and scientific journals. Blosser
(47) presented the source of industry effluents by manufacturing category
where bacterial supplementation of biological waste treatment systems has
been conducted. The Kraft category represents the majority of experience
followed by mechanical pulping and paper production categories. Wastewater
treatment at various mill locations represents a rather even distribution
between the activated sludge (AST) and aerated stabilization basin (ASB)
processes. With the exception of one mill location, pulp and paper industry
experience with bacterial supplementation reflects the "before" and "after"
approach to product evaluation, in that historical data provided the basis
for comparison. This, according to Blosser (47) reflects the lack of
opportunity within the physical design of the treatment systems to provide
the control feature necessary for conducting a parallel examination.
Numerous studies documented in the literature report on the successful
application of bioaugmentation products in the improvement of various aspects
of aerobic wastewater treatment (4, 7-9, 32, 52-64). Similarly, improvements
in the anaerobic sludge digestion and enhancement of anaerobic treatment
(65-69) with the use of biocatalytic additives have been reported. Litera-
ture abounds with references regarding the implementation of microbial
supplements in waste water systems for the improvement of sludge digestion
(70-74); sludge management (75-77); sludge settling improvement (78; sludge
filtration (79); and nitrification (80,81).
Many studies have also shown the useful application of enzyme prepara-
tions as biocatalytic additives (79, 82-91) in the waste treatment processes;
in the treatment of primary sewage sludge; for the improvement of biosludge
filtration; in the pretreatment system of industrial wastes and in treatment
of specific organic toxic compounds; as well as in the food industry waste
processing.
II. Research Results and Critiques by Critics of
the Bioaugmentation Approach
Review of the studies of bacterial augmentation of biological treatment
in the literature indicate that the majority of experience has resulted from
examination of performance data from full scale processes comparing the
trial period with before-and-after performance data. In those cases where
the use of parallel treatment trains have been cited, no reference has been
made to characterizing the process to determine what differences may exist
between parallel treatment units. In addition, failure to identify process
conditions in the trial and control systems during the course of the studies
limits the utility of the data generated.
The literature also contains reports on a number of laboratory studies
where bioaugmentation has been evaluated. Some researchers who criticize
the bioaugmentation technology laboratory research studies, argue that .
although laboratory scale activated sludge studies provide a desired degree
461
-------
of control, they generally operate under dissolved oxygen levels and hydrau-
lic clarification loadings far removed from conditions existing in full
scale application of the process. In addition, it is indicated that such
studies do not normally incorporate the variability of influent conditions
to which a full scale process must respond.
Blosser (47) summarizes experience with bacterial augmentation of bio-
logical treatment process to improve performance in the treatment of pulp
and paper industry wastes and reports on supplemented bacteria inoculations
of a full scale activated sludge process treating a bleached kraft mill
effluent. Pulp and paper industry experience with bacterial supplementa-
tion has predominantly resulted from full scale application using the
before-and-after assessment approach and a number of such situations have
been identified.
Blosser (47) reports a 150-day study conducted at a bleached kraft
mill to examine bacterial augmentation of its activated sludge process.
Specific goals assigned to the study were: improvement in effuent BOD5 and
suspended solids, with a greater emphasis on effluent suspended solids qual-
ity; an accelerated rate of recovery of the activated sludge from a process
upset; reduction of biological sludge generation rate; and sustaining or
improving activated sludge settleability and thickening properties.
The response of the control and trial (augmented) activated sludge proc-
esses at this specific bleached kraft mill have resulted in the following
observations:
• The BOD5 quality in the treated discharge from the activated
sludge process was neither improved nor deteriorated by the use of
bacterial augmentation.
• The data collected following augmentation trial indicated that the
differences in suspended solids discharged from the trial and control
portions of the treatment system were in effect, typical of system
performance and not necessarily the result of bacterial supplemen-
tation.
• Activated sludge settling velocity and thickening potential were
not influenced by bacterial augmentation.
• A reduction in biological sludge production, of approximately 0.10
Ib TSS produced per Ib of 6005 removed, observed in the trial acti-
vated sludge unit when compared to the control, was shown to be an
inherent aspect in the performance of the dual train activated
sludge process and could not necessarily be attributed to the use
of the specialized bacteria.
Quasim and Stinehelfer (39) described the use of control and trial
laboratory activated sludge units to determine the impact and performance
of a special bacterial product. Batch reactors were used to select the
462
-------
optimum dosage of the freeze-dried microorganisms, based upon the maximum
reduction of Q£ demanding substrate in a domestic wastewater. Evaluation
of the bacterial additive was conducted in 2 identical continuous flow
activated sludge units operating under the same conditions.
The utility of bioaugmentation was evaluated by two techniques. One
was a statistical analysis of effluent BOD5 quality for the dosed and
control units, while the second approach compared the biological kinetic
coefficients observed in the 2 units. The kinetic coefficients compared,
were: biological sludge yield (Y), maximum rate of bio-oxidation (k),
microorganisms decay (kj) and the half velocity constant (Ks). Statistical
analysis of the effluent quality indicated no significant difference between
the control and dosed activated sludge units.
The authors concluded that addition of the special bacterial product
had no effect on substrate utilization, k, and half velocity Ks coefficients.
The dosed unit did exhibit modest differences in sludge yield and sludge
decay coefficients, being slightly greater in the former and slightly less
in the latter. The effect of the kinetic coefficients in the control and
trial units was estimated using a Monod mathematical model. The model
projected that the activated sludge unit receiving the special bacterial
culture would produce a slightly lower BOD5 concentration than the control
unit, depending upon the sludge age of the process. There was little effect
of the product on the overall performance of a well operated activated
sludge plant. Perhaps the product may have some benefit in those plants
that are already overloaded and are operating at poor organic removal
efficiencies.
While most of the product manufacturers talk about the successes they
have had with their products, little has been published about testing
methodology. The intent of Quasim and Stinehelfer (39) studies was to
present a methodology for evaluating such products. The procedure presented
in the report provides a systematic approach in evaluation of bacterial
culture products and is a widely used technique in developing design para-
meters for industrial and joint industrial-municipal wastewater treatment
facilities. Because the kinetic coefficients developed for bench scale
reactors are so valuable in the design of treatment plants, it is felt by
the authors that the same procedure can be used to evaluate the performance
of bacterial culture products for special applications such as the effect
of these products on biological treatment plants under different operating
conditions.
In his assessment on the use of enzymes and biocatalytic additives for
wastewater treatment process, Young (91) maintains that adding commercially
available enzyme products to both aerobic and anaerobic systems might not
cause a dramatic change in the system's performance, since a full complement
of different enzymes is generally needed to mediate the many complex series
of biochemical reactions, beginning with the change of the parent substrate,
going through many intermediate steps and finalizing in the end product.
Such full complement composition of enzymes in any of the commercial bioaug-
mentation products is neither available nor economically feasible.
463
-------
The exact composition of commercially available biocatalytic additives
is often unavailable. As either dried bacterial solids, by-products of
bacterial fermentation reactions or a combination of microbial cells and
their fermentation products, these additives are not pure enzymes and
possibly could contain about 1% enzyme by weight even though no separation
or purification step is involved. According to Young and others, it is
difficult to see how the small dosage of dried cultures recommended by the
suppliers can overcome the effect of the large amounts of bacteria already
present in the wastewater treatment process. If the environment was favor-
able for their existence, they would already be there in large quantities.
Again, according to the critics, it is difficult to see how a commercial
biocatalytic formulation containing by-products of microbial fermentation
reactions, even if used at full strength, can cause a dramatic increase in
the breakdown of organic materials, and how these preparations can help
decompose materials which can not be decomposed by the bacteria which have
been in constant contact with the waste materials.
In addition, the use of the bioaugmentation products at quantities
needed and for the duration of time recommended by suppliers to enhance a
specific treatment of waste, may be quite cost prohibitive. In a number of
cases, where enzymes have aided treatment, it is the feeling of the critics
that there is no way to tell if the enzyme would have been more effective,
if some of the operating process control strategies had been modified,
before addition of enzyme, to effect a better treatment system.
Biocatalytic additives frequently have been misapplied, according to
critics of bioaugmentation approach. The critics point out that it is
almost impossible for biocatalytic agents added in quantities normally
recommended by suppliers to improve the performance or capacity of properly
designed and operating treatment plants, including anaerobic digesters.
When fats, proteins and carbohydrates accumulate to such an extent that
problems occur, biocatalytic agents may be beneficial if they can be
justified economically.
According to Young (91) a number of significant recommendations were
made by the Special Committee on Enzyme and Biocatalytic Additives of the
Water Pollution Control Federation. These recommendations are summarized
as follows:
• Biocatalytic additives should not be purchased until after they
have performed as claimed in accordance with a written guarantee.
• Advertising literature should be rewritten clearly and correctly
so as not to misrepresent the merits of the product.
• Distributors should deal directly with engineers, superintendents
or operators rather than attempt to influence mayors, councilmen,
or members of boards of public works.
464
-------
• Before the additive is considered, it should be tested so that
fundamental and reliable data can be obtained.
• When digestion problems develop, major efforts should focus on
proven methods of analysis and recovery.
These recommendations have been published in the Water Pollution
Control Federation Manual of Practice No. 16 on Anaerobic Sludge Digestion.
The Manual concludes that biocatalytic additives cannot be added economically
in effective quantities.
Rittman (92) in his discussion of the relationship between genetic-
control and process control strategies in municipal wastewater treatment
points to the following problems in current biological processes: poor
reliability of treatment and poor performance; characteristics of sewage
input: high volumes, variable flow rates, low and variable constituent
concentrations and sudden shock loads caused by precipitation or industrial
discharges; and high cost of treatment. He outlines the following approaches
to enhance treatment performance and enhancement of reliability and/or
economics of biological treatment: make a process related improvement;
apply an existing process in a new situation; select novel microorganisms;
and genetically engineer appropriate microbiota to have a new desired
function or trait.
Rittman stresses the need for coordinating the process control with
genetic engineering control. He maintains that genetically controlled
microorganisms must continually proliferate and must be able to grow and
out-compete its rival competing strains of microbiota for the various
substrates. He states that addition of exogenous biomass at the tons/day
rate would be expensive if not practically Impossible. Initial and
maintenance additions of mutant bacteria are insignificant compared to the
total production of biomass. Only if mutant bacteria can grow and out-
compete the indigenous strains would bioaugraentation be valuable and if
such growth occurs, maintenance additions would then be unnecessary.
Rittman also points out that causes of process control problems are never
identified in bioaugmentation literature.
465
-------
EVALUATION STUDIES OF BIOAUGMENTATION PRODUCTS
Bioaugmentation products are manufactured by an increasing number
of industries to enhance specific wastewater treatment and sludge hand-
ling capabilities. With the increasing use of these products, the En-
vironmental Protection Agency (EPA) recently initiated a program to
evaluate the bioaugmentation technology and to assess the legitimacy
and effectiveness of the manufactured biocatalytic products through the
use of standard testing protocols. The goals of this program are three-
fold: first, to place the use and characterization of these types of
products on a scientific basis; second, to arrive at consensus-testing
standards for label verification/package contents and for performance of
these bioaugmentation products; third, to assess capability of the bio-
augmentation technology through the use of the standard testing proto-
cols on biocatalytic products. Accordingly, the National Sanitation
Foundation (NSF) in Ann Arbor, Michigan was funded by the EPA under the
Cooperative Agreement No. CR-811884 to meet these goals.
The specific objectives of the Agreement are: to review and cate-
gorize the current array of bioaugmentation products by type of generic
category and by application; to develop and validate the standard pro-
cedures to insure that package contents are consistent with package
labeling; to develop and validate standardized test procedures to eval-
uate product performance against manufacturer claims; to bring product
manufacturers together to launch the development of a national voluntary
concensus standard for the bioaugmentation products; and, to implement an
ongoing, self-supporting program of independent testing and evaluation
based on these standards and the developed test protocols with the aim
of achieving a product certification system.
The project was divided into four general Phases. A completed Phase
I consisted of gathering information on bioaugmentation products and test
methods and the formation of a Technical Advisory Committee and an Indus-
try Committee. Both committees were given responsibility to review and
comment on data summaries and reports. The Technical Advisory Committee
was assigned responsibility for test protocol development. The Industry
Committee was asked to draw on its experience with test methodologies to
assist in protocol refinement. The ongoing Phase II includes the devel-
opment of detailed protocols, product testing, protocol refinement and
validation and review of results. In the future, Phase III will dissem-
inate results and develop the final reports. Phase IV, which does not
rely on project funds, will develop a national voluntary consensus stand-
ard for bioaugmentation products.
In Phase I, the Technical Advisory Committee established, (Table 8
provides a list of committee members) five major categories of potential
product applications. These were: aerobic treatment; anaerobic treat-
ment; sludge handling and digestion; the in-sewer treatment; and specific
466
-------
TABLE 8. BIOAUGMENTATION ADVISORY COMMITTEE
CTl
Dr. James V. Chambers
Smith Hall
101 D
Dept. of Food Sciences
Purdue University
West Lafayette, IN 47907
Mr. Donald Kerr (observer)
US Department of Agriculture
Food Ingredient Assesment Div.
Washington, DC 20250
Mr. Eugene DeMichele
Water Pollution Control Federation
2626 Pennsylvania Ave. , NW
Washington, DC 20037
Mr. William Hill
V.P. Operations and Maintenance
Camp Dresser & McK.ee Inc.
One Center Plaza
Boston, MA 02108
Dr. Henryk Melcer, Chief
Biological Treatment
Environmental Protection Service
Canada Centre for Inland Waters
Burlington, Ontario L7R-4A6
Phone Number
317-494-8279
202-447-7680
202-337-2500
617-742-5151
416-637-4546
Mr. Howard Selover 517-373-0397
Chief
Operator Training Unit
Community Asistance Div.
Department of Natural
Resources
Lansing, MI 48909
Dr. Jospeh Trauring 215-822-8123
Pollu-Tech Inc.
PO Box 77
Chalfont, PA 18914
Dr. Leon. Weinberger 301-340-7990
Peer Consultants Inc.
1160 Rothville Pike
Suite 202
Rockville, MD 20852
Dr. David R. Zenz 312-780-4060
Metropolitan Sanitary
Dist. of Greater Chicago
R & D Laboratory
5915 W. 39 St.
Cicero, IL 60605
Dr. Joe Rang 313-665-6000
McNamee, Porter and Seeley
3131 South State Street
Ann Arbor, Michigan 48104
-------
substrate removal. Top priorities (Table 9) included all categories ex-
cept in-sewer treatment. The NSF Staff and the Technical Advisory Com-
mittee also established a Bioaugmentation Industry list. This list in-
cludes those identified by NSF as being involved and/or interested in the
bioaugmentation industry: manufacturers, reformulators, vendors and con-
sultant/users. This list is not inclusive and does not contribute an en-
dorsement by the EPA or NSF of the companies listed.
Also in Phase I, at the first Bioaugmentation Industry Meeting, held
on November 27, 1984, the industry recognized the value of a certifica-
tion program and was fully appreciative of the EPA's attempt to improve
the credibility of the bioaugmentation technology and the biocatalytic
products. They also recognized that the existing assays were not uniform
across the industry for specific bioaugmentation activities. They became
aware of a need for the industry to financially support field testing,
after the completion of the NSF laboratory investigation, and eventually
to develop a certification process. After meeting with the Industry Com-
mittee, the Technical Advisory Committee of the Bioaugmentation Products
Evaluation Studies, working closely with the NSF research team and the
EPA project officer, began the development of the label verification and
performance testing protocols. At a second Bioaugmentation Industry
Meeting on March 1, 1985, attended by a cross-section of producers and
suppliers as well as technical representatives of the bioaugmentation
industry, the NSF research team and members of the Technical Advisory
Committee, a general consensus and agreement was reached on the proposed
draft label verification and efficacy testing protocols for the project.
It was also agreed that the voluntarily obtained bioaugmentation products
would be tested by the NSF. A separate program to develop policies and
procedures for national voluntary consensus standards was initiated con-
currently with the EPA project.
After the Second Bioaugmentation Industry Meeting, Dr. Robert Wolfe,
Professor in the Biostatistics Department, University of Michigan,
School of Public Health, developed a statistically sound experimental de-
sign for the Label Verification and Performance Protocols. He is also
performing statistical analysis of the generated data throughout the
evaluation study at National Sanitation Foundation. With the appropri-
ate quality control, Dr. Wolfe's statistical analysis experimental de-
sign for the continuous flow phase of the performance (efficacy) proto-
col is based on the Latin Square approach. The Latin Square statistical
experimental design approach, reviewed by the U.S. EPA statistical team,
is very appropriate for the use in performance-testing of biocatalysts
in both the synthetic and raw wastewater media in either sequential or
concurrent experimental runs.
DESCRIPTION OF TESTING PROTOCOLS FOR EVALUATING BIOAUGMENTATION PRODUCTS
Protocol testing was divided into three areas of effort: label
verification/product classification, bench scale performance (efficacy)
testing, and field performance testing. A flow diagram for the overall
approach involving the protocol development process is presented in
Figure 16.
468
-------
TABLE 9. MAJOR CATEGORIES OF POTENTIAL BIOAUGMENTATION PRODUCT APPLICATION
Four major categories of potential product applications were defined. These
were: in-sewer treatment, aerobic treatment, anaerobic treatment, and sludge
handling. The committee chose aerobic treatment and sludge handling as the
two categories requiring the most attention. The major categories and their
applications, as prioritized by the group, are outlined below.
AEROBIC
TREATMENT
SLUDGE
HANDLING
ANAEROBIC
TREATMENT
5.
Reduce sludge 1. Sidestream
production improvement
Improve 2. Thickening/
efficiency dewatering
Improve 3. Composting
clarification
Specific sustrate
removal (including nitrification and denitrification)
Improve startup time
1. Digestion
improvement
2. Septic tanks
3. Improve
mainstream
treatment
IN-SEWER
TREATMENT
1. Odor control
2. Cleaning
3. Corrosion
control
4. In-line
treatment
469
-------
Figure 16. Bioaugmentation Study Design,
-------
PRODUCT CLASSIFICATION AND LABEL VERIFICATION PROTOCOL
The objectives of the Label Verification/Product Classification
portion of the bioaugmentation products evaluation study are listed as
follows:
• Develop and test reliable protocols to determine the potency
of bioaugmentation products with potency measured in terms of
number of viable organisms grown on eight different agar/
supplement combinations, enzyme activity, and/or whole cell
activity (e.g. specific substrate reduction, oxygen uptake,
TOG removal).
• Determine the consistency of products submitted for evaluation
with consistency measured as the reproducibility of product
potency determined on replicate subsamples of different lots
of samples secured from the same industry at different times.
• Enumerate and identify the active microbiota and correlate the
findings with information provided by industry.
• Develop an overview of the bioaugmentation products as measured
by the range of potency of the bioaugmentation products.
The Product Classification and Label Verification Protocol includes
a procedure for plate counts (aerobic and anaerobic) using Standard
Methods agar media and Reasoner agar media (R2A) to quantify, character-
ize and identify viable microbiota in all products. Composition of R2A
medium is outlined in Table 10. The protocol also includes total plate
counts, supplemented with specific substrates to detect enzyme activity
(protease, lipase, amylase, etc.) and to observe specific substrate re-
moval (cellulose, phenol, phthalates, chlorinated organics, etc.).
Naturally occurring microbial populations in soil, activated sludge and
other samples are evaluated in the classification protocol and used as
controls to establish baseline criteria for characterizing the bacterial
additives.
The bioaugmentation products could either exhibit or not exhibit
a significant difference from the controls. If there is no significant
difference, then two possibilities are pursued:
• the protocol is not capable of discriminating between effective
products and naturally occuring biota, or
o the products demonstrate no significant difference from natur-
ally occuring biota.
Products with active biocultures are further evaluated with aerobic
and/or parallel anaerobic activity tests, while non-bacterial products
are evaluated with non-bacterial tests (nutrient analysis, tests for
"free" enzyme activity). Appropriate short-term tests of enzyme acti-
vity are performed on homogenized cell-free extracts and filtrates (free
471
-------
TABLE 10. COMPOSITION OF R2A MEDIUM
Ingredient Concentration, g/L
Yeast Extract 0.5
Proteose Peptone No. 3 0.5
Casamino Acids 0.5
Glucose 0.5
Soluble Starch 0.5
Sodium Pyruvate 0.3
K2HP04 0.3
MgS04.7H20 0.05
Agar 15.0
Final pH 7.2, adjust with K2HP04 or KH2P04 before adding agar.
Add agar, heat medium to boiling to dissolve agar and autoclave
for 15 minutes at 121°C, 15 psi.
472
-------
enzymes) using procedures for determining enzyme activity contained in
Methods of Enzymology as referenced in a manual published by Worthington
Biochemical Corporation, Freehold, NJ 07728. A filter/ultrafiltration
approach is used to isolate free enzymes from whole cells and cell frag-
ments. Short-term tests of enzyme activity are completed on the fil-
trate. Longer term activity measurements on specific substrates are
compared to "equivalent activated sludge" using whole cell tests based
on disappearance of substrate and oxygen uptake determinations. The
overall proposed approach for characterization and verification
of product quality is shown in Figure 17.
The majority of bioaugmentation products have labels with general
information concerning ingredients and broad performance claims. On
labels where numbers of bacteria or levels of enzyme activity are listed,
the procedures used for quantification are not reported. In addition,
due to the lack of specificity, there have been reports by users of prod-
ucts that the additives were nothing more than inert fillers or dried
activated sludge.
The specific goal of the label verification protocol is to provide
simple and reproducible methods for measuring general microbial and
enzymatic activity. Label verification is conducted "label blind," as if
no product ingredients were specified. Protocols are being used to
screen products where labels provide variable information on content.
The verification of specific ingredients is an important part of the
study.
The statistical experimental design developed with an appropriate
quality control effectively addresses the testing of potency, consist-
ency and microbial identification/enzyme characterization of the bio-
catalyst additives. Reproducibility of the testing methods to determine
the potency and consistency of the biocatalysts is established by the
use of ten replicates of standards, natural samples and products. The
replicate samples are run initially to determine the variance of sub-
sampling and analytical methods. Routine analyses during product test-
ing are run in duplicate. The range of duplicate samples are compared
to the standard deviation of the ten replicates. Results with a range
exceeding two times the standard deviation are reported and the product
sample is reanalyzed.
Product consistency is measured by analyzing ten replicate sub-
samples of ten different lots of the product requested from industry for
the repeated testing. Suitable quality control in the Label Verifica-
tion Protocol is established for the test methodologies used for plate
counts, enumeration and identification of the active microbiota, enzyme
activity, and whole cell tests.
The specifics of the Label Verification experimental design are pre-
sented in Figure 18. The design calls for three test sequences for all
products; plate counts, enzyme activity, and whole cell activity tests.
473
-------
Proposed approach For Characterization
and Verification of Product Dually
EOT All Products
Identify Viable Cells-
Battery of Plate Counts
(Aerobic & Anaerobic)
Establish Baseline
Criteria for Defining
"Bacterial" Additives* *
Aerobic
Heterotrophic
Bacterial
Products
(1) Label Blind
•Total Plate Using
2 Media
•Protein (Protease)
•Starch (Amylase)
•Fats (Llpase)
(2) Other Not Captured
Above •
Appropriate Non-volatile
Substrates Based On
Label Claims
(Phenol, Phthalate, etc.)
*Run plate counts on all products submitted
Anaerobic
Bacterial <
Products
Other
Products
(Non-Cellular)
Appropriate Longer
Term Activity Measure-
ments on Specific
Substrates (Protein
Hydrolysis, etc.)
Compared to "Equiva-
Activated Sludge
Compar
lent" Ai
Activity Assays
Whole Cell Tests •
Disappearance of
Substrate/0, Uptake
(Longer Term)
Homogenized Cell-Free
Extracts
(Short Term)
Filtrate (Free Enzymes)
(Short Term)
Tests For Nutrients and
"Free" Enzyme Activity
* *Use soil, activated sludge and other samples to
perform these analyses and provide "baseline" data to
discriminate between naturally occurring organism and
engineered strains.
Figure 17. Proposed Approach For Characterization and Verification of Product Quality,
-------
All
Products
Received
Activated
Sludge
Soil
ExtraceUar
»
Aerobic
Heterotrophic
• Plate
Counts
/Esqmt
W Activity
\ Tests
Total
>-
Whole
Ceil
Activity
Tests
o
»•
Results
ami
Protocols
Reviewed
i
Selected Belesis ami New Products
Correlate
with
Performance
and Field
Studies
Final
Report
1.
Enumeration of organisms grown on
general heterotrophic media and media
which will enhance growth of organisms
with selective enzymatic activity.
Products are homogenized for total
activity.
Products are ultrafiltered, titrate is
analyzed for extracellular activity
3.
Batch activity tests include 02 uptake
rate, specific substrate disappearance, etc.
Figure 18. Label Verification Experimental Design,
-------
Plate Counts
Protocol - Each product is analyzed initially with a series of plate
counts. Samples are prepared by completely mixing 1 mL of liquid product
or 1 gm of dry product to 99 mL of autoclaved distilled water. Dilutions
are made as required. The plate counts are completed using nine differ-
ent media:
Standard Plate Count Agar (STP)
STP Agar minus glucose (STPG-)
STPG- supplemented with milk
STPG- supplemented with Tween 80
STPG- supplemented with starch
Dilute strength STPG- (DSTPG-)
DSTPG- supplemented with milk
DSTPG- supplemented with Tween 80
DSTPG- supplemented with starch
Quality Control - Duplicate plates are prepared for each type of agar and
each incubation time/temperature. Plates are incubated at two tempera-
tures and two incubation periods. Duplicate plates with each of the
media are incubated at 23°C for 48 hours, 23°C for 72 hours, 35°C for 48
hours and 35°C for 72 hours. Results are reported as organisms per unit
weight or volume.
A positive and negative control is used with each analytical run.
Positive controls are plates of agar and supplement with appropriate or-
ganisms (e.g. E_. coli, Pseudomonas). Negative controls constitute plates
of agar and supplement with no bacteria or products added. Routine qual-
ity control also consists of two subsamples (replicates) of a product.
Ten replicates of a liquid and a dry product are plated for each of
the nine media. The ten repetitions are incubated at 25°C for 48 hours
and 35°C for 48 hours. The results are used to develop control charts
for 23°C and 35°C incubations.
The control charts are used to measure the variability of subsam-
ples taken from a product. If the relative range of two subsamples
exceeds three coefficients of variation for like samples (e.g. liquid
samples incubated for 48 hours at 23°C), the samples are reanalyzed.
Ten replicates of five natural samples (e.g. activated sludge and
soils) are analyzed. Three activated sludge samples are used (North-
field Township, Chelsea, and Ann Arbor). Two soil samples are analyzed.
The results for these natural biomasses are compared with product sample
results.
Enzyme Activity
Protocol - Extracellular and total enzyme activity are measured.
Liquid and dry samples are prepared in the same way as samples for plate
counting. Extracellular activity is measured on samples which have
476
-------
been filtered through a 0.45 micron filter. Total enzyme activity is
measured on samples which have been completely homogenized in a high
speed blender and then passed through a 0.45 micron filter.
Enzyme activity tests vary based on the enzyme of interest. How-
ever, most rely on measurement of the disappearance of a substrate or
appearance of an enzymatic product. Results are reported in a number of
ways, but generally relate to moles of substrate used or product pro-
duced per unit time. Standard enzymes are available for comparison.
Quality control - Routine analyses are completed in duplicate. A.
blank, three standards, and a matrix and control spike are included with
each analytical run. Samples which exceed two times the activity of the
highest standard are appropriately diluted and reanalyzed. Samples with
activity less than 40 percent of the lowest standard are rerun either as
a concentrated sample or with a diluted standard.
Ten replicates of the standard at three concentrations are analyzed
to establish the reproducibility of the method. Ten replicates of a
liquid and dry sample are analyzed to establish the reproducibility of
subsampling procedures.
Whole Cell Tests
Protocol - The protocol depends on the product and its claims.
Products making claims for specific substrate reductions are analyzed by
the static-culture flask screening procedure (93, 94). Products with
claims of more general performance are analyzed for oxygen uptake rate or
TOG reduction. For either of these tests, products are added to real
wastewater at the prescribed dose and at one-half and two times the pres-
cribed dose. Three different wastewaters are used.
Quality Control - The reproducibility of the method is established
with 10 replicates. Standard NSF laboratory analytical and quality con-
trol procedures are used when appropriate (e.g. BOD, TOG).
PERFORMANCE (EFFICACY) PROTOCOL
The specific tests in performance evaluation of bioaugmentation
products in the municipal and industrial wastewater treatment include
tests for enhancing: overall aerobic and anaerobic wastewater treatment;
biodegradation of specific substrates; nitrification/denitrification;
sludge reduction and settling; and, oxygen update by active microbiota in
waste treatment.
Consequently, experimental design for performance testing protocol
has two levels: to develop methodology for evaluating bioaugmentation
product performance and to determine whether bioaugmentation products do
improve performance of bench-scale extended aeration reactors and batch
treatment systems.
477
-------
The Performance (Efficacy) Protocol consists of: continuous flow
studies for products claiming sludge reduction and better overall bio-
logical treatment; and batch biodegradability studies for products
claiming improved treatment of specific organic priority pollutants and
other hazardous and/or toxic compounds.
Continuous Flow Studies
The continuous flow studies incorporate in 20-liter capacity
aeration reactors with separated clarifiers, synthetic and raw waste-
water as media, specific pollutant substrates in several concentrations,
and the bioaugmentation products. These products include specific sub-
strate oxidizing products, sludge reduction products, sludge settling
aid products, nitrification enhancement products, and 02 uptake enhance-
ment products. The tests at several product concentrations include
several incubation temperatures as well as control systems without addi-
tives using the same specific substrates and activated sludge biomass in
the synthetic and raw wastewater media.
The performance protocol experimental design is outlined in Figure
19. The experimental design features: studies at 3 consecutive MCRT
series (A, 8, and 12 days) (steady state achieved after 3 MCRT periods)
to develop kinetics of specific substrate removal (substrate MCRT biode-
biodegration rates) and kinetics of sludge reduction (sludge reduction
rates); and studies to determine the effects of shock loads of substrates
after completion of full test sequence with specific substrate.
The laboratory analyses include measurement of TOC/COD, 8005, SS/VSS
of the effluent; D.O., pH, MLSS/MLVSS, ATP, dry solids, filterability and
settleability of reactor liquor; as well as microscopic examiniation of
the biomass of reactor liquor and waste sludge to characterize and iden-
tify the microbial population. Specific substrate analyses are performed
on feed media, reactor liquor and effluent samples. Figure 20 describes
the continuous flow testing phase of performance protocol sampling scheme
and the reactor requirements.
The continuous flow studies with the synthetic and raw wastewater
and twenty liter capacity aerated reactors with appropriate feed tanks,
return sludge systems, clarifiers and collection tanks incorporate con-
trol systems with and without substrate, but without a bioaugmentation
product; experimental reactors either for sludge reduction or sludge
settling biocatalysts, nitrification/denitrification enhancement prod-
ucts, or oxygen uptake enhancement additives; but all with a specific
substrate oxidizing product in them.
Experimental Design —
Steady state evaluation made at each of 3 consecutive MCRT
series (steady state achieved after 3 MCRT periods), with
measurements of effluent TOG; effluent specific substrate
in the synthetic wastewater feed; effluent 8005; effluent
SS and VSS; reactor biomass (MLVSS and ATP and microscopic
478
-------
Feed
Waste
1. Synthetic
2 Real
1
-»•
r1
+ Phenol 1. 2
L3
r4
^.Phonnl Lc
1 6
OK , r7
+ Phenol (-8
U
-.10
Control
Reactors
Oxident
Reactors
Sludge
Reduction
Test
Reactors
-------
03
o
Waste
Sludge
Sludge
1. Chloramines
2. SS/VSS, BOD5 (total & soluble), TOC (total & soluble), NH4, N03, TKN, COD
3. ATP, SS/VSS, Filterability, Settleability
4. BOD, (total & soluble), TOC (total & soluble), NH<, N03,TKN, COD, SS/VSS
-»-\ Clarifiers
Collection
Tanks
Figure 20. Performance Protocol Sampling Scheme .
-------
characterization of organisms); waste sludge dry mass; waste
sludge filterability; sludge settling characteristics using
settleometer (ZSV).
• Shock load, after completion of full test sequence (period of 6
MCRT) with specific substrate.
The elements of the testing method using 10 reactors include:
• Consecutive MCRT series of 4 days, 8 days and 12 days with reactor
hydraulic retention times of 6 hours;
• Steady state established after 3 MCRT periods, with 3 samples
analyzed per each MCRT period before steady state is achieved;
• After steady state is attained, 9 samples collected during the
next 3 MCRT periods;
• Five gallon (20 liter) aeration reactors with an external clari-
fier, and pumps to return settled sludge to the reactor;
• Sludge settling tests with the withdrawn sludges that are returned
to the reactor;
• Sludge wasting from the reactor, with the volume and frequency
based on the MCRT and pump characteristics;
• Synthetic wastewater for the continuous flow studies as formulated
and used by U.S. EPA, Cincinnati, Ohio (95, 96).
The composition of the synthetic wastewater is described in Table 11.
A detailed sampling schedule and description of analysis to be performed
on wastewater feed, reactor liquor and effluent samples is shown in Table 12.
A statistical experimental analysis design for the continuous feed
portion of the efficacy testing protocol, based on Latin Square approach,
comprises studies with synthetic and raw wastewatwer systems at 3 MCRT test
sequences (4, 8 and 12 days), each to be run for six MCRT periods; 3 to
achieve steady state and 3 during steady state operation. The study
variables in the experimental statistical design are as follows: type of
media: synthetic wastewater (S); raw wastewater (W); treatment methods:
control (C); sludge reduction (R); substrate oxidation (0); sludge set-
tling action (A); aerobic treatment (AT); anaerobic treatment (ANT);
nitrification/denitrification activity (N); oxygen uptake activity (U);
and 3 different MCRT series to be run sequentially (4-, 8- and 12-day
series.
Reactors with synthetic and raw wastewater media are run either
sequentially or concurrently. Because of the different wastewater con-
ditions, the results for the synthetic and raw wastewater tests are
481
-------
TABLE 11. SYNTHETIC WASTEWATER FORMULATION
Used in the Continuous Flow'Biodegradation Studies at U.S. EPA,
WERL, Wastewater Research Division Laboratories
In order to eliminate the variable nature of municipal wastewater and prevent possible
introduction of toxic materials during the acclimation period, a synthetic waste was used in
most of these studies. This feed solution was formulated after a review of prior published
formulations (10 through 23).
The final synthetic waste solution, prepared by diluting stock solutions with distilled
water had the following composition:
Constituent Concentration, mg/L
KH2P04 8.5
K2HP04 22
Na2HP04 33
NH4C1 2
MgSC4 22
CaCl2 36
FeCl3 0.3
Urea 50
NaHO>3 300
Yeast Extract (Difco) 55
Baco Peptone (Difco) 50
Meat Extract (Difco) 50
Fish Meal Extract (Purina Trout Chow) 0.5 ml
The fish meal extract was prepared by grinding lOg of Purina Trout Chow with 200 ml of
distilled water in a high speed blender for 3 min. The mixture was allowed to settle for
10 min. and then 0.5 ml added to each liter of synthetic waste. Experience has shown this
additive helps control bulking of sludges produced from synthetic feeds.
The synthetic feed typically has the following characteristics:
Item Concentration, mg/L
Chemical oxygen demand 160
Total organic carbon 76
Suspended solids 2
Total Kjeldahl nitrogen 47
Ammonia nitrogen 2
Nitrite and nitrate nitrogen <0.1
Total phosphorus 13
Alkalinity, as calcium carbonate 211
pH 7.3 units
The nitrogen and phosphorus content is slightly higher than most municipal wastewaters.
However, it was deemed necessary to have these nutrients in excess to insure the waste was
not growth limiting. The suspended solids are low because the waste is composed of soluble
materials for ease of preparation of the feed from stock solutions.
482
-------
TABLE 12. PROPOSED LAB ANALYSES
ANALYSIS
TOC
BODs
SS/VSS
D.O.
pH
MLSS/MLVSS
ATP
Dry Solids
Filterability
SAMPLE
LOCATION
Clar. Eff.
Clar. Eff.
Clar. Eff.
Reactor
Reactor
Waste Sludge
Waste Sludge
Waste Sludge
Waste Sludge
Or Reactor
SAMPLE
TYPE
Grab
Grab
Grab
Probe
Probe
Grab
Grab
Grab
Grab*
SAMPLE
FREQUENCY
(1), (2)
(1), (2)
(1), (2)
Daily /React or
Daily/Reactor
Daily /Reactor
(1), (2)
(1)
(1)
WASTEWATER
FEED
(3), (4)
(3), (4)
(4)
-
-
-
-
-
-
Settleometer Reacter Grab*
(Sludge Settleability)
Slide for Microscopic
Examination Waste Sludge Grab
(1), (2)
(1)**, (2)
Key to Sample Frequency/Wastewater Feed;
(1) Analyses to be completed 18 times over a period of 3 MCRT's to achieve a steady
state and 3 MCRT's at a steady state.
(2) Analyses to be completed during transition periods at a frequency of 3 times/MCRT.
(3) Periodic check for synthetic waste - 1 time/month or less if uniform results
obtained.
(4) Perform tests as needed to characterize wastewater feed. Use WWTP data as a QC.
* Sample returned to reactor upon completion of test.
** Prepare 1 slide for each MCRT at steady state (total of 3).
483
-------
analyzed as two separate experiments. In order to increase the statis-
tical power of the experimental design, the results of the separate
analyses are combined, where possible, to yield an overall estimate of
efficacy for a specific treatment and an overall test statistic compar-
ing the experimental systems to control systems.
In the Latin Square design, ten reactors allow for testing of two
different biodegradation products at the same time per experiment.
Three reactors are operated for each bioaugmentation product and three
as controls. The design includes 9 experimental reactors with a tenth
reactor as an additional control for the sludge reduction product.
Currently there are two bioaugmentation products tested for per-
formance in comparison to the control systems, a sludge reducing product
and a specific substrate oxidizing product. Accordingly, an example of
the experimental and control reactor strategy for the specific oxidant
(i.e. phenol) and sludge reduction product in one experiment is illus-
trated as follows:
123456789 10
0 G! R 0 GI R 0 GI R C2
GI = control with substrate 0 = specific substrate oxidant
C2 = control without substrate R = sludge reduction additive
The tenth reactor, used as a control in the monitoring of the impact of a
specific substrate on a sludge reduction product, is operated with a
wastewater that does not have an added substrate. The control reactors
2, 5 and 8 receive wastewater spiked with a specific substrate.
The sampling schedule in the proposed performance testing protocol
consists of 18 samples during each MCRT series (4, 8 and 12 days). An
MCRT series consists of 3 MCRT periods to approximate steady state
followed by 3 MCRT periods for product evaluation. The analysis series
of 18 samples for each series provides kinetic data for a treatment
system with and without the use of a specific biocatalytic additive.
The general Latin Square design (as shown in Figure 21) provides ade-
quately an interchange between control and test reactors at the end of
each MCRT series and generates statistically sound experimental data.
The specific oxidant used in the first series of experiments is phenol
oxidizing products with a phenol substrate at 50 mg/L.
The MCRT is used as a blocking factor. The design calls for three
distinct MCRT series of 4, 8 and 12 days in succession for each reactor.
This gives a range of MCRT within a reasonable period of time for the
overall experiment. The treatment used in a reactor is changed when
the MCRT changes.
484
-------
LATIN SQUARE EXPERIMENTAL DESIGN
FOR THE PERFORMANCE PROTOCOL.
Reactor ID
MCRT
(days)
A
8
12
1
S
C
R
2
C
R
S
3
R
S
C
4
S
C
R
5
C
R
S
6
R
S
C
7
S
C
R
8
C
R
S
9
R
S
C
C • control
R - sludge reduction product
S - specific substrate oxidizing product
Figure 21. Latin Square Experimental Design for the Performance Protocol
485
-------
The study design is a repeated measures design with successive meas-
urements taken for each reactor under differing conditions. Each reactor
is run at the three MCRT in succession during an experiment. For a given
waste type, the experimental conditions are the same for all reactors in
operation at a given time, except the treatments will be different for
different reactors. Although the conditions of operation for the various
reactors are controlled to be as similar as possible, except for the
treatment, it is appropriate to consider the reactor to be a blocking
factor.
Quality Control — Routine performance testing quality control in-
cludes analytical quality control and reactor operation quality control.
Analytical quality control procedures are specified in the NSF Analyt-
ical Services Quality Control Manual and the NSF Chemistry Laboratory
Methods Manual. Reactor operation quality control procedures are speci-
fied in the NSF quality control document.
Batch Biodegradability Studies
The Batch Biodegradability Studies are performed to measure possible
enhancement of the specific substrate disappearance under aerobic and
anaerobic conditions through the use of substrate oxidizing or reducing
bioaugmentation products. In aerobic batch biodegradation testing, the
electrolytic respirometry approach in which oxygen uptake by the micro-
biota is used for indirect determination of percent biodegradation, and
the batch-shaker flask method are employed. The batch-shaker flask
method testing uses specific substrate analyses or indirect methods such
as TOG and/or DOC determination of culture samples.
With the electrolytic respirometry method, generated oxygen uptake
velocity data are used to calculate BOD values for respective incubation
times, at several incubation temperatures of culture media containing
the organic test compound. The percent biodegradation of the substrate
is determined for same-time intervals from the BOD and ThOD (theoretical
oxygen demand) ratio. In addition, oxygen uptake data as well as the
concentration of active biomass are used to develop BOD kinetic coeffi-
cients for determination of specific growth rates and biodegradation
rates of the substrate with or without the bioaugmentation products.
The electrolytic respirometry biodegradation testing experimental
design is outlined in Figure 22. Control biomass used in the studies is:
activated sludge; soil inoculum; or enrichment culture. The studies in-
corporate the use of controls with substrate without biomass, or controls
without substrate with either biomass and product together or with bio-
mass alone. The experimental reactor systems with the substrate oxidi-
zing product use substrate, biomass and the product at several concentra-
tion levels. The experimental reactor systems without the product will
incorporate reactors containing substrates and biomass at the same con-
centration levels for parallel monitoring of substrate disappearance in
reactors with or without the bioaugmentation product. The biodegradation
products are compared against: controls with substrate; controls without
486
-------
-pi
oo
Experimental Reactor I
Systems Without Product I
Biomass With
Product and
Without Substrate
Determination
of On Uptake
curvet
Calculation of
BOD Values
Detemrinallen ef
% Blodegradatlen
From BOrVThOD
ratio
Development ef
BOD Kinetic
Coefficients
Reators
1. Activated Sludge
2. SoU Inoculum
3. Enrichment Culture Inoculum
500 ml and 1,000 ml capacity
resproinetric flasks
Figure 22. Batch Electrolytic Respirometry Biodegradability Experimental Design „
-------
substrate; controls without product; activated sludge and/or soil inoc-
ulum of "equivalent dose"; and, enrichment cultures as inoculum of
"equivalent dose."
Synthetic medium (Table 13) used in the studies was formulated to
meet Organization for Economic Cooperation and Development (OECD) stan-
dards and consists of mineral salts, trace salt solution, vitamin
solution and/or yeast extract. In studies utilizing the cometabolism
approach to discern biodegradability of the substrate, vitamin and/or
yeast substrate serves as the primary substrate for growth and energy
source of microorganism. As that is depleted, the microbiota turn to
secondary substrate (test compound) for metabolic activity. In studies
using the test compound as sole carbon source, the primary substrates
are deleted from the synthetic medium.
The batch-shaker flask method, incorporating activated sludge or
soil or enrichment culture inoculum as control biomass in synthetic
medium, is an approach for testing biodegradation as developed by OECD
and modified at USEPA, Cincinnati, Ohio. Essentially the experimental
approach of the shaker flask batch system is the same as that for the
electrolytic respirometry studies. The flask batch method employs
Erlemeyer flasks (2-liter capacity) containing one liter volumes of
synthetic medium spiked with substrate at several concentration levels,
and biomass at several concentration levels, with or without the bio-
augmentation product. Three concentration levels of the product are
considered in the test. Activated sludge, soil or enrichment culture
inoculum serve as control biomass.
The batch-shake flask biodegradation testing experimental design is
outlined in Figure 23. At appropriate time intervals, samples of thor-
oughly mixed liquid media are analyzed lor residual specific substrate
and residual TOC/DOC in the system. For the same time intervals, levalg
of active biomaas are determined in the same gamplei te eitimate growth
rate biokinetic coefficients. The generated TOC/DOG data an w«ll ai
specific residual substrate data are used to develop substrate Ions
curves and to determine the percent by degradation of the substrate.
The generated data on the levels of active biomass are used to provide
growth rate curves and estimate biokinetic coefficients. The substrate
removal and growth rate data for the Shaker Flask Batch System are then
corelated with data generated from electrolytic respirometry studies to
make a final assessment of the rate of biodegradation of the substrate
in the reactors with and without the substrate oxidation enhancement
bioagumentation product.
FIELD TESTS
The NSF will not conduct field evaluations of bioaugmentation
products within the scope of this project. However, independent data
from treatment facilities already using bioaugmentation products will be
obtained and reviewed by NSF.
488
-------
TABLE 13. COMPOSITION OF SYNTHETIC MEDIUM
Manometrlc Respirometry (Rev 5)
1.6.1.2. Nutrient solution. •
The nutrient solution contains per litre, 10 ml of solution (a)
and 1 ml of each of the following solutions (b) to (f) in water
(1.6.1.1.) (A.R. means Analytical reagent)
(a) KH2P04 AR 8.50 g
K2HP04 AR 21.75 g
Na2HP04.2H20 AR 33.40 g
NH4C1 AR 2.50 g
dissolve in and made up to
1000 ml with water (1.6.1.1.)
The pH value should be
pH: 7.2
(b) MgS04.7H20 AR 22.50 g
dissolve in and make up to
1000 ml with water (1.6.1.1.)
(c) CaCl2 AR 27.50 g
dissolve in and made up to
1000 ml with water (2.6.1.1.)
(d) FeCl3.6P.20 AR 0.25 g
dissolve in and make up to
1000 ml with water (1.6.1.1.)
This solution is freshly prepared immediately before use.
(continued)
489
-------
TABLE 13. COMPOSITION OF. SYNTHETIC MEDIUM
Manometric Resplrometry (Rev 5) (con-
(e) Trace element solution
MnS04.4H20 AR "~ 39.9 mg (»30.23mg
H3B03 AR 57.2 ing
ZnS04.7H20 AR 42.8 rag
AR 34.7 mg
(=36.85mg (NH4)6M07024.4H2
Fe - chelate:
(FeCl3.EDTA) AR 100 mg
dissolve in and make up to
1000 ml with water (1.6.1.1.)
Sterilization of the trace element stock solution at (120 °C),
2 atm. 20 min.
(f) Vitamin solution
Biotin AR 0.2 mg
Nicotinic acid AR 2.0 mg
Thiamine AR 1.0 mg
p-Aminoben«olc *cid AR 1.0 mg
PantQthenie acid AR 1,0 mg
Pyridoxamine AR 5,0 rag
Qyaneeobalamine AR 2,0 n§
Folie seid AR 5,0 ng
in sn4 wake up te 10Q »j. wish wat§ff (l.i.l.
The sslHeien is filtered sttfile threugh 0,2 B Beabfane
Inaeasd of solution 1.6.1.2. (f) 15 ng of yeait extract say be
used per 100 ml of water (1.6.1.1.)
Solutions (e) and (f) may be omitted.
490
-------
*
Controls Without
Biomass
f
Substrate Without
Biomass
Controls Without
Substrate
Biomass Without
Substrate and
Product
Biomass With
Product and
Without Substrate
*
I
I Experimental Reactor 1
Systems With Product 1
Substrate
1 Substrate I
[ Cone. 1 H
| Substrate 1
1 Cone. 2 1
1 Substrate 1
Cone. 3 1
^
1 Cone" 1 |-
I Biomass 1
IConc. 2 1
1 Biomass 1
4 Cone. 3 1
t
Product
I Pr oducl 1
1 Cone. 1 I
I Product I
| Cone. 2 1
Ll Product |
1 Cone. 3 1
t
*
(Experimental Reactor 1
Systems Without Product _J
Substrate BiPJUIi
• L*L, *
Substrate
Cone. 1 I
Substrate
Cone, 2 I
Substrate I
Cone. 3 I
H Biomass i
Cone. 1 I
rEfomass i
1 Cone. 2 |
1 Biomass 1
U Cone. 3 1
1
J
-»•
-*•
-*
ouM be: Gyrotory Shaker
ed Sludge 2 Nter
culum 1 liter
tent Culture Inoculum
capacity Erienmyer Flasks-
of media
Final
Report
Correlate Data
With Resprirometric
Results
Measure Specific
Substrate Residual
at Appropriate
Sampling Times
Measure TOC/DOC
at Appropriate
Sampling Times
Calculate %
BiodegradaUon
From TOC/DOC Data
Determine Levels
of Active Biomass
to Develop growth
Rate Kinetics and
Growth Curves
Figure 23. Batch Shaker Flask Biodegradability Experimental Design
-------
CURRENT STATUS IN BIOAUGMENTATION EVALUATION STUDIES
Currently 12 bioaugmentation products are being tested by the Label
Verification testing protocol. The biocatalysts represent bacterial cul-
tures alone, bacterial cultures with nutrients, and mixtures of bacterial
cultures and enzymes. These bioaugmentation products, as reported by the
industries, serve as aerobic and anaerobic treatment enhancement addi-
tives, sludge handling aid products, in-sewer treatment enhancement
products, sludge settling aid products, deodorizers, septic tank treat-
ment enhancement products, and biocatalysts for the enhancement of the
treatment of phenolics, aromatics, petrochemicals (aliphatic), pulp and
paper mill wastes, dairy wastes, and animal and vegetable oils.
The successful development and confirmation of these testing proto-
cols will provide an approach for evaluating not only bioaugmentation
products currently produced by natural selection and in situ mutation,
but also those produced by genetic engineering.
REFERENCES
1. Davis, L.T. Focus on: New Technologies. Bioaugmentation: Assessing
the Possibilities. Waterworld News, p. 9-10, 1985.
2. Grubbs, R.B. Reducing Energy Needs Through Biotechnology. 5th
Annual Convention of the Hawaii Water Pollution Control Association,
January 28, 1983. 6 pp.
3. Grubbs, R.B. Environmental Applications With Biotechnology: The
Current State of the Art. The Genetic Control of Environmental
Pollutants Conference, University of Washington, July 31-August 3,
1983. 5 pp.
4. Grubbs, R.B. Value of Bioaugmentation for Operations and Main-
tenance of Wastewater Treatment Facilities Symposia Proceedings.
Wastewater Treatment Plant 0 & M. Conference Sponsored by the U.S.
EPA. Presented by The Water & Wastewater Technical School of Neosho,
Missouri in Atlanta, Georgia, July 9-11, 1974. 12 pp.
5. Allen, M. Correspondence to G.C. Bower of Bower Industries, Inc.
May 21, 1973.
6. Baig, N. and Grenning, E.M. The Use of Bacteria to Reduce the
Clogging of Sewer Lines by Grease in Municipal Sewage. In; Biological
Control of Water Solution. Ed. Tourbier and Pierson, University of
Pennsylvania Press. 1976. Flows Laboratories, Inc., Env. Cultures
Division, Implewood Co., 90301. p. 245-252.
492
-------
7. Bower, G.C. Bacteria - Their Role in the Sewage Treatment Process.
Joint Annual Meeting, Chesapeake Water Pollution Control Assn. and
The Water and Waste Operators Assn. of Maryland, Delaware, and
District of Columbia, Ocean City, Maryland. June 7-9, 1972.
8. Gardner, C. Bacterial Supplementation: Aids Wastewater Treatment.
Public Works Magazine, March 1980. p. 71-72.
9. Gasner, L.L. Chapter 10. Microorganisms for Waste Treatment. In;
Microbial Technology. Ed. H.J. Peppier and D. Pearlman, 2nd Edition,
Vol. 2, Academic Press, Inc., 1978.
10. Grubbs, R.B. The Role of Bacteria in Sewer Maintenance. Joint
Annual Meeting. Pesented to the Chesapeake Water Pollution Control
Assn. and The Water and Waste Operators Assn. of Maryland, Delaware,
and District of Columbia, Ocean City, Maryland, June 17-19, 1981.
10 pp.
11. Jensen, R.A. Control of Nuisance Odors From Ponds by the Use of
Bacteria Cultures. 83rd Annual Meeting of the American Institute of
Chemical Engineers, Houston, Texas, March 1977. 6 pp.
12. Kirkup, R.E. and Nelson, L.R. City Fights Grease and Odor Problems
in Sewer System. Public Works Magazine, October 1977. 2 pp.
13. McPhee, A.D. and Geyer, A.T. Aeration, Bottom Turbulence, and
Bacteriological Studies of Naval Ship Sewage Collection, Holding and
Transfer Tanks. Report 77-0043, April 1977, Available from Commander,
Naval, Ship Eng. Ctr., Washington, D.C., 20362.
14. Spraker, P. and Telepchak, M. Foam Control and Degradation of Nonionic
Detergents. Industrial Wastes, 26: No. 1, 16-67, January/February
1980.
15. Hoover, J.R. Montgomery County Environmental Planners Suggest
Bacterial Cleanup of Sewers. Montgomery County Planning Board,
Silver Spring, MD, July 12, 1974.
16. Grueninger, H., Sonnleitner, B. and Fiechter, A. Bacterial Diversity
in the Thermophilic Aerobic Sewage Sludge - III. A Source of Organisms
Producing Heat-stable Industrial Useful Enzymes, e.g. ot emylases
Applied Microbiol and Biotechnol. 19: 414-421, 1984.
17. Grubbs, R.B. Biotechnology is Taking its Place in Wastewater Treatment.
Presented at the Innovative and Alternative "Emerging" Technology
Seminars. Sponsored by the USEPA and Water and Wastewater Treatment
Equipment Manufacturers, 1981.
18. Huff, J.E. and Muchmore, C.B. Treatment of High Strength Fatty Amines
Wastewater. Journal of Water Pollution Control Federation. 54: No. 1
p. 94-102, January, 1982.
493
-------
19. Poduska, R.A. and Anderson, B.D. Successful Storage Lagoon Odor
Control. Journal WPCF 53; No. 3. p. 299-310, 1981.
20. Uziel, M. Energy Wasted to Overcme Hydraulic Head Losses as Result
of Biodegradable Deposits on Sewage Pipes Innerwalls. Report to
R.B. Grubbs of Flow Laboratories, Inc., August 8, 1979.
21. Bitondo, M. Cold Weather Waste Treatment. Industrial Wastes,
January/February, p. 29. 1980.
22. Chambers, J.V. Bioengineering an Activated Sludge Microflora to
Improve Waste Removal Performance. In: Proceedings of the 5th
Annual Industrial Pollution Conference of the Water and Wastewater
Treatment Equipment Manufacturing Association, Atlanta, Georgia,
April 1977. 10 pp.
23. Chambers, J.V. Improving Waste Removal Performance Reliability of
a Waste Treatment System Through Bioaugmentation. In: Proceedings
of the 36th Industrial Waste Conference, Purdue University, West
Lafayette, Indiana, p. 631-643. 1981. Published 1982.
24. Willson, G.B., Parr, J.F., Taylor, J.M., and Silora, L.J. Land
Treatment of Industrial Wastes. Principles and Practices (Part I).
Biocycle Journal of Waste Recycling 23; No. 1, 37-42, January/February
1979.
25. Worne, H.E. The activity of Mutant Microorganisms in the Biological
Treatment of Industrial Wastes. Paper presented at Aqua Sana Lenten-
beurs van Vlaanderen, Ghent, Belgium, 1972. March 7, 1972. 11 pp.
26. Zitrides, T.G. Mutant Bacteria for the Disposal of Hazardous Organic
Wastewaters. Presented at the Pesticide Disposal Research and Develop-
ment Symposium. Sponsored by the U.S. EPA Municipal Enironmental
Research Laboratory. Solid and Hazardous Waste Research Division.
Reston, Virginia. September 6-7, 1978. 8 pp.
27. Coe, R.H. Bench-Scale Biological Oxidation of Refinery Wastes With
Activated Sludge. Sewage and^Industrial Wastes 24;731, June 1952.
28. Crame, L.W. Pilot Study on the Enhancement of the Refinery Activated
Sludge Process. API Publication 953. October 1977. p. 953.
29. Reitano, A.J. Start-up and Operation of a Refinery Activated Sludge
Plant. In; Proceedings of 36th Annual Purdue Industrial Waste
Conference, May 12-19, 1981. p. 310-319. Published 1982.
30. Thibault, G.T., Steelman, B.L. and Tracy, K.D. Enhancement of the
Refinery Activated Sludge Process With Powdered Activated Carbon.
Paper presented at 69th Annual Industrial Pollution Conf., Water and
Wastewater Equip. Mfrs. Assn., St. Louis, MO, April 1978.
494
-------
31. Tracy, K.D. and Zitrides, T.G. Environmental Management - Mutant
Bacteria Aid Exxon Waste System. Hydrocarbon Processing, October
1979. 4 pp.
32. McDowell, C.S. and Zitrides, T.G. Improved Biological Wastewater
Treatment Through Biomass Engineering. Presented at the 1979 Annual
Meeting of the California Water Pollution Control Assn. San Diego,
California, April 25-27, 1979. 6 pp.
33. Tracy, K.D. and Shah, P.S. Application of Bacterial Additives to
Refinery Activated Sludge. 50th Annual Meeting of the California
Water Pollution Control Assn., Sacramento, California, Friday, April
21, 1979. 8 pp.
34. Christiansen, J.A. and Spraker, P.W. Improving Effluent Quality of
Petrochemical Wastewaters With Mutant Bacterial Cultures. In; Proc.
of 37th Annual Purdue Ind. Waste Conf. 37:567-576, 1982. Published
1983.
35. Hughes, D.E. and McKenzie P. Microbial Degradation of Oils and Petro-
chemicals in the Sea. In; Microbiology in Agriculture, Fisheries,
and Food. (ed) F.A. Skinner & J.G. Carr. Society for Applied
Bacteriology Symposium Series 4, 1976. p. 91-107, Academic Press,
London.
36. Thibault, G.T. and Zitrides, T.G. Biomass Engineering of Petro-
chemical and Refining Wastewaters. Presented at the 1979 Annual
Mid-Atlantic Industrial Waste Treatment Conference Penn State Uni-
versity, University Park, Pennsylvania, July 16-20, 1979.
37. Thibault, G.T. and Tracy, K.D. Demonstration of a Mutant Bacterial
Additive for Enhancement of Operational Stability of Oxygen Activated
Sludge. In; Proceedings of 34th Annual Purdue Industrial Waste
Conference, Purdue University, West Lafayette In, May 8-10, 1979.
p. 235-243. Published 1980.
38. Grubbs, R.B. Effect of a Bacterial Culture Product on Biological
Kinetics. Journal WPCF, 54; No. 12, 1614-1616. December 1982.
39. Quasim, S.R. and Stinehelfer, M.L. Effect of the Bacterial Culture.
Product of Biological Kinetics. Journal WPCF, 54: No. 3, p. 255-260.
March 1982.
40. Adamse, A.D. Bacteriological Studies on Dairy Waste-activated Sludge.
Meded. Landb Hogesch. Wageningen, 66: No. 6, p. 1-79. 1967.
41. Chambers, J.V. and Harper, W.J. Loading Characteristics Affecting
the Performance of a Model Dairy Activated Sludge System. In;
Proceedings of the Cornell Agricultural Waste Management Conference.
Food Processing Waste Management, p. 260-273. 1973.
495
-------
42. Chambers, J.V. Improving Waste Removal Performance Reliability of a
Waste Treatment System Through Bloaugmentation. Water and Wastes
Digest, January/February 1984. 22 pp.
43. Edmonson, J.E. and Brazis, A.R. Microbial Fermentation of Dairy
Wastes. Journal of Dairy Science 38; No. 6, 597. 1955.
44. Harper, W.J., Blaisdell, J.L. and Grosskopf, J. Dairy Food Plant
Wastes and Waste Treatment Practices. A State-of-the-Art Study for
the Quality Office of the Environmental Protection Agency. Wat.
Cont. Res. Series 12060 ECU 3/71, 1971.
45. Lembke, A. Biological Basis for the Removal of Dairy Waste Water.
Kieler milchw. ForschBer. 8: No. 3, 305-375. 1956.
46. Barnes, T.S. The Effectiveness of Mutant Bacteria in Pulp and Paper
Wastewater Treatment System. Georgia Kraft Company, Krannert Division
Report. 1981.
47. Blosser, R.O. Technical Bulletin No. 438 - Evaluation of Bacterial
Augmentation of an Activated Sludge Process. National Council of
the Paper Industry for Air and Stream Improvement, Inc. July 11,
1984. 32 pp.
48. Davis, L.; Blair, J.E.; and Randell, C.W. Development of Color
Removal Potential in Organisms Treating Pulp and Paper Wastewater.
Journal WPCF, 50: No. 2. p. 382-386, February 1978.
49. Zitrides, T.G. Mutant Bacteria Control Filamentous Growth in Mill
Wastewater Treatment. Pulp and Paper, February 1980. 3 pp.
50. Nichols, W.E. The Use of Freeze-Dried Microorganisms for Bulking
Control and Operational Improvements. Appleton Papers, Inc. Pre-
sented at NCASI Central-Lake States Regional Meeting, September 19-20,
1978.
51. Dougherty, E. Mutant Bacteria Undergo Field Evaluation - A First in
North America Could Mean Cleaner Water for Alberta Town of Peace
River. Pulp and Paper Canada, 83: No. 4, p. 32-33. 1982.
52. Bradford, H.T.; McDowell, C.S. and Zitrides, T.G. Mutant Bacteria
Improve Wastewater Treatment Performance. Presented at Florida
Section American Waterworks Association & Florida Pollution Control
Assn., Miami Beach, Florida, November 5-8, 1978.
53. Davies, B.T., et al. Custom Designs Cut Effluent Treatment Costs:
Case Histories of Chevron USA, Inc. In; Proceedings of the 32nd
Industrial Waste Conference, Purdue University, Ann Arbor, MI. Ann
Arbor Science Publishers, Inc. 1978.
54. Himebaugh, R.R. The Use of Filabac-M in Municipal Wastewater Treatment
Polybac Corporation, November 1981. 22 pp.
496
-------
55. Horsfall, F.L. III. Biochemical Augmentation of Wastewater Treatment
Deeds & Data. Journal WPCF, February 1977. 2 pp.
56. McDowell, C.S. Biomass Engineering Improves Wastewater Treatment
System Performance. Presented at the 39th Annual Meeting Inter-
national Water Conference, Pittsburgh, Pennsylvania, October 31-
November 2, 1978.
57. Nyer, E.K. and Bourgeoig Jr., H.J. Operational Trouble Shooting in
Industrial Biological Treatment Systems. 35th Industrial Waste
Conference, Purdue University, West LaFayette, Indiana, May 13-15,
1980. p. 849-854. Published 1981.
58. Saunders, F.J. Biotechnology and Waste Treatment. Effluent and
Water Treatment Journal. 24: No. 11, pp. 421-425, November 1984.
59. Sawada, H., Parr, R.C. and Rogers, P.L. Photosynthetic Bacteria in
Waste Treatment. J_. Ferment. Technology, 55: No. 4, 326-336, 1977.
60. Stinehelfer, M.L. The Efects of a Bacteria Culture Product on a
Continuous Activated Sludge Process Using a Laboratory Scale Model.
M.S. Thesis, The University of Texas at Arlington. December 1974.
61. Thibault, G.T. and Tracy, K.D. Controlling and Monitoring Activated-
Sludge Units. Chemical Engineering. September 11, 1987. p. 6.
62. Young, R.A. Mutant Bacteria Gives Fast Start-up for New Process.
Pollution Engineering Magazine, 11: No. 7, 34-35, July 1979.
63. Zitrides, T.G. Using Customized "Bugs" for Biological Waste Treatment.
Plant Engineering, June 23, 1977. 3 pp.
64. Zitrides, T.G. and Nyer, E.K. The Role of Bacteria in Control and
Destruction of Hazardous Organic Wastes. Presented at the AIChE
Cleveland Conference, August 1982.
65. Ghosh, S. et al. Anaerobic Sludge Digestion with a Biocatalytic
Additive. In; Proceedings of the 55th Annual Conference Water
Pollution Control Fed., St. Louis, MO. 1982.
66. Longaker, J.J. Liquid Mixed Culture Bacteria Application to Anaerobic
Digesters. The Tennessee Valley Authority Office of Power. Office
of National Resources, Chattanooga, Tennessee. April 1982. p. 1-38.
67. Straley, P.J. Bioaugmentation of Anaerobic Digesters. General Info.
and Bibliography. 1981.
68. Torien, D.F. Population Description of the Nonmethanogenic Phase
of Anaerobic Digestion - Isolation Characterization and Identifica-
tion of Numerical Important Bacteria. Water Research, 4: 129-148.
1970.
497
-------
69. Traurig, J.C. Anaerobic Digestion. Pollu-Tech., Inc. Special
Publication, Chalfort, PA, 1983.
70. Bryan, A.C. How Enzymes Improved Sludge Digestion. Public Works
83: No. 12, 69, 1952.
71. Grune, W.N. and Sload, R.Q. Biocatalysts in Sludge Digestion.
Sewage Industrial Wastes 26: p. 1425, 1954.
72. Himebaugh, R.R. and Hentrich, Jr., R.L. Waste Sludge Reduction by
Selectively Adapted Mutant Bacteria. Residuals Management of the
55th Water Pollution Control Federation, St. Louis, Missouri.
October 3-8, 1982. 15 pp.
73. Pearson, E.A. et al. Biocatalytic Additives in Sludge Digestion.
Sew. Ind. Wastes 29: 1066, 1957.
74. Wells, W.N. and McKinney, R.E. Plant-Scale Test of a Biocatalyst
on Sludge Digestion. Sew. Ind. Wastes 27; 871, 1955.
75. Hasit, Y. Optimization of Municipal Sludge Handling and Disposal
Systems. Unpublished Ph.D. Dissertation, Duke University, Durham,
NC. October 1978.
76. Hasit, Y. and Vesilind, P.A. Economics of Sludge Management. Journal
WPCF, 53: No. 5, 560-564, May 1981.
77. Nakasaki, K.; Sasaki, M.; Shoda, M. and Kubota, H. Effect of Seeding
During Thermophilic Composting of Sewage Sludge. Applied and Environ-
mental Microbiology, 49: 724-726, March 1985.
78. Blair, J.; Christenen, J.; Spraker, P and Telepchak, M. Improved
Settling of Wastewater in the Activated Sludge Process Using Bacterial
Augmentation. Tappi 65: No. 4, 128-130, 1982.
79. Carlson, Carl-Goran Improved Filtration of Biosludges by Enzyme
Treatment. Filtration & Separation, p. 82-86, January/February 1979.
80. Christiansen, J.A.; Dohner, B. and Marcheski, R. Achievement of
Complete Nitrification in Rendering Waste Through Use of Mutant
Cultures - Food Processing. Sybron Biochemical, Chemical Division.
Reprinted from Water Pollution Control Association of Pennsylvania
Magazine. March/April 1982. 4 pp.
81. Poduska, R.A. A Dynamic Model of Nitrification for the Activated
Sludge Process. Doctorial Thesis, Clemson University, 1973.
82. Atlow, S.C.; Bonadonna-Aparo, L. and Klibanov, A.M. Dephenolization
of Industrial Wastewater Catalyzed by Polyphenol Oxidase. Biotech-
nology and Bioengineering, 26: 599-603, June 1984.
498
-------
83. Carroad, P.A. and Wilke, C.R. Enzymes and Microorganisms in Food
Industry Waste Processing and Conversion to Useful Products: A
Review of the Literature. Resource Recovery and Conservation 3,
165-178, 1978.
84. Knapp, J.S. and Howell, J.A. Treatment of Primary Sewage Sludge
With Enzymes, Biotechnology and Bioengineering, 20: 1221-1234, 1978.
85. Lamed, R.J.; Keinan, E. and Zeikus, J.G. Potential Application of
an Alcohol-aldehyde/ketone Oxidoreductase from Thermophilic
Bacteria. Enzyme Microbiol. Technol, 3: 144-148, 1981.
86. Ray, M. Modesto Experiments with Enzyme Catalyst for Industrial
Waste. Western City, 29. No. 4, 52, 1953.
87. Robinson, R.R. Enzymes Give Good Results in Sewage Treatment Plant.
Public Works. 85: 116, 1954.
88. Rudolfs, W. Enzymes and Sludge Digestion. Sew. Works J., 4: 782,
1932.
89. Slote, L. Development of Immobilized Enzyme Systems for Enhancement
of Biological Waste Treatment Processes. Report No. 5501-0113, U.S.
EPA, Washington, D.C., 1970.
90. Wojnowska-Baryla, I. and Young, J.C. Measuring the Effect of Bio-
catalytic Additives on Treatment Process Performance. Journal WPCF.
55: No. 11., 1373-1380, November 1983.
91. Young, J.C. The Use of Enzymes and Biocatalytic Additives for Waste-
water Treatment Processes. Water Pollution Control Federation
Highlights. Deeds & Data 13: May 1976. 5 pp.
92. Rittman, B. The Relationship Between Genetic Control and Process
Control Strategies in Municipal Waste Treatment. In; "Proceedings
of Bioexpo 85" an International Conference in Biotechnology. Boston,
Mass. Cahners Exposition Group. 1985. p. 401-409.
93. Tabak, H.H.; Quave, S.A.; Mashni, C.I. and Barth, E.F. Biodegrad-
ability Studies with Organic Priority Pollutants. Journal WPCF
53: No. 10, 1503-1518, October 1981.
94. Tabak, H.H. and Hannah, S. Static Procedure for Biodegradability
Determination. Internal EPA Report, 1979.
95. Tabak, H.H.; Barth, E.F. and Mashni, C.I. Biodegradation Studies of
Carboxy Methyl Tartronate. Internal EPA Report. EPA-600/2-78-115,
1979.
96. Tabak, H.H. and Hannah, S. Continuous Flow Procedure for Biodegrad-
ability Determination. Internal EPA Report, 1979.
499
-------
MUNICIPAL SLUDGE OXIDATION WITH THE
VERTICAL TUBE REACTOR
by
Gilbert B. Morrill
McCall-Ellingson & Morrill, Inc.
Denver, Colorado 80218
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
501
-------
MUNICIPAL SLUDGE OXIDATION WITH THE
VERTICAL TUBE REACTOR
by: Gilbert B. Morrill
McCall-Ellingson & Morrill, Inc.
Denver, Colorado 80218
ABSTRACT
The first full-scale vertical tube reactor (VTR) was built at Longmont,
Colorado in 1983 for wet oxidation of municipal wastewater treatment
sludge. The VTR consists of coaxial pipes about 2400 meters long suspended
vertically in a deep well.
The concentric annular spaces form very long U-tubes. Sludge and air
are pumped down through an annular space and return up through an adjacent
annular space.
When preheated to a sufficiently high temperature, wet oxidation
occurs, destroying up to 80% of the chemical oxygen demand (COD) of the
sludge. Readily settleable ash and short chain organic acids are formed.
The clarified VTR effluent liquor is returned to the municipal wastewater
treatment plant where it is treated through the existing biological
secondary treatment units.
Temperatures above 260°C can be attained at the bottom of the VTR
without boiling because of the high hydrostatic pressure developed by the
column of liquid. Higher temperatures produce greater percentage
reductions of the COD. The wet oxidation reaction is autogenic after the
reactor is initially heated using an external heat source.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved
for presentation and publication.
502
-------
INTRODUCTION
The City of Longmont, Colorado entered into cooperative agreements with
the United States Environmental Protection Agency (USEPA) and a private
industry (VerTech) in 1982. By these agreements, a prototype vertical tube
reactor (VTR) for wet oxidation of the City's municipal wastewater
treatment plant sludge was designed and constructed by VerTech's
predecessor in 1983. After initial operation, VerTech redesigned and
retrofitted the reactor in 1984. Demonstration of the VerTech System with
USEPA grant assistance is scheduled to be completed by December 1985.
Prior to the construction of the prototype VTR, VerTech's predecessor
had developed and operated a bench scale laboratory reactor and a small 2"
pilot reactor which they constructed in an abandoned oil well. The results
of this earlier work encouraged Bow Valley, VerTech's parent company to
finance the construction of the prototype facility. Early operating
results further encouraged Bow Valley to acquire the process patent rights
and to form VerTech Treatment Systems to commercialize the process.
The City of Longmont, Colorado was chosen as the site for the prototype
installation for a number of reasons. The City is located thirty miles
north of Denver, a short travel distance for private industry. Due to
rapid growth, the City was facing a growing sludge disposal problem. The
City with a population of 46,000 has a representative combination of
residential, commercial and industrial users of the City's municipally
owned wastewater treatment system. The size of the reactor the industry
wanted to build for the demonstration had the potential to become the
primary means of treatment for all of the City's sludge. The City is
within easy haul distance of a number of diverse industrial waste producers
so that the VTR at that location may be used to demonstrate the process to
handle a variety of wastes. As an additional imoortant consideration,
the City staff and elected officials were enthusiastic about being a part
of a cooperative effort between the local community, VerTech, and the USEPA
to develop and demonstrate this innovative process.
Under the cooperative agreements, the City provided the land at its
wastewater treatment plant, plant modifications to accommodate the VTR,
laboratory facilities and personnel, prepared the grant application, and
administered the grant funds provided by USEPA.
VerTech designed and built the VTR at its own expense and furnishes the
personnel to operate it.
The USEPA is furnishing grant funds to help offset the City's costs and
VerTech's cost of operation during the demonstration period.
503
-------
The firm of McCall-Ellingson and Morrill, Incorporated, is the City's
consulting engineer in this project which has been designated as
principal investigators to prepare the final report.
THE VERTICAL TUBE REACTOR
VerTech's below-ground system (Figure 1) is housed in a
conventionally drilled, cement encased well. Tubes of various diameters
are concentrically arranged within the well to create annular spaces for
two-phase flow and heat exchange.
The influent, composed of diluted liquid sludge and air, is pumped
into the inner annular space under low pressure. The influent stream is
pressurized to 9.65 MPa (1400 psi) at the bottom of the well by the
height and density of the fluid column above. Oxidation takes place when
oxygen and organic constituents are present at a sufficient temperature.
Temperatures above 260°C (500°F), can be attained without boiling because
of the high hydrostatic pressure that is developed. The result is wet
air oxidation with a dramatic reduction of organic solids to a small ash
residue. Excess heat liberated during the process is transferred to the
surface by an internal hot oil heat exchanger.
Tube diameters and overall lengths are designed to provide sufficient
residence time to complete the oxidation process. The high
length-to-diameter ratio of 30,000:1 provides efficient counterflow heat
exchange, mass transfer, and plug flow to maximize chemical reaction.
A significant advantage of a long reaction vessel installed in a deep
well is the influent sludge and air or oxygen can be introduced near the
top of the reaction chamber at nominal pressure of 2.1-3.4 MPa (300-500
psi) with maximum pressures equal to 9.7-10.3 MPa (1400-1500 psi)
developed at the bottom of the reaction chamber due to hydrostatic
pressure of the column of waste.
THE CITY OF LONGMONT FACILITY
Figure 2 is a flow diagram of the Longmont installation. Sludge from
the City plant is stored in a mechanically mixed holding tank from which
it is pumped at a controlled rate to the vertical tube reactor.
Diluent in the form of City plant process water, City plant effluent
or reactor effluent recycle is mixed with the sludge to produce the
desired influent concentration to the reactor.
504
-------
MEAT TRANSFER OIL
nuwt-oo Min
R OXYGEN
IS
iS®$£:
• • • •,'*••'•'"..
'/'< K iff /'#)*> J.
•$
,
£
'I
"*J
V
^
~
^
j
"1 IRFQ^ • i .. '.
uoc. o / _
TUbc/ 5
RETE *
£
i
—
^~
; \
i
^V
t
t
1
t
,j-^
11
t
i
1
u
^~ *•'
X i»*
v •:
v?
v 'i'.«
V 4
^i
$1
S t.'[
S &
v 5v
V >')
ii
s! J*»*
5£
i^ *
~~ A*-
^ V
^ .7*
il
i ^ '\
'! "" .^*
; i
v '-r
I
\
\
I
*.'
\
i
;L'
U^J
^
\
::.
.',
^•iSj;^
•*- INFLUENT
1
I
i
V
1
I
1
I
-»• EFFLUENT
f
I
.^
-«-•
t
1
jr
^^^^Sil^
"T
V^
• ".;'•; I".'.'- i '..'•>.•'.'•';'
'•.'•/^•CONCRETE ':'•..»
"'"***"" *. " * ^* • " *.•'»"
-***'.'.*** "••*." '*'•
^
|o^'^^*:?
•*•/*»,* • •""'l S*t« f « " • *
^VeV'^'^'^WH
[
r
^
'/
^^^~*~ ^^fTT
"» /^^^
^ /f^
1 li vvs
xkx:
5- x%^
f ^^
s
^ TYPICAL
_
-?
2.
7
~4
'2
13 V8" CASING
10" TUBE
L-INSULATED TUBULAR
Figure 1. Longmont reaction vessel
down-hole cross sectional schematic.
505
-------
(Ji
O
LONGMONT
SLUDGE FEED
SLUDGE
GRHOER
DILUENT te--OO
- — J
q ^'
j , r
AGIO WASH BYPASS '
-j-D<3-sr^
MAIN
FEED PUMP
\CID WASH
ROM AGIO TANK
EFFLUENT TO CITY WASTE
TREATMENT PLANT
REACTION
VESSEL
Figure 2. Process flow diagram.
-------
The reactor feed pump is a positive displacement variable speed pump
which is used to control the flow rate through the reactor. The sludge
pump is also a positive displacement variable speed pump by which the
rate of sludge feed to the reactor is controlled. The diluent pump is a
centrifugal constant speed pump. The diluent delivered to the system
makes up the difference between the volume pumped by the sludge pump and
the volume pumped by the reactor feed pump. The amount of chemical
oxygen demand (COD) in the feed to the VTR ,is adjusted simply by changing
the speed of the sludge pump.
Compressed air was initially used as the only source of oxygen. An
oxygen source (Figure 3) has since been added so that oxygen-enriched air
can be provided. The oxygen source is introduced at 30 or 122 meters
(100 or 400 feet) from the surface. Temperature and pressure to start
the oxidation reaction is developed at about 300 meters (1,000 feet)
below the surface.
A heat exchanger is provided to add heat to or extract heat from the
reactor. Heat transfer oil is circulated through the heat exchanger
which extends essentially the full depth of the reactor (Figure 1). Heat
can be added by a natural gas fired heater or wasted through a fan
powered radiator to the atmosphere.
Energy recovery has been demonstrated at Longmont, but the excess
heat provides no beneficial use.
For startup of the reactor it is necessary to add heat to bring the
reactor up to operating temperature and to heat the rock surrounding the
reactor. At Longmont it was necessary to add heat continuously through
the early stages of the demonstration project. In fact, the system did
not operate autogenically until after the oxygen-enrichment system was
added.
Effluent leaves the reactor at a temperature of about 11-14C0 above
the influent temperature. Much of the reaction zone heat is given up to
preheat the incoming waste. The VTR effluent is routed through a gas
separator and a solids separator and is then returned to the City
wastewater treatment plant.
Ash from the bottom of the separator is pumped as slurry to ash
pits. Clarified liquor is decanted and returned to the City wastewater
treatment plant. Settled ash is periodically removed and taken to the
City landfill or other ultimate disposal. The ash is essentially inert.
The combination of wet air oxidation to remove biological suspended solids,
coupled with ash settling to 40 wt%, provides for a 25-fold reduction in
solids volume.
507
-------
cn
o
CO
-CRYOGENIC
PUMP
ATMOSPHERIC
VAPORIZERS
HOLDING TANK
AIR-
<"':V.-:U'-.X:.A
-INFLUENT
-EFFLUENT
Figure 3. Oxygen-enriched air flow diagram.
-------
The liquor returned to the City wastewater treatment plant contains
low molecular weight organic acids (C] to 04), primarily acetic
acid. The strength of the liquor varies from 2,000 mg/1 to 6,000 mg/1 of
COD, depending on the reactor feed concentration. The liquor strength
varies directly with the VTR loading. The COD percentage reduction
increases with inlet concentration and with higher reactor operating
temperature, as shown on Figure 4. Higher operating temperatures also
form shorter chain acids in the reactor effluent.
Provision is made to recycle clarified VTR effluent to partially or
totally replace the other diluent sources. Recycling, initially
practiced to conserve process heat, appears to reduce the organic acids
to shorter chain acids and also to oxidize some of the ammonia formed in
a single pass operation.
VerTech has operated bench scale completely mixed activated sludge
bio-reactors at the Longmont facility from August, 1984 through May,
1985. This has confirmed the treatability of the clarifier reactor
effluent. The reduction of biochemical oxygen demand (BOD) and COD
through the VTR and bench scale bio-reactor, is shown by the schematic
diagram on Figure 5. The overall BOD reduction from the diluted
municipal sludge at 11,800 mg/1 to the bench scale bio-reactor effluent
of 76 mg/1 is greater than 99%.
There is no reason to believe the treatment through the municipal
plant is not comparable to that through the bench units during steady
state runs of the VTR. Abrupt changes in the operation of the VTR do not
significantly affect the performance of the municipal plant. However,
when the VTR is shut down to acid wash the reactor, the municipal
secondary units which are trickling filters and rotating biological
contactors, slough the bio-mass built up during the steady state
operation when the VTR return flow is providing additional nutrients to
the biological system. Sloughing results in an increase in the municipal
plant effluent suspended solids and total BOD, but does not greatly
affect the effluent dissolved BOD.
PROCESS CAPACITY
The Longmont VTR has a 25.4 cm (10") diameter outside reactor tube
which with the 17.8 cm (7") and 12.7 cm (5") concentric tubes form a
reaction chamber with a total volume of about 41.6 cubic meters (11,000
gallons). About 60% of this total volume or 25 cubic meters (6600
gallons) is in the reaction zone where the temperatures are high enough
for wet oxidation to proceed. At a diluted sludge flow design rate of
0.45 cubic meters per hour (120 gallons per minute) and a gas volume
fraction in the two-phase stream equal to 50%, the residence time in the
oxidation zone is about 45-50 minutes,taking into consideration the
compression of the gas phase.
509
-------
019
C
(D
o
o
-h
O
3
O
O
o
C
n
% COD REDUCTION
(n-g->i-J-g-NiCDCD(DOOGoCDCDU>
co
&
en
o
O)
o
I
o
-------
C00 = 50,400
800=35,300
DILUTION WATER
COO =16, 800
C00 = 2,700
300 = 11,800
t
VTR
i
800 = 2,230
Figure 5. Bench scale completely mixed activated sludge COD and BOD removal (mg/L),
-------
Using air as the oxygen source, the Longmont VTR has normally treated
4500 kilograms (10,000 pounds) per day of COD or 4100 kilograms (9,100
pounds) per day of dry sludge solids. At a 4500 kilogram per day rate,
it was necessary to continuously add heat to the system to overcome heat
losses to the rock and to replace heat flushed out with the reactor
effluent.
The equipment to provide oxygen for air enrichment was added in
October 1984. With oxygen-enriched air, the influent COD strength can be
significantly increased. Influent streams with strength ranging from
20,000-45,000 mg/1 COD have been treated. These compare to the 10,000
mg/1 COD using air alone. With the higher strength influent more heat is
generated per unit volume of influent oxidized, permitting the system to
operate autogenically.
Figure 6 shows VerTech's projections of capacities of two different
sized reactors. The Longmont reactor is 10" in diameter and the
capacities indicated are from operating experience with that facility.
The capacities of the 6.3 cubic meters (16") reactor is the result of
model analysis by VerTech.
80-
6O-
40-
20-
80
Q
UJ
UJ o
-
20
18
cm
10" 16"
REACTION VESSEL DIAMETER
AIR
OXYGEN ENRICHED AIR
Figure 6. Capacity vs diameter effect with oxygen.
512
-------
One of the early concerns was that the Longmont reactor might prove
to be too small to serve the total sludge disposal needs of the Longmont
Wastewater Treatment Plant. Indeed, until the oxygen-enrichment
equipment was added this appeared to be the case. The wastewater
treatment plant was producing sludge in amounts up to 4000 kilograms
(8800 pounds) per day of COD prior to the start up of the VTR. The
return flow from the VTR increased the wastewater plant sludge production
by 25-30%. Sludge production for the twelve months that the VTR has been
in operation has averaged 4990 kilograms (11000 pounds) per day of COD
and was nearly 6350 kilograms (14,000 pounds) per day for the entire
month of April 1985. The maximum loading rate on the VTR with air only
was 6530 kilograms (14,400 pounds) per day of COD. This would about
handle the present City plant needs but would not allow for any growth.
Growth in the City is occurring at a 6% annual rate.
With oxygen-enrichment of the air supply the reactor has been
operated with loads of 13600-20400 kilograms (30,000-45,000 pounds) per
day of COD. Autogenic operation has been sustained at the lower loading
rate but that is still more than double the present sludge production of
the City wastewater treatment plant. The Longmont VTR which was
initially feared to be too small, is actually too large when operated
with oxygen-enriched air.
An operational solution to this pleasant problem has been to
accumulate sludge for several days and then run the reactor for several
days until the sludge supply is depleted.
To operate the reactor at the higher loading rates for a short time,
VerTech has hauled sludge by tank trucks from two nearby municipal
wastewater treatment plants.
COST OF OPERATION
Before autogenic operation was achieved, the largest operating cost
of the VTR was the cost of fuel to maintain reactor temperature. With
autogenic operation one of the largest operating costs for the VTR is the
periodic nitric acid wash of the reactor for scale control. The buildup
of scale produces an increased roughness of the pipe walls which causes
greater friction losses, which in turn requires increased pumping
pressures to maintain a given flow rate through the VTR. The scale does
not appear to significantly reduce the cross section area of the flow
path and acid washing restores lower pumping pressures. Acid washing to
date has been performed after about 200,000 kilograms (220 tons) of COD
processed.
513
-------
Under present conditions at the Longmont wastewater treatment plant,
which has very shallow final clarifiers, the wastewater plant operators
have to feed aluminum sulfate as a coagulant aid to avoid violation of
their plant effluent limitations. The aluminum ends up in the plant
sludge. Part of the scale build up in the sludge reactor is aluminum
phosphate and aluminum silicate.
Improvements now under construction at the wastewater treatment plant
will eliminate the need to feed aluminum sulfate. This is expected to
reduce the reactor scaling problem. If in fact it does, the cost of acid
washing the reactor for scale control will be reduced.
Some of the pumping pressure build-up which was first thought to be
totally the result of scaling has been found to be partially due to
organic fouling which does reduce the cross-sectional areas in the top
section of the reactor downcomer. Hydraulic backflushing of the reactor
without acid has produced reductions in the pressure build-up by as much
as 80 psi. If the pumping pressure build-up can be even partially
controlled by simple back flushing, the frequency and cost of acid
washing will be further reduced.
The cost of operation of the Longmont facility under steady state
autogenic operating conditions is not yet available; however, VerTech is
very optimistic that they can achieve operating costs of $200 or less per
dry ton of solids treated. The VTR will be quite competitive with
conventional methods of municipal sludge disposal if this estimated
operating cost is confirmed.
The capital cost of the Longmont VTR is not a good measure of what
the commercial price of similar units will be. It is a prototype
designed to demonstrate the technology and operability of the process.
Capital cost estimates will be included in the final report prepared for
the City and the USEPA.
SUMMARY
The USEPA funded operation of the Longmont facility is essentially
complete; however, VerTech plans to continue their operation and study at
Longmont and we will continue to receive data. What we know at this time
is:
o The Longmont VTR is a technical success. Municipal sludge
is wet oxidized; more than 75% of the COD is destroyed
in the sludge feed.
o The VTR effluent contains organic acids which may be
returned to and be readily treated through the municipal
wastewater plant secondary treatment units.
514
-------
The ash is easily separated and concentrated by gravity
settling and may be disposed of by a number of acceptable
methods because it is essentially inert.
The Longmont VTR with oxygen-enriched air operates
autogenically. Such operation is essential for the process
to be cost competitive with other sludge disposal methods.
The Longmont VTR with oxygen-enriched air can accept the
total sludge production of the Longmont municipal
wastewater treatment plant with substantial reserve
capacity for growth of the City.
Operating costs and capital costs are still being analyzed;
however, VerTech is projecting operating costs at Longmont
of $200 per ton of dry solids processed.
ACKNOWLEDGEMENT
The author has been in constant communication with the Engineering
Staff of VerTech Treatment Systems and with Howard Delaney of the City of
Longmont during the preparation of this paper. Their generous sharing of
their time and knowledge is gratefully acknowledged.
515
-------
MUNICIPAL WASTEWATER TREATMENT USING THE CAPTOR PROCESS
by
James A. Heidman
U.S. Environmental Protection Agency
Water Engineering Research Laboratory
Cincinnati, Ohio 45268
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati Ohio
517
-------
MUNICIPAL WASTEWATER TREATMENT USING THE CAPTOR PROCESS
by: James A. Heidman
U.S. Environmental Protection Agency
Water Engineering Research Laboratory
Cincinnati, OH 45268
ABSTRACT
The CAPTOR process is a variation of the activated sludge process
in which small porous polyurethane pads are added to the reactor to provide
surface area for biomass colonization. Effluent screens retain the pads
within the reactors. The pads are periodically cleaned by air-lifting
them onto a conveyor where they pass through rollers providing a tight
compression. Cleaned pads are returned to the reactor and a seperate
waste sludge stream is produced. This patented system is marketed by
Simon-Hartley Ltd. and originated from studies at the University of
Manchester, England. It is under evaluation on municipal wastewaters in
Freehold and Stevenage, England and the U.S.EPA Test and Evaluation Facility.
The original optimistic claims for process performance and advantages
have not been realized to date. A pilot study of CAPTOR at Downingtown,
PA eliminated the system from consideration as an upgrading approach there.
The system at Freehold has undergone several modifications to obtain more
uniform pad distribution and mixing and thusfar the effluent quality has
not been near the original projections. Previous CAPTOR results reported
in the literature are based on settled effluent samples whereas no effluent
settling is envisioned for most of the applications where the process has
been claimed to be advantageous; original literature references do not
mention that samples were settled prior to analysis. Thusfar results at
the EPA facility have not confirmed either an enhancement in oxygen
transfer or process performance comparable to original projections for
Freehold. While the concept underlying the CAPTOR system is quite
attractive, much remains to be learned to make a realistic assessment of
it's potential usefulness in wastewater treatment.
This paper has been reviewed in accordance with the U.S. Environ-
mental Protection Agency's peer and administrative review policies and
approved for presentation and publication.
518
-------
CAPTOR DESCRIPTION
The CAPTOR process is a fixed film modification of the activated
sludge process. The trade mark CAPTOR is derived from Captivated Sludge
Process. The process is patented (1) and is marketed by Simon-Hartley
Ltd. of Manchester, England. The U.S. representative is Ashbrook-Simon-
Hartley in Houston, TX.
The CAPTOR process contains a number of novel features. Polyurethane
foam support particles (25 mm x 25 mm x 12 mm) are added to activated
sludge tankage to provide surface area for biomass colonization. The
foam pads have a porosity of 97 percent and a specific gravity slightly
greater than water. Forty pads/L is a design value considered reasonable
to allow for adequate pad circulation and provide sufficient surface area
for biomass growth. At 40 pads/L, 33 percent of the tank volume is
occupied by support pads. Because of the high pad porosity, the actual
reduction in available liquid volume is only 1 percent. The pads are
manufactured by a thermal process with a pore size specification of about
0.7 mm.
Screens are utilized to prevent the support pads from exiting with
the reactor effluent. Currently screens are constructed with 6 mm mesh
with 60 to 70 cu m/hr/sq m considered an acceptable liquid effluent
flux rate.
Excess biomass production within the pads is removed by cleaning
as illustrated in Figure 1. The pads are air-lifted onto a conveyor where
they are first subjected to an initial squeeze to remove water that does
not freely drain from the pads. This pre-squeeze may use one or more
rollers and is intended to remove excess water but not the attached
biomass. Finally the pads are tightly compressed by the squeeze rollers
and the cleaned pads are returned to the aeration system. A concentrated
waste sludge stream is also produced.
CAPTOR ORIGIN
The CAPTOR process is an extension of studies with biomass support
particles which were conducted by the University of Manchester. A number
of biomass support particles were evaluated by Atkinson et al., (2,3) for
use in industrial fermentation processes. In the initial studies, the
most satisfactory particles were prepared from a single strand of narrow
gauge steel wire and crushed into 6 mm spheres of about 80 percent
porosity. Studies were expanded to include polypropylene toroids and
reticulated polyester foams of 25 x 25 x 10 mm and 30 pores per inch (4).
Walker and Austin (5) described the use of both polypropylene toroids
and polypropylene pads in pilot studies using a narrow 0.6 cu m column fed
settled sewage and aerated with a single diffuser. Studies over a 4-month
period with the 5.3 cm toroids with 92 percent porosity showed that the
entrapped biomass became progressively more mineralised towards the pad
center. The compromises accepted for ease of manufacture yielded a
519
-------
Dewater
Influent
en
ro
o
O
O
REACTOR
O
o
S Judge
Effluent
Air
Figure 1. CAPTOR System Principles
-------
disproportionately large amount of anaerobic sludge inside the toroids
requiring very low food:mass ratios to give good BOD removal or nitrif-
ication. The foam pads were desirable because they could be removed for
cleaning by squeezing and it was indicated that a cleaning device was
designed to yield surplus sludge at 4 to 6 percent solids. A month of
operation at a 70-minute hydraulic retention time was indicated to show 85
percent total 300 removal and 92 percent soluble BOD removal. It was
also stated that laboratory trials with a scaled-down CAPTOR unit had
shown that 96 percent BOD removal could occur within 24 hours on an
industrial waste with a BOD of up to 6000 mg/1 and suspended solids of
around 4000 mg/1. A 12.5 cu m system was designed and was expected to go
on stream in early 1981.
By 1981, Simon-Hartley Ltd. was promoting CAPTOR as a process ideally
suited for pretreatment of strong industrial wastes prior to sewer dis-
charge (6). A typical installation would consist of aerated reactor,
pad cleaner and sludge holding tank. No secondary settlement or biomass
recycle was required. Waste sludge was claimed to contain up to 6 percent
solids and biomass concentrations were stated to be equivalent to 10 to
15 grams/L.
INITIAL U.S. EPA INVOLVEMENT WITH CAPTOR
In February 1983, a technology exchange meeting was held in Cincinn-
ati with representatives from the Water Research Center (WRC) in Great
Britain, Environment Canada and the U.S. EPA. CAPTOR was one of several
systems discussed. Simon-Hartley, the WRC, the Severn-Trent Water Author-
ity and the Swiss Federal Office for Environmental Protection were initi-
ating an evaluation of CAPTOR at Freehold, England. Financial support
from both the U.S. EPA and Environment Canada was solicited ')y the WRC.
The 18,700 cu m/day (5 mgd) plant at Freehold contains five parallel
aeration basins and two final clarifiers (7). The plant could not achieve
an acceptable degree of nitrification for more than three months per year
and upgrading to acheive year-round nitrification was required. It was
proposed that this could be achieved by modifying the first 25 percent
of each of the five aeration lanes to the CAPTOR process. Prior to a
complete plant conversion, however, it was proposed to modify just two of
the existing lanes to include CAPTOR and to retain two lanes as an
activated sludge control. Each system was to receive 50 percent of the
primary effluent flow. An economic evaluation indicated that the
approach with CAPTOR was more cost effective than the addition of nitrify-
ing trickling filters provided the rate of pad replacement was not in
excess of about 20 percent per annum. This economic analysis took no
credit for claimed improvements in aeration efficiency with CAPTOR.
Preliminary studies in a 0.6 cu m column (7) indicated that the
presence of the pads significantly increased the oxygen transfer efficiency
of both fine and coarse bubble diffusers with a 5 to 6 fold increase in
K|_a being reported for fine bubble aeration. Increased oxygen transfer
efficiencies (OTE's) would be an additional economic benefit for CAPTOR.
521
-------
A summary of process operation and projected performance for the
Freehold pilot study is presented in Table 1. The projected CAPTOR BOD
removal was indicated to be based on data collected from three seperate
studies where BOD removal was related to volumetric loading by
Percent BOD Removal = 98 e-°-12(Volunietn'c Loading, kg/cu m/day)
or to biomass loading by
Percent BOD Removal = 105 e-0-67^""0"1355 Loading, kg BOD/kg biomass/day)
However it was not made clear that these relationships were only applicable
to CAPTOR effluent samples which had been settled for one hour (8) and
therefore were not appropriate for calculating the organic load to a
CAPTOR system discharging directly into a subsequent biological system.
TABLE 1. DESIGN! VALUES FOR FREEHOLD, ENGLAND CAPTOR EVALUATION
Average CAPTOR Flow 9350 cu m/day (2.5 mgd)
Primary Effluent BOD 144 mg/1
F:M (CAPTOR only) 0.56 kg BOD/kg pad MLSS
Pad Concentration 40/L
Pad Solids 180 mg/pad
Equivalent MLSS 7200 mg/1
BOD Removal (CAPTOR only) 75 % (70 % winter, 80 % summer)
CAPTOR HRT 0.84 hours
Total HRT per Lane 3.4 hours
Waste Sludge Concentration 5 %
OTE 10 %
Based on the potential benefits which were presented for a CAPTOR
system and an interest in high biomass processes, the U.S. EPA awarded a
cooperative agreement to WRC in September, 1983 to cover a portion of the
costs and gain access to the data at Freehold as well as construct a
pilot-scale CAPTOR system at the EPA Test and Evaluation Facility (T&E)
in Cincinnati, Ohio.
CAPTOR PILOT STUDIES IN THE U.S.
The first pilot study of the CAPTOR process on municipal wastewater
in the U.S. was in Downingtown, Pennsylvania with the period of formal
data collection from 16 January thru early May, 1984. The study was
undertaken by the city at their expense because they were required to
upgrade their plant to achieve nitrification, and it was felt that an
522
-------
approach similar to that proposed for Freehold would be cost effective.
Data from this study were made available to the WERL-WRD.
The CAPTOR pilot reactor was 2.4 x 7.3 x 2.4 m in depth (8 x 24 x 8
ft) with the direction of flow parallel to the 7.3 m length. Reactor
volume was 43.5 cu m (11,480 gal) providing a hydraulic residence time
(HRT) of approximately 1 hour at an influent flow of 757 L/min (200 gpm).
Overall pad concentration was 42 pads/L although pad density was higher
at the effluent end of the reactor. Aeration and mixing were provided
with 54 ceramic diffusers.
Data from various periods of operation are presented in Table 2.
The influent wastewater to this plant was relatively weak with a mean
influent BODs concentration for the entire period of 101 mg/L. This
corresponds to an applied loading of 0.84 kg BODs/day/kg pad MLSS at
757 L/day (200 gpm) and pad solids of 72 mg/pad. Mean filtered BODs
removals shown in Table 2 varied from 20 to 28 mg/L and seemed to increase
somewhat as the waste strength increased. Considering only the data where
the influent filtered BODs's were in excess of the median value of 43 mg/L,
gives filterable BODs removal of 28 mg/L or 43 percent and total BODs
removal of 31 mg/L or 24 percent. It is clear, however, that BOD removals
were not strongly influenced by the range of influent and operating
parameters at this site.
Pad solids did not display any consistent variation and displayed
little sensitivity to the range of conditions encountered. The pad solids
concentration from 7 to 28 March, when the pad cleaner was operated most
frequently, were essentially the same as for 29 March to 23 April when
the pad cleaner was only operated 1/6 as often. Effluent SS were consist-
ently somewhat higher than the influent concentration reflecting excess
biomass growth in the system not removed via sludge wasting.
From 1 February to 16 March, the average pad cleaner run time was 4.74
hours/day and the waste sludge concentration was measured routinely (21
measurements). Data from this time period showed an average of 75 mg/L
of SS leaving in the process effluent. The pad cleaner produced waste
sludge flow with a mean solids concentration of 10700 mg/L. This is
equivalent to an additional 21 mg/L of SS that were removed via wasting
rather than in the process effluent. The sum of these values (96 mg/L)
is higher than expected indicating some variation in the waste sludge
flow (a constant 7.57 L/min (2 gpm) assumed) and/or some error in the
solids measurements. Nonetheless, this analysis shows that even during
the period of most extensive pad cleaner operation at this site approx-
imately 78 percent of the combined influent SS plus biomass growth were
escaping in the process effluent. The mean 44 mg/pad solids removal
and waste sludge estimates suggest that about 30 percent of the pads were
cleaned each day under these operating conditions. Boyle and Wallace (8)
reported 8 to 9 percent pads cleaned/hour corresponding to 38 to 43
percent of the pads cleaned/day.
523
-------
TABLE 2. CAPTOR PERFORMANCE AT DOWNINGTOWN, PA
16 Jan - 6 May3 1 Feb -22 Aprb 7 Mar - 28 MarC sBOD > 43mg/l^ 29 Mar -23 Apr*
No. Mean No. Mean No. Mean No. Mean No. Mean
Samples Value Samples Value Samples Value Samples Value Samples Value
Total BODs, mg/1
Influent
Effluent
Filtered BODs, mg/1
Influent
Effluent
Suspended Solids, mg/1
Influent
Effluent
Pad Solids, mg/pad
Before Cleaning
After Cleaning
Waste Sludge, mg/1
91
92
89
92
97
98
78
79
30
101
78
48
26
66
73
72
35
9980
72
72
70
72
78
78
58
59
21
93
72
43
22
62
68
71
34
107 00f
21
21
21
21
22
22
15
14
~~
84
70
37
17
64
77
66
33
— —
45
45
45
45
45
45
37
37
— —
127
96
65
37
74
81
77
37
— —
23
23
23
23
26
26
14
14
— —
86
61
40
17
56
60
63
37
— -
a Entire period of data collection
b Period of steady 757 L/min (200 gpm) influent flow
<• Mean pad cleaner run time of 6 hours/day; DO 2 to 5 mg/1
d Data when filtered influent BODc > median value of 43 mg/1
e Mean pad cleaner run time of 0.9 hours/day
f Last measurement was March 17, 1984
-------
Based on the results of this pilot study, a decision was made to
exclude the CAPTOR process from further consideration for upgrading at
the Downingtown facility.
The second pilot study of the CAPTOR process on a municipal waste-
water in the U.S. was at Marion, Illinois in the Spring of 1984. This
pilot unit also consisted of a first-stage CAPTOR unit fed primary efflu-
ent and discharging directly into a second-stage activated sludge unit.
Sodium aluminate and polymer were added to the raw wastewater resulting
in a primary effluent BOD of roughly 75 mg/L (8). Pad biomass solids
during the entire period of operation ranged from 45 to 70 mg/pad. The
data available for process characterization are quite limited and not of
a consistent high quality. Shortly after the study began, the State of
Illinois EPA decided that an innovative process such as CAPTOR was unac-
ceptable regardless of the pilot study results.
CAPTOR STUDY IN FREEHOLD ENGLAND
During the same period of disappointing CAPTOR pilot plant results
in the U.S., a paper presented by Cooper et al., (9) described selected
results from the study at Freehold. It was reported that a number of
unsteady-state step-change tests has shown that the kLa values obtained
in CAPTOR were higher by 2 to 3 fold than in the conventional activated
sludge system with fine bubble aeration. Oxygen utilization (OTE's) in
the CAPTOR system with 27 pads/L was estimated at 18 percent. Data from
various experimental runs are summarized in Table 3. BOD removals
varied between 45 to 60 percent and it was indicated that 70+ percent
BOD removal was still an achievable objective. Walker and Austin (10)
summarized plant performance at Freehold by indicating that a feed BOD of
140 mg/L yielded CAPTOR effluent BOD's of 50 to 60 mg/L in a 45-minute
retention time corresponding to mean BOD removals of 61 percent.
TABLE 3. CAPTOR TEST RESULTS AT FREEHOLD, ENGLAND REPORTED BY COOPER
ET AL. (9)
22 Nov - 1 June - 15 Sept -
31 Dec, 1982 7 July, 1983 4 Nov, 1983
Pad Concentration, pads/L 27 18 28
Pad Biomass Concentration, mg/pad 143 260 162
HRT in CAPTOR, min 43 46 47
Influent BOD, mg/1 114 142 136
CAPTOR Effluent BOD, mg/L 45 78 74
Influent SS, mg/L 101 122 144
CAPTOR Effluent SS, mg/L 81 90 97
525
-------
It was not until November, 1984 that the U.S. EPA in Cincinnati
began receiving copies of the CAPTOR system progress reports which had
been prepared for the CAPTOR Steering Committee. These reports contain
tabulated data and brief summaries of results from Freehold as well as
results from a number of pilot scale studies at Stevenage, England.
These reports provided further evidence that there were problems with
CAPTOR operation at Freehold as indicated by Boyle (11). Boyle and
Wallace had visited the Freehold facility in the summer of 1984 as part
of their assignment under U.S. EPA contract to evaluate porous biomass
support systems (8).
The Freehold CAPTOR system has experienced a variety of problems
including pad cleaner reliability and pad mixing. There was severe
pad maldistribution and typically one would find pad concentrations of
5 to 10 pads/L at the inlet end and 40 to 60 pads/L at the outlet end (12),
There were frequent problems with large "rafts" of pads floating on the
tank surface. Various modifications with different diffuser configurat-
ions were made in an attempt to overcome the mixing problem. At the
beginning of 1984 a decision was made to temporarily switch the major
effort from full-scale experimentation at Freehold to smaller scale
development at Stevenage (13).
Given all of the modifications and operating problems at Freehold,
it is difficult to identify periods of operation in the progress reports
which would represent stable, trouble-free operation. Table 4 summarizes
selected performance data in the progress reports for periods which
appeared to have no extreme performance problems. These data show BOD
removals of 20 percent or less and no suspended solids removal. These
results are in marked contrast to those previously reported for this same
facility by Cooper et al. (9) and Walker and Austin (10). In particular,
the results from June, 1984, which reflect an additional year of plant
modification compared to the June/July data in Table 3, show BOO removals
of only 8 and 15 percent.
TABLE 4. CAPTOR PERFORMANCE AT FREEHOLD, ENGLAND
7 Nov-15 Dec, 83
March, 84
June, 84
Unit Designation
Influent Flow, cu m/day
HRT, min
Influent BOD, mg/L
Effluent BOD, mg/L
Influent SS, mg/L
Effluent SS, mg/L
Pad Biomass, mg/pad
a 17 pads/L
b 28 pads/L
c Number of Samples in
C2a
261
38
156(16)c
123(16)
140(22)
149(19)
151(29)
Parenthesis
Clb
218
46
120(10)
115(12)
105(19)
112(21)
137(22)
Cl C2
223 223
45 45
155(9) 155(9)
133(9) 144(9)
105(19) 105(19)
105(18) 136(18)
114(17) 161(17)
526
-------
Major modifications have been made to the facility in Freehold
within the last few months and data generated in 1985 will represent a
hydraulic and staging configuration of the CAPTOR system substantially
different than previously utilized.
U.S. EPA CAPTOR PILOT PLANT STUDIES
Three pilot plant reactors were constructed for CAPTOR evaluation at
the EPA T&E facility in Cincinnati. The reactors have a liquid volume of
approximately 15.3 cu m (540 cu ft) and can be operated in series or
parallel. Feed is primary effluent from the Mill Creek Wastewater
Treatment Plant. Aeration is provided by EDI Reef fine bubble diffusers.
Because of small differences in tank geometry, 6 diffusers are located in
one tank whereas the remaining two contain 5 diffusers each. The diffus-
ers are located down one wall of the tank (beneath the effluent screens)
and produce a strong spiral flow in the completely mixed reactors. A
schematic of Tank A is shown in Figure 2.
OXYGEN TRANSFER EFFICIENCY
Because of the reported large enhancement in oxygen transfer effi-
ciencies with a CAPTOR system (7,9,10), the pilot studies were designed to
confirm and quantify this effect. Specifically, clean water OTE's with no
pads added, OTE's with a mixture of recycle activated sludge and primary
effluent (no pads), and OTE's in the CAPTOR system itself have all been
determined.
Clean water efficiencies were determined in accordance with the ASCE
Standard Procedure (14). Three DO probes with one ml thick membranes and
strip chart recorders were used for each determination. Tap water from
the City of Cincinnati was used and changed as needed to remain below the
allowable limit for total dissolved solids. Data were analyzed by the
non-linear regression method. The mean clean water OTE's in the three
reactors determined from the results of duplicate tests at each air flow
rate are shown in Table 5.
TABLE 5. CLEAN WATER TRANSFER EFFICIENCIES IN CAPTOR SYSTEM REACTORS
Date Tank No. Mean Air Water OTE Diffusers
Dec. Flow Temperature 20°C and 0 DO Per Tank
1984 scfm °C %
11 A 25.5 17.5 15.0 6
12 A 63.8 17.5 15.9 6
13 A 45.3 18.0 16.3 6
16 B 25.6 16.8 14.5 5
16 B 64.1 17.0 14.8 5
20 B 45.3 17.0 15.6 5
527
-------
INFLUENT
CO
DIFFUSER
6'
EFFLUENT
>
•(
c
. >
t
ft
| INFLUENT \
i DISTRIBUTOR \
' \
-1 * I-
1 t
/ EFFLUENT i
SCREEN \
\
V
h-2'H
i 1
4"
Figure 2. Schematic Diagram of Reactor No. A. at the T&E Facility
528
-------
Dirty water OTE's were measured by offgas analysis (15) using an
analyzer manufactured by Ewing Engineers and a 0.9 x 1.2 m (3 x 4 ft)
hood for gas capture. From 29 January to 1 February, 1985 the pilot
system was operated with series flow through the three reactors. Primary
effluent and return activated sludge flows were approximately 170 and
57 L/min (45 and 15 gpm), respectively providing an HRT of about 67
minutes per reactor. Parallel flow with 95 L/min (25 gpm) primary and
57 L/min (15 gpm) recycle was used for the offgas testing on 12 February
resulting in an increase in HRT to 100 minutes in Tanks 2 and 3. From
5 February to 10 February there was so much foam on the reactor surfaces
that it rose above the 0.6 m (2 ft) of freeboard and overflowed the top
of the tanks. Under these circumstances it was not possible to use the
offgas analyzer. The offgas results are presented in Table 6 where they
are grouped by stage of treatment. The alpha values in the lead stage
range from 0.20 to 0.51. Except for the data from 11 February, the second
and third stages exhibit the typical increases in alpha that one would
expect from fine bubble systems operating in a reactor-in-series con-
figuration where most of the soluble BOD removal was occurring in the
lead stage. Whatever entered the sewer system and caused the severe
foaming problem, resulted in significant alpha depression on 11 to 12
February compared to the values obtained prior to that time.
TABLE 6. DIRTY WATER OFFGAS ANALYSIS RESULTS GROUPED BY STAGE OF
TREATMENT
Reactor
No.
A
A
A
A
B
C
B
B
B
B
C
C
C
Date
Air Flow
scfm
DO
mg/L
29/1
30/1
4/2
11/2
12/2
12/2
FIRST STAGE OF TREATMENT
52.4
56.8
35.9
24.9
11.3
11.1
0.9
2.0
1.8
1.7
2.9
4.7
30/1
1/2
4/2
11/2
SECOND STAGE OF TREATMENT
14.9
14.5
14.7
11.6
2.0
3.7
3.4
2.5
31/1
4/2
11/2
THIRD STAGE OF TREATMENT
12.6 3.7
13.0
10.3
3.4
2.7
Alpha3
0.36
0.30
0.51
0.20
0.25
0.40
0.56
0.66
0.71
0.30
0.69
0.77
0.33
% OTE at
20°C and 0 DO
5.
4.
,64
.70
8.04
3.17
3.70
5.94
8.33
9.92
10.6
4.54
10.3
11.5
4.92
Based on mean clean water OTE for this tank
529
-------
CAPTOR PERFORMANCE AT THE T&E FACILITY
Following pad addition and system startup, the first period of
steady state CAPTOR data collection was from 19 May to 23 June, 1985.
During this period the three reactors were operated in parallel with
influent flows of 95 L/m1n {25 gpm) each. The pad cleaners were operated
to clean 1/5 the number of pads in each reactor per day. Inlet air
flows of 2.1, 1.4 and 0.71 cu m/min (75, 50 and 25 scfm) were provided to
Tanks A, B and C, respectively. The impact on daily mean DO levels is
shown in Figure 3. Offgas testing was performed on two occassions with
the results indicated in Table 7. Based on comparing the first stage
OTE's in Table 6 with the values in Table 7, there is no evidence thusfar
of enhanced OTE's with this system. Several more determinations are
required to confirm the lack of enhanced oxygen transfer.
9.0
a Reactor A
+ Reactor B
O Reactor C
0.0
. i i i i i i i i i i i i i i i i i i i i i i i i i i i i ' i
19 21 23 25 27 29 31 2 46 8 10 12 14 16 18 20 22
MAY
JUNE
DATE, 1985
Figure 3. Mean Dissolved Oxygen Levels in the CAPTOR Systems
530
-------
TABLE 7. OTE'S FOR THE CAPTOR SYSTEM
Reactor
No.
A
B
C
A
B
C
Date
1985
12/6
18/6
18/6
1/7
3/7
3/7
Air Flow
scfm
72.0
50.2
24.8
74.9
49.8
24.4
Reactor
DO, mg/1
3.1
1.5
1.0
5.0
4.0
1.0
% OTE
0 DO and 20°C
3.67
4.74
4.99
7.39
4.93
3.67
Selected data from the CAPTOR systems are presented In Table 8.
Although the mean DO in Reactor C was 0.8 mg/L, this system frequently
operated at lower DO levels (Figure 3) and it is clear that insufficient
DO was present in this system on some occassions. At the start of these
studies there was considerable speculation that the low pad solids levels
which had been observed in U.S. pilot studies were the result of relativ-
ely weak wastewaters and/or excessive turbulence and resulting shearing
of biomass from the pads. The results in Table 8 show no correlation of
air flow and pad solids levels. Furthermore, the waste strength at this
facility is roughly the same as that at Freehold, England (Table 1).
Where DO was not limiting, total BODs removals were about 54 percent and
filtered BODs removals were 84 to 90 percent. The 2.7 hour HRT was 3.2
times greater than the design value for Freehold where total BOD removals
of 75 percent were projected.
TABLE 8. MEAN VALUES FOR CAPTOR OPERATION AT THE T&E FACILITY FROM
19 MAY to 23 JUNE, 1985
Total BOD5 mg/1
Filtered 8005, mg/1
Suspended Solids, mg/1
Pad Solids, mg/pad
Dissolved Oxygen, mg/1
Waste Sludge Concentration, %
Influent
144(25)a
106(26)
123(35)
Reactor No.
A
68(26)
11(26)
154(35)
73(32)
4.1(216)
0.64(36)
B
65(26)
17(25)
153(35)
82(32)
2.3(216)
0.68(36)
C
91(26)
40(26)
141(35)
71(32)
0.8(216)
0.97(35)
Number of 24-hour composite samples or number of measurements
SUMMARY
The concept of porous biomass support systems such as CAPTOR offers
a number of potentially attractive advantages. Other approaches with
531
-------
porous pad media, such as that of Linde AG of West Germany, are also
under development (8). Extensive pilot studies at Stevenage and Freehold
and the U.S. T&E facility in Cincinnati should provide an adequate basis
to assess true process potential for CAPTOR. The potential of this system
for nitrification applications also needs to be investigated.
To date CAPTOR results have been nowhere near the optimistic early
projections for the process. The initial performance projections were
based on settled samples from the CAPTOR reactor when, in fact, no settling
occurs. Claims such as enhanced OTE's are also unverified. Fundamental
issues such as the best basin configuration and diffuser arrangement are
still unanswered. It is clear that much remains to be learned about the
true potential and limitations of this process.
532
-------
REFERENCES
1. Atkinson, B. et al., United States Patent 4419243, 6 Dec., 1983.
2. Atkinson, B. et al., Biological Particles of Given Size, Shape, and
Density for Use in Biological Reactors, Biotech, and Bioeng., 21,
193, 1979.
3. Atkinson, B. et al., Process Intensification Using Cell Support
Systems, Process Biochem., pg. 24, May, 1980.
4. Atkinson, B. et al., The Characteristics of Solid Supports and Biomass
Support Particles When Used in Fluidised Beds, Chapter 5, in
Biological Fluidised Bed Treatment of Water and Wastewater, edited by
Cooper, P. F. and Atkinson, B., Ellis Norwood Ltd., Chichester,
England, 1981.
5. Walker, I. and Austin, E. P., The Use of Plastic, Porous Biomass
Supports in a Pseudofluidised Bed for Effluent Treatment, Chapter 16,
in Biological Fluidised Bed Treatment of Water and Wastewater, edited
by Cooper, P. F. and Atkinson, B., Ellis Horwood Ltd., Chichester,
England, 1981.
6. CAPTOR Waste Water Treatment Plant with CSP, Simon-Hartley Ltd.,
1981.
7. Boon, A. G. and Cooper, P. F., Project No. 1800 CAPTOR, Water Research
Center, December, 1982.
8. Boyle, W. C. and Wallace, A. T., Status of Porous Biomass Support
Systems for Wastewater Treatment: An Innovative/Alternative Technology
Assessment, Final Report for EPA Contract No. 68-03-3130, In Press.
9. Cooper, P. F. et al., Evaluation of the CAPTOR Process for Uprating
an Over-Loaded Sewage Works, Paper presented at I. Chem. E. Symposium,
Bradford, West Yorkshire, 15 March, 1984 and I. Chem. E. Symposium,
Manchester, 28-29 March, 1984.
10. Walker, I. and Austin, E. P., Process Intensification by the CAPTOR
System, Presented to PIRA seminar SPB/6, Biological Treatment of Paper
and Board Mill Effluents, 17 April, 1984.
11. Personal Communication with W. Boyle, 14 August, 1984.
12. CAPTOR Progress Report No. 6-1 March to 8 May 1984 for the Steering
Group Meeting, Water Research Center, 15 May, 1984.
13. Letter from P.F. Cooper, Water Research Center, to R. C. Brenner of
U.S. EPA, 2 Nov., 1984.
533
-------
14. A Standard for the Measurement of Oxygen Transfer in Clean Water,
ASCE Standard, American Society of Civil Engineers, July, 1984.
15. Redmon, D. et al., Oxygen Transfer Efficiency Measurements in Mixed
Liquor Using Off-Gas Techniques, Jour. Water Poll. Control Fed.,
55, 1338, 1983.
534
-------
BIOLOGICAL AERATED FILTER PERFORMANCE
by
H.D. Stensel
University of Washington
Seattle, Washington
and
R.C. Brenner
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
535
-------
BIOLOGICAL AERATED FILTER PERFORMANCE
by: H.D. Stensel
University of Washington
Seattle, WA
R.C. Brenner
U.S. EPA
Cincinnati, OH
ABSTRACT
The Biological Aerated Filter (BAF) system is a high rate downflow fixed
film treatment process applied after primary treatment of domestic waste-
waters. No secondary clarification is required after BAF treatment and the
system requires about one-fifth the land area of the conventional activated
sludge process. The system has been studied in an EPA Demonstration Project
in Salt Lake City, Utah. The study showed that a critical operating and
performance parameter for the system is the size of the media used. An
effective media size of 3.4 to 4.4 mm is recommended for wastewaters with a
high proportion of BOD in the particulate form and with a significant influent
suspended solids concentration to minimize backwash requirements. Organic
loading rates to achieve secondary treatment of domestic wastewaters may range
from 3 to 4 kg BOD/m -day (180 to 250 Ib BOD/1000 cu ft-day) depending on the
wastewater temperature. The organic loading rate could be higher if the
wastewater has a more soluble BOD form to allow the use of a smaller media
size. Oxygen consumption rates and sludge production rates are a function of
the wastewater characteristics. The oxygen consumption rate could vary from
0.6 to 1.0 kg oxygen per kg of BOD removed. The sludge production rate could
vary from 0.75 to 1.1 kg per kg of BOD removed.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
536
-------
INTRODUCTION
There has been considerable interest in the past few years in developing
biological wastewater treatment processes that reduce reactor volume and land
area requirements. These biological process designs, generally employing a
fixed film treatment mechanism, include plastic media trickling filters,
rotating biological contactors, the Captor process (1), fluidized bed reactors
(2), the PER system (3), the Acticontact process (4), and the Biological
Aerated Filter (BAF) system (5,6). The last four of the above processes are
characterized by a relatively high fixed film surface area per unit volume, a
relatively high reactor biomass concentration, and mechanical oxygen
dissolution methods to support the higher volumetric biological oxidation
rates.
DESCRIPTION OF BAF SYSTEM
In the BAF system wastewater is passed downflow through a 1.5-1.8m bed of
vitrified clay media. The effective size of the media used has ranged from
3.0 to 4.0mm. For domestic applications, primary treatment is normally
required ahead of the BAF system. Pretreatment is necessary to reduce
suspended solids and grease to acceptable levels. No additional clarification
is used after the BAF treatment system. Oxygen is supplied by air sparging
through an air header system located about 30cm. above'the bottom of the
media. The media is supported by an underdrain plate, and plastic nozzles are
used to collect the treated water and distribute backwash water and air. BOD
removal is accomplished by solids filtration and biological metabolism of
soluble organics. Periodic backwashing is necessary to flush out excess
biological solids and filtered solids to minimize the operating headless. The
system operation is completely automated by a microprocessor that initiates
and controls a backwash sequence of air, air/water and water.
Backwashing is initiated by the microprocessor on a timed basis or by a
level probe sensing a certain headless build-up above the media. The first
step of the backwash sequence involves the addition of air to the bottom of
the bed at a rate of 0.9 m/min (3 scfm/ft ) for about one minute. Then air
and water are applied simultaneously with a water application rate of 0.73
m/min (189 gpm/ft ) for about 1 minute. At this point a level probe detects
the higher water level and both the air and water applications are stopped for
one minute to allow the media to settle on the bed. Water is then applied
again at the previous rate for about 20 seconds. The water level in the BAF
cell is then high enough to cause the backwash water to flow through siphon
pipes located in the cell. About two minutes elapse before the backwash water
is siphoned out of the cell. This procedure is repeated four to five times
for a complete backwash event. During some of the earlier operating
537
-------
experiences, the described backwash cycle was completed two times when the
backwash was initiated by a headloss build-up and five times for a timed
backwash event.
The claimed design loadings by the BAF developers to achieve secondary
treatment have been in the range of 3 to 5 kgBOD/m -day (190-310 lbs/1000
cu.ft.-day). In January 1983 a BAF system with a nominal flow of 1893m /day
(0.5 MGD) was started up at the South Davis Sewer Improvement District south
plant near Salt Lake City under an EPA demonstration plant project. The
objective of this project was to evaluate the BAF system relative to BOD and
suspended solids removal, oxygen requirements, energy requirements, sludge
production, backwash requirements, media characteristics, operational
requirements and mechanical reliability. This paper will summarize the
performance results for a one year operating period relative to the above
objectives for this study.
DEMONSTRATION PLANT DESCRIPTION
BAF FACILITY
Figure 1 shows a schematic of the BAF facility. The BAF system processed
a portion of the primary effluent stream at the South Davis plant that treats
a total average flow of approximately 11,370 m /d (3 MGD) through a two stage
trickling filter system. The wastewater consists of 70 to 80 percent domestic
wastewater with industrial contributions originating from an oil refinery,
a small tannery, and a laundry. The BAF system has two self priming
centrifugal pumps rated at 2.2 and 3.7 kw (3 and 5 Hp) each to independently
feed primary effluent to each of the two cells, designated East and West.
Each cell is 3.7 m (12 ft) x 3.2 m (10.5 ft) and has a media depth of about
1.7m (5.7 ft). The pumping rate of each feed pump was controlled by variable
frequency drive (VFD) controllers. A custom built signal generator directed a
signal to the VFD controllers to vary the flow each hour to simulate diurnal
flow patterns. A typical flow pattern used for each cell is shown in Figure 2
for a 1.3 peak to average (P/A) flow variation. The signal generator also
controlled influent and effluent sample pumps to provide hourly flow
proportioned samples that were stored in a 4°C refrigerator at the site.
Propeller type flow meters were installed in the effluent lines from each cell
to record the instantaneous flow rate on strip chart recorders. The
accumulative flow was indicated by a digital readout on the meter.
Automatic backwashing was initiated by either a signal to the micro-
processor from a level probe as the headloss built up in the BAF cell or by
programming for a timed backwash cycle. The microprocessor would then shut
off the appropriate feed pump and one of the two 7.5 kw (10 Hp) process air
blowers, activate the 15 kw (25 Hp) backwash air blower and the 15 kw (25 Hp)
backwash water pump, and then open and close the necessary valves at
programmed time intervals to carry out the backwash procedure. When the
backwash water clear well volume was depleted, a level probe provided a signal
to open an automatic drain valve to direct process effluent to the clear well.
The backwash water was collected and drained to the plant headworks. An
automatic sampling system was also operated by the microprocessor, to collect
a backwash solids sample during each backwash flush.
538
-------
BW
H2°
EFFL.
BACKWASH COLLECTION
CLEAR
WELL
PRIMARY
EFFLUENT
WEST
BAF TANK
EAST
H-
K-
CONTROL
SHED
FEED
J X
CLARIFIER
- BW BLOWER
PROCESS AIR
BW
AIR
PROCESS
AIR
Figure 1. Biological aerated filter demonstration plant.
539
-------
1.9 -i
1.8 -
1.7 -
.6 -
u.
O
01 ^ 1 2 -
O
i 1-1 -
1 -
0.9 -
0.8 -
0.7 -E
_ ,
H
\A/r~CT j
1 1 i
C
1-
!
[
^^^^w
i
jr
H
i
i
1 — c
i
U
1
P n
'
3 D D 0 1
EAST
1 2AM 4AM SAM
12PM
Dr-i r
, i | i i i | i i I
4PM 8PM
TIME
Figure 2. Diurnal flow variation pattern.
-------
OPERATING CONDITIONS AND DATA COLLECTION
As shown in Table 1, the West cell was operated with a much higher
hydraulic application rate than the East cell._ The original operating
objective was to operate at about a 5 .kg BOD/m -day (310 lb/1000 cu ft-day)
organic loading in the West cell to meet 30/30 mg/1 BOD and total suspended
solids (TSS) effluent concentrations. A 3 kg BOD/m -day (190 lb/1000 cu
ft-day) operating load was set for the East cell to meet 10-10 mg/1 BOD and
TSS effluent concentrations. The actual loadings achieved were generally
about 20 percent lower because infiltration to the plant was higher than
normal during this study period due to record breaking precipitation in the
Salt Lake City area. The average hydraulic application rate was limited under
these conditions by allowable peak flow hydraulic application rate that would
not result in a rapid headless build-up.
TABLE 1. BAF OPERATING CONDITIONS
Hydraulic Application Rate (m/hr) Peak to Average Flow
Date West Cell East Cell West Cell East Cell
June-October 1983 3.9 2.4 1.3 1.0
October-November 1983 3.9 2.4 1.0 1.0
December-January 1984 2.4 2.4 1.0 1.0
February 1984 3.4 2.4 1.4 1.3
March-June 1984 2.4 1.2 1.0/1.5 1.0/1.5
NEW MEDIA
Empty Bed Volume Detention Time: 26 min. at 3.9 m/hr.
43 min. at 2.4 m/hr.
During October and November 1983, a constant flow operation was used,
because of problems encountered with excessive headless build up and a high
backwash frequency requirement. This appeared to be caused by two successive
problems. During the last week of September and early October, one of the
plants digesters was cleaned, resulting in primary effluent TSS concentrations
in the range of 200-300 mg/1. This caused a rapid headloss build-up and
shortened the BAF operating time between backwashes. The second problem
causing a high backwash frequency appeared to be due to the presence of a
fiber or hair-like material in the influent wastewater. Thin fibers, which
were about 2-9 cm (0.8 to 3.5 in) in length, were collected on the top of the
BAF cell media and caused a formation of a mat of fiber and solids. The fiber
problem diminished some time during December. The hydraulic application rate
for both cells was similar in December since different backwash techniques
were used to attempt to minimize the backwash frequency.
541
-------
During January, the hydraulic loading to the West cell was not increased
due to a reduced BOD removal efficiency. This was affected by both a much
lower wastewater temperature and higher wastewater strength. As the
wastewater temperature increased in February, the hydraulic application rate
to the West cell was increased. The diurnal flow variation was also resumed
in both cells.
In early March 1984, the media in both cells was changed to one with a
larger effective size to determine if this would result in a lower backwash
frequency. A lower treatment efficiency was expected with the larger media
since the biofilm surface area per unit volume was decreased. The previous
media had an effective size of 2.8 mm and a 1.6 uniformity coefficient. The
replaced West cell media had an effective size of 4.4 mm with a 1.6 uniformity
coefficient, and the replaced East cell media effective size was 3.4 mm with a
1.5 uniformity coefficient. Based on the effective size, the new media in the
West cell had about 63 percent of the surface area per unit volume as the
original media.
The 24 hour flow proportioned composite samples were analyzed for total
and soluble BODr and COD and total and volatile suspended solids to document
overall treatment performance. Soluble BOD and COD analyses were performed on
the filtrate of samples filtered through glass fiber filter paper. All BOD
analyses were inhibited to prevent nitrification. Analytical methods were
done in accordance with the 15th edition of Standard Methods (7).
The net solids production (kg TSS/kg BOD removed) was determined by
measuring the mass of solids exiting in the backwash water and in the system
effluent. The quantity of backwash water siphoned from a BAF cell during each
backwash cycle was relatively constant. Backwash water was collected in a
tank whose volume had been calibrated as a function of liquid depth. A
microprocessor provided a counter and LED readout of the number of backwash
cycles. A backwash sample of approximately 1 liter was pumped from a cell at
the beginning of each backwash cycle siphon. A signal to the sample pump was
controlled by the microprocessor. At the time of sampling, the backwash water
was well mixed due to the water flush through the cell.
Two methods were used to determine the amount of oxygen consumed. The
first and more direct method involved measuring the air sparge rate and
sampling and analyzing the oxygen content of the off gas escaping through the
liquid at the top of the BAF cell. The air supply rate was measured by the
use of an annubar and manometer located on each of the air supply lines to
each BAF cell. Temperature and pressure measurements were used to calculate
the air rate at standard conditions. A 0.61 m (2 ft) x 0.91 m (3 ft) fiber-
glass tank was immersed in the liquid with the open end down to collect the
gas. A tube at the center of the gas collection device directed the gas to a
gas sampling tube. The oxygen and carbon dioxide content of the gas samples
were measured in the laboratory using an Orsat apparatus and procedure (8).
Gas samples were taken at nine different locations uniformily located across
the top of each BAF cell. During the off gas sampling period, the feed flow
rate was maintained constant and influent and effluent grab samples were
obtained for COD analysis. The second method was to conduct a COD balance on
542
-------
the system based on the COD of the influent, effluent, and backwash water.
While less accurate, this method allowed a daily oxygen utilization rate
calculation for a complete 24 hour period and the COD data were available
daily. In the off gas method, oxygen utilization rates were only observed
over a 1-2 hour test period.
BAF TREATMENT PERFORMANCE
BOD AND TSS REMOVAL
The primary effluent wastewater characteristics are divided into four
periods in Table 2 according to the influent temperature conditions. From
June through October, the influent temperature was relatively warm ranging
from 18 to 23°C. During November, the wastewater temperature steadily
declined from 17 to 12°C. From December to early March, the wastewater
temperature remained relatively cold and then gradually increased from 12 to
17°C from April through May. The primary effluent was 20 to 30 percent weaker
than normal during this study phase as a result of unusually high
precipitation. As shown, the soluble BOD concentration was only 20-25 percent
of the total BOD concentration. This shows that a large portion of the BOD
removed in the BAF system was due to solids removal by filtration with the
media bed.
TABLE 2. PRIMARY EFFLUENT AVERAGE WASTEWATER CHARACTERISTICS
TBOD SBOD TSS
DATE Temp °C (mg/1) (mg/1) (mg/1)
June-October 1983 21 91 20 123
(18-23)
November 1983 15 89 20 108
(12-17)
December-March 1984 12 100 24 113
(10-13)
April-May 1984 15 92 21 117
(12-17)
West cell average organic loading rates, temperatures and effluent BOD
and TSS concentrations are summarized in Table 3 for the initial operating
period prior to changing the media in March 1984. During the warmer operating
period from July to.,September, the average weekly organic loading ranged from
3.5 to 6.0 kg BOD/m -day (220 to 370 lb/1000 cu ft-day) and the effluent BOD
and TSS concentrations averaged 13.0 and 14.3 mg/1, respectively, for that
period. During the latter part of September, when the digester cleaning
operation resulted in very high influent solids concentrations and higher
organic loadings, the BAF effluent TSS concentration increased but the
effluent BOD concentration was not significantly affected. During the high
backwash frequency in October due to an apparent influent fiber problem, the
543
-------
effluent TSS and BOD concentrations decreased, even though the organic loading
was higher.
TABLE 3. WEST CELL AVERAGE PERFORMANCE
AVERAGE ORGANIC
PERIOD kg
July - September 1983
September-Digester
Cleaning (2 weeks)
October
November-December
January-February 1984
LOADING
BOD /m -day
4.7
5.0
4.9
4.3
4.4
AVERAGE EFFLUENT CONCENTRATIONS (mg/1)
TEMP °C BOD TSS
21.5
21.3
19.6
12.0
11.3
13.0
11.6
10.0
12.6
22.7
14.3
16.5
8.8
10.8
20.4
During the colder operating months, both organic loading and wastewater
temperature were changing on a daily and weekly basis. The data shows that
the system effluent BOD and TSS concentrations increased significantly during
the colder temperature operation in January and February.
As Table 4 shows, the lower loaded East cell produced effluent BOD and
TSS concentrations close to 10 mg/1, with organic loadings in the 3.0-3.5 kg
BOD/m-day (180-220 lb/1000 cu ft-day) range. As in the West cell operation,
a higher effluent TSS concentration occurred during the digester cleaning
operation. During the high backwash frequency operation associated with the
appearance of fibers, the East cell effluent BOD and TSS concentrations were
lower even though the organic loading was not decreased. As observed for the
West cell effluent the effluent BOD and TSS concentrations .increased during
the colder temperature operation in January and February.
TABLE 4. EAST CELL AVERAGE PERFORMANCE
AVERAGE ORGANIC
LOADING AVERAGE EFFLUENT CONCENTRATIONS (mg/1)
PERIOD kg BOD/m-day TEMP °C BOD TSS
July - September 1983
September Digester
Cleaning
October (High Backwash)
November— December
January-February 1984
3.2
3.5
3.2
3.5
4.1
20.6
21.1
19.6
13.2
11.4
11.8
9.6
7.2
11.3
17.8
9.6
12.9
6.4
9.1
16.1
544
-------
Since wastewater characteristics, strength, and temperature could all
vary simultaneously, it was not possible to separate performance variables to
precisely determine relationships describing treatment performance. However,
organic loading and temperature are parameters that affect performance. The
effect of these parameters on effluent BOD concentration is shown in Figure 3.
The data points are separated into two categories: wastewater temperatures
greater than 18°C and wastewater temperatures less than 15°C. The October and
November performance data are not included in this analysis because of the
changing wastewater temperatures and the presence of fibers in the influent.
The April to May data are also not included in Figure 3 because a different
media size was used during that operating period. There was less scatter in
the data at temperatures greater than 18°C, and a linear regression yielded
the straight line shown with a correlation coefficient of 0.89. A similar
analysis for the lower temperature data yielded a correlation coefficient of
0.80. The data show that the average weekly effluent BOD concentration is
higher and more sensitive to organic loading at the lower temperature range.
At lower organic loadings the effluent BOD concentration is not as
significantly affected by the wastewater temperature ranging from 10 to 23°C.
At an organic loading of 5.0 kg BOD/m-day (310 lb/1000 cu ft-day) the
effluent BOD concentration increased by about 40 percent when the temperature
was decreased to the lower operating range. Effluent TSS concentrations
followed the same trend as effluent BOD concentrations throughout the study.
As indicated in Figure 4, lower effluent BOD concentrations were achieved
in October and November over a wide range of organic loadings than during
other operating periods. Performance was also relatively insensitive to
wastewater temperature which dropped from 18 to 12°C during this period.
Previous performance data obtained for temperatures greater than 18°C are
included in the shaded area on the figure. The lower effluent BOD and TSS
concentrations during this period are due to improved filtration performance.
A more rapid headloss buildup and higher backwash frequency was also observed.
These results appear to be related to fibers in the influent wastewater to the
BAF cells. Visual examination at the top of the media bed showed that the top
few inches of media contained a mat like structure consisting of an
agglomeration of media, small fibers, filtered solids and attached biological
solids. This material provided excellent filtration at the top of the bed at
the expense of a very rapid headloss build-up.
These performance results indicate the importance of wastewater
temperature and wastewater characteristics on treatment system efficiency.
BOD removal efficiency can not be assumed to be only a function of organic
loading, which has been used previously as the basic performance parameter.
For the BAF system, filtration is an important particulate BOD removal
mechanism along with mass transfer and biological oxidation for removal of
soluble degradable organics in the wastewater. Thus, the effluent BOD that
may be achieved at a given organic loading will depend on the proportion of
particulate BOD in the influent and the filterability of these solids. As
illustrated by the fibers occurrence, the backwash requirements are also
affected by influent solids characteristics.
Soluble COD removal profiles obtained from grab samples at taps located
along the media depth, indicated that most of the soluble organic removal
545
-------
\
O>
E
o
z
o
u
o
o
m
34
32 -
30 -
28
26 -
24 -
22 -
20 -
18 -
16 -
14 -
12 -
10 -
8 -
6 -
4 -
2 -
0
T<15D
D
D
—i 1 1 , , n
345
ORG. LOADING RATE (KG. BOD/CU.M.-DAY)
T>18
Figure 3. Effluent total BOD versus organic loading rate-weekly averages.
-------
tn
\
O)
E
u
z
O
u
Q
O
m
24
22 -
20 -
18 -
16
14 -
12
10 -
8 -
6 -
4 -
2 -
PREVIOUS
D
D D
n
024
ORG. LOADING RATE (KG. BOD/CU.M.-DAY)
Figure 4. Influent fibers affect effluent BOD.
-------
occurred in the first 0.9 to 1.2 m (3 to 4 ft) of media. This suggests that
the same effluent quality may have been achieved if the organic loading was
increased by an increase in the influent soluble BOD concentration. Unless
very conservative organic loading design value are used for the BAF System,
pilot plant studies would be required to determine the effect of different
wastewater characteristics on performance.
BACKWASH REQUIREMENTS
As Figure 5 shows, before and after the digester cleaning and influent
fiber occurence, the percent of product water used to backwash the West cell
generally ranged from 10 to 15 percent. The East cell requirements were
generally in the 15 to 20 percent range for the same period. Both cells were
backwashed with five cycles on a 24 hour timed basis each day. In addition,
one or two additional two cycle backwashes occurred in October and November
with product water requirements reaching as high as 20 to 30 percent for the
West cell. The East cell requirements were in the 30 to 45 percent range for
one week of operation during this period. In such cases, backwashing was
occurring every 2 to 4 hours during the day. The source of the fibers was not
determined. A laundry and small leather tannery were suspected as possible
sources. After November, the problem did not occur again and the backwashing
requirements returned to normal in December.
In early March 1984, the media in both cells was changed to a larger size
and the backwash product water requirements dropped significantly. The
backwash frequency was reduced to once per day, which is desirable to minimize
hydraulic load variations to a BAF system due to recycle streams. The higher
percent backwash water for the East cell in late May 1984 was a result of
decreasing the influent hydraulic application rate to promote nitrification in
the cell.
These results show that there is a tradeoff between media size and
backwash requirements. A larger media size may be desirable in the BAF system
to minimize the amount of product water recycled through the primary clarifier
and BAF cells. Increased backwash frequency can impact system economics
several ways. The size of primary clarifiers and the number of BAF cells are
affected by increased system hydraulic loads. In addition, increased
backwashing requires that each BAF cell be out of service for a longer time
each day, thus requiring a larger system to handle wastewater flows during
backwash periods. Energy requirments also increase with increased backwash
requirements since backwash air and water must be supplied over a longer
duration each day. However, when the media size is increased to minimize
backwash requirements, the system's treatment efficiency is reduced due to a
lower filtration efficiency and less biofilm surface available per unit of
reactor volume. Thus, an optimal media size likely exist for different
wastewaters as a function of the influent solids concentration and solids
characteristics, and treatment level needed.
MEDIA COMPARISON
As Figure 6 shows, to meet desired treatment levels, the organic and
hydraulic loading rates were decreased to both BAF cells after changing the
548
-------
6175
BACKWASH (% OF PRODUCT WATER)
•n
H-
OQ
fi
(B
(t
i-t
n
n
n
rt
fD
i-l
en
m
cr
o>
D
CO
fD
fD
P
tu
fD
i-l
01
OQ
fD
-------
00
m
c^
o
o
05
CO
o
O
H
(W
O
03
0?
CO
"8
O
ID
i-h
O
m
it
09S
BOD CONCENTRATION (mg/l)
O N)
I I
I I I
I I I I I
O
o
•
r~
O
g
o
H
m
o
o
o
\
o
c
o
>
I
-------
media. The West cell was operated at about the same organic loading rate used
previously for the East cell and the East cell organic loading rate was
reduced to 1.5 to 2.0 kg BOD/m -day (90-125 Ib BOD/1000 cu ft-day) to achieve
a treatment goal of both BOD removal and nitrification. Table 5 compares the
BAF performance for the two different media sizes for operating periods with
similar wastewater temperatures and organic loadings. Both systems were
operated under constant flow conditions during these periods. Even though the
new media was operated under a slightly lower organic loading rate than used
for the earlier media the effluent TSS and BOD concentrations were signifi-
cantly greater. However, the backwash water requirement was significantly
reduced with the new media.
An overall comparison between media BOD removal performance is further
shown in Figure 7. A linear regression of effluent BOD versus organic loading
for the new media yielded a straight line with a correlation coefficient of
0.85. This is compared to the straight line obtained for data collected on
the initial media. The sensitivity of effluent BOD concentration to organic
loading was similar in both cases. The figure clearly shows the reduced
treatment capacity for the larger sized media. An effluent BOD concentration
of 20 mg/1 may be achieved at an organic loading of 3.0 Kg BOD/m -day (190 Ib
BOD/1000 cu ft-day) for the new media versus about 4.5 kg BOD/m day (280 Ib
BOD/1000 cu ft-day) for the old media. The lower treatment efficiency for the
new media is attributed to poorer filtration efficiency and lower biofilm
surface area per unit volume.
TABLE 5. BAF MEDIA EVALUATION
Date
Wastewater Temperature (°C)
Organic Loading
(kg BOD/m -day)
Average
Performance
TBOD (mg/1)
TSS (mg/1)
Percent Backwash
INITIAL MEDIA
11/28 to 1/19
12.4
(12.0 - 12.6)
3.8
(3.4 - 4.4)
Influent Effluent
88
107
13.4
10.8
NEW MEDIA
3/26 to 4/30
12.7
(12.0 - 13.5)
3.4
(3.1 - 3.8)
14.8
Influent
94
123
6.8
Effluent
21.7
22.5
OXYGEN UTILIZATION AND SLUDGE PRODUCTION
Oxygen consumption per unit of BOD removed was determined from off-gas
tests and from the COD balance calculations. The average values and
corresponding ranges of values determined with each method are summarized in
Table 6. The oxygen consumption per unit of BOD removal was similar for both
cells even though the organic loadings were different. This may be related to
551
-------
o
o
m
3O
o>
E
o
z
o
o
a
o
m
11 16 21 26 31 36 41
WEEK * - MAY 3D. 1983 - MAY 3O. 1984
46
Figure 7. Effluent BOD versus organic loading rate for new media.
552
-------
the fact that the solids filtered out in both systems may have remained in
each cell about the same amount of time, since both cells were backwashed at
least every 24 hours. The oxygen consumption ratios are low compared to
conventional or highly loaded activated sludge treatment systems. Since much
of the influent BOD was in the form of particulate matter, this suggests that
the solids filtered were not highly degraded. Filtered solids may be retained
in the BAF for a matter of hours before backwashing compared to days for
activated sludge systems. Thus, less solids degradation may be expected. The
portion of soluble BOD in the influent may also affect the oxygen consumption.
As the soluble fraction of the influent BOD increases, the oxygen consumption
ratio can be expected to increase since a greater portion of the influent BOD
should be oxidized.
TABLE 6. OXYGEN CONSUMPTION IN BAF CELLS
Oxygen Consumption (kg 02/kg BOD Removed)
East Cell West Cell
Off Gas Tests 0.55 0.51
(0.43 - 0.69) (0.42 - 0.80)
COD Balance Method 0.66 0.63
(0.40 - 0.96) (0.44 - 0.82)
The solids production per unit of BOD removed was similar for both cells
as shown in Figure 8. These observations agree with the previously noted
similar oxygen consumption values for each cell. The solids production values
were higher during the digester cleaning operation that resulted in
substantially increased influent solids concentration. The lowest value
occurred for the week of January 16. During that week, the influent soluble
BOD fraction was higher than previous weeks. This further supports the
hypothesis that soluble BOD is more readily degraded in the BAF cell than
particulate BOD.
Based on the observations in Figure 9, the solids production yield factor
was plotted against the ratio of TSS to BOD in the BAF influent. As Figure 9
shows, a relationship was apparent with a correlation coefficient of 0.89.
Figures 8 and 9 indicate that the solids production and oxygen consumption per
unit of BOD removed are a function of the influent wastewater characteristics.
SUMMARY AND CONCLUSIONS
This paper summarizes a one year evaluation of a full scale BAF system
treating a municipal wastewater plant primary effluent. The operation was
completely automated with a microprocessor controller, and operator
requirements were limited to equipment maintenance and sampling. Treatment
553
-------
en
en
-p.
Q
LU
LJ
OH
Q
O
m
o
y.
t/i
t/i
o
0.6
9-5
1-23
2-20
WEEK BEGINNING DATE (9-5-83 TO 2-20-84)
Figure 8. Solids production rate for East and West BAF cells,
-------
959
KG TSS/KG BOD REMOVED
p
b>
OQ
i-l
(C
o
d.
rt
H-
O
1-1
PJ
(D
D-
p
Ml
M
c
H
to
CO
o
o
i-t
(U
O
(D
o
^J
[
o
bo
i
o
ID
i
-» -»
b —
i i
-»
fo
i
-*-»-»
w ^ en
i i i
-»
en *•
i
c
rn
to
\
CD
O
O
!
n
-------
performance parameters evaluated were BOD removal, TSS removal, oxygen
consumption, sludge production and backwash frequency.
The evaluation showed that the BAF, BOD and TSS removal efficiencies as a
function of organic loading are highly dependent on the media size used in the
BAF cell. Besides biological oxidation of soluble BOD diffusing to the
biofilm, filtration is also an important removal mechanism for the BAF system.
Filtration accounted for a major portion of the BOD removal in this study
because the influent wastewater had a relatively high particulate BOD
fraction. Use of a smaller media size improves the filtration efficiency and
provides more surface area per unit volume for biofilm growth to improve the
soluble BOD removal efficiency as well. However, smaller media sizes can
create a more rapid headloss build up during operation and increase backwash
frequency to undesirable levels. A once-per-day backwash frequency is
desirable to provide a more stable BAF operation and to minimize the effects
of high recycle flows. In this study, the media effective size was increased
from 2.8 to 4.4 mm for the West cell and from 2.8 to 3.4 mm for the East cell.
The wastewater characteristics may also influence the media size
selection for a given BAF application. A more soluble wastewater with less
influent solids would likely favor a smaller media size. The level of
influent solids and the characteristics of those solids will affect filtration
and BOD removal performance as well as backwash requirements for a given
media. Headloss problems occured when the influent contained an unusual level
of small fibers in this study.
Expected BOD and suspended solids removals for a BAF system may not be
predicted from the design organic loading rate only. Removals will also be
affected by the influent particulate and soluble BOD fractions, the nature of
the influent solids, the media size, and wastewater temperature. At lower
loadings, system performance is less sensitive to temperature changes. BAF
system effluent BOD and TSS concentrations increased significantly as the
loadings increased for temperatures in the range of 12 to 15°C. At higher
temperatures (above 18°C), the performance decreased less rapidly with
increasing organic loadings. With the larger media used in this study,
effluent BOD and TSS concentrations of less than 25 mg/1 could be achieved at
organic loadings of 3.5 to 4.0 kg BOD/m -day (220 to 250 Ib BOD/1000 cu
ft-day) with wastewater temperatures ranging from 12 to 17°C. This loading
range is about five times that used for conventional activated sludge
treatment. In addition, final clarifiers and sludge recycling equipment are
eliminated. The land area savings advantages of the BAF system are readily
apparent.
Oxygen requirements and sludge production were also affected by the
influent wastewater characteristics. Sludge production rates averaged about
1.1 kg TSS/kg BOD removed but increased relative to the influent TSS/BOD
ratio. When the soluble portion of the influent BOD increased, the sludge
production rate decreased. The lowest value observed in this study was 0.75
kg TSS/kg BOD removed. The oxygen consumption rate per unit of BOD removed
varied over a wide range (0.40 to 0.96 kg Oxygen/kg BOD removed) and averaged
from 0.62 to 0.66 for the two BAF cells. The particulate and soluble BOD
fractions of the influent wastewater likely affected these values.
556
-------
The results of this study provide useful information on factors that
affect the BAF system performance and also provide a range of design
parameters for municipal wastewater treatment applications. Conservative
values for design should be used unless pilot plant studies are performed to
evaluate the treatment performance and backwash requirements for a specific
wastewater and media selection.
557
-------
REFERENCES
1. Walker, J. and Austin, E.P., "The Use of Plastic, Porous Biomass Supports
in Pseodu-Fluidized Bed for Effluent Treatment" Biological Fuidised Bed
Treatment of Water and Wastewater ed. by P.P. Cooper and B. Atkinson, p.
273 Ellis Horwood Limited, 1981.
2. Jeris, J.S. et al., "Biological Fluidized Bed Treatment for BOD and
Nitrogen Removal," Journal Water Pollution Control Federation, 48, 816,
(1977).
3. Young, J.C. and Stewart, M.C., "PBR - A New Addition to the AWT Family,"
Water and Wastes Engineering, 20, August (1979).
4. Isohata, Y., "Development and Field Experience of Acticontact Packed Bed
Reactor for Wastewater Treatment" Presented at the Second International
Conference on Fixed Film Biological Processes" Arlington, Virginia, July
10-12, 1984.
5. Stensel, H.D. and Lee, K., "Biological Aerated Filter for Full Scale EPA
Demonstration Project" Proceedings of the 1983 Annual Meeting, Water
Pollution Control Association, Park City, Utah. pg. 56, April 1983.
6. Stensel, H.D., Brenner, R.C. and Lubin, G.R., "Aeration Energy
Requirements in a Sparged Fixed Film System" Presented at the Second
International Conference on Fixed Film Biological Processes, Arlington,
Virginia, July 10-12, 1984.
7. Standard Methods for the Examination of Water and Wastewater, 15th ed.
Am. Public Health Assoc., Washington, D.C., 1980.
8. Sawyer, C.N. and McCarty, P.C. Chemistry for Environmental Engineering,
3rd ed., McGraw-Hill, 1978, pp. 506-508.
558
-------
FINE PORE AERATION PRACTICE
by
William C. Boyle
Department of Civil and Environmental Engineering
University of Wisconsin
Madison, Wisconsin 53706
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
559
-------
FINE PORE AERATION PRACTICE
by: William C, Boyle
Dept, of, Civil & Environmental Engineering
University of Wisconsin
Madison, Wisconsin 53706
ABSTRACT
This paper describes the current state-of-the-art of fine pore aeration
of municipal wastewaters. It represents a condensation of an interim guide-
lines report that will be published by United States Environmental
Protection Agency (U.S. EPA) in the Fall, 1985. The material presented
herein is the result of effort by the ASCE Subcommittee on Oxygen Transfer.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agencies peer and administrative review policies and approved for
presentation and publication.
560
-------
INTRODUCTION
Aerobic biological processes continue to be one of the more popular
methods employed to treat municipal and industrial wastewaters. The supply
of oxygen to the biomass in activated sludge and aerated lagoons represents
the single largest energy consumer in the treatment plant. Recent studies
indicate that from 50 to 90 percent of the net power demand for a treatment
plant lies within the aeration system (1). A rather general survey of data
available in 1982 on municipal and industrial wastewater treatment installa-
tions suggests that in the North American Continent there are approximately
1.75 million installed horsepower of aeration equipment in place at an
installed value of 0.6 to 0.8 billion dollars (2). Operating costs for
these systems may be expected to be about 0.6 billion dollars per year.
Originally, oxygen was diffused into wastewater through perforated
pipes located at the bottom of the aeration tank. The development of the
porous plate for aeration was considered an important advance in the
diffused air process because of the high transfer efficiency of this fine
pore device(3). Porous diffuser plates were used as early as 1916 and
became the most popular method of aeration in the 1930s and 1940s (4,5). It
was clear shortly after the development of porous diffusers that clogging
could be a problem. Early work on clogging led to the use of coarser media
(6) and eventually to large orifice devices (7). Use of mechanical aeration
devices was another answer to the clogging problem although these devices
were normally applied to small treatment facilities and industrial waste
applications (7).
The energy crisis of the early 1970s brought new awareness within the
sanitary engineering community relative to the efficiency of oxygen transfer
systems. Although there is some controversy over which aeration device pro-
vides the most favorable performance, the fine pore diffusion of air has
gained renewed popularity as a very competitive system. Yet considerable
concern has been registered regarding the performance and maintenance of
fine pore systems owing to their susceptibility to clogging. Diffuser
clogging, if severe, may lead to deterioration of aeration efficiency
resulting in an escalation of power costs. Furthermore, troublesome
maintenance of diffusers may consume considerable amounts of operator time
and plant operations money.
The term "fine bubble" aeration is elusive and difficult to define in
specific terms. For purposes of this report, the term fine pore diffusers
is used instead of fine bubble to more nearly reflect the characteristic of
the diffusers themselves. Fine pore diffusers are defined herein as porous
air diffusing devices that produce an initial component of surface tension
equal to or greater than 5 cm (2 in) water gauge.
The objective of- this report is to provide current information on the
performance, operation, maintenance, and retrofitting of fine pore aeration
in municipal wastewater treatment systems.
561
-------
TYPES OF FINE PORE DIFFUSERS
FINE PORE MEDIA
There exists in the marketplace today a number of porous materials
capable of serving as effective aeration devices. In general, a wide range
of products which were initially developed to filter air or liquid have also
been found to act as a satisfactory air diffusion device. Because of cost
and specific characteristics, only a relatively few types of materials are
actually being used in the wastewater treatment field.
Ceramic Materials
The oldest and still the most common type of porous material on the
market is the so-called ceramic type. As a general description, it consists
of crystalline or irregular shaped particles bonded together into various
shapes to produce a network of interconnecting passageways through which the
air flows. As the air emerges from the surface pores, pore size, surface
tension, and flow rate interact to produce the characteristic bubble size
(8). There are a number of types of ceramic diffusers including glass and .
resin bonded silica and alumina. The two most popular types are the glass
fused silica and alumina.
The silica material is produced from naturally occurring sand
particles. After screening to obtain the desired uniform particle size, an
amorphous glass binder is added. The aggregate and binder mix is then
pressed in a mold to produce the desired shape. After pressing, the
material is fired at approximately 980°C (1800°F). At this temperature, the
binder material encapsulates the sand particles. When the mix is cooled, a
glass bond is formed by the binder material at the contact points between
the individual particles.
The fused alumina material is made from aluminum oxide. The actual
grains are produced by melting bauxite ore at approximately 1120°C (2050°F)
to form large pigs. The pigs are then crushed and the resulting particles
screened to select the desired size. For the alumina, a more elaborate
binder resembling procelain is used. After pressing, the grit and binder
mix is then fired at 1425°C (2600°F) which upon cooling creates the glass
bond at the contact points. The final product is typically 80-90$ aluminum
oxide.
There are a few minor differences between the two types of material.
Because of the crushing process, the alumina grains are more angular and
jagged in shape compared to the silica. Because it is a mined material with
a limited particle size range, the pore size that can be produced with
silica is limited by naturally occurring grain sizes. This prevents its use
in applications where a large pore opening is desired.
Today, the majority of ceramic diffusers being marketed are manu-
factured from aluminum oxide. The alumina material is harder and possibly
562
-------
somewhat stronger than the silica, but this alone may not be the reason for
its widespread use.
Plastic Materials
A more recent development in the fine pore field is the use of porous
plastic materials. Similar to the ceramics, a material is created con-
sisting of a number of interconnecting channels through which the air can
pass. Advantages of the plastic material over aluminum oxide are its weight
(the light weight makes it especially suited to lift out applications),
cost, durability, and depending upon the actual material, its resistance to
breakage. Disadvantages include strength, creep resistance, and variable
dynamic wet pressure with time due to a change in contact angle.
Porous plastics are made from a number of thermoplastic polymers
including polyethylene, polypropylene, polyvinylidene fluoride, ethylene-
vinyl acetate, styrene acrylonitrile, and polytetrafluoroethylene (14).
Probably the most common type of plastic materials in use are high
density polyethylene (PE) and styrene acrylonitrile (SAN). PE is used
because it is relatively easy to process compared to other thermoplastics.
Shrinkage is low, a uniform quality product can be obtained, and small pore
sizes can be produced. The actual material is manufactured by a proprietary
process, and thus little information is available.
The major advantages of the PE media is that it is lightweight
(approximately 560 kg/m^ (35 Ibs/ft^), essentially inert, and will not
break, even when frozen. Potential disadvantages are strength, creep, and
variable contact angles. Furthermore, it is a relatively new product (at
least as an air diffusion device) and all of the long-term effects may not
be known.
The second most common type of thermoplastic material is styrene
acrylonitrile (SAN) copolymer. The raw material is a mixture of four
different molecules. Physically the media is made up of very small resin
spheres which are fused together under pressure. The SAN media has a
density only slightly greater than PE. The presence of the styrene,
however, makes the material brittle and the media can break if dropped, even
at room temperature. A major advantage of the SAN material is that it has
been in use for approximately 15 years without known deleterious effects.
Flexible Membranes
Flexible type diffusers have been in use for approximately 40 years.
They initially were referred to as "sock diffusers" and made from materials
such as plastic or synthetic fabric cord or woven cloth. As with plastic
tube diffusers, a metallic or plastic core material is necessary for
structural support. Although sock diffusers were capable of achieving
relatively high oxygen transfer rates, fouling problems were often severe.
Today, there is essentially no market for the early sock design.
563
-------
Within the last several years, a new type of flexible diffuser has been
introduced. It consists of a thin flexible sheath made from soft plastic or
rubber material. Air passages are created by punching minute slots in the
sheath material. When no air is being applied, a check valve at the inlet
prevents backflow into the distribution piping. When the air is turned on,
the sheath expands. Each slot acts as a variable aperture opening, the
higher the flow rate the greater the opening. The sheath material is
supported by a plastic tubular frame.
PHYSICAL CHARACTERISTICS OF MEDIA
There are a number of physical properties of the media that are
important to the design engineer. These include permeability, porosity,
pore size, uniformity, dynamic wet pressure, strength, chemical stability,
heat resistance, and density.
Permeability
"Permeability is defined as the volume of air in cubic feet per minute
which is passed through one square foot of diffuser, tested dry, at 5 cm (2
in) water differential pressure, under standard conditions of temperature
and humidity. This permeability rating has been the accepted standard of
comparisons for some 30 years, for want of a better practical test.
However, it does not give a true basis for comparisons of performance,
because the same permeability rating could be obtained from a diffuser with
a few relatively large pores or a multitude of fine pores. Also, two
diffusers with exactly the same pore structure would have different ratings
if of different thickness. It appears that a supplementary specification is
necessary to give a practical measure of the number and size of the pores in
a diffuser of given permeability. This is currently being studied ..."
(5).
The quotation above from the 1952 Air Diffusion in Sewage Works Manual
still reflects the status of permeability measurements in 1985, although the
study of supplementary methods has greatly intensified and should lead to a
more rational procedure for specification of porous diffusers.
As best can be determined, the ceramic industry has not "standardized"
this test procedure. The early specifications were developed for 30.5 x
30.5 cm (12 in x 12 in) plates of 2.5 cm (1 in) and 3.8 cm (1.5 in)
thickness. Today, specifications are needed for products of various shapes,
densities, and wall thicknesses, often of ill-defined effective area.
Attempts have been made to apply the principles of the test through a
parameter known as specific permeability (8). In its determination, an
airflow is measured through a diffuser mounted on a fixture similar to the
one used in service at a pressure differential of 5 cm (2 in) water gauge.
From this measurement and the geometry of the diffusers, estimates are then
made as to what the airflow would have been had the dimensions of the test
diffuser been 30.5 cm x 30.5 cm x 2.5 cm(12 in x 12 in x 1 in).
The specific permeability procedure has served to improve the utility
of this test but does not overcome the remaining deficiencies which include:
564
-------
• Clamping and sealing details are not well enough defined to provide
acceptable precision.
• Correction factors to account for pressure, temperature, and
humidity have not been developed.
• The test indirectly measures a characteristic of real interest which
is only slightly more difficult to measure directly and with greater
accuracy (Dynamic Wet Pressure on Bubble Release Vacuum (11)).
The major redeeming feature of the existing test procedure is its
simplicity. Its use should probably be retained, at least temporarily,
until more suitable standardized procedures are available.
Porosity and Pore Size
Porosity is defined as the ratio of pore volume to total volume of the
material. Porosity is relatively difficult to measure and does not have
well-defined relationships with the functional characteristics of porous
diffusers (other than density); consequently, it is rarely specificed by the
engineer or owner. A porous media will have pores which can be classified
as open, closed, or open only at one end. Since the porosity includes the
volume of all pores, even though only the open pores will transmit gas, the
porosity value can be misleading.
For ceramic diffusers, the porosity is usually in the 30 to 40/S
range. For a specific manufacturer, the porosity is about the same
regardless of the grade of material. Different grades are simply made by
varying the combination of size and number of pores. As a result, porosity
and permeability are not synonymous or directly related. Yet, pore size or
effective pore sizes is a characteristic of considerable significance since
it gives some indication of wet pressure and uniformity.
Uniformity
Uniformity of individual diffusers and the entire aeration system is of
extreme importance if high efficiency is to be maintained. On an individual
basis, the diffuser must be capable of obtaining uniform air distribution
across the entire surface of the media. If dead spots exist, chemical or
biological foulants may form and eventually lead to premature fouling of the
diffuser. Also, if small areas of extremely high flow are present, larger
bubbles may form. The diffuser will "coarse bubble" and oxygen transfer
efficiency will be reduced.
Like permeability, a porous diffuser specification should include a
requirement for testing to assure that the media will distribute air uni-
formly. A method which will quantify this uniformity actually measures the
rate of air release from different areas of the diffuser (10). With the
diffuser submerged in 5 to 20 cm (2 to 8 in) of water and at an airflow rate
of approximately 36 m^/hr/m (2 scfm/ft ), the rate of air release is deter-
mined by measuring the displacement of water from an inverted cylinder.
565
-------
Based on air volume, time, and the area of the collection cylinder, a flux
rate is determined. A comparison of the flux from various points will give
a true indication of the uniformity. Although procedures have been
presented (10), no guidelines have yet been developed in regard to the
variations between points which could exist before the diffuser would be
rejected as nonuniform. Development of such guidelines should be
undertaken.
In addition to individual diffusers, flow characteristics throughout
the system must also be uniform to assure equal airflow rates throughout the
aeration tank. System uniformity should not be a problem if the diffusers
pass the permeability requirements described in the preceding section,
provided suitable individual control devices are provided.
Wet Pressure Loss (Dynamic Wet Pressure)
Wet pressure loss can also be an important consideration in evaluating
or selecting a porous media. As a general rule, the lower the permeability,
the smaller the bubble size, and the higher the headless. While a small
bubble may increase the oxygen transfer, the additional power required to
overcome the headless may negate any potential savings.
The porous media currently in use today has a wet headless over the
typical operating flow range of 10 to 25 cm (4 to 10 in) of water. The
specific value depends on the flow rate, type of material, thickness, and
surface properties. For the ceramic and porous plastic materials, the
headloss versus flow curve is linear over the typical operating range and
the slope is relatively flat. A fourfold increase in airflow 1 to 3 m-vhr
(0.5 to 2.0 scfm) per unit for some diffusion elements will result in only a
2 to 5 cm (1 to 2 in) increase in headloss across the media itself. For the
flexible material, the small holes act like an orifice. As a result, the
headloss versus flow curve is steeper (11). Over the typical operating
range of 3 to 10 nP/hr. (2 to 6 scfm) per unit, the headless may increase
from 13 to 38 cm (5 to 15 in).
For the ceramic and plastic material, the majority of the wet headloss
is associated with the pressure required to form bubbles against the force
of surface tension. Only a small fraction of the total headloss is required
to overcome the frictional resistance (10). Thus, the thickness of the
material is not directly related to the overall wet headloss.
Wet headloss for comparison purposes may be measured in the laboratory
or the field (10). It is important that the porous diffusers be allowed to
soak for several hours (plastics require several months) prior to testing to
assure that they are completely saturated. Since the actual headloss will
be a function of the degree of water saturation in the diffusers, a slightly
different curve will be obtained if the airflow is started at a low flow
rate and is increased or vice versa. Standard practice is usually to purge
the media at the upper flow value for a predetermined time interval (5 to 10
minutes), then record additional headloss values as the flow is decreased.
566
-------
Other Characteristics
Other characteristics of porous media which may also be of importance
are strength, chemical stability, heat resistance, and density. Diffusion
media must be strong enough to withstand static head of water over the
diffuser (when air supply is off), forces applied to media under construc-
tion, and stresses of reasonable handling and shipping. All the materials
described herein are impervious to normal concentrations of chemicals used
in wastewater treatment including periodic exposure to strong acid solutions
used for cleaning. Temperature resistance is not a problem under normal
operating conditions. Density is of concern where diffusers are to be
lifted out of the tank for routine maintenance.
TYPES (SHAPES) OF FINE PORE DIFFUSERS
The nature of the majority of fine pore media (ceramic and porous
plastic) is such that it can be molded into practically any shape that is
desirable. There are, however, a few practical guidelines which tend to
influence the design. These include minimizing vertical edges to prevent
bubble coalescing, keeping the shape simple to prevent molding problems, and
some restrictions of size.
Today there are four general shapes of porous bubble diffusers on the
market, they include plates, tubes, domes, and discs.
Plates
The original fine pore diffuser design was a flat rectangular plate.
The plates are typically 30 cm (12 in) square and 2.5 to 3.8 cm (1 to 1.5
in) thick. They are manufactured from either glass bonded silica or glass
bonded aluminum oxide. The plates are installed in the tank by grouting
into recesses in the floor, cementing into prefabricated holders, or clamped
into metal holders. A chamber underneath the plates acts as an air
plenum. The number of plates fixed over a common plenum is not standard and
can vary from only a few to 500 or more. In current designs, individual
control orifices are not provided on each plate.
In the early activated sludge plants (1910 to 1920s), fine pore plates
were used almost exclusively as the method of air diffusion. Today, other
than in some of the original plants, fine pore plates are not being
installed.
Tube Diffusers
Like the plates, fine pore tubes have been used in wastewater treatment
for a number of years. The early tubes were Saran wound or made from
aluminum oxide and have been followed by the introduction of styrene
acrylonitrile copolymer (SAN), porous plastic (PE), and, most recently, the
new generation of flexible media.
567
-------
All the tube diffusers on the market are of the same general shape.
Typically, the media portion is 51 to 61 cm (20 to 24 in) long and has an
O.D. of 6.3 to 7.6 cm (2.5 to 3.0 in). The thickness of the media is
variable. Flexible membranes are very thin, commonly in the 0.5 to 1.3 nun
(0.020 to 0.050 in) range. The PE media is usually supplied in a 6.3 mm
(0.25 in) thickness, the SAN approximately 15.2 mm (0.6 in), and fused
ceramic material in the 9.5 to 12.7 mm (0.375 to 0.5 in) range.
The holder designs for the ceramic and plastic media are very similar.
However, plastics need structural reinforcing. Most consist of two end caps
held together by a connecting rod through the center. For the flexible mem-
brane diffusers, the end caps and support frame are one piece. The membrane
material is installed over the support frame and clamped on both ends.
Tube diffusers are designed to operate in the 3 to 17 m-vhr (2 to 10
scfm) range. Because of their inherent shape, it is sometimes difficult to
obtain air discharge around the entire circumference of the tube. The air
distribution pattern will vary with different types of diffusers. In
general, the prevalence of inoperative area will be a function of the air-
flow rate and the headloss across the media. Because dead areas can provide
sites for slime growth, it would be beneficial prior to selecting a particu-
lar design of tube, to observe its performance on a laboratory or pilot
scale basis to assure that proper air distribution will be obtained at the
design flow conditions.
Most tube assemblies are fitted with a control orifice inserted in the
inlet nipple to aid in air distribution throughout the system. Typically
the orifice is approximately 12.7 mm (0.5 in) in diameter, although dif-
ferent sizes can be used for various design flow rates. Also, some
assemblies include check valves to prevent the backflow of liquid into the
air piping.
Dome Diffusers
The fine bubble dome diffuser was developed in Europe in the 1950s and
introduced in the U.S. market in the early 1970s (12). Considered as the
standard in England and some parts of Europe, domes are now installed in a
number of U.S. plants.
The dome diffuser is essentially a circular disc with a downward turned
edge. Currently, these diffusers are 17.8 cm (7 in) in diameter and 3.8 cm
(1.5 in) high. The media is approximately 15.2 mm (0.6 in) thick on the
edges, and 19.0 mm (0.75 in) on the top or flat surface.
domes presently are being made only from aluminum oxide.
The dome diffuser is mounted on either a PVC or mild steel saddle type
base plate. The diffuser is attached to the base plate by a bolt through
the center of the dome. The bolt can be made from a number of materials
including brass or stainless.
568
-------
The slope of the headloss versus airflow curve for a ceramic diffuser
is very flat. It has been reported that a variation from the average of 4^
in the specific permeability can result in a 200% change in airflow for the
same headloss (13). To better distribute the air throughout the system,
control orifices are placed in each diffuser assembly to create addition
headloss and balance the airflow. The fastening bolt is hollowed out and a
small hole drilled in the side or the orifice is drilled in the base of the
saddle. The size of the orifice is typically 5.0 mm (0.2 in).
Dome diffusers are usually designed to operate at 1.7 m^/hr (1.0 scfm)
with a range of 0.8 to 3 m-Vhr (0.5 to 2.0 scfm). In designing a system,
careful consideration should be given to the desired airflow range. Testing
has shown that the oxygen transfer efficiency (OTE) is dependent on the
airflow rate per diffuser increasing as the airflow rate decreases. This
can lead to the temptation to design systems to operate at 0.7 to 0.8 m-vhr
(0.4 to 0.5 scfm)/diffuser. Although favorable in terms of oxygen transfer,
this practice can lead to operational problems. At low airflow rates,
uniform air distribution across the entire diffuser surface may be difficult
to obtain. Also, at 0.8 m^/hr (0.5 scfm), the headloss across the control
orifice will be less than 25 mm (1.0 in) water. At low airflow rates, the
orifice will not serve its intended purpose of balancing the air throughout
the system. In any case, if either the entire area or portions of the
diffuser are not discharging air, foulant development can begin which could
lead to a premature fouling of the system.
The upper limit for airflow through a dome diffuser is usually
considered to be 3 nr/hr (2.0 scfra). Operation above this level is
possible, but is not very economical. Increasing the flow rate results in a
continued decrease in OTE and may require a larger control orifice.
Disc Diffusers
Disc diffusers are a relatively recent development. Discs are flat, or
relatively so, and are differentiated from the dome diffuser in that they do
not include a downward turn peripheral edge. While the dome design is
relatively standard, currently available disc diffusers differ in size,
shape, method of attachment, and type of material.
Disc diffusers are available in diameters which range from
approximately 18 to 24 cm (7 to 9.5 in) and thicknesses from 13 to 19 mm
(0.5 to 0.75 in). With the exception of two designs, all consist of two
flat parallel surfaces. Although the majority of disc diffusers are made
from aluminum oxide, a porous plastic (PE) disc is also available.
Like the dome diffusers, the disc is mounted on a plastic (usually PVC)
saddle to type base plate. Two basic methods are used to secure the media
to the holder, a center bolt or a peripheral clamping ring. The center bolt
method is similar to that used with the domes. The more common method of
attaching the disc to the holder is to use a screw on type retainer ring.
With the threaded type collar, a number of different types may be used.
569
-------
In general, the retainer ring method of attaching the diffuser to the
holder has two potential advantages over a center bolt. It has been
reported (15) that as the diffusers become fouled, excessive amounts of air
are discharged from the edges and the area around the center bolt washer.
Although not specifically documented under controlled conditions, this
nonuniform airflow could reduce the OTE of the system. The retainer ring
will tend to minimize these problems. A second advantage is that breakage
of diffusers from over tightening the bolt or air leakage problems from
stretching a nonmetallic bolt can be eliminated.
For the disc diffuser designs, there are two methods of attachment to
the air piping. The first is to solvent weld the base plate to the PVC
header prior to shipment to the job site. The second technique uses a
mechanical method of attachment. This can be either a bayonet type holder
which is forced into a saddle on the pipe or a wedge section which is placed
around and clamps the holder to the pipe.
Disc diffuser assemblies also include individual control orifices in
each assembly. Designs employing the bolt method of attachment usually will
use a hollow bolt with a orifice drilled in its side. The other designs
will either have the orifice drilled in the bottom of the diffuser holder or
the base will include a threaded inlet where a small plug containing the
desired orifice can be inserted. The diameter of the orifice is similar to
that used with the dome diffusers.
Disc diffusers have a design flow rate of from 0.8 to 5 m^/hr (0.5 to
3.0 scfm)/diffuser. The most economical operating range will, however, be
somewhat dependent on the size. The 18 cm (7.0 in) diameter discs are
usually operated in the 0.8 to 3.H m^/hr (0.5 to 2.0 scfm) range, similar to
the dome diffusers. For the larger discs 22 to 24 cm (8.5 to 9.5 in), the
typical lower limit may be 1.3 to 1.5 m-vhr (0.75 to 0.9), up to an upper
limit of 4.3 to 5.1 nP/hr (2.5 to 3.0 scfm). Prolonged operation at flow
rates less than 1.3 m^/hr (0.75 scfm) is not desirable with a large disc
because insufficient air is available to assure good distribution across the
entire surface of the media. In those applications where operation above
3.4 nP/hr (2.0 scfm) is desirable, the control orifice should be sized
accordingly so that the headless produced does not adversely affect the
economics of the system.
Clean water testing has shown that the oxygen transfer efficiency is
related to diffuser size (12,14). A fewer number of large diameter discs
are required to achieve the same oxygen transfer results. Apparently, the
additional surface area results in a lower flux rate, thus the higher
transfer. There is, however, no generally accepted ratio for comparing the
various size diffusers. As a range, one 23 cm (9 in) diameter disc has been
found to be approximately equivalent to 1.1 to 1.4 - 18 cm (7 in) discs.
The actual ratio is related to airflow rate and diffuser submergence. Since
the surface areas are nearly the same, a 17.8 cm (7 in) diameter disc and
dome should achieve comparable results.
570
-------
AIR PIPING AND DIFFUSER LAYOUT
Fine bubble plates are normally installed in either total floor
coverage or spiral roll. Total floor arrangements may include closely
spaced rows (transverse or longitudinal) or incorporated in a ridge and
furrow design. Total floor layouts produce higher oxygen transfer
efficiencies than the more efficient mixing spiral roll.
Most tube diffusers can be attached to existing left out assemblies
which make them especially suited for retrofit. They are often installed
parallel along one or both long sides of the aeration basin. Newer designs,
however, employ cross roll or full floor coverage patterns.
Both domes and discs are installed in a total floor coverage or grid
type pattern. In some cases where mixing may control the design (near the
end of long narrow tanks), the diffusers can be placed in tightly spaced
rows along the side or middle of the basin to create a spiral type mixing
pattern. The diffusers are usually mounted as close to the tank floor as
possible, within 23 cm (9 in) of highest point being typical.
PERFORMANCE CHARACTERISTICS
In the late 1960s and early 1970s, consulting engineers began
specifying clean water performance tests to be conducted by the aeration
equipment suppliers as a means of verifying.aerator performance. Various
engineers developed their own testing criteria.
In April 1978, a "Workshop Toward An Oxygen Transfer Standard" (15) co
to sponsored by the U.S. Environmental Protection Agency and the American
Society of Civil Engineers was held in an effort to obtain concensus
standards for the evaluation of aeration devices in both clean and process
water. The outcome of the workshop was the formation of an Oxygen Transfer
Standards Subcommittee under ASCE (16) and progress toward the development
of process water test procedures (17).
CLEAN WATER PERFORMANCE
This section presents a distillation of clean water performance data on
fine pore diffusion devices. Some but not all of the data was generated
using the current ASCE recommended clean water standard (16). Thus, the
oxygen transfer results presented in this section reflect the utilization of
the current nonlinear least squares method of analysis as well as a prior
procedure using the linear least squares log deficit analysis (16). The
latter method permitted data truncation. Both methods produce comparable
results under ideal testing conditions. Every effort has been made to
screen the data reported herein and to omit data of questionable validity.
The results of clean water transfer tests are reported herein as
Standard Oxygen Transfer Efficiency (SOTE), Standard Oxygen Transfer Rate
(SOTR), or Standard Aeration Efficiency (SAE). Standard conditions are
571
-------
defined as clean water (alpha = 1.0, beta = 1.0), Temperature = 20°C,
Atmospheric Pressure = 1.0 atmosphere and D.O. = 0.0 mg/1.
One of the critical parameters required for the calculation of oxygen
transfer rates is the equilibrium D.O. saturation concentration, C*. For
submerged aeration applications, C* is significantly greater than the
surface saturation value, C*, tabulated in most standard tables (17). It
is therefore necessary to either calculate C* (17), or to measure it during
clean water tests (16). This value is primarily dependent upon diffuser
submergence, diffuser type, tank geometry, and gas flow rate. One of the
more comprehensive evaluations of C* in clean water tests was reported by
Yunt et al. (18).
The performance of diffusers under clean water test conditions are
dependent upon a number of factors in addition to those standardized in the
calculation of SOTS, SOTR, or SAE. Among the most important factors are
diffuser type, diffuser placement and density, gas flow rate per diffuser,
and tank geometry.
Typical oxygen transfer efficiencies for fine bubble diffused air
systems are presented in Table 1 and Figure 1. These data are reported for
a 4.6 m (15 ft) diffuser submergence. The effect of diffuser type, place-
ment, and airflow per diffuser are clearly delineated from this summary of
eight different clean water studies. In general, it can be observed that
ceramic domes and discs demonstrate slightly higher clean water transfer
efficiencies than typical plastic porous media or flexible membrane tubes in
a grid placement. Both tubes and discs/domes are significantly superior to
all coarse bubble placements. Within a given diffuser type, spreading the
diffusers more uniformly along the tank bottom area (spiral to dual or grid)
tends to improve clean water performance.
TABLE 1. CLEAN WATER OXYGEN TRANSFER EFFICIENCY COMPARISON
(SUBMERGENCE - 15 FT)
Diffuser Airflow
Type & Placement scfm/dlffuser SOTE - % Reference
Ceramic Discs-Grid 0.6 to 2.9 36 to 25 14
Ceramic Domes-Grid 0.5 to 2.5 39 to 27 14,18.19,20
Plastic Porous Media Tubes
Grid 2.4 to 4.0 32 to 28 21
Dual 3.0 to 9.7 28 to 18 14,18,22
Spiral 2.0 to 12.0 25 to 13 14,22
Flexible Membrane Tubes
Grid 1-4 29 to 22 23
Quarter Point 2-6 24 to 19 23
Spiral Z-6 19 to 15 23
Coarse Bubbles
Dual 3.3-9.9 13 to 12 18,24
Midwidth 4.2-45 13 to 10 18,24
Spiral 10-35 12 to 9 18,24
572
-------
§
v- IT > 5,
-------
are higher (thereby resulting in a greater driving force), and there is
opportunity for longer bubble residence time in the aeration tank. The SAE,
however, remains relatively constant (or may decrease) for the fine bubble
diffusers as depth increases since power requirements to drive the same
volume of air through diffusers at the greater depths will increase. Note,
however, that coarse bubble diffusers exhibit a gradually increasing SAE
with submergence but never reach efficiencies achieved by the fine bubble
systems.
50
40
-» 30
5?
i
20
10-
20'x 20'Tank
-f-
10
-t-
25
15 20
Water Depth (ft.)
Note: Domes at 0.5 hp/1000 cu.ft.; other systems at 1.0 hp/1000 cu.ft.
Figure 2. Oxygen transfer efficiency vs depth (4).
574
-------
10
8-
6- •
4-
2-
20ft. x 20ft. Tank
Dome Grid
Coarse
Bubble
-H-
20
-I-
25
10
15
Water Depth (ft.)
Note: Domes at 0.5 hp/1000 cu.ft.; other
systems at 1.0 hp/1000 cu.ft.
Figure 3. Wire aeration efficiency vs depth
The clean water transfer efficiencies reported in Table 2 are for
polyethylene or plastic porous media tubes. The dual and spiral rolls data
is from Yunt et al. (18), Huibregtse et al. (14), and Paulson (23). The
grid results are from Popel (22). The tube performance is typical of fine
bubble diffusers and exhibits the effect of increase diffuser density.
Popel observed that increased diffuser density in grids decreases upward
flow velocities and therefore increases the retention time of bubbles.
575
-------
TABLE 2. CLEAN WATER OXYGEN TRANSFER EFFICIENCY -
POROUS PLASTIC TUBES
Placement
Grid*
Dual Roll
Spiral Roll
Airflow
scfm/dlffuser
2.4 to 4.0
3.2 - 6.3
9.0 - 9.7
2.0 - 6.7
8.0 - 12.0
10 ft
—
16-11
14-10
15-12
15-10
SOTE at
Water Depth - X
15 ft
32-28
24-17
15
20-15
17-10
20 ft
-
32-22
26-21
25-22
22
*7.7 ftz/tube at 13.6 ft water depth. Tank 14.4 ft x 108.2 ft
Typical performance of a flexible membrane diffuser from Wyss (23) is
presented in Table 3.
TABLE 3. CLEAN WATER OXYGEN TRANSFER EFFICIENCY -
FLEXIBLE MEMBRANE TUBES
Placement
Floor Cover
(Grid)
Quarter Points
Center Roll
Spiral Roll
Airflow
scfm/dlffuser
1 to 4
2 to 6
2 to 6
2 to 6
10 ft
18-14
15-13
11-9
11-7
SOTE at
Water Depth - X
15 ft
27-21
22-18
18-15
18-14
20 ft
35-29
29-24
27-23
28-21
•Density 4-8 ft2/tube
This diffuser also exhibits a decreasing transfer efficiency with increasing
airflow rate. The effect of diffuser placement is also shown. The increase
for quarter-point placement in a rectangular basin is greater than the mid
to width placement.
The clean water oxygen transfer efficiency of disc/dome grid systems
are illustrated in Table 4 and Figure 4. The results are from studies by
Huibregtse et al. (14), Yunt et al. (18), Sullivan and Gilbert (20), and
Paulson (21). This type of system has produced the highest transfer
efficiencies reported for fine bubble devices. The density of placement is
greater than the tube-grid systems and the airflow rates per diffuser are
lower. Huibregtse (19) reported a slightly increased transfer efficiency
with a 238 mm (9.4 in) dia disc versus a 178 mm (7 in) dia dome. Houck and
576
-------
lower. Huibregtse (19) reported a slightly increased transfer efficiency
with a 238 mm (9.4 in) dia disc versus a 178 mm (7 in) dia dome. Houck and
Boon (12) have also reported a similar relationship between dome or disc
diameter and oxygen transfer per diffuser.
TABLE 4. CLEAN WATER OXYGEN TRANSFER EFFICIENCY -
CERAMIC DISC/DOME GRID SYSTEMS
Diffuser Density
sq. ft/dfffuser
DUc-9.4"
6.4
4.1
3.2
Airflow
scfm/dlffuser
0.9-3.0
0.8-2.9
0.7-2.6
SOTE at
Water Depth
10 ft
22-20
24-21
25-22
15 ft
31
34-30
34-31
- X
20 ft
37-34
41-35
41-38
Ref.
5
5
5
Oome-7"
5.6
4.2-4.4
3.2-3.3
2.2-2.5
1.5
0.5-2.0
0.5-2.5
0.5-2.0
0.5-2.5
0.5-2.5
—
23-16
24-20
23-17
26-18
31-25
32-25
37-27
35-27
34-27
40-28
41-30
44-31
47-33
6
6.7
4,5,6
6,7
7
As the disc or dome density increases (decreasing area per diffuser),
the transfer efficiency increases for a given airflow rate. The studies of
Sullivan and Gilbert (20) show a nonlinear decrease in transfer efficiency
while the results of others indicate a linear decrease in transfer
efficiency with increasing airflow rate.
PROCESS WATER PERFORMANCE
In 1978, Eckenfelder opened the Workshop Toward an Oxygen Transfer
Standard (15) by indicating that compared to clean water testing, "it is far
more important to define aerator performance under field operating
conditions." Prior to 1981, there was a general lack of consistency in the
methods used to evaluate aerator performance under process conditions, and
there existed a paucity of coherent data on dirty water performance. Since
that time, significant data under process conditions have been developed;
however, there is a continuing need to expand this data base.
Another area where coherent data are lacking in the literature pertains
to measurement of alpha values for various aeration devices. A great amount
of reported alpha data were obtained from bench scale units which do not
properly simulate mixing and Kla levels, aerator type, water depth, and the
geometry affects of their full-scale counterparts. As a result, much of the
reported data on alpha values, particularly for diffused aeration systems is
of questionable validity. Reliable full-scale test procedures for use under
process conditions coupled with good clean water performance data are
required to overcome these difficulties.
577
-------
45--
40--
35--
UJ
5
(0
30--
25--
20
Water Depth = 15.0 ft.
Increasing Density of Placement
2.2 sq.ft./Diffuser
5.4 sq.ft/Diffuser
0.5 1.0 1.5 2.0
Air Flow Rate Per Diffuser (scfm)
2.5
Figure 4. Effect of diffuser density on oxygen transfer.
In 1981, the ASCE Oxygen Transfer Standards Committee undertook a study
to evaluate, in parallel, four principal methods for estimating oxygen
transfer under process conditions at seven sites (25). These methods
include the steady to state method, the nonsteady state method, an off-gas
analysis procedure, and two tracer techniques. Table 5 summarizes the
factors which affect the selection of the best method to use to measure
process OTE values. It is the intention of this Committee to prepare a
manual of methods for testing oxygen transfer devices under process
conditions.
578
-------
TABLE 5. SELECTED FACTORS AFFECTING OXYGEN TRANSFER FIELD TESTING
FOR ESTIMATION OF OXYGEN TRANSFER EFFICIENCY (26)
Oxygen Transfer Teats
FACTORS Inert Gas
SS NSS OG Tracers
Sensitivity To: ~~ ~~
Variations in -
Influent Waatewater Flow
Oxygen Uptake Rate
Alpha
Dissolved Oxygen Conoentr. -
Product of (air rate x Kla)-
Accurate Measure of -
Oxygen Uptake Rate - +
Dissolved Oxygen
D.O. Saturated Value + +
Air Flow Rate
Other
Costs:
Manpower
Analytical
Capital Invest.
Calculations:
Estd. Precision:
-
+
+
+
+
+
-
_
+
0
+
+/0
0
0
1
2
0
0
0
0
+
_
3
0
_
_
0
*
1 Calculate OTE directly S3 = Steady State
2 Requires accurate measurement of C02 in gas NSS = Nonsteady State
3 Requires accurate estimate of ratio OG = Off-gas
Kj._aoer/Kla - especially in diffused air systems
+ Positive response... (eg. not sensitive, less costly,
more precise, easier)
0 Intermediate response
Negative response
The factors which affect the performance of diffused aeration systems
under process conditions include wastewater characteristics, process type,
flow regime, loading, geometry, equipment configuration, diffuser fouling,
operational control, mechnical integrity, and preventative maintenance. In-
process testing of selected diffusers under a variety of test conditions
will eventually provide the data base needed by engineers to intelligently
transfer standardized clean water performance data to field conditions.
This data is now being collected. A partial list of that data base is pre-
sented below.
Table 6 presents process water performance data for thirteen individual
evaluations of various sites employing a variety of diffused aeration de-
vices. Each set of data represents the observed performance of a particular
system over a period of several hours and is in no way suitable for the
design of similar systems. The intention of this table is to give the
reader a general feeling for the relative performance of the systems listed
under a variety of operating conditions.
In all cases, the OTE values reported have been converted to 20°C and
zero residual D.O. "Apparent" values of alpha have been estimated from
clean water performance data for similar geometry, air rate per unit, and
equipment placement. Since the performance of most porous diffusion devices
is likely to change with time due to fouling, the term "apparent alpha" has
been adopted to distinguish between the total difference in performance from
clean water as compared to differences due to waste characteristics only
(alpha).
The first three data sets originate from off-gas testing at Madison,
Wisconsin (26,27). The ceramic grid data represents the overall performance
of a three pass system.
579
-------
TABLE 6. IN PROCESS OXYGEN TRANSFER DATA
CO
o
In Process SOTEf (%)
Madison,
Madison,
Madison,
Site
WI (26,27)
WI (26,27)
WI (26,27)
Whittier-Narrows, CA (28)
Whittier-Narrows, CA (28)
Brandon,
Brandon,
Orlando,
Seymour,
Lakewood
Lakewood
Brewery
Brewery
WI (27)
WI (27)
FL (29)
WI (29)
, OH (29)
, OH (29)
(29)
(29)
System
Ceramic
Grid
Ceramic &
SAN Tubes
Wide Band -
Coarse Bubble
Ceramic
Jet
Jet
Jet
Grid
Wide Band -
Coarse Bubble
Ceramic
Ceramic
Ceramic
Ceramic
Grid
Grid
Grid
Grid
Static Aerator
Flow Regime
Step Feed
Step Feed
Step Feed
Plug Flow
Plug Flow
CSTR
CSTR
CSTR
Plug Flow
Plug Flow
Plug Flow
CSTR
CSTR
Average
17.8
11.0
10.0
11.2
9.4
10.9
7.5
7.6
16.5
14.5
8.9
14.2
7.4
Range
12.
7.
7.
9.
7.
9.
7.
6.
12.
12.
7.
12.
5.
6/26.2
5/13.4
9/10.9
3/15.2
8/10.8
7/12.1
4/7.7
8/8.4
0/18.8
4/15.9
0/11.1
5/15.2
7/8.8
Apparent Alpha
Est. Mean Range
0.64
0.62
1.07
0.45
0.58
0.45
0.47
0.75
0.66
0.52
0.31
0.37
0.50
.42/.
.46/.
.83/1
.35/.
.48/.
.40/.
.46/.
.67/.
.49/.
-44/.
.26/.
.32/.
.36/.
98
85
.19
60
72
50
48
83
75
57
37
37
51
Mean Air
Flux Rate
(L/m2/sec)
1.42
2.69
0.53 2.69
1.07
1.88
0.66
1.98
4.67
0.35
0.71
0.46
1.52
2.69
-------
Figure 5 illustrates local performance in terms of OTE, air flux rate
(flow per unit surface area of tank), and residual D.O. as a function of
tank length. At this facility, alpha appeared to vary from about 0.4 at the
inlet to near 1.0 at the discharge point. Note the reduction in apparent
alpha at each point of primary effluent addition. The second data set for
ceramic and SAN plastic tubes applied in a dual spiral roll configuration
represents performance for the first pass of a three pass system. Passes
two and three are represented by the third data set employing wide band
coarse bubble diffusers also in a dual spiral roll placement. The higher
relative alpha of the latter system is strongly affected by its favorable
position at the rear of the process where significantly higher alphas are
encountered compared to the inlet end of this facility.
14
10
100 200
Tank Length (ft.)
Note: PE = Primary Effluent
RAS = Return Activated Sludge
300
0.5
•4
0.3
O.UO
2 g
Figure 5. Gas transfer analysis along tank length
floor coverage, ceramic domes, tapered air (17).
581
-------
The two data sets for Whittier-Narrows, California, were obtained in
August 1981 after approximately nine months of operation (28). The two
systems presented were part of a three system process water evaluation,
conducted by Los Angeles County Sanitation District for the U.S. EPA in
parallel trains. This evaluation compared the performance of a ceramic grid
system to that of a jet aeration system, wherein the jets were installed on
one side of the basin along the entire tank length, with the nozzles being
directed across the basin, similar to a spiral roll.
It is of interest to note that the mean apparent alpha of the jet
system was approximately 1.3 times that of the grid system. The terminal
apparent alpha at the Whittier-Narrows facility was approximately 0.7 versus
almost 1.0 at Madison. The presence of nonbiodegradable surfactants is one
explanation for the low apparent alphas at the California site. This
observation points out the danger of extracting data of this type for design
purposes. Each treatment facility has unique characteristics, which must be
considered individually.
The Brandon, Wisconsin, data depicts performance of a 9.1 m (30 ft)
long by 4.6 m (15 ft) wide by 4.6 m (15 ft) deep complete mix aeration tank
using jet aerators at two different airflow rates. This municipal facility
treats a combination of domestic and industrial wastewater.
The Orlando system, which employs wide band coarse bubble diffusers, is
included because the system treats domestic wastewater only. This system is
currently being retrofitted with a ceramic grid system in an effort to
improve aeration efficiency and increase aeration capacity at this site.
Data from Seymour, Wisconsin, a site analyzed by Houck (30) in his
North American survey of disc and dome systems and also studied by Vik et
al. (3D was lightly loaded at an SRT in excess of 25 days at the time of
the tests.
The two data sets from Lakewood, Ohio, demonstrate the relative
performance of two parallel basins, one recently cleaned and the other
operating for approximately a year with no diffuser cleaning. The entire
system was retrofitted with ceramic disc diffusers in a grid configuration
during 1982 and 1983. In this instance, the uncleaned system as found was
performing at a mean weighted SOTE of 8.9? versus 14.5$ for the cleaned
system. During the operating period of about a year, it appears that in to
process performance deteriorated by roughly 40$. Part of the deterioration
in performance may be due to several periods of air interruption which
occurred during retrofit.
The last two data sets provide information on both a ceramic grid and a
static tube system that were tested within the same complete mix basin. Of
interest is the relative performance of both systems as it pertains to rela
to tive gas phase efficiency and apparent alphas. The ratio of alpha of the
static tube system to that of the ceramic grid system was observed to be
0.50/0.37 or 1.35. The ratio of field OTEs at zero D.O. are roughly 1.9
to 1.
582
-------
Table 7 presents data from a field evaluation of selected tubular
diffusers at Madison, Wisconsin, within the last pass of a three pass system
employing a dual spiral roll configuration (26). Since alpha approached
unity at this location, direct use of the data is not suitable for design
purposes. The relative performance of new and used ceramics and the SAN
plastic tubes is interesting. The used ceramic and SAN plastic tubes were
in service continuously for about three years in a different tank prior to
relocation for this test. It should be pointed out that analysis of
multiple systems within a given tank cannot be conducted by any other
technique than off-gas analysis. Known data on tube systems and membrane
systems is scant.
The data described above represents a diverse cross section of in-
process diffuser performance under a variety of conditions. No attempt has
been made here to correlate performance to loading, process criteria,
wastewater characteristics, etc. It is evident that several gaps in our
current knowledge still exist for which additional indepth study is needed
to address designer concerns.
TABLE 7. IN-PROCESS TUBULAR DIFFUSER COMPARISONS WITH OFF GAS
PROCEDURES AT MADISON, WISCONSIN (27)
Diffuser P.O. (mg/1) Mean
Wideband Fixed Orifice
Flexible Membrane Tube
Used Pearlcomb Tube
Wide Band Fixed Orifice
New Ceramic Tube
Used Ceramic Tube
Wide Band Fixed Orifice
0.9
1.7
1.7
2.0
1.7
1.7
1.2
8.29
14.18
11.33
10.28
15.96
11.00
8.29
On the basis of the data presented in Table 6, it appears that the
relative alpha values between ceramic grid and other more turbulent systems
such as jets, static tubes, and fixed orifice systems may not be as great as
previously reported in the literature (32). In addition, the overall
average apparent alpha values presented in Table 6 and elsewhere (7,33,3*0
appear lower than many that are normally used for design purposes.
Another major factor affecting aerator performance in wastewater is
system loading and effluent criteria, especially with respect to nitrifying
and non-nitrifying aeration designs. A recent study at Rye Meads, U.K.,
(35) exemplifies the impact of process goals relative to aeration
efficiency. In this study optimization of the nitrification process in
conjunction with an anoxic zone, tapering the aeration system to meet oxygen
demand, and the use of D.O. to control airflow to the system resulted in
overall transfer of 2 kg 02/KWh versus 1.2 kg 02/KWh for an unmodified
control basin. A third parallel train employing tapered air and D.O.
control in a non-nitrifying operational mode averaged about 1.4 kg 02/KWh
583
-------
during the study phase. Both the proper dispostion of ceramic diffusers and
automated D.O. control were identified as essential elements of aeration
efficiency optimization.
Another factor of concern was observed in a recent long-term study of
ceramic grid systems (36) where it was observed that the slope of log OTE
versus log applied air rate under process conditions had a significantly
steeper negative slope with increasing air rate than observed under clean
water conditions. Other investigators have not observed this phenomenon
(35). It is evident that generalizations on this characteristic cannot be
made due to the variable nature of the phenomenon under a multitude of
operating conditions. It is an area worthy of additional study.
MAINTENANCE CONSIDERATIONS
One important factor that will influence the selection of a diffuser
system for new or retrofit facilities deals with system maintenance
requirements. A major concern voiced regarding fine pore diffused aeration
is the fouling of the diffuser and the impact of fouling on performance.
DIFFUSER FOULING
Porous ceramic diffusers, introduced in the U.S. in the 1920's were the
predominant air diffuser at mid-century (5,7). Types of fouling were
identified by early investigators and the list expanded by recent
investigations to include (10).
Air Side
1. Dust and dirt from unfiltered air
2. Oil from compressors or viscous air filters
3. Rust and scale from air pipe corrosion
4. Construction debris due to poor cleanup
5. Wastewater solids entering through diffuser or pipe leaks
Liquor Side
1. Fibrous material attached to sharp edges
2. Inorganic fines entering media at low or zero air pressure
3. Organic solids entering media at low or zero air pressure
4. Oils or greases in wastewater
5. Precipitated deposits, including iron and carbonates
6. Biological growths on diffuser media
The rate of fouling was typically gauged by the rise of back pressure
in service. Since significant microbiological fouling can take place with
little attendant rise of back pressure, this provided a crude and
qualitative measure at best.
584
-------
It was common practice at that time to operate a number of diffusers
off a common plenum. This resulted in less uniformity than is obtained
today under present day practice of individual flow control. The lack of
uniformity probably augmented the rate of microbiological fouling.
In the sixties and early seventies, the relative cost of energy to
maintenance labor was low. As a consequence, many of thoseinstallations
were replaced with less efficient fixed orifice diffusers. In the middle
seventies, this trend was reversed and many of those installations were
replaced by porous media diffusers with individual air controls.
In the early eighties, better methods of measuring the degree of
fouling and the effects of cleaning became available. These methods include
dynamic wet pressure, bubble release vacuum, the ratio of one to the other,
and chemical, as well as microbiological analysis. The practice of
employing pilot diffusers which could be removed from the tank and
individually analyzed came into use (9,10).
Concurrently, better methods were developed to measure the performance
of operating systems which permitted better appraisal of the effects of
fouling, facilitating better scheduling and maintenance. These methods
included the use of gas tracers, off-gas analysis, dual nonsteady state with
peroxide, and D.O. and respiration rate profiles (17,25,27).
It is not surprising, however, that with respect to a complex
phenomenon such as fouling, much remains to be learned in spite of the work
that has been done. On the other hand, a number of aspects are better
understood and a number of hypotheses have been developed that better
explain the observed effects than heretofore possible; a general discussion
of some of these follow.
One type of fouling that has been observed leads to a significant
increase in headless across the diffuser wherein fouling rates appear to be
greater in the regions of high localized flux rates (operating pores).
Examples of this type of fouling are believed to be air side particulate
fouling and water side precipitate fouling such as calcium carbonate and
iron hydroxide. This type of fouling appears to be typically accompanied by
an increase in back pressure that may be greater than the capabilities of
the air supply system. This may further result in insufficient air for
process requirements.
It is also of interest that with this type of a relationship between
air flux rate and fouling rate, increases in back pressure can be
accompanied by improvements in transfer efficiency (10). Figure 6 is an
idealized plot of DWP and OTE versus time with a foulant of this type.
It appears that another type of fouling exists that can lead to
significant reductions in oxygen transfer efficiency with modest, if any,
increases in back pressure. There is some evidence to suggest that this
type of fouling tends to take place in areas Qf relatively low flux rates
(nonoperating pores), such as the underside of tubular diffusers and along
585
-------
the less pervious areas of planar diffusers. As fouling progresses, flux
rates in the more pervious areas are believed capable of increasing to a
level at which a number of diffuser pores function in concert. This serves
to reduce the component of resistance to flow provided by surface tension,
and correspondingly increases the component provided by frictional
resistance to flow through the porous media. It is believed that the net
change in resistance may be either positive or negative. Since the fouling
occurs at a greater rate in areas of low flux rate, in time, the uniformity
of air distribution is imparied as is the oxygen transfer efficiency. An
example of this type of fouling is believed to be the progressive growth of
microbiological slimes in areas of low flux rates (10,29). Figure 7 is an
idealized plot of DWP and OTE versus time with this type of foulant.
CL
1
I
UJ
5
Time
Figure 6. High local flux rate fouling.
586
-------
The variables that appear to affect the rate of biofouling are not
fully understood. Experience and test data (10,29) give some indication
that the rate of biofouling is increased by operation at high organic
loading (expressed on either a volumetric or per unit biomass basis) and/or
low air rates. Some experience indicates that the rate of biofouling may be
accelerated by the presence of certain types of soluble industrial wastes,
particularly high strength, readily biodegradable, or nutrient deficient
ones (37,38).
Other types of fouling can also be factors. Air system interruptions
can allow inorganic and organic solids to enter the diffusion medium through
the diffuser surface. There is no unanimity of opinions as to the
consequence of this type of fouling. Fouling of the air side of diffusers
i i
Time
Figure 7. Low local flux rate fouling.
587
-------
by the presence of mixed liquor in the air system can be serious but is
considered to be preventable through design and specification.
The origin of many of the foulants which have been determined by
analysis are understood. One exception is silica, which as been found in
quantity at a number of sites (29,38).
It is believed that under service conditions, all the types of fouling
discussed about and some others in addition can occur singly or in com to
bination and with variable dominance from plant to plant and within the same
plant from time to time.
A good deal of data is available and is in the process of being
assembled regarding fouling and its effects. Unfortunately, consistent
methods of reporting have not as yet been developed.
Table 8 is a compilation of fouling rate data from a number of ceramic
diffused air municipal plants considered representative of experience today
(29).
TABLE 8. REPRESENTATIVE CERAMIC FOULING DATA - MUNICIPAL PLANTS (29)
PLANT DIFFUSER TIME OF EXPOSURE ABRV ADWP ADWP/ABRV
(Days) (1n/yr) (in/yr)
A
A
A
A
A
B
C
D
E
G
G
K
K
L
L
M
M
N
N
Disc
Disc
Disc
Disc
Disc
Disc
Disc
Dome
Disc
Disc
Disc
Dome
Dome
Dome
Dome
Dome
Dome
Disc
Disc
120
120
133
360
90
365
365
1100
210
93
93
210
210
360
360
350
350
900
900
9.1
17.3
8.2
7.6
20.0
10.5
17.0
10.3
20.6
_
_
13.9
10.8
4.8
6.7
6.0
13.5
25.5
77.8
6.1
5.8
3.3
3.8
17.4
2.2
4.0
3.7
8.2
32.6
19.6
7.3
11.1
2.3
1.8
2.5
5.1
13.0
50.3
.67
.33
.40
.51
.87
.21
.29
.36
.40
_
.53
1.01
.48
.27
.42
.38
.51
.65
Table 9 presents a similar data base from industrial origin (29). The
fouling rates were calculated from the measurements made assuming a linear
increase with time. Since the true rate is not usually constant, the values
presented are not directly comparable, but are presented to give general
orders of magnitude.
** '
It should be noted that foulants at plants A, M, N, H1, and J1 were
known to contain higher than uaua^ fractions of inorganic constituents as
well as higher ADWP/ABRV ratios. $
**.
588
-------
TABLE 9. REPRESENTATIVE CERAMIC FOULING DATA - INDUSTRIAL (29)
PLANT DIFFUSER
TIME OF EXPOSURE aBRV iDWP ADWP/ABRV INDUSTRY
(Days) (1n/yr) (1n/yr) TYPE
AI
BI
CI
CI
DI
01
El
GI
GI
HI
HI
JI
JI
KI
KI
KI
LI
Dome
Disc
Disc
Disc
Disc
Disc
Disc
DISC
Disc
Disc
Disc
Disc
Disc
Plate
Plate
Plate
Dome
720
120
16
92
218
110
21
34
31
420
420
90
90
30
77
58
110
141
59.3
962
341
35.3
83.4
393
280
73.0
39.1
53.0
219
270
1470
208
30
83.4
22.6
18.6
132
56.5
5.3
18.3
128
75.2
24.1
32.0
37.4
186
194
-
17.2
-
18.3
.16
.31
.14
.17
.16
.22
.33
.27
.33
.82
.71
.85
.72
.
.10
.
.22
Pulp/Paper
Pulp/Paper
Pulp/Paper
Pulp/Paper
Pulp/Paper
Pulp/Paper
Mun1c/Ind.
Pharm.
Pharm.
Food
Food
Brewery
Brewery
Hunlc/lnd.
Munlc/Ind.
Mun1c/Ind.
Mun1c/Ind.
Based on the preliminary work done to date, the following tentative
observations may be proposed:
* Municipal plant fouling rates appear to be variable from plant to
plant
* Fouling rates are also widely variable both spatially and
temporally.
* Industrial plant fouling rates were higher and somewhat more
variable than municipal plants.
* No significant or persistent differences in fouling rates were
observed to be attributed to the shape or composition of the ceramic
diffusers tested.
* There are possibly differences in fouling between the plastic and
glass bonded ceramics since their affinity for water (and contact
angle) differ as well as the relationship between bubble release
vacuum and pore diameter.
* Current work tends to support earlier belief that the major factors
contributing to liquid side fouling are - high organic load; low
local flux rate; relatively small pore diameter.
* Airside fouling has not been found to be a significant factor in
fouling in the 50 plants studied to date.
* Visual appearance of foulant has failed to consistently provide
reliable basis for identifying the nature or origin of fouling.
589
-------
• In plug flow systems fouling rates are usually greatest at the
influent end.
* No comparative data is available on fouling rates between plug flow
and completely mixed systems.
* Either inorganic or biological fouling can produce reductions of OTE
by 50 percent in a matter of weeks under extreme conditions of
fouling. In other situations little reduction in OTE may be noted
over an extended period of time.
* Silica has frequently been found as a major constituent of diffuser
foulents.
* Ferrous sulfate (and, likely other metallic salts), added to
aeration tanks for process reasons may aggravate fouling.
* Most diffusers that have been tested can be restored to
substantially original conditions by selected in to situ cleaning
methods including hosing, steam cleaning, gas cleaning, and hose to
acid cleaning.
* Little data is available yet on flexible membrane fouling.
SYSTEM MAINTENANCE
An important element in system maintenance is process monitoring. Air
side and liquid side fouling of the type favored by high air flux rate cause
an increase in the headless through the diffuser at constant airflow rate,
and such increases in headless may be detected by operating conditions in
the air supply system. Depending on the specific design approach, increases
in the pressure in the air supply system (monitored, for example, in the
blower discharge header or by increased opening of the flow control valves)
indicates an increase in diffuser headloss. These factors, along with the
airflow rate, should be monitored on a daily basis.
While overall system monitoring provides an indication of extreme
fouling, it does not provide a very sensitive indication of increased
headloss nor does it necessarily indicate significant fouling of the type
inversely effected by air flux rate. For example, a 10 percent increase in
system headloss (an apparent minor increase in total system pressure) may
represent a significant increase in diffuser headloss. And more
importantly, can indicate fouling that may have had a significant adverse
effect on OTE. Moreover, fouling of only a portion of the diffusion system
may lead to a significant redistribution in airflow but little increase in
system pressure. Consequently, use of a more sensitive technique may be
desirable or necessary. This is provided by measuring the dynamic wet
pressure (DWP)(9,10).
Individual diffusers are outfitted with a series of manometers that
allow measurement of the headloss across the air distribution control
590
-------
orifice and across the diffuser. The headloss across the orifice allows
determination of the airflow through the diffuser, while headloss across the
diffuser indicates the degree of fouling. By outfitting individual
diffusers throughout the system, the conditions of various portions of the
diffusion system can be monitored.
Since transfer efficiency can be significantly reduced by fouling
without attendant significant increases in back pressure, effective
monitoring will include other parameters in addition to headloss including
OTE and bubble release vacuum (9). Savings in power obtainable by
optimizing cleaning schedules are believed to frequently justify the modest
equipment and labor required for such monitoring. A number of other
candidate parameters for monitoring efficiency exists. One such parameter
is the specific airflow, which is the volume of air supplied per unit of
pollutant removed. Airflow is measured in standard cubic meters (standard
cubic feet), while pollutant loadings are measured in terms of the BOD
removed plus the ammonia transformed to nitrate-nitrogen in the plant
effluent. The ratio (m^Air/kg pollutant transformed ) should be monitored,
along with the aeration basin D.O.. Either an increase in the specific
airflow, a decrease in D.O., or both indicate a decrease in OTE. OTE can
also be measured directly using a variety of techniques previously described
earlier. One of the easiest is the off-gas technique (26).
These quantitative measures of system performance should also be
coupled with visual observations of the system. The surface pattern on the
aeration basin should be smooth with no air "boils". These arise because of
breaks in the air supply piping that allow large quantities of air to exit
the system. Such leaks should be repaired as quickly as possible, both
because of the decrease in OTE due to the release of coarse bubbles and
because of the possibility of further damage to the diffusion system.
The uniformity of the surface pattern may indicate plugging of a
portion of the diffusion system. An unusually low degree of surface
turbulence in a portion of the aeration basin may indicate that the
diffusers are partially fouled, thus restricting airflow to that portion of
the basin. Cleaning of the diffusers in that portion of the basin may be
required.
The size of the air bubbles exiting the aeration basin also provide an
indication of fouling, particularly loosely adherent biomass that may cause
the formation of large bubbles. Typically some degree of "coarse bubbling"
is observed at the inlet end of an aeration basin, generally thought to
occur because of large bubble formation as a result of high local flux rates
caused by surfactants contained in the influent wastewater. These materials
are quickly absorbed and/or degraded by the activated sludge, however, which
restricts the size of the "coarse bubble" zone. On the other hand, if
biological fouling occurs, the coarse bubble zone can expand until, in the
worst cases, it covers the entire surface of the aeration basin. It is
desirable that the surface of the aeration basin be inspected when initially
placed in service so as to become familiar with the size and appearance of
the fine and coarse bubbles exiting the inlet and outlet portions of the
591
-------
aeration basin. This familiarity will provide a basis for recognizing
coarse bubbling, should it occur later.
A major finding of the studies of dome/disc plants in England and
Holland by Houck and Boon in 1981 (12) was that the exellent O&M performance
of these grid systems was due to both the knowledge and diligent care of
treatment plant operators. Routine draining, tank and grid washdown, and
hardware checking was standard operating procedure at all plants surveyed.
Operators were also aware of symptoms of problems in the diffuser system and
were quick to respond.
Preventative maintenance is necessary to keep the system in proper
working order and at an optimum level of performance. Proper equipment
maintenance and system operation are necessary to maintain equipment
efficiency and to reduce the rate of diffuser fouling. It should also
eliminate the need for emergency maintenance resulting from system
failure. Certain types of fouling will occur, however, requiring periodic
diffuser cleaning.
Proper maintenance of the air filtration and supply system can signifi-
cantly reduce airside fouling of the diffusers. Proper operation and main-
tenance will generally exclude atmospheric dust sufficiently to eliminate
concerns over airside plugging from this source. The guidance provided by
the equipment manufacturer is generally sufficient in this area. Proper
maintenance will also reduce interruptions in air supply that can lead to
the entry of solids into the system, as discussed above. The deposition of
solids on the liquid side of the diffuser and penetration into the upper
pores is also reduced.
Proper system operation can also minimize the rate of liquid side
diffuser fouling. Minimum airflow rates per diffuser must be maintained to
prevent the deposition of solids that can later penetrate and plug the
surface pores of the diffuser (12).
Experience indicates that these approaches will be successful in
reducing the rate of liquid side diffuser fouling. However, fouling will
still occur (although at a lower rate), and the diffusers must be cleaned
periodically. Diffuser cleaning may be accomplished according to a regular
preventative maintenance schedule that balances the cost, of diffuser
cleaning against the power cost savings resulting from higher system OTE
(resulting in lower airflow requirements) and lower system pressure.
Figure 8 illustrates the concept of the optimum OTE. System power
costs decrease with higher OTE due to lower system air requirements. On the
other hand, the cleaning costs required to maintain a certain average OTE
increase as the target OTE increases. This results because an increased
cleaning frequency and, perhaps, a change in the cleaning method will be
required to maintain a higher OTE. The optimum OTE is the one that
minimizes the sum of the power cost and the cleaning costs required to
maintain a higher OTE. The optimum OTE is the one that minimizes the sum of
the power cost and the cleaning costs required to maintain the OTE, thus
592
-------
Cleaning Cost to Maintain
Specified OTE
OTE
Figure 8. Idealized plot of optimum OTE to balance
power and diffuser cleaning costs.
producing the lowest overall operating cost.
applied to system pressure.
This same concept can be
It should be recognized that Figure 8 is an idealized plot. It
presumes, among other things, that the fouling rate and its effects remain
constant with time and that the relationship of cleaning versus OTE does not
progressively change. It would therefore be most effective to regularize
monitoring programs and develop cleaning intervals around set point changes
in OTE or DWP.
593
-------
CLEANING TECHNIQUES
A variety of diffuser cleaning techniques are currently available, and
they can be broadly classified as process interruptive and nonprocess
interrupt!ve. Process interruptive cleaning techniques require that the
aeration basin be taken out of service to provide access to the diffusers,
while nonprocess interruptive techniques do not require such access. A
further distinction is between techniques that do not require removal of the
diffusers from the basin (in-situ) and those that do require diffuser
removal (ex-situ). All ex-situ techniques are process interruptive, while
only some in-situ techniques are process interruptive.
Among the important in-situ cleaning methods in use today are hosing,
steam cleaning, and in-situ acid cleaning. Hosing with either high pressure
or low pressure sprays and steam cleaning will effectively dislodge loosely
adherent liquid side biological growths. The application of ^1^% HC1 (18
Baume1 inhibited muriatic acid, 50% by volume) with a portable spray
applicator to each ceramic diffuser following hosing or steam cleaning and,
then, rehosing the spent acid is effective in removing both organic and
inorganic foulants (29,30).
Gas cleaning consists of the injection of an aggressive gas (HC1 or
formic acid) into the air feed to the fouled diffusers. The oxidizing agent
is transported to the diffuser by the airflow where it can oxidize most
types of foulants. The exception is atmospheric dust deposited on the
airside of the diffuser which has not been found to be a significant source
of fouling as previously reported.
Refiring is the most expensive technique and applies only to ceramic
diffuser elements. It involves removal of the diffuser from the aeration
basin, placing them in a kiln, and heating them in the same fashion
originally used to manufacture the element. The result is an element with
most foulants removed (or incorporated into the diffuser element) and
essentially restored to its original condition.
Currently, the effectiveness of the various cleaning methods being used
today for the variety of different foulants encountered on different fine
pore media is not well documented. Furthermore, the costs for these methods
are not generally available. Current research by EPA/ASCE will develop a
sound data base on cleaning technology.
Several methods are available to measure the effects of diffuser
cleaning on the characteristics of the diffuser. One approach is to apply
the process monitoring procedures discussed above. Thus the effects are
measured as a decrease in system pressure or diffuser DWP, or by an increase
in OTE or decrease in specific airflow (i.e., airflow per unit of pollutant
removal). Techniques can also be applied to directly measure the
characteristics of individual diffusers. These include OTE (26), chemical
analysis of foulants, and measures of airflow capacity of individual
diffusers. These latter techniques include specific permeability and bubble
release vacuum (BRV), all of which either measure the airflow at a specified
594
-------
applied diffuser headloss or the applied headless required to induce airflow
through the diffuser (9,38).
As discussed above, the effectiveness and costs of the various diffuser
cleaning techniques is an area of active research and the development of
detailed information in this area should be forthcoming. However, there is
little doubt that the incorporation of an effective cleaning schedule is a
necessary component of any fine pore diffused aeration preventative
maintenance program.
RETROFIT CONSIDERATIONS
DESIGN
The principal advantage of placing fine pore diffusers in municipal
wastewater treatment facilities is to reduce the airflow required to provide
the oxygen necessary for effective activated sludge treatment. This
reduction in required airflow can result in significant energy savings in
operation if proper attention is given to all the system components. Energy
costs are escalating rapidly. Estimates of electrical energy cost increases
of 25 to 35 percent in excess of inflation by the year 1989 have been made
(39). Other reasons for fine bubble retrofit include:
• Replacement of existing equipment which has reached the end of its
useful life
• Increased treatment capacity required by increased influent flow
and/or organic load
• Increased level of treatment required by more stringent NPDES Permit
limits
Enhanced oxygen transfer capability in itself cannot change plant
treatment capacity. Aeration tank volumes and consequent hydraulic
detention time must be great enough to support increases in flow and/or
nitrification. If air supply capacity is a limiting factor, however, fine
bubble retrofit can affect plant treatment capacity.
Wastewater characteristics will effect the design of retrofit fine pore
systems. In addition to oxygen demand (carbonaceous plus nitrogeneous),
some wastewater constituents may facilitate diffuser fouling. Furthermore,
as described earlier, surfactants will play a dramatic role in process
oxygen transfer.
The operating characteristics of fine pore diffused air systems are
different than those of other oxygen transfer devices, and these differences
affect process design. While diffused air systems can produce strong
vertical mixing components, horizontal components will generally either be
unidirectional (in cases where the diffused aeration equipment is located
along only one portion of the basin as in spiral roll) or largely
595
-------
nonexistent (in the full floor coverage example). Consequently, the flow
pattern is likely to be plug flow in character above certain length to width
ratios, resulting in a gradient in process oxygen demands from the aeration
basin influent to effluent.
The airflow range over which a particular fine pore diffuser can
effectively operate must be identified. In general, the lower limit is set
by the flux rate required to maintain uniform flow across the diffuser.
Operation below this value may result in accelerated fouling. The lower
gassing rate must also be high enough to provide adequate suspension of the
mixed liquor. The upper limit corresponds to the airflow rate beyond which
a significant decrease in the oxygen transfer efficiency is observed.
Taken together, these factors generally mean that fine pore diffused
aeration systems frequently are designed with tapered aeration capabilities
in tanks with high length-to-width (aspect) ratios. At a minimum, the
diffuser density (i.e., effective area of per unit aeration basin floor
area) should be varied, with the highest density near the tank inlet and the
lowest at the tank outlet. Such variations should be designed to meet
expected variations in air requirements, considering both variations in
process oxygen requirements and alpha factors along the length of the
aeration basin. It may also be desirable to section the diffusion system
into grids, with independent air supply control to each grid. A total of
three grids might typically be provided in an aeration basin with a length
to to to width ratio of 3 to 1 or greater.
Failure to provide proper tapering in tanks with high aspect ratio or
in staged tanks can result in inadequate oxygen transfer capacity at the
inlet end of the aeration basin, resulting in low dissolved oxygen con to
centrations. Such conditions have been found to result in accelerated
biofouling of fine pore diffusers (discussed in more detail below) and may
also lead to other process and/or operational problems (12). Overdesign (in
terms of the number of diffusers provided) can lead to system inefficiency
if the airflow rate to meet the minimum requirement per diffuser exceeds
that to meet process oxygen requirements.
Most air supply blowers in municipal treatment plants are single or
multistage centrifugal types, or rotary positive displacement units. The
energy savings available with fine bubble diffusers result directly from a
reduction in the air required to provide the process with necessary
oxygen. This reduction in airflow will result in operating fewer blowers
and/or operating the same blowers at different points on their performance
curves.
The efficiency of both single and multistage centrifugal blowers can
vary from more than 70% to less than HQ%, depending on the blower itself,
and the operating combination of discharge volume and discharge pressure.
Estimating input horsepower for these units should always be done using the
actual certified performance curves of the blowers or estimated performance
curves supplied by the blower manufacturer.
596
-------
Potential power savings resulting from reduced airflows with fine pore
diffusers can be completely negated by a change in blower operating
efficiency as a result of reduced airflow. It is, therefore, of primary
importance to accurately estimate blower horsepower under the actual
conditions that will be required to operate and not by using compression
formulae which require an estimate of blower efficiency to determine power
for a given discharge condition.
Centrifugal blower capacity should be regulated to the extent possible
by throttling on the inlet side, as significant power savings are available
at any duty point. Rotary positive displacement blower capacity cannot be
varied by throttling for all practical purposes. Airflow to aeration can be
changed only by operating more or less units, or by "blowing off" some of
the air to atmosphere. Wasting air to atmosphere may reduce the actual
airflow to aeration, but it will not reduce the power consumption of the
blowers because the wasted air must first be compressed to the system
pressure before it is discharged. Here again, it is crucial that blower
horsepower be estimated using actual blower performance curves and under the
anticipated actual conditions of operation because efficiency for rotary
positive blowers is by no means constant from one unit to another.
The existing system of air distribution piping can in general be reused
with some reservations. Because airflow rates will be reduced as a result
of enhanced oxygen transfer efficiency, the size of the existing blower
discharge headers and air mains which deliver air to the tanks will normally
be sufficient. Depending on the type and arrangement of fine pore
equipment, the individual drop pipes into the tanks may also be large
enough. The air distribution system should be checked for adequacy,
however, as part of the design of the project.
Because the small air passage orifices in fine pore diffusers can be
easily clogged by inlet air particulates, air filtration is an important
prerequisite (5,7). Air filters can be located on the inlet to the blowers
or in-line in the air distribution system. Blower inlet filters will
effectively remove contaminants from the outside air but will not protect
the diffusers from dirt, rust, scale, or other debris which might already be
in the downstream piping. It is recommended, therefore, that careful
consideration be given to the use of in-line filters. In some cases, it may
even be desirable to locate filters adjacent to the air drops into the
aeration system so that only clean, new, corrosion-resistant pipe need be
located between the filters and the diffusers. Existing piping systems
composed of galvanized steel or stainless steel pipe may present little
danger of present or future rust or scale particles plugging the diffusers
if blower inlet filters are selected. Painted or uncoated steel or iron
pipe should be used with extreme caution unless in-line filters are
installed downstream.
One drawback of in-line filters is the incremental increase in blower
discharge pressure required to overcome losses in the filters. This
consumes some power, but the effects can be minimized by properly sizing and
maintaining the filters. A good rule of thumb to minimize air filter
597
-------
pressure drop is to limit face velocities to 91 to 107 m/min (300 to 350
ft/min) through the filter at peak airflow rates.
ECONOMIC ANALYSIS
The factors used in the design of a fine pore diffuser system also
comprise the basis for determining its economic viability. Installation of
fine pore equipment should be undertaken only if a reasonable return on
investment can be foreseen. The cost effectiveness of fine pore retrofit is
most appropriately based on present day flow and loading to the plant rather
than anticipated future increases and should consider the total present
worth of the investment as well as "simple payback."
Simple payback indicates payback in years but neglects the effects of
inflation and the time value of money. A more realistic approach to
economic evaluation is based on the present worth of the the cost savings of
the retrofit project at some time in the future. This type of analysis
accounts for the fact that energy costs and maintenance costs will increase
due to inflation and that the money invested in a fine pore retrofit project
would provide a return if used for other types of investments, as well.
A more sophisticated determintion which would allow energy costs to
increase at a faster rate than the general inflation rate can also be
performed. An estimate of the energy cost inflation rate should be made in
conjunction with the local power utility as this phenomenon could vary
widely. As the payback period becomes longer, these analyses become very
sensitive to the assumptions made for inflation and discount rates. If the
apparent payback period exceeds five or six years, the assumptions used
should be reviewed. If they are found to be reasonable and prudent, the
economic viability of the retrofit should be scrutinized carefully.
ONGOING STUDIES
A great deal of progress has been made in the last five years to better
delineate the design, testing, maintenance, and control of oxygen transfer
equipment. Yet, it is clear from the discussion in this report that there
are still many gaps in our knowledge of fine pore aeration systems to
their performance in process water and how to translate clean water data to
field conditions; the behavior of fine pore diffusers with respect to liquid
side fouling; the strategies needed to maintain and control these systems in
order to secure highest possible efficiency from the device.
There are a number of research programs now under way in the U.S.,
Canada, and the U.K. dealing with the design, performance, operation,
maintenance, installation, control, and costs of fine bubble aeration
systems.
The U.S. EPA has recently funded a cooperative research agreement with
ASCE and the Subcommittee on Oxygen Transfer to evaluate the existing date
base on the performance of fine pore diffused aeration in clean and process
598
-------
water and to carry out field studies at municipal treatment facilities as
required to fill these perceived gaps. Diffuser cleaning studies will also
be carried out at a number of installations in order to delineate factors
affecting the selection of a particular method and to establish economic
bases for that selection. Two manuals will also be prepared. The first, an
interim guidelines report on the state to of to art of fine pore diffusion,
from which this paper is derived and the second, a comprehensive design
manual as fine pore aeration to be published in 198?.
599
-------
REFERENCES
1. Wesner, G. M. et al. Energy conservation in municipal wastewater
treatment. EPA-430/9-77-011, Office of Water Programs Operations, U.S.
EPA, March 1978.
2. Barnhart, Edwin L. An overview of oxygen transfer systems. In;
Proceedings of Seminar Workshop on Aeration System Design, Testing,
Operation, and Control. EPA-600/9-85-005, January 1985.
3. Roe, F. C. The installation and servicing of air diffuser mediums.
Water and Sewage Works. 81: 115, 1934.
4. Committee on sewage disposal. The operation and control of activated
sludge sewage treatment works. Sewage Works Journal. 14: 3, 1942.
5. Committee on Sewage and Industrial Waste Practice to Subcommittee on
Air Diffusion. Air diffusion in sewage works. Federation of Sewage
and Industrial Wastes Association, 1952.
6. Bushee, R. J. and Zack, S. I. Test on pressure loss in activated
sludge plants. Engineering News Record. 93: 21, 1924.
7. Subcommittee on Aeration in Wastewater Treatment. Aeration on
wastewater treatment - MOP-5. Water Pollution Control Federation,
1971.
8. McCarthy, J. J. Technology assessment of fine bubble aerators. EPA-
600/2-82-003. Municipal Environmental Research Laboratory, U.S. EPA,
February 1982.
9. Redmon, D. T. Operation and maintenance/trouble to shooting. In;
Proceedings of the Seminar Workshop on Aeration System Design, Testing,
Operation, and Control. EPA-600/9-85-005. Municipal Research
Laboratory, U.S. EPA, January 1985.
10. Boyle, W. C. and Redmon, D. T. Biological fouling of fine bubble
diffusers: State-of-art. ASCE Journal of Environmental Engineering.
109, No. 5, October 1983.
11. Personal communication with Parkson Inc. Fort Lauderdale, Florida.
12. Houck, D. H. and Boon, A. G. Survey and evaluation of fine bubble dome
diffuser aeration equipment. EPA-600/2-81-222. Municipal
Environmental Research Laboratory, U.S. EPA, September 1981.
600
-------
27. Broohtrup, J. A. A study of the steady-state and off-gas methods of
determining oxygen transfer efficiency in mixed liquor. Masters
Thesis, Department of Civil and Environmental Engineering, University
of Wisconsin-Madison, 1983.
28. Yunt, F. W. et al. Aeration equipment - phase II Whittier-Narrows
water reclamation plant. Los Angeles County Sanitation District -
Operations and Off-Gas Test Results. U.S. EPA (to be published).
29. Ewing Engineering Company. Milwaukee, Wisconsin. Personal
Communications, 1985.
30. Houck, D. H. Survey and evaluation of fine bubble dome and disc
aeration systems in North America. U.S. Environmental Protection
Agency, (in press).
31. Vik, T. A., Lamers, D. J., and Roder, D. L. Full-scale operating
experience utilizing fine bubble ceramic aeration with in place gas
cleaning at Seymour, Wisconsin. Presented 57th Annual Conference,
WPCF, October 1984.
32. Gilbert, R. G. Measurement of alpha and beta factors. In; Workshop,
toward an oxygen transfer standard. EPA-600/9-78-021, 1979. p. 179.
33. Lister, A. R. and Boon, A. G. Aeration in deep tanks: An evaluation
of a fine bubble diffused-air system. , J. Institute of Water Poll.
Control. 72: 590, 1975.
34. von der Emde, W. Aeration developments in Europe. In; Advances in
water quality improvement. Univ. of Texas Press, 1968. p. 237
35. Chambers, B., Robertson, P., and Thomas, V. K. Energy saving -
optimization of fine bubble aeration, final report and replicators
guide - May 1984. 300-S, Water Research Centre, Stevenage Laboratory,
U.K.
36. Allbough, T., Benoit D. J., and Spangler, J. Aeration System Design
Using Off-Gas Oxygen Transfer Testing. (To be presented Annual WPCF
Conference, Kansas City, 1986.)
37. Alkema, T. Slime growth on ceramic fine bubble dome diffusers. In;
Proceedings High Efficiency Aeration in Wastewater Treatment. The
Canadian Institute of Energy Seminar, May 1983.
38. Boyle, W. C. et al. Investigations of biological fouling of ceramic
fine bubble diffusers. U.S. EPA, WERL (report in progress).
39. Consulting Engineer. Vol. 64, No. 4, April 1985. p. 10.
601
-------
13. Wren, J. D. Diffused aeration-types and applications. In;
Proceedings of the Seminar Workshop on Aeration System Design, Testing,
Operation, and Control. EPA-600/9-85-005. Water Engineering Research
Lab, U.S. EPA, January 1985.
14. Huibregtse, G. L., Rooney, T. C. and Rasmussen, D. C. Factors
affecting fine bubble diffused aeration performance. JWPCF. Vol. 55,
No. 8, August 1983.
15. Proceedings of the Workshop Toward an Oxygen Transfer Standard. EPA-
600/9-78-021. April 1979.
16. ASCE Oxygen Transfer Standards Subcommittee. Measurement of oxygen
transfer in clean water. Standard ASCE, New York, July 1984.
17. ASCE Oxygen Transfer Standards Subcommittee. Development of standard
procedures for evaluating oxygen transfer devices. EPA-600/2-83-102.
October 1983.
18. Yunt, F., Hancuff, T., Brenner, R., and Shell, G. An Evaluation of
submerged aeration equipment - clean water test results. A summary of
photographic slides, WWEMA Industrial Pollution Conference (Houston),
June 1980.
19. Sullivan, R. C. and Gilbert, R. G. The significance of oxygen transfer
variables in the sizing of dome diffuser aeration equipment. Paper at
WPCF Conference (Atlanta), October 1983.
20. Paulson, W. L. Oxygen absorption efficiency study - Norton Co. dome
diffusers. Norton Company Report, March 1976.
21. Popel, J. H. Improvements of air diffusion systems applied in the
Netherlands. In; Proceedings Seminar Workshop on Aeration System
Design, Testing, Operation and Control. EPA-600/9-85-005, January
1985. pp. 156-176.
22. Paulson, W. L. and Johnson, J. K. Oxygen transfer study of FMC
pearlcomb diffusers. FMC Corporation Report, August 1982.
23. Wyss Inc. Wyss flex-a-tube Diffuser Systems. Report (Results by
Shell, G. and Paulson, W. L.), October 1983.
24. Paulson, W. L. Selected communications from various transfer
efficiency studies. Personal Communication, 1985.
25. Boyle, W. C. Oxygen transfer under process conditions. (To be
published by U.S. EPA.)
26. Redmon, D. T., Boyle, W. C., and Ewing, L. Oxygen transfer efficiency
measurements in mixed liquor using off-gas techniques. Journ. Water
Pollution Control Federation, Vol. 55, No. 11, p. 1338, 1983.
602
-------
COMPOSTING PRACTICE IN THE UNITED STATES TODAY
by
Arthur H. Benedict
Brown and Caldwell
Pleasant Hill, California 94523
This paper has been reviewed in ac-
cordance with the U.S. Environmental
Protection Agency's peer and adminis-
trative review policies and approved
for presentation and publication.
Prepared for Presentation at:
Tenth United States/Japan Conference
on Sewage Treatment Technology
October 17-18, 1985
Cincinnati, Ohio
60:
-------
COMPOSTING PRACTICE IN THE UNITED STATES TODAY
by: Arthur H. Benedict
Brown and Caldwell
Pleasant Hill, California 94523
ABSTRACT
Current municipal sludge composting practice in the United States is
described based on investigations at selected operating facilities. An
inventory of over 40 municipal sludge composting facilities is presented
and a breakdown of these facilities by size, years of operation, and
dewatered sludge total solids content processed is given.
Operations at three aerated static pile facilities and one conven-
tional windrow facility are described. One aerated static pile facility
treats anaerobically digested sludge at a nominal loading of approximately
45 wet metric tons (50 wet tons) per day (Mg/d), while the other two
process limed and unlimed raw sludge at nominal loadings of 360 and 180 wet
Mg/d (400 and 200 wet tons per day (wtpd)), respectively. The conventional
windrow facility treats 450 wet Mg/d (500 wtpd) of anaerobically digested
sludge. Site features, sludge characteristics and loading variations,
mixing procedures, active composting requirements, and procedures for
drying, curing, and screening are reviewed, and successful odor control
techniques are described. Finished compost quality and distribution is
discussed, and operating costs are presented.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved
for presentation and publication.
604
-------
INTRODUCTION
Interest in composting as a means of municipal sludge treatment in the
United States began in the early 1970s. At that time, the Los Angeles
County Sanitation Districts initiated windrow composting of sewage sludge
at the Joint Water Pollution Control Plant in Carson, California, and
the U.S. Department of Agriculture developed large-scale studies of static
pile composting at the Agricultural Research Station in Beltsville,
Maryland. Since that time, interest and activity in municipal sludge
composting in the United States has increased dramatically.
This paper presents the results of an investigation of municipal
sludge composting as currently practiced in the United States. The study
was initiated in the spring of 1984 and was completed in September 1985.
STUDY OBJECTIVES
Objectives of the municipal sludge composting study were as follows:
1. To investigate aerated static pile and windrow composting
technologies based on operating experiences at full-scale
facilities;
2. To compare and contrast features of the aerated static pile,
aerated windrow, and conventional windrow processes based on
this experience;
3. To evaluate performance relative to design and operation,
including cost;
4. To identify key problems associated with municipal sludge
composting using these technologies; and
5. To define methods which have been used or are being considered to
resolve these problems.
The investigation focused on three composting processes: the aerated
static pile process, including the extended static pile technique; the
conventional windrow process; and the aerated windrow process. In-vessel,
mechanical processes were not considered in the evaluation.
FACILITIES INVENTORY
The municipal sludge composting study was initiated with an inventory
of operating facilities in the United States. A summary of facilities
inventoried is presented in Table 1.
605
-------
premixing using a front-end loader (20 to 25 minutes) followed by fine
mixing (15 to 20 minutes) with a manure spreader or rototiller. Fine
mixing at another site is performed with three passes of a mobile composter.
If only front-end loaders are used for mixing, careful monitoring of the mix
must be performed. Observations made during the on-site investigations
revealed that for the quantities of material mixed at the Columbus facility,
40 minutes was the minimum time necessary for effective mixing with
front-end loaders alone.
Various mixtures at the Hampton Roads facility were independently
tested for pore space and homogeneity as part of the on-site investigations
at this location. The rototilling method provided the most uniform mixture,
as presented below:
Pore space, Unmixed clumps,
Mixing method percent percent
Front-end loader only 44 5 to 20
Front-end loader/manure spreader 60 10
Front-end loader/rototiller 62 5
ACTIVE STATIC PILE COMPOSTING
Effective static pile composting requires proper pile construction
techniques, sufficient aeration, and adequate process monitoring and
control during the 21-day active composting period. Together with a
homogeneous, sludge-wood chip mixture at a moisture content less than
60 percent, these factors combine to ensure uniform pile aeration, prevent
short-circuiting, and minimize heat loss for efficient stabilization and
pathogen inactivation.
Static Pile Configurations
Static pile aeration system facilities and pile construction
techniques will vary depending on site-specific conditions. However,
generally one of two basic schemes is employed. One configuration involves
placing perforated aeration piping of an appropriate length directly on a
paved compost pad, after which the piping is covered with a base material,
such as wood chips, 30 to 45 centimeters (12 to 18 inches (in.)) deep.
Sludge-wood chip mixture is then placed on the base material in an extended
pile configuration (see Figure 3), after which insulating cover material is
applied. The aeration piping is connected by a manifold to a blower which
provides either positive or negative aeration. Generally, one such extended
pile compartment is constructed for each daily loading of dewatered sludges.
The second basic static pile configuration is similar to the first
except that perforated aeration piping is placed in belowground troughs
which are formed as part of the paved compost pad. The troughs are then
filled with a material such as wood chips, after which base material,
sludge-wood chip mixture and cover material are placed in a manner similar
to the first configuration. The aeration piping is again connected to a
blower via a manifold.
606
-------
STUDY METHODS
Schematics for the aerated static pile and conventional windrow
composting processes are presented on Figures 1 and 2. The aerated static
pile process involves mixing dewatered sludge with a bulking agent, such as
wood chips, followed by active composting in specially constructed piles
such as shown on Figure 3. Typically, both recycled bulking agent and new
(external) bulking agent are used for mixing. Induced aeration, either
positive (blowing) or negative (suction), is provided during the active
composting phase. Temperature and oxygen are monitored during active
composting as a means of process control. The active composting period
lasts 21 days, after which alternate pathways to produce finished compost
may be employed.
RECYCLED
WOOD CHIPS
-
r
ACTIVE
COMPOSTING
t t
- INDUCED AERATION
T*
1
CURING
1
-»
DRYING
I
SCRE
' n
.NING — »• FQR OISTRIBUTION
T
COVER BASE I __.
MATERIAL MATERIAL r~ ~~ — — *•
I
I ALTERNATE MATERIALS MANAGEMENT STEPS
FIGURE 1. PROCESS SCHEMATIC FOR AERATIC STATIC PILE COMPOSTING
If at the end of the 21-day active composting period, composted
material is sufficiently dry, screening may be performed directly to
separate bulking agent for recycle. The recycled bulking agent is
generally stored prior to reuse in the mixing operation. Screened compost
is restacked and cured for at least 30 days and then stockpiled as finished
compost prior to distribution.
If at the end of the 21-day active composting period, the compost
material is not sufficiently dry for screening, a separate-stage drying
step is required prior to screening, curing, and stockpiling. Alterna-
tively, unscreened compost may be restacked for the 30-day curing
period, after which separate-stage drying, screening, and stockpiling
is performed.
607
-------
809
OA"^voo
<&
-------
MIXING AND AERATION
BY TURNING
ACTIVE COMPOSTING
AND DRYING
FINISHED COMPOST
FOR DISTRIBUTION
FIGURE 2. PROCESS SCHEMATIC FOR CONVENTIONAL WINDROW COMPOSTING
The conventional windrow process (Figure 2) involves initial mixing of
dewatered sludge with a bulking agent such as finished compost, often
supplemented with an external amendment, followed by formation of long
windrows such as shown on Figure 3. Formation of the windrows is generally
performed with a specially designed, mobile composter which, in addition to
giving the windrow its triangular (or sometimes trapezoidal) cross section,
is capable of turning the windrow in place.
An active windrow composting period of 30 days (or more) is provided
following initial mixing and formation. During this period, the windrows
are periodically turned to aerate and remix the material. A turning
frequency of two to three times per week is typical. Temperature is
monitored for process control. Following the active windrow composting
period, the composted material is allowed to cure for 30 days; then, a
portion of the finished compost is recycled and a portion is stockpiled for
distribution.
KEY FACILITY FEATURES
Key features of the composting facilities investigated are presented
in Table 2. The Peninsula Composting Facility of the Hampton Roads Sanita-
tion District (Hampton Roads facility), located in Newport News, Virginia,
is a 45-wet-Mg/d (50-wet-ton-per-day (wtpd)) aerated static pile operation
processing a mixture of anaerobically digested primary and waste activated
sludge. Two facilities process raw sludge. The Washington Surburban
Sanitary Commission Site II Composting Facility (Site II facility) located
in Silver Spring, Maryland, and the Columbus, Ohio, Southwesterly Composting
Facility (Columbus facility) in Franklin County, Ohio. The Site II facility
treats limed, raw sludge at a nominal loading of 360 wet Mg/d (400 wtpd),
and the Columbus facility treats unlimed, raw sludge at a nominal rate of
180 wet Mg/d (200 wtpd). The only conventional windrow facility investi-
gated was the Joint Water Pollution Control Plant Composting Facility of the
Los Angeles County Sanitation Districts (Los Angeles facility), located in
Carson, California. This facility processes about 450 wet Mg/d (500 wtpd)
of anaerobically digested primary and secondary sludge.
609
-------
TABLE 1. INVENTORY OF OPERATING MUNICIPAL SLUDGE COMPOSTING
FACILITIES, SPRING 1984
Facility charactertistics
Facility composting process
Aerated
static pile
Conventional
windrow
Aerated
windrow
Total number
Note: Tons per day x 0.9072 - Mg/d.
28
12
Nominal size, dry tons per day
Less than 10
10 to 25
26 to 50
51 to 100
Greater than 100
Years of operation
Less than 1
1 to 5
6 to 10
Greater than 10
Sludge total solids, percent
Less than 15
15 to 20
Greater than 20
19
3
3
2
1
3
21
4
0
10
13
5
9
2
0
0
1
0
7
1
4
5
5
2
1
1
0
0
0
0
1
1
0
0
1
1
A total of 42 locations composting municipal sludge were identified
as of spring 1984. Twenty-eight utilize the aerated static pile process,
twelve use conventional windrow techniques, and two employ aerated
windrowing. Most of the facilities are under 9 dry metric tons per day
(Mg/d) (10 dry tons per day (dtpd)) and have been operating about 5 years
or less.
A variety of sludges are processed by the facilitiBS, including raw,
anaerobically digested, and aerobically digested sludges. Ten of the
aerated static pile facilities inventoried report representative dewatered
sludge total solids concentrations below 15 percent, and five of the
conventional windrow systems were in this category. Thirteen static
pile facilities report dewatered solids typically in the 15 to 20 percent
range, although note that daily values are sometimes under 15 percent.
Five of the conventional windrow and one of the aerated windrow facilities
report sludge solids concentrations in this category. Eight of the
municipal sludge composting facilities report dewatered sludge total
solids concentrations greater than 20 percent. One of these, an aerated
windrow facility, receives a heavily limed sludge having a total solids
concentration of about 40 percent.
The remainder of this paper describes representative United States
practice for aerated static pile and conventional windrow technologies
based on investigations at three static pile facilities and one conven-
tional windrow facility. The aerated windrow practice is not presented.
Information on this technology and on other features of the municipal
sludge composting investigation may be found in Reference 1.
610
-------
TABLE 3. KEY SITE FEATURES
Sice feature
Total operational area, ac
Mixing operations
Area, ac
Paved
Equipment
Bulking agent
Active composting
Area, ac
Paved
Covered
Aeration piping
Drying operations
Area, ac
Paved
Covered
Method
Screening operations
Area, ac
Paved
Covered
Equipment
Curing operations
Area, ac
Paved
Covered
Aerated
Storage operations
Bulking agent
Area, ac
Paved
Covered
Unscreened compost
Area, ac
Paved
Covered
Finished compost
Area, ac
Paved
Covered
Runoff collection
Ponds
Area, ac
Hampton Roads facility
6.2
0.3
Concrete
Yes
Front-end loaders uith manure
spreader or rototlller
Wood chips
1.1
Concrete
No
Laid in troughs in pad
Positive and negative
1.7
Concrete
No
Spread and mixed on
open slabs
0.2
Concrete
Yes
Mobile screens
-*
No
No
No
0.4
Concrete
No
1.6
No
No
0.4
Concrete
No
No
-
Site II facility
40
1.1
Concrete
Front-end loaders and
mobile composter
Wood chips
2.5
Concrete
Yes
Laid on pavement
Negative
0.8
Concrete
Yes
Induced aeration
1.0
Concrete
Fully enclosed
Fixed screens
3.3
Concrete
No
Yes
4.5
Asphalt
No
_
-
_
-**
No
No
Yes
10.0
Columbus facility
37
0.4
Asphalt
Partially enclosed
Front-end loaders only
Wood chips
3.7
Concrete
No
Laid on pavement
Positive
_
_
_
_
0.5
_
No
Mobile screens
4.6
Concrete
No
Yes
0.9
Concrete
No
10.0
Concrete
No
0.7
Asphalt
Yes, partially enclosed
Yes
0.9
•Included as part of unscreened compost storage.
••Included as part of curing area.
Note: ac x 0.40469 - ha.
SLUDGE LOADINGS AND CHARACTERISTICS
Representative sludge loadings and dewatered sludge characteristics
for the three aerated static pile facilities investigated are presented in
Table 4. Current loadings on an operating day basis vary from 57 to 327 wet
Mg/d (63 to 360 wtpd), which translate to 10 to 56 dry Mg/d (11 to 62 dtpd),
based on a mean dewatered sludge total solids content, as received, of
17 percent at each plant. The number of operating days per week vary from
5 to 7. Peak-to-average day loading ratios vary from 1.4 to 1.9. Although
the mean sludge total solids content at each facility is 17 percent,
variations as low as 13 percent are routinely received at one and as high
as 22 percent at another.
611
-------
TABLE 4. SLUDGE LOADINGS AND CHARACTERISTICS
Item Hampton Roads facility Site II facility Columbus facility
Type of sludge processed Anaeroblcally digested primary Raw, limed primary and Raw, unlimed primary and
and secondary solids secondary solids secondary solids
Loadings per operating day
Average, uet tons
Average, dry tons
Peak-to-average ratio
Operating days per week
Total solids, percent
Mean
Range
Total volatile solids, percent
Mean
Range
Density, Ib/cu yd
pH
6.3
11
1.6 to 1.8
6
17
14 to 22
55
43 to 61
1,600
8.2 to 8.3
360
62
1.4 to 1.6
5
17
15 to 22
43
34 to 51
-
12.0 to 12.5
170
29
1.9
7
17
13 to
74
63 to
1,900
-
22
79
Note: tons x 0.9072 - Mg
Ib/cu yd x 0.59325 - kg/m3
Volatile content of the anaerobically digested sludge processed
at the Hampton Roads facility is typically 55 percent, with a range of
43 to 61 percent. Raw, unlimed sludge received at the Columbus facility
is typically 74 percent volatile, with a range of 63 to 79 percent.
Liming of raw sludge to a pH of 12.0 to 12.5 prior to composting results
in volatile solids contents typically in the range of 34 to 51 percent
(Site II facility).
MIXING OPERATIONS
Wood chips are used as a bulking agent at all of the aerated static
pile facilities investigated. Both hardwood and softwood chips are
employed, though personnel at one facility prefer hardwood chips because
they do not decompose as rapidly as softwood chips and are cheaper to
purchase. Chips are generally stored in piles on paved surfaces without
any cover. Paved surfaces prevent wood chip losses, minimize moisture
control problems during inclement weather, and keep chips clean.
Dewatered sludge is mixed with recycled and new wood chips at
ratios of 3.5 to 4.5 (volume/volume) at the total solids concentrations
typically received at the composting facilities studied. The mix ratio
typically varies with chip quality, moisture content of sludge and chips,
season, and the proportion of fresh and recycled chips used. All facilities
employ mobile equipment for the mixing operation. A minimum initial total
solids concentration of 40 percent in the wood chip-sludge mixture is
considered essential for effective static pile composting.
For the quantities of material handled at the facilities studied,
about 40 to 45 minutes is required to achieve a well-mixed material for
active static pile composting. All facilities employ mobile equipment for
mixing as noted in Table 3. At one facility, mixing is accomplished by
612
-------
TABLE 2. KEY FACILITY FEATURES
Facility name and location
Composting
process
Nominal size,
wet tons per day
Sludge
characteristics
Hampton Roads Sanitation District Peninsula Static pile
Composting Facility, Newport News,
Virginia
Washington Suburban Sanitary Commission Static pile
Site II Composting Facility, Silver
Spring, Maryland
City of Columbus, Ohio Southwesterly Static pile
Composting Facility, Franklin County,
Ohio
Los Angeles County Sanitation Districts Conventional
Joint Water Pollution Control Plant windrow
Composting Facility, Carson, California
50
400
200
500
Anaerobically digested
primary and secondary
Limed, raw primary and
secondary
Unlimed, raw primary
and secondary
Anaerobically digested
primary and secondary
Note: Tons per day x 0.9072 - Mg/d.
AERATED STATIC PILE OPERATIONS
SITE FEATURES
Key site features for each of the aerated static pile facilities
investigated are summarized in Table 3. Total operational areas for the
Hampton Roads, Site II, and Columbus facilities are 2.5, 16, and 15 hectares
(ha) (6.2, 40, and 37 acres (ac)), respectively. Operational area is
defined as site area actually required for composting operations, including
access roads, buffer zones, and support facilities such as administrative
buildings, maintenance structures, and truck scales. Storage ponds for
runoff control, if required, are also included as part of the operational
area. Land which is available on the site but is not a part of active
operations, e.g., land available for expansion, is not included as part of
the operational area. Total operational areas employed at the Hampton
Roads, Site II and Columbus facilities are equivalent to 0.06, 0.04, and
0.08 ha per wet Mg/d (0.12, 0.10, and 0.18 ac per wtpd), respectively.
Areas employed for various operations vary depending primarily on
facility size, materials management procedures, and process requirements.
For example at the Columbus facility, a 180-wet Mg/d (200-wtpd) plant, 4 ha
(10 ac) are provided for unscreened compost storage because screening
operations are affected by weather, and the market for distribution of
finished compost is just being developed.
Key operations such as mixing are often covered to minimize moisture
control problems, and at two of the three facilities the screening operation
is also protected for this purpose. Most operating areas are paved either
with concrete or asphalt, in part for runoff collection and in part for
moisture control and equipment access during the composting process.
Equipment and/or operating methods employed at various steps vary, as noted
in Table 3.
613
-------
Although many facilities in the United States have been constructed
with a single blower to service more than one extended pile compartment,
the trend is now toward providing one blower of sufficient capacity for
each daily extended pile compartment constructed. One blower may be
provided to service the entire length of a pile compartment, or one may be
provided at each end to service one-half of each pile compartment.
Extensive testing by plant personnel was required at two of the
facilities investigated during this study to arrive at reliable pile
construction and aeration system techniques for routine operation. The
Site II facility constructs daily extended pile compartments which are
about 34 meters (m) (110 feet (ft)) long. Each pile is serviced by one
11-kilowatt (kw) (15-horsepower (hp)) blower located at one end of the
compartment which is operated in the suction aeration mode. An aeration
manifold consisting of four, 15-cm (6-in.)-diameter, flexible plastic hose
laterals placed at 1.5-m (5-ft) intervals on the paved compost pad extends
the length of each compartment.
Each aeration lateral consists of 4.6 m (15 ft) of solid pipe at the
blower end and about 26 m (86 ft) of perforated pipe. No piping is provided
the last 2.7 m (9 ft) under the toe farthest from the blowers. Perforations
are tapered as presented below:
Perforations,
Distance from sq cm/linear m
blower, m (ft) (sq in./linear ft)
0-4.6 (0-15) -0-
4.6-9.8 (15-32) 8.87 (0.415)
9.8-15.8 (32-52) 17.57 (0.830)
15.8-21.9 (52-72) 56.30 (2.660)
21.9-30.8 (72-101) 90.59 (4.280)
30.8-34.0 (101-110) No pipe
This perforation arrangement has been found to produce uniform aeration
during routine operations.
Base material at the Site II facility is new wood chips placed to
a depth of 30 cm (12 in.), except under the toe nearest the blower.
Unscreened compost base is used for a distance of 3.6 m (12 ft) at this
location to reduce short-circuiting. Each 34-m (110-ft) long by 6-m (20-ft)
wide by 3.6-m (12-ft) high pile compartment includes a daily 45-cm (18-in.)
blanket of screened or finished compost as an insulating cover.
At the Hampton Roads facility, a 73-m by 62-m (240-ft by 204-ft)
concrete composting pad is sectioned into quadrants. Each quadrent contains
12 troughs serviced by two 2.2-kw (3-hp), variable-speed blowers at each
end. Each trough contains 30 m (100 ft) of 15-cm (6-in.), Schedule 40,
polyvinyl chloride (PVC) aeration pipe connected to a manifold; one blower
services six troughs. The pipe is perforated, with 10-millimeter (mm)
(3/8-in.) holes spaced 3.8 cm (1 1/2 in.) apart, beginning 6.7 m (22 ft)
from the end of the compost pile. The end of the pipe is capped.
614
-------
The perforation configuration was determined from extensive testing
during normal operation and provides the most uniform oxygen levels and
temperatures throughout each pile. Airflow is controlled by a damper
located at each trough.
Pile construction at this location includes a wood chip base, 30 to
45 cm (12 to 18 in.) deep, except at the toes, and cover material of dry,
recycled (cured, screened) compost. Cover material depth is 45 cm (18 in.)
in the summer and 60 cm (24 in.) in the winter. Base material at the toes
is dry, recycled compost to prevent short-circuiting, and 10 to 15 cm
(4 to 6 in.) of screened compost is placed on the face of the pile for added
insulation and to prevent short-circuiting. Each pile is constructed to an
initial depth of about 4 m (13 ft) and widths of 2.5 to 3.0 m (8 to 10 ft).
Aeration and Process Control
The Site II facility which processes raw, limed sludge applies the
following typical aeration process control parameters for a 21-day active
composting period:
1. First week—negative aeration at 110 to 125 nrVh per dry metric
ton (3,500 to 4,000 cubic feet per hour per dry ton of sludge
(cfh/dt)) to promote material decomposition and for odor control.
2. Second and third weeks--aeration is controlled to maintain
temperatures between 50 and 60 degrees C, including a minimum
of 3 days above 55 degrees C to meet pathogen inactivation
requirements; then, aeration is increased to 187 m-^/h per dry
metric ton (6,000 cfh/dt) for drying and odor control at teardown.
Each blower used for aeration turns off for a 5-minute period every so
often to enable the aeration pipes to drain. All leachate and condensate
flows to drains leading directly to a sewer.
Temperature controllers are installed on each blower. The controllers
are set for the desired pile temperature, and the blowers automatically
turn on and run until the desired pile temperature has been reached. If
the pile temperature is below its minimum temperature set point, the blower
will automatically turn on in response to a timer control and aerate at
a rate of at least 31 m-Vh per dry metric ton (1,000 cfh/dt) so that the
pile will not go into an anaerobic state. If the pile temperature exceeds
the maximum temperature set point, the controller overrides the timer
and turns on the blower until the pile temperature is reduced to the set
point. The timers and temperature controllers enable the site operators to
maintain high oxygen levels in the piles while still maintaining desired
temperatures.
Each pile is monitored at five locations for temperature and three
locations for oxygen. Temperature probes are left in place during the
21 days of active composting and temperatures are recorded daily. Oxygen
readings are taken daily from oxygen-monitoring tubes. A temperature and
oxygen monitoring report is maintained for each pile denoting time of
measurement, location, and blower mode.
615
-------
During the early stages of active composting of anaerobically digested
sludge at the Hampton Roads facility, airflow rate is set to maintain
an oxygen content of less than 5 percent and temperatures greater
than 55 degrees C within the piles. Blowers rated to 34 m-Vh (1,200 cubic
feet per minute (cfm)) operate on timers. Typical seasonal operating modes
are presented below:
1. Winter operation--negative aeration, cycled 10 minutes on and
20 minutes off at 31 to 34 m-Vh per dry metric ton (1,000 to
1,100 cfh/dt).
2. Summer operation—positive aeration, cycled 10 minutes on and
20 minutes off at 34 to 38 m^/h per dry metric ton (1,100 to
1,200 cfh/dt).
For both operating modes, once 55 degree C temperatures are met for
3 consecutive days, the blowers are put on positive aeration continuously
at 78 to 87 m3/h per day metric ton (2,500 to 2,800 cfh/dt).
In the negative aeration mode, exhausted air must be treated for odors
by filtering through a finished compost pile. The positive aeration mode
requires no odor control; additionally, in this mode, drying of compost to
55 percent total solids is sometimes achieved during dry weather, enabling
direct screening of the compost after teardown.
With the positive aeration mode, leachate does not form as it does
with negative aeration. Because of this, material which collects in the
aeration troughs does not get wet under positive aeration, and odor is
not generated. Periodic trough cleaning is still required with either
aeration mode, however.
Each daily pile at the Hampton Roads facility is monitored at
six locations for temperature and at three locations for oxygen.
Temperature probes are left in place during the 21 days of composting.
Temperatures are recorded daily until 55 degrees C are obtained for
3 consecutive days, and then once per week. Oxygen readings are taken daily
using a portable oxygen meter.
The Columbus facility uses the following process control procedure:
the 0.75-kw (1-hp) aeration blowers are designed for reversible operation to
blow air into or exhaust air from the compost pile. Each blower is equipped
with a system to control the operation either by a timer, temperature probe
in the compost pile, or manually. A timer and manual control of the blowers
is utilized as required by the stage of the composting process, which
changes according to the season. In winter, aeration is not started until
the pile reaches 40 degrees C; then it is on for 20 minutes and off for 10.
In summer, aeration starts as the pile is constructed. Positive aeration is
now used. The process is monitored by the use of temperature probes that
are inserted into the pile as the pile is set up. A reading is taken
and recorded and this monitoring is continued until three daily readings
of 55 degrees C or greater are recorded. At such time, blower rate is
increased to 100 percent to remove moisture from the material. Oxygen
content is not routinely measured for process control at the Columbus
facility.
616
-------
DRYING AND CURING
Material composted by the aerated static pile method often is not dry
enough to efficiently screen directly (total solids greater than or equal
to 50 to 55 percent is required). In these cases, a separate drying step
is employed. As previously noted, personnel at the Hampton Roads facility
found that during the summertime, when positive aeration occurs for the
entire 21-day period, the compost may be dry enough for screening, and
the drying step is bypassed. Screened compost is then cured to provide
additional pathogen inactivation and odor removal prior to storage for
distribution. Benefits of this mode of operation are immediate recycling
of wood chips and reduction in the area required for curing. At other
times, however separate drying is performed. Normally, unscreened compost
is stored in the curing area until weather permits drying. Drying
operations entail transfer of the cured compost by front-end loader to a
drying slab where the material is spread out to a depth of 38 to 45 cm
(15 to 18 in.). Periodically, the compost is rototilled until a minimum of
50 percent total solids is achieved. Typically, the drying time is less
than 3 days in the summertime when temperatures are high. After drying,
material is piled under a covered shed until screened.
Overall curing time at the Hampton Roads facility is a minimum of
30 days, as required by state regulations. Additional storage time is
used, however, in both the curing area and in the final product storage
area, depending on materials handling considerations.
At the Site II facility, each pile is torn down by a front-end loader
after the active compost period, and composted material is transported to a
drying area where it is restacked for an aerated drying step prior to
screening (prescreen drying). For 24 to 48 hours prior to teardown, the
piles are aerated full time, reducing temperatures to remove most of the
heat. Therefore, steam or odor release is minimal. Compost is then
restacked over 10-cm (4-in.) perforated pipe and connected to a 0.75-kw
(1-hp) blower. The same pile dimensions as those for active composting are
used for prescreen drying operations.
Constant aeration is applied and a goal of 52 percent total solids
after prescreen drying has been set for process control. This can
generally be met within 4 to 5 days, even under wet-weather conditions.
Temperature and oxygen are monitored during prescreen drying. Generally,
pile temperature increases rapidly to 55 degrees C (within 1 day) after
restacking, then drops to ambient. Oxygen content is maintained at
>15 percent during the prescreen drying step.
After screening, front-end loaders transport screened compost to
storage for a minimum 30-day aerated curing period. Maximum storage
capacity of this area is limited to a 6-month production of screened
compost. Three-quarter-kilowatt (1-hp) blowers are used for curing pile
aeration and aeration pipes are 10-cm (4-in.) laterals, three per blower,
located on 2.5-m (8-ft) centers. Aeration maintains high oxygen levels in
the curing piles and prevents release of odors. Aeration continues until
617
-------
the compost is distributed. Temperature and oxygen are monitored during the
aerated curing period, with a similar process response as that described for
prescreen drying.
SCREENING OPERATIONS
Screening is an important step in the static pile process as this
operation is necessary for effective recovery and recycling of bulking
agent (wood chips). Finished compost uses, common in the United States,
typically require screens which produce finished compost grades in the
6- to 10-mm (1/4- to 3/8-in.) size range.
Under optimum conditions, wood chip recoveries of 80 to 90 percent
can be obtained in these size fractions using the types of screening
equipment employed at the three aerated static pile facilities investi-
gated. To do so requires that the total solids content of unscreened
compost be between 50 and 65 percent, depending in part on screen size and
design. Exceeding the upper limit can result in losses from excessive
dust generation, while not meeting the lower limit inhibits the separation
of fines from the wood chips. The latter situation leads to wood chip
contamination, reduced screening rates, and often mechanical problems with
the screening equipment.
Under sustained routine operation, wood chip recoveries of 70 to
85 percent are typical given the heterogeneity of compost material,
operational variability, and other factors. At one facility investigated,
three replicate observations using a 6-mm (1/4-in.) screen at screening rates
of 75 to 82 m^/h (99 to 107 cubic yards per hour (cu yd/hr)) and an
unscreened compost total solids of 56 percent, bulking agent recoveries of
64, 76, and 86 percent were obtained. By comparison, review of facility
records for 3 months of routine operation showed recoveries of 42 to
93 percent, with an average of 64 percent. Unscreened compost total solids
content was between 46 and 79 percent.
At the Site II facility, which employs fully enclosed screening
equipment, wood chip recoveries of 85 to 87 percent can be consistently
achieved during dry weather. However, under wet, cold weather conditions,
screening problems occur and recoveries are less even with effective
moisture control in the incoming unscreened compost.
MATERIALS FLOW CONSIDERATIONS
Representative bulk densities, total solids, and total volatile solids
of materials at various steps during static pile composting are presented
in Table 5. Data for facilities processing anaerobically digested sludge
(Hampton Roads) and raw sludge (Columbus) are shown.
FINISHED COMPOST QUALITY AND DISPOSITION
Table 6 presents a comparison of representative dewatered sludge
and finished compost characteristics at the three aerated static pile
facilities investigated in this study. Trace metal concentrations in the
dewatered sludge varied, and in general, this variation is reflected in the
618
-------
finished compost. This is also true for TKN and total phosphorus. Total
volatile solids of finished compost produced from raw sludge (without
lime) is slightly lower than that produced from digested sludge; finished
compost bulk densities are between 535 and 900 kilogram per cubic meter
(kg/m3) 900 and 1,500 Ib/cu yd).
TABLE 5. REPRESENTATIVE PHYSICAL CHARACTERISTICS OF MATERIALS
DURING STATIC PILE COMPOSTING
Process conditions Component
Anaerobically digested sludge*
summer operation Dewatered sludge
Wood chips
Recycled
New
Initial mix
Unscreened compost
Dried cured compost
Screened compost
Raw sludge, winter operation Dewatered sludge
Wood chips
Recycled*
Neu
Initial mix
Unscreened compost
Dried/cured compost**
Screened compost***
Bulk density,
Ib/cu yd
1,544
769
456
830
933
875
066
1,943
1,087
671
1,230
1,123
1,120
1,164
Total solids,
percent
16
56
55
41
48
56
52
16
44
58
39
40
41
52
Total volatile
solids, percent
54
71
82
69
68
59
49
71
66
98
73
63
71
64
*Recycled unscreened compost.
"After 31 days of curing.
***Data taken from previously screened material.
Note: Ib/cu yd x 0-59325 - kg/m3
Finished compost is sold to local users at all of the facilities
investigated. Demand varies with season and thus materials management
procedures for low-use periods are employed. The price of finished compost
varies from about $4/m^ to $12/m3 ($3/cu yd to $9/cu yd), depending on
facility location and quantity purchased. Typical finished compost
production rates are about 0.4 to 0.8 m3 per wet metric ton (0.5 to
1.0 cu yd per wet ton) of sludge processed.
ODOR CONTROL
The "earthy" smell of stabilized compost is often detectable in the
immediate vicinity of municipal sludge composting operations. Key design
and operational procedures for aerated static pile facilities to minimize
more objectionable odors from being generated and/or transported off site
include the following:
1. Trucks used to haul dewatered sludge should be covered and cleaned
frequently. This is particularly important with raw sludge.
2. Deliveries of dewatered sludge to the composting site should be
managed such that mixing and other operations can be performed
without sludge accumulating for long periods of time. This is
particularly important with raw sludge and in hot weather.
619
-------
3. The importance of moisture control during static pile composting
cannot be overstressed. Whether moisture is present in incoming
dewatered sludge, or the bulking agent or amendment; whether it is
added during processing from inclement weather; or whether it is
generated as a by-product of the composting process itself,
moisture must be controlled for effective odor control as well as
for stabilization, pathogen inactivation, and finished compost
quality control. An initial mix moisture content of 60 percent
or less (total solids >40 percent) is a key process design
criterion for effective performance. Enclosing the mixing opera-
tion, or other operations, and scrubbing the exhaust gas from the
aerated pile can be an effective step for odor control in some
situations.
4. Initial mix uniformity and porosity are also important for odor
control during the active composting period. The presence of
clumps of unmixed sludge can lead to anaerobic, and thus odorous,
conditions during the active composting period, as well as
incomplete stabilization and pathogen inactivation.
5. Proper pile construction techniques are important to ensure
porosity, aerobic conditions, minimize heat loss, and prevent
short-circuiting. Construction of daily pile compartments, with
cover material, is also important.
6. Positive aeration during active composting can minimize odor
generation potential since the pile cover material acts as an odor
scrubber. Negative aeration can require the use of a separate
exhaust scrubber system such as a finished compost filter pile.
Regardless of the aeration mode employed, control of aeration
rate, oxygen content, and temperature is critical for effective
odor control (and proper composting).
7. Pile teardown can be managed to minimize release of odors.
At one facility studied, piles are not torn down during wet
weather or in the early morning hours when air inversions may
result in odor generation. At another, high-rate aeration for
24 to 48 hours prior to teardown has been effective in odor
control.
8. An effective leachate, condensate, and runoff collection and
disposal system will minimize odor generation potential. Proper
drainage of aeration piping and troughs, and means of collecting
and transporting leachate and/or condensate to points of disposal
such that liquid does not accumulate and stagnate, are important
odor control features. Proper site drainage is required to
prevent ponding which can generate odor.
9. Effective housekeeping procedures such as washing equipment and
flushing or sweeping working areas such as mixing pads will also
reduce odor generation potential.
620
-------
TABLE 6. COMPARISON OF DEWATERED SLUDGE AND FINISHED
COMPOST CHARACTERISTICS
Characteristic*
Solids, percent
Total
Total volatile
Density, Ib/cu yd
Trace metals, mg/kg
Cd
Cu
Cr
Pb
Ni
Hg
Zn
TKN, as N, percent
Total phosphorus, as P,
Hampton Roads
facility
Sludge
17
55
1, 600
10
714
102
197
57
2. 21
1,651
5.4
t2.8
Compost
52
49
900
6.5
430
-
140
30
-
900
2.6
2. 1
Site II
facility
Sludge
17
43
-
3.0
186
-
196
24
1.8
487
2.5
1.2
Compost
51
-
1, 100 to
1, 500
2.0
112
-
40
46
0.4
176
1.0**
0.6
Col umbus
facility
Sludge Compost
17
74
1,900 1,
1,
38
272
279
287
-
-
2,583
-
-
52
64
100 to
400
14
216
67
178
-
-
965
1.3**
3. 3
percent
Fecal coliform per 100 gm
Salmonella
pH
8.2
to 8.3
<2 to 60
Negative
5.9
to 6.8
Negative
12.0
to 12.5
Negative
4.6 to
9.0
*Units are based on dry weight.
**Total nitrogen.
Note: Ib/cu yd x 0.59325 - kg/ra3
STATIC PILE OPERATING COSTS
Annual costs for on-site operations were reviewed as part of the
municipal sludge composting study. As shown in Table 7, over 75 percent of
the annual operating cost for aerated static pile systems is associated
with labor, bulking agent purchase and aeration pipe replacement. The
annual operating breakdown does not include amortized capital cost for
facilities and equipment or costs for sludge transport and general
administration.
Annual on-site operating costs (1984) for the Hampton Roads, Columbus
and Site II facilities based on current operations are about $33, $26, and
$24 per wet metric ton ($30, $24, and $22 per wet ton), respectively.
Annual average total solids content of the dewatered sludge processed at
the three facilities is about 17 percent. Therefore, corresponding unit
costs on a dry-weight basis are $194, $153, and $141 per dry metric ton
($176, $141, and $129 per dry ton). Sludge loadings associated with these
unit costs at Hampton Roads, Columbus, and Site II are 43, 104, and 338 wet
621
-------
metric tons (47, 115 and 373 wet tons) per operating day, respectively.
The Hampton Roads, Site II, and Columbus facilities are operated 6, 5, and
7 days per week, respectively.
TABLE 7. OPERATING COST BREAKDOWN FOR AERATED
STATIC PILE COMPOSTING
On-site operating cost category
Percent on total annual cost*
Labor, including fringe benefits
Bulking agent and aeration piping
Power
Fuel
Equipment maintenance
Laboratory and other expenses
Total on-site cost
40 to 50
22 to 42
1 to 10
5
11 to 15
1 to 8
100
*Range based on three facilities investigated.
Labor costs account for about 40 to 50 percent of the annual operating
costs at each facility (Table 7). Hampton Roads operates with 8 on-site
personnel, Columbus employs 19, and Site II utilizes 42.
Revenue generated from the sale of finished compost at the facilities
investigated ranges from about $1 to over $2 per wet metric ton «$1 to
over $2 per wet ton) based on current operations. All facilities have
established marketing programs to increase revenue in the future.
CONVENTIONAL WINDROW OPERATIONS
The conventional windrow composting facility at Los Angeles [2-10]
processes anaerobically digested sludge from an adjacent wastewater
treatment plant consisting of advanced primary sedimentation using anionic
polymers and secondary treatment using pure oxygen. Digested sludge is
dewatered by both basket and low-speed scroll-type centrifuges. Total
daily production of dewatered sludge at the treatment plant is 1,350 wet
Mg/d (1,500 wtpd), with approximately 900 wet Mg/d (1,000 wtpd) hauled to a
landfill and approximately 450 wet Mg/d (500 wtpd) processed at the adjacent
composting facility.
SITE FEATURES
The Los Angeles composting facility employs an asphalt-paved, 10-ha
(25-ac) composting field which is the main active windrow composting area.
All operations are performed outdoors. The layout also provides areas for
equipment storage, research operations, and expanded composting operations.
Several external amendments are used as a bulking agent in conjunction
with recycled finished compost, depending on the nature of the finished
compost product desired. The sludge-based products produced for sale at
this facility are:
622
-------
1. Nitrohumus—a general soil amendment product which is 90 percent
composted sludge and 10 percent sawdust.
2. Topper—a top dressing (new lawn covering, mulch) product that is
60 percent sawdust and 40 percent composted sludge.
3. Amend—a product recommended for vegetable and flower gardens; it
is 75 percent rice hulls and 25 percent composted sludge.
4. Gromulch—an outdoor planting mix which is similar to Topper
but contains other proprietary ingredients.
Dewatered sludge from the adjacent treatment plant is conveyed to
storage silos, and then from the silos to a truck-loading station. Twelve
sludge storage silos have a capacity of 500 wet metric tons (550 wet tons)
each, for a total capacity of 6,000 wet metric tons (6,600 wet tons).
Sludge is stored in the silos during nighttime hours, over weekends, and
during rainy periods when disposal (either composting or landfilling) is not
possible.
Normal dry-weather operation calls for all silos to be empty by the
end of the day shift on Fridays. By the end of the nighttime shift on
Sundays, the silos are at their maximum normal dry-weather storage level.
From Monday to Friday, storage levels oscillate downward, reaching zero by
Friday afternoon.
One operator from a central control panel can control the withdrawal
of sludge from all 12 silos and the operation of conveyor belts from silos
to two truck-loading stations. Each truck-loading station requires one
operator.
Each truck which hauls sludge from the storage silo to the composting
field is loaded with approximately 13.5 Mg (15 tons) of wet cake and
then moved to a loading area where amendment (rice hulls, sawdust, and/or
finished compost) is added from a stockpile. The amount of material added
is measured by the bucket volume of a front-end loader. Loaded trucks are
then driven to the 10-ha (25-ac) active windrow compost pad where piles are
formed.
SLUDGE CHARACTERISTICS
Characteristics of dewatered sludge processed at the Los Angeles
facility are summarized below:
Characteristic Value, percent
Total solids 22-25
Volatile solids 50
Nitrogen (as N) 2
Phosphorus (as ?205) 3
Potassium (as l^O) 0.1
Dewatered sludge density is about 1,050 kg/m^ (1,800 Ib/cu yd) at the range
of total solids concentrations indicated.
623
-------
WINDROW FORMATION
To form a windrow, trucks loaded with sludge and amendment drive to
a compost pad and discharge their load. Unloading time for power ram
trailers used is approximately 2 minutes. Combining the sludge and
amendment in the trailer prior to transport provides only limited initial
mixing; thus, the present method of forming a windrow consists of laying
two small windrows side by side as the truck unloads. Each small windrow
is first turned with a front-end loader, and then the two are pushed into
a single windrow with the front-end loader.
Two front-end loaders are used for mixing, amendment loading, and
windrow formation. One has a 3.5-m-' (4.5-cu yd) bucket, and the other has
a 4.5-m3 (5.75-cu yd) bucket. After the preliminary mix during windrow
formation, initial windrow mixing is performed using a conventional mobile
composter. The conventional mobile composter has the capacity to turn
approximately 6 Mg (7 tons) per minute of wet mixture having a bulk density
of 900 to 1,000 kg/m3 (1,500 to 1,700 Ib/cu yd). A schematic of windrow
formation operations has been presented previously on Figure 3. The
cross-sectional dimensions of a typical windrow at the completion of this
stage in the composting process are 1.2 to 1.5 m (4 to 5 ft) high and 4.3 m
(14 ft) wide at the base.
ACTIVE WINDROW COMPOSTING
The active composting period takes place over a period varying from
30 to 90 days, depending upon ambient temperatures and drying rates. A
typical operation for a 56-day active composting period is shown on Figure 4
and described below. Six small windrows such as those described in the
previous section are constructed in the first week. Each windrow is
about 250 m (800 ft) long with up to 475 wet metric tons (525 wet tons) of
dewatered sludge per windrow. At the end of the second week, the six small
windrows are combined into four intermediate size windrows. After the sixth
week, the four intermediate windrows are combined into a very large windrow.
Internal windrow temperatures are monitored in the very large windrow to
demonstrate compliance with the criterion of 15 days at 55 degrees C or
greater. A large mobile composter is used to form and turn the large
windrow, which minimizes heat loss.
A turning frequency of three turns per week has been found to provide
adequate pathogen inactivation without affecting drying. The cross-
sectional dimensions of the large windrow are 2 m (7 ft) high and 7 m
(23 ft) wide at the base. The large windrow is broken down in the ninth
week, and the finished compost is delivered to a private company located
next to the composting facility which markets the product.
The large mobile composter has the capacity to turn approximately
10 Mg (11 tons) per minute of wet mixture having a bulk density of 900 to
1,000 kg/m3 (1,500 to 1,700 Ib/cu yd). Both mobile composters are self-
propelled machines designed specifically for windrow composting. Each
machine straddles the windrow and has a high-speed rotating drum at ground
level. The drum has flails which lift the sludge up and over the drum,
624
-------
depositing it behind the machine in windrow form. Turning mixes the wet
cake and amendment, increases porosity in the windrow to maintain aerobic
conditions, promotes drying of the sludge by exposure to air and sun,
and ensures that all of the sludge is subjected to the high internal
temperatures of the windrows. Table 8 is a summary of windrow properties
for the Los Angeles facility.
CONSTRUCT SIX SMALL WINDROWS
WEEK 1
WEEK 2 (END)
WEEK 3-6
WEEK 7-8
WEEK 9
REMOVE COMPOST
CLEAR FIELD
FIGURE 4. ACTIVE WINDROW COMPOSTING OPERATIONS AT
LOS ANGELES FACILITY
625
-------
TABLE 8. WINDROW PROPERTIES
Mobile composting
machine*
Coventional size
Conventional size
Large size
Volume, cu yd/
100 linear ft
80
125
335
Windrow property
Volume,
cu yd/ac
1,000
1,800
3,500
Surface to volume
ratio, ft/ft
0.83
0.60
0.32
*See text.
Note: cu yd/100 linear ft x 2.508 - m3/100 linear m
cu yd/ac x 1.889 - m3/ha
ft/ft = m/m
Recently, the process described above has been modified to eliminate
construction of the intermediate-size windrows. Instead, the large windrows
are built directly from three small windrows. Experience has shown that
using three, rather than six, small windrows to construct the large windrow
gives sufficiently low surface to volume ratios to conserve heat and meet
recommended time-temperature performance standards.
Experience at the Los Angeles composting facility has established that
for composting to proceed satisfactorily, windrows should have an initial
total solids content of at least 40 percent. This is achieved by mixing
dewatered cake with previously finished compost material or other
amendment. A total solids content of less than 40 percent results in a
mixture of low porosity which inhibits oxygen transfer within the windrow.
Initial volatile solids content of windrows at the Los Angeles facility
is typically about 45 to 50 percent. When the volatile solids have been
reduced to 40 to 45 percent, and the total solids have increased to between
60 and 65 percent, the composting process is completed. After the 8-week
active windrow composting period, the large windrow is broken down during
the ninth week using a front-end loader, and the finished compost is
trucked to a stockpile area. After stockpiling, delivery is made to a
private company which screens all material for objects 10 mm (3/8 in.) and
larger prior to bagging and sale.
PRODUCT QUALITY
The general properties of finished compost from the Los Angeles
composting facility are listed below:
Item
Total solids
Compost with recycled sludge amendment
Compost with sawdust or rice hull
amendment
626
Value
60-65 percent
50-55 percent
-------
Item
Volatile solids
Compost with recycled sludge amendment
Bulk densities
Compost with recycled sludge
Compost with sawdust amendment
Compost with rice hull amendment
Carbon to nitrogen ratio
Compost with recycled sludge
Compost with rice hulls
Compost with sawdust
Particle size
Median size
90 percent of particles
Value
40-45 percent
770 kg/m3 (1,300 Ib/cu yd)
520 kg/m3 (875 Ib/cu yd)
500 kg/m3 (850 Ib/cu yd)
12
30
135
1.5 mm (0.06 in.)
less than 5 mm (0.20 in.)
Table 9 presents a summary of cadmium, lead, and PCS concentrations
detected in compost products from the Los Angeles composting facility.
Heavy metal content of the finished compost has increased since 1977. Data
prior to 1977 show that finished compost was less contaminated with heavy
metals than at present because only large digested sludge particles were
captured in the sludge cake with old dewatering equipment used at the
treatment plant. Cadmium is associated with smaller digested sludge
particles now being captured with new dewatering equipment. Data from the
period prior to 1977 show that on a dry-weight basis finished compost
cadmium content averaged 26 mg/kg, compared to 50 to 70 mg/kg currently.
TABLE 9. CADMIUM, LEAD AND PCB CONTENT OF WINDROW COMPOST PRODUCTS
Concentration, mg/kg dry weight
Compost product
Nitrohumus
Amend
Topper
Year
1983
1984
1983
1984
1983
1984
Cadmium
60
70
26
57
32
47
Lead
470
510
250
350
230
330
PCBs
0.4
0.3
0.3
Recycled compost
1982-83
627
-------
ODOR CONTROL
Odor measurements at the facility indicate that 83 percent of windrow
odor emissions are the result of ambient surface emissions and 17 percent
are the result of windrow turnings. This is equivalent to 320,000 odor
units per square meter (ou/m^) (30,000 odor units per square foot (ou/
sq ft)) for ambient surface emissions and 65,000 ou/m^ (6,000 ou/sq ft)
for the windrow turnings. These values are based on a 40-day active windrow
composting period with 20 turning cycles.
Ambient surface emissions from a windrow decrease significantly
as the active compost cycle progresses. This is evident from the chart
presented on Figure 5. Emissions are also the greatest immediately after
windrow turning, as shown on Figure 6.
I
o
V)
O
5 4
in
cc
O
Q
o
LU
O 0
< 2
u.
oc
D
V)
o
cc.
Q r»
NOTE: ou/min/sq ft x 10.764 = ou/min/m2
10
20
30
40
FIGURE 5,
ACTIVE COMPOSTING TIME, days
AVERAGE WINDROW SURFACE ODOR EMISSIONS
DURING A COMPOST CYCLE
Experience at the Los Angeles composting facility has indicated that
the best way to control odors and minimize complaints is to limit the size
of the composting operation. The maximum amount of sludge which can be
composted in the summer without odor complaints is 450 wet Mg/d (500 wtpd).
Due to lower productivity in the winter, the annual average is 385 wet Mg/d
(425 wtpd), calculated on a 7-day-per-week operation.
628
-------
I
1
co
O
UI
§
LU
O
U.
cc
(O
20
15
10
NOTE: ou/min/sq ft x 10.764 = ou/min/m2
I
FIGURE 6
0 20 40 60 80
MINUTES AFTER TURNING
SURFACE ODOR EMISSIONS AFTER WINDROW TURNING
100
Ambient surface emissions from windrow composting cannot be eliminated
but are subject to some degree of control based on the type of amendment
used. Therefore, when these maximum sludge quantities are exceeded, the
odor threshold is exceeded and local residents complain. Even when
composted quantities are within these limits, certain meteorological
conditions (e.g., inversion layers) can cause odor complaints. Various
chemical-masking agents have been tried but have been found to be
ineffective.
WINDROW OPERATING COSTS
Operating costs for the latter part of 1984 and the early part of 1985
were reviewed as part of the study. This period reflects current routine
operations at the Los Angeles composting facility and specifically excludes
a period during the early part of 1984 when current operating procedures
were initiated, productivities were low, and supervisory costs were high as
personnel were trained in new operations. Composting costs per operating
day are presented below:
629
-------
Dollars per
Item operating day
Equipment capital recovery 1,050
Fuel 475
Maintenance 500
Wages (salary plus fringe benefits) 2,000
10 operators plus 1 foreman
Total costs 4,025
The total cost shown above is equivalent to $37 per dry metric ton
($34 per dry ton) or $9 per wet metric ton ($8 per wet ton) of sludge
composted based on a sludge loading of 450 wet Mg/d (496 wtpd) on an
operating day basis at 24 percent total solids. External amendment is used
during conventional windrow composting at the Los Angeles facility but is
available at no cost. Revenue equivalent to $4 per dry ton is generated
from finished compost sale. Subtracting this from the unit operating cost
yields a net cost of about $33 per dry metric ton ($30 per dry ton) or
$8 per wet metric ton ($7 per wet ton) of sludge composted.
Approximately $300,000 per year is spent on research to better
understand the process and to investigate improvements. Labor costs
for solids analyses, chemical analyses, and pathogen determinations are
approximately $60,000 per year or $1.40 per dry metric ton ($1.30 per dry
ton) of sludge cake composted. There is no cost for bulking agents used at
the Los Angeles facility.
SUMMARY
Over 40 municipalities in the United States employ composting
as a means of treating dewatered sludge prior to ultimate disposal.
The majority of these communities employ either the aerated static
pile or conventional windrow technique. Raw, aerobically digested and
anaerobically digested sludges are processed at average loadings which vary
from below one wet Mg/d (1 wtpd) to 450 wet Mg/d (500 wtpd).
Three aerated static pile facilities and one conventional windrow
facility were investigated to identify physical characteristics such as
site features and equipment, and to define operating procedures. One
aerated static pile facility studied processes anaerobically digested
sludge, two process raw sludge, and the conventional windrow facility
treats anaerobically digested sludge. Nominal dewatered sludge loadings at
these facilities vary from 45 to 450 wet Mg/d (50 to 500 wtpd). Current
practice at these facilities provides insight into design and operating
requirements for applying the municipal sludge composting technologies at
other locations.
ACKNOWLEDGMENTS
Field investigations at the aerated static pile facilities described
were performed with the assistance of Dr. E. Epstein and Dr. J. Alpert of
630
-------
E&A Environmental Consultants, Inc., Stoughton, Massachusetts. M. McLemore,
T. Williams, and D. Finley, of the Hampton Roads Sanitation District
Peninsula Composting Facility, C. Murray, G. Crosby, and J. Thompson, of
the Washington Suburban Sanitary Commission Site II Composting Facility,
and D. Rodgers and D. Goodridge of the City of Columbus Southwesterly
Composting Facility provided valuable assistance and insights into static
pile composting at their respective facilities, as did R. Caballero of the
Los Angeles County Sanitation Districts conventional windrow facility.
REFERENCES
1. Brown and Caldwell. Municipal sludge composting technology evaluation.
Prepared for the U.S. Environmental Protection Agency, Cincinnati,
Ohio, 1985.
2. Caballero, R. "Experience at a Windrow Composting Facility: Los
Angeles County Site. In Sludge Composting and Improved Increrator
Performance, EPA Technology Transfer. July 1984.
3. Hay, J.C., et al. "Disinfection of Sewage Sludge by Windrow
Composting." National Science Foundation Workshop on Disinfection,
Coral Gables, Florida. May 7-9, 1984.
4.
lacoboni, M.D., et al. "Windrow and Static Pile Composting of
Municipal Sewage Sludges." EPA, Municipal Environmental Research
Laboratory, Cincinnati, Ohio. May 1982.
5. Garrison, W.E. "Composting and Sludge Disposal Operations at the
Joint Water Pollution Control Plant." County Sanitation Districts of
Los Angeles County, Whittier, California. June 23, 1983.
6. LeBrun, T.J., et al. "Overview of Compost Research Conducted by the
Los Angeles County Sanitation Districts." National Conference on
Municipal and Industrial Sludge Composting, Philadelphia, Pennsylvania.
November 17-19, 1980.
7. Hay, J.C., et al. "Forced-Aerated Windrows Composting of Sewage
Sludge." Virginia Water Pollution Control Association Conference,
Williamsburg, Virginia. April 30-May 2, 1984.
8. lacoboni, M. "Compost Economics in California." Biocycle, July-August
1983.
9. lacaboni, M. , et^al. ;"Deep Windrow Composting of Dewatered Sewage
Sludge." National Conference .oC-Mvrnicipa,!;. and Industrial Sludge
Composting, Philadelphia., Pennsylvania'.., ^Nto^fembjdr 17-19, 1980.
',- '•„, ••!*<; i.; v ' j,' /•J.-J-iii. ,
10. Horvath, R.W. "Operating and Design Criteria for Windrow Composting of
Sludge." National Conference on Design of Municipal Sludge Compost
Facilities, Chicago, Illinois. August 29-31, 1978.
631
US. GOVERNMENT PRINTING OFFICE 1986 - 646-0! 4/40020
------- |