T!
r \
* !
4
I .1
| 1
H
. I
?!
I
I i
I
M'.
I
h
Ki
n-
3!
¦
PB86-156965
Status of Porous Biomass Support
Systems for Wastewater Treatment
An Innovative/Alternative Technology
Assessment
Dynamac Corp., Kockville, MD
Prepared for
Environmental Protection Agency, Cincinnati, OH
Jan 86
U& fapartmcnt of Centre®
ftsfisrJ Ttchmeal yomtstion Z&vx*
"j i %
-------
;iV
University of Idaho
&
"w
^^Jg1oT<4 PB86-156965
•)th ?Xo°r St.S-^-
100 30303
^rtanta' EPA/600/2-86/019
January 1986
STATUS OF POROUS BIOMASS SUPPORT SYSTEMS FOR WASTEWATER TREATMENT:
AN INNOVATIVE/ALTERNATIVE TECHNOLOGY ASSESSMENT
by
William C. Boyle
University of Wisconsin
Madison, Wisconsin 55706
Alfred T. Wallace
nlversity of Idahc
Moscow, Idaho 838A3
Contract No. 68-03-3130
Project Officer
James A. Heidman
Wastewater Research Division
Water Engineering Research Laboratory
Cincinnati, Ohio 45268
WATER ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
St' HPftODUClO Br
a- NATIONAL TECHNICAL
i INFORMATION SERVICE
US. piPfR.IWKJ « C0MUERCE
ft
C,
Ste-\
SPBINGflElD. VA. 22161
S:'' '
87-2JT
-------
TECHNICAL REPORT DATA
(Please read IhutjcHons un the reverie before completing}
1. REPORT NO. 2.
EPA/600/2-86/019
3. RECIPIENT'S ACCESSION NO.
PB8 6 I5b9b5/A<
4. TITLE AND SUBTITLE
Status of Porous Siomass Support Systems for Haste-
water Treatment: An Innovative/Alternative Technology
Assessment
6 RETaRnTuDaAr7yE 1986
S. PERFORMING ORGANIZATION CODE
7. AUTHOR(S) 1
William C. Boyle ?
Alfred T. Wallace^
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME ANO ADDRESS
^ University of Wisconsin, Madison, UI 55706
^ University of Idaho, Dept. of Engineering
Moscow, ID 03343
10. PROGRAM ELEMENT NO.
li. fcV1s1ts were made to laboratories of the original process developers
in the United Kingdom and in West Germany;> Data were gathered through
interviews with academic and commerical investigators in both countries
and through a review of all available literature and data, both published
and unpublished.
-~ihe study concluded that PBSS technology does not presently qualify
as a fully developed technology, but that the technology offers some
attractive potential benefits and very little risk for some intended
appHcations.—-Thorough pilot plant and full-scale studies are needed to
answer many remaining questions about the process and to provide design
data and guidance.
17. KEY WORDS ANO DOCUMENT ANALYSIS
L DESCRIPTORS
b. 1DENT 1F1ERS/OPEN ENDED TERMS
c. cosati Field/Group
IB. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
105
20. SECURITY CLASS fThispa^e)
Unclassified
22. PRICE
¦PA Perm 2220*1 (9-72) £
-------
DISCLAIMER
The information in this document has been funded wholly or in part by the
United States Environmental Protection Agency under Contract No.
68-03-3130 to Dynamac Corporation. It has been subject to the Agency's
peer and administrative review, and it has been approved for publication
as an EPA document. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
11
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with
protecting the Nation's land, air, and water systems. Under a mandate of
national environmental laws, the agency strives to formulate and implement
actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life. The Clean Water Act,
the Safe Drinking Water Act, and the Toxic Substances Control Act are three of
the major congressional laws that provide the framework for restoring and
maintaining the integrity of our Nation's water, for preserving and enhancing
the water we drink, and for protecting the environment from toxic substances.
These laws direct the EPA to perform research to define our environmental
problems, measure the impacts, and search for solutions.
The Water Engineering Research Laboratory is that component of EPA's
Research and Development program concerned with preventing, treating, and
managing municipal wastewater discharges; establishing practices to control
and remove contaminants from drinking water and to prevent its deterioration
during storage and distribution; and assessing the nature and controllability
of releases of toxic substances to the air, water, and land from manufacturing
processes and subsequent product uses. This publication is one of the
products of that research and provides a vital communication link between the
researcher and the user community.
This report evalutes the technology of porous biomass support systems and
the potential application of this process to biological wastewater treatment
systems. This new technology offers some attractive potential benefits to
wastewater treatment technology.
Francis T. Mayo, Director
Water Engineering Research Laboratory
iii
-------
ABSTRACT
A study was conducted to assess the emerging wastewater treatment
technology of porous biomass support systems (PBSS). These systems use
large numbers of small, open-cell or reticulated polyurethane foam pads
to support high concentrations of biomass in an aeration basin.
Visits were made to laboratories of the original process developers
in the United Kingdom and in West Germany. Data were gathered through
interviews with academic and commercial investigators in both countries
and through a review of all available literature and data, both published
and unpublished.
The study concluded that the PBSS technology does not presently
qualify as a fully developed technology, but that the technology offers
some very attractive potential benefits and very little risk for some
intended applications. Thorough pilot plant studies &re needed to answer
many remaining questions about the process and to provide design data
and guidance.
This report was submitted in partial fulfillment of Contract No.
68-03-3130 by Dynamac Corporation, Rockville, Maryland, under the sponsor-
ship of the U.S. Environmental Protection Agency.
iv
-------
CONTENTS
Page
Foreword iii
Abstract iv
Figures vii
Tables ix
Acknowledgments xi
1. Introduction 1
2. Conclusions 3
3. Recommendations 4
4. Technology Description 9
Brief summary description 9
The Captor process 12
Basic process description 12
Captor process modifications 15
The Linpor process 17
Linpor-C , 17
Linpor-N, effluent polishing and nitrification 20
5. Development Status 22
Summary of Research Findings 22
U.K. Experiences - UMIST/Simon Hartley Ltd 22
West Germany - Linde AG . . 23
Summary of Pilot and Full-Scale Facilities 28
Experience in the United States 28
Experience in the United Kingdom 41
6. Technology Assessment 53
Process Theory 53
Porous Support Particles 56
Fluidization and Hydraulic Distribution 57
Biomass Hold-up Characteristics 59
Substrate Uptake Relationships 63
Other Considerations 69
Process Capabilities, Applications, and Limitations 69
Design Considerations 70
PBS Particle Characteristics 71
PBS Particle Number and Biomass Concentration 72
Substrate Uptake/Loading 73
Air Flow - Oxygen Transfer/PBSP Mixing 74
PBS Particle Cleaning/Sludge Characteristics 76
Screening 77
v
-------
Energy Analysis 77
Operation and Maintenance Requirements 77
Costs. 78
7. Comparison with Conventional Technologies 79
General 79
Cost 79
Energy. . 81
Performance/Reliability 81
Environmental Benefit 83
Toxics Management 83
Joint Treatment Potential 84
Residuals Generated 84
8. National Impact Assessment 86
Present and Future Market 86
Market Penetration Potential 87
Risk Assessment 87
REFERENCES 90
vi
-------
FIGURES
Number Page
1 Typical Porous Biomass Support Pads 10
2 Aeration Basin Containing Pads 10
3 Flow Schematic - Basic Captor Process 13
4 Diagram of a Captor Pad Cleaner 13
5 Effluent Polishing and Nitrification Process Using Captor ... 16
6 Modified Solids Contact Process Using Captor 17
7 Flow Schematic - linpor-C 19
8 Flow Schematic - Linpor-N 21
9 Marion, Illinois, Pilot Plant Schematic Showing Sampling
Locations 31
10 Marion, Illinois, Pilot Plant Results, Total BOD Removals -
7 Day Averages 32
11 Marion, Illinois, Pilot Plant Results, Soluble BOD Removals
7 Day Averages 32
12 Downingtowu, Pennsylvania, Pilot Plant Results, Total
BOD Removals - 7 Day Averages 36
13 Downingtown, Pennsylvania, Pilot Plant Results, Soluble
BOD Removals - 7 Day Averages 36
14 Performance of an Industrial Captor System
(Employing Sludge Recycle) 40
15 Captor System for Bakery Wastewater Pretreatment 41
16 Flow Schematic to Evaluate Captor vs. Activated Sludge
at Freehold 43
17 Freehold WR Plant - Monthly Average Effluent BOD 48
vii
-------
Number Page
18 Freehold WR Plant - Monthly Average Effluent Suspended
Solids 48
19 Flow and Soluble BOD Load During 24-Hour Survey 49
20 Performance of PBSS Using Glucose 66
21 BOD Removal vs. Volumetric Loading Rate 67
22 A Joint Industrial-Domestic Flow Diagram
Indicating Potential Economic Advantages 85
viii
-------
TABLES
Number Page
1 Pilot Plant Results - Linpor 24
2 Pilot Plant Results - Linpor-N 25
3 Linpor-C Test Results at Site F 26
4 Linpor-C Test Results at Site M 27
5 Performance of the Marion, Illinois, Pilot Plant with a
Pad Concentration of 40/L 33
6 Characteristics of Industrial Captor Treatment System 39
7 Captor Pretreatment System for a Bakery - Post Break-in
Performance 42
8 Design Loading Conditions - Freehold Water Reclamation
Works 44
9 Design Details - Freehold 44
10 Predicted Loadings and Performance - Freehold Water
Reclamation Works 45
11 Freehold 24-Hour Survey, June 1984 50
12 Statistical Comparisons - Freehold Captor Lanes 1 and 2,
2^-Hour Survey, June 1984 52
13 Characteristics of Reactors Containing PBS Particles 55
14 Characteristics of Typical PBS Particles 57
15 Some Reported Applications of PbSS's 60
16 Biological Rate Equations for PBSS. 65
17 Reactor Configurations for PBSS's and Their Characteristics. . . 71
18 Selected PBSS Design Parameters 72
ix
-------
Number Page
19 PBSS Operation ami Maintenance Requirements 78
20 Performance Assumptions - Economic Analysis of Captor vs.
Conventional Activated Sludge 80
x
-------
Acknowledgments
This report was prepared by Dr. William C. Boyle of the University
of Wisconsin-Madison and by Dr. A. T. Wallace of t.he University of Idaho.
Project Officer for the U.S. Environmental Protection Agency was Dr. James
A. Heidnian of the Innovative and Alcernative Technology Staff {IATS), with
assistance from Mr. Richard C. Brenner of the IATS, both located at the Water
Engineering Research Laboratory at Cincinnati, Ohio.
A large number of individuals contributed time, ideas, and data to the
development of this documet-t. Prominent among these wer3:
o From Scotfoam Corporation, Eddystone, Pennsylvania:
Dr. Louis R. Matlack, Vice President
Richard Hankel, Process Engineer
William Hatzell, Technician
o Glen Clarida, Clarida Engineering Co., Marion, Illinois
o Ron McKinney, Wastewater Superintendent, Marion, Illinois
o Daniel H. Houck, D. H. Houck and Associates, Silver Spring, Maryland
o David N. Immendorf, Engineering Management, Inc., Reading, Pennsylvania
o Gene DeCarlo, Assistant Wastewater Superintendent, Downiagtown,
Pennsylvania
o Dr. Werner Hegemann,Technical University of Munich at Garching
o From Linde AG • Werksgruppe TVT, Munich, Germany:
Dr. Hans Reimann
Dr. Eberhard Lassmann
xi
-------
From the Sewage Processes Development Group, Water Research Centre,
Stevenage, Herts, United Kingdom:
Paul F. Cooper, Section Leader
Ian 1. Thompson, Research Scientist
From Simcn-Hartley, Ltd., Stoke-on-Trent, United Kingdom:
Eric P. Austin, Technical Director
Paul Adams, Principal Process Engineer
Dr. Iar. Walker, Principal Research Engineer
Mr. Cnailes E. Tharp, Ashbrook-Simon-Hartley, Houston, Texas
From the University of Manchester, Institute of Science and Technology:
Dr. Bernard Atkinson
Dr. Colin Webb
xii
-------
SECTION 1
INTRODUCTION
The Clean Vater Act of 1977 and the Municipal Wastewater Construction
Grant Amendments of 1981 include provisions that encourage the use of innova-
tive and alternative (I/A) wastewater treatment technologies. To identify
candidate technologies that might qualify as innovative or alternative, tech-
nology assessments of developing technologies are required. These assessments
provide technical feedback to the technology development community and to the
Federal, state, local, and private agencies involved in using emerging tech-
nologies to enhance the quality of the Nation's waters.
One of the emerging technologies selected by the U.S. Environmental
Protection Agency (EPA) for assessment is porous biomass support systems
(PBSS). These systems use large numbers of small, open-cell or reticulated
polyurethane foam pads to support high concentrations of biomass in an aera-
tion basin. The basic technology hap. been developing along two different
lines. The British developer, Simon-Hartley, Ltd., has concentrated on exter-
nal pad cleaning devices to waste the excess biomass in an effort to avoid the
use of secondary clarifers in some applications of the process. The West
German developer, Linde AG, has chosen to Torego external pad cleaning and has
developed their process more along the lines of conventional activated sludge
treatment. In their ?rcccs6, they grow biomass both on the pads and in
suspension. The function of the pads is to retain a large quantity of biomass
in the aeration basin, thus reducing the solids loading on the secondary
clarifiers and maintaining a higher effective mixed liquor suspended solids
(EMLSS) concentration than could be maintained without the use of the pads.
1
-------
To accomplish thi': assessment, visits were made to the laboratories of
the original process developers in the United Kingdom and in West Germany, and
the status of research and development efforts was discussed with academic and
commercial investigators in both countries. Several pilot-scale and full-scale
development projects were also visited in both countries during this visit.
Two pilot-scale systems and one small industrial-scale system in the United
States were also visited as part of this assessment. All available literature
and data, both published and unpublished, were reviewed. This report is the
result of these efforts, most of which took place during the spring and summer
of 1984.
2
-------
SECTION 2
CONCLUSIONS
Based on the visits to operating facilities and the review of all the
available data, we have concluded that PBSS technology does not presently
qualify as a fully developed technology. Definite risks are associated with
adopting it for several of the applications for which its developers intended
it. On the other hand, the technology offers some very attractive potential
benefits and very little risk for some intended applications.
The manufacturers and suppliers of PBSS technology would never reconspend
installation of a system at present without thorough pilot plant investigation
to provide design data and guidance. As a result, adoption of the
technology on a full scale will proceed slowly, but it should not be marred by
spectacular failures. The really promising applications of PBSS at this time
appear to include:
1. Pretreatment systems for high-strength industrial waste,
2. Upgrading overloaded activated sludge plants, especially those that
are habitually plagued by filamentous bulking, and
3. Addition to existing systems to produce nitrification without the
need for additional clarifers.
3
-------
SECTION 3
RECOMMENDATIONS
GENERAL
The Linpor-C system (Section A) appears almost ready to >>eccme classified
as one of the regular members of the available spectrum of biological treatment
systems capable of consistently meeting secondary treatment requirements. Two
full-scale lield projects near Munich have demonstrated the efficacy of the
Linpor C concept for upgrading existing municipal wastewater treatment plants
(Section 5). The Captor process, used as the 6ole biological process without
final clarifiers, is not ready to become so classified. Both systems require
extra attention to the quality and reliability of preliminary and primary
treatment processes to avoid severe operation and maintenance problems with
pad cleaners (Captor) and retaining screens (both processes). Furthermore,
design requirements to ensure proper mixing and distribution of porous biomass
support (PBS) particles need to be better defined. Porous biomass support
system (PBSS) technology should be treated as an innovative technology with a
degree of risk that varies with the specific application. PBSS projects
receiving consideration for I/A funding should be those for which the risks
are minimized through careful consideration of the specific details of the
project. As the technology continues to evolve and be better understood, it
should be possible to attempt some projects with higher risk factors, providing
that high potential benefits are possible as well.
FURTHER R&D EFFORTS
A tremendous number of questions about PBSS's retain. Answers to these
questions could increase the attractiveness of this technology to potential
4
-------
users. Some of the research and developmeat areas that have been pinpointed
through the work of this technology assessment are discussed in the following
sections.
The Po-ous Biomass Supports
Very little remains to be done in this area. We assume that the foam
manufacturers will continue to look at new materials that might result in less
pad wear and hence an improvement in system economics. The Captor pad dimen-
sions and pore size characteristics have evolved as about optimum for this
process, so few changes, if auy, are foreseen for these. A6 the Linpor pads
do not have to be handled in the same way as the Captor pads, it is con-
ceivable that they might be altered somewhat to try to improve the cost-
effectiveness of this system. For example, although the Linpor pads must be
large enough to be retained by screens, they could easily be of random shape
and not necessarily cut to size, as they are at present. Such a shift in
manufacture might result in lower cost for the pads.
One very interesting possibility for basic research would be the use of
specially manufactured or pretreated pads for enhanced removal of toxic or
other materials when PBSS's are used as the first stage of a multi-stage
biological system. A simple example would be the pre-coating of pads with a
polyelectrolyte after pad cleaning to enhance the removal of suspended solids
or to assist in colonization during system start-up.
Biomass Properties
A need exists for substantial fundamental research on the characteristics
of PBSS biomass. The first basic question concerns the physiological equivalence
of biomass contained within porous supports and freely suspended biomass at
identical loadings. Inherent differences may exist in the distribution of
physiological states between PBSS's (especially Captor) and activated sludge
because of the equal and independent probability of wasting sludge of any age
with conventional activated sludge and the (potentially) higher probability of
wasting younger biomass from a PBS particle. This assumption presumes that
5
-------
the younger biomass is always found uear the surface of the pad and that the
older, nore mineralized biomass will be found at the interior. With the
development of highly effective pad cleaners, this could cease to be of any
concern.
The following areas also need intensive evaluation: the effect of pad
cleaning rates on biomans physiology, specific substrate uptake, biomass
sludge yields, sludge dewaterability, diffusional resistance to oxygen and
substrate transport, biomass hold-up, free suspended solids, and fluidizing
properties of PBS particles. Similar attention should also be paid to Linpor
systems in which PBS particle biomass hold-up is in equilibrium between growth
aad hydraulic attrition.
Pad Cleaning
Althoughsludge concentrations of 4 to 5 percent have been achieved in pad
squeezing during several laboratory and pilot plant studies, it has been
difficult to reproduce these results routinely in full scale. Modifications
to the pre-squeeze roller system are currently being studied to help bridge
this gap. The current thinking is to replace the pre-squeeze roller with a
belt stretched between two rollers to increase contact time for the pre-
squeeze function in hopes of removing more interstitial water.
Substrate Uptake
The independent parameters that influence substrate uptake and removal
for conventional suspended growth systems will likely influence PBSSs. Unfor-
tunately, to date, no definitive variable or set of variables have been
defined as a predictive parameter. Volumetric loading and hydraulic residence
times are currently used. Efforts to use F/M or SRT have not yet been
entirely successful. A part of this problem is the analytical determination
of active biomass from the particles. Additional development of effective
standardized procedures of measurement are important in solving this problem.
The role of diffusion in process performance is also critical. Evidence from
pilot plants suggests the importance of diffusional transport of oxygen in
6
-------
nitrification processes. Similar phenomena are likely to exist in carbonaceous
systems as well. The impacts of bulk substrate and oxygen concentration,
particle characteristics (pore size, dimensions), hydraulic regime, biomass
hold-up, and reactor temperature among others on substrate uptake may provide
leads as to the importance of diffusion in FBSS's.
Finally, some data are needed to indicate the level of sensitivity of
PBSS's to reactor temperature, pH, toxic shock, and hydraulic and organic
shock.
Air Flow Requirements
Air from submerged diffusers provides oxygen to support basic biomass
respiration and provides turbulence to mix and distribute PBS particles.
Off-gas oxygen transfer studies under process conditions must be conducted in
order to effectively evaluate the rate of mass transfer of oxygen to PBSS's.
Process conditions under which these tests are conducted must be such as to
provide proper PBS particle distribution. Mass transfer comparisons between
PBSS's and conventional suspended growth systems must be made under process
conditions which are optimal for each system if they are to be meaningful in
any kind of economic or energy audit.
The effective distribution of PBS particles in the reactor is of paramount
importance to the success of this process. Every indication is that this can
be achieved in diffused air systems if proper downflow regions are created.
Pad distributions may also be effectively mixed by mechanical means. Mechanical
power inputs to properly distribute pads may be lower than those used in
diffused air systems. To date, no definitive research has been performed to
assess the economic value of mechanical pad distribution.
Currently, there is research under way on PBS particle mixing with air in
both West Germany and the United Kingdom. Investigators at the Water Research
Centre in the U.K. also indicate that oxygen transfer studies will take place
at Freehold in the near future.
7
-------
Sludge Settling Properties
The West German experience has been favorable with respect to improving
biomass settling properties with PBS particle additions. Whether the particles
improve biomass SVI by virtue of reducing reactor F/M, by changing the
predominant suspended growth population via attrition of more settleable
biomass or by some other phenomenon is not entirely clear at this time.
Side-by-side full-scale studies at plants with endemic sludge settling problems
would be useful. One such study at Freising, Weft Germany is under way at
this time, but experimental control at this facility is uncertain.
Further basic research on biomass properties in Linpor-type systems may
also be very useful for better describing what has been seen in West Germany.
In concert with biomass settling is the effect of P8SS on the desigc of
final clarifiers. Information on the set'ling and thickening properties of
the freely suspended biomass from PBSS (Captor and Linpor) would be most
useful.
8
-------
SECTION 4
TECHNOLOGY DESCRIPTION
BRIEF SUMMARY DESCRIPTION
The general class of wastewater treatment processses falling into the
category of porous, biomass support systems (PBSS's) have a singular common
denominator. They all exploit the natural ability of microorganisms commonly
found in wastewater treatment plants to colonize the surfaces and cavities of
inert materials when such materials are contacted with wastewater. In their
current stage of commercial development, PBSS's employ large numbers of small
3 3
{ca. 1.5 cm to 8 cm ) open cell or reticulated polyurethane foam pads* of
very high porosity (ca. 95% to 97%). The pad material has a specific gravity
very near that of water, so by inducing the correct flow patterns in a basin,
the pads can be mixed and circulated within the basin. Figure 1 shown some of
the pads and Figure 2 shows the pads in an aeration basin.
As the pads move through the wastewater, the wastewater also moves through
the pads, bringing nutrients, oxygen and particulate matter into contact with
the biological growth which may be either attached to the pad material or
entrapped within the pores. The most obvious advantages of such a system,
assuming the use of screening devices to prevent pads from leaving the basin,
are:
*In the literature being developed on this subject, the biomass support pads
are also referred to as particles and elements. The letters BSP, standing
for biomass support particles, have sometimes been used.
9
-------
Figure 1. Typical porous, biomass support pads.
Le£t photo - Captor pads, 25mm x 25mm x 12rom
Right photo - Linpor pads, 15mm x 12mm x 12mm
Figure 2. Aeration basin containing
30 percent.
pads at a volume concentration of
10
-------
1. The maintenance of high biomass concentrations within the aeration
basin. The pads have been used successfully at volume concentrations
of up to 40 percent and have been colonized, in field installations,
at biomass concentrations up to 40 g/L (based on pad gross volume),
giving an "equivalent" biomass concentration of 16,000 mg/L, neglecting
any biomass in the free liquid phase. More typically, equivalent
biomass concentrations are expected to be in the 7,000 to 10,000
mg/L range. Under conditions of very low strength wastewater,
biomass levels may decrease to the 3,0C0 to 4,000 mg/L range.
2. The absence of a need for recycling microorganisms to the aeration
basin in order to maintain a given level of biomass.
3. The removal of a thickening constraint from final clarifier design
and in fact, in certain applications, the elimination of the need
for the final clarifier at all.
Other advantages, both potential and/or observed, may be realized in
specific applications of the technology. These will be introduced in subsequent
sections.
To date, two somewhat different paths of development of the basic concept
outlined above have occurred. The Captor Process, marketed by Simon-Hartley,
Ltd. (Ashbrook-Simon-Hartley in the USA), was developed from research hy Dr.
Bernard Atkinson and his colleagues at the Department of Chemical Engineering.
University of Manchester Institute of Science and Technology (UMIST) and is
being commercially developed with technical assistance from the Water Research
Center, Stevenage, U.K. The Linpor Process is being developed by Linde AG in
West Germany from in-house research conducted by Dr. Hans Reimann and Uwe
Fuchs. Linde AG has had technical assistance from the Technical University of
Munich at Garching.
11
-------
THE CAPTOR PROCESS
Basic Process Description
The pads are reticulated (substantially all windows removed) polyurethane
foam, cut to cuboidal form with dimensions of (closely) 25 mm x 25 mm x 12.5
ima. Pore size is controlled through manufacture at dimensions which result in
approximately 11.8 pores per lineal centimeter (30 per inch) with a porosity
of 97 percent. Laboratory studies have indicated that these gross dimensions
and internal characteristics are about optimum for the Captor concept (see
Section 3).
3
The normal design level for pad concentration is AO,000 per m , a volume
concentration of 31 percent. This concentration has been determined to be
near the practical limit for consistent mixing without special provisions in
both pilot- and full-scale trials. In practice to date, pad concentrations
3
ranging all the way from 16,000 to 43,000 per m have been successfully used.
In the basic Captor process, shown in Figure 3, settled primary effluent
enters the aeration basin with its complement of pads. Depending upon the
strength of the wastewater, a certain level of biomass will become established
in and on each pad. Current experience indicates that this may be from 60 mg
to 300 mg per pad, giving an "equivalent" mixed liquor suspended solids
concentration (neglecting non-pad solids) of from 2,400 no 12,000 mg/L (at a
pad concentration of 40/L). Typical target levels are 7,000 to 10,000 mg/L.
To provide the balance between growth on the pads and wasting, pads are
withdrawn from the aeration basin at some pre-determined rate and most of the
accumulated biomass is removed by a pad cleaner, as shown schematically in
Figure 4.
As the pads move up the conveyor (at a rate of about 2,400 per minute on
a 300 mm (12-inch) wide belt), some interstitial water, practically solids-free,
is drained from the pads by gravity. Additional solids-free water is removed
by a pre-squeeze roller at point 3 in Figure 4. The compressible pads are
12
-------
Retaining
Figure 3. Flow schematic - basic Captor process.
Figure 4. Diagram cr Captor pad cleaner.
13
-------
thea squeezed ba>--" be>.'*2en two rollers (point 5 on Figure 4), which removes
the now-concentrated biomass at up to 4 to b percent (dry) solids under favorable
conditiocs. Lower concentrations have been observed in practice, generally
because of weak wastewater and low pa.1 solids levels. The sludge is diverted
from the process by gravity flow or by pumping, and the cleaned pads are
returned to the aeration ba^in to accumulate more biomass.
Pads cannot be allowed to leave the aeration basin with the treated
wastewater, thus some restraint, in the form of screens, is required across
the exit of the basin. The screen openings must be small enough to prevert
passage of pads but large enough to allow passage of most smaller-dimensioned
particles in order to avoid blockage, requi.ing excessive maintenance. In
addition, there needs tc be sufficient screen area to prevent velocities which
are high enough to hold large numbers of pads on the screen surface. Trial-
and-error development has resulted in the adoption of 4-mesh screens (5.13-mm
square apertures), large enough to limit the peak hydraulic loading to less
3 2
than 78 m /m -h based on gross submerged screen area.
Aeration system design has proven to be a much more critical element for
the success of the Captor process than for conventional activated sludge. The
development work has clearly shown the necessity of producing discrete upward
and downward currents which are strong enough to counter the buoyant charac-
teristics usually displayed by the pads. Thus a proper arrangement of diffusers
on the tank bottom is vital for producing the mixing pattern necessary to keep
the pads suspended and moving through the liquid. This has proven even more
critical when fine bubble diffuser systems are employed for aeration and in
aeration basins of high aspect (L/W) ratio. In the latter, spiral roll patterns
have proven superior. For a given successful diffuser configuration, there is
also a lower and an upper limit on air flow rate. Based upon basin floor
2
area, the presently accepted range of air flow rates is 100 L/min-m to 300
2 2
liters/min-m . Less air flow uhan 100 L/min-m allows "dead" spots and
encourages either settling or floating (or both) of large numbers of pads.
2
More air flow than 300 L/min-m creates too much shear on the pads, preventing
accumulation of the optimum amount of biomass per pad for cleaning purposes.
Depth of the basin may be an important element and is currently being investi-
gated as it affects aeration and mixing.
14
-------
Effluent from the basic Captor process (Figure 3) may follow one of
several possible routes. These include:
1. Discharge to a municipal sewer if Captor is being used as industrial
pretreatment process.
2. Discharge to a subsequent treatment process of another type, e.g.,
activated sludge, facultative lagoons or a land-based treatment
system if Captor is being used as an upgrading or load reduction
process.
3. Discharge to another Captor stage if Captor is being used as a
multi-stage complete treatment system.
This is certainly not an all-inclusive list. In addition, modifications
to the basic Captor flow diagram are also possible and may show soue advantages.
Some of the more obvious ones, suggested by other successful processes, are
presented in the next section. At present none of these has been the subject
of any bench-scale or pilot-scale study.
Captor Process Modifications
Effluent Polishing and Nit rification—
'.lie flow schematic for this modification is shown in Figure 5. The
initial, carbonaceous BOD removal process could be of any type, including
Captor. If it was Captor, there simply is not enough background data available
to rule out the need for an interstage clarifer; thus it is shown in the
diagram as typical for any process. Based upon recent WRC results, no clarifier
would be needed after the polishing and nitrification stage, primarily because
of the low-yield coefficients associated with the nitrifying bacteria. The
ability to accumulate a relatively high concentration of nitrifiers in the
pads and to "absorb" an increment of suspended solids could make this
modification competitive with many nitrification - effluent filtration schemes.
15
-------
Cleaner M'*y Not
Figure 5. Effluent polishing and nitrification process using Captor.
Modified Solids Contact—
In the flow diagram shown in Figure 6, waste solids with a fairly high
sludge age arc added to a contact tank along with screened, de-gritted raw
sewage. The effect is an enhanced removal of BOD and suspended solids in the
subsequent clarifier. The clarifier overflow becomes the feed to the Gaptor
process. The clarifier underflow would, of course, be a particularly unstable
sludge and would have to be stabilized by either anaerobic or aerobic digestion.
This modification may have some advantages over the basic Captor flow diagram
for pre-treaUnent of certain strong industrial wastes. The contact and clari-
fication could be accomplished in a clarifier of the flocculating feedwell
design.
16
-------
Sludge Westing To
Flocculation Feechvell
Retaining
Screen
u
Q
• i
Raw
Sewage
Pad
Return
Pad
Delivery
Captor
Effluer^
O
¦o o
¦o
o
¦o o
o
Aeration
System
Sludge To
Stabilization
Figure 6. Modified solids contact process using Captor.
Miscellaneous—
Additional modifications which have been suggested to date include:
1. Use as an adjunct to batch sequencing reactors, and
2. To increase the biomass in mixed anaerobic reactors treating
(predominantly) soluble wastewaters.
THE LINPOR PROCESS
Linpor-C
The pads used in the Linpor process are a bit more heterogeneous in size
and shape and manufactured to less exacting standards than those used in the
Captor process, for reasons which will soon be made clear. The process
17
-------
developers, Linde AG, have the same concerns as Simon-Hartley Ltd., however,
relative to pad wear and pad life expectancy, and a great deal of attention is
paid to this aspect of pad manufacturing. Less attention is given to final
dimensions. An "average" pad is a parallelepiped with dimensions about 12 ma x
12 am x 12 mm. Some pads may have a dimension down to 10 mm and others may have
a dimension up to 17 ran. The pads are of open cell type foam and have a pore
size somewhat smaller than Captor pads, usually running about IS to 20 per
lineal centimeter (38 to 50 per inch). In terms of the volume concentration
of pads used, the Linpor process may employ from 10 to AO percent. Most of
the preseat investigations are being carried out near the upper level of AO
volume percent.
Just as in the Captor process, screens arc needed across the aeration
basin exit to prevent loss of the pads. To this point the two processes
appear almost identical. However, the Linpor process employs no external pad
cleaning device, representing a radical departure from Captor. The Linpor
process relies upon the turbulence and shearing action ic the aeration basin
to control the amount of biomass fixed in and on the pads. Biomass in excess
of this amount exists as freely suspended floe particles. In the Linpor-C
process, shown in Figure 7, where the objective is carbonaceous BOD removal,
these freely suspended floes represent a substaucial amount of both BOD and
suspended solids. Thus a secondary clarifier is necessary if good quality
effluents are to be produced. As long as the clarifier must be included, one
might as well recycle biomass to the aeration basin to further increase the
treatment capacity, and this is part of the Linpor-C process.
Typically, as in Captor, there might be 8 to IS g/L of biomass associated
with the pads in Linpor. In addition, there might be 2.5 to 4 g/L of biomass
associated with the free liquor phase at steady state. Using 8 g/L and 3 g/L,
respectively, as representative values, and a pad concentration of AO percent
by volume, the resulting "effective" mixed liquor suspended solids (EMLSS)
concentration would be (see Tables 1, 2, and 3)
(8 x .A) + (3 x .6) = 5 g/L (5,000 mg/L).
18
-------
Return Sludge Was,e s,udfle
Figure 7. Flow schematic, Linpor - C.
This may not appear to represent a significant improvement over activated
sludge processes (unless both the pad and non-pad solids could be increased,
which experience shows is an attainable goal); however, there are other
important considerations. The most obvious relates to the differences in
secondary clarifer capacity required for an activated sludge process operated
at a MLSS level of 5,000 mg/L as opposed to one operated at 3,000 mg/L (as the
latter value is what the Linpor clarifier will "see"). Also, and this can
only be learned from an analysis of actual Linpor system performance data
(refer to Section 5), the sludge volume index (SVI) of the solids wasted from a
Linpor system is considerably lower than that of a side-by-side activated
sludge plant treating the same wastewater at a comparable food-to-microorganism
(F/M) loading. The combination of the above factors will result in either a
19
-------
reduction in clarifier volume for equivalent effluents or a better Linpor
effluent at equivalent clarifier capacity.
Linpor-N, Effluent Polishing and Nitrification
This idea was actually the first application conceived for the foam pads
during the development work by Linde AG. The flow diagram is shown in Figure
8. Any biological process which can reduce the BOD to the 20 to 30 mg/L range
without oxidizing nitrogen cay be used as the first step. The second step is
Linpor, operated at 20 to 60 volume percent pads and a hydraulic residence
time of 2 to 4 hours. The dissolved oxygen must be maintained in the 4 to 5
mg/L range at a minimum, as the nitrification rate appears to be dependent upon
dissolved oxygen concentration below this level when the nitrifying biomass is
fixed within the pads. No clarifier is needed downstream of the Linpor-N
reactor.
Miscellaneous--
The Linpor process has also been suggested as an adjunct to well-mixed
anaerobic reactors treating predominantly soluble wastes, a pilot system has
been operated on winery wastewaters.
20
-------
Return Sludge (A.S.)
N>
Primary
Effluent
Carbonaceous
BOO Removal
Process
Figure 8. Flow schematic, Linpor - N.
Waste Sludge
Clarifier
Retaining
' Screen
1
j
1
1
1
O 1
% 1
•
Aeration
Nitrified
Effluent
-------
SECTION 5
DEVELOPMENT STATUS
SUMMARY OF RESEARCH FINDINGS
U.K. Experiences - UMIST/Siroon-Hartley Ltd.
Much of the early development work with PBS particles wes performed at
the University of Manchester (UMIST), U.K. Detailed descriptions of this
early work are delineated in Section 6. As a result of this early work, a
patent was applied for and granted in the U.K. (1) and later in the U.S. (2) by
UMIST and Simon Hartley Ltd. for the application of porous supports (called
Captor) for the treatment of municipal and industrial wastes.
Walker and Austin (3) reported on preliminary work performed by Simon-
Hartley Ltd. to assess the potential of PBSS for wastewater treatment. Initial
3
work was performed in 0.6 m pilot columns, 0.5 m in diameter and 3.5 m high.
PBS particles studied included polypropylene toroids with a specific gravity
of 0.91 and porosity 92 percent; polypropylene pads of several thicknesses
with porosities ranging from 95 to 97 percent; and plastic foam pads with
a specific gravity of 1.1 and porosity of 97%. Samples of wastewaters from
two domestic sites were used as influent to the pilot units.
This early research with PBS particles indicated that the lighter-than-
water polypropylene particles had undesirable fluidization properties and were
often found packed at the top or bottom of the column. The more porous
polypropylene particles produced somewhat more favorable fluidization properties
due to the higher gross densities of the particles. Low biomass hold-ups were
obtained, however, with the higher porosity pads due to the small pore sizes
22
-------
(approximately 2.9 g dry weight/L of pad volume versus 8 g/L for the toroids).
The more dense plastic foam pads produced the best "fluidization" properties
of the particles studied and the particles "retained sufficient biomass of
high activity." No data on particle biomass hold-up were actually presented,
however. The study also indicated that the biomass mineralized within the pad
structure over time, aud it was felt that in order to maintain an effective
biomass activity, cleaning of the PBS particles was necessary. Squeezing of
the plastic foam pads produced a sludge of about U to 6 percent.
Extensive studies were performed by Simon-Hartley to perfect the properties
of the PBS particle and to evaluate important process control variables (A).
Detailed experimental studies were not published in the scientific literature,
however. A reticulated polyurethane foam particle, 25 mm x 25 mm x 12.5 mm in
size with 97% void volume was eventually found to possess excellent properties
of wear, "fluidization," biomass hold-up, and "squeezability." In-house
studies led to preliminary recommendations relative to aeration and mixing
requirements, oxygen transfer, screening, cleaning, and performance capabilities.
These parameters were subsequently used to design pilot studies at Freehold,
U.K., which are reported in Section 5. In parallel with the pilot work at
Freehold, the Water Research Centre is currently performing pilot studies to
better evaluate particle integrity, mixing requirements and biomass properties.
These results are not yet available for review.
West Germany - Linde AG
The use of PBS particles for wastewater treatment was apparently develop -.d
somewhat parallel to the UMIST/Simon-Hartley Ltd. studies by Linde AG/Technical
University of Munich (5,6,7,8). Linde AG also adopted an open cell polyurethane
foam which was somewhat more dense with smaller pores (38 to 50 pores/inch)
than the Captor particles. The particle sizes used in both laboratory and
field studies range from 10 to 15 mm in size, normally in the shape of a cube
or parallelepiped.
The first application of PBS particles by Linde AG was for the purpose of
nitrification (6,7). These nitrification experiments were conducted with a 1
23
-------
3
m reactor filled with PBS cubes to 37% by volume. No clarifier or sludge
return were used in these studies. Intluent to tbe unit was secondary
sffluent from a domestic wastewater treatment process. The results of this
study are presented in Table 1. Effluent solids were always under 20 mg/L.
Within four weeks of start-up, a high degree of nitrification had occurred.
Over the experimental study of "some" months, no deterioration of particle
biomass was noted. An equilibrium apparently was developed between growth and
attrition in this system. Other pilot tests conducted at still another domestic
treatment plant with lower PBS particle concentrations (24 to 35%) and higher
influent TKN (24 to 60 mg/L N) at a hydraulic retention time of 1.1 hours
produced about 80% TKN oxidation. No details of effluent solids, biomass
hold-up or system stability were presented (5,6).
TABLE 1. PILOT PLANT RESULTS - LINPOR (6)
Experimental phases
Parameter
Units
1
2
3
4
Influent
BOD
mg/L
16
16
27
9
TKN
mg/L
34.8
25.2
23.5
17.6
no3-n
mg/L
0.7
0.7
0.5
1.4
Effluent
BOD
mg/L
7
6
12
8
TKN
mg/L
2
3.8
9.6
1.2
NO -N
mg/L
28.5
17.4
9.8
19.0
SSJ
mg/L
8
16.7
16
10
Hyd. ret. time
hrs
2.5
1.7
1.2
3.2
Temp, reactor
°C
14-17
15-19
15-18
16-20
DO reactor
mg/L
3-7
3-5
3-5
4-6
PBSP biomass
g/L
8.7
10.1
14.6
12.5
EMLSS
g/L
3.2
3.8
5.4
4.6
TKN load
kg/m3 d
0.33
0.36
0.45
0.13
24
-------
Pilot planl studies on nitrification are also being conducted at the
present time at Abwafserzweckverband Munchen-Ost at Poing, West Germany by
Linde AG (7). Effective nitrification is being achieved in this pilot unit,
which receives secondary effluent from the domestic wastewater treatment
plant. Results of pilot plant experiments at this facility to date are
presented in Table 2. Six experimental studies were conducted using Linpor pads
at a volumetric concentration of 30 percent. NH.-N loadings ranged from 0.37
3
to 0.65 kg/m d and PBSP biomass concentrations ranged from 13 to 15.7 g/L.
Of substantial interest in these studies was the rinding that dissolved oxygen
concentrations played an important role in the nitrification process, likely
due to diffusional limitations. DO values above 5 mg/L were required to
achieve effective nitrification.
TABLE 2. PILOT PLANT RESULTS - LINPOR H (7)
Experimental phases
Parameter
Units
1
2
3
4
5
6
Influent
BOD
NH.-N
4
mg/L
mg/L
21
56
23
69
21
63
25
65
17
73
16
67
Effluent
BOD
nh4-n
mg/L
mg/L
9
4
13
19
17
35
17
32
12
7
10
15
Hyd. ret. time
hrs
3.5
3.7
4.1
3.3
3.5
2.5
Temp, reactor
°C
19
16
20
18
20
20
DO reactor
mg/L
5.3
5.6
2.2
3.8
5.9
5.7
PBSP biomass
g/L
14
13
15.7
15.7
14.7
14.3
EMLSS
g/L
4.2
3.9
4.7
4.7
4.4
4.3
NH.-N load
Volumetric
Biomass
kg/m^ d
kg/kg EMLSS
.39
d .09
.44
.11
.37
.08
.47
.10
.50
.11
.6!
.1!
25
-------
The second general type of experiments conducted by these investigators
have dealt with PBS particle addition to aeration tanks for the purpose of
intensifying biomass activity and improving effluent settling properties. PBS
particles were employed in combination with suspended flocculent growths in
aeration basins. Clarification and sludge return were employed as in conven-
tional activated sludge processes (Linpor-C).
Hegemann and Englmann (6) reported on several pilot plant studies at
3
domestic treatment facilities in West Germany. A 1-m pilot plant filled with
40 percent by volume of PBS particles was operated at one site (F). The
results of these different pilot test runs with increasing volumetric loading
are presented in Table 3 along with average results of full-scale plant
operation without PBS particles. Two important features of PBS particles can
be observed from these tests. First, the addition of the PBS particle to the
aeration tank substantially increased the reactor biomass concentration. This
effectively reduced the F/M ratio for a given volumetric loading. Second, the
addition of the PBS particles dramatically reduced the SVI of the reactor
suspended solids. Both features contributed to an improved effluent.
TABLE 3. LINPOR-C TEST RESULTS AT SITE F (6,8,9)
Test Volumetric MLSS PBSP EMLSS F/M bVI B0D5 eff.
loa
-------
Titese authors found similar results at another domestic wastewater
treatment facility. Parallel tests in 20-L pilot units with and without PBS
particles indicated lower F/M, lower SVI and superior effluent BOD^ in the
plant containing 40 percent by volume PBS cubes. They reported PBS particle
biomass hold-up values of 15 g/L in these tests (6,9).
In another test series, side-by-side experiments were conducted over a
12-month period to evaluate the long-term stability and effectiveness of
PBSS's. The pilot plant used at city M employed an aeration tank with a
3
volume of 6 m . The tank was filled with 30% by volume of PBS cubes. The
full-scale plant did not receive PBS particles but was operated at approximately
the same volumetric loading. Results of this study appear in Table 4. As was
fouad in the previous tests, the addition of PBS particles effectively reduced
F/M for the same volumetric load and reduced SVI as compared with the full-scale
plant.
TABLE 4. LINPOR-C TEST RESULTS AT SITE M (8,9)
Test Volumetric MLSS PBSP EMLSS F/M SVI BOD eff.
loaij biomass
kg/m d mg/L mg/L mg/L kg BOD/kg d ml/g mg/L
Pilot tests
(PBSP)*
2b 2.34 4640 9,950 6180 .41 148 15
3 3.07 4290 13,210 6830 .45 128 13
Full-scale
plant
(no PBSP)
2b 2.06 2520 — 2520 .88 222 24
3 2.12 2160 — 2160 .98 1.88 10
*Linpor cubes, 30% by volume in pilot tests.
27
-------
It is significant to note the observations of these West Germany inverti-
gators relative to the PBS particle biomass. Hegemann and Erglmann (6)
reported that an equilibrium was reached after a period of several weeks
between biomass growth and attrition. Furthermore, they indicated, as did
Reimann (8), that no loss in biomass activity, as measured by specific oxygen
uptake rate, was noted up to 12 months of continuous operation without pad
removal. Our observations of two full-scale plants in operation for many
months (Freising and Munich 1) indicated that PBS particle biomass was fresh
and dark brown throughout the particle volume. There were no signs of
anaerobosis or mineralization within the pad stiucture.
Although comparisons between pilot-scale and full-scale performance are
not directly one-to-one and considerably more full-scale experience is warranted,
these early West Germany studies do demonstrate two important characteristics
of PBSS's which make them likely candidates for treatment plant upgrading.
Currently, there are at least two full-s^le studies using Linpor-C being
conducted in West Germany. At Freising, Linpor particles (25% by volume) have
been added to one tank for the primary purpose of reducing SVI. Currently,
pads are now being added to remaining tanks. Results of this work will be
available in 1985. Research at Munich I, which is being supported by the town
and the federal government, is being conducted in both pilot plants and full-
scale tanks. Results of this work will be available in the near future.
SUMMARY GF PILOT- AND FULL-SCALE FACILITIES
Experience in the United States
In the .U.S., Captor experience is limited to two pilot plants, each
operated for about four months, and one small-scale industrial system. While
there are some valuable lessons to be learned from the U.S. experience, it
would be grossly unfair to judge the potential of the Captor process on the
basis of this experience alone. The p?1ot plant studies were not particularly
well conceived or carried out. In neither case was there a well-defined
control, as for example in the Freehold, U.K. studies, to compare Captor
28
-------
against. In one case (Marion, Illinois) the state regulatory agency informed
the city and its consultant that Captor, being an untried process in the U.S.,
would not be considered for plant upgrading regardless of its performance at
pilot-scale. This disheartening intelligence cane just at about the time of
system startup. It naturally had an adverse effect on the performance of the
study, especially the quality of the analytical work.
At Downingtown, Pennsylvania, the Captor process was thwarted by the
nacure of the wastewater load, which concomitantly also caused severe problems
for the existing activated sludge system. Extremely high infiltration/inflow
and occasional slugs of toxic wastes, together with a peculiar periodic loading
of starchy materials from a large commercial bakery, was sufficient to discourage
the consultant charged with Captor evaluation.
The industrial installation is achieving excellent BOO and COD removals.
However, it would be a mistake to attempt to evaluate this installation as a
"pure" Captor process. Because of a poor initial aeration system design, most
of the experience at this plant has been gained with the majority of the pads
collected in a large floating raft at the surface of the aeration basin.
Consequently, the bulk of the biomass has been in suspension rather than in
the pads and was recirculated to the reactor from a conventional clarifier.
The operation has been more akin to activated sludge than Captor.
Marion, Illinois—
This pilot plant consists of a Captor section in series with complete-mix
3 3
activated sludge (CMAS). Both basins have a volume of 14.5 m (512 ft );
however, the hydraulic detention time of the CMAS section is 3.2 hrs, three
times that of the Captor section. A constant feed rate of 3.78 L/sec (60 gpm)
(determined by bucket and stor-watch measurement) is fed to the plant. Fol-
lowing the Captor section, 2.52 L/sec (40 gpm) is diverted to the effluent
sump and 1.26 L/sec (23 gpm) receives second-stage biological treatment in the
activated sludge section. This arrangement was selected because it simulated
closely the design detention times which would have occurred in the Marion
wastewater treatment plant if it was converted to a Captor-CMAS process
29
-------
flow scheme. The pilot plant was brought on-line in March of 1984 using
primary clarifier effluent as feed. As sodium alurainate and polymer arc added
to the raw sewage to produce additional removals of suspended solids and BOD
(to decrease the loadings to a faltering RBC system) and the Marion sewerage
system receives substantial infiltration and inflow during the spring season,
the strength of the pilot-plant feed was quite low during the first three
months of operation. Extensive sampling and analysis was carried out during
the first two months of operation. The Marion POTW is manned 24 hours per day
using three shifts. Originally, the composite samples (see Figure 9 for
sampling locations) were collected and analyzed by the second (afternoon)
shift, whereas the first (day) and third (night) shifts collected grab samples
only for analysis. During this initial testing period, observations made by
the superintendent caused him to alter this schedule so that the composites
were collected and analyzed by the first shift. Crab samples were collected
and analyzed by the second and third shifts. Soon after start-up, a decision
was made by the Illinois EPA that the Captor process would not be allowed to
be used at Marion regardless of the outcome of the pilot-scale test program.
This decision was evidently prompted by the severe problems experienced with
the RBC system at Marion leading to a reluctance on the part of the state to
approve any other innovative technology to replace one which had failed to
live up to performance claims. At this juncture, testing of the pilot-plant
was cut back to one shift per day (day shift). In addition to these complica-
tions, it should be mentioned that interviews with the laboratory staff showed
that several other problems may exist relative to the data obtained from the
pilot-plant operation. Although all the plant personnel involved in the
testing program were "schooled" in proper procedures by a qualified professional
chemist prior to the start-up of the pilot plant, there were inconsistencies
noted among analysts on procedures and calculations, especially with regard to
the BOD tests. There were also continuing problems noted with the quality of
dilution water and the seeding techniques employed. All things considered,
the accuracy of the BOD data is probably somewhat worse than that indicated as
reasonable in Standard Methods (±15%), and numerous outright anomalies exist.
For example, on many occasions total BOD values were substantially less than
the soluble BOD value on the same composite sample. Where such obvious problems
arose, the data for that day were not included in the "weekly" averages.
30
-------
Figure 9. Marion, Illinois schematic.
Here, as in the case of Downiagton, "weekly" averages means an average of
seven consecutive usable (believable) data points. The seven days are not
necessarily seven consecutive days, as some data are missing and some are
purposely omitted. Within these rather severe limitations, Figures 10 and 11
show the performance of the Captor section and the overall plant relative to
total and soluble BOD removal.
As shown on the performance graphs, the pilot plant had a "full" (design)
complement of pads (40/L) for only two of the eight weeks for which usable
data exist. Pad 6olids during the entire period of operation were very low,
ranging from 45 to 70 iug per pad. During the last few weeks of the study,
with the wastewater temperature averaging 18°C (65°F), the pilot-plant per-
formance was fairly consistent and is summarized in Table 5. During this
period the F/M load to the Captor section ranged from 0.6 to 0.9 kg BOD/kg
MLSS'd and that to the activated sludge section ranged from 0.09 to 0.13. The
overall system effluent was excellent, averaging less than 5 mg/L of BOD and
31
-------
Figure 10. Marion, Illinois pilot plant results
Total BOD, 7-day averages.
WMfc Of Opwriion
Figure 11. Marion, Illinois pilot plant results
Soluble BOD, 7-day averages.
32
-------
TABLE 5. PERFORMANCE OF THE MARION, ILLINOIS, PILOT PLANT WITH A PAD
CONCENTRATION OF 40/L (16 DAYS)
Effluent
Average percent
removals
Tot. BOD
Soluble BOD
SS
nh3n
Captor
49
54
-36
29
Final
94
94
92
100
suspended solids and zero ammonia. Unfortunately, there is no reliable control
available to compare these data against, as the full-scale plant was suffering
severe mechanical problems with the RBC installation during this period.
The pad cleaner was operated for one hour per day during this part of the
study, providing an (approximate) pad cleaning rate of 20 to 25 percent per
day.
Dovmingtown, Pennsylvania—
A Captor pilot plant was installed and tested at the Downingtown Regional
Water Pollution Control Center to see whether the Captor ^ricess could be used
to economically upgrade trie existing activated sludge plant. This plant has
had a history of filamentous bulking, washout of MLSS during high rainfall
events, periodic excursions in the effluent SOD (presumed due to industrial
toxic discharges*), and inability to achieve consistent nitrification. The
original pilot plant was 9.76 m (32 ft) in length with a cross-section 2.44 m
by 2.44 m (8 ft x 8 ft). During start-up, problems were observed with pad
distribution along the length of the tank. By consensus it was decided to
reduce the aspect ratio from 4:1 to 3:1 by blocking off the first 2.44 m (8
ft) of tank length, leaving an effective length of 7.32 m (24 ft). It is
~Operating personnel are convinced that toxic material was being dumped
at night, probably directly into manholes. When they let it be known that
they were out to catch the dumperk the problem stopped immediately.
33
-------
generally agreed that the pad distribution iinproved slightly upon making this
modification, although no clear understanding of the relationship between
aspect ratio and pad distribution seems to exist at present.
The aeration system originally installed consisted of 72 ceramic dome
diffusers installed in four lines down one side of the tank to produce a spiral
roll pattern. This arrangement was subsequ: utly modified by removing the line
nearest the tank centerline, leaving 54 diffusers. This produced a more
effective mixing pattern.
Other characteristics of the pilot plant of interest were as follows:
1. The feed was primary clarifier effluent at a constant flow of
12.6 L/sec (200 gpm).
2. Maximum aeration capacity was 118 L/sec(250 cfm), but was generally
held between 38 to 82 L/sec(80 to 175 cfm) using a bleed-off valve.
3. The pad cleaner was located at the effluent end of the tank and
discharged cleaned pads to the feed trough, which carried them to
the influent end. Pad cleaning occurred once per day for one hour
during much of the four-month (approx.) study. During part of March
the pad cleaner was operated for four and later for eight hours per
day. Rough counts of the pad cleaning rate produced an average
value of 150,000 per hour. This figure would translate into a
cleaning rate of about 8-9 percent for one hour per day, 34-35
percent for four hours per day and 69-70 percent for eight hours per
day.
4. During the course of the study the wastewater temperature ranged
from 10.5°C to 12.5°C.
5. Pad distribution along the length of the tank during the study
varied, but generally could be expected \.o be about 30/L in the
first third, 40/L in the second third and 50/L in the final third of
the basin. The overall average was 42/L.
34
-------
6. The pad solids during the study varied between 32 mg/pad and 127
mg/pad, with an average of 71 mg/pad, resulting in an "effective"
MLSS of only 2990 mg/L. Microscopic examination and oxygen consump-
tion measurements showed the pad biomass to be quite inactive at the
beginning of each week due to very low plant loadings on weekends.
Activity increased through the week, reaching a peak by Friday.
Exceptions were noted when high rainfall events occurred to dilute
the incoming sewage.
7. Effluent suspended solids from the Captor plant averaged 74 mg/L,
compared to an average influent suspended solids of 66 mg/L. This
loss of solids is obviously a major factor in the poor total BOD
removals achieved by this plant. A rather low waste sludge concen-
tration was achieved by the pad cleaners; averaging about 0.95
percent solids with a maximum recorded value of 1.32 percent.*
Weekly** average values of total and soluble BOD into and out of the
Captor pilot plant are plotted in Figures 12 and 13. Comparing the data from
the Captor section alone to that from Marion, IL, shows a worse performance at
Downingtown at about the same F/M loading and strength of feed. The only
obvious contributing factor is the lower wastewater temperature at Downingtown.
There were, however, other possible reasons for the mediocre performance. The
operation was plagued with some purely mechanical problems which tended to
erode solids from the pads. In particular it was noted that the combination
of placement of the pad retaining screen and aeration system geometry was such
that large numbers of pads impinged heavily upon the screen, resulting in
physical abrasion of pad solids and subsequent loss through the screen before
they could be picked up by other pads.
* Recent data obtained with the same pilot plant at Moundsville, WV, after
several adjustments to the pad cleaner, have shown that thicker waste sludges
can be obtained, even with low pad solids concentrations.
**The "weeks" actually consist of seven consecutive BOD values. They were not
always continuous and did not always consist of data from seven consecutive
days, especially near the end of the study, when data were taken about every
other day.
35
-------
"»U*T- loT-ll t* "
- • Cipttr cmu«f«
2b%
0 I I I I I I ¦ ' I 1 ' ' 1 1 ' • •
o 1 4 6 » 10 U M '•
•MaOIOmMi
Figure 12. Downingtown, Pennsylvania pilot plant results
Total BOD, 7-day averages.
I - Primary Effluent (Feed)
L - Captor Effluont
Average F/M - 0.88
Temperature • 10.5-12.5 °C
Average Removal - 48%
-I—
10
—1—
14
~r
18
12
16
Weeks of Operation
Figure 13. Downingtown, Pennsylvania pilot plant results
Soluble BOD, 7-day averages.
36
-------
Possibly more important, however, it appears that the unique characteris-
tics of the waste loading pattern simply did not allow the establishment of a
reasonable steady state with regard to system operating parameters. It is
certain that no special steps were taken to attempt to operate the Captor
process in response to the varying loading conditions. In fact, to do so
would have been presumptuous on the part of the operating staff, as they had no
established background of operating experience with Captor to guide them. To
take these thoughts a step .further, in examining all the development work done
to date, there has been so much effort required to solve basic process problems,
including hardware improvement, and just getting Captor to "work", that there
has been very little time available to the process developers to devote to the
fine points of both Captor and overall process optimization. One must realize
that the second does not necessarily follow the first when Captor is used as
the first stage of a multi-stage biological treatment system.
The downstream process, whatever it is, will probably behave in an
optimum way when it receives its "design" load in the absence of substantial
pertubations. If an upstream Captor process of finite removal capability, in
kilograms of BOD, for example, was always operated to achieve 'hat removal,
incoming pertubations would be reduced in amplitude and increased ia length,
but would still exist. These would be passed on to create stress in the
downstream process. If, on the other hand, the Captor process was designed
and operated to pass on a more constant load, the overall process result might
improve even though the Captor performance was variable and not always at its
optimum level. The ability to implement this sort of process operating strategy
is an inherent property of Captor, primarily because of the ability to clean
pads at rates varying from zero to 100 percent per day (or higher). The
wastewater treatment field has not previously had a process whereby the growth
rate of the system could be so quickly and conveniently manipulated. Iu
future research and development work on the Captor process, a lot of thought
must go into trying to harness this capability.
The application of these somewhat philosophical remarks to Downingtown is
straightforward. If Captor had been adopted at Downingtown, it was to have
been as a first stage followed by a nitrifying activated sludge system. In the
absence of other complications such as high infiltration/inflow or toxic
37
-------
materials entering the system, the BOD loading occurred as a fairly predictable
weekly cyclic pattern. Because of the lack of industrial activity (particularly
a large commercial bakery) on weekends, the BOD loads to the plant were quite
low through Friday night, Saturday, Sunday and into mid-morning on Monday.
During these periods loads were received at a rate of 500 to 2000 kg BOD/d
(1,100 to A,400 lb/d). During the remainder of the week loads were more
typically 2,900 to 5,000 kg BOD/d (6,400 to 11,000 lb/d). How should the
overall system be optimized? We do not have the kind of data base yet which
allows us to answer this precisely. However, the following scenario seems
intuitively satisfactory.
1. During the low load periods, clean pads at a high frequency and
waste sludge to the downstream activated sludge system.
2. As the load increases, decrease the rate of pad cleaning to allow
pad biomass to accumulate to near its highest steady-state level.
Gradually switch the point of sludge wasting from the downstream
process to out of the system.
3. When the desired pad solid- level is '.tained, increase the rate of pad
cleaning while continuing to waste solids entirely out of the system.
Employing an operating strategy of this type, it should be possible to
keep a fairly steady load on the downstream nitrification system, maintain
maximum activity in the nitrification system, equalize aeration requirements
and equalize solids loadings from the nitrifying activated sludge tanks to
their secondary clarifiers. All these factors should result in improved
overall plant performance, but at the expense of performance rating for the
Captor portion of the process.
Industrial System—
This system was designed for a low flow (2.6 L/sec [42 gpm]) of strong
industral wastewater from the manufacture of plastic specialty products. The
chemical oxygen demand (COD) of the wastewater ranges from 1,000 to 2,500 mg/L
after chemical pre-treatment, settling and flow equalization. Tiie major
characteristics of the Captor system are summarized in Table 6.
58
-------
TABLE 6. CHARACTERISTICS OF INDUSTRIAL CAPTOR TREATMENT SYSTEM
Reactor (circular)
Water depth
Hydraulic detention time
Average flow
Blower capacity
Normal aeration rate to control
D.O. @ 2 rag./L
Pad concentration - from 8 Feb.
Diameter
4.1 m (lj.5 ft)
4.6 m (15 ft)
1C.7 h
1.6 L/sec (25 gpai)
89.6 L/sec (190 cfm)
40 to 71 L/sec
(85 to 150 cfm)
1984
40 per liter
The system is equipped with a final clarifier because of the amount of
growth anticipated during the t reatment of such a high-strength wastewater.
During system start-up, changeb were made to the piping to allow the recycle
of settled biological sludge back to the Captor reactor. During the period
for which most of the analytical data are available, sludge was being recycled
and pad mixing conditions within the reactor were quite poor, with large
quantities of pads collected together at the surface in "rafts." Especially
bad episodes of pad flotation seemed to correspond to high reactor temperatures,
especially those in excess of 32°C (90°F). However, the problem has been
somewhat chronic and is believed to be associated primarily with the design of
the aeration system. The original design consisted of 12 radial laterals
emanating from a centrally located drop pipe, each lateral at 30° from its
neighboring laterals. Each lateral had three coarse bubble diffusers spaced
at equal distances along its length. This pattern did not provide for the
discrete up-and-down fluid movement which experience has now shown is necessary
for adequate pad mixing. An attempt to create a "roll" in the reactor by
blinding off five adjacent laterals did not succeed in solving this problem;
therefore, the decision was made to completely redesign the aeration system.
In early September, the aeration system was completely overhauled. Twelve
large porous plastic plate diffusers, each equipped with a control valve, were
installed around the aeration tank perimeter'. At an air flow rate of 51 L/scc
(108 cfm) and a pad concentration of 38/L, the pad circulation appears to be
excellent and the floating pad rafts are gone.
39
-------
Host of the performance data for the system is in terms of COD values.
BOD/COD correlations have been developed for the influent and effluent and are
checked often eaough for consistency that they can be used with reasonable
confidence to estimate performance relative to BOD removal. These BOD/COD
ratios are 0.36 for the feed and 0.14 for the effluent. For most of the
performance history available on this system, the pad solids averaged about
100 mg/pad and the MISS averaged about 2,400 mg/L. At 40 pads/L, this would
translate to an EMLSS of 5,630 mg/L except that there vas limited contact
between the pad biomass and the wastewater. Most of the treatment, except on
occasions which are not well documented, was being accomplished by the suspended
biomass. A substantial level of treatment even under these conditions was
being achieved, but it is clear from an inspection of Tigiire ]4 that once the
pad mixing problem at this plant is overcome, a very high level of performance
may be expected. Beyond that, little cat be said about this system as an
example of Captor experience.
1 2 3 4 B 6 7 e 9 10 11 12 13 14 IS >6 17 18
WMk» Of Operation
Figure 14. Performance of industrial Captor 3yetem.
40
-------
Experience in the United Kin&dom
There is a much longer history of Captor experience in the U.K., most
(but not all) of which has been subject to far better controls than in the U.S.
Some of the work has been at rather small scale, and this work will be referred
to in Section 4 of this assessment. In the present section, only two systems
will be reviewed.
Pretreatment of Bakery Wastewater--
The first system designed from the earlier bench- and pilot-scale work
was an industrial pretreatment system installed at a commercial bakery. The
flow diagram is shown in Figure 15. Discharge of the effluent is to a municipal
sewer.
Effluent To
Sewer
Feed
Blower
Sump
Figure 15. Captor system for bakery wastewater pre-treatment.
41
-------
3 3
The circular reactor has a volume of only 12.5 m (440 ft ). Performance
stcndards for this plant are < 550 uig/L for COD and < 225 nig/L for BOD. A&
with many small industries, a lot of time and attention has not been lavished
on this facility, including the monitoring function. However, a reasonable
amount of reliable analyses were obtained during and just after system break-in.
The data obtained after achieving a steady-state condition are sumnarized in
Table 7 (4).
Unfortunately, there are no complete records on the day-to-day operation
of this facility. Thus, critical quertions as to pad bioraass concentration,
pad cleaning rates, waste sludge production, oxygen utilization, and especia Ily
costs of operation cannot be answered. In this regard it is perhaos inappro-
priate to include this example in the present document. However, on balance
the example does provide some additional background data on the process and
the reader can judge how effective this performance is relative to available
alternatives for comparable degrees of pretreatment.
TABLE 7. CAPTOR PRETREATMENT SYSTEM FOR A BAKERY - POST BREAK-IN PERFORMANCE
COD tng/L BOD mg/L SS mg/L pH
Feed
Final
Feed
Final
Feed
Final
Feed
Final
Maximum
3540
1000
2025
510
926
918
6.8
7.0
Mininum
600
336
350
1C1
340
440
4.2
6.0
Average
2084
554
1246
255
720
653
5.2
6.7
Average %
Removal
74
80
Average volumetric load 4.3 kg BOD/m -d (270 lb BOD/1000 ft -d)
Average hydraulic retention time 6.9 hrs.
BOD and COD values based on unsettled samples.
42
-------
Freehold Water Reclamation Wotks--
The Freehold plant, located near Stourbridge, West Midlands, was an
activated sludge system which needed upgrading to achieve year-round nitrifi-
cation. The plant consisted of five lanes of basins aerated by a fine bubble
diffuser system. To provide the proper controls for the study, one basin was
removed from service and the flow was equally divided among the other four
basins. Two of these were kept as activated sludge basins and two were con-
verted to two -stage units with Captor in the first quarter and activated
sludge in the final three-quarters. The two systems discharged to two separate
clarifiers so that effluent coinparisons between the Captor-activated sludge
and conventional activated sludge systems are possible. The basic flow schematic
following the conversion of two lanes is shown in Figure 16.
Figure 16. Flow schematic to evaluate Captor vs. activated sludge
at Freehold.
43
-------
The Freehold study, begun in September of 1982 and originally planned
for two years' duration, represents the most comprehensive investigation of the
Captor process as the initial stage of a multi-stage biological system ever
attempted. The study was designed for both development and evaluation.
Before presenting and discussing the data accumulated thus tar on this system,
it seems appropriate to deal with some preliminaries which assist in putting
the design of the conversion and the data into a proper perspective. The
basic design data are presented first in Tables 8, 9, and 10.
TABLE 8. DESIGN LOADING CONDITIONS FREEHOLD WATER RECLAMATION WORKS
Flow, m?d ^^5 Load, kg/lane-d
Avg. annual 18,700 144 673
Avg. winter 20,000 148 740
Avg. February 20,000 174 870
Max. day — — 1261
TABLE 9. DESIGN DETAILS
Basin geometry 61 m x 3.6 n x 3.0 i SWD
Total volume/basiu 659
First 25% (15.5 m) converted to Captor
Captor volume 165 m «
Activated sludge volume 494 m _
Captor support concentration 40,000/m
Predicted biomass/pad 200 mg
Predicted MLSS in activated sludge section 3,200 mg/L
44
-------
TABLE 10. PREDICTED LOADINGS AND PERFORMANCE OF CAPTOR/ACTIVATED SLUDGE
SYSTEM - FREEHOLD WATER RECLAMATION' WORKS
CapLor BOD load
Period kg/m -d kg/kg-d
Predicted BOD"*
removal in
Act'd sludge
load
kg/kd-d
Predicted
NH--N in
final eff., mg/L
Captor, %
Winter 4.5
Sumner 3.5
Annua1 4.0
0.56
0.44
0.50
70
80
75
0.14
0.073
0.08
10
1
5
*See comments on the design equation which follows.
Fir&t, some comments on the physical facilities and conversion design.
The choice of Freehold appears to have been a good one, even though screw
pumps had to be added after the primary clarifiers in order to guarantee equal
flow distribution to the twp comparative systems. The plant was not able to
produce consistent nitrification under existing loading conditions, but with
the addition of some additional carbonaceous BOD removal at the front end,
could be expected to show a dramatic improvement in nitrification. The ability
to keep the sludges from the two sides separate was an absolute necessity for
a successful study and, although difficult, this was accomplished. Given the
unique demands of the sti'dy, one aspect of the plant which did not receive
adequate a'tertion was the primary treatment. Throughout the course of the
study, the special hardware required to make Captor function—pad cleaners and
retainer screens—were plagued by plastics, especially plastic strips, which
vcie j continual problem. The plastic materials were never noted as a problem
before the conversion, as they did not interfere with aeration basin operation
undjr an activated sludge mode. Several attempts were made to exclude these
materia.s using fine screens in the feed chaunils, but the screens clogged too
rapidly and created a continual maintenance problem. Work to solve this
problem continues.
It is possible, ev*n probable, that the problems with oil and plastics at
Freehold, although not unique to that plant, are intensified hy the design and
operation of their particular primary clarifier system. First, the inlet
45
-------
condition seemed to encourage, rather than discourage, t>hort-circuiting.
Next, the sludge and scum scrapers were operated intermittently and at perhaps
too low a frequency (they did not operate during the three-hour period of our
visit). Last, and probably most important, there is no scum baffle downstream
of the rotating scum pipe. Thus, it is very easy for floating material to
ride underneath the scum pipe and there is nothing to stop the material from
exiting the basin once it does pass this point.
Turning to Table 10, the loadings and performance predicted for the
Captor-activated sludge side of the plant, there is little doubt that the BOD
removals predicted for the Captor section are overly optimistic. The BOD
removal correlation developed by Simon-Hartley, Ltd. from previous pilot-scale
work was
Percent removal = 100 exp (-.67 F)
M
and has appeared in several of their internal publications (4,10). However,
in the development cf this correlation, the effluent BOD values used were
based upon the BOD remaining after a one-hour period of settling (11) and are
thus not indicative of Captor operation as the first stage of a multi-stage
process. Without benefit of an intermediate clarifier, the correlation
overestimates the BOD removal by a variable, but unknown, amount. Unfortunately,
no corrections were made to the removals calculated for the Freehold conversion
to correct for the suspended solids carried into the activated sludge section
of the aeration basin from Captor. Thus the actual loadings to this section
would be higher than those predicted in Table 10. This error is partially
compensated for by the fact that a part of the BOD load to the activated
sludge is respiring biornass; that is, some of the feed, F, is also biomass, M.
This presents an interesting problem in the proper interpretation of F/M in a
multi-stage process, a problem which does not seem to have been addressed
before in the literature. It do?s seem clear that F/M does not have the same
meaning here as it does for a process where there is only a small inoculum
contained in "F." Thus F/M should probably not be the basis of performance
calculations for the second stage of treatment when Captor is used as a first
stage. Nor does sludge age appear to offer an easy way out of this dilemma,
46
-------
primarily because part ot the sludge inventory has a low probability of being
removed from the pads during cleaning because of its location deep inside the
pads. If this material deep within the pads is the most mineralized biomass
and low in activity so that the part that gets routinely wasted is of a younger
physiological age, calculation of a true sludge age would be difficult.
Without some research directed towards defining these differences, there is no
obvious way of distinguishing "old" from "young" biomass and the concept of
sludge age suffers consequently. If highly effective pad cleaners become the
norm, the above point becomes le«s important.
The available oota base from the Freehold investigation covers the period
from early November, 1983 until March, 1985. Because of the purely mechanical
problems previously referred to, the Captor sections in both converted lanes
did not have their design complement of pads installed until March, 1985. The
first lane (CI) has operated at 28 pads per liter while the second (C2) has
operated with only 16/L. At present, both lanes have pad concentrations of 40
per liter, with good pad mixing conditions being experienced. Progress reports
summarizing the raw data and the operating log have been written by the Joint
Project Manager. These reports have covered periods of from one to three
months. The data obtained are extensive and quality control of the analytical
techniques is believed to be excellent. Along with BOD and suspended solids
data for all key points in the system(s), there are extensive data on sewage
flow, air flow, and pad biomass, and limited data on pad cleaning rates and
specific respiration rates for both pad biomass and suspended biomass.
Looking first at the overall aspects of the study, Figures 17 and 18
provide comparisons of the monthly average final effluent BOD and suspended
solids for the side-by-side systems.
The differences during the first six months are dramatic, especially
considering all the mechanical problems detailed in the progress reports as
the pad cleaners went through several stages of development and the investiga-
tors learned about balancing the mixing and aeration. The differences begin
to disappear through the next three months as the effluents for both systems
are seen to continually improve. Because the first indications of nitrification
in the Captor-activated sludge system are observed in mid-July, the increasing
47
-------
40 -
30 -
0
O 20 •
10 •
B - Control (Activated Sludge)
£ - C«ptor (Activated Sludge)
0 1 | | »' I I I I I I
No* Dec Jin Frto Mar Ap» May June July
1983 1984
Month
Figure 17. Freehold wr plant - monthly average effluent BOD.
40
s
5
30 -
20
10 ¦
B ¦ Control (Activated Sludge)
A - Captor (Activated Sludge)
I I I I 1 I I I I
Nov Doe Jan Feb Mar Apr May June July
1083 1884
Month
Figure 18. Freehold wr plant - monthly average effluent suspended solids.
48
-------
wastewater temperature is likely the primary cause for the improvement noted
in both systems; activated sludge benefiting equally with the multi-stage
system. As for the significant differences noted from November through April,
there are several possible explanations. One relates strictly to the differences
in the settling characteristics of the solids produced during the colder
periods of operation. Data from the Linpor work (sec Section 5) and also from
a recent survey of coupled trickling filter-activated sludge processes (12,
13) lend support to this idea. Unfortunately, there are no SVI or other
settling rate data which can be used to test this possibility. Another theory
is that the Captor process is simply more robust than activated sludge at
lover temperatures. This possibility cannot be tested either, since frequent
respiration rate measurements are only available for February, 1983.
In June of 1984, under the flow and loading conditions shown in Figure 19
(loading based upon filtered BOD values), a 24-hour survey of the two Freehold
Captor sections was carried out by the project development team (14). The raw
data from this survey are given in Table 11. One pad cleaner in each lane was
run for the first five hours of this test and approximately 180,000 pads from
each lane were cleaned.
Figure 19. Flow and soluble BOD load during 24-hour survey.
49
-------
Tine
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1
2
3
4
5
6
7
8
Miii.
Max.
Avg.
TABLE 11. FREEHOLD 24-HOUR SURVEY, JUNE 1984 (ALL BOD DATA BASED ON FILTERED SAMPLES)
Sewage Air BOD. Values D.O. Biomass/pad Respir. rate Spec. rem. rate
$low £lov feea Cl Eff C2 Eff CI C2 CI C2 og 0 /g-hr mg BOD/g-hr
m /hr m /hr mg/L mg/L mg/L mg/L mg/L mg mg Cl C2 Cl C2
276
1329
30
20
20
4.7
4.5
129
197
--
54.5
7.5
7.3
312
1303
36
27
27
5.0
4.7
86
178
85.0
10.8
8.2
306
1342
41
30
28
5.0
5.0
112
188
56.9
11.6
11.1
282
1347
47
33
31
4.2
4.6
162
186
49.2
--
7.3
12.0
253
1095
56
40
37
3.2
3.0
167
205
--
63.8
8.3
11.8
247
1041
71
48
44
3.0
2.8
160
204
57.5
--
11.6
20.0
253
1187
80
52
47
3.7
3.6
200
234
—
63.1
11.7
17.0
243
1365
90
56
51
3.6
2.6
167
193
55.1
—
16.3
25.8
234
1360
103
59
58
2.8
4.9
175
218
—
68.0
18.4
24.0
256
1355
118
63
68
3.7
2.5
175
189
52.3
29.9
34.1
251
1563
119
66
65
4.5
4.4
180
198
—
78.4
24.7
39.2
256
1870
120
69
64
4.5
4.4
171
227
56.5
26.7
34.5
283
1850
113
65
60
4.6
4.5
174
230
--
73.9
27.6
32.5
263
1848
139
92
82
4.4
4.5
173
220
50.7
25.1
33.7
261
1636
137
101
100
3.8
3.5
175
252
--
72.9
20.0
18.3
217
1320
134
101
96
1.6
1.5
166
218
50.3
19.8
20.0
199
1304
137
100
101
2.0
2.6
168
239
--
61.5
12.7
14.2
166
1304
135
102
89
2.3
2.5
200
227
46.7
--
9.8
19.3
138
1315
136
99
81
1.7
2.2
185
239
59.2
9.6
16.7
140
1306
122
87
75
1.4
1.8
156
233
46.0
10.0
16.4
160
1290
122
86
80
1.4
1.7
142
227
—
65.4
13.8
17.7
134
1295
102
76
70
1.4
1.7
148
214
5.0.7
7.8
12.3
125
1291
88
68
62
1.4
2.1
177
208
—
68.1
4.2
9.1
214
1289
68
62
56
1.0
2.0
178
219
41.8
2.8
7.4
125
1041
30
~20
20
1.0
1.5
86
178
41.8
54.5
2.8
7.3
312
1870
139
102
101
5.0
5.0
200
252
85.0
78.4
29.9
39.2
227
1383
98
67
62
3.1
3.3
163
214
53.5
65.5
14.5
19.3
-------
Using the data produced during the 24-hour survey, several standard
statistical tests were run to compare the two Captor lanes. In all cases,
this involved testing the null hypothesis that the means of measured and
calculated parameters over the 24-hour period were equal at the 0.05 level of
significance. Table 12 summarizes the results of the hypothesis tests.
First, the effluent from CI is virtually identical to that of C2 although
CI contains many more pads than C2, and in fact has a higher EMLSS concentration.
However, the biomass in C2 shows a higher respiration rate and a higher specific
BOD removal rate, and the products of these parameters and the EMLSS concentra-
tion are not significantly different when comparing the two lanes. As would
be expected, the aeration tank dissolved oxygen levels are not significantly
different either, as long as there is roughly equal distribution of air to
both lanes. These tests indicate that the activity of the pad biomass is at
least as important as the pad concentration (number/L) and the pad biomass
level (mg/pad). The Simon-Hartley, Ltd. analysis of these data speculated
that a dissolved oxygen limitation may exist at Freehold, critical levels
being below 4 mg/L. This seems a distinct possibility. In addition, there is
not much question that the substrate supply rate (input BOD flux) may also
limit growth and activity in this system.
51
-------
TABLE 12. STATISTICAL COMPARISONS - FREEHOLD CAPTOR LAMES 1 AND 2, 24-HOUR SURVEY, JUNE 1984
(BOD DATA FROM FILTERED SAMPLES)
Parameter
Capto
Mean
r Lane 1
St'd dev.
Captor Lane 2
Mean St'd dev.
Results of testing difference of
means at 0.05 signif. level
(n) = no. of samples
Eff. BOD, mg/L
66.5
25.:
62.0
23.5
No. diff. (24)
Susp. pads/L*
17.9
1.6
11.4
1.0
Signif. diff. (24)
Biomass/pad, mg
163
25.6
214
19.7
Signif. diff. (24)
Equiv. MLSS, mg/L
2919
517
2446
333.5
Signif. diff. (24)
Mid-tank D.O. mg/L
3.12
1.36
3.26
1.18
No. diff. (24)
Biomass resp. rate, R
mg02/g-hr
53.5
10.9
65.6
7.2
Signif. diff. (12)
Specif. BOD rem. rate, K
mg BOD/g-hr
14.5
7.8
19.3
9.5
Signif. diff. (24)
EMLSS x R, mg/L-hr
153.2
20.1
160.6
27.6
No. diff. (12)
EMLSS x K, mg/L-hr
43.1
25.5
47.3
23.3
No. diff. (24)
*The measured means represent 63% (CI) and 69% (C2) of the actual numbers of pads placed in these lanes. The
unaccounted-for pads are collected in floating rafts. Each of the 24 samples consisted of sub-samples from
14 separate locations.
-------
SECTION 6
TECHNOLOGY ASSESSMENT
PROCESS THEORY
Traditionally, wastewater has been biologically treated at relatively low
volumetric rates of reaction in large tanks requiring high separation costs
and high power consumption for mixing and aeration. A strategy to alleviate
this situation has been referred to as process intensification (1). Briefly,
process intensification seeks to:
o increase volumetric rates of reaction,
o reduce reactor size,
o improve separation processes,
o enhance operational control capability, and
o enhance design capabilities.
Beth biological floes and films provide certain properties which could be
engineered to achieve at least some of these goals. Atkinson et al. (2) have
concluded that this engineering woulJ be best facilitated, however, by the
ability to produce biomasd of any size, shape and density using any substrate-
microbe system. This can be be achieved by immobilization of the biomass in
some form.
It has long been established that the mixed microbial cultures in wastewater
treatment have excellent adhesion characteristics and readily form layers of
immobilized biomass on stone, glass, plastic and other commonly available
materials. These immobilized systems have the advantage of producing high
53
-------
biomass concentrations which are la gely independent of flow-through and do
not suffer from washout. In simplest form, these processes include trickling
filters and the rotating biological contactor. Although greatly improved upon
in the; past decade, these fixed-film systems suffer the disadvantages of
unpredictable biomass hold-up and variability in local hold-up, which inevitably
lead to variable performance and uncertain design capability.
A more recent development, the fluidized bed employing immobilized biomass,
may be designed to overcome some of these difficulties. In a completely
stirred reactor containing uniform, suspended, immobilized biomass particles,
biomass hold-up may be controlled by attrition which results from particle/
particle and particle/wall contacts. Steady biomass hold-up may thereby be
achievable through proper hydraulic design. Conversely, the fluidized bed
could be designed to provide suspended particles of different sizes, shapes or
densities so as to control attrition within the reactor apd enhance recovery
of attached biomass from the system with minimal disturbance (2).
In an effort to develop a system of biomass particles of controlled size,
shape and density, Atkinson (2) believed that the system should allow all of
the natural mechanisms of adhesion and flocculation to take place in a low-
shear environment. Control of the particle size would then be achieved through
either physical or fluid mechanical means. Early work in this direction with
the fungi, Aspergillus foetidus, was initially performed at the University of
Manchester (U.K.) (JMIST) by Perez (3) who used layers of stainless steel mesh
wound about a stainless steel draft tube in an air lift fermenter. Later this
work was extended to suspended parcicles of stainless steel mesh by Atkinson
and Lewis (4). Since that work in 1977-1978, support particle size, shape and
material has markedly changed and the concept has been extended to encompass a
wide range of biochemical reactions employing both pure and mixed cultures.
PBS particles can be placed in fixed, expanded or fluidized beds or in
agitated tanks. In each system, the control of biomass bold-up, recovery of
biomass, performance and design will vary. Table 13 summarizes some of these
Characteristics of PBS reactors (5).
54
-------
TABLE 13. CHARACTERISTICS OF REACTORS CONTAINING PBS PARTICLES (5)
Fixed/expanded Fluidized beds/agitated beds
beds
Variable parti-
Variable parti'
Steady particle
cle biomass
cle bioaass
bicciass hold-up
hold-up
hold-up
Biomass control
Back-washing
Continuous parti-
Particle/particle/
cle removal and
wall attrition
cleaning (parti-
within fluidized
cles could be
beds; particle/
removed period-
particle/agitator
ically
attrition
Biomass recovery
Washings
At high concen-
Conventional
collected in
tration external
clarifier
a humus tank
to reactor
Reactor performance
Variable
Steady (variable
Steady
efficiency, e.g.,
for periodic
carbon removal
particle removal)
Design
Requires pre-
Requires predic-
Facilitated by the
diction of
tion of the varia-
fact that all as-
the variation
tion of the bed/
pects of perform-
of the bed
particle biomass
ance are steady
biomass hold-
hold-up with
and both particle
up with time
time
and bed biomass
hold-ups pre-
selected
The new field of PBS particle technology is still in its early stages of
development and tne biochemical, chemical and physical theories dealing with
systems mechanisms are still being formulated. A brief theoretical synopsis
of the current state-of-the-art follows.
55
-------
Porous Support Particles
3iomass support particles, interwoven three-dimensional webs providing a
random latticework, have evolved from hand-wound stainless sttel spheres to
the present polyurethane foams. It is not certain from a review of the published
literature when the first work with porous biomsss support particles (PBSP)
was introduced or what those materials might have been. The important physical
characteristics of these particles include high porosity and an open structure.
The particles have been prepared from a variety of materials with specific
gravities both greater and less than that of water, and in a range of sizes and
shapes.
PBSP are necessarily open enough to allow normal biological flocculation
mechanisms to take place in a low-shear environment and, subsequently, to
entrap small floes within the lattice structure. Biomass stability is thought
to be aided by adhesion between organisms and the particle surfaces within the
particle, the winding of filamentous organisms about the elements of the
lattice, and the entrapment of flocculent biomass structures (2).
Since biomass completely permeates the lattice structure, particle size
and shape as well as porosity will dictate diffusional control on the entrapped
biomass. Finally, the physical properties of the PBSP will influence the
fluidization, hydraulic distribution, wear and handling characteristics of
the PBSS. For example, spherical particles are easily fluidized and are well
characterized with respect to their fluidization properties and diffusional
limitations. In contrast, cuboids have a large dimension which aids in handling
and a smaller dimension which may be independently adjusted for diffusion
limitation, but fluidization properties are not well defined and hydraulic
distribution can be a problem. Since pad costs represent a significant portion
of PBSS costs, pad wear is of great concern. Particle/particle and particle/wall
contacts as veil as pad handling mechanisms may result in significant wear of
pads.
Table 14 summarizes the characteristics of typical biomass support particles
(5).
56
-------
TABLE 14. CHARACTERISTICS OF TYPICAL PES PARTICLES (5)
Matrix Method of Shape and size Density Porosity
material preparation (i) of matrix of
(ii) of biomass particle
filled
particle
_3
Stainless Single strand Spheres, 3 to (i) 7.7 g/cm 0.80
steel wire of wire, weft 30 mm diameter (ii) 2.42 g/cm
knitted and
compressed in
e die
_3
Polypropylene Single strand Overall diameter (i) 0.9 g/cm .
strands of polypropy- 53 nun. Torus (ii) 1.084 g/cm 0.92
lene, weft diameter 20 mm
knitted and
hand rolled
into a toroid.
Heat sealed.
_3
Reticulated Expanded foams Cuboids (ij 1.2 g/cm .
polyether with lamellae 10 x 10 x 2 mm (ii) 1.103 g/cm 0.95-0.97
foams collapsed by to
sonic shock 25 x 25 x 12.5 mm
waves or Cubes
other methods 12 to 17 mm
-3
Matted Matted in the Cuboids (i) 0.9 g/cns ,
reticulated manner of non- 25 X 25 x 6 mm (ii) 1.09 g/cm 0.95
polypropylene woven fabrics
and heat bounded
Fluidization and Hydraulic Distribution
The fluidization and hydraulic distribution of PBSP is an important
elsment in successful PBSP operation. The fluidization of stainless steel
wire PBSP spheres was studied by Webb et al. (6). They found that the fluidi-
zation behavior of filled spherical PB3P was qualitatively similar to that of
solid spheres (6 mm diameter) and that the Ergun (7) correlations modified for
57
-------
surface roughness gave satisfactory predictions of minimum fluidizing velocities
for these PBSP. Partially filled or empty PBSP required substantial modification
of these correlations to allow for fluid flew through the particles.
The fluidization of low-density polypropylene toroids (sp. gr. - 0.91,
porosity - 92%) was reported by Walker and Austin (8). A pseudo-fluidization
o'i these particles was achieved by introducing fine bubbles of air below the
bed, which wos sufficient to cause the bed to expand downwards into the less
dense mixture of air and water. Initially, the unfilled particles entrapped
air jnd floated; but once biomass was established in the particle, air entrap-
ment became less cf a problem. Air flows required to induce pseudo-fluidity
of these particles .increased as particle numbers per unit volume increased and
as dry weight of biomass per particle decreased. Furthermore, there was an
upper limit of particle numbers per unit voluue where no amount of air flow
could move the bed. An extension of this work to a more porous polypropylene
pad (95 to 97%) showed similar fluidization properties, with an important
difference. Higher porosities resulted in higher biomass concentrations and
concomitant higher particle densities which approached or exceeded that of
water. Air flow requirements for fluidization, therefore, were dramatically
reduced and more stability of fluidization was noted.
With the introduction of higher-density reticulated fo'im pads (sp. gr. -
1.1, porosity - 97%), the fluidization and distribution of Lhe pads did not
require the higher superficial velocities needed to fluidizo the stainless
steel PBSP or the high gas flows to induce pseudo-fluidity of the lighter-
than-water polypropylene particles. Particle motion was controlled by the
intensity of gas flow and particles behaved as elements of the liquid in which
they were submerged. The behavior of PBSP in this type of system is difficult
to describe and has not yet been sufficiently modeled. Observations by Black
et al. (9) pointed out that:
o when gas flow was introduced over the entire bottom area of the reactor,
circulation was poor and there was a tendency for particles to float;
and
58
-------
o good liquid/PBSP mixing was most effectively achieved by introducing
gas over only a portion of the realtor bottom jrea.
These observations were further supported by the full-scale experience at
Freehold in the U.K. and at an industrial system in the U.S. (Section 5).
Based on these experiences, an aeration system was devised to provide for
regions of downward liquid ilow ant effective recirculation by aeans of proper
diffuser distribution (U.K. Patent Application, 1982). Work continues on this
important design feature both at the Water Research Centre ir. the U.K. and at
Linde Aktiengesellschaft G1G) in Munich.
Biomass Hold-Up Characteristics
The FBSP provides a site for the development of biomass in flocculent,
filamentous and attached forms. It has been demonstrated that these particles
can be filled with a wide variety of organisms, including those that do not
flocculate, at biomass concentrations up to 100 g (dry weight)/Iiter (free
particle volume) (2,9). Applications of PBSS in a variety of biological
reactor systems under aerobic and anaerobic conditions are pr»sf*nted in Table
15 (10).
The actual mechanisms of biomass development and maintenance within PBSP
are not well understood. It is clear that the floes formed within the particles
are robust enough to allow a sufficient degree of particle movement without
loss until they extend beyond the limit of the particle structure. Control of
biomass hold-up may then be affected by a balance between biomass growth and
attrition. The open structure of the reticulated foams permits gas evolved by
the biomass to escape freely without disruption of the biomass integrity and
enhances diffusion of substrate, oxygen and waste products throughout the
particle matrix. Although some researchers reported biomass "aging" and
mineralization with time in PBSS (8), stable biomass activity appears to be
the more prevalent finding for a variety of substrate systems (9,16,17,18).
The overall density, P , of a PBSP may be estimated by:
P = P (1 - 0) + P (6.1)
pw m ew
59
-------
TABLE 15. SOME REPORTED APPLICATIONS OF BSPs
Organism
Product
PBS system*
Comment
Reference
Hixed culture
(aerobic)
Waste
treatment
FOAM/CBF
3
Up to 160 m
retrofit to
existing
activated
sludge plants
Walker & Austin
(1981) (£).
Richards &
Wilson (1983)
(11).
Hixed culture
(anaerobic)
Methane
FOAM/CBF
1 liter
Fynn & Whitmore
(1982) (12).
Yeast
Ethanol
Stainless
spheres/FBF
FOAM/CBF
2.0 liter
Black et al.
(1984) (9).
Acetobacter sp.
Acetic acid
Stainless
spheres
Fixed bed
da Fonseca,
(1983) (13).
Aspergillus sp.
Citric acid
Stainless
spheres/FBF
Atkinson & Lewis
(1980) (4).
Trichodenna sp.
Cellulase
Stainless
spheres/SBF
Approx. 7
liters
Fukoda et al.
(14).
Capsicum sp.
(plant cell)
Capsaicin
FOAM
Shake flasks
Lindsey et al.
(1983) (see (10))
Humulus sp.
Hop flavors
FOAM
Shake flasks
Rhodes & Kirsop
(1982) (see (10))
Mixed cultures
(aerobic)
Waste
treatment
FOAM/CBF
Up to 45 m3
polishing,
improve SVI
Hegemann (1983)
(15). Hegemann
& Euglmann (1983)
(16). Reimann
(1984) (17).
Mixed cultures
(aerobic)
Nitrification
of secondary
effluent
FOAM/CBF
1 m3
Reimann (1984)
(17).
*CBF - Circulating bed fermenter
FBF - Fluidized bed fermenter
SBF - Spouted bed fermenter
60
-------
where is the density of the support matrix, is the effective wet biomass
concentration based on the total particle volume, V , and jJ is the particle
porosity. If the immobilized biomass is assumed to consist of a water phase
(density Pw) and a biomass phase of constant wet density, P^, then
P = (P (1 - x) + P. x) 0 (6.2)
ew w bw
where x is the fraction of the porosity filled with biomass of density P^.
Overall particle densities calculated from Equations 6.1 and 6.2 for a
variety of PBSP shown in Table 14 are relatively insensitive to the density of
the support matrix because of the high PBSP porosities, and the similarities
between the density of water and biomass. Thus, a tendency toward particle
classification is relatively small for most PBSS.
The dry biomass hold-up, Pg, of the PBSP of high porosity may be estimated
by:
P = in /V (6.3)
e p' p
where m^ is the dry weight of biomass in the particle. When particles are
completely filed with biomass at a solids concentration, P^ (dry weight per
unit wet available volume) corresponding to that of drained wet biomass, then
Pe = Pb0 (6.4)
and the amount of biomass (dry) per particle may be estimated for a given PBSP
as
®p = Pb • Vp <6'5>
Finally, the overall concentration contributed by the PBSP to the reactor,
P^, would be calculated as
Pf = Np mp/VR (6.6)
61
-------
where N is the total number of particles in the reactor and the reactor
P
volume is Vjj. Added to the value calculated by Equation 6.6 is the free
biomass suspended in the reactor anJ that attached to the reactor walls as
film. The contributions to the entire reactor system made by the suspended
portion of the biomass can be significant, as in the Linpor systems designs.
Recent experiments with 6 mm polyester foam PBSP cubes with a porosity of
97% and varying pore sizes ranging from 30 to 60 pores per inch revealed that
biomass hold-up, m^, for a yeast culture increased with the numoer of pores
per inch (9). Under all conditions, the 60 ppi particles contained more
biomass than those at 30 ppi for the same particle size and porosity. Similar
unreported studies by Siaon Hartley on 25 x 25 x 12.5 mm polyether foams with
mixed cultures indicated an optimal pore size of 30 ppi with decreases in
biomass hold-up at higher or lower pore sizes. Reasons for these findings are
still unclear, but these r.tudies suggest that pore size and pad size are
important variables in designing PBSS.
For particles with steady biomass hold-ups, it is reasonable to suggest
that aip represents a balance between overall growth and degree of attrition.
Thus, when the overall specific growth rate, G, exceeds the specific rate of
attrition, T, the bioirass hold-up will increase as:
wflp - (G ~T)n)p (6.7)
dt
According to the biological rate equations (19), it can be deduced that as m
P
increases G decreases due to the greater diffusional limitation on entering
substrate until a balance (steady state) is achieved between growth and attri-
tion at a giveu reactor substrate concentration. Atkinson et al. (18) have
deduced that T, under a given set of conditions of particle and bed hydraulics,
will either be a function of a or a constant. Thus, for steady state conditions
P
o will be either proportional to G or be a constant.
P
62
-------
In analysis of data collected on a glucose-mineral salts media, Atkinsoa
et al. (18) indicated that a relationship existed between biomass hold-up and
reactor glucose concentration. Assuming that
= f(G) and
r c
[ o
p+c
P C
G ~ f( o ) (6.8 a)
Atkinsoa showed that
n = m max ( o ) (6.8 b)
where C is the reactor substrate concentration and fi is an constant,
o r
This work also suggested that biomass hold-up increased with increased
influent glucose concentraticn up to some limiting value. Thus, for a given
reactor turbulence, there is likely some limiting biomass concentration that
can be effectively maintained within a PBS particle under steady-state condi-
tions.
Waste strength alone does not dictate biomass hold-up, as demonstrated in
3
tests conducted in the 25 m pilot plant system at the Wat?r Research Centre.
Biomass hold-up measured in this pilot plant with a wastewater BOD ot al.cut
200 mg/L ranged from 100 to 120 mg biomass per pad, as compared with hold-ups
in excess of 200 mg/pad at the Freehold facility, where wastewater BOD values
were about 140 mg/L. It would appear from this preliminary data that wastewater
characteristics, organics loading rate and hydraulic regime all contribute to
the accumulation of biomass within the PBSP.
Substrate Uptake Relationships
The overall rate of substrate removal, R^, in PBSS is the result of the
combined metabolic effects of the support particles, films and suspended
floes. The latter is the result of biomass erosion from PBSP and reactor
63
-------
walls as well as growth of the suspended biomass within the reactor. The
general rate equation may be written:
R = R H +NA+RM (6.S)
1 p p s s
where R^ and Rg are the overall specific rates of substrate uptake in the
reactor by the FBSP and by the suspended floes, respectively, N is the
overall rate of substrate uptake per unit area of microbial film, and
are the total quantities of biomass on the support particles and freely
suspended floes, respectively, and A is the area of microbial film. Atkinsoq
et al. (18) have developed rate equations for this system assuming that the
intrinsic microbial kinetics are described by the Honod equation and that there
is a strong diffusional limitation within the PBSP and the attached reactor
films but net within the freely suspended floes.
In summary, these equations are presented in Table 16 (18). Results of
experiments using stainless steel spherical PBSP with glucose as a substrate
indicated that the overall rate of glucose uptake by a mixed microbial
population followed half-order kinetics in C over a range of substrate concen-
trations up to 4 g/L. A typical result of these experiments is shown in
Figure 20 (18).
3
Studies currently under way at the Water Research Centre employing a 25 m
pilot facility have produced a relationship as depicted in Figure 21. Attempts
to develop a simple, useful design relationship for BOD^ uptake have not yet
been entirely successful, since the defining independent variables are not yet
well delineated. A»: important missing element in the relationships presented
by Atkinson et al. (18) as well as in Figure 21 is the biomas history (usually
characterized by mean cell residence time).
In the CSTR operated by Atkinson et al. (.8), "steady-state" biomaes was
claimed to exist after three days of constant biomass hold-ups and conversion
efficiencies. Under steady-state conditions, it is assumed that biomass
wasted from the system has the same activity as that in the PfcS paitides. If
this is not the case, steady biomass hold-up would not imply steady mean cell
64
-------
TABLE 16. BIOLOGICAL RATE EQUATIONS FOR PBSS (20)
R =
max
<*k, d/b
2p p
2
-------
3.0
20
1.0
0
0.4
0.3
0.2
0.1
0
t.O
o 0.5
¥ V
?
V
V
V
V
V V
v V
V
16000 Particlts
,
< «
1
«
~
~
¦
CXI Y
T
WW
% ~
~ ~
~
-
~ °
*
-
%
8
~
16000 Particle*
~
~
~ Particle Biomass
-
Hold-up
•
~ Freely Suspended
Biomass
1
,
-
o'o *
••
o
. |
•
?•*
16000 Particles
O Glucose
~ "
A
* TOC
Cf
*
5*
• COD
•
•
i
II I. .1
-1
'
1.0
4.0
2.0 3.0
Hydra _ Ratldertcs Tims
(Hours: Bated on Liquid Volume in Fermenter)
6.0
Figure 20. Performance of PBSS using glucose (18).
66
-------
80
70
Water Research Centre
Pilot Plant - 25 ms
40 pads/L (avg.)
123 mg/pad (avg.)
DO > 3.0 mg/L
Oct. 1984 - Feb. 1985
Figure 21. BOD removal relationships
T
2
Organic Load - kg/m*, d
Captor.
-------
residence times. That Atkinson et al. (18) were able to achieve a reasonable
degree of reproducibility over a ?.4-month period of continuous operation
suggests that the assumption of steady-state mean cell residence times is
probably valid for thei.. system.
A similar question must be raised for systems employing mechanical removal
of biomass (Captor). Here, mean cell residence time is purportedly controlled
by PBSP squeezing rates. There is some evidence to suggest that as squeezing
rate increases, PBSP biomass concentrations decrease, biomass activity increases
(as measured by oxygen uptake rate) and more efficient suspended solids "clean-
up" occurs. There is a need for additional research on the impacts of PBSP
cleaning on biomass activity before useful design models can be developed for
these types of systems.
The impact of diffusional limitations on system performance must also be
addressed in modeling efforts. The Atkinson et al. (19) biological rate
equations account for diffusion through effectiveness factor correlations.
These corrections cannot be ove-looked, especially in scaling up pilot studies.
An important design parameter used in modeling biological systems is the
biomass yield coefficient. There is a paucity of data on biomass yield for
PBSS at this time. The reason for this deals primarily with methods of
analytical measurement. Yields must account for PBSP biomass, suspended
biomass and fixed films along reactor walls (in laboratory-scale systems).
Very preliminary data from wastewater treatment facilities indicate biomass
yields ranging from 1.2 to 0.3 kg/kg BOD,, removed, depending upon cleaning
rates (20).
One final point should be made relative to substrate uptake modeling.
The fundamental work reported by Atkinson et al. (18) was conducted with a pure
substrate glucose, and glucose uptake was directly monitored. Attempts to
monitor TOC or COD uptake in these studies produced poor correlations, as might
be expected, since both measurements are nonspecific. In wastewater systems,
these "lumped parameters" of measurement are commonly used. Furthermore,
filtered, settled or total samples may be analyzed. These measurement con-
straints add to the difficulty of effective PBSS performance modeling.
68
-------
Other Consideratirns
Oxygen Transfer--
Early repo.ts from pilot plant work with Captor (20) indicated that PBSS
significantly increase the efficiency of oxygen transfer for both coarse and
fine bubble systems. This enhanced transfer may be ascribed to air bubble
hold-up and attachment io the PBS particles as veil as to interfacial transfer
between the attached bubble and the active bicmas?. surface. Reiber and Stensel
(21) reported increases in oxygen transfer through a sparged fixed-film
biological reactor, and attributed this transfer to both bulk liquid and inter-
racial transfer. Both laboratory and field work with PBSP are under way to
evaluate 'his phenomenon and to quantify it more fully. Preliminary results
from these tests indicate that the original claims were exaggerated.
Off-gas oxygen transfer tests under field conditions appear to be '.he
most elective way to evaluate PBSS. Unfortunately, direct comparison with
comparable suspended growth systems under identical conditions may be very
difficult. In some configurations, gas flow rates required to hydraulically
mix PBSP may be significantly higher than optimal for oxygen transfer.
Furthermore, diffusion limitations bv DO way require higher sixeJ-liquor DO
concentrations than employed in conventional suspended growth systems (Section
S, page A3). Thus PBSP mixing and ent.rainment requirements, as well as higher
DO requirements, may offset any gains achieved in oxygen transfer.
PROCESS CAPABILITIES, APPLICATIONS AND LIMITATIONS
The PBSb lias been found to lead to the following advantages for treating
wastewaters:
o Very high concentrations of biomass cau be retained in the reactor, as
compared with suspended growth systems.
o Biomass nay be retained within the reactor, obviating the need for
external biomass recycling.
69
-------
o Biomass may be recovered directly from the TBS particles, thereby
reducing the solids loading to the clarifier.
o The biomass recovered from the pads may be higher than that obtained
from gravity settling and thickening.
These advantages suggest the following immediate applications of this new
technology:
o Upgrading existing overloaded treatment facilities, and
o Pretreatment of high strength industrial wastes.
Current PBSS technology is considering a number of different reactor
configurations. Table 17 presents the characteristics of four general reactor
arrangements. These systems may be single inits or may be staged in combination
with other treatment processes. In general, PBSS reactors are mixed by air
diffusion with fine or coarse bubble devices. Diffuser placements vary from
one system to another, but are critical in providing proper PBSP distribution.
PBSSs are designed with hydraulic configurations ranging from completely mixed
to plug flow systems with aspect ratios as high as 10 to 1.
Details of reactor systems now being tested are described in Section 4.
The primary limitations to the current PBSS technology, as described in
Sections S an«i 7, deal with:
o PBSP distribution,
o Presence of high suspended solids in the effluent, and
o The need for a fundamental understanding of PBSP mechanisms which will
lead to more effective modeling of the process.
DESIGN CONSIDERATIONS
The criteria for designing PBSS for wastewater treatment applications are
currently being developed. rlany of the important design parameters for these
70
-------
TABLE 17. REACTOR CONFIGURATIONS FOR PBSS's AND THEIR CHARACTERISTICS
Characteristic
PBSP retained
or removed
Biomass control
Effluent biomass
recovery
Removed
Reactor type
Removed
Retailed
Retained
Squeezed Squeezed Attrition Attrition
None Clarification Clarification None
Biomass wasting Squeezed solids Squeezed solids Underflow Effluent
Biomass recycle
No
Yes
Yes
No
systems have been identified and are outlined in Table 18. A brief discussion
of a number of these parameters and their current status is presented below.
PBS Particle Characteristics
Both Simon-Hartley Ltd. and Linde AG have expended substantial time and
money to develop PBS particles that are wear-resistant and that meet the
requirements of their particular process. Reticulated or partially reticu-
lated polyurethane (polyether) foams have been selected over polypropylene or
polyester outerials based upon the better wear-resistant properties of the
former materials. High porosities are desirable, as described earlier,
typically being between 95 and 97 percent. Pore size apparently affects total
biomass hold-up, but insufficient research has been performed with wastewaters
to indicate the importance of this characteristic. Simon-Hartley Ltd. feels
that the 30-pore-per-inch size is optimal for its pad size and configuration.
They feel that biomass hold-up is dependent upon the ratio of pad size to pore
size.
71
-------
TABLE 18. SELECTED PBSS DESIGN PARAMETERS
PBSP characteristics:
Material, pore size,
density, dimensions
PBSP number/reactor volume
Biomass hoid-up
Reactor MLSS (total)
Biomass specific substrate uptake
Reactor loading:
F/M
SRT
Volumetric load
Hydraulic residence time
Air flow:
PBSP distribution
Oxygen transfer efficiency
Aeration efficiency
Oxygen demand (SOUR)
Reactor D.O.
PBSP cleaning rate
Biomass yield
Sludge concentration
Screen opening
Screea hydraulic load
PBS particle dimensions depend to a large extent upon the need to handle
the particle (i.e., to remove and clean them). A maximum diffusional dimension
of about 6 to 8 mm has been adopted by both Simon-H^rtlcy Ltd. and Linde AG.
Currently Simon-Hartley Lid. employs a 25 am x 25 mm x 12.5 ram particle,
whereas Linde AG has elected to go with cubes ranging from 10 to 15 mm on a
side.
PBS Particle Number and Biomass Concentration
To date, PBS particle concentrations have been used in wastewater appli-
cations ranging from 10 t.o AO percent by volume. Increasing particle concen-
trations increase potential biomass concentrations and concomitant volumetric
rates of reaction. Unfortunately, upper limits on particle concentrations are
dictated by basin hydraulics, which are, in turn, dependent upon basiu geometry,
diffuser type and placement and gas flow rate (current PBSS are all mixed by
air). Upper limits on PBS particle concentrations in practice today range
from AO to 50 percent. Substantial research remains to determine optimal
basin geometry, diffuser placement and gas flows for PBSSs.
As described earlier, particle biomass hold-up is dependent upon substrate
concentration and characteristic particle pore size, hydraulic regime and
cleaning rates. Wastewater characteristics and loading variations have also
72
-------
been implicated as potential controlling variables as well. Current experience
of Simon-Hartley Ltd. with British wastewaters indicates that typical particle
hold-ups range from U to 8 g/L (based on dry weight per gross particle volume)
with an average of, perhaps, 6 g/L. Initial Simon-Hartley Ltd. design predic-
tions were for particle hold-ups of 20 to 25 g/L. Linde AC predicts particle
hold-ups ranging from 10 to 20 g/L.
Total reactor biomass concentrations are dependent upon particle biomass
plus suspended biomass. Simon-Hartley Ltd. predicts typical reactor biomass
concentrations in the 7,000 to 10,000 mg/L range (particle biomass only).
Linde AG may expect concentrations tanging from 5,000 to 10,000, assuming
particle biomass (10 to 20 g/L) plus suspended biomass (2 to U g/L).
Substrate Uptake/Loading
Currently, the loading parameter that best depicts PBSS performance is
not well de'ir^ated. The fundamental independent variables that affecr 2ny
biological treatment performance, namely, active biomass concentration,
specific substrate uptake rate, substrate concent ration/characteristics,
dissolved oxy&in concentration, temperature and pH, should affect PBSS. The
mathematical form of the design parameters, however, is not yet clear.
Currently, PBSSs arc designed on the basis of BOD loading (mass/time, mass/
volume, time) and hydraulic detention time (reactor volume/forward flow,
reactor volume/forward + recycle flow). Initial Simon-Hartley design loadings
3
of 4 to 5 kg BOD/m -d and a retention time of 45 to 6C minutes appear to be
optimistic. Recent pilot plant work performed at the Water Research Centre
indicate that a loading rate of 2 to 3 kg/BOD m . d is more reasonable, giving
total BOD removals of about 50%. Linde AG design carbonaceous systems in the
3
range of 2 to 3 kg B0D^/iu- d at residence times of 100 to 150 minutes. Addi-
tional research is needed to develop more definitive loading parameters for
these systems.
Adding to the complexity of this performance parameter formulation is the
contribution of suspended biomass to system kinetics (positive) and effluent
quality (negative). This problem is not unique to PBSS's. Fixed-film and
fluidized-bed systems as well as disperseo growth reactors produce dispersed.
73
-------
solids that negatively contribute to effluent quality. Analytical procedures
must be carefully specified in order to properly evaluate this contribution.
The fraction of active biomass associated with the nonparticle suspended
solids in the reactor will contribute in a positive fashion to substrate
uptake. Unfortunately, the activity of this biomass fraction is likely dif-
ferent than that associated with Lhe particle. (Diffusion limitions in the
particle may or may not affect specific substrate uptake r£tes( and attrition
or cleaning of particles may produce a disproportionately high amount of
inactive biomass.) Clearly, suspended biomass contribution to reactor per-
formance is more important in Type 2 and 3 reactors (Table 17).
Air Flow - Oxygen Transfer/PBS Particle Mixing
Currently, all PBSSs are mixed with diffused air. As discussed earlier,
the mixing requirements of PBSSs are not yet well developed but are highly
dependent upon mixing patterns within the basin. In order to effectively mix
PBS particles within the reactor, a region of downward flow must be provided.
Furthermore, in tanks that produce a more nearly plug flow condition, back
mixing of pads countercurrent to influent flow must be provided in order to
prevent a build-up of PBS particles at the effluent end of the reactor.
Currently, the Captor project team of Simon-Hartley Ltd. and the Water
Research Centre are considering completely mixed tank configurations with
diffusers placed such as to provide a region of downward liquid flow. Using a
3
25 m test tank with an aspect ratio of 2:1, investigators at the Water
Research Centre laboratories studied three distinct diffuser placement patterns:
o Longitudinal placement with all diffusers along one side occupying 25
percent of the bottom area.
o Longitudinal placement with all diffusers along one side occupying 46
percent of the bottom area.
o Transverse placement with all diffusers along the centerline occupying
26 percent of the bottom area.
74
-------
Air flows required to achieved effective roixing of 36 pads/L for the
2
three configurations above were 75, 133 and 141 L/m • min, respectively. Full
2
floor coverage required air flows of up to 300 L/m . min. Linde AG uses a
cross roll configuration in tanks of higb aspect ratio operated as nearly plug
florf reactors. Air flow rates in full-scale basins appear to be high, but
particle distributions from influent to effluent end are fairly uniform. No
technical design data are currently available fron Linde.
Air is also provided to satisfy the respiratory requirements of the
biomass. Although few measurements of specific oxygen uptake rate (SOUR) have
been made for PBSSs, values reported range from 30 to 120 mg O^/g hour. These
are in the range of those reported for conventional activated sludge. Total
oxygen requirements (mg/L/hr) for PBSSs, however, could be substantially
higher than conventional suspended growth systems because of the higher active
biomass per unit volume.
As discussed earlier, there is sone evidence tli3t oxygen transfer
efficiencies of PBSS's are higher than suspended growth reactors with comparable
geometry, diffuses placement and gas flow rate. These increases in efficiency
are reported to range from 10 to 15 percent (Linde AG) and higher (Simon-
Hartley Ltd.). Substantial attention must be given to quantifying these
increases through off-gas analysis. Since air flow rates required to effectively
mix particles may be higher than those for optimal oxygen transfer, it is
likely that tue advantage of increased efficiency may be substantially lost in
PBSSs. Conservative design, assuming conventional transfer efficiencies, is
warranted at this time until more definitive data are available.
Finally, in designing the oxygen transfer system, it is important to
provide reactor dissolved oxygen concentrations that are sufficient to prevent
diffusion limitations. Recent field experience at Freehold and at the Water
Research Centre with pilot plants using carbonaceous removal systems, and by
Linde AG with nitrification systems, indicate that specific rates of substrate
removal are Strongly correlated with dissolved oxygen concentration (Section
5, page 22). The high DO concentrations require higher air flow than needed
75
-------
for mixing pads in the the Captor carbonaceous configurations. Recent studies
at Water Research Centre indicate air flow requirements in the region of 2C0
2
to 240 L/m • rain for proper maintenance of dissolved oxygen.
PBS Particle Cleaning/Sludge Characteristics
Cleaning of PBS particles is an important feature of the Simon-Hartley
Ltd. PBSSs. The mechanical design of PBS particle cleaners is now well
developed (Section 5). The impact of particle cleaning frequency on
PBSS performance is not well understood, however. Excessive cleaning appears
to severely reduce particle solids and result in poor particle distribution
(flotation). Little or no cleaning may result in excessive biomass build-up,
diffusional limitations, and inert solids build-up in the particles. Simon-
Hartley Ltd. recommends cleaning rates in the range of 10 to 20 percent per
day (number cleaned/day/total number in system).
Intuitively, one would expect that an increase in cleaning rate would
produce a younger, more active particle biomass, a higher sludge yield and
perhaps a greater capacity of the particle to "mop up" dispersed suspended
solids. Furthermore, it is theoretically reasonable to assu/:e that the
cleaning rate should be established at a point where no attrition will occur.
Currently, there are insufficient data to show any relationship between
cleaning rates and specific substrate uptake rates, effluent suspended solids,
biomass activity (SOUR) and sludge yield.
Sludge solids concentrations obtained from PBS particle cleaners have
varied with the generation of cleaner. The technology is now available to
effectively drain water away from pads by gravity prior to the final squeezing.
Under these circumstances, particle sludge concentrations in the range of 4 to
5 percent may be achievable.
The yield of biomass in PBSSs is not well documented to date. It is
reasonable to assume that yield would be similar to conventional biological
treatment systems. But again, t.he defining independent variables affecting
yield in PBSSs are not yet known.
76
-------
Screening
Screen openings are dependent upon fBS particle size. Currently,
Simon-Hartley Ltd. has found that woven mesh, A mesh (5.13 mm aperture) works
very well for the 25 mm x 25 mm x 12.5 mm PBS particles. Screens are sized to
3 2
handle a wastewater flow of 1A m /h/m . It is imperative that PBSS's be preceded
by adequate screening, however. Build-up of detritus escaping from primary
processes can cause serious screen blockages in the PBSS if adequate prescreening
is not provided. No data are currently available on Linde AG screens.
ENERGY ANALYSIS
The energy requirements for PBSS are probably very similar to those of
conventional activated sludge systems. Currently, there are insufficient data
to accurately assess energy differences that may exist. Basically, energy
differences between these processes, if they exist, will be in providing
mixing and oxygen requirements. As described above, the PBSSs may achieve
higher oxygen transfer efficiencies than comparable suspended growth systems,
but air flow requirements may be higher in order to effectively mix the particles.
Particle cleaners will also consvune some eneigy in the Type 1 and 2 processes
(Table 17).
OPERATION AND MAINTENANCE REQUIREMENTS
Operational requirements for PBSS's are listed in Table 19. In general,
they parallel requirements for activated sludge systems, with the exception of
particle cleaning rates in Type 1 and 2 processes. Laboratory testing includes
influent and effluent monitoring and reactor monitoring including particle
biouass hold-up, particle distributions, SOUR, suspended solids, cleaner
solids, dissolved oxygen, temperature and pH.
Maintenance of PBSS is also similar to conventional activated sludge
systems (Table 19), with the exception of PBS cleaners and screens.
77
-------
TABLE 19. PBSS OPERATION AND MAINTENANCE REQUIREMENTS
Maintenance
Operation
o Screens
o PBS cleaner
o Blowers
o Diffusers
o Reactor D.O.
o Pad cleaning rate (Type 1 & 2)
o Sludge return (Type 2 and 3)
o Sludge wasting
o Influent/effluent monitoring
o Reactor monitoring
COSTS
Cost data on PBSS are not well developed at this time. A detailed Cost
analysis performed by the Water Research Centre appears in Section 7.
78
-------
SECTION 7
COMPARISON WITH CONVENTIONAL TECHNOLOGIES
GENERAL
The studies performed on porous biomass support systems to date have
been aimed more at development than demonstration or evaluation. Lengthy
records of steady-state performance under optimum or even near-optimum opera-
tions do not yet exist. The technology cannot be considered as satisfying
the "fully developed" I/A criterion au this time. Some of the comparisons
against conventional technologies which are normally called for in an assess-
ment of this type are, therefore, premature. Nevertheless, comparisons will
be attempted, using the existing data base, in seven general areas. These
are:
1.
Cost
2.
Energy
3.
Performance/Reliability
4.
Environmental Benefit
5.
Toxics Management
6.
Joint Treatment Potential
7.
Residuals Generated
Cost
Only two cost analyses of the Captor process have been prepared to date.
One of these was by D. H. V. Wheeldon of the Water Research Centre in July, 1981
using third-quarter 1980 costs to compare Captor versus activated sludge for a
79
-------
new plant at the scale of 20,000 ra^/d (5.3 mgd). Two options were considered:
a carbonaceous BOD removal system satisfying U.S. secondary treatment standards
of performance, and a system capable of year-round nitrification. The assump-
tions used in that anaij^is were as shown in Table 20. Secondary clarifiers
were assumed to je required for the Captor systems. With there assumptions,
the conclusion was that the most significant cost element in a present worth
analysis of Captor was the replacement rate of the pads. Captor was judged
more cost-effective than a nitrifying activated sludge syst?m when the pad
replacement rate was less than 20 percent per year. Captor was more cost-
effcctive than an activated sludge system designed for carbonaceous BOD removal
when the pad replacement rate was less than 30 percent per year. It is
interesting to note that current estimates of pad life are about 12 years, a
replacement rate of 8.3 percent per year.
More information is, of course, available now. primarily from the Freehold
study, which can be used to test the rest of Wheel doii's assumptions and perhaps
expand the scope of the present worth comparison somewhat.
An economic analysis was also performed in connection with the planning
for the Freehold study. This analysis, performed by H. Crabtree of the Severn
Trent Water Authority, compared the addition of CapLor as pretreatment in all
TABLE 20. PERFORMANCE ASSUMPTIONS - ECONOMIC ANALYSIS OF CAPTOR
VS. CONVENTIONAL ACTIVATED SLUDGF.
Captor
Activated sludge
Biomass, g/L
9
3.5
F/M, kg BOD/kg-d
Nitrification
0.15
0.15
Non-nitrifying
0.25
0.25
Aeration efficiency
kg 02/kwhr
1.8
1.8
baste sludge cojc. (%)
5
cosettled with
primary sludge to 7
Pad replacemevt rate
(%/yr)
0-100
N/A
80
-------
five aeration basins to the addition of biological towers. Again, the conclusion
was that Captor would be more economical than the nitrifying filters if the
replacement rate of the pads was less tban 20 percent per year.
Energy
Preliminary data on oxygen absorption in a Captor cell do point to probable
increases over activated sludge designed for the same effluent quality. The
few oxygen transfer experiments performed by Hegemann (1) using the Linpor
foam elements also point to higher oxygen transfer; however, the increases
found by Hegemann (about 10 to 15 percent) were much more modest than t.he
increases reported from some of the early Simon-Hartley, Ltd. pilot-plant
work. Until the performance of off-gas testing at the Freehold facility is
completed, we are reluctant to quote specific figures relating to possible
energy savings over conventional activated sludge. We are willing to state
that there is no possibility of decreasing the efficiency of energy utilization
by employing foam elements in an aeration basin. The quantity a K^a will be
at least as high as one would find in an activated sludge system of the same
geometry and aeration system design.
Performance/Reliability
Porous bioaass support systems have some unique characteristics which,
under certain conditions, will provide greater performance reliability ..than
equivalent conventional technologies. The qualifying conditions are important,
however, because when they are not met, PBSS's may prove less, rather than more,
reliable. First, preliminary and primary treatment systems must be designed
and operated to protect the PBSS, especially against grease and solids, the
latter of which can interfere with the hardware (pad cleaners in the case
of Captor and retaining screens for both Captor and Linpor). Second, these
processes do not seen, to have special advantages for carbonaceous BOD removal
from weak wastewaters. No "break-even" economic analysis has yet been per-
formed to determine the approximate cross-over point (concentration of BOD)
where PBSS's show advantages over conventional technologies, if such a point
exists. With the conditions of at least medium-strength wastewater and good
31
-------
preliminary/primary treatment satisfied, the advantages of PBSS's are as
follows.
For certain applications there is les?, sometimes no dependency on secondary
clarifier operation as a sine
-------
There is also an improved reliability of PBSS's over activated sludge
relative to their ability to handle "shock" loads (severe but temporary per-
tubations in concentrations, flows, or both together). PBSS's would show no
advantages over trickling filters in this regard, however.
The advantages stated previously in this section can be taken as
reasonably well-established fact. Given the potential economic advantages as
well, an engineer might well ask whether activated sludge systems are on their
way out, to be replaced by PBSS's wherever the two basic conditions are met.
Careful analysis of the data, however, indicates that there are trade-offs to
consider. In particular, a somewhat lower performance of PBSS's is indicated
when compared against activated sludge at comparable F/M loadings. It is not
correct to say that because you can maintain 8,000 mg/L of EMLSS in a PBSS and
only 3,000 mg/L in a conventional activated sludge plant, that you can reduce
the volume of the PBSS aeration basin to 3/8 of that required by the activated
sludge for the same end point. Because of differences in physiological age
and condition of the respective sludges, the PBSS is likely to show a slightly
lower specific activity (substrate conversion rate in mg BOD/gram EMLSS-hr.).
Only additional research will answer questions concerning the relative
effectiveness of biomass in PBSS's as compared to activated sludge.
Environmental Benefit
PBSS's will involve less commitment. of land then trickling filters,
without the potential for odors or filter fly production, and with less
commitment of land than activated sludge. There may be substantial environmental
improvements when PBSS are used to retrofit an existing overloaded plant,
since this is one method of upgrading which does not require extensive new
construction or a lengthy disruption of normal treatment practices. Additional
possible benefits are discussed in the next section.
Toxics Management
In general, there are probably no substantial differences relative to
conventional technologies. An exception may exist where PBSS's are used as
83
-------
pretreatment processes for wastewaters which only occasionally show slugs ot
toxics, and in which the toxic materials tend to partition onto both the
biomass and the pad material. From the Freehold studies it is known that
Captor did provide protection for the downstream activated sludge process
against the effects of three toxic discharges. Two known episodes of copper/
nickel dumping and one of waste oil have been documented since the Captor
study began at Freehold. The heavy metal discharges affected the control
activated sludge system severely, and on the second occasion the sludge became
so inactive that most of it had to be wasted from the system to reestablish
activity. The activated sludge systems downstream from the Captor processes
were not so affected. During the oil discharge episode it was found that the
pads, being hydrophobic, had a natural affinity for the oil and sorbed most of
it, preventing its escape to the downstream activated sludge process. Although
difficulties were experienced with the pad cleaning operation because of the
oil, the overall process stability was far better than that of the control,
which experienced sludge settleability problems tor some time after the discharge
had occurred.
Jrjnt Treatment Potential
No special advantages are seen at this time relative to conventional
technologies in terms of process capability, except as noted in the Performance/
Reliability section above, for cases where the industrial wastewater component
tends to produce filamentous bulking. An economic advantage should certainly
be gained whenever the flow diagram of Figure 22 can be employed. In the
diagram, a strong industrial waste is pretreated, using Captor, and the concen-
trated waste sludge is diverted to the same sludge treatment system used for
the domestic waste sludges. Joint biological treatment (of any sort, including
perhaps the use of Captor) is provided for the combined effluent.
Residuals Generated
For a comparable level of treatment, the dry weight of sludge generated
in a PBSS will be practically identical to that produced by any conventional
process. The advantage will lie in comparison of the wet volumes generated
relative to activated sludge. Both dry weights and wet volumes will be closely
comparable to the trickling filters.
84
-------
To Joint Sludge
Treatment System
Municipal
Wastewater
22. A joint industrial - domestic flow diagram Indicating potential economic advantages.
-------
SECTION »
NATIONAL IMPACT ASSESSMENT
PRESENT AND FUTURE MARKET
The primary market areas for porous biomass support systems appear to
fall into two categories. The first is for upgrading of existing secondary
plants, primarily those using activated sludge or one of its modifications.
PBSS's might be considered as an add-on to the front end to attain compliance
in carbonaceous BOD removal or as an add-on to the rear end to provide nitri-
fication. It is known from summary reports (1, 2) that approximately 25 to 30
percent of all municipal and industrial secondary treatment plants in the U.S.
are experiencing significant problems in meeting their permit conditions.
Further, there were about 10,300 secondary and advanced secondary municipal
treatment plants and over 40,000 industrial treatment plants operating in 1982
(3). The number of municipal plants in these categories is expected to increase
to about 18,400 by the year 2000, primarily through upgrading of existing
primary plants to secondary status. These numbers indicate that a tremendous
market exists for inexpensive and reliable upgrading techniques, ani PBSS's
certainly represent one approach for supplying this need.
The second large potential markst is for pretreatoent of industrial
wastewaters, especially high-strength ones, for a variety of reasop.s. Some
specific applications would be the following uses of the PBSS technology:
1. Before discharge to municipal sewers to decrease user charges.
*2. Before land treatment systems to provide a low-cost "roughing"
treatment, alleviating problems with biological clogging of the
86
-------
infiltrative surface while conserving heat (important for cold-
temperature operation).
3. Before RBC units and biological towers, with plastic media to reduce
growth on these units and avoid structural failures.
4. Before activated sludge systems operating at high F/M ratios, to
alleviate filamentous bulking problems.
MARKET PENETRATION POTENTIAL
The extent to which porous biomass support systems penetrate the broad
market available in the U.S. will probably depend on the future success of
demonstration projects that are on-line or planned for the near future. Past
efforts to prove the value of Captor in the U.S. have not been particularly
successful (see page 28), and the more successful experience in the U.K. (see
page 41) has, until this assessment, been withheld from potential users in the
U.S. Similarly, the more promising data from Linpor systems being tested in
West Germany has not reached the U.S. as yet. At this point in time, very few
U.S. engineers have even heard of PPSS's, and among those that have, there is a
great deal of understandable skepticism regarding its potential. Given the
limitations of the U.S. pilot-plant trials at Marion, Illinois and Downingtown,
Pennsylvania, and the quality of the data generated at these installations, an
engineer would have to be relatively optimistic to commit to Captor or Linpor
without some pilot-plant work. Potential users of PBSS's will often be
constrained by budget, time-frame or both, and forced to choose from among
those processes which appear to h*no risk associated with them.
RISK ASSESSMENT
A rational evaluation of risk must begin with the proviso which was
discussed on page 81—there must be adequate preliminary and primary treatment
of the wastewater to remove materialc which can physically interfere with pad
cleaning (Captor) and retaining screens (Captor and Linpor). Failure to
provide adequate facilities at the head end of the plant will result in a
substantial risk of system malfunction and can be considered a- minor dis-
advantage of PBSS technology as compared with some, but not all, conventional
technologies.
87
-------
The first area with some associated risk is that of the mixing charac-
teristics of the aeration system. Although guidance is rapidly evolving
through the pilot and field studies, there presently are no infallide design
guidelines to guarantee proper pad mixing and distribution. Given the
importance of these aspects of operation, this is a serious deficiency,
especially when considering the adoption of PBSS technology in an existing
aeration basin. The owner cannot always be expected i.u bc^r the costs of
optimizing mixing patterns by trial and error. Another area of risk is the
uncertainty of the role of secondary clarification when a PBSS is used alone
and not coupled to a downstream activated sludge process. There are not
enough full-scale results upon which to confidently size a clarifier for all
the process modifications which might be considered. Omission of a secondary
clarifier from a process flow scheme utilizing Captor as the sole biological
treatment process would be a serious mistake at this stage of process devel-
opment. Although claimed or implied as an advantage of Captor in some early
literature on the process, omission of a secondary clarifier is definitely not
recommended at this time. Exceptions are noted for industrial pretreatmenc
and perhaps also for nitrification. On the other hand, pilot plant experience
in Vest Germany with Linpor N indicates that this process may succeed without
the use of final clarifiers. Less risk would be associated with I/A funding
of full-scale demonstrations of this concept.
Some areas of minor risk which need to be addressed in further research
programs include the impact of DO concentrations on process performance, the
effects of certain biological growths (i.e., fungi) on PBS particles with
respect to pad cleaning and biological activity, and factors affecting biomass
yields from these processes.
Although these areas of risk are of concern and require additional study
and development, they should not be allowed to jeopardize the evolution of the
technology. As described in previous sections of this report, there are
several inherent advantages of PBSS technology, some already demonstrated and
some potential but not as yet demonstrated. The technology has proven
successful in certain specific applications, for example:
88
-------
1. Industrial wastewater pre-treatment
2. Upgrading an activated sludge system plagued by high SVI
3. Nitrification
On the other hand, the technology has not fulfilled all of the expectations of
its developers. For example, it seems poorly suited for use with weak waste-
waters when nitrification is not an objective and has not beea capable of
meeting effluent objectives without a final clarifier when employed as the
sole biological process unit.
89
-------
REFERENCES
SECTION 5
1. British Patent Application No. 38267/78 and similar foreign applications,
1978.
2. U.S. Patent No. 4,419,243, Atkinson et al., December 6, 1983.
3. Walker, I. and E. P. Austin, "The Use of Plastic, Porous Biomass Supports
in a Pseudo-Fluidized Bed for Effluent Treatment," In Biological Fluidized
Bed Treatment of Water and Wastewater (Ed. P.F. Cooper and B. Atkinson).
Ellis Horwood Ltd., Chichester, Sussex, U.K., Chapter 16, 1981.
4. Austin, E.P. and I. Walker, "Process Intensification by Captor System,"
Conference on Uprating Water and Wastewater Systems, London, U.K., October
11, 12, 1983
5. Fuchs, U., "The Linpor Process for Biological Wastewater .Treatment"
("Linpor-Verfahren zur biologischen Abwasserreinigung"), Berichte der
Abwassertechnischen. Verp-igigung, e.v. Nr 34. 457. 1982.
6. Hegemann, W. and E. Englmann, "The Activated Sludge Process Using Foam
Pellets to Increase Biomass Concentration" (Belebungsverfahren mit
Schaumstoffkorpern zur Aufkonzentrierung von Biomasse"), GWF-
Wasser/Abwasser, Nr. 5, 233, 1983.
7. Wolf, 11., "Linpor-H Process Pilot Plant Studies at Munich East Wastewater
Treatment Facility Neufinsing W. Ger.," Bayer. Landesamt fur
Wasserwirtschaft, Jan. 1985.
8. Reimann, H., "The LINPOR Process with Fixed Biomass on Plastic Foam -
Practical Aspects and Results," Use of Fixed Biomass for Water and
Wastewater Treatment, 37th International Conference CEBEDEAU Liege, Fr.,
353, May 1984.
9. Hegemann, W., "A Combination of the Activated Sludge Process with Fixed
Film Biomass to Increase the Capacity of Wastewater Treatment Plants,"
IAWPRC Workshop on Large Wastewater Treatment Plants, Vienna, Austria,
October 1983.
10. Simon-Hartley, Ltd., "Principles of Design for a Captor Unit," Stoke-on-
Trent, U.K. (1983).
11. Simon-Hartley, Ltd., "Proposal for the Uprating of Freehold Effluent
Treatment Plant," WRC Project No. 1800, Stoke-on-Trent, U.K. (1982).
12. Stenquist, R. J. et al., "The Coupled Trickling Filter - Activated Sludge
Process: Design and Performance," EPA-600/2-78-116, July 1978.
13. Harrison, J. R. et al., "A Survey of Combined Trickling Filter and
Activated Sludge Processes," Jour. Wat. Poll. Cont. Fed. 56 (10):1073-79,
October 1984.
90
-------
14. Simon-Hartley, Ltd. "Captor Assessment - Freehold E.T.P.," Captor Project,
September 1984.
SECTION 6
1. Atkinson, B. and N.W.F. Kossens, "Fermenter Design and Modeling of
Continuous Processes," Proceedings of First European Congress of
Bacteriology, Dechema Monographien, Vol. 62, 37, 1978.
2. Atkinson, B., G.M. Black, P..I.S. Lewis, and A. Pinches, "Biological
Particles of Given Size, Shape, and Density for Use in Biological
Reactors," Biotechnology and Bioengineering, Vol. XXI, 193, 1979.
3. Perez, F., "Measurement of the Ideal Microbial Film Thickness for Product
Formation," M.Sc. Dissertation, University of Manchester, U.K., 1977.
4. Atkinson, B. and P.J.S. Lewis, "The Development of Immobilized Fungal
Particles and their Use in Fluidized Bed Termenters," British Mycological
Society Symposium Series No. 3 - Fungal biotechnology (Eds. J.E. Smith,
D.R. Berry and B. Kristiansen), Academic Press, 153, 1980.
5. Atkinson, B., G.M. Black, and A. Pinches, "The Characteristics of Solid
Supports and Biomass Support Particles When Used in Fluidized Beds," In
Biological Fluidized Bed Treatment of Water and Wastewater (Eds. P.F.
Cooper and B. Atkinson), Ellis Horwood Ltd., Chichester, Sussex, U.K.,
Chapter 5, 75, 1981.
6. Webb, C., G.M. Black, and B. Atkinson, "Liquid Fluidization of Highly
Porous Particles," Chem. Eng. Res. Pes., Vol. 61, 125, March 1983.
7. Ergun, S. "Fluid Flow Through Packed Columns." Chem. Eng. Progr.,
Vol. 48, 91, 1952.
8. Walker, I. and E.P. Austin, "The Use of Plastic Porous Biomass Supports in
Pseudo-Fluidized Bed for Effluent Treatment," In Biological Fluidized Bed
Treatment of Water and Wastewater (Eds. P.F. Cooper and B. Atkinson) Ellis
Horwood Ltd., Chichestre, Sussex, U.K., Chapter 16, 272, 1981.
9. Black, G.M., C. Webb, T.H. Matthews, and B. Atkinson, "Practical Reactor
Systems for Yeast Cell Immobilization Using Biomass Support Particles,"
Biotechnology and Bioengineering, Vol. XXVI, 134, 1984.
10. Black, G.M. and C. Webb, "A Practical Technology for Cell Immobilization
Based on Novel Biomass Support Particle," Vlth Australian Biotechnology
Conference, Brisbane, Australia, September 1984.
11. Richardson, S.R. and Wilson, F.V., "Rapid Scanning Electron Microscope
Techniques to Investigate Colonisation of Biomass Support Particles in
the CAPTOR Process," Env. Tech. Lett., Vol. 4, 183, 1983.
12. Fynn, G.H. and T.N. Whitmore, "Colonisation of Polyurethane Reticulated Foam
Biomass Support Particle by Methanogen Species," Biotechnology Letters,
Vol. 4, No. 9, 577, 1982.
91
-------
13. da Fonseca, M.M.R., "Assessment of the inaction Properties of Biomass
Support Particles Using AcetobacterPh.D. Dissertation, University of
Manchester, U.K., 1983.
14. Fukoda, H., C. Webb, B. Atkinson, "Continuous Cellulase Production in a
Spouted Bed Fermenter Using Cells Immobilized in Biomass Support
Paricles," Third European Congress on Biotechnology, Munich, 1984.
15. Hegemann, W., "A Combination of the Activated Sludge Process with Fixed
Film Biomass tc Increase the Capacity of Wastewater Treatment Plants,"
IAWPR Workshop on Large Wastewater Treatment Plants, Vienna, Austria, 1983.
16. Hegemann, W. and £. Englmann, "The Activated Sludge Process Using Foam
Pellets to Increase Biomass Concentration," GWF - Wasser/Abwasser, Vol.
124, No. 5, 233, 1983.
17. Reimann, H., "The Linpor Process With Fixed Biomass on Plastic Foam -
Practical Aspects and Results," Use of Fixed Biomass for Water and
Wastewater Treatment, 37th International Conference CEBEDEAU, Liege,
France, 353, May 1984.
18. Atkinson, B., J.D. Cunningham and A. Pinches, "Biomass Hold-Ups and
Overall Rates of Substrate (Glucose) Uptake of Support Particles
Containing A Mixed Microbial Culture," Chem. Eng. Res. Des., Vol. 62,
156, 1984.
19. Atkinson, B., Biochemical Reactors, Pion Press Ltd., London, U.K., 1974.
20. Report to Support Design of Freehold Pilot Plant, Appendix B, Original
Work on Captor, March 8, 1983.
21. Reiber, S. and H.D. Stensel, "Biological Enhanced Oxygen Transfer in a
Fixed Film System," Jour. Wat. Poll. Cont. Fed. (In Press).
SECTION 7
1. Hegemann, W. Technical University of Munich at Garching. Unpublished
data (1984).
SECTION 8
1. Ucbrin, C. G. and R. L. Caspe. Operational characteristics of federally
funded wastewater treatment plants. Jour. Wat. Poll. Contr. Fed.
55(7):935-40. July 1983.
2. Schaeffer, D.J., et al. Municipal compliance - another view. Jour.
Wat. Poll. Contr. Fed. 56(8):924-27. August 1984.
3. Fact sheet for wastewater treatment. Jour. Wat. Poll. Contr. Fed.
54(10):1346-48. October 1982.
92
-------
DATE DUE
j
------- |