Louisiana State University LSU Digital Commons
LSU Historical Dissertations and Theses Graduate School
1988 Development and Design of a Fluidized Bed/ Upflow Sand Filter Configuration for Use in Recirculating Aquaculture Systems. Daniel G. Burden Louisiana State University and Agricultural & Mechanical College
Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses
Recommended Citation Burden, Daniel G., "Development and Design of a Fluidized Bed/Upflow Sand Filter Configuration for Use in Recirculating Aquaculture Systems." (1988). LSU Historical Dissertations and Theses. 4619. https://digitalcommons.lsu.edu/gradschool_disstheses/4619
This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected]. INFORMATION TO USERS
The most advanced technology has been used to photo graph and reproduce this manuscript from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are re produced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. These are also available as one exposure on a standard 35mm slide or as a 17" x 23" black and white photographic print for an additional charge.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Order Number 8917798
Development and design of a fluidized bed/upflow sand filter configuration for use in recirculating aquaculture systems
Burden, Daniel G., Ph.D.
The Louisiana State University and Agricultural and Mechanical Col., 1988
300 N. Zeeb Rd. Ann Arbor, MI 48106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DEVELOPMENT AND DESIGN OF A FLUIDIZED BED/UPFLOW SAND FILTER CONFIGURATION FOR USE IN RECIRCULATING AQUACULTURE SYSTEMS
A Dissertation
Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy
in
The Department of Civil Engineering
by Daniel G. Burden B.S., Harding University, 1978 M.E., Louisiana State University, 1985 December 1988
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS
This research was supported by the Louisiana Sea Grant College
Program, an element of the National Sea Grant College Program, under the
auspices of the National Oceanic and Atmospheric Adminstration, U.S.
Department of Commerce. Special thanks goes out to Mr. Ron Becker,
Associate Director of the Louisiana Sea Grant College Program, who has
continually supported this research effort throughout the past two
years. Invaluable advice was received from Drs. Dudley D. Culley and
Leon Duobinis-Gray at LSU's School of Forestry, W'ildlife, and Fisheries.
Much of the initial research depended on crawfish from off-season
production ponds provided by these individuals. The author would also
like to acknowledge Herman "Bubby" and Steve Oufnac of Baton Rouge,
Louisiana, whose commercial facility was used for verification testing
of design criteria on full-scale filter units. Crabs and invaluable
technical advice were provided by Mr. Cultus Pearson through his
commercial operation in Lacombe, Louisiana. Special thanks also goes
out to Ms. Ann Mannino and Mr. Jim Robin who gave invaluable laboratory
and field assistance.
ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FORWARD
I would like to extend my sincere thanks to Dr. Ron Malone for his
assistance, guidance, and encouragement as my major professor in
addition to being a good friend. The numerous accomplishments I have
achieved over the past six years in no doubt reflect our working
relationship. I would like to extend my thanks to the other members of
the committee, Drs. Flora Wang, Marty Tittlebaum, Dipak Roy, and Dudley
Culley.
I would like to thank my brother and close friends whom I have
enjoyed working with over the past years: David S. Burden 'soon to be
Dr. David S. Burden), Dr. Constantine "Dean" Mericas, Don Manthe,
Kevin Cange, and Paul Gremillion. I also would like to acknowledge
members of the research group: Dr. Walter "Wink" Zachritz, deEtte
Ferguson, and Kelly Rusch, for their support and editorial assistance
over the last year. I also owe a great deal of gratitude to Dr. David
N. Richardson, a former employer and now a faculty member at the
University of Missouri at Rolla, whose encouragement to attend
graduate school in civil engineering was a turning point in my career.
Most of all, I am deeply indebted to my "closest" friend— my wife,
Betsy. Without her love, understanding, and constant companionship,
this accomplishment would have never been possible.
Some individuals have referred to this occasion as the end of a
"dynasty", but in the words of T.S. Eliot "In the end is my new
beginning". Well, call it whatever you want, I'm glad its over (and
so is my wife). Farewell Tiger fans.
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ...... ii
FORWARD ...... iii
LIST OF TABLES ...... vii
LIST OF FIGURES ...... viii
ABSTRACT ...... xi
CHAPTER
I. Introduction ...... 1
II. The Effect of Sand Grain Size on Fluidized Bed Filter Performance in Recirculating, Soft-Shell Crawfish Production Systems ...... 6
Ab s t r a c t...... 6
Introduction ...... 7
Background ...... 9
Methodology ...... 12
Results ...... 18
Fine Sand (20/40) 18
Medium Sand (12/20) 21
Coarse Sand ( 8 / 1 6 ) ...... 24
Discussion ...... 26
Conclusions and Recommendations ...... 29
III. Evaluation of a Fluidized Bed/Upflow Sand Filter Combination Proposed for a Recir culating Blue Crab Shedding S y s t e m ...... 31
Abstract ...... 31
Introduction ...... 32
Background...... 33
Nitrification ...... 33
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER Page
Oxygen Consumption ...... 34
Fluidized Bed Filter ...... 36
Upflow Sand Filter ...... 37
M e t h o d s ...... 38
Filter Studies ...... 39
Waste Characterization Study ...... 45
Results . 4 7
Filter Studies ...... 47
Fluidized Bed/Upflow Sand Filter Combination ...... 48
Fluidized B e d ...... 50
Upflow Sand F i l t e r ...... 54
Excretion S t u d y ...... 56
D iscussion ...... 56
Oxygen Consumption ...... 58
Nitrification ...... 62
Relative Filter Ability ...... 65
Summary ...... 66
Conclusions ...... 67
IV. Fluidized Bed/Upflow Sand Filter Performance in a Recirculating Soft-Shell Crawfish Facility ...... 69
Abstract ...... 69
Introduction . . 70
Background ...... 71
Commercial Facility Description...... 80
Methodology...... 82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER Page
Results ...... 84
Discussion...... 89
Conclusions and Recommendations ...... 94
BIBLIOGRAPHY ...... 96
APPENDICES ...... 102
A. Water Quality Data ...... 103
1. Grain Size Determination D a t a ...... 104
2. Filter Evaluation Data ...... 116
B. Crab Waste Characterization Data ...... 128
C. Filter Fluxrate D a t a ...... 130
D. Commercial Facility D a t a ...... 137
VITA ...... 140
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES
Table Page
1 Water quality parameters and methods used with the filtration research ...... 17
2 Analytical data summary for the steady-state analyses using the fine, medium, and coarse s a n d s ...... 20
3 Dimensions of the experimental setup used for determining filter carrying capacity ...... 41
4 Water quality methods and instruments used with the filtration research ...... 46
5 Steady-state data obtained on the fluidized bed (FB) and upflow sand (UFS) filters when running in parallel operation ...... 51
6 Steady-state data obtained on (a) the single fluidized bed filter and (b) the single upflow sand filter ...... 53
7 Statistical summary for all waste characterization data collected on the blue c r a b ...... 57
8 Major components of a recirculating shedding system ...... 74
9 Summary of interim design criteria for shedding systems employing fluidized bed/upflow sand filter configurations (taken from Malone and Burden, 1988) ...... 75
10 Summary of water quality and field data collected at the commercial soft-shell crawfish facility in Baton Rouge, Louisiana. All measurements of water quality data were taken in sump # 2 ...... 86
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES
Figure Page
1 Experimental apparatus used with the grain size determinations ...... 13
2 Detailed schematic of the fluidized bed filter (not to scale) ...... 14
3 Loading, ammonia, nitrite, effluent dissolved oxygen, and pH data observed on the fine (20/40) grain sand when used as a media with the fluidized bed fi l t e r ...... 19
4 Flowrates expressed as a percentage of the initial flowrate observed on the fine, medium, and coarse sands ...... 22
5 Loading, ammonia, nitrite, effluent dissolved oxygen, and pH data observed on the medium (12/20) grain sand when used as a media with the fluidized bed f i l t e r ...... 23
6 Loading, ammonia, nitrite, effluent dissolved oxygen, and pH data observed on the coarse (8/16) grain sand when used as a media with the fluidized bed f i l t e r ...... 25
7 A schematic of the experimental setup used with the three filter configurations. The fluidized bed/upflow sand filter combination and the single upflow sand filter (circumvented by the dashed line) configurations are denoted in (a). The single fluidized bed filter is illustrated in (b) ...... 40
8 A detailed illustration of the fluidized bed filter used with the study (not to scale) ...... 42
9 Normal and intermittent operation of the upflow sand filter (not to scale) ...... 44
10 Water quality monitoring results and loading curves obtained on the fluidized bed/upflow sandfilter combination ...... 49
11 Monitoring results obtained on the fluidized bed filter using floss pre-filtration...... 52
vm
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure Page
12 Loading curves and water quality monitoring data obtained on the single upflow sand filter con figuration ...... 55
13 Gamma distributions for OLR data on the fluidized bed and submerged rock filter configurations. Mean values are indicated in parentheses and denoted by the horizontal lines on each distribution ...... 59
14 Pie diagrams illustrating the representative portions of total OLR on the upflow sand and fluidized bed filters ..... 61
15 Results of the linear regression analyses performed on total nitrogen accumulation in the three filter configurations. R-square values indicated in parentheses ...... 64
16 Components of a recirculating system for a soft-shell crawfish facility ...... 73
17 Unpressurized fluidized bed filter designs for diameters ranging from 10 to 20 inches (25 to 50 cm) ...... 76
18 Unpressurized upflow sand filter designed for diameters ranging from 10 to 20 inches (25 to 50 c m ) ...... 78
19 Schematic of the commercial soft-shell crawfish facility located in Baton Rouge, L o u i s i a n a ...... 81
20 Total ammonia, nitrite, and pH monitoring data obtained on the commercial soft-shell facility in Baton Rouge, LA. Filter design capacity (150 lbs/ft3) indicated by the vertical line ...... 87
21 Total ammonia and nitrite removal efficiency data obtained on the commercial facility in Baton Rouge LA. Solid diagonal lines represent removal efficiencies ...... 88
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure Page
22 Fluidized bed and upflow sand filter OCF observations taken at the commercial facility. Reported values for the upflow sand filter represent a mean with upper and lower standard deviations for the three filters at a given filter loading ...... 91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT
A fluidized bed/upflow sand filter configuration, was developed and
designed for utilization in recirculating aquaculture systems, specifi
cally the soft-shell crab and soft-shell crawfish industries. These
filters were selected and designed because of their ability to withstand
clogging and still maintain high levels of water quality for aquaculture
production. The effectiveness of sand grain size was used to evaluate
fluidized bed filter performance with filter loadings ranging from 16 to
1285 pounds of crawfish per cubic foot of filter sand. A coarse sand
grain size was recommended as a filter media because of it's ability to
shear excessive biofilm growth from the sand, thus prohibiting clogging
from occurring within the filter bed. The fluidized bed/upflow sand
filter combination was evaluated in terms of nitrification and oxygen
consumption when used with a recirculating crab shedding system. The
filter combination's carrying capacity (700 crabs per cubic foot of sand
media) exceeded that observed with the submerged rock filter by more
than 20 times and was largely explained by the filter's solids removal
ability which significantly reduced the filter's oxygen loading rate
(OLR). Nitrification rates with the filter combination were extremely
high as total ammonia and nitrite levels remained below 1.0 mg-N/L.
Verification of a volumetric loading criteria (150 pounds per cubic
foot) for this filter combination was further established with
performance data obtained from a commercial soft-shell crawfish
facility. Water quality monitoring results indicated that the filters
maintained total ammonia and nitrite levels below 1.0 mg-N/L under
typical operating conditions. Shock loading, pH control, and over-
xi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. feeding, rather than filter capacity, dominated water quality fluctua
tions, thereby indicating that the loading criteria was sufficient for
commercial operation.
xii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER I
INTRODUCTION
The Gulf Coast states have recently experienced a re-growth and
re-expansion of the soft shell crab industry as reported landings have
increased tremendously over the last few years. Most of this increase
was attributed to the development of engineering designs for closed,
recirculating systems to hold and shed peeler crabs. These systems
centered on using low-rate biological treatment, such as the submerged
rock filter.
More recent activities have centered on culturing soft-shell
crawfish as a potential aquaculture industry in Louisiana and other Gulf
states. This past year (1988) marked the first year when several
large-scale producers, ranging from 500 to 2000 pounds, went into
commercial operation using flow-through systems. However, economic
constraints and the lack of reliable design criteria for maintaining
adequate water quality in shedding systems prohibited the initial
growth of the industry. Consequently, efforts focused on using new
technology, specifically recirculating systems, for producing the
soft-shells.
The low-rate submerged rock filter used with the crab industry was
limited in its inability to withstand clogging and still maintain
adequate water quality during periods of high loadings. For the soft-
shell crab industry to maintain its current growth and for the newly
developed soft-shell crawfish industry to increase its efficiency,
other filter designs were necessary to provide filters of a reasonable
size which would still maintain water quality. Consequently, research
■1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. efforts were directed toward developing high-rate biological filters
that both industries could incorporate into recirculating designs.
Dissertation Objective and Approach
This dissertation consists of three manuscripts, each describing an
integral part of a two-year research project directed toward developing
and designing high-rate biological filters for recirculating systems.
The first manuscript focuses on the fluidized bed, a high-rate biologi
cal filter. Because of the limited amount of research conducted on the
filter, inconclusive data were available concerning the appropriate sand
grain size to employ as a filter media. Consequently, the initial
experimental work centered on investigating several sand grain sizes to
substantiate design criteria. Filter testing was carried out using
scaled-down recirculating systems for holding red swamp crawfish
(Procambarus clarkii). Chapter II, "The Effect of Sand Grain Size on
Fluidized Bed Filter Performance in Recirculating, Soft-Shell Crawfish
Production Systems", describes the results of this research. This
manuscript has been accepted for publication in The Journal of
Aquacultural Engineering.
The second portion of the research involved determining the most
effective filter design to use with a recirculating system. Following a
comprehensive literature search, a combination of the fluidized bed and
upflow sand filter appeared most appropriate for maintaining adequate
water quality in both soft-shell crab and crawfish shedding systems.
The upflow sand filter was added to the filter configuration because of
its ability to remove solids and consequently reduce BOD loading in the
system. Therefore, filter testing centered on analyzing this con
figuration, as well as its individual components, to determine filter
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. performance in terms of both nitrification and oxygen stability. Filter
evaluations were performed with blue crabs (Callinectes sapidus) , those
commonly used for shedding operations in the Gulf Coast region. In
addition to this study, a separate waste characterization analyses was
conducted to determine the relative effect of solids removal on
crab-excreted wastes. Chapter III, "Evaluation of the Fluidized
Bed/Upflow Sand Filter Combination Proposed for a Recirculating Blue
Crab Shedding System", describes the results of this work and makes
recommendations as to why the filter combination is well suited for
recirculating system designs.
Finally, Chapter IV, "Fluidized Bed/Upflow Sand Filter Performance
in a Recirculating Soft-Shell Crawfish Facility", presents verification
data based on the developed design criteria. Filter designs depended
upon several different aspects: (I) fed or unfed systems, (2) filter
media and size, (3) pumping requirements, (4) filter hydraulics, and
(5) filter carrying capacity. Design verification was carried out in a
1600 pound commercial crawfish shedding operation located in Baton
Rouge, Louisiana. The relative ability of the filter combination was
tested and analyzed in terms of both nitrification and dissolved oxygen
requirements and covered in Chapters III and IV. Both chapters have
also been submitted for publication in The Journal of Aquacultural
Engineering.
Conclusions
The high-rate fluidized bed and upflow sand filter combination was
proven as an extremely effective method to achieve water treatment in a
recirculating aquaculture system. Filter carrying capacities exceeded
those historically observed with the submerged rock filters by more than
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4
20 times while still maintaining adequate water quality for shedding
operations. Total ammonia and nitrite levels remained below 1.0 mg-N/L
without experiencing any harmful impacts on molting crabs or crawfish.
The filter combination's higher efficiency was explained by the high
kinetic transport rates resulting from surface abrasion and hydraulic
distribution of blofilm in the fluidized bed and by the enhanced solids
removal ability of the upflow sand filter.
From the results of this research effort, one can readily see the
immediate impact that these high-rate filters are having on the aqua
culture industry. Because of the industry's continued growth rate,
efforts should be made to improve and enhance current system designs.
Continued research should focus on improving kinetic transport rates,
reducing water/volume ratios, and investigating the impact of CO^
accumulations, as applied to these filters. A more thorough under
standing of these concepts will increase the efficiency of the
recirculating system when utilized in the aquaculture industry.
Industrial Impact
Commercial dissemination of filter designs was initiated in April,
1988 when an advisory agent workshop, sponsored by the Louisiana Sea
Grant College Program, was conducted on the Louisiana State University
campus. Information on design of recirculating systems incorporating
the new filter technology was distributed to extension agents in
Louisiana, Florida, Minnesota, Mississippi, South Carolina, and Texas
for their use in the public sector. Consequently, the development of
recirculating soft-shell crawfish facilities in Louisiana and other Gulf
states has rapidly expanded and continues to do so. Commercial systems
ranging in size from 500 to 6000 pounds have been built and new
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5
facilities are currently in construction for the 1988-1989 season. The
new filter technologies are also being implemented in other aquaculture
areas outside the Gulf Coast region as well. These areas include Morgan
Hill Aqua Farms in California, Sea Hatcheries Limited in Australia,
Norton Brothers, Inc. in Florida (tropical fish hatcheries), Early Times
Distillery in Kentucky (catfish production), and the Public Service of
Colorado (co-generation talapia grow-out system) just to mention a few.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER II
THE EFFECT OF SAND GRAIN SIZE ON FLUIDIZED BED FILTER PERFORMANCE IN RECIRCULATING, SOFT-SHELL CRAWFISH PRODUCTION SYSTEMS*
Daniel G. Burden and Ronald F. Malone Department of Civil Engineering, Louisiana State University Baton Rouge, Louisiana 70803
ABSTRACT
The effectiveness of media size was investigated when using a
fluidized bed filter in a recirculating system for soft-shell crawfish
production. Three different sand grain sizes, 0.42-0.84 mm, 0.84-
1.68 mm, and 1.19-2.38 mm were used to evaluate filter performance with
3 3 crawfish loadings ranging from 256 to 20,600 kg/m (16 to 1285 lbs/ft ).
Results of the study indicated that the hydraulic behavior of the fine
and medium grain sands prohibits using them as filter media at high 3 crawfish loadings, typically greater than 4633 kg/m of sand 3 (300 lbs/ft ). During these high loading periods, excessive floe
developed within the filter causing the sand media to "gel" and effluent
oxygen levels to collapse. Conversely, the coarse sand exhibited no 3 3 clogging at loadings as high as 20,600 kg/m (1285 lbs/ft ). The coarse
sand's ability to shear excessive biofilm growth from the sand prohi
bited clogging from occurring within the filter bed and therefore made
it a superior media to use with a fluidized bed in a soft-shell crawfish
system.
*In Press, The Journal of Aquacultural Engineering
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7
INTRODUCTION
Soft-shell crawfish have recently been considered a seafood
delicacy although availability has been severely limited by production.
Newly developed technology, however, now permits commercial production
of soft-shell crawfish on a small scale. Total production for this
growing industry rose from an estimated 2950 kg (6500 lbs) in the
1985-1986 season to over 6800 kg (15,000 lbs) during 1986-1987 season.
Current projections indicate somewhere between 22,700 and 34,000 kilo
grams (50,000 and 75,000 pounds) will be produced during the 1987-1988
season. Production levels in excess of 454,000 kilograms (one million
pounds) per year are expected within a few years only if the industry
can overcome certain obstacles. These problems include: (1) guaran
teeing a source of immature hard crawfish and extension of season,
(2) difficulties with acclimating pond-raised or wild immature crawfish
into soft-shell facilities, and (3) the lack of reliable design criteria
for maintaining adequate water quality in a shedding system.
Industry has placed its current focus on improving water quality by
developing reliable design criteria for shedding facilities. The
earliest commercial systems employed flow-through systems for holding
and separating soft-shell crawfish. Operational costs associated with
this technology, however, are high because of the water use and
heating costs. Flow-through systems require freshwater, usually drawn
from wells or municipal water systems, at a continuous flushing rate
ranging from 0.42 to 0.83 L/min-kg (0.05 to 0.1 gpm/lb). Thus, a
small facility with a 454 kg (1000.1b) holding capacity demands 8 to
16 billion liters (17 to 35 million gallons) of heated groundwater
over an eight month season. Furthermore, the annual costs associated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8
with heating the groundwater from 21 to 27°C (70 to 80°F) will exceed
$15,000. Assuming a 2 percent daily production rate and the current
wholesale market price of $17.6/kg ($9.00/lb), a 454 kg (1000 lb)
facility can generate approximately $43,000 a year in gross income. If
heating costs account for nearly 30 percent of this income, the
economical viability of the entire operation becomes questionable.
A more cost-effective means of producing the soft-shell crawfish
can be achieved by using a closed recirculating system. These systems,
characterized by their reuse of water, contain filtration components
which process animal wastes to a relatively inert and harmless state
(Manthe et al. 1984). Annual heating costs are relatively low ranging
from $1,000 to $2,000 for an insulated recirculating system. Research
has therefore centered on developing high-rate filters which can support
exceedingly large amounts of biomass and still maintain adequate water
quality. One of the biological filters proposed for this purpose is the
fluidized bed filter.
The fluidized bed filter operates as a "fixed-film" biological
filter, but involves an upward movement of fluid, or wastewater, through
a sand bed at a velocity sufficient enough to "fluidize" the sand. Sand
has been historically used as a granular media because it is durable and
relatively inexpensive. The sand grain size defines the hydraulic
environment within a fluidized bed and thus, is the major factor con
trolling the filter's effectiveness. Therefore, research was directed
toward determining an appropriate grain size to employ with a fluidized
bed filter when used in a recirculating system. This article compares
the effectiveness of three sand grain sizes and makes recommendations on
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. an effective grain size to use with a fluidized bed filter in a recircu
lating soft-shell crawfish system.
BACKGROUND
In a typical soft-shell crawfish culture system, pond-raised
immature crawfish (Procambarus clarkii) are placed in shallow, indoor
culture trays and fed until they develop color patterns which indicate
the approach of molting (or ecdvsis). Premolt crawfish are isolated
in "molting trays" and then removed by hand within a few hours of the
molting process and immediately frozen (Culley et al. 1.985). Systems
are designed based on holding capacity (kilograms or pounds). Most
systems in Louisiana are small with holding capacities ranging from 227
to 908 kilograms (500 to 2,000 pounds). These systems currently produce
1.5 to 2.0 percent of capacity per day when properly operated at 27°C
(80°F).
While in the system, crawfish continuously excrete ammonia
(Hartenstein, 1970). This direct excretion of total ammonia associated
with the crawfish, as well as that produced from the decay of uneaten
feed, continually contribute total ammonia to the recirculating system.
Cange (1987) demonstrated that crawfish total ammonia excretion rates
vary from 58 to 86 mg-N per pound of crawfish per day depending on the
protein content of the feed. High protein foods (43 percent) can
contribute as much as 130 mg-N per gram of food if left to decay in a
system (Cange, 1987). Consequently, high mortalities have been often
associated with total ammonia and possibly nitrite accumulations in both
flow-through and recirculating systems.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10
Reported total ammonia toxicity limits for crawfish (P^ clarkii)
range from a maximum of 2.65 mg-N/L (pH=8.6), associated with juvenile
crawfish (based on 96-hr LC-50's), to as low as 1.00 mg-N/L observed
during the molting process (Hymel, 1985; Cange, 1987). Hymel (1985)
also reported mean nitrite toxicity levels of 5.94 mg-N/L (96-hr LC-
50's) on intermolt crawfish while no mortalities were observed when
concentrations were maintained at 1.00 mg-N/L. Additional nitrite
toxicity research on intermolt crawfish (|\_ clarkii) by Gutzmer and
Tomasso (1985) indicated a 96-hr LC-50 of 8.5 +0.5 mg-N/L. Observa
tions made on commercial facilities lead the authors to conclude that
molting losses become significant at levels well below the indicated
acute toxicity levels (96-hr LC-50's) because of the intermolt's
increased sensitivity and long exposure periods (up to 50 days).
Considering the lack of data on chronic toxicity levels, the authors
assumed that ammonia (total) and nitrite concentrations should be
maintained below 1.0 mg-N/L to avoid high molting losses. The success
of a recirculating system thus depends on designing filtration
components, which can adequately maintain total ammonia and nitrite
levels when used in an aquaculture setting.
Weber et al. (1970) first recognized the ability of the biological
fluidized bed filter when he used fluidized beds of activated carbon to
remove dissolved organic compounds from chemically-treated sewage.
Weber noticed the biomass growth occurring on the particles and likewise
the reduction in nitrite concentrations through the bed. Similar to a
fixed-film filter, the sand or carbon particles become evenly coated
with both heterotrophic and chemoautotrophic bacteria. Heterotrophs
obtain their energy from organic compounds (excreted by the crawfish)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11
and convert them to an inorganic state while the chemoautotrophs obtain
their energy by converting the inorganics. Nitrosomonas and Nitro-
bacter, the principal genera of the autotrophic bacteria, convert total
ammonia to nitrite and nitrite to the relatively harmless state of
nitrate, respectively. The active biomass concentration in the filter
is large because of the greater amount of surface area available for
biological growth.
The range of media sizes used with fluidized bed filters varies by
almost 200 percent. Jeris et al. (1974) employed fluidized beds for
denitrification of domestic and industrial wastewaters using sand
ranging in diameter from 0.8 to 1.1 mm as a medium for biological
growth. Short (1973) also developed a fluidized bed filter for
oxidizing low concentrations of total ammonia (0.5 - 2.0 mg NH^-N/L)
found in river water when used as a water supply. Short assumed that by
providing a larger surface area for bacterial growth more treatment
could be achieved and, therefore, used sand ranging in diameter from
0.05 to 0.15 mm. Additional research on fluidized bed filters, where
sand is employed as the granular media, has been based on particle
diameters ranging in size from 0.3 to 0.9 mm (Cooper, 1981; Nutt et al.
1984).
Selecting an appropriate grain size insures that (1) a suitable
amount of surface area is available for biological growth, (2) an
adequate oxygen supply is maintained across the filter bed, and (3)
a sufficient amount of abrasion is provided to avoid filter clogging
as a result of excessive biofilm growth. The methods used for
determining the effect of grain size on filter performance are
presented in the following section.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12
METHODOLOGY
Three independent fluidized bed filters were used to examine the
effect of grain size on the removal of total ammonia and nitrite in a
recirculating system. As implied by Standard Methods, the test for
ammonia nitrogen (NH^) measures all nitrogen in the form of NH^, NH^OH,
and NH^. Therefore, throughout the remainder of the text, reference to
'ammonia’ or 'NH^' implies 'total ammonia'.
Each system consisted of a fluidized bed reactor with a holding
tank containing a known weight of live crawfish. Live crawfish were
used to assure that the waste load imposed on each system was
representative of a commercial facility, thus assuring that the
filters would hold representative populations of both hetertrophic and
autotrophic bacteria. Crawfish were not fed while in the system so
that loading information could be accurately determined for
comparative purposes.
The holding tank consisted of a 521 mm (21.5 inches) square plastic
tub with a 330 mm depth (13 inches). A 25.4 mm (1 inch i.d.) drain pipe
was located in the tank's front center (Figure 1). Total system volume
was equivalent to 98 liters (26.0 gal.). A 250 x 500 mm sheet of
plastic louvering was placed inside the tank and elevated at a height of
approximately 70 mm to allow additional surface area for the crawfish
during periods of high loading.
Each fluidized bed filter was constructed using 38.1 mm (1.5 inch
i.d.) diameter clear PVC tubing (762 mm long) with appropriate PVC
fittings on the intake and outlet similar to those illustrated in
Figure 2. Recirculation of water through the filter was maintained
using a Cole-Palmer 1/25 HP centrifugal pump with magnetic drive.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13
TO AIR PUMP
M S*ne FLUIDIZED MEDIA HOLDING TANK
PUMP
Figure 1. Experimental apparatus used with the grain size determinations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14
19 mm (3 /4 " )
DISCHARGE BACK TO SYSTEM
3 8 mm
100 PERCENT EXPANSION
PACKED HEIGHT
— 2 5 m m ( I ) OF SUPPORT PLATE GRADED GRAVEL 3.2 mm HOLES
SCREENED IN FLOW
BALL VALVE SECTION A-A
Figure 2. Detailed schematic fo the fluidized bed filter (not to scale).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15
Flowrates were based on maintaining 100 percent expansion of the bed at
all times by using a 1.9 cm (3/A") ball valve on the filter's intake
side. Aeration was provided to the system by using two Whisper 600
pumps with airstones placed in each holding tank.
Three different sand grain sizes were used for comparative work in
this study: (1) a 1.19-2.38 mm coarse filter sand (graded to pass a #8
mesh and retain on a #16 mesh screen), (2) a 0.8A-1.68 mm medium filter
sand (graded to pass a #12 mesh and retain on a #40 mesh screen) , and
(3) a 0.42-0.84 mm fine filter sand (graded to pass a #20 mesh and
retain on a #40 mesh screen). All sands were pre- washed with ammonia-
free dechlorinated water to remove excess debris. Packed bed height of
each filter was equivalent to 30.5 cm (1 ft.) corresponding to a total 3 3 media volume of 347 cm (0.012 ft ). Because the fluidized bed operates
in a turbulent environment, excreted solids (which are continuously
released by the crawfish) remain within the system. Therefore, a small 3 floss bed filter (volume = 709 cm ) was positioned at the effluent
discharge and changed on a daily basis. The floss bed, which removed
excreted solids, was used to simulate commercial operations which employ
some means of solids removal in the treatment scheme.
Experimental protocol consisted of filling each tank with 41 liters
(10.83 gal.) of ammonia-free, dechlorinated tap water. A known "wet
weight" of crawfish, ranging in carapace width from 18 to 20 mm
(14 + 1 gram), were then loaded into each system. Crawfish were not fed
while in the experimental tanks and were replaced every 48 to 72 hours
with crawfish taken from a separate 7600 liter (2000 gallon) recircu
lating holding system where feeding was conducted. Purina Trout Chow
Checkerettes #4 (w/ 43 percent protein) were used for feeding in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16
support system. Ammonia excretion rates for these crawfish were
estimated at 36 mg-N/lb-day (Cange, 1987). Mortalities were removed
every 3 to 4 hours and replaced with healthy crawfish from the main
holding system. Total system weights were checked once a day and
adjusted accordingly. All systems were maintained at 24 + 1°C.
Physical and chemical parameters were monitored daily between
9:00 am and 11:00 am. Parameters included pH, dissolved oxygen (DO),
temperature, flowrate, ammonia nitrogen (NH^-N) and nitrite nitrogen
(NOn-N). When each system stabilized at a given loading of crawfish
weight, a "steady-state" analysis was conducted to indicate typical
filter performance. Steady-state analyses included biochemical oxygen
demand (BOD), and alkalinity (ALK) in addition to those parameters run
on a daily basis. At the end of a steady-state analysis, each system
was completely drained; replaced with fresh ammonia-free, dechlorinated
tap water; and adjusted for optimum pH by adding sodium bicarbonate.
The in-tank crawfish weight was then increased and maintained until the
next steady-state condition. The methods and instrumentation used with
each analysis are outlined in Table 1. All analyses were conducted
according to procedures outlined in Standard Methods (APHA, 1985).
Bicarbonate addition was used to maintain pH levels (7.5-8.3), as
well as alkalinity levels (>100 mg/L as CaC03), in an optimum range for
nitrification (Paz, 1984; USEPA, 1975). Below a pH of 7.0, a reduction
in nitrification occurs but does not cease completely (Hirayama, 1970).
pH levels were therefore adjusted at the beginning of each steady-state
analysis by chemical adding 2.4 to 4.8 mg/lb of sodium bicarbonate
(NaHC03) per day (Malone and Burden, 1987). All adjustments were made
following each day's sampling routine.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1. Water quality parameters and methods used with the filtration research.
Water Quality Parameter Analysis Method
Total Ammonia Nitrogen Distillation, Nesslerization
Nitrite Nitrogen Sulfanilamide-based Colorimetric
Nitrate Nitrogen Orion Nitrate Ion Electrode - Model 93-07, Orion Double Junction Reference Electrode - Model 90-02
Total Kjeldahl Nitrogen Digestion, Distillation, Nesslerization
Biochemical Oxygen Winkler Method Demand
Alkalinity Titration/Potentiometric
Total Suspended Solids Gravimetric
Dissolved Oxygen Yellow Springs Instrument Model 57 Oxygen Meter
PH Cole-Palmer Mini pH Meter
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18
RESULTS
Filter performance between the three grain sizes was evaluated
based on relative nitrification ability, maintaining adequate dissolved
oxygen in the effluent water, and maintaining adequate pH. A comparison
of the three grain sizes was made based on a crawfish loading rate of 3 3 4633 kg/m (289 lbs/ft ), the highest load successfully sustained by all
three sands.
Fine Sand (20/40). A total of four steady-state analyses (June 16,
July 3, July 22, and August 7), with loadings ranging from 256 to 3 3 4633 kg/m (16 to 289 lbs/ft ), were performed on the fine media. Time
trace data for both daily ammonia and nitrite data are compared with
corresponding loadings in Figure 3. Steady-state values for ammonia and
nitrite all increased with carrying capacity (Table 2). At a loading of 3 3 4633 kg/m (289 lbs/ft ), steady-state ammonia and nitrite concentra
tions were equivalent to 1.26 and 0.08 mg-N/L, respectively. The
volume/weight ratio, defined as the ratio of total water volume to the
crawfish weight (loading) held in the system, was equivalent to
25.5 L/kg (3.0 gal/lb). The previous loading rate of 946 kg/m^ 3 (59 lbs/ft ) displayed moderate steady-state ammonia levels
(0.62 mg-N/L) and non-detectable nitrite concentrations.
Monitoring was only performed for a 12-day period following the 3 steady-state analyses at the maximum loading of 4633 kg/m (August 7). 3 System failure occurred with a capacity equivalent to 9266 kg/m 3 (578 lbs/ft ) when the excess floe development caused the filter to
"gel" and oxygen levels in the effluent water collapsed. The dramatic
increase in ammonia and nitrite concentrations (11.99 and 0.68 mg-N/L,
respectively) on August 16 reflects this two-fold increase in loading.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19
pH
O X 0 0
» 7 -«£ 6
000
3 00
400
500
100
E2AY29 JU5I0 JUS) 50 JU L13 J u t 30 &UGI3 AUG 31 K P T 13 SEPT 30 SAUPLB40 DATE
Figure 3 Loading, ammonia, nitrite, effluent dissolved oxygen, and pH data observed on the fine (20/40) grain sand when used as a media with the fluidized bed filter..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 537 21.4 20,600 • 49 964 1285 15,454 10 12 263 723 1.68 2.15 1.61 0.47 0.46 0.30 1.45 7.50 23.8 11,590 7 10 540 304 578 7.35 7.75 7.90 8.2 23.3 24.0 23.0 1.78 1.38 1.23 7.05 0.19 0.19 23.1 9266 4 9 86 242 289 7.85 6.35 6.06 5.71 6.04 6.31 1.16 2.64 0.12 23.4 0.08 0.69 1 1 4 2 79 102 348 59 7.40 5.67 0.01 0.16 0.78 3.19 23.7 23.7 0.01 <0.01 • • . 30 23.97.000.77 23.4 7.40 1.02 8.00 0.61 0.62 1.26 5.80 0.64 0.68 1.03 1.96 23.9 23.7 23.7 3.22 0.540.02 0.64 < < 0.01 • . . 1 4 •• , , • 5.49 7.600.47 6.85 24.1 0.55 7.50 6.95 7.50 7.15 24,0 23.9 23.9 0.49 256 475 946 4633 denotes no data to B0D5 (mg/L) PH NH3-N (mg/L) Flowrate (L/min) Coarse Sand: N02-N (mg/L)ALK (mg/L as CaC03) < 0.01 0.02 (lbs/ft (lbs/ft ) 16 (kg/m (kg/m )„ Temp. Temp. (°C) B0D5 (mg/L) Flowrate (L/min) 3.25 ALK (mg/L as CaC03) pH N02-N (mg/L) 0.02 HH3-N (mg/L) Temp. Temp. (°C) BOD5 (mg/L) Fine Sand: Flowrate (L/rain) 1.27 Medium Sand: ALK (mg/L as CaC03) Temp. (°C) pHNH3-N (mg/L) N02-N (mg/L) < 0.01 7.50 Loading Table 2. Analytical data summary for the steady-state analyses using the fine, medium, and coarse sands.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Although the data indicates rapid conversion of ammonia to nitrite, the
filter became inoperative when clogging occurred approximately three
days later. Flowrate during this short interval averaged less than
0.6 L/min, a 50 percent decrease from the initial flowrate observed on
June 16 (Figure. 4). Additionally, effluent dissolved oxygen concentra
tions decreased with loading as the flowrate decreased and bacterial
activity (biofilm growth) increased. Mean effluent DO remained con-
sistently below 1.0 mg/L at the maximum loading (578 lbs/ft ).
Medium Sand (12/20). Six steady-state analyses (June 16, July 3 and 22,
August 4 and 19, and September 2), with loadings ranging from 256 to 3 3 11,590 kg/m (16 to 723 lbs/ft ), were performed on the medium grain
sand. Time trace data for daily ammonia and nitrite levels in the
filter are compared with respective carrying capacities in Figure 5.
Nitrite levels in the system remained consistently below 0.50 mg-N/L for
3 the entire 113 day period. When carrying capacities exceeded 4633 kg/m
3 (289 lbs/ft ), steady-state ammonia levels generally exceeded 1.5
mg-N/L. For the duration of testing, steady-state values for nitrite
steadily increased with loading (Table 2). Steady-state ammonia,
however, was reduced by almost 20 percent when loading was increased
from 9266 to 11,590 kg/m'* (578 to 723 lbs/ft'*). The higher ammonia
3 level observed at 9266 kg/m appears to result from the system's pH
(7.05) falling below the optimum range for nitrification (7.5-8.3).
Ammonia and nitrite levels peaked around August 15 when a pump
failure occurred. The largest ammonia peak of almost 16 mg-N/L
(September 11) occurred some seven days after the final loading of 3 3 15,454 kg/m (964 lbs/ft ). Mean flowrates steadily decreased from
3.8 to less than 1.0 L/min with increased loading (see Figure 4).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ho ro COARSE (8 /1 6 ) MEDIUM (12/40) FINE ( 20/40) L O A D I N G(lbs/ft3 ) observed on the fine, medium, and coarse sands. 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1.0 0.9 0.8 0.7 0.3 0.6 0.2 0.1 0.0
£ £ -J 0.5 _! 0.4 CC Figure 4. Flowrates expressed as a percentage of the initial flowrate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23
3 o
DO.
PUUP
1000
coo
« 700
o 600
5 0 0
200
100
UAY20 JUWI3 AW 50 A l i a JUL30 &U0t3 E£PT >3 SEPT 50 CAMPLING DATE
Figure 5 Loading, ammonia, nitrite, effluent dissolved oxygen, and pH data observed on the medium (12/20) grain sand when used as a media with the fluidized bed filter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Filter failure (caused by excessive floe development and clogging)
occurred with a loading of 15,454 kg/in'* where mean flowrate decreased to
25 percent of the initial flowrate. Effluent dissolved oxygen levels
were consistently low during this period averaging less than 1.2 mg/L.
At the maximum steady-state loading (11,590 kg/mJ) , where filter
clogging was not observed, nitrite levels remained below 0.3 mg-N/L,
although ammonia levels exceeded 1.0 mg-N/L. The volume/weight ratio at
this loading was equivalent to 10.2 L/kg (1.2 gal/lb).
Coarse Sand (8/16). A total of eight steady-state analyses (June 4;
July 3 and 22; August 7 and 23; September 2, 19, and 28) were performed 3 3 with loadings ranging from 256 to 20,600 kg/m (16 to 1285 lbs/ft').
Figure 6 illustrates daily ammonia and nitrite levels with corresponding
loading values observed over the entire 129 day study period. After the
seventh steady-state analysis (15,454 kg/m ), system volume was
increased from 41 to 82 liters by plumbing a second identical tank to
the existing system and increasing the pump size from 1/25 HP to 1/10
HP. This additional volume increase was required to provide more
surface area for crawfish at the next loading level of 20,600 kg/m"* 3 (1285 lbs/ft ). By doubling the volume, the density of crawfish was 2 2 reduced by almost 35 percent (from 2.6 lbs/ft to 1.7 lbs/ft ) and the
volume/weight ratio was increased by more than 50 percent (from 0.9
gal/lb to 1.4 gal/lb). Aeration in the system was essentially doubled
by placing an additional sprayhead, which discharged effluent water back
into the system, along the tank's side. This setup also provided
additional circulation. A siphon, constructed of 38.1 mm (1.5" i.d.)
PVC, was placed across the two tanks to maintain equivalent head.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25
DO
t o <
■ N H.'N
1300 -
1200 - UGO -
1000 - - £00 - n 800- £ 3 Z TOO ' © 600 • | 5 0 0 - § 4 0 0 - “ * 3 0 0 -
£ 0 0 -
100 •
CAY 20 «A#(I3 J U N 30 JU L15 JIB. 3 0 A U 313 AUG 31 E£PTIS C2PT 3 0 8M9PUNQ DATE
Figure 6 Loading, ammonia, nitrite, effluent dissolved oxygen, and pH data observed on the coarse (8/16) grain sand when used as a media with the fluidized bed filter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Steady-state ammonia levels ranged from 0.47 to 2.15 mg-N/L while
nitrite levels ranged from below detection limits to 0.78 mg-N/L
(Table 2). As with the two previous grain sizes, the dramatic peaks in
ammonia and nitrite concentrations (13.1 and 0.68 mg-N/L, respectively)
observed on August 16 resulted from a two-fold increase in loading and a
lack of dilution volume in the system. However, the filter's ability to
quickly recover demonstrates the extreme robustness of the fluidized bed
under high loading regimes. With the exception of this loading period
and a pump failure on August 27, effluent nitrite levels remained, for
the most part, consistently below 0.50 mg-N/L. High ammonia levels,
however, were evident in the latter part of the analyses. Carrying 3 3 capacities which exceeded 4633 kg/m (289 lbs/ft ) consistently
demonstrated steady-state ammonia levels in the range between 1.6 and
2.1 mg-N/L. Analyses were abruptly halted on September 30 when the
seasonal supply of crawfish ended.
Results indicated the coarse media sustained a maximum carrying 3 3 capacity of 20,600 kg/m (1285 lbs/ft ) without experiencing filter
clogging. Mean flowrate through the filter was 5.7 L/min and demon
strated no significant reduction throughout the entire study (Figure 4).
The larger grain size appears to contribute to abrasion which shears
excessive biofilm growth from the sand thereby prohibiting clogging in
the loading regime evaluated in this experiment. Unlike the fine and
medium grain sands, effluent dissolved oxygen levels remained above
2.0 mg/L for the majority of the study reflecting, in part, the constant
flowrate maintained throughout the study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION
During normal operation, the fluidized bed expands and contracts in
response to flowrate changes and/or biomass (biofilm) growth on the sand
particles. As the biofilm thickens, the effective volume of the sand
particle increases and lowers the terminal velocity. Thus, the flow1
required for maintaining 100 percent expansion decreases. Further, if
filter loading becomes excessive, the biofilm can increase in thickness
to the point that particle movement gradually becomes restricted, and
the filter bed gels and becomes inoperative. This event was readily
apparent with the medium (12/20) and fine (20/40) grain sands as the two
media demonstrated a slow, yet gradual, flow restriction through the
filter bed as loading was continually increased (Figure 4). Likewise,
the decline in flowrate allowed effluent dissolved oxygen concentrations
to decrease below 2.0 mg/L, thereby limiting the filter's nitrification
ability. Conversely, the coarse (8/16) grain sand showed no effect of
clogging whatsoever even at higher loadings. The larger grain size
appears to contribute to the filter's ability to shear excessive biofilm
growth on the particle thereby reducing its chances of clogging (at
least within the loading conditions examined in this experiment).
Based on the results of this study (Table 2), no appreciable
differences were apparent when comparing the relative abilities of each
grain size with respect to nitrification. Results from this study do
not substantiate findings from past research which indicates that media
surface area has a significant effect on the overall nitrification
ability. In fact, the larger grain size, which had less surface area
for biofilm growth in comparison to the two smaller grain sizes, had an
equal, if not slightly better, nitrification ability based on compari-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28
sons made at equivalent carrying capacities. More importantly, the
coarse (8/16) grain sand appears most suitable as a media since it has a
more predictable hydraulic behavior with respect to filter clogging.
Additionally, the coarse sand can safely maintain loadings equivalent to 3 3 4800 kg/m (300 lbs/ft ) of unfed crawfish while keeping ammonia and
nitrite concentrations below 1.0 mg/L and 0.5 mg/L, respectively. Based
on these observations, the coarse sand must be considered the superior
media to use with a fluidized bed filter in a recirculating aquaculture
system.
Several experimental factors must be considered as limitations if
one attempts to use these results for determining actual filter carrying
capacity. First, the results of this research are based on an unfed
system. Commercial soft-shell crawfish systems, which employ feeding,
require more nitrification ability as a result of the additional
nitrogen load contributed by uneaten food which remains in the system.
Second, the extremely high loading rates which were used with the medium
and coarse grain sands were not anticipated in the original experimental
design. Consequently, insufficient volume/weight ratios (less than
2 gal/lb) were maintained for the shock ammonia loadings which occurred
within each individual system when crawfish loadings were suddenly
increased. Finally, an inadequate amount of aeration was apparent with
the higher loading rates which led to intermittent oxygen depressions,
pH declines, and CC^ accumulation within the system. Loadings that 3 2 exceeded 4800 kg/m (300 lbs/ft ) appear to show no significant
differences in effluent quality (particularly ammonia), thus leading
the authors to believe that either a trace element deficiency, C02
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29
accumulation, pH decline, or combination of these factors has a direct
effect on the kinetics governing the nitrification process.
While the fluidized bed has enormous advantages over the conven
tional treatment methods such as the submerged rock filter (Manthe,
1984), the filter also has three major disadvantages. The first
involves the required chemical addition needed to stabilize pH levels.
In this case, sodium bicarbonate was used to maintain an adequate pH and
assure that sufficient amounts of bicarbonate ion were available for
nitrification. Without chemical addition, the filter will not achieve
optimum levels of nitrification and consequently fail to maintain
adequate water quality. A second disadvantage involves the filter's
inability to capture solids. Since the fluidized bed specializes
primarily in nitrification, the filter must be used in conjunction with
a filter which can remove solids, such as an upflow sand filter.
Finally, a higher filter flowrate is required to maintain sufficient
fluidizatlon, thus increasing the energy costs required for operation.
CONCLUSIONS AND RECOMMENDATIONS
Based on the results of this study, the three sand grain sizes
demonstrated no appreciable differences in terms of carrying capacities
for maintaining ammonia levels below 1.0 mg-N/L. However, with respect
to observed maximum carrying capacities, only the coarse sand (1.19-
2.38 mm) maintained its ability to withstand clogging and consequent
oxygen failure. The larger grain size appears to contribute to the
filter's ability to shear excessive bacterial growth from the sand
particles thereby making the media more hydraulically predictable. The
coarse sand must therefore be considered a superior media to use with a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30
fluidized bed filter in a recirculating aquaculture system. Fluidized
bed filters employing the coarse sand as a media should be capable of
3 3 safely supporting loadings equivalent to 4800 kg/m (300 lbs/ft ) of
unfed crawfish while maintaining ammonia concentrations at or below
1.0 mg/L and nitrite levels below 0.5 mg/L.
Future research regarding fluidized bed filters and recirculating
aquaculture systems should investigate specific ammonia and nitrite
toxicity limits for molting crawfish. Chronic toxicity testing should
be conducted on conditions that realistically simulate commercial
operating requirements for recirculating systems. Additionally, factors
which limit filter effluent quality (such as trace elements, CC^, and
pH) should be determined so that more efficient filter designs can be
implemented in commercial facilities.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER III
EVALUATION OF A FLUIDIZED BED/UPFLOW SAND FILTER COMBINATION PROPOSED FOR A RECIRCULATING BLUE CRAB SHEDDING SYSTEM*
Daniel G. Burden and Ronald F. Malone Department of Civil Engineering, Louisiana State University Baton Rouge, Louisiana 70803
ABSTRACT
A fluidized bed and upflow sand filter combination were evaluated
in terms of nitrification and oxygen consumption when used in a
recirculating crab shedding system. Systems using the filter
technology can expect carrying capacities of over 700 crabs per cubic
foot of sand media while maintaining adequate water quality for a
shedding operation. This filter combination's carrying capacity
exceeded that observed with the submerged rock filter by more than 20
times and was largely explained by the filter's ability to remove
solids to reduce the oxygen loading rate (OLR). The fluidized bed
exhibited a 56 percent decrease in OLR when compared to the best
submerged rock filter design. This substantial reduction was
principally attributed to biomass attrition occurring within the
individual filter itself. Nitrification rates with the filter
combination were extremely high as total ammonia and nitrite levels
remained below 1.0 mg-N/L.
*Submitted to The Journal of Aquacultural Engineering for publication.
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32
INTRODUCTION
The use of recirculating systems for producing soft-shell crabs has
gained wide acceptance in Louisiana and the surrounding Gulf States.
Filtration components are utilized to maintain high water quality as
required by molting crabs. The historical filtration method used by the
industry has been the submerged rock filter. The submerged rock filter
combines the advantages of nitrification, solids removal, and pH control
into one unit, thus eliminating the need for additional treatment.
Furthermore, these filters reauire .little or no maintenance over an entire
shedding season.
Current design criteria for the submerged rock filter stipulate a 3 3 carrying capacity of 33 crabs/ft (132 kg/m ) of filter media although 3 3 capacities as high as 50 crabs/ft (200 kg/m ) have been observed under
experimental conditions (Manthe et al. 1988). As their carrying
capacities indicate, these filters can only operate at low loading
rates. Bacteria in the filters are maintained in a state of endongenous
respiration which assures that there is no net production of bacterial
biomass. As loading rates are increased, filter clogging generally
results from excessive bacterial growth which fill void spaces in the
rock bed. This clogging prevents a uniform distribution of oxygen
throughout the filter bed, causing localized oxygen deficiencies which
lead to filter failure and eventual loss of crab stocks.
Aside from clogging, pH control becomes increasingly difficult
because of rapid fluctuations in carbon dioxide concentrations. As the
system's buffering capacity becomes limited, the nitrifying bacteria's
ability to adequately reduce ammonia and nitrite levels in a recircu
lating system becomes severely inhibited (Gaudy and Gaudy, 1978; Paz,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33
1984). Consequently, toxic nitrite concentrations rapidly increase to
exceedingly high levels, sometimes greater than 20 mg-N/L (Manthe et al.
1984). Molting crabs were found to be adversely affected by nitrite
levels as low as 0.5 mg-N/L. When toxic nitrite concentrations were
present during molting, crabs often died halfway out of their old shell.
As a result, water quality often controls the economic viability of the
shedding operation.
Research efforts have therefore been directed toward developing
high-rate filters which are small in volume, can withstand clogging, and
yet maintain adequate water quality. This manuscript presents results
of laboratory studies which evaluate the carrying capacity of a
fluidized bed and upflow sand filter combination when used in a crab
shedding facility. Independent evaluations of carrying capacity on the
fluidized bed and the upflow sand filter, as well as waste characteriza
tion studies on the blue crab, were also conducted to provide supporting
data.
BACKGROUND
Nitrification. Biological filters are designed to optimize natural
waste degradation processes whether aerobic or anaerobic. Although
anaerobic filters have been proposed for removing nitrate from system
waters, most biological filtration schemes are based upon enhancing
aerobic processes. The most critical degradation step, nitrification,
depends upon bacteria genera (principally Nitrosomonas and Nitrobacter)
that are strictly aerobic. These bacteria transform the toxic forms of
ammonia and nitrite to the relatively harmless form of nitrate. The
overall reaction is summarized as follows:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 34
NH? + 20„ '-*■ NO" + 2H+ 4- Ho0 Cl) 4 2 3 2
On a stoichiometric basis, 4.57 milligrams of oxygen are required for
each milligram of ammonium-N which is oxidized to nitrate-N. In a
recirculating system's biological filter, however, nitrification is not
the only oxvgen-consuming phenomenon which takes place (Manthe et al.
1985). Other processes such as mineralization of animal wastes and
respiration by carbonaceous organisms during BOD assimilation
simultaneously compete for oxygen. Filter designs must therefore
satisfy not only the oxygen demand for nitrification, but that amount
for these additional processes as well.
While the nitrification process requires oxygen, its end products,
hydrogen ions and carbon dioxide, reduce both pH and alkalinity.
Alkalinity reduction occurs at a rate of 7.14 moles for every mole of
ammonium oxidized during nitrification (USEPA, 1974). Recirculating
systems must therefore maintain adequate alkalinity levels to appease
the inorganic carbon levels required by the nitrifying bacteria.
Typical water quality guidelines for recirculating systems require a
minimum alkalinity of 100 mg/L as CaC03 (Malone and Burden, 1988).
Optimum pH levels for soft-shell crab production systems lie between 7.5
and 8.3 (Haefner and Garten, 1974). This pH range will adequately
sustain aquatic life, as well as the nitrifying bacteria, and minimize
the dissolution of ions in a recirculating system.
Oxygen Consumption. The amount of oxygen consumed by a biological
filter indicates the amount of bacterial activity, which in turn,
relates to the waste loading imparted to a recirculating system.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35
The relationship of oxygen consumption to waste loading, coupled with
the fact that oxygen frequently limits filter performance, has led to
the development of filter designs based on these concepts.
Manthe et al. (1988) more recently presented submerged rock filter
designs for blue crab shedding systems based on oxygen supply. The
authors based their methodology on a mass balance of dissolved oxygen
using two design parameters: the oxygen consumed during filtration
(OCF) and the oxygen loading rate (OLR). The OCF parameter was based on
oxygen consumption and flowrate across the filter:
OCF = Q * (C± - Co) (2)
where
OCF = oxygen consumed during filtration (mg-O^/day),
Q = flowrate through the filter (L/day),
C^ = filter influent DO concentration (mg-O^/L), and
Cq = filter effluent DO concentration (mg-O^/L).
The OCF measures the amount of oxygen consumed in the filter as a result
of BOD exertion and bacterial respiration. The OCF varies continuously,
even with a constant population of aquatic species in the system, as a
result of subtle water quality changes.
The mean oxygen demand on the filter on a per crab basis is
expressed as:
N OLR = Z (OCF./R.) (3) 1 = 1 3 3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26
where
OLR = the mean oxygen loading rate (mg-O^/crab-day);
R = the number of crabs in the system;
N = the number of OCF observations; and
j = observation number.
The OLR parameter estimates the oxygen demand (BOD) on the filter as
wastes from a unit number of crabs are degraded. Submerged rock filters
provide for degradation of captured solids, dissolved wastes, and
bacterial biomass; thus, the potential oxygen demand is completely
expressed. Addition of pre-filtration components or upflow sand filters
reduces the OLR by removing solids from the system before they
stabilize.
The ability of the submerged rock filter to operate at high loading
rates is constrained by the bacteria's spatial demands and oxygen
transfer rates (Haug and McCarty, 1972; Manthe et al. 1984). The
proposed fluidized bed/upflow sand filter combination is designed to
alleviate clogging problems at high loadings, provide a robust oxygen
supply, and maximize solids removal in an effort to minimize oxygen
demand (or OLR) on the filter. Each filter component is briefly
discussed in the following section.
Fluidized Bed Filter. Fluidized bed filters have been historically used
in the chemical engineering field but recently emerged in the late
1970's as a means of upgrading existing sewage treatment plants (Cooper
and Sutton, 1983). Design criteria for these biological reactors have
been primarily developed for treating high influent BOD concentrations
commonly found in overloaded facilities.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37
The fluidized bed simulates the submerged rock filter in many
respects, but the packing medium, typically sand, expands by an upward
movement of fluid through the bed. Fluidization of the media increases
the effective surface area available for bacterial growth per unit of
reactor volume and provides a favorable environment for kinetic
transport in a relatively small reactor volume. Furthermore, the
fluidized bed minimizes operational problems such as clogging since the
bed can expand to accommodate the extra volume necessary for bacterial
growth (Jeris et al. 1974; Andrews and Tien, 1979; Cooper, 1981). The
limiting factor with using the fluidized bed filter lies in its
inability to capture solids. Since the filter operates in an expanded,
turbulent environment, biofilm which adhere to the sand grains are
continually abraded as a direct result of fluid shearing and particle
abrasion (Powell and Slater, 1982). This continual abrasion of biofilm,
or "biomass attrition", combined with the filter's inability to capture
solids, therefore mandate combining the fluidized bed with some
mechanism for solids removal. The upflow sand filter provides one such
option.
Upflow Sand Filter. Because of the desire to maintain suitable water
quality in a recirculating system and the inability of the fluidized bed
to remove solids, filtration for a recirculating system must be
supplemented with additional treatment. A granular-medium filter such
as the upflow sand filter serves such a purpose. The upflow sand filter
consists of a packed bed of sand through which water flows upward from
the filter bottom and discharges back to the system or to a waste line,
depending on the mode of operation. During normal operation, the filter
remains packed, capturing solids within or on the filter surface,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38
returning clean effluent waters back to the system. Solids capture
occurs through the process of physical straining and sedimentation
within the filter media (USEPA, 1980). Although solids capture is the
filter's primary objective, a high degree of nitrification also takes
place within the filter bed (Rich, 1980; Wheaton, 1985).
As solids buildup occurs within the filter, flowrate declines,
headloss increases across the filter bed, and effluent quality, in
general, decreases. Consequently, intermittent backwashing (or
cleaning) of the upflow sand filter is required on a routine basis.
Backwashing is accomplished by increasing the flowrate to expand the
bed. As the flowrate increases, trapped solids within the filter bed
are removed by fluid shearing and particle abrasion. Backwashing is
normally directed to a slow sand filter which removes solids discharged
from the backwash waters prior to re-entering the recirculating system.
Backwash water is applied to the filter bed surface through a distri
bution manifold which insures uniform flooding of the bed surface.
Solids are removed in a surface mat which forms on the filter's upper
layer of fine sand. The trapped layer of solids and sand is periodi
cally removed, discarded, and replaced with new sand.
METHODS
Two different experiments were conducted in this study: filter
studies and waste characterization studies. The filter studies were
conducted to determine carrying capacities on the fluidized bed/upflow
sand filter combination, as well as each individual component, using
unfed premolt crabs. The waste characterization studies were performed
to provide supporting data on the amount of ammonia and excessive solids
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39
material which were excreted by blue crabs. The methods used with each
experiment are described below.
Filter Studies. Three different filter configurations were tested:
(1) a fluidized bed/upflow sand filter combination , (2) a fluidized bed
with floss pre-filtration, and (3) an upflow sand filter. Each filter
was constructed from clear acrylic tubing, 4 inches (10.2 cm) in
diameter, using PVC fittings on the inlet and outlet. All systems,
consisting of three plexiglass holding tanks, a sump, and a pump, were
plumbed with polyvinylchloride (PVC). An American Products 1/2 HP pump,
capable of delivering a maximum of 53 gpm (200 L/min) at 20 feet of
head, was used to assure a flowrate which exceeded the system's
requirements throughout the experimental regime. Flowrates were
controlled using 3/4 inch (1.9 cm) ball valves on each filter.
A general schematic and dimensions of the experimental setup are
presented in Figure 7 and Table 3, respectively. Aeration was supplied
by plumbing spray nozzles into each holding tank from a manifold off the
pump output line. Each independent system had a total volume equivalent
to 67.4 gal. (255 liters). The single filter configurations, systems
utilizing a single fluidized bed (Figure 7a) or a single upflow sand
filter, had a filter media volume equivalent to 0.116 ft"". The con
figuration consisting of both filters operating in parallel (Figure 7b) 3 had a filter media volume equivalent to 0.233 ft or twice that of the
single filter configuration.
The fluidized bed filter (Figure 8) consisted of a sand bed, 16
inches (40.6 cm) in depth, through which a constant upflow of water
continuously maintained the filter in its expanded mode. Typically,
bed expansion was maintained between 25 and 50 percent. A medium
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. O -E~ TANK 3 TANK TANK! FLOSS 1 FILTER o "O- -J o PUKPHP 1/2 » » SUUP HOLDING TANK 2 TANK HOLDING X HOLDING QALLVALVE FLUIDIZED DED T > 1/2 HP PUMP HP 1/2 * * SUtSP * * l— — h r HOLDING TANK I HOLDING (C) 1 HOLDING TANK3 HOLDING HOLDINGTANK 2 -pj -pj SAND SLOW FILTER (b ) u p f l o w SAND FILTER OALL VALVE OALL FLUIOIZE0DED _ra- (a) (I filter (circumvented by the dashed line) configurations are denoted in (a). A schematic of the experimental setup used with the three filter configurations. The single fluidized bed filter is illustrated in (b). The fluidized bed/upflow sand filter combination and the single upflow sand Figure 7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission 41
Table 3. Dimensions of the experimental setup used for determining filter carrying capacity.
Cross Water Sectional Component Length Width Depth Depth Area Volume (ft) (ft) (ft) (ft) ( f O (ftJ)
Holding Tanks 3 2 1 0.479 6 2.875
Sump 3 ? 1 0.542 6 3.250
Fluidized Bed - - 1.33 - 0.087 0.116 Filter (Media)
Upflow Sand - - 1.33 - 0.087 0.116 Filter (Media)
Fluidized Bed/ — . — 1.33 0.174 0.233 Up flow Sand Filter (Media)
Slow Sand 1.167 1 0.67 ~ 1.167 0.778 Filter (Media)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCHARGE TO SYSTEM
100 PERCENT EXPANSION
PACKED BED HEIGHT ( 16“)
S OP GRADED GRAVEL (1/2^- I/O") SUPPORT. PLATE
SECTION 4 -4 DALL VALVE
Figure 8. A detailed illustration of the fluidized bed filter used with the study (not to scale).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 43
grain sand, ranging in diameter from 0.84 to 1.68 mm, was used as a
filter medium. Solids removal was accomplished by placing polyester
aquarium filter floss in a mechanical filtration box at the filter's
discharge. Floss was changed on a daily basis.
The upflow sand filter (Figure 9) consisted of a 16 inch (40.6 cm)
packed bed of dolomite (4.76 - 7.93 mm in diameter). Backwashing was
performed once a day by temporarily shutting off the normal flow dis
charged back to the sump, thereby directing the backwash waters to a
slow sand filter (Figure 7a) via a 1.5 inch (3.8 cm) drain line. Clean
effluent water was returned to the sump.
Each system was initially filled with ammonia-free, dechlorinated
tap water and adjusted for salinity to 5 ppt using artificial sea salts
(Instant Ocean). A specified number of medium crabs (Callinectes
sapidus), ranging in carapace width from 10 to 15 cm, were then loaded
into each system. Crab loadings were increased only after each system
stabilized at the specified crab loading so that representative water
quality samples could be collected. Stabilization merely indicated that
the population of nitrifying bacteria within the filter had approached
an equilibrium state with the ammonia and nitrite concentrations present
in the system.
Crabs were not fed while in the experimental tanks to simulate a
commercial shedding operation. To prevent cannibalism, weak animals and
mortalities (approximately two percent daily) were removed and replaced
every two to four hours. Crab populations in each system were routinely
replaced with a healthy crab population every 5 days. Replacement crabs
were held in a separate 600 gal. (2250 L) recirculating system where
feeding was conducted. Purina Trout Chow Checlcerettes i t4 (w/ 43 percent
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. - - p p 3/0 Holes
SECTION 4 -4 BALL VALVE EXPANSION -5 0 PERCENT
{ 3C BALL VALVE
SLOW SAND FILTER V INTERMITTENT EXPANSION VWASTE WATER , / ' > / ' > DISCHARGE TO. INFLOW ( 1 /2 - 1/B") BREAK-BAR SUPPORT-PLATE
PACKED BED HEIGHT 3" OF GRADED GRAVEL TO SYSTEM DISCHARGE NORMAL OPERATION Normal and intermittent operation of the upflow sand filter (not to scale) BALL VALVE Figure
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45
protein) and live crawfish were used for feeding in the support system.
All crabs were obtained from a commercial soft-shell crab fisherman
(Lacombe, Louisiana) two to three times a week. Both the holding and
experimental systems were maintained at room temperature (23 + 1 °C).
Water loss due to evaporation was replaced daily using dechlorinated tap
water.
Monitoring was performed on a daily basis for total ammonia-
nitrogen (NH^), nitrite-nitrogen (NO^)» dissolved oxygen (DO), pH,
temperature (Temp), flowrate (Q), and salinity. All monitoring was
conducted between 9:00 and 11:00 am. When each system stabilized at
the specified crab loading, an expanded "steady-state" analysis was
conducted to determine filter performance. Steady-state analysis
included nitrate-nitrogen (NO^) , biochemical oxygen demand (BOD,.),
alkalinity (ALK), and total suspended solids (TSS) in addition to those
parameters run on a daily basis. The methods and instrumentation used
with each analysis are outlined in Table 4. All analyses were conducted
according to procedures outlined in Standard Methods (APHA, 1985). At
the end of a steady-state analysis, the crab loading in each system was
increased and maintained until the system stabilized at the new loading.
Bicarbonate addition was necessary to maintain pH levels (7.0-
8.3), as well as alkalinity levels (> 100 mg/L as CaCO^), in an optimum
range for oxidation rates of the Nitrosomonas and Nitrobacter bacteria
(Paz, 1984; USEPA, 1975). pH levels were adjusted by adding 2.4 to 4.8
mg of sodium bicarbonate (NaHCO^) per pound of crab following each day's
sampling routine (Malone and Burden, 1987).
Waste Characterization Study. Separate from the individual filter
studies, a waste characterization study was conducted on the blue crab
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 46
Table 4. Water quality parameters and methods used with the filtration research.
Water Quality Parameter Analysis Method
Total Ammonia Nitrogen Distillation, Nesslerization
Nitrite Nitrogen Sulfanilamide-based Colorimetric
Nitrate Nitrogen Orion Nitrate Ion Electrode - Model 93-07, Orion Double Junction Reference Electrode - Model 90-02
Total Kjeldahl Nitrogen Digestion, Distillation, Nesslerization
Biochemical Oxygen Winkler Method Demand
Alkalinity Titration/Potentimetric Endpoint
Total Suspended Solids Gravimetric
Dissolved Oxygen Yellow Springs Instrument Model 57 Oxygen Meter
pH Cole-Palmer Mini pH Meter
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to determine ammonia excretion rate and subsequent solids loading.
These data allow generalization of design findings by defining carrying
capacities in terms of parameters that could be used in any application.
Crabs used in the study were taken from a main holding system where they
were fed whole crawfish and Purina Trout Chow Checkerettes #4 prior to
their isolation. Methods used with the study are similar to those
followed by Cange (1988). A total of six water quality parameters were
analyzed: ammonia nitrogen, nitrite nitrogen, total Kjeldahl nitrogen
(TKN), biochemical oxygen demand, total suspended solids, and volatile
suspended solids (VSS). Analyses for the BOD and TKN parameters were
conducted on both filtered and unfiltered samples, using a 0.45 micron
filter, to determine the impact of suspended solids upon water quality.
All analyses were conducted in triplicate on the same day of sampling
according to procedures outlined in Standard Methods (APHA, 1985).
RESULTS
Filter Studies.
Initial filter acclimation was accomplished with each system
configuration in 20 days (June 19) after successive loadings of 3 3 43 crabs/ft (9 days) and 86 crabs/ft ('11 days). Startup curves for
ammonia and nitrite followed the typical pattern for biological filters
as described by Manthe et al. (1984). Subsequent to acclimation, crab
loadings were increased and steady-state analyses were conducted
accordingly. pH levels gradually declined within each system
illustrating the inability of the silica-based sand and dolomictic media
to effectively buffer pH changes. Following approximately 40 days of
monitoring, chemical addition of sodium bicarbonate was implemented in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48
all three filter systems to maintain pH levels in an effective range for
nitrification. Results of the daily and steady-state monitoring for the
combination and individual filter configurations are presented below.
Fluidized Bed/Upflow Sand Filter Combination. Results of the daily
monitoring of ammonia, nitrite, effluent dissolved oxygen, and pH data
with respect to crab loading are illustrated in Figure 10. Since the
filter combination essentially doubled the filter media volume,
3 volumetric carrying capacities, ranging from 86 to 773 crabs/ft , were
consequently lower in contrast to the individual filter configurations.
Monitoring was discontinued after September 16 as a result of external
factors.
The inability of the sand and dolomite media to provide adequate
buffering capacity, as well as pH control, was readily apparent after
40 days of continuous operation. At this point, pH levels dropped below
7.0, a critical level with regard to high nitrification rates. The high
ammonia peak observed on July 17 (day 49) resulted from a steady decline
in pH levels (prior to chemical addition) coupled with shock loading the
3 system to a total of 172 crabs/ft . This loading period also had a
short-term, diminishing effect on the effluent DO levels in the upflow
sand filter as illustrated in Figure 10. Smaller ammonia peaks observed
throughout the remainder of the study appear to result from shock
loading on the system. Effluent DO in the upflow sand filter briefly
fell below 2.0 mg/L during this period, but remained, in general, above
3.0 mg/L while DO levels in the fluidized bed typically stayed above
5.0 mg/L.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49
pH
h-< Z9 N. g O < ^ MH_. N
1600
o 1200
1000
GOO
° GOO
SAY50 JUNEI9 JUfS50 JUIYI3 JULT30 AUG 15 AUG91 GCPIG SAMPLING OATE
Figure 10. Water quality monitoring results and loading curves obtained on the fluidized bed/upflow sand filter combination.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50
A total of six steady-state analyses, beyond that of initial filter
acclimation, were conducted on the system (Table 5). Ammonia and
nitrite levels for the most part, remained below 1.0 and 0.5 mg-N/L,
respectively, for the entire study. These low levels appear to reflect
a somewhat less rigorous test on the filter combination since maximum
carrying capacities were not reached in this experimental regime.
Fluidized Bed. Daily ammonia, nitrite, pH, and effluent dissolved
oxygen data are illustrated along with crab loading data in Figure 11. 3 Following filter acclimation with a loading capacity of 86 crabs/ft , 3 crab loading on the system was doubled to 172 crabs/ft . Steady-state 3 analyses were conducted on loadings ranging from 172 to 688 crabs/ft
corresponding with sampling dates June 28, July 9, and August 7
(Table 6a). Steady-state ammonia, nitrite, nitrate, and BOD levels
steadily increased with loading. Mean flowrate over the entire study
was equivalent to 9.8 L/min. As a result, oxygen transport in the
filter never became limited as effluent dissolved oxygen levels remained
above 4.0 mg/L.
The large ammonia peak (8.2 mg-N/L) observed on July 17 (day 49)
resulted from both shock loading and a steady decline in pH levels
(< 7.0) before bicarbonate addition was implemented on July 18. Crab
loading was essentially doubled five days previous to this ammonia peak.
Furthermore, with the increased loading the water/volume ratio fell
below 1 gal/crab. Shock loadings, like those experienced here, can be
avoided if a ratio of 2 gal/crab or greater is provided based on
previous recommendations by Manthe et al. (1987).
Routine monitoring of ammonia and nitrite was carried out for
an additional 20 days beyond that of the last steady-state analyses
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 5. Steady-state data obtained on the fluidized bed (FB) and upflow sand (UFS) filters when running in parallel operation. Flowrates denoted by 'Q'.
Loading (crabs/ft3) 86 172 339 430 645 773 (lbs/ft3) 20 40 80 100 150 180
Q-FB (L/min) 14.1 11.0 10.3 10.8 16.8 16.8
Q-UFS (L/min) 3.3 1.6 3.6 2.6 2.3 ^O • 4. 0
NH3 (mg-N/L) 0.08 0.23 0.28 0.29 0.31 0.60
N02 (mg-N/L) 0.05 0.12 0.33 0.28 0.29 0.46
N03 (mg-N/L) 87 148 348 313 395 459
BOD,. (mg/L) 3 n 7 5 3 4
ALK (mg/L as CaCO^ ) • 14 192 190 135 144
pH 7.8 6.6 7.5 7.5 6.7 6.9
TSS (mg/L) 10 9 16 3 14 12
Temp. (°C) 22.0 22.0 23.5 24.0 22.5 23.0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o «
< ?
coo
zoo
UAYSO JUNEI9 AURESO JULYI9 JULYIO AUOI9 AU03I CEP 16 8AUPMN0 DATE
Figure 11. Monitoring results obtained on the fluidized bed filter using floss pre-filtration.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6. Steady-state data obtained on (a) the single fluidized bed filter and (b) the single upflow sand filter.
Loading (crabs/ft3) 172 344 688 1289 1547 (lbs/ft3) 40 80 160 300 360
(a) FTuiaized Bedr
Q (L/min) 9.0 9.5 10.6
NH3 (mg-N/L) 0.19 0.13 0.74
NO£ (mg-N/L) 0.16 0.18 0.63
N03 (mg-N/L) 63 110 333
BOD5 (mg/L) 4 4 13
ALK (mg/L as CaC03) • 43 289
pH 7.9 7.0 7.6
TSS (mg/L) 8 8 21
Temp. (°C) 22.0 22.0 23.5
(b) UpFlow Sand Filter:
0 (L/min) 4.7 2.2 3.6 1.9 1.7
NH3 (mg-N/L) 0.28 0.59 0.61 0.77 0.98
N02 (mg-N/L) 0.27 0.33 0.50 0.53 0.55
. NO (mg-N/L) 69 204 296 342 421
BOD5 (mg/L) 3 8 5 6 9
ALK (mg/L as CaC03) • 112 262 155 198
pH 7.9 7.5 7.6 7.6 7.6
TSS (mg/L) 7 17 8 18 4
Temp. (°C) 22.0 22.0 23.5 22.5 22.5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54
3 on August 7 (day 70). Crab loadings which exceeded 158 lbs/ft resulted
in heavy mortalities, typically greater than 10 percent/day, as ammonia
and nitrite levels reached levels as high as 5.5 and 1.3 mg-N/L,
respectively. The filter failed to exhibit any signs of recovery and
consequently, the system was shut down.
Upflow Sand Filter. Five steady-state analyses beyond initial
filter acclimation were conducted on the upflow sand filter with 3 loadings ranging from 172 to 1549 crabs/ft (Table 6b). Ammonia,
nitrite, and nitrate levels steadily increased with loading. Flowrates
and consequently effluent dissolved oxygen levels consistently decreased
with increased loading as biofilm growth increased throughout the
dolomitic filter bed. Although backwashing was performed on a routine
daily basis, the ever increasing solids loads resulting from biofilm
growth and crab loading largely contributed to the flowrate decline.
Prior to the last steady-state analyses, effluent dissolved oxygen
levels remained below 1.0 mg/L.
Routine daily monitoring of ammonia, nitrite, pH, and effluent
dissolved oxygen concentrations are illustrated in Figure 12. Beyond
filter acclimation, three large ammonia peaks ranging from 2.5 to
5.7 rag-N/L were observed in the data. The first peak (July 17)
directly ensues the day when system pH (6.5) was at a minimum.
Immediately following that day's sampling, bicarbonate addition was
implemented in the system. The two subsequent peaks, one of which
displayed the highest ammonia level (5.7 mg-N/L), were observed on
August 3 and 27 (days 66 and 90) with loadings of 344 and 1289 3 crabs/ft , respectively. Effluent dissolved oxygen levels were
extremely low, below 1.0 mg/L, prior to these periods. Nitrification in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55
or
DO
o < ?
1400
g 1200
1000
0 0 0 ■
2 0 0
MAY 3 0 Figure 12. Loading curves and water quality monitoring data obtained on the single upflow sand filter configuration. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56 the filter was undoubtedly inhibited to a large extent as a result of these oxygen deficiencies. Excretion Study. Results of the excretion study are summarized in Table 7. Mean excretion rate for the blue crab was equivalent to 0.219 +0.076 mg NH^+N^/g-day based on a mean crab weight of 105.6 grams. The large amount of variability displayed by the excretion data was attributed to three factors: (1) crabs used in the study were based on "wet weight", (2) the wide range of crab weights used throughout the study, and (3) the physiological state of the crabs as a result of time spent in the holding system. Additional factors such as the natural variability in both sampling and laboratory procedures, as well as the crab's feeding and metabolic rates, may have also contributed to the observed variability. The unfiltered and filtered BOD and TKN samples represent the solids partitioning portion of this study. Solids removal reduced BOD loading by 0.169 mg/g-dav (62 percent) while only a small decrease (7 percent) in nitrogen loading was apparent. This large reduction of BOD indicates the need for a solids removal mechanism such as the upflow sand filter in a recirculating shedding operation. DISCUSSION The fluidized bed/upflow sand filter combination provided superior nitrification treatment when compared to the submerged rock filter. 3 Maximum carrying capacity (773 crabs/ft ) observed on the filter com bination exceeded the rock filter by more than 20 times while main taining suitable water quality for a crab shedding operation. Higher Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57 Table 7. Statistical summary for all waste characterization data collected on the blue crab. Standard Parameter n X Deviation Range Crab Weight (g) 15 104.1 11.48 83 - 125 Excretion 14 0.219 0.076 0.102 - 0.383 (mg/g-day) Nit^ (mg-N/g-day) 14 0.194 0.083 0.083 - 0.375 N02 (mg-N/g-day) 14 0.025 0.016 0.005 - 0.059 Unfiltered 15 0.330 0.132 0.160 - 0.610 TKN (mg/g-day) Filtered 14 0.315 0.139 0.094 - 0.581 TKN (mg/g-day) Unfiltered 15 0.262 0.120 0.083 - 0.542 BOD5 (mg/g-day) Filtered 15 0.093 0.055 0.026 - 0.216 BOD5 (mg/g-day) TSS (mg/g-day) 14 0.711 0.376 0.309 - 1.77 VSS (mg/g-day) 14 0.468 0.186 0.218 - 0.887 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58 carrying capacities may have been possible, but were unobtainable because of a limited crab supply. Oxygen Consumption. Examination of the oxygen loading rates (OLR) for each filter configuration demonstrates why the fluidized bed/upflow sand filter combination yields superior performance. Oxygen loading rates were determined on each filter configuration based on the following criteria: (1) effluent dissolved oxygen concentrations must be greater than 2.0 mg/L to assure the expressed OLR was not oxygen-limited and (2) ammonia concentrations must not exceed 1.0 mg-N/L to avoid the effect that shock loadings may have on system operation. These criteria follow the guidelines set forth by Manthe et al. (1988) in their OLR evaluation on the submerged rock filter. Results of the OLR analyses revealed mean values of 68, 139, and 140 mg-0o/crab-day for the fluidized bed w/ floss pre-filtration, the upflow sand filter, and the fluidized bed/upflow sand filter combina tion, respectively. To gain a more thorough understanding of the OLR evaluation, the mean values (and their standard deviations) for the individual filters were examined by using probability distributions. Distributions used in these analyses were chosen in the form of a gamma density function based on data skewness (Ward et al. 1982). Gamma distributions for the individual fluidized bed filter w/ pre-filtration along with those for the two submerged rock filter configurations are compared in Figure 13. Observation of the distributions illustrates the highly variable nature of the OLR parameter. The two separate distributions for the submerged rock filter are based on data taken from Manthe et al. (1988): (1) an OLR of 500 mg-0?/crab-day without pre-filtration and (2) an OLR of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. \D Ul Id Id I (x =500) W/0 PRE-FILTRATION SUBMERGED ROCK FILTER ROCK SUBMERGED m g-02/crab- ) day =344) (x W/ PRE-FILTRAT ION SUBMERGED ROCK FILTER ROCK SUBMERGED SOLIDS REMOVAL ) 68 = 1 ( BIOMASS ATTRITION OXYGEN LOADING RATE ( W/PRE-FILTRATION FLUIDIZED BED 100 200 300 400 500 600 700 800 900 1000 O.OI2r 0001 0.002 OOOO filter configurations. Meanvalues are indicated in parentheses and denoted by the horizontal lines on each distribution. Ui E 0.004 ^ 0003 U- U- 0.006 m m 0.008 2 0.005 S 0.007 lil > u 0.009 Figure 13. Gamma distributions for OLR data on the fluidized bed and submerged rock Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 344 mg-O^/crab-day using floss pre-filtration. The 31 percent dif ference in the mean OLR values results from the floss bed filter's ability to remove the larger solids which would otherwise decompose and exert additional, oxygen demand. When compared with the results from the excretion study, the floss filter is relatively inefficient in solids capture since a BOD reduction of nearly 65 percent was percent was possible with total solids removal (see Table 7). As indicated by the vertical line to the right of the fluidized bed distribution, the mean OLR was reduced an additional 55 percent beyond the original reduction realized by floss pre-filtration with the submerged rock filter. This additional reduction in OLR was principally attributed to biomass attrition, thus supporting the filter's ability to continually abrade biofilm when operating in a turbulent environment. In contrast to the submerged rock filter w/o pre-filtration, the fluidized bed/floss combination realizes an impressive 86 percent reduction in OLR reflecting both biomass attrition and solids removal. Consequently, the relative amount of OLR exerted (OLR^) by the fluidized bed was equivalent to 13 percent of the total, or 68 mg-On/crab-day. Each component is readily illustrated in Figure 14a. Thus, the total, or unimpaired, OLR (OLR^) on the system can be expressed as follows: OLR = OLR, + OLR + OLR (4) t b s e where OLR^ = OLR removal due to biomass attrition, OLR = OLR due to solids waste removal, and s ’ OLR = OLR exerted in the filter, e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 5 % (OLRb) = OLR DUE TO BIOMASS ATTRITION 31 (OLRj,) = WASTE REMOVAL OLR DUE TO SOLID 41% (O LRb)= OLR DUE TO BIOMASS ATTRITION °/< OLR on the upflow sand and fluidized bed filters. 28 3 1 % (OLR0= ) UPFLOW SAND FILTER FLUIDIZED BED FILTER (O L R a) = EXPRESSED OLR WASTE REMOVAL OLR DUE TO SOLID Figure 14. Pie diagrams illustrating the representative portions of total Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 62 Based on a similar analysis with the upflow sand filter, the mean OLR decreased by 41 percent as a result of biomass attrition thus, making the combined OLR (OLR^ and OLRs) reduction equivalent to 72 percent or 359 mg-O^/crab-day (Figure 14b). Under the assumption that OLR^ (floss bed) equals the OLRg (upflow sand filter), estimates indicate the OLR^ (upflow sand filter) is less than the OLR^ since solids are not removed instantaneously as they are held for longer detention times. In summary, the higher OLR^ observed with the upflow sand filter (139 mg-O^/crab-day) justifies using the fluidized bed in combination as a treatment method. Although not illustrated, the OLR^ of the fluidized bed/upflow sand filter combination was reduced by almost 60 percent when compared to the best submerged rock filter design. Nitrification. Further scrutiny of the nitrogen data lends some insight as to why the fluidized bed/upflow sand filter provides superior water quality. The most apparent advantage of the filter combination stems from the relatively short time (20 days) required for filter accli mation. Because of the fluidized bed's favorable environmental for bacterial growth, kinetic transport rates are higher and acclimation periods are consequently shorter. In contrast, acclimation for the submerged rock filter typically takes between 30 to over 100 days (Hirayama, 1974; Manthe and Malone, 1987). By decreasing the acclima tion period, productivity and economic viability of the shedding system increases. Additionally, system water quality remains more stable because of the filter's enhanced ability to handle heavy crab loadings. These factors have been a major obstacle with using the submerged rock filter in commercial systems. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 Although maximum carrying capacities were not reached with the filter combination, acceptable water quality was maintained throughout the experimental run. Ammonia and nitrite levels, for the most part, remained below concentrations considered safe for commercial operation. Furthermore, these conditions were maintained at pH levels considered less than optimum since inhibition of the nitrifying bacteria generally occurs at pH levels below 7.0. Thus, the authors firmly believe that higher carrying capacities could have been attained with the filter combination. From the results of this study, the system's reserve capacity (in terms of nitrification) rests in the fluidized bed filter when the upflow sand filter is near its pratical operational capacity. Higher ammonia and nitrite levels would have nonetheless impacted the entire system had only the upflow sand filter been in operation. The fluidized bed also complements the upflow sand filter in its inherent ability to impose a lower oxygen demand (OLR^) on the system as previously discussed. The effect of nitrogen loading on each filter configuration system was further examined by using a regression analysis. Total nitrogen (summation of ammonia, nitrite, and nitrate) was designated as the dependent variable and crabday as the independent variable for each filter system. Crabday was defined as one crab held in the system for one day; thus, 10 crabs held for 5 days would equal 50 crabdays. Results of the regression analyses are illustrated in Figure 15 for each filter configuration. Of the three filter configurations tested, the fluidized bed filter was best represented by a linear model (P<0.0i, R2=0.998). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ■C- o\ _l _L {(■2 = 0.92) = {(■2 / , UPFLOW SAND FILTER COMBINATION FLUIDIZED BED/ 0.86) 1 (r2 _1_ UPFLOW SAND FILTER SAND UPFLOW y s ' / s ' ' CRABDAY (r2 =0.99) _L_ FLUIDIZED BED 1000 2000 3000 4000 5000 6000 T000 8000 9000 100 500 300 200 indicated in parentheses. accumulations in the three filter configurations. R-square values E o> Z < I- O U1 CD cc O K- „ 4 00 L5. L5. Results of the linear regression analyses performed on total nitrogen Figure Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65 Based on the slope of each regression line, the average nitrogen accumulation rate in each system was determined based on a system volume of 255 liters. Total nitrogen accumulation rates were equivalent to 48, 52, and 103 mg-N/lb-day for the upflow sand filter, the fluidized bed/upflow sand filter combination, and the fluidized bed using pre-filtration, respectively. The accumulation rate observed on the fluidized bed compares favorably well with the excretion rate observed on the crab (102 mg-N/lb-day) while those the remaining filter configurations do not. The lower accumulation rates observed with the fluidized bed/upflow sand filter combination and the upflow sand filter reflect an approximate 50 percent reduction in accumulated total nitrogen, principally nitrate. Assuming this reduction occurs as a result of denitrification, this appears to result from using the slow sand filter with the two upflow sand filter configurations. In contrast, denitrification does not take place with the individual fluidized bed and upflow sand filter configurations since the floss bed system (used for pre- filtration) maintains an aerobic environment as it was changed on a daily basis. While no significant toxicity effects of accumulated nitrate are known to occur in a recirculating system, this reduction improves the overall water quality in the shedding system. Relative Filter Ability. The clogging problems associated with the submerged rock filter during periods of high loading did not hinder performance of either filter until maximum carrying capacities were obtained. The upflow sand filter provided acceptable water quality for maintaining a shedding facility up to a maximum carrying capacity of 3 1547 crabs/ft . Beyond this point, the filter failed to support the system as a result of flow restriction and dissolved oxygen limitations Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66 in the effluent water. Higher carrying capacities may have been possible if baclcwashing had been performed on a more frequent basis, perhaps three or four times a day. As discussed earlier, the fluidized bed operates in an environment conducive to high-rate nitrification treatment principally because of the greater surface area available for biofilm growth. However, based on the results of this study, the fluidized bed exhibited a lower carrying capacity in comparison to the upflow sand filter, which was initially Included for solids removal. Two possible reasons may explain this markedly large difference. First, the highly turbulent environment in which the fluidized bed filter operates may in fact contribute an excessive amount of shear, as well as abrasion, on the biofilm thereby causing more harm than good. The percentage of bed expansion, pre dominated by fluid dynamics, may therefore control substrate removal, thus explaining the lower carrying capacity of the fluidized bed. Secondly, the dramatic differences in filter performance may also be linked to excessive biofilm detachment but for reasons other than shearing (Turakhia et al. 1983). Thus, while both filters provide nitrification, the fluidized bed appears to function more as a roughing filter while the upflow sand filter functions as a polishing filter. Summary. The fluidized bed/upflow sand filter combination provided superior water quality in comparison to the submerged rock filter designs typically used with crab shedding systems. Both filter acclimation time and the actual amount of oxygen demand exerted (OLR^) by the filter combination were reduced significantly. The reduction in OLR is particularly significant since system designs are commonly based on oxygen consumption. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 67 A design criteria based on a fluidized bed/upflow sand filter combination would provide a significantly higher treatment capacity for a recirculating system. Holding systems can be configured to hold the maximum density of crabs without degrading water quality and virtually eliminate loss of crab stocks due to filtration failures. With this ability, the production of soft-shell crabs in can be increased tremendously. Production rates can also be more reliable and predictable permitting the soft-shell crab industry to become more competitive on a national basis. CONCLUSIONS The following conclusions can be made with regards to the research conducted on fluidized bed and upflow sand filters when utilized with recirculating crab shedding systems: (1) The fluidized bed/upflow sand filter combination provided superior treatment in comparison to the submerged rock filter. These high-rate biological filters have carrying capacities of at least 20 times higher than historically observed with the submerged rock 3 filter and were observed to support 773 crabs/ft while still main taining satisfactory water quality for a blue crab shedding operation. (2) The fluidized bed/upflow sand filter combination increased carrying capacity over that of the submerged rock filter can be explained in large by the enhanced solids removal ability and corre sponding OLR reduction. Total oxygen loading rates (0LRt> were reduced an additional 41 and 56 percent for the upflow sand filter and fluidized bed filter, respectively when compared with the submerged rock filter using floss pre-filtration. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68 (3) The upflow sand filter alone provides acceptable water quality 3 treatment up to a maximum of 1547 crabs/ft at which time the filter failed as a result of flow restriction and dissolved oxygen failure in the effluent water. Bacterial respiration rates in the filter compare favorably with that of the fluidized bed filter. (4) Bicarbonate addition is a mandatory requirement for pH control in recirculating systems using these high-rate filtration components over extended periods of time, as the system's alkalinity will rapidly decrease. (5) Mean excretion rate for fed crabs was equivalent to 0.219 mg/g of total nitrogen per day. Solids partitioning studies indicated that solids removal accounts for an approximate 64 percent reduction in BOD loading. (6) The slow sand filter appeared to contribute a significant amount of denitrification to the system by reducing total nitrate accumulation to approximately 50 percent of those levels found in the fluidized bed using only floss pre-filtration. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER IV ASSESSMENT OF A VOLUMETRIC LOADING CRITERIA FOR A FLUIDIZED BED/UPFLOW SAND FILTER CONFIGURATION* Ronald F. Malone and Daniel G. Burden Department of Civil Engineering, Louisiana State University Baton Rouge, LA 70803 ABSTRACT A volumetric loading criteria of 150 pounds of fed crawfish per cubic foot of filter sand was used for designing fluidized bed and upflow sand filters in several commercial recirculating soft-shell crawfish (Procambarus clarkii) facilities. One particular facility was monitored to provide representative data on filter performance. Water quality results indicated the filters sufficiently maintain total ammonia and nitrite levels below 1.0 mg-N/L under typical commercial conditions. The upflow sand filter was principally responsible for solids removal and nitrification in the system. The fluidized bed was biologically active, but appeared to support heterotrophic bacteria which reduce the biological oxygen demand. Normal aeration in the system provided sufficient oxygen levels for the crawfish and bacteria, but failed to control carbon dioxide accumulations causing difficulties in pH control. Shock loading and overfeeding, rather than filter capacity, appeared to dominate water quality fluctuations, thereby indicating the volumetric loading guideline was sufficient for commercial adoption. *Submitted to Aquacultural Engineering for publication. 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70 INTRODUCTION The red swamp crawfish (Procambarus clarkii) experiences the growth process of molting (or ecdysis) a minimum of eleven times before reaching maturity (Huner and Barr, 1984). The crawfish remains in the vulnerably "soft-shell" state for several hours after each molt. At this stage, the soft-shell crawfish is considered a seafood delicacy although availability has been severely limited by production as well as technology. Recent designs of crawfish shedding systems, however, have stimulated the rapid growth of this emerging industry (Cullev et al. 1985a) . Production during the 1987-88 season (November to June) appears to have increased from 50,000 to 75,000 pounds in comparison with just 7,000 pounds in the 1985-86 season. Although the production process appears straightforward, the process is labor intensive and production rates are still somewhat low, ranging from 1.5 to 2.0 percent per day of the system's total holding capacity. Earlier shedding systems were based primarily on flow-through technologies utilizing fresh water, usually drawn from wells or municipal water systems, at a continuous flushing rate ranging from 0.025 to 0.1 gpm/lb of crawfish (Cange, 1987). However, water quality problems and energy costs associated with flow-through designs confronted operators. Iron and ammonia levels in many groundwaters lead to mortalities when crawfish were in the sensitive molting stage, thus prohibiting the use of this technology. Additionally, the cool groundwater (typically 70°F or 21°C) required that flow-through systems bear the high cost of pre-heating water. Soft-shell crab (Callinectes sapidus) production facilities in the Gulf States, to a large extent, use recirculating systems which employ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 71 submerged rock filtration for maintaining water quality (Perry et al. 1982, Manthe et al. 1983). The most inherent advantage of this technology lies in using filtration components which process metabolic wastes to a relatively harmless, non-toxic state. Recirculating systems also reduce heating costs by approximately a factor of ten, providing a strong financial incentive for adopting the technology. Because of these advantages, the recirculating system was immediately considered as an alternative to the flow-through technology currently employed for soft-shell crawfish production. However, the submerged rock filters, used for soft-shell crab production, proved inadequate for crawfish shedding systems where feeding was necessary to encourage growth and molting. Design criteria based upon fluidized bed and upflow sand technologies under development for the soft-shell crab industry (Malone and Burden, 1987) were, therefore, modified and refined in cooperation with the commercial sector. This paper presents recirculating design criteria which have proven acceptable for producing soft-shell crawfish. Monitoring results from a commercial facility constructed on these criteria are presented and relative filter performance is evaluated. Emphasis is placed upon the suitability of a volumetric loading criteria (pounds of crawfish per cubic foot of sand) which is used for designing the fluidized bed and upflow sand filters. BACKGROUND Commercial production of the soft-shell crawfish requires several critical steps: capture, transport, acclimation, sorting, shedding, processing, and sale. These steps have been described in detail by Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72 several authors (Bitterner and Kopanda, 1973; Culley et al. 1985a, 1985b; Culley and Duobinis-Gray, 1988a, 1988b; Huner, 1980; Huner and Avault, 1976; Scudamore, 1948; Stephens, 1955). This article focuses on the shedding aspect where recirculating systems are used for commercial production. The recirculating system used for producing soft-shell crawfish consists of six functional components: culture and molting trays, biological filters, screen boxes, a reservoir, a sump, and a pump (Figure 16). The function of each component and the criteria used for design are provided in Tables 8 and 9, respectively. The biological filters (specifically the fluidized bed and upflow sand filters), which provide water purification, alleviate oxygen supply and clogging constraints that limited the submerged rock filter (discussed in Chapter III). The fluidized bed was intended to provide rapid BOD reduction and high-rate nitrification. The upflow sand filter was originally included to provide a mechanism of solids removal. Design guidelines for these filters have proven critical to commercial adoption (Malone and Burden, 1988). Figure 17 presents a generic configuration for unpressurized fluidized beds that can be used with filters ranging in diameter from 10 to 20 inches (25 to 50 cm). Water flows into the filter bottom, up through the underdrain plate and support gravel bed, finally fluidizing the filter media (sand) prior to exiting the overdrain port. Bacteria attached to the sand grains effectuate nitrification and BOD reduction. Fluxrates through the filter are high, typically ranging from 45 to 65 2 3 2 gpm/ft (2640 to 3820 m -day/m ). Filter media selection (1.2-2.4 mm 3 filter sand) and the loading criteria (150 lbs/ft of sand) preclude Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73 HOLDING OR MOLTING TRAYS BIOLOGICAL FILTERS WASH WATER DISCHARGE SCREEN BOXES PUMP MAKE-UP SUMPS R RESERVOIRS WATER Figure 16. Components of a recirculating system for a soft-shell crawfish facility. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 74 Table 8. Major components of a recirculating shedding system. Function Design Factors Culture and Molting Trays Provide easily accessed space Tray surface area, aeration for holding, separating, and rates, water circulation, molting of crawfish lighting, access Biological Filter Capture solids, degrade Volume, media composition, dissolved wastes, remove media surface area, aeration, ammonia and nitrite water flowrate Screens Capture debris, prevent Mesh size, box size, clogging of spray heads placement Reservoir Provide dilution volume to Volume, circulation rate stabilize the system Sump Provide water for pump intake Volume, turbulence, and control system water levels circulation rate Pump Provide circulation and Pump type, flowrate, aeration of system water pressure Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 75 Table 9. Summary of interim design criteria for shedding systems employing fluidized bed/upflow sand filter configurations (taken from Malone and Burden, 1988). Parameter Value Comment Tray Area 1 lb/ft2 Normal loading density for trays Water Depth 1 inch Recommended water depth in trays Sand Size 1.2-2.4 mm Diameter of 8/16 filter sand Sand Volume 150 lbs/ft3 At least 50 precent of the total sand volume must be in the upflow sand filter Bed Depth 15 inch Assumed bed depth in sand filters Volume/Weight 5 gallons/lb Total operational water volume Ratio ratio for all components Flowrates 0.07 gpm/lb Minimum flowrate to trays 65 gpm/ft2 Normal operational fluxrate for fluidized bed filter i 10 gpm/ft“ Normal operational fluxrate to upflow sand filter 65 gpm/ft2 Expansion fluxrate for upflow sand filter Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76 w oooooooooo 5/16* HOLES ooooooooo >NIET & OUTLET PIPES CENTERED ON SIDE WALLS BRASS BAR "s> 3 8 HCLES BOTTOM- h a .F OF PiPF ONI V Figure 17 Unpressurized fluidized bed filter designed for diameters ranging from 10 to 20 inches (25 to 50 cm). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 77 clogging and/or oxygen depletion problems during periods of high loading. However, the filter bed is incapable of capturing solids and, in fact, adds to the solids loading by continually abrading bacterial biomass generated in the waste conversion process. The generic upflow sand filter configuration (Figure 18) is similar to the fluidized bed. In contrast, the upflow sand filter captures and removes waste solids from the system. The filter also supports bacterial populations on the sand grain surfaces which contribute to BOD reduction and nitrification in the system. The coarse sand (1.2-2.4 mm ? 3 ? filter sand) permits a moderate fluxrate (9.4 gpm/ft“ or 551 m -day/m“) without expanding the bed. Thus, solids are captured in the bed while the solids-free water returns to the system via the two-inch discharge 2 pipe. Once or twice a day, the fluxrate is increased (45-65 gpm/ft ) by backwashing (or expanding) the filter bed to flush out accumulated solids. The two-inch discharge line is closed forcing the wash waters out of the system. Following the backwashing sequence, the fluxrate is decreased returning the filter to normal operation. Replenishment of wash waters from an external source (groundwater or tap water) permits periodic water overturn in the system. Low and variable fluxrates through the upflow sand filter preclude sequential filtration. A single fluidized bed typically operates in parallel with two or three identically sized upflow sand filters receiving influent waters from a centralized sump (see Figure 16). The fluidized bed's flow can be diverted to expand each sand filter during the backwashing cycle. Water treated by the fluidized bed discharges directly back to the sump while solids-free water from the upflow sand filter discharges into the large reservoir. The reservoir's high Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 78 ty 5/18* HOLES OOOOOOOOOO 5/10* HOLES (oeooooeood oooooooooo] w oooooooooo * loooeoeeooo oooooooooo ooeooooo o o T ■ oooooooooo oooooooooo Oft ft _fl 6 Oft ft n Q SECTION AA SECTION A-A BOX FILTERS ty CYUNOER FILTERS TOP OPEN INLET fi OUTLET PIPESCENTEREO ON SIDE WALLS 48 8/10 SAND GRAVEL 1/2-1/0* i \ 3/0" HOLES BOTTOM HALF OF PIPE ONLY Figure 18. Unpressurized upflow sand filter designs for diameters ranging from 10 to 20 inches (25 to 50 cm). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 79 dilution volume provides stability following shock loading when large quantities of immature crawfish are introduced into the system. The volumetric loading criteria relates the pounds of fed crawfish (measured in pounds or kilograms) that can be safely supported by a unit volume of sand (measured in cubic feet or cubic meters). By estab lishing a a practical volumetric loading criteria, the recirculating shedding system can sustain a maximum load of crawfish for extended periods with total ammonia and nitrite levels at or below 0.5 mg-N/L. These quality objectives are conservative, particularly since mortality problems observed in both laboratory and commercial shedding systems over the last three years appear to be associated with concentrations above 1 mg-N/L for each parameter. Research defining the chronic effects of long term exposure (30-45 days) must be conducted before quality objectives can be more specifically defined. The criteria does not discriminate between sand placed in the fluidized bed or upflow sand filter. However, at least fifty percent of the sand volume must be in the upflow sand filters to assure solids capture and removal. A system may be designed employing the upflow sand filter only; however, filter failure, caused by oxygen depletion, can be catastrophic. Systems subjected to high waste concentrations, caused by overfeeding or sudden increases in animal populations (shock loadings), benefit from the robust reserve capacity that the fluidized bed offers. The volumetric loading criteria assumes feeding rates of about one percent of crawfish weight per day based on food protein levels between 25 and 35 percent. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80 COMMERCIAL FACILITY DESCRIPTION A 1600 pound soft-shell crawfish production facility (Figure 19) located in Baton Rouge, Louisiana was used for monitoring the perfor mance of the fluidized bed and upflow sand filters. The commercial facility, constructed in February 1988, employs four equivalently sized filters, one fluidized bed and three upflow sand filters, operating in a parallel configuration. All filters were unpressurized, nearly square in shape, constructed with marine plywood, and fiberglassed to prevent leakage. Filter dimensions consisted of a 17-inch cross-sectional area at the bottom with slightly slanted sides ("V 3 deg.) to facilitate media expansion and backwashing. Filter designs were based on design guidelines presented in Table 9. Each filter contained a coarse filter sand, 15 inches in depth, 3 3 with a volume of 2.62 ft (0.297 m ). Based on the recommended 3 150 lb/ft design capacity (Malone and Burden, 1988), the total combined 3 3 filter volume (10.477 ft or 0.297 m ) could support 1600 pounds of crawfish. All the upflow sand filters in the system were backwashed twice-a-day, once in the early morning and again in the early evening. Makeup water used to compensate for backwashing and evaporation was taken from a quarter-acre pond located adjacent to the facility. pH control was maintained in the facility by periodically adding sodium bicarbonate (1 to 5 lbs) to the water. Aside from the filters, the shedding system consisted of two sumps, four 1.5 HP pumps, 112 culture trays, 14 molting trays, and a gas water heater. During normal operation, the system held approximately 2700 3 gallons (10.2 m ) of water making the volume/weight ratio equivalent to 1.7 gallons per pound of crawfish. This ratio was significantly lower Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. WASTE SAND filter UPFLOW SUMP HO. 2 2 SUMPHO. 550GALLONS SAND filter UPFLOW SAND UPFLOW FILTER 4"0PVC GATE VALVE GATE TRATS 4"<* PVC 4"<* GATE VALVE GATE SUMP KO I I SUMP KO 1150 GALLONS 1150 0A3 HEATER 0A3 0TU 100,000 I3HP PUUP .OACKUr in Baton Rouge, Louisiana. 10 HP HP 10 PUtJP QALL VALVE Q VALVE QALL FROtf POND FROtf UAREUPWATER Figure 19. Schematic of the commercial soft-shell crawfish facility located Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82 than the 5.0 gallon per pound criteria recommended by Malone and Burden (1988). Four rows of culture trays (28/row) were constructed with an indi- ? vidual tray surface area of 10 ft*". These tray dimensions were smaller 2 than those normally recommended for commercial use (24 ft^/tray). One row of molting trays (14 total) was constructed, although only six to 2 eight trays (8 ft /tray) were normally used during peak loadings. Thus, the smaller molting trays were used for culture trays whenever necessary. All trays were loaded at density of one pound per square foot. Sprayheads discharging water to the trays originated from a 1-inch diameter PVC manifold extending the length of each tray run. Stainless steel ball valves installed on the manifold head were used for controlling flowrates. Both culture and molting trays emptied to common 4-inch PVC drainlines which returned water to the sump. METHODOLOGY Two different sampling programs were conducted at the commercial facility: (1) a general water quality survey of the facility and (2) an evaluation of filter performance. The sampling methodology used with each survey, in addition to the analytical methods, are presented below. Data obtained during the first portion of the study was used to evaluate system water quality in terms of nitrification, pH, and oxygen stability. A total of 13 sampling events spread over a 45 day period (April 26 - June 6) were monitored at the facility. All sampling was conducted between 7:00 and 9:00 am. Ideally, the volumetric loading criteria should be evaluated under a constant loading. However, at this stage, such conditions are rarely Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 83 observed in the soft-shell crawfish industry. Source of supply for high quality immature crawfish remains a problem for most operators and pur chases tend to be made in large quantities. Evaluation of the volumetric loading criteria in this commercial facility was complexed because the system lacked sufficient volume to dilute accumulated wastes following intermittent shock loading (200-400 pounds crawfish per day). Approximately three days are required for the nitrifying population to adjust to increased ammonia production. Efforts were made to avoid sampling within 72 hours of a population increase of more than 10 percent per day. However, residual effects of shock loading undoubtedly influenced a number of the observations. Samples used to evaluate the system's nitrification ability were analyzed for both total ammonia-nitrogen (NH^) and nj.trite-nitrogen (NO^). Filter performance in terms of oxygen stability was obtained by measuring influent and effluent dissolved oxygen (DO) and corresponding flowrates (Q) on each filter. Influent DO measurements were made near the pump intake in sump //2 (see Figure 19) to provide a representative sampling of filter influent. Field analyses also included monitoring pH, temperature, and alkalinity (ALK). pH measurements were taken in each individual filter, the sump ( i l l ) , and randomly selected culture and molting trays. Temperature monitoring was conducted in the trays as well as the sump (//2) . Alkal inity samples were taken from the sump (112). Loading estimates were made in the system by randomly selecting holding trays and measuring "wet weights" of crawfish. Based on the number of trays currently online, a loading estimate was calculated. Records were maintained by the operators on the amounts of immature Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 84 crawfish added to the system, mortalities removed, and the daily quantities of soft-shells produced in the facility. Separate from the water quality survey, six evaluations of filter performance were conducted over a 20 day period. These evaluations were performed to lend further insight into the relative nitrification ability of each filter. Influent and effluent samples were taken on each filter for total ammonia-nitrogen and nitrite-nitrogen analyses. All influent samples were taken near the pump intake in sump f/2 while effluent samples were taken from each filter's discharge line. Loading estimates were made using the procedure previously discussed. All analyses (for both studies) were conducted according to the procedures outline in Standard Methods (APHA, 1985). Total ammonia and nitrite laboratory analyses were performed using the nesselerization and the sulfanilamide-based colorimetric procedures, respectively. Alkalinity analyses were conducted on-site using the titration/potentio- metric endpoint method. Titrant standardization was conducted with each sampling event. Dissolved oxygen measurements were made using a Yellow Springs Model 57 oxygen meter. pH monitoring was conducted using a Cole-Palmer Mini pH meter. RESULTS The commercial facility was monitored on thirteen different occasions while the system was near or slightly above its design capacity of 1600 pounds or, in terms of the filter loading criteria, 150 lbs/ft'" of filter sand. Filter design capacity was exceeded on 3 two occasions with loadings of 160 and 165 lbs/ft . Mortalities in the system were normal for commercial system (between 2 and 3 percent Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 85 per day) and soft-shell production was moderate, averaging approximately 2 percent per day of the system's loading. A statistical summary of the loading and water quality data is presented in Table 10. Ammonia and nitrite levels observed at the various filter loadings are presented in Figure 20. Although the total ammonia and nitrite levels fluctuate above the authors recommended guideline of 0.5 mg-N/L, the facility operated reasonably well during the entire monitoring period. The filters consistently maintained total ammonia and nitrite below the 1.0 mg-N/L threshold level. The highest ammonia concentra tion, 2.3 mg-N/L, was observed following a combined shock load of 760 pounds of crawfish over two consecutive days. Ammonia levels, averaging 0.57 mg-N/L, were slightly lower than the nitrite levels, averaging 0.62 mg-N/L. Alkalinity in the facility averaged 216 mg/L (as CaCO^) and remained above the minimum recommended level of 100 mg/L (as CaCO^) for safe commercial operation. Despite efforts by the operators, however, pH levels (Figure 20c) remained below the desirable range of 7.5 to 8.0. Normal aeration in the system maintained dissolved oxygen levels in the sump between 4.9 and 7.0 mg/L throughout the monitoring period. Mean water temperature (30.1 + 2.1°C) in the system remained consistently higher than normally recommended (22 - 28°C). Figure 21 illustrates ammonia and nitrite removal efficiency data obtained from the independent filter evaluations conducted on the fluidized bed and upflow sand filters. Removal efficiencies are bounded by the diagonal lines marked 0 and 50 percent removal while the x-axis denotes 100 percent removal. Close examination of the data indicates that the ammonia conversion efficiencies for the fluidized Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 86 Table 10. Summary of water quality and field data collected at the commercial soft-shell crawfish facility in Baton Rouge, Louisiana. All measurements of water quality data were taken in sump #2 (see Figure 19) Standard Parameter n X Deviation Range Loading (lbs) 13 1351 238 900 - 1725 NH3 (mg-N/L) 13 0.57 0.56 0.20 - 2.30 NO^ (mg-N/L) 13 0.62 0.22 0.24 - 0.93 Alkalinity 12 216 91 100 - 423 (mg/L as CaCO,,) 00 (mg/L) 13 5.8 0.6 4.9 - 7.0 pH 13 7.1 0.2 6.7 - 7.7 Temp. (°C) 13 30.1 2.1 27.5 - 34.0 Flowrate (gpm) - Fluidized Bed 13 103 24 33 - 126 Filter - Upflow Sand 39 18 6 5 - 29 Filters Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 87 9 Or 4 1 ' 4 O ! 3 9 0 2 0 a a < I 9 . I»«*N /L THRESHOLD • • • • . • • • • • FILTER 4 0 - DESIGN CAPACITY G39 Xv ■ t»• 30 E —u 2 9 H£20 5 15 „ l*»-N /L THRESHOLD 0 I 0 T 09 o o 1 ... —i------» ------i 1— ------1— :— - .i------j DESIRABLE RANGE V • • • • • • • • • t • • 60 90J— K50 ‘ no ■ ISO ■ 130 i 140 • 180 190 ITO FILTER LOAONQ CAPACITY (IbO / ft * ) Figure 20. Total ammonia (a), nitrite (b), and pH (c) monitoring data obtained on the commercial soft-shell facility„in Baton Rouge, LA. Filter design capacity (150 lbs/ft ) indicated by the vertical line. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88 Oi C 7 »- 00 09 cp 0 .7 0 4 U. 0 2 0 2 0.3 0 4 0 6 IN FLUENT AMMONIA OR NITRITE (mg-N/L) Figure 21. Total ammonia (a) and nitrite (b) removal efficiency data obtained on the commercial soft-shell facility in Baton Rouge, LA, Solid diagonal lines represent removal efficiencies. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89 bed are erratic with reductions averaging only 12 percent. Similarly, net nitrite reductions in the fluidized beds averaged less than one percent. Conversely, important contributions were made by the three upflow sand filters with regard to system nitrification. The filters consistently demonstrate reductions in both ammonia (averaging 34 per cent) and net nitrite (averaging 18 percent). Thus, the upflow sand filters not only play an important role in solids removal (as discussed in Chapter III) but have a major influence on controlling ammonia and nitrite levels in the recirculating system. DISCUSSION During the monitoring period, the system was repeatedly subjected to shock loadings. Crawfish loadings, ranging from 200 to 400 pounds per day (12.5 to 25 percent of capacity), were imposed to maintain the system at design capacity. For example, the highest observed ammonia level (2.3 mg-N/L) occurred following the addition of 760 pounds of crawfish to the system over a two day period. This increased loading not only doubled the filter loading in a three day period, but also 3 exceeded the system's 150 lbs/ft design capacity. With a volume/weight ratio of less than 2 gallons per pound of crawfish, this facility lacked the dilution capacity necessary for stabilization while the filters acclimated to the increased loading. Difficulties with pH control (as illustrated in Figure 20c) and overfeeding, caused by operator inexperience, contributed additional instability to the system. Consequently, no apparent relationship was evident between filter loading and system water quality. However, these filters appear to maintain ammonia and nitrite below 1.0 mg-N/L Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 90 under typical commercial operational conditions when loaded to the 3 criteria level of 150 lbs/ft . Further improvements in water quality will, undoubtedly, result as the facility's managing personnel gain experience operating the filters, managing loading patterns, controlling feeding rates, and controlling pH. While the upflow sand filters demonstrated consistent reductions in ammonia and nitrite concentrations, contributions of the fluidized bed filter are not readily discernible as demonstrated in Figure 21. A- more in-depth analyses of the filter's ability considers the oxygen consump tion rate (OCF), a direct measure of the bacterial activity of the filter (Hirayama, 1965; Manthe et al. 1988). Results of the analyses for both the fluidized bed and upflow sand filter are illustrated in Figure 22. Single OCF observations are illustrated for the fluidized bed filter while mean OCF values, with upper and lower standard deviations, represent the three upflow sand filters at a given loading. Based on these results, the fluidized bed appears bacteriallv active and, intermittently, responsible for most of the waste conversion, particularly when filter loading exceeds the design carrying capacity. Results of filter removal efficiencies (Figure 21), however, suggest that heterotrophic bacteria, rather than the slower growing nitrifiers, dominate the filter. Thus, the upflow sand filters are reducing ammonia and nitrite concentrations below those levels required to support nitrifiers in the abrasive environment provided by the fluidized bed. The BOD reduction effectuated by the fluidized bed filter, however, decreases the load on the upflow sand filters contributing indirectly to their nitrification rates. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. l ______160 160 170 _1 CAPACITY CARRYING 'DESIGN FILTER LOADING CAPACITY ( Ibs/ff3) t Moan upper w. B 'ovsor std. d e v .) o Upflow Sand Filters O Fluidized Bad Filter 80 90 100 110 120 130 140 150 200 800 600 400 1400 1000 1600r 1200 ■ CJ E O o» 5* u. o O O facility. Reported values for the upflow sand filter representloading. a mean with Fluidized bed and upflow sand filter OCF observations taken at the commercial upper and lower standard devications for the three filters at a given filter Figure 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Failure of submerged filters is most often associated with oxygen depletion (Hirayama, 1965, 1974; Manthe and Malone, 1988). Comparison of the mean oxygen consumption rates to the oxygen fluxrates indicates that these filters are operating safely below their maximum carrying capacity. Nitrifying bacteria are severely inhibited when dissolved oxygen levels in the filters are allowed to drop below 2.0 mg-09 (Gaudy and Gaudy, 1978). Thus, the useful oxygen flux or oxygen carrying capacity (OCC) can be computed as the product of the flowrate and the difference between the influent oxygen level and the minimum allowable filter effluent level of 2.0 mg-0o (Manthe et al. 1988). Design i. flowrates (Table 9) coupled with maintenance of a sump dissolved oxygen level of 6.0 mg-On/L define the OCC as 429 and 2966 g-0?/day for a upflow sand filter and fluidized bed, respectively. The average OCF for the upflow sand filters (292 g-09/day) was observed when the system operated between 75 and 100 percent of capacity (113 and 150 lbs/ftJ), thus indicating that 68 percent of the OCC was utilized. In contrast, the fluidized bed's OCF averaged only 461 g-09/day, or 16 percent of capacity, in the same operational range. Recognizing that the some filter capacity must be reserved to absorb daily fluctuations, the upflow sand filters are near their practical operational capacity. The majority of the system’s reserve capacity rests in the fluidized bed (as intermittently illustrated in Figure 22) . Based on these observations, the fluidized bed and upflow sand filters compliment each other when used in a recirculating shedding system. First, the upflow sand filters appear responsible for most of the waste conversion and removal (as solids) during normal loading conditions. However, the fluidized bed contributes to the waste Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93 removal as a result of biomass attrition in the abrasive, fluidized environment. A second means by which the two filters complement each other involves the upflow sand filter's inability to maintain oxygen supplies during peak loadings. During these periods, the fluidized bed appears to support the bulk of the waste load, thus protecting the system from catastrophic failure. The high level of filter bacterial activity contributes to dif ficulties with pH control in commercial systems. The crawfish, Procambarus clarkii, tolerates a wide pH range without adverse effect (Huner and Barr, 1984). However, above 8.0 the toxic un-ionized ammonia species becomes increasingly dominate over ionized ammonia (ammonium), thereby increasing the animals sensitivity to total ammonia accumula tions. Conversely, pH values below 7.0 inhibit the activities of the nitrifying bacteria in biological filters (Paz, 1984). Recognizing that pH values within a filter can decline significantly as a consequence of nitrification, the influent (sump) pH level should be maintained between 7.5 and 8.0. Despite an average alkalinity of 216 mg-CaCO^, pH in the sump ranged from a low of 6.7 to a maximum of only 7.7, averaging 7.1. Although normal aeration levels in the system adequately maintained dissolved oxygen levels in the sump above 5.0 mg/L, they were not sufficient to contend with the large amounts of carbon dioxide produced by respiration activities of the crawfish and the bacterial population. Theoretical pH/alkalinity calculations reveal carbon dioxide levels, ranging from 17 to 174 mg-CO^/L, persisted throughout the monitoring period. Given the known sensitivity of the nitrifying Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 94 bacteria to depressed pH levels, some type of supplemental aeration (or air stripping) apparatus would seem justified for pH control. CONCLUSIONS AND RECOMMENDATIONS 3 The recommended volumetric loading criteria of 150 lbs/ft is a sufficient criteria for designing fluidized bed and upflow sand filters when employed with recirculating soft-shell crawfish production facilities. The filters are robust, controlling total ammonia and nitrite levels below 1.0 mg-N/L. Results indicated that water quality fluctuations were not related to filter carrying capacity. Water quality in systems which follow design guidelines are controlled mostly by system management rather than by the maximum carrying capacity of filters. Careful attention must be given to managing pH levels in the system, reducing shock loadings, and avoiding overfeeding. The upflow sand filter provides most of the waste processing during periods of normal loading. In addition to its crucial solids removal function, the upflow sand filter appears responsible for most of the system's nitrification. Effluent waters from the upflow sand filters were not completely nitrified indicating that additional water quality improvements can be realized if factors controlling the nitrification rates can be identified. The fluidized bed remained active throughout the study, but appeared primarily responsible for BOD reduction rather than nitrification. Addition of sodium bicarbonate was not sufficient for pH control jn the system. Normal aeration adequately supplied oxygen to the crawfish and biological filters but failed to control carbon dioxide accumula tions. The pH declined rapidly during periods of peak biological Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. activity and high carbon dioxide production. Methodologies capable stripping carbon dioxide from the recirculating waters should be investigated and included in future designs. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BIBLIOGRAPHY APHA, 1985. Standard Methods for the Treatment of Water and Wastewater, 16th ed. , American Public Health Administration, New York. Andrews, G.F. and C. Tien, 1979. The expansion of a fluidized bed containing biomass, Amer. Inst. Chem. Eng. Journ., 25(4): 720- 723. Bitterner, G.D. and R. Kopanda, 1973. Factors influencing molting in the crayfish Procambarus clarkii, Journ. of Exp. Zoology, 186(1): 7-16. Cange, K.M., 1987. Water quality management strategies for production of the soft-shelled red swamp crawfish (Procambarus clarkii), Master's thesis, Louisiana State University, Baton Rouge, Louisiana 70803. Chidester, F.E., 1912. The biology of the crayfish, American Naturalist, 46: 279-293. Cooper, P.F., 1981. The use of biological fluidized beds for the treatment of domestic and industrial wastewaters, Chem. Eng., 371(2): 373-376. Cooper, P.F. and P.M. Sutton, 1983. Treatment of wTastewaters using biological fluidized beds, Chem. Eng., 392:65. Culley, D.D., N.Z. Said, and E. Rejmankova, 1985a. Producing soft-shell crawfish: A status report, Louisiana Sea Grant College Program, Baton Rouge, Louisiana, 16 pp. Culley, D.D., M.Z. Said, and P.T. Culley, 1985b. Procedures affecting the production and processing of soft-shell crawfish. Journ. World Mariculture Society, 16: 183-192. 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97 Culley, D.D. and L. Duobinis-Gray, 1988a. Effects of temperature on molting and mortality rates of red swamp crawfish (Procambarus clarkii) in a soft-shell culture system, In Review. Culley, D.D. and L. Duobinis-Gray, 1988b. Effects of dissolved oxygen on molting and mortality rates of red swamp crawfish (Procambarus clarkii) in a soft-shell culture system, In Review. Culley, D.D. and L. Duobinis-Gray, 1988c. Molting, mortality, and the effects of density in a commercial soft-shell crawfish operation. Journ. World Mariculture Society, in Press. Gaudy, M.S. and Gaudy, E.T., 1978. Microbiology for Environmental Engineers, Wiley-Interscience, New York. Gutzmer, M.P. and J.R. Tomasso, 1985. Nitrite toxicity to the crayfish Procambarus clarkii, Bull. Environ. Contain. Toxicol. 34: 369-376. Haefner, P.A. and D. Garten, 1974. Methods of handling shedding blue crabs, Callinectes sapidus, Marine Advisory Series No.8, Virginia Institute of Marine Science, Gloucester Point, Virginia. Hartenstein, R., 1970. Nitrogen metabolism in non-insect arthropods, in Comparative biochemistry of nitrogen metabolism, Vol. 1, ed. J.W. Campbell, Academic Press, New York, pp. 299-372. Haug, R.T. and P.I. McCarty, 1972. Nitrification with submerged filters, Journ. WPCF, 44: 2086-2102. Hirayama, K., 1965. Studies on water control by filtration through sand beds in a marine aquarium with a closed circulating system - I. Oxygen consumption during filtration as an index in evaluating the degree of purification of breeding water, Bull. Jap. Soc. Sci. Fish. , 31: 877-892. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 98 Hirayama, K., 1970. Studies on water control by filtration through sand bed in a marine aquarium with closed circulating system - VI. Acidification of marine water, Bull. Jap. Soc. Sci. Fish., 36(1):26-34. Hirayama, K., 1974. Water control by filtration in closed systems, Aquaculture, 4: 369-385. Huner, J.V. 1980. Molting in the red crawfish, Proc. of the First National Crawfish Culture Workshop, D. Gooch and J. Huner, ed., pp. 63-76. Huner, J.V. and J.W. Avault, 1976. The molt cycle of subadult red swamp crawfish. Journ. World Mariculture Society, 7: 267-273. Huner, J.V. and J.E. Barr, 1984. Red swamp crawfish biology and exploitation. Louisiana Sea Grant Publication, Louisiana Sea Grant College Program, Center for Wetland Resources, Louisiana State University, Baton Rouge, Louisiana 70803. Hymel, T.M., 1985. Water quality dynamics in commercial crawfish ponds and toxicity of selected water quality variables to Procambarus clarkii. Master's thesis, Louisiana State University, Baton Rouge, Louisiana 70803. Jaspers, E.J., 1969. Environmental conditions in burrows and ponds of the red swamp crawfish, Procambarus clarkii (Girard), near Baton Rouge, Louisiana, Master's thesis, Louisiana State University, Baton Rouge, Louisiana 70803. Jeris, J.S., C. Beer, and J.A. Mueller, 1974. High rate biological denitrification using a granular fluidized bed, Journ. WPCF, 46(9): 2118-2128. f ■ ■ ' ' ' ' ' Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99 Johnson, S.K., 1982. Water quality in crawfish farming. Proc. Crawfish Production and Marketing Workshop, J.T. Davis, ed., pp. 16-30. Kumar, S., 1984. Closed recirculating nitrification systems for soft-shell crabs, Master's Thesis, Louisiana State University, Baton Rouge, Louisiana 70803. Malone, R.F. and D.G. Burden, 1987. High rate nitrification in blue crab shedding systems, Proc. of the 2nd Nat'l Symposium on the Soft-Shell Blue Crab Fishery, M. Oesterling and C. Plummer, eds., in Press. Malone, R.F. and D.G. Burden, 1988. Design of recirculating soft craw fish shedding systems, Louisiana Sea Grant Publication, Louisiana State University, Baton Rouge, Louisiana, 74 pp. Manthe, D.P., R.F. Malone and H. Perry, 1983. Water quality fluctua tions in response to variable loading in a commercial closed blue crab shedding system, Journ. of Shellfish Research, 3(2): 175-182. Manthe, D. P., R. F. Malone, and S. Kumar, 1984. Limiting factors associated with nitrification in closed blue crab shedding systems, Aquacultural Engineering, 3: 119-140. Manthe, D. P., R. F. Malone, and S. Kumar, 1985. Elimination of oxygen deficiencies associated with submerged rock filters used in closed, recirculating-aquaculture systems, in Closed Blue Crab Shedding Systems. National Symposium on the Soft-Shelled Blue Crab Fisheries, P.M. Perry and R.F. Malone, eds., pp. 49-55. Manthe, D.P. and R.F. Malone, 1987. Chemical addition for accelerated biological filter acclimation in closed blue crab shedding systems, Aquacultural Engineering, 6 : 227-236. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 Manthe, D.P., R.F. Malone, and S. Kumar, 1988. Submerged rock filter evaluation using an oxygen consumption criterion for closed recirculating systems, Aquacultural Eng., 7: 97-111. Nutt, S.G., H. Melcer, and J.H. Pries, 1984. Two-stage biological fluidized bed treatment of coke plant wastewater for nitrogen control, Journ. WPCF 56(7): 851-857. Paz, J.D., 1984. The effects of borderline alkalinity on nitrification in natural water systems, Ph.D. Dissertation, Polytechnic Institute of New York, 125 pp. Penn, G.H., 1943. A study of the life history of the Louisiana red crawfish, Cambarus clarkii Girard, Ecology, 24(1): 1-18. Perry, H.M., J.T. Ogle, and L.C. Nicholson, 1982. The fishery for soft crabs with emphasis on the development of a closed recirculating seawater system for shedding crabs, Proceedings of the Blue Crab Colloquium 1979, No. 7, H.M. Perry and W.A. Van Engel, eds., Gulf States Marine Fisheries Commission, pp. 137-150. Powell, M.S. and N.K.H. Slater, 1982. Removal rates of bacterial cells from glass surfaces by fluid shear, Biotech. Bioeng., 24: 2527-2537. Rich, L.G., 1980. Low-Maintenance Mechanically Simple Wastewater Treatment Systems, McGraw Hill, New York. Scudamore, H.H., 1948. Facts influencing the molting and the sexual cycle in the crayfish. Biological Bull., 95: 229-237. Short, C.S., 1973. Removal of ammonia from river water, Water Research Association, Technical Paper TP101, Medmenham, Bucks. Spotte, S., 1979. Fish and Invertebrate Culture - Water Management in Closed Systems, 2nd ed., John Wiley and Sons, New York. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101 Stephens, G.C. , 1955. Induction of molting the crayfish Cambarus by modification of daily photoperiod, Biological Bull., 103(2): 235- 241. Turakhia, M.H., K.E. Cooksey, and W.G. Characklis, 1983. Influence of a calcium-chelant on biofilm removal, Appl. Environ. Microbio., 46: 1236-1238. USEPA, 1975. Process design manual for nitrogen control, Sections 2-4, Office of Res. & Dev., Cincinnati, Ohio. USEPA, 1980. Process design manual for onsite wastewater treatment and disposal systems, Office of Res. & Dev., Cincinnati, Ohio. Ward, R.C., G.M. Smillie, J.C. Loftis, and T.C. Sanders, 1982. Relating stream standards to water quality monitoring practices, Final report to Nat'l Sci. Foundation, Washington, D.C., Report No. NSF/PRA-82001. Weber, W.J., C.B. Hopkins, and R. Bloom, 1970. Jour. WPCF 42(1):83—99. Wheaton, F.W., 1985. Aquacultural Engineering, 2nd ed., Wiley-Inter- science, New York. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDICES 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX A WATER QUALITY DATA This appendix provides a listing of the raw water quality data used with Chapter II (Grain Size Determinations) and Chapter III (Filter Evaluations). Appendix A-l presents the grain size data as sort by sand size: Fine (0.42 - 0.84 mm), Medium (0.84 - 1.68), and Coarse (1.19 - 2.38 mm) while Appendix A-2 presents filter evaluation data as sorted by filter design: Fluidized Bed, Upflow Sand Filter, and Fluidized Bed/Upflow Sand Filter Combination. Total ammonia nitrogen (NH3), nitrite nitrogen (N02), nitrate nitrogen (N03), total kjeldahl nitrogen (TKN), dissolved oxygen, total suspended solids (TSS), volatile suspended solids (VSS), and biochemical oxygen demand (BOD) are reported in units of milligrams per liter (mg/L). Alkalinity (ALK) units are mg/L as calcium carbonate. Temperature units are in degrees Celsius. pH data is reported in standard units. Unreported data is denoted by a 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX A-l GRAIN SIZE DETERMINATION DATA Influent dissolved oxygen data is denoted under the heading 'DOINF' while effluent dissolved oxygen is designated under 'DOEFF'. Total wet weight of crawfish (reported in grams) held each system is designated as 'WGT'. 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced HED!A=Flf4£ i/i o o ^o -^o o o o T 0 -4 o o o it o e o ^o o o in ooonoo ooooooo o ^ tc o n a t f > . . > f t a ’ O« ^ O O l f O ^ O O O O ^ O ^ O O O O O ^ O O ^ - ^ O O O O O O O C O O O S’ O « 2 rg w> 165 6 29 32 165 26.0 0.723 0.030 3.10 7.90 7.05 1.06 9 9 9 166 6 30 33 165 26.0 0.535 0.006 3.10 7.30 6.95 1.10 167 7 1 36 165 25.0 0.560 0.006 7.95 7.30 6.90 1.03 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced ME 0 IA=FIM6 a IU U »- l- z o Z r r z O LU «r 3 Ja2 a -J j ID a n © m © X> «X a: a QC o a. u. z * _J r CL 2 j a m X a. .• U. © x © z p n©r»«O © O' r*» « © in a o o r o A“ © r- © oro ro PA aj pH © ia H © in o O o O r*~ r- S O KS AJ nj o r o ph o © j- .y & • • • 0 0 t H 9“* O' A- A> ro AJ o o © o r- o r- o o o o o o © in «o o 03 a T p f p in AJ AJ AJ AJ ne in in © © A* © P- p- © 4- K or ro ro ro p o rw h- K> o m ro p © o AJ o © O AJ © © -S' O © o o <■4 Jm AJ O' O o & H H e H 0 0 0 0 0 0 o 0 pH 0 A* o © in p © © o in pH o y JAJ AJ pH <0 in ' O 0* © in aj o r aj AJ H 0 0 0 0 p 0 © A- in p in p f- o © o in o r p 0) in p O -in t- in h* 03 © CO 0) in © 03 © O pH O o 4* H 0 0 H 0 0 H 0 H H © AJ A- © p P- p- © © ro p- o -in p- 0 i/* T t U.I O O H o w roat- ro ^ ^ m j* r\j < j ^ ••••• w *•••• >- * z ••••• OVOQ • • « • « r-n mh» «0u j k(T1 ^sX) il J O TJ n •* m .* .*■ o« o • n* o o o • o o o m o ^ >4 f4 O N 0. I • • • • • J K N K o O IT IT o f- \& & in © e o a» u. u. • • • • • ~ 4 O O O •-• o in in o o h- *» in * * & * a o *■* z u. • • • • • cm fn nj m •* a n in ii) co o uJ ^ «C f» f ^ m «-4 ^ ^ ^ ff> N H « •O G* «Q (N n j M M N M IM M IM IM M 3 (j v- w njn rM r jm Q «J V Z IT N C O o W «e vn ^ h«j ^ a «t > s a z v - x in m & in vo o co in moooo Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108 o CT» -< <0 H *4 h J J m o o o *^*-4 o o o «-' fla5'j0vfl\a^)|J'^)j5,O'D'fl H H H H H H H »»4 & sO o c o o o o o J’LT J’LT ec ^ •• h o o o « f*''flc®N'0<3it3'Cj,^oiri'£irJHirir^PH'NJnNwc*-< f*''flc®N'0<3it3'Cj,^oiri'£irJHirir^PH'NJnNwc*-< ••••*• •• •••«••«••«••••••• ••• •••«••«••«••••••• •• ••••*• 0 sflQOX'h»ui ^o o M f l f k«j.ffl'ino^Hwa'rtMtoi/iNj)in ^o o M f l f k«j.ffl'ino^Hwa'rtMtoi/iNj)in sflQOX'h»ui 0 « *••• ooooooooooo r * r*n O *-4»-4f'jm,xir\>a^«og' W NKN^KNNNNKK|i»h.Kffc^>.SN^.|/(|^trtiniflUlini/'l/1WI/lir>lflin W NKN^KNNNNKK|i»h.Kffc^>.SN^.|/(|^trtiniflUlini/'l/1WI/lir>lflin X CL U. > 2 J u, o « o e « • h L— l/) t l/) L— ana » Ofci/^ ajfooff'trj'O'O'ff'ff’cru'crooocoooQOO'H^HHHi^rt^H^ f'jOfci/^ UJ zc Q. zc UJ xoncfc- xoncfc- - n«T l Q C* Q O O Q 1 C 2 ^ I ^ 2 (Pt^lpv0t£^*ifitjOv0'A'£v0'£>£viO'£vfi*-O'£'£k£v£<£gOv0 C 1 i # m o ui cocoonooooaooooo^ooocoo^ooo^^oooQC)ui o m # i 10 HOID3H=VI09W Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced MED Ifl= MED I U M X u i oo*^oooooooooooooooo*- 43 A O ro 0 © O A* 0 0 0 0 0 0 0 ! or roro ro 0 PM PM O © n4 )4) 4) in 43 O' o 04 in in 3 A O' © 43 A* ) 4 34 )4 M3 43 4) 43 43 4) 43 -4 **4 A- O ro _i AJ O' ro o ^ 43© A* PM o o © O ro ro A-© 04 AJ A» © n OO in O 0 0 0 0 0 0 0 0 0 000 0 0 0 7 I 0 0 0 7 •*< pm aj 7 0 0 7 AJ A- 04 H © © O © 4 0 MAJ PM zr in 04 in 04 O 43 m AJ o 04 »A A* A* A» 43 oAJ ro O 04 O n43 in nPM in 0 0 0 0 0 0 0 0 0 MAJ PM A* AJ H © 04 O o O' 04 zr O *4 Jm AJ o©ro © ro o O •S' © 43 © «? O O '©4) © O' 0 0 0 ro 0 O ro & ro AJ 04 A- 40 ^4 04 04 4) in in 'OO' O O' PM ro o 0t zr 04 04 n© in © O in tO in AJ O oO o o o 0 0 0 0 0 0 0 0 0 0 AJ 0 ro A* )4) 4) 04 ro AJ H ro in m 4p4 04 nin in in 43 O' K Mro PM © in in 0 0 0 0 0 MPM PM j 0<3 <0 4 0 in o O O or ro ro in ro o o AJ o 34 3J* 43 © 43 43 43 © o o © in in c <0 43 43 *r- A* ro I04 AI © O' 0 0 0 0 0 0 0 0 0 PM 43 zr & 04 43 O' 404 «4 AJ ro o «4 nin in o* O' nin in m o o 43 in O PM «3 O 0 0 0 0 0 0 n43 in AJ & 0C3 <0 ro 43 A- 43 ^4 AJ 4H 04 o in 04 ro O' O' r- P* in m rr> o 04 0 0 0 0 0 0 0 0 0 0 0 4 0 AJ 43 © -J* t4 43 p O AJ & O 4 0 *in 4* •4 AJ o A* H © in *© J* 43 "A © A* A" O PM o 43 O nin in * 4 *r- J* O 0 0 0 0 0 0 0 PM © A* © 43 4 0 4) 4 0 JP AJ PM AJ in n 43 04 4o 04 in ro o 04 J* o f— 04 in O © O a oA AJ AJ ro © o O © O © 43 0 0 0 • 0 AJ A* o tO © 43 © O ro r4 H © o ro O' P* o 43 O o O' O' © Jro AJ 3in 43 'O O' O' O' o in OPM •O -A 0 0 0 AI A- © O' © A- P» 'o O' © 04 04 © o © O ro 04 4 0 o •d O' 04 o o O' o © o © *A- A* O © 0 0 0 0 0 • 0 0 0 0 0 AJ nin in A- © PM 43 ro 4 0 PM ro O m © A- 4) nA* in O in 43 in in o o o o A- © O' 0 4 0 0 0 0 0 AJ ro 04 F*« © © © 04 *4 ro MAJ PM o 4 0 A* o O Q AJ j in O in 4 0 4? in A» A» o O 0 0 « 0 0 0 AJ in 04 Jro AJ © O A* ro 4 0 4 0 o 4* o o 4 0 o in O A- in in PM j* AJ © *A A- A- A* o 'o O' 04 © o • 0 PM in in 0 04 r- to © 4PM 04 oro ro MPM PM ro AJ PM m MPM PM ro 04 O ro 4© *4 © A* oPM ro O © O © O in PM © in 0 0 e 0 0 0 AJ 0 r4 r- © .0 © PM in ro AJ AJ in 4 0 H O' in ro © © © o PM ro AJ in 4t « 0 a 0 0 0 0 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced HEDIA=MEDIUM i / ' I i f U . Q I / 1 x a m m a o o a Q - u x a a x iu ►- u _j a h t C X cc C a SC O * f - _J * ^ ul ►- * z co a a u. j o «« a O a u. x • 3 g a m z u, z II (T C tf > 7 n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced MED IA = C OUft S; / J D^ 0 < * « 3 e ^ O O < - i < - > * H - O ^ O H * 0 < - * Q 0 0 0 0 0 0 0 0 0 0 0 3 I/' C > C O O l C Q O I « OJ D ^ ©2 X IS^(/'tP©\£v£' aj • aj 9* aj aj • • • • ...... aj j A j A / A j A j A j A j A j A j A i A j A g r j A j A j A j A j A j A / A j A g r 9 • • • * o » I N O & ■ © ...... «I ..... « « « ( •I • | | • H 0 « 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced NEDIfl=COU«SE i n t I O ) i Q / l 2 H o D O a 3 ^ 2 O Z _j O 3 O j _ . u a »r > >- Uio. c k e f Q J Ih ft tf ^ J a cceft o MED I ft = CQUS SE / ri i> i N N u r u r J i M H M N O O H O O O H H O C P c i/1 ir i/>iu o I a a k- Ju. UJ a a .-j - u. a - -©oi-p,*©*Cp P-»-*r'>cp»«o.©sOm©onj«op* p-©©.ooi»-»p-,j*?©o*'C©p“ »-oa: r «r c z a IJJ 3 z kM z ««••*•««••••»•«•»•••••«•«••••••• • • • • • • « • « • • • • • » • « • » • • X • • « « • * • • o « « o k- 4 £0 e > ar a z X a a. > / 0“ 1/1 -OJ k- z X .• U. go rg IL Q. ro 5 X vO OJ ro r4 0- <0 M OJ o O 0*1 JOJ OJ ro vO JP nin in P« OJ *4 to Jm OJ vO N. in vX) © ro in © o P* in JfO OJ O e O Ol OJ «0 H p4 © OJ in o sO ro O in O Ol p- © op- ro ro v0 o O O in o H OJ 9 4 9 4 4 4 • P- 0 tn ro CO OJ o cn OJ OJ OJ OJ IT in P» in s6 o OJ ro o*4 ro O© vO 3? p o O■0 sO in 7 © (P © «3 *7' © o 4* o o O O 4 4 4 9 4 4 JOJ OJ 0 eO in ro nin in n jO jO JO OJ OJ OJ OJ oj OJ oj ’© O’ 4 c0 JO OJ OJ OJ «o ro OJ 4H ^4 CO OJ o <0 p* P- OJ o in o o 9 4 9 0* - r © ro © © sfi to ro P- O •■J JH OJ P- *4 - H r-4 o O o oOJ ro v£ v£ n© in O O © 9 4 4 4 9 9 9 9 94 9 ' P- © p- ro © OJ j o O o O <0 ro nin in p- P- J* -4 IT © •4 in © •M © O OJ in in P» o o in o 0 9 4 4 0 0 9 ' 4 4 P- IT © © OJ © ot oro ro to ro ~T OJ © C J* 4*4H *•4 ^4 *4 OJ ro IT ro J© OJ 0J in n© in O p* © o in © OJ O ro © OJ © o © in IT ro O ■T & # Jro OJ OJ © o © OJ CO 0“ O n-p in ro o P* in © © -3- © o P» © o 4 4 O' ro O o © © © © © OJ o © © lin Ol OJ O in OJ o © p- in o oj & 0' 4 4 4 4 4 © o^4 ro oj ro ro in *■4 IT XT 0J no in Ol in © p- © OJ O *p- p* o o © © © o j 4 9 o 4 > © © oro ro © © *•4 OJ Ol © H O' J* o © © OJ o ro .* o © © © in T © O o in in -©o © P- © o © 4 4 4 4 4 © J* j o O p- © OJ in O' ro OJ ?in >? O *•4 mi © o ©■ OJ ro © o © O' tn in p» p *“4 .© in 0 9 0 9 0 - . n© in O -0- 0- © © © o © TITIT O' Ol o © o ro in in ro © r4 © © H o © *4 ro rO o o © © n©i i © in in © in 9' 9 ro n©0-© 944 4 9 4 4 > o4 H 4 * ot4 *>4 0 © © 0 IT o -S' ro 0* ro m n i © in in M ^4 •M © o n © in ro O o ro ro o ^ ro -4 ? 'r- r O' g? © o © F— ro o o o o i *© ©to © © 0* in m ro © 999 9 9 4 © 1 0J OJ 0 7 4* 9*4 *4 »4 0 2 T T 0 9-4 o IT M * f ro n ©N© © © © © N © in o © Ol O o 4* J* oro ro © o l© Ol © 0* o © ^4 ^4 0 9 4 0 0 0 O' OJ mi © o -o- 0- m in •4 ^4 © ro oro ro OJ © ro OJ © IP •H . o o © *4 O o ro H 4 0 ro © Jr ^ © ^ ro OJ © © © © o tO © n H in © Ol J JO JO OJ OJ OJ OJ OJ OJ © «4 © o © © m © r* 0- © h* o *o© o O* ^ O K o 'K p* K O' Ol in ^ © © 9 9 4 4 0 4 4 4 9 0 © © © 0 © © ^ 4 >4 H >4 o o ro ro ro © © © © © O H | p O © O - 0 - 0 - 0 J p* j o OJ J l OJ Ol OJ - j eo Oj P- © © m o o .ro p. ro to a a a J OJ © OJ © © © n o m in 4 9 4 9 4 4 0 9 0 4 9 4 9 4 9 9 9 4 0 2 7 0 © ro O O O O O © o o •T -T 5 7 ro ^ s mi 0 q 5 7 © O -« © «} P- © © © mi © o O © © © P" rrn •*4 im mi m4 mi ■<4 imimi JOJ OJ o © ro © 3o© o *3 . © o o © © © mi 1- o © © 4 4 9 4 4 P- © ro -p* P- OJ •>4 © m e3 © *© p* © © <0 © © o P- OJ © o O © -O© O P- O MED IA = C OJ R SE n a UJ z pH p o o ro04 OJ 04 rg O o p o OJ O in in o «n ro tr h- p rg UJ -U O- U) o- o ro ro p J VO UJ* nin in UJ 3« © «0 <3 pH p ro ro fO tn oj OJ *n tn UJ O' H H • H « • • a a • • a • • H • H o 0 pH ro O' in 'o© o O' O' p ro pH tn pH & & & & OJ OJ r- s£ o o in pH o* uj o o U) pH o * H • a a • a • • o pH 0 p p r*» pH xJ3r- p tn COO' O' i o 04 p h- © © o o -U> u> © r- gO rg in p UJ pH O' 04 uj OJ UJ o 0in *0 a a a a a a • H H H H H p pH pH OJ ro pH p P- © OJ O' rg p in OJ OJ in >u> u> U3 h- UJin o in O o in o tn o <8 03 O' UJ OJ H © a a H H a H p 0 p © p p OJ Oj O' © in H ro €0 p o p OJ N- in O in o J* OJ « O in p ■H «oo- UJ H a H a H H a a a H H H pH 0 p ro p o OJ in p J- O' OJ p OJ p f** OJ in o r- UJ O' ro O > n* O' «o * p H a a a a a a J H H H H 115 APPENDIX A-2 FILTER EVALUATION DATA Effluent dissolved oxygen data for the single filter configurations and for the fluidized bed filter, when used in combination with the upflow sand filter, are denoted by the symbol 'DOEFFB'. Effluent dis solved oxygen data for the upflow sand filter when used in combination with the fluidized bed are denoted by 'DOEFSF'. Influent dissolved oxygen for all configurations is denoted by 'DOINF'. All dissolved oxygen data are reported in units of mg/L. Flowrate data for the single filter configurations are reported as 'FL0W1'. Flowrates for the combination filter design are designated as 'FL0W1' for the fluidized bed and 'FL0W2' for the upflow sand filter. Number of crabs held in the system are denoted by 'POP' while the number of crab mortalities are reported under 'MORT'. 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced OHSICN=FBED VOZ&UJQ^ C» C3 I'*1 U © c. ►** U. X C O i, 2 H 2 O J , li a -J l X © X U t UL to U. u C i i a of k X X X X ' u 'y a, Q. C CL C Q. Q O i CO WO Hi J < c 2 - X G- 2 £ e b- O O C ^ C rj h £ * CO * c X N ^ O H O N H H n N l f t r N O O H H O H H O l C N O O O O O O O O O O O H ir p * - i * - c o < r < o r - c o c o o o io - T O H t r o m ^ n f O J r o o o c o o o o o o a o o o o o o o o o o O © © O < CM •O O *H i co f*N C © c o o rH ' UN it rH CM © o CO rH h» xO ^ CM CJ CM fvjm O O c O O 0 o o o o o o o 0 0 0 0 J « h h OOOlf' IT. t rU * NU NU Nt / N HfH tH pH o o CD © CD CO © CD I O CM CM CM O' M3 o ri x© ir sr m ro rJ O 0 0 I-$ • f M ■ C J c C o M r < N M m xC 0 3 r * ' c r > o o o r i o o O »H fH UN o o O O O O C o O o o o o o-o o 0 O oooo fH O f-*P» 0 0 rH xO O 0 o o o O HpH fH fH CM H xO CO o O CM O c CD c r» CO 0 0 « 0 0 0 0 0 0 0 « Xtf* UN ^DrHf'u, >cDcroHNf^*4’iri>oNa3 H N WNrNNNfnc»1fn r'N N NW NNN HHH H H H H H H H rH m xO r- O *H m fH o tr h- CC CO 0 000 0 o 0 0 0 0 0 0 0 0 UN O UN H rH r» ca CM «H CD CO CD o O • HHHHHHHHHNCMMNNPJr: c' a»oo • H CM rH m xO u* O UN tn D C 3CO 03 o O c rH O cr CM c P- > 7 HfH rH O UN O xO <* M3Is pH CD UN 0 0 0 0 0 0 0 7 7 8 fH UN rH xO c © 1 '$'*HUNmnNfnfMCDma' 7 7 a xO p CM CM o r- xD © CO CO O o o H 0 o 0 0 o 0 7 p fH M3 f- fH tr rH O m c © x^ H 0 0 0 0 «*®«000**®*0®0*®0 0 « V3* xO in p xfr < MM3 CM x xO © CM *H X^ UN O rH rH O' xO n h- in o in rr r- o Co CC ' UN O' O C © q 0 0 0 9 0 © o * ? t xfr UN xtf *? 9 rH rH rH © © © © r** O' O' co r~ O CM © 0 0 0 0 0 0 7 r- 9 r- cm o © o o o • 0 8 fH fH fH fH fH H O f l f O xO 0 0 P ® O' ' O n i O CD P O P CT*O CM O' CM UN CM O' CM n i o o o o o o o o o iHrH 0 O' o 9 0 o 9 in xO fH p H 0 0 UN rH O C?CO CM CO fH 0 000*0 r \ rH xO 0 NUN UN >0 o* a o O' n > 0 rH UN o m c O' h 0 O c o fH o o fH o* O O O fH CM 0 0 • 0 0 158 7 1 33 40 22 0.481 0.450 7.45 7-10 . 7.50 0 159 7 2 34 20 22 0.174 0.136 7.75 7.40 . 7.30 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced i'.. c a DES T »M= wiICJUJQVJ c c . j i . u - u . r o . cm -J 3 C h s o j £ £ O K •" z c_ z c 7 • o O O c rH a. 4ro 04 OJ 'O Ol v5- 0 ft • • 7 9 •c cn m - P >o OJ cr r-' ® O' ® o 04 o o o t9 ft ft • 9 e e H r i o c o ro m O *0 rs r- r*- *N o 04 o a r < o ft • • • • 9 f t f t f t f t f t f t f t ® f t f t ® * ® f t * H * - #* m vO in CC - 0 ! O H r M? rHOJOJ O' ? s C 0404 m 04 cr- OJ ro O o cr o o « ft 9 « • in H H H H H H H H r H » H f H p P- m IT. o O' f-H 0 r H r H r rH H r H r OJ r- OJ cC. r- o O O O O O o « a ft ft * 0 o- n i 0 <- ^ Hr H r rH rH m m 'flN'O CO o ^'O'OvOirirmy, H r O' - 0 sO p» - o- p- ft * ft •e ro 0 tn 0 © v4- OJ rH ft 0 0 ft 0 O f <3 OlOl 04 ^ o O o o o O *0 ' O fO ro ro o o o o O - a ft ft ft ft • p oj 0- r- 0- r- IT fO 0 m *o Ol o HrH rH h n c* 04 ^ O o O o ^ ^ m H ? -f*- S* 0 ft 0 rH rH rH H r H r H r ir. a . • 0 in H r a? o rH a: o rr 04 e ft © - C* IT N? r* H r 'O 4 r h- rH r« r* in CC 04 m o rH m c 0 o sO e ft e • * 000 p cnsO(7NiAora'NN-or. cnsO(7NiAora'NN-or. h© rHOOOOOCOOOOOOO OCCOQOOCOOOOOO m OiO(r/^(rx,H 'Ori'Oh(Mr‘/O^H(7rjxu, ooocooooooococ* H OftftftftOOftftOOftft f t f O O t f t f O O t f t f t f t f O • fffott •«ftftf«oftft * • o • « *"H H * H f M C H r n O "O to so #od ««■••*■••••• '^lOOOffiOCHNfn 0 o' m cn 0* * 0 n c m ' o 1 ^ ^ o H K fi ^ f M O J m rH rH C m M J O M f O J Coooin M ( M a *'3''O *£>o irm 'O m fl®*## ftO ft 0 e o • « c n h ^ i * r r * v f j o h r * c * ««•••« fH r^ O CO -v O' H 4 M 4 O r 04 CM 04 r j r ® o o c O O IT (?r>^ •■•••-* ft ftft ft ftft 8 <6 *®® 0 ft >£ ^ j 5 5 68 80 23.5 0.724 0.471 6.20 5.30 . 7.10 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced PFSIGH=FBF t r t x o r u . o «/» «/» o . u r o x t r t c •i- u c . T s s a ctH . u r: c »- -J -J C U- cc Z VS - H c z < mim H- z a z H- L. a CL 7 1 7 U 3 OJ 3 U o z > cr v, cr 0 r • r o H H 0 cl H u a3f^aDr^^'4,'^'^r^cor~'4*0'0 9 in - f 3 0 u vO o © © CO CD 'T in sO'ONQO'CiT'^'tf'ofir-u*. nC C IT. «£ <0 ff O0'0CDN»fCl CDr’i»i(rlhl>Q3 r ( ^ ' ^ N ^ rH O' ir tr fw ^ sC “ P - r IT r 0 * U' n i m o w (*** r* m *o o *o 'O »o *cr« o O ' O ' O •sf* ' O ' O ' O v O ' A i ^ ' x u i T i ^ u o s o *o r- o o * r ' u o « p o r H ,C Ti ( 0 " * f ^ ,n ' ( l f f O N n D ( H i f h 0 3 Q t r 3 *-4000 a 0' 0 rH O' ' O n i L 03 “ P n* © C O O O O C O s f l O o n i O O O O O r i m tr f— IT o c U' in O s IT c rH in M I nJ H IM O C co © < f o u ' r g r o c O ' O r - r H j ^ ^ a ' C J r H CO O O ' c m *£> H 0 O rH O >0 vO 0*c >0 r» m o o *0HI m 4 0 4 4 0 « • 4 0 4 4 4 4 4 * 4 0 44 4 0 4 4 0 4 0 0 4 4 0 0 4 4 0 4 4 4 4 4 4 0 0 4 0 • 0 0 0 0 0 m rH IT in © rH rj fH o 'O r* o > CD c r« r«- rH vTl NO o O . 0 a 0 0 9 o 9 0 0 H If* O O W 0 ' 0 ' £ ' O s £ ' ' U O ^ ' H - ^ - h o o o o o o N o n i o o n i o o ^ i o o o ^ o o 4 4 4 4 4 0 4 4 4 4 4 4 4 4 0 4 0 4 0 0 *1 O1 ’O n i N ^ l r H r H * H * *-1 H H « rH HrtHrt r H t r H H H H H H H H l r H *0 0 O ' O ' O ' n i n i o » o « n i o * m m O* o r< n o *0 H H H H H I M fM M I p- p- P- o o O M O H r C4 *o *o m X in in - O f if* in Cp* O H r O O O fn *-* r> > O ' >0 A I m in in r H X H r H r O' ' O - r n f o o O O H r rH H r H H r •o *o *o *o *o 0 0 9 0 9 • 0 e 9 0 0 0 0 9 999 9 9 O 0 0 8 5 0 0 ' o* m O' h f»3 pH - b cc rH 6 9 0 IM * r H r 0 © o o '•O CO o o O m H r - r O O o IN o o m 0 0 0 9 0 0 0 m O' O m 00 m © in P* O o rH m fn rH © O rH P- O O o © H) e 0 0 0 9 0 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced c* z I=SF ILT W I IUK Q U1 OOOOOOOOOOOOOOOOOOOOO^tO«9-rH»-IOOOOOOrtiH r r z z r r c C» (j u t. c c M N C M C M fMoOt-ir.:rr-' < in 0 © o c © O m CO rj cm m CDcc e © iH 9 9 0 9 o 0 0 0 N^^r^f^s^fn^vnsn^CNTirir i c m i m in m m ir r ic m J r m i r i T N C ^ n s n v ^ n f ^ s ^ f ^ r ^ ^ N '£l' in p^oooocnonNONhm-ON'O^cD-jHOP'Oc p^oooocnonNONhm-ON'O^cD-jHOP'Oc rg^rvj in 0 cn rg r*- cn © © 9HCo r- o fH r: CD o m 9 o 9 0 m '£^^)^'C, in O o o o r*- o ITin m © m © © ir- cn 9 9 in HfH fH o ^ ® r— co ^ © o r—in m in SO 9 9 0 0 9 0 9 9 r i o a CD cn r* CM■** in sj* If* o 0 9 9 0 o e©o ©oo©o o©o©©© © oo© ooo© o© oooo© o© o© eo© O m eg SO %o o i s nC M C ? - M r i M s j ’ ?in s f r m DCD CD in CMm m © © o o ' 9 9 e 9 9 0 9 0 0 0 o o © CM sO rg 9 9 0 ^}^O^)O'£( fH in © O CD nmin m in O CM in in * 0 9 9 0 O o r- ifin fM o in if. fH c c M o © s* CD ' O 9 9 9 9 9 CM o in o © CO CM© © »H 0 9 0 0 o 0 0 0 0 9 0 o H !»• © m © © DCD CD fH 9 0 9 9 o o H © o «o rH NO o o 9 0 o s * > 0 r H I f CO o rH M3 O '•0 H-o rHr- 0 9 9 0 9 0 - M M H o © 0 in if' in ri O O o © © © vO If r~ m o o r- m h- O 0 9 9 99 9 9 0 0 9 0 0 9 ^)' 8 cm 9 8 O Hino 0 o © o 0 9 0 0 9 0 0 '£k e o o CM o o CD O m o in o © 9 9 0 e 0 0 9 0 9 s m 0 o ' O m CD fH nco c in o sOm c 9 9 9 9 0 0 '£,£' O r- n n o o in cnmrg © FH C © © © © 0 o 0 0 O O h rj K m o © fHfH fH m CD h- o 0 0 0 ^ r ^ 0 cn CM fH fH O' O o o r»If fH 0 9 9 9 0 o 0 ' r r 0 O H CD © is. O 0 9 0 9 0 0 c *n inn ri n r* c - h i o •H in 0k0 rH fH o IT © CM o m 9 9 9 e 0 9 O CM H U' CM PH CD O o o o o a? m 9 9 0 0 9 0 0 'D fn fn 0H sO u.O fH pH o © o CD © *o o 9 9 0 8 0 9 0 9 cmm O © o jL*.rj o PH CM T'*■ >T M)in DFS !ON=SFILT QCufucc C C* u a. QV) O O H * H * m Q O O H r O n f O O O O O H f 1 - 9 Q H i O M < O H f i « t O O M f O H f 0 0 4 - r H f ) V JQ U T C I ) U C*I- X X X rh7a. 7 h r c JC2H 2 C J L it. C < >- j cnah atc w r O C «4 >» « i . n. x Q - , < \ j < M r M f M r H < M < \ j f ' ' ' $ - p " , ' 4 * f n r , v » < 9 * v ? ,f n < M r M f M f M < M r M r v i r r > < M < M r , > » H r v f r i \ m N ? . 4 * X. X n cr r»fON?(flor*.f>.fp'r.rg zc p* vrcor-f-«mf*-ma irr'rjpjHN-o^MHnfo^ir r*. Q- X *s> r c u. a 2 < m -sC'^otnif ''i^rnf sj* ir- if. 'O'fin^irinif f i n t g? o ^ ' C u* 1/ c-Cs '$,'tf'rg*4-(\i^>j*sj'»Hfv,rHroin ’^psiif\'i,*4,'Cmf«s!,'«^m'Ofn*HO'0^,fnfncocD'OHm^«4*^ • • • 9 9 9 9 * a « 9 o o 9 e o o 9 o o o o o 9 * 0 g 9 o o o o 9 »ffr'©(>>f’^QDa,'j,*4ivO'fla,HU'M^vOfsiHNH^.Nir. u. o e o g o o o o a o o e t i o e e o o o o o o o o o o o a g ^ o e o ir-pp iT''flCr-'OMr^rJCMrttHH^Nv},’«'^(7'n{CCrtfiN«£inOO coooootnooinoo 1 >0'<0O'0'0 * 0 s 0 v 0 ' © g D h - P “ * f , “ , P ‘ , ' f ’ , “P i - P ‘ - Pf c ' P , * f > ' * C D C 0 C 0 c p a 3 c 0 C D rlHrlHHrlHHHHNMNrJNPJNrjNNmn ^ottinooNor. cm •••. Mr f fH fH CM rH cm cm ••••o9«o cm cm cm cm cm (T'vOP-rH 9 9 9 9 cm 2 r-i r-m 3 co cm 1 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced PITS TGN=Sp TLT C* . U O t l l J a D C w I o ' u j O w OU; O U. U. £fi .Or. O a o *- H S O J . O N 5 J C J . U a 2 □ cc h > - - > Ccr 'Z. < > o .oa a. o s: a : s - u cr. o o < < JiC ►" X 2 0 . X u ts> z ' r s j r ^ n , ’ ’ , n ^ r j s r CC CP CC O* O . T l ^ n C H T T ( \or-cDC>oHrjfr o o o c ^ o - n i ^ n f j r H O ^ : a * * P O ' r i ^ r' f j r H o > C D c - r o \ . fc* f r O \ 0 v in ^ c n f N m 1 r H M O , H O C O c ^ r i 'O ir t i^ t r N h H 'O O C M cocDcccEcoci'CDCDaDcccoQr'CGcrcr'cococDcocDcocrcocDcocD©©©©©©^© 'O in ^ « ( N f * 0 O «ON®O‘ n' 'p*-“f“r--‘hcc 0 •o • • • • « hj 0 0 ix v xi ri i r « 'C^n' 0 ' 0 ' ' ' ' i ' U ' v w 0 0 h h n n n h h h , ' f > w » - ^ ' £ ) h - » n ' O c c f * ~ i r r- r - f © o »oo ' 94999 0 cD^h-hc>NNO' ''Cu o CO o o O O o in '1 p . i r n v C t r O IT. O O h n h 9944 H *H r* H > r n r t> <>00 r < ' r S o o « _ N h 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced PS SI GN= Sp ILTER “ O^- 5“ N h Ot^V-«-* M H N 1 2 n o z o Cu H i vn > INJ X • • H z CL a _• u_ u. CJ INI 2 3 in CO in H? H* 0^ o o © rH H CO O o Os o rH m in IT O o in o o rg eg eg rg rg rg cn JN?DJ rH ■o o* o CO O o t**- - r - r © o o o o o o ' • I r r j hr - O O • • n • • • • • • ■ OJ tn H rH in Mp- •M H O O M N n rg p- o O oto to p» r^ • . ■o o on y r O ID © D» O • • • • « • • • H rH rH rH P- <0 in o - r © o o © c • n©o © rn cn o O h~ O- o o o ■0 '• « • • • • . • • • fvj r-4 SJ1 H ©*■ rH H o O H5 © CO CO o n u\ en © m id o o - r «n rH *H r> • • • « • « • • • t • • ng> in rH Os rH rH m rH rH in in rH O o n o o O co m m p“ m 0 o Cl • • • • • • rH >9 rg cn gf © O o © CO CO CO rH H5 © rH r- in CO o H H3 •O o r*"» h* n ID © vT 9 » • « • 123 124 i/i I or uj Q W OOOOOOOOOOOOOOOOOOOOOi-tOfnfnfUOOOOOOfvirH T O « OOOOOOOHOOOOHMHOOOOOHHH^rnNHHNOOHfl'fl 00000 eo ooo ooo eeoooooooootoooeooooo fM O H- l/> \Sl ooooooaoooooeooooooooooorooooooooo © o fM c> 'O OJ N-OOlM(\JCD>tO ninoh-o O * O ' f CO P - H s fl fO CT* rl O-'OOIA IA 'S’ fM fA o fsj <• m >90900909 I 9 9 • • 0 1 0 0 0 0 9 c-4 lr. lA U\ m m h* o n O'OnHfTlH'OOCOhOO 'C'»r0'0O'j‘nNrjH£>‘&"0Nr»oh. v i-1 'ey v.1 x »'i ix i 4 n ir w" xj ix r— u ON'ON uj in mj ininnC'COhOH^r-'Oir 90990099909009090909 >000909999099 •0 r, innroi'JMN'flicir. o r n c M T h in p- >o mrjoN'cn'fl-^rifnot? I. -J O * «-» HHHHHHHHHrlrirlrtHHH H H H H H H rt r1 -J -* -* * - -* fiNMriMrlfnMOMNHHrlHHlTOHrlOCDO'CDCOIDNC^-©©ooooooom m oooooooooooooooooooo IAo fOin «*?o 9009999990999900900009090090909900 ®CDCCXCDOOa5eD ► *nC'0'0^^^ffiwcDH^a,'?oninHh-ccC'0''0h0''0c0hh m o® r‘i^fjm fM s0r' Ccrr^OMOOCOOOOO"P*ONP-i-0'ffP'fT''fOvflNff>Offl in ® ® Of^h-»-«O0OOOO0OO®®OOrHO Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced DPSTGN=FB/Sr . <>u * -r-r r * - OvOHhN N h £ H O v j 'O O • f** h-r*- r*- r- > Lr- u. u f- o u c _ - l uL U C C0900000000999999090090009000990000 ^ a j u r i S d e z z nee f Z C ' •K < Z U - 99900000000000 9000990009000 9 * 090909 u. D uJ H X O J L USC r^fJHJ*^'H2H'<\jfMr-i'^N>yj*mrn>4‘WU'SIC. '4*fr'rv4rOMrJrofMrs -riH o omof o OOHON t ov»m( Nn nN (M 0 ?r;sO'OOif 'sJ-rommh-Mo-aoomo r ntmtNO>ONC' o* *o 'O C N O > O trN trm tn trm m n ^ ^ o o - o o ^ c n CP N * t I f , O e M ( »fH ^ m i f v ooooooooo * h » ^ h r4 r'w tr o 3 4 - vO m ai 'O' 0 «fi 125 126 Jm CC Hi D 1/1 HQQHMHQ^OHO^N^HNHNQOONtn^HOQOOQOOOH»H O o rH OITUMf'U>HfSJOl'^0‘UMAOOsOO IT tn in N N v O H ,^P*Nh-l^03lf»inCOsOOCDP*(,0’«r X a O' V— fH (Nj h NHHNriNHHHH ©CM rH © o 0 0 0 0 < -j ^ CM O* + O' CO cn H fH e 0 0 0 h— tn t/> tn rH o o o o © 2 o o • 0 0 0 o CM u2 q a 0 0 0 0 0 0 0 0 0 9 0 0 0 0 0 0 0 h» rH O' CM o* © ' D O ' O ' o © o o o © CM CM p - © o H? © '4- vO CM •4* o O ' CM © rH © © p - i n m o i n i © r i © CM ^ n O 3C cm • t 0 # 0 0 0 0 0 0 0 0 # 0 0 0 0 0 0 0 0 0 © m © © CM © CM © © © CM CM p^ h - M3 M3 CM M CM © CM CM >o v3- rH © r - fH © *0* V? © O c m o CM CC vO l CC o U"* O' IT* f> © tn CM f J i n o u*. in CM © M3 CM H? © © «r m © I f n. .J CJ JS fH • 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 O O o rH O CM CM rH r u © © CM © CM CM CM © If. in >0 in © © © ft rH rH rH rH rH © rH © © © © © © © 1 © © © © © ja a k- * 2 r* - > O 3 ru a t r ft t a uj 2 2 u. u. ■rft o o o o o ,,,lOOO0J<0JftJOftJftJft»ftJfft4tft©*n»* o ' o O O o J " a o * i - o n * i - o ' * / o J i t ' o f / J o i - H * o H O O Q o H O o t O r C o © ^ O N o O H O f o t o r M o d i o M j o n o ^ t o r j o t o f M o N ' o / o ' i o ' o o o o o o o o Q O f*.®«o*o(j>a(?'00© 0 P “ O ' H O * O ' f O C D > 0 0 3 T l O - C j ' * - r ~ r23 “p * ;3 ' , o ,Uo “P sc - rDir , “f x * ro * " iooooooooocotnaocinmoooooooooo r ‘ Oofw'£' r f c ' O r , , - f , * - f f c - f s * r ^ « o < 3 v r r * ”i * " ' l u » i r i ' £ j ^ ' C N 5 “' o r - O0«O099 O»9«9OO9OOO9OOO0O9OBOO999O9 o o o o o o n i n i o o o o o o o o o o o o m n i o c n i o . r i o o o o n i n i D O 9 * 9 ' ’ O 9 Q O O e e Q O O Q O e O « 0 O o O O O O B O « « O O O ••••(••••• • • • • • 9 9 g 0 « . 0 9 9 . 9 9 » 9 9 9 9 9 « 0 « 0 * 9 0 9 * 0 9 0 9 0 9 9 9 9 9 « ' 9 0 0 9 9 9 • ••099900999• 0 9 0 9 9 9 0 0 0 9 9 9 9 0 » 9 g 9 9 0 0 0 9 9 0 9 9 9 9 9 0 0 0 • v cMfMfMfMcMfMrMfncnfMfM^yy^ninfncMco j ' n o f n n n n n H ^ i p n r