Paller 1982 Aquacultural-Engineering PDF

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Aquacultural Engineering 1 (1982) 139-151

R E C I P R O C A T I N G B I O F I L T E R F O R W A T E R R E U S E IN
AQUACULTURE

M. H. PALLER and W. M. LEWIS

Fisheries Research Laboratory and Department of Zoology, Southern Illinois University,


Carbondale, Illinois 62901, USA

ABSTRACT

The reciprocating biofilter is automatically dewatered at regular and frequent intervals,


in contrast to the conventional upflow submerged biofilter which is continually in-
undated. Reciprocating biofilters were compared with submerged biofilters in terms of
ability to maintain water quality in small-scale fish homing units. In the first trial the
reciprocating filter systems averaged 35 % more fish in terms o f numbers, 59 % more
fish in terms of weight, and a 45 % greater feeding rate. In the second trial the recipro-
cating filter systems averaged 29 % more fish in terms of numbers, 33 % more fish in
terms of weight, and a 29 % greater feeding rate. Superior performance of the
reciprocating filters appeared to be the result o f resistance to clogging and improved
aeration o f the filter substrate.

INTRODUCTION

The submerged type of biofilter is commonly used to maintain water quality in closed
fish culture systems (Burrows and Combs, 1968; Meade, 1974), despite the fact that it
often suffers from two problems (Lewis et al., 1981a). One problem is that the sub-
merged design is rather inefficient at supplying oxygen to the aerobic triter bacteria.
The second problem is that the interstices of the filter substrate become obstructed by
particulate matter and bacterial growth. Prior work in our laboratory (Lewis et al.,
1978) suggested that these problems could be reduced by the use of a biofilter with a
reciprocating pattern of water flow (i.e. alternate flooding and draining of the bio-
filter). To test this hypothesis, we compared biofilters with a reciprocating flow
pattern (hereafter referred to as 'reciprocating biofilters') and conventional submerged
biofilters with an upflow pattern, in terms of ability, to maintain water quality in
small-scale fish holding units.
139
Aquaeultural Engineering 0144-8609/82/0001-0139/$02.75 © Applied Science Publishers Ltd,
England, 1982
Printed in Great Britain
140 M.H. PALLER, W. M. LEWIS

METHODS AND MATERIALS

This study consisted of two trials. Three reciprocating biofilter systems were compared
with three submerged biofilter systems in each trial. All systems were 100% recycle,
with the exception o f the water exchange that resulted from backflushing the
biofilters. Water flow rates and biofilter dimensions were varied between trials.
The procedure for each of the trials was to regularly monitor water quality in each
system, while gradually increasing the stocking and feeding rates until the carrying
capacity of each system was determined. The systems were stocked with laboratory-
acclimated goldfish fed Silver Cup number 2 trout food. The systems were completely
drained between the trials, which were separated by a little under four months.
Shallow biofflters were used during the first trial (Fig. 1). Each gravel bed was
32 cm wide, 43 cm long, approximately 8 cm deep and contained 1%5 kg of gravel.
The gravel particles averaged 9.7 mm in diameter. Water from a fish tank entered each
biofilter through a manifold located below the gravel bed. An empty space below the
manifold permitted the particulate matter to settle. The water then flowed upward

• .. ..o. //

oF I GF
J J
Fig. 1. Design of the reciprocating and submerged biof'flter systems used in trial 1. Components
common to both systems include: fish tank (A), air stone (B), standpipe (C), valve (D), biof'dter
(E), gravel bed (F), manifold (G), settling basin (H), biof'flter standpipe (I), and pump (J). The
reciprocating systems also ineludes a perforated pipe housing a float switch and float (K).
RECIPROCATING BIOFILTER FOR WATER REUSE 141

through the gravel and subsequently drained to a pump. The water was pumped up to
a 75 litre fish tank, where it was vigorously aerated with compressed air. Water from
the fish tank flowed down to the biofilter at the rate of 8.5 litre min -1.
The pattern of flow in the reciprocating filters was different from that in the sub-
merged filters. Water flowed into and was pumped out of each submerged biofilter at
equal and constant rates, thus keeping each submerged filter bed constantly inundated.
Water flowed into each reciprocating biofilter at a constant rate, but was periodically
pumped out at a much higher rate, thus dewatering each reciprocating triter bed at
regular intervals. A float switch deactivated the pump when the filter bed was fully
drained, and activated the pump when the water rose to a level several centimetres
above the gravel. Each cycle of filling and draining took approximately 2 rain to com-
plete. The total water volume in each submerged falter system (i.e. filter and fish tank)
was approximately 100 litres, and approximately 93 litres in each reciprocating filter
system.
Deep biofilters were used in the second trial (Fig. 2). Each gravel bed was cylindrical,
49 cm deep, 15.35 cm in diameter, and contained 15.9 kg of gravel. Water flowed into

;0
E /I

i) l
Fig. 2. Design of the reciprocating and submerged biofilter systems used in trial 2. Components
common to both systems include: fish tank (A), air stone (B), standpipe (C), valve (D), biofilter
(E), gravel bed (F), settling basin (G), biofilter outlet (H), and pump (I). The reciprocating system
also includes a separate container attached to the biofilter which houses a float switch and
float (J).
142 M.H. PALLER, W. M. LEWIS

each biofilter bed at a relatively low r a t e - 2.5 litre min -1. A single reciprocating
cycle took 3-4 min to complete. Total water volume in each filter system was
approximately 87 litres. The other features of the deep bed biofilters were similar to
those of the shallow bed biofilters, with one exception: the float and float switch of
each deep bed reciprocating biofilter were contained in a separate housing which was
connected to the container enclosing the filter bed. This connection permitted the
water levels to equalize within the paired containers.
It was necessary to backflush the deep bed biofilters to maintain unimpeded flow.
Backflushing was accomplished by connecting an external water source to the bottom
of the triter housing and directing a relatively powerful water stream upward through
the gravel. This procedure dislodged accumulated detritus, which flowed up and over
the edge of the filter housing. The gravel was stirred with a rod during backflushing to
help dislodge particulate matter.
The chemical and physical parameters measured in the study are listed below.
(1) Dissolved oxygen concentration (DOC) (rag litre -1) was measured with a YSI
polarograph. Measurements were made in the fish tanks during the first trial,
and in the fish tanks and biofilter effluents during the second trial. Filter
effluent was collected from a sampling port located at the base of each deep
bed reciprocating biofilter column to avoid contamination by the unfiltered
water in the associated container housing the float and sump pump switch.
This permitted comparison with filter effluent from the submerged biofflters
which was collected from the outflow into the fish tank.
(2) Total ammonia-nitrogen (mg litre -1) was measured with an Orion ammonia
electrode 95-100 and Orion 901 Microprocessor Ionalyzer (USEPA, 1974).
Samples were taken from the tanks and from the biofilter effluents during both
trials. Filter effluent samples were drawn from sample ports in the deep bed
reciprocating biof'flter columns.
(3) Nitrite-nitrogen (mglitre -1) was measured with a Hach spectrophotometer.
Samples were taken from the fish tanks.
(4) Nitrate-nitrogen (rag litre -1) was measured with a Hach spectrophotometer.
Samples were taken from the fish tanks.
(5) Temperature (°C) was measured in the fish tanks.
(6) pH Was measured with a Coming Model 12 pH meter and electrodes. Samples
were taken from the fish tanks.
(7) Alkalinity (acid binding capacity) was determined according to Standard
Methods procedure 4C (APHA, 1975). The end point was pH 4.4. Samples
were taken from the fish tanks.

RESULTS

Ammonia-nitrogen levels remained relatively low in each system until a critical loading
rate was exceeded but then rapidly increased to 10-40mglitre -1 (Figs 3 and 4).
RECIPROCATING BIOFILTER FOR WATER REUSE 143

Number of fish 0 2 7 12 17 22 22 22 27 32 3742


Food (g) 0 :15 6 8.5 11 17.5 20 23 26 29 29

--~"
•~. 5.0 t II
It I
.EE II I
c I
.0 40 I ~ :1
I 1

I
/ I

t t Ii
2.0 /~ l I:
._ d':: i ill
z : 'I :j,
V
10 I !1

< oo --- i l ::i


0 10 20 30 40 50 60 70

Days
Fig. 3. Changes in total a m m o n i a - n i t r o g e n concentration in the fish tanks of the reciprocating
biofilter systems (e), and in the fish tanks of the submerged biofilter systems (o), versus time and
changes in loading rate (trial 1). Each data point is an average derived from three systems. Changes
in loading rate are indicated by broken lines.

Number of fish 0 2 5 15 2.5 30 3.5 40 50 45


F o o d (g~ 0 I 2.5 Z5 12.5 15 1.75 20 25 22.5

~" so

e-
._0 4.o

213

e !
e- /
, ,

< o.o
0 10 20 30 40 50 60 70 80
Days
Fig. 4. Changes in t o t a l a m m o n i a - n i t r o g e n concentrate, o n in the fish t a n k s o f t h e r e c i p r o c a t i n g
b i o f i l t e r systems (o), and in the fish t a n k s o f t h e submerged b i o f i l t e r systems (o), versus t i m e and
changes in loading rate (trial 2). Each data point is an average derived f r o m three systems. Changes
in loading rate are indicated by broken lines.
144 M.H. PALLER, W. M. LEWIS

Therefore, the carrying capacity of each system is defined as the maximum stocking
density and feeding rate attained before the development of a large increase in
ammonia-nitrogen concentration. This definition could not be applied to submerged
filter system 1 o f trial 2, where severe clogging o f the gravel bed prevented normal
water flow despite repeated backflushing. In this case, carrying capacity is defined as
the maximum stocking density and feeding rate attained before the occurrence of
severe irremedial filter clogging. The carrying capacity of reciprocating filter system 2
was not determined during trial 1 because of a mechanical failure which forced its
premature shutdown.
To prevent their loss, the fish were usually removed from each system soon after
the development o f the ammonia-nitrogen increase. However, this was not done in the
three reciprocating filter systems during trial 2, where stocking densities and feeding
rates were instead decreased to carrying capacity levels soon after the ammonia-
nitrogen increase began to develop. The return of ammonia-nitrogen concentrations
to low levels following this action (Fig. 4) confirmed the carrying capacity assessment
in these systems.
Carrying capacities in the reciprocating filter systems were higher than in the sub-
merged filter systems (Table 1). In trial 1, the reciprocating filter systems averaged
55 % more fish in terms of numbers, 59 % more fish in terms o f weight and a 45 %
greater feeding rate. In trial 2, the reciprocating filter systems averaged 29 % more fish
in terms o f numbers, 33 % more fish in terms of weight and a 29 % greater feeding rate.

TABLE 1
Relative capability of reciprocating and submerged biof'flters in terms of carrying capacitya

Reciprocating Submerged
filter systems filter systems

1 2 3 1 2 3

Trial I b
Number of fish 37 c 52 27 27 32
Cumulative weight of fish (g) 774 _c 1099 559 533 675
Quantity of food per day (g) 26 _c 35 20 20 23
Trial 2 d
Number of fish 45 45 45 35 35 35
Cumulative weight of fish (g) 1006 1049 995 754 769 767
Quantity of food per day (g) 22.5 22.5 22.5 17.5 17.5 17,5
a . .

b Higher loading rates caused ammoma concentrations to increase to potentially toxic levels.
Shallow biofilters employed.
e Mechanical failure caused shutdown before the carrying capacity could be determined; 37 fish
~(eighing 737 g fed 26 g at time of shutdown.
Deep biofilters employed,
RECIPROCATING BIOFILTER FOR WATER REUSE 145

The average ammonia-nitrogen values in the reciprocating filter system fish tanks
after the reciprocating filters became activated, but before the filter carrying capacities
were exceeded, were 0.28mglitre -1 during trial 1 and 0-58mglitre-' during trial 2;
corresponding values for the submerged filter systems were 0.25 and 0.07 mg litre -1,
respectively. During trial 2, ammonia-nitrogen concentrations in the reciprocating
filter systems are initially somewhat higher than in the submerged f~ter systems
because of occasional clogging of the gravel interstices in the reciprocating filter
columns, which caused water to be channeled into the columns housing the float
switches. This greatly reduced the flow rates through the reciprocating biofilter heds
and resulted in ammonia-nitrogen peaks at days 38 and 55 (Fig. 4). This problem is
not inherent in the design and operation of reciprocating filter systems and resulted
in a conservative estimate of the relative performance of the reciprocating filters
during trial 2.
Dissolved oxygen concentrations in the fish tanks of the reciprocating filter systems
were higher than in the fish tanks of the submerged filter systems (Figs 5 and 6). This
difference became progressively greater as stocking densities and feeding rates became
greater. Additionally, dissolved oxygen measurements made in the biofilter effluents
during trial 2 indicated that oxygen levels within the reciprocating biofilters remained
above 3-0 mg litre -1, but fell to approximately 0-5 mg litre -1 within the submerged
biofilters (Fig. 7).
Average nitrite-nitrogen concentrations were similar in the reciprocating and sub-
submerged filter systems during trial 1 (1.41 and 1.22 mglitre -1, respectively).
Nitrite-nitrogen concentrations peaked in both the reciprocating and submerged filter
systems after the first few fish were stocked, but peaked higher and later in the
reciprocating filter systems (Fig. 8). Nitrite-nitrogen concentrations during trial 2

N u m b e r o f fish 0 7 12 17 22 22 22 27 32 37 42
F o o d (g) 0 a5 6 ~.5 11 lz5 20 23 26 29 29

- i ! ! ! i i!

~ 6.0 -.-e

~C : : i i : : : i i i:
1.2 OO 0 }0 20 30 40 50 60 70

Days
Fig. 5. Changes in dissolved oxygen concentration in the fish tanks o f the reciprocating bioffiter
systems (e), and in the fish tanks o f the submerged biofJlter systems (o), versus time and changes in
loading rate (trial 1). Each data point is an average derived f r o m three systems. Changes in loading
rate are indicated by broken lines.
146 M.H. PALLER, W. M. LEWIS

Number of fish 0 2 5 15 25 3 35 40 50 45
Food (g) 0 1 2.5 7.5 12.5 15 1z5 20 25 22.5

9 i ~
/._-8 o--L ~ : - - :~

o 8 5 ~ ' ~ - " " - - )'

g82
1

0
10 20 30 40 50 60 70 80

Days
Fig. 6. Changes in dissolved oxygen concentration in the fish tanks of the reciprocating biofilter
systems (o), and in the fish tanks of the submerged biofilter systems (o), versus time and changes
in loading rate (trial 2). Each data point is an average derived from three systems. Changes in
loading rate are indicated by broken lines.

N u m b e r o f fish 0 2 5 ]5 25 30 35 40 5o 45
Food(g / 0 1 ; 5 .75 125 15 lz5 20 25 22.5

~. 8 Q. : :

o_
= 5

-~
a3~2
0 i i
0 10 20 30 40 50 60 70 80

Days
Fig. 7. Changes in dissolved oxygen concentration in the biofilter effluents of the reciprocating
biofilter systems (o), and in the biofilter effluents of the submerged biofilter systems (o), versus
time and changes in loading rate (trial 2). Each data point is an average derived from three systems.
Changes in loading rate are indicated by broken lines.
RECIPROCATING BIOFILTER FOR WATER REUSE 147

Number of fish 0 2 7 12 17 22 22 22 27 32 37 42
Food {,g) o ~ ~.5 6 8.~ , 17.52o23 26 ~929
B~lo ! :: :: i

8 8 i! i ~i
7 /e\\ :: i !

',a / "~i : : : : : '

O i ! : : i :
t-- . ; \ : : : : :
3 : i k • i : : '
: \ : : :
2 : i tk : : : ! :
z ~: %.. ; i~ i

~__ 10 20 30 40 50 60 70

Days
Fig. 8. Changes in nitrite-nitrogen concentration in the fish tanks of the reciprocating biofilter
systems (o), and in the fish tanks of the submerged biofilter systems (o), versus time and changes in
loading rate (trial 1). Each data point is an average derived from three systems. Changes in loading
rate are indicated by broken lines.

were uniformly low in both the reciprocating (mean = 0.34mglitre -1, range =
0.01-1.00) and submerged (mean = 0-18 mg litre -1, range = 0-03-0.85) filter systems.
Nitrate-nitrogen concentrations showed the same pattern of change in all recipro-
cating filter systems during both trials. Concentrations were low, then climbed steadily
to approximately 15 mg litre -1 and subsequently declined to approximately 3-6 mg
litre -1. The decline was probably due to the occurrence of denitrification in the
anoxic settling basins located beneath the filters. Nitrate increases also occurred in the
submerged filter systems but were not followed by conspicuous decreases as in the
reciprocating filter system.
Temperatures in the submerged and reciprocating filter systems were similar within
trials. Mean temperatures during trial 1 were 22.7°C (21.5-23.5°C) in the recipro-
cating filter system, and 23.7°C (22.0-25.0°C) in the submerged filter systems. Mean
temperatures during trial 2 were 25.8°C (23.5-31.0°C) in the reciprocating systems
and 26.1°C (23.5-29.0°C) in the submerged filter systems.
Alkalinity changes followed the same pattern in all systems during both trials.
Alkalinity averaged 45 mg litre -~ in the reciprocating filter systems at the beginning of
trial 1, 57 mg litre -1 in the submerged triter systems at the beginning of trial 1,86 mg
litre -1 in the reciprocating filter systems at the beginning of trial 2 and 64 mg
litre -1 in the submerged filter systems at the beginning of trial 2. As the trials
continued, alkalinity levels progressively declined to values as low as 16 mg litre -1.
When low levels were reached, alkalinity was restored by the addition of 5-50 g of
sodium bicarbonate to each system. Alkalinity was restored three times in trial 1 and
148 M. H. P A L L E R , W. M. L E W I S

twice in trial 2. Alkalinity levels rose as high as 310 mg litre-' after restoration with
sodium bicarbonate.
Fluctuations in alkalinity were associated with fluctuations in pH. pH Varied
between 7.2 and 8.4 in the reciprocating filter systems during trial 1, between 6.7
and 8.1 in the submerged filter systems during trial 1, between 6.7 and 8-2 in the
reciprocating filter systems during trial 2 and between 6.7 and 7-7 in the submerged
filter systems during trial 2. pH Usually remained above 7.2 in all systems. Fluctua-
tions in pH and alkalinity were not associated with other changes in water quality or
with changes in fish behavior.
None of the shallow biofilters employed during trial 1 were backflushed, since
water flow through the filters appeared to be unimpeded. However, the much deeper
columnar filters employed during trial 2 often became blocked and required repeated
backflushing. The submerged columnar biofilters required more backflushing than the
reciprocating columnar biofdters. Forty-three backflushing operations were distributed
among the three submerged biofxlters. Clogging was so severe in submerged filter 1
that it became impossible to maintain the water flow, despite repeated backflushings
at 8 h intervals. In contrast, only 24 backflushing operations were necessary among the
three reciprocating filters. Since the reciprocating filters operated longer than the
submerged filters, this amounts to 0.29 backflushings per day among the reciprocating
filter systems, and 0.63 backflushings per day among the submerged filter systems.
Microbial growth appeared to differ between the submerged and reciprocating
filters. This was most noticeable when viewing the tops of the gravel beds in the
shallow biofilters. Microbial growth appeared to form a rather closely adhering layer
around individual gravel particles in the reciprocating filter beds, but formed loose
associations collectively covering a number of gravel particles in the submerged filter
beds.

DISCUSSION

Lack of oxygen can limit the performance of aerobic biofilters (Semmens, 1976).
Recognition of this fact has led to the development of various methods for supplying
oxygen to biof'dters. Two successful methods are the addition of air or oxygen to the
biofilter influent (Meade, 1974; Lewis et aI., 1981b), and the use of revolving circular
plates which cyclically expose the attached microbial growth to the filter influent and
to the atmosphere (Lewis and Buynak, 1976). The reciprocating biof'llter provides
oxygen to the filter bacteria by periodically dewatering the filter bed. This process
creates a vacuum which draws air into the gravel interstices and oxygenates the thin film
of water surrounding the gravel particles. Oxygenation occurs uniformly throughout the
filter bed. As a result, oxygen levels in reciprocating f'dter effluents tend to be higher
than in submerged filter effluents.
The oxygenating action of the reciprocating filter is particularly important at high
loading rates. Under such conditions, extensive oxygen consumption by fish and
RECIPROCATING BIOFILTER FOR WATER REUSE 149

bacteria in the fish tank results in low oxygen levels in the biofilter influent. Con-
currently, oxygen demand within the biofilter increases because of the large quantities
of organic matter and ammonia entering the filter. As oxygen concentrations approach
limiting levels, the efficiency of the f'flter bed declines and the quality of the filter
effluent deteriorates. Therefore, an even greater load is imposed upon the filter the
next time the water is recycled. The result is a vicious circle causing the collapse of
filter performance and a resulting rapid increase in ammonia concentration. This
collapse is delayed by the aerating action of the reciprocating filter, which helps
replenish the oxygen consumed by the filter bacteria. Furthermore, as oxygen levels
become progressively undersaturated, aeration becomes more efficient due to faster
transport across the air-water interfaces in the dewatered filter bed. As a result of
these factors, the aerating effects of the reciprocating filter are not fully manifested
until high filter loading rates are attained.
The clogging of biofilters with microbial growth and particulate organic matter is
another factor that typically limits biofilter performance. Clogging often causes
channelization of the water through localized areas of the filter bed and can
completely obstruct the water flow. Clogging is typically controlled by occasionally
reversing the water flow and backflushing the filter bed. Clogging is apparently
reduced in the reciprocating filters by the two-way water flow, which exerts a con-
tinuous low level backflushing action and reduces the occurrence of channelization. It
is also possible that the reciprocating pattern of water flow leads to forms of microbial
growth and organic matter accretion which minimize clogging and channelization.
These factors are probably responsible for the reduced backflushing requirements in
the columnar reciprocating filters.
Unlike the columnar biofilters, the shallow biofilters did not appear to require
backflushing. Reduced occlusion in the shallow biofilters was probably due to the
distribution of incoming water and suspended solids over a relatively large area. How-
ever, we believe that the deep columnar biofilters more accurately simulate actual
conditions in full-scale biofflters. The superior performance of the reciprocating
filters under both types of conditions suggests that the reciprocating mode of
operation may be widely applicable to many types of fish culture bioffltration
systems.
In comparing the results of the two trials, it is noteworthy that the superiority of
the reciprocating filters was more evident during trial 1, when neither the reciprocating
nor the submerged filters were backflushed. The pronounced superiority of the
reciprocating filters during trial 1 may have been the result of channelization in the
submerged filters, which resulted in under-utilization of some of the filter substrate.
Conversely, the fact that the submerged biofilters were backflushing more than twice
as often as the reciprocating biofilters during trial 2 may have resulted in the
imposition of a heavier organic load upon the reciprocating biofdters, since back-
flushing removes large amounts of organic matter. Therefore, in trial 2 the
reciprocating systems might have shown greater superiority if they had been back-
flushed as often as the submerged systems.
150 M.H. PALLER, W.M. LEWIS

Another factor which can influence the performance of the reciprocating filter is
the duration of the reciprocating cycle, i.e. the time needed to fill the empty filter and
then completely drain it again. Prior experience in our laboratory suggests this is a
flexible design parameter and reciprocating cycles lasting 1 h have been successfully
employed. However, the dissolved oxygen content and oxygen demand of the filter
influent ultimately impose limitations on the length of the reciprocating cycle, since
reciprocation should be frequent enough to maintain filter oxygen levels above
3 mg litre -1.
Reciprocating filters may be inferior to submerged filters in terms of the amount of
time required for activation. This is suggested by the result of trial 1, which show that
ammonia-nitrogen concentrations initially rose to fairly high levels in the reciprocating
filter systems, but remained low in the submerged filter systems. Additionally, nitrite-
nitrogen concentrations showed early periods of elevation in both reciprocating and
submerged filter systems, but the period of elevation in the reciprocating filter systems
occurred later and lasted longer. Similar increases in ammonia and/or nitrite-nitrogen
concentration were not observed in any system during trial 2, probably because the
filter substrate used in trial 1 was also used in trial 2. Although the substrate was
washed and drained between trials, it is probable that living bacteria in an inactive
form remained in the gravel. If the inoculum was heavy enough preactivation would
occur and an initial breaking-in period would not be observed.
Reciprocating filter systems may have a potential engineering advantage in terms of
insuring the even distribution of water to all parts of the filter bed. Water distribution
systems such as manifolds and revolving armatures are a major expense in the con-
struction of most biofilters. In contrast, the water is automatically distributed to all
parts of the reciprocating biofilter as it enters the dewatered filter bed. This feature
might permit the construction of relatively simple, economical, yet functional
biofi!ters.
There are various other engineering aspects of reciprocating biofilters which were
not investigated during this study. A mechanical float valve can be used to flush the
reciprocating biofilters (Lewis et al., 1981a). This permits the elimination of the
electronic float switch, and, more importantly, permits the utilization of a lower
capacity pump. It is also possible to obtain a continual discharge from the biofilters,
rather than the intermittent discharge characteristic of the reciprocating filter systems
used in this study. This is accomplished by situating one reciprocating biof'ilter above
another (Lewis et al., 1981a). Water is continually pumped to the upper filter. When
the filter is filled, a float valve rapidly drains it though a large diameter pipe into the
lower filter. The lower filter is designed to drain more slowly so that it approaches
emptiness as the upper filter enters its next flushing cycle.
This study indicates that the desirable attributes of the reciprocating biofdter
include: (1) ability to maintain relatively high oxygen levels within the filter substrate,
(2) resistance to clogging, and (3) automatic distribution of the water to all parts of
the filter bed. The reciprocating filter performs these functions simply, easily and
automatically by efficient use of the energy of flowing water. Other methods of
RECIPROCATING BIOFILTER FOR WATER REUSE 151

accomplishillg these tasks usually require additional energy expenditures and/or more
elaborate and expensive designs. Therefore, the reciprocating design may be most
advantageous in simple, economical systems, and least advantageous in highly
engineered systems incorporating such extras as bottled oxygen and elaborate filter
cleaning systems, although further research with large-scale systems is needed to clarify
these points. It is also noteworthy that reciprocating filters may possess inherent
disadvantages in terms of the time required for activation. This could be important
in applications involving the intermittent use of biofiltration.

REFERENCES

APHA (1975). Standard Methods for the Examination of Water and Wastewater, 14th edn.
American Water Works Association, and Water Pollution Control Federation, American Public
Health Association, Washington, DC, USA.
Burrows, R. E. & Combs, B. D. (1968). Controlled environments for salmon propagation. Prog.
Fish Cult., 30 (3), 123-36.
Lewis, W. M. & Buynak, G. L. (1976). Evaluation of a revolving plate type biofilter for use in
recirculated fish production and holding units. Trans. Am. Fish. Soc, 105 (6), 704-8.
Lewis, W. M., Yopp, J. H., Schramm, H. L., Jr & Brandenburg, A. M. (1978). Use of hydroponics
to maintain quality of recirculated water in a fish culture system. Trans. Am. Fish. Soc., 107
(1), 92-9.
Lewis, W. M., Yopp, J. H,, Schramm, H. L., Jr & Brandenburg, A. M. (1981a). On the maintenance
of water quality for closed fish production systems by means of hydroponically grown
vegetable crops. In Aquaculture in Heated Effluents and Aquaculture Systems, ed. K. Tiews.
H. Heenemann GmbH and Company, Berlin, pp. 121-9.
Lewis, W. M., Heidinger, R. C. & Tetzlaff, B. L. (1981b). Tank Culture of Striped Bass, Illinois
Department of Conservation, Project F-26-R. Fisheries Research Laboratory, Southern Illinois
University, Carbondale, Illinois, USA.
Meade, T. L. (1974). The Technology of Closed System Culture of Salmonids, Marine Technical
Report 30, University of Rhode Island, Kingston, Rhode Island, USA.
Semmens, M. J. (1976). A Feasibility~Development Study for the Removal of Ammonia from
Wastewater using Biologically Regenerated Clinoptilolite, Illinois Water Resources Center
Research Report 115, USA.
USEPA (1974). Manual of Methods for Chemical Analysis of Water and Wastes, Methods Develop-
ment and Quality Assurance Research Laboratory, National Research Center, Cincinnati,
Ohio, USA.

You might also like