600J88010
TRICKLING FILTER/SOLIDS CONTACT PERFORMANCE
WITH ROCK FILTERS AT
HIGH ORGANIC LOADINGS
by
Raymond N. Matasci
David L. Clark
James A. Heidman
Denny S. Parker
Bruce Petrik
Darrel Richards
Presented at the 59th Annual Conference of the
Water Pollution Control Federation
October 9, 1986
Los Angeles, California
-------
-------
TRICKLING FILTER/SOLIDS CONTACT PERFORMANCE
WITH ROCK FILTERS AT
HIGH ORGANIC LOADINGS
by
Raymond N. Matasci
David L. Clark
James A. Heidman
Denny S. Parker
Bruce Petrik
Darrel Richards
The trickling filter process was formerly the most popular
method for treating municipal wastewater. In 1975, about 4,300
trickling filter plants were operating in the United States, which
was almost two times the number of activated sludge plants. The
popularity of the trickling filter process declined in the 1970s
and early 1980s because the existing plants were often exceeding
the newly established secondary treatment standard of 30 milligrams
per liter (mg/1) for total suspended solids (TSS) and 5-day bio-
chemical oxygen demand (BOD^).
Recent 4work on the trickling filter/solids contact (TF/SC)
process2'3' ' has demonstrated that trickling filters can reliably
achieve secondary treatment and advanced treatment standards with
relatively simple plant modifications. The popularity of the TF/SC
process has increased dramatically since its development in 1979
because it produces significantly better effluent than the trick-
ling filter process alone. The process is applicable to new plants
as well as existing trickling filter plants.
The U.S. Environmental Protection Agency (USEPA) estimated that
approximately 55 percent of the existing 2,700 trickling filter
plants will be upgraded or abandoned by the year 2000. Many of
these changes are necessary to treat increasing plant loads or
improve effluent quality. Substantial savings would be realized if
relatively simple modifications could enable existing trickling
filter plants to operate at higher loads with greater reliabi-
lity. The TF/SC process is an important topic of study because it
shows promise for producing just these results—significant
improvements in effluent and plant reliability, even at higher
organic loadings.
Study Objectives
Most of the operating trickling filter plants use rock media.
To test TF/SC performance with rock filters at high organic
-------
loadings, USEPA sponsored full-scale studies at the Morro Bay-
Cayucos TF/SC plant. The studies also included an assessment of
trickling filter performance with flocculator-clarifiers and
reaction rate coefficients for soluble carbonaceous BODr (SCBODc)
removal in rock trickling filters.
The studies included 9 weeks of field investigations at the
Morro Bay-Cayucos facility in Morro Bay, California. The field
investigations data were supplemented with operating records from
the Morro Bay-Cayucos plant and from plants in Coeur d'Alene,
Idaho; Corvallis, Oregon; and Oconto Falls, Wisconsin.
PLANT FACILITIES AND EXPERIMENTAL PROCEDURES
The investigators conducted full-scale studies at the Morro
Bay-Cayucos plant between March and June 1986. The plant .and the
experimental procedures are described below.
Plant Description
The Morro Bay-Cayucos plant is located on the Pacific coast in
central California. The original plant was constructed in 1954.
It was expanded in 1964 and again in 1983. The recent expansion
was designed to comply with the 1978 Water Quality Control Plan
for Ocean Waters of California. In 1985, the City of Morro Bay and
Cayucos Sanitary District obtained a waiver of full secondary
treatment in accordance with Section 301(h) of the 1977 Clean
Water Act.
The plant uses split treatment to meet its discharge require-
ments. Figure 1 shows the plant's flow schematic. Raw influent
undergoes preliminary treatment (screening and aerated grit
removal) and then primary sedimentation. A portion of the primary
effluent receives secondary treatment with the TF/SC process. The
remaining fraction of primary effluent is blended with secondary
effluent and chlorinated before discharge to the ocean outfall.
Table 1 lists -the plant's design data. The Mode III TF/SC
process, which includes return sludge aeration and aerated solids
contact provides secondary treatment. The two rock filters are
relatively shallow with a mean loading of 750 grams per cubic meter
day (g/irr-d) [47 pounds per day per 1,000 cubic feet (ppd/1,000 cu
ft)] at the secondary treatment design flow of 0.042 cubic meters
per second (nr/s) [0.97 million gallons per day (mgd)]. The return
sludge aeration and aerated solids contact times at design flow are
13 and 3.3 minutes, respectively.
-------
RETURN SLUDGE
AERATION
AERATED SOLIDS
CONTACT
PRELIMINARY
TREATMENT
RAW
INFLUENT
FLOCCULATOR-
CLARIFIER
PRIMARY
SEDIMENTATION
TANKS
OUTFALL
Figure 1. Plant Schematic
The secondary effluent TSS for the TF/SC process for the year
before the full-scale studies is shown on Figure 2. The TF/SC
process generally produced effluent with TSS between 5 and 10 rag/1
at the design flow of about 0.043 m /s (1.0 mgd). In August 1985,
the plant staff changed operation from split treatment to full
secondary treatment. The average secondary effluent TSS increased
to 19 mg/1 when the secondary flow increased 50 percent to
accommodate full secondary treatment.
Experimental Program
The plant has unusual flexibility because of its split
treatment feature. This flexibility allowed the study team to
conduct relatively closely controlled experiments on the full-scale
secondary treatment process.
The experimental program consisted of three phases. In
Phase 1, the secondary process operated in the TF/SC mode with both
rock filters in operation. In Phase 2, the study team stopped
secondary solids recirculation, and the secondary process operated
in the trickling filter mode. In Phase 3, the secondary process
operated in the TF/SC mode with only one filter to determine TF/SC
performance at high organic loadings.
-------
Table 1. Design Data
Parameter
Value
PLANT CAPACITY
Flow, n3/s (ngd)
Average dry-weather flow (ADWF)
Peak seasonal dry-weather flow (PSDWF)
Peak dry-weather flow (PDWF)
Peak wet-weather flow (PWWF)
Waste strength, iag/1
Biochemical oxygen demand (BODj)
Total suspended solids (TSS)
PRIMARY TREATMENT
Sedimentation tanks
Number
Diameter, m (ft)
Tank 1
Tank 2
Average side water depth, m tft)
Overflow rate at PSDWF, m3/mz-d (gpd/sq ft)
TRICKLING FILTERS (in partial secondary
treatment mode of operation)
Flow distribution at PSDWF, raVs (ngd)
Filter 1
Filter 2
Diameter, m (ft)
Filter 1
Filter 2
Average media height, m (ft)
Filter 1
Filter 2
Hydraulic loading rate, 1/m -s (gpm/sq ft)
Filter 1
Filter 2
BODj loading rate, g/m^-d (ppd/1,000 cu ft)
SOLIDS CONTACT CHANNEL
Channel length, m (ft)
Reaeration portion
Contact portion
Channel depth, m (ft)
Channel width, m (ft)
Reaeration time (based
Aerated solids contact
including 33 percent return), minutes
SECONDARY CLARIFIER
Diameter, m (ft)
Average sidewater depth, m (ft)
Overflow rate at PSDWF, mj/n -d (gpd/sq ft)
on 33 percent return), minutes
time (based on total flow
0.090
0.103
0.291
0.289
280
280
IS
12
2.7
29.7
0.017
0.025
18
21
1.4
1.5
0.23
0.26
750
7.6
7.6
1.2
1.2
13
3.3
17
4.6
16.6
(2.06)
(2.36)
(6.64)
(6.60)
(50)
(40)
(9)
(730)
(0.39)
(0.58)
(60)
(70)
(4.5)
(5.0)
(0.34)
(0.38)
(47)
(25)
(25)
(4)
(4)
(55)
(15)
(408)
The filter organic loading for the TF/SC process was increased
from 480 g/m3-d (30 ppd/1,000 cu ft) in Phase 1 to 960 g/m -d
(60 ppd/1,000 cu ft) in Phase 3. A study of 63 trickling filter
-------
plants in the northern United States showed that 86 percent of the
plants had organic loadings less the 960 g/m -d (60 ppd/1,000 cu
ft) and 62 percent had loadings less than 480 g/nr-d (30 ppd/
1,000 cu ft). Thus, the filter organic loadings tested at
the Morro Bay-Cayucos plant are above average to high values.
30
o>
•t
to
CO
H
H
UJ
-1
u.
u.
UJ
Q
o
o
HI
CO
20
10
SPLIT TREATMENT : SECONDARY FLOW
r 1.0 mgd
FULL SECONDARY TREATMENT:
SECONDARY FLOW = 1.5 mgd
MAR APR MAY JUN JUL AUQ SEP OCT NOV DEC JAN FEB
1985 1986
Figure 2. Secondary Effluent Suspended Solids at the Morro Bay-
Cayucos TF/SC Plant
-------
The filter organic loading rate during Phase 2 (trickling filter
mode) was 445 g/nr-d (28 ppd/ 1,000 cu ft). Phases 1, 2, and 3
lasted 4, 2-1/2, and 3-1/2 weeks, respectively. The study tean
allowed 1 to 2 weeks between phases before resuming sampling to
allow the secondary process to equilibrate.
Figure 3 shows a typical diurnal flow curve for plant inflow.
Plant inflow varies significantly within each hour because of the
intermittent operation of large constant speed pumping stations
nearby in the collection system.
The flow controller ^Figure 3) maintains a flow to the filters
of approximately 0.043 nr/s d mgd) at all times. To maintain the
desired filter wetting rate, the plant also incorporates recycle of
a fraction of unsettled trickling filter effluent back through the
filters. This operating approach kept the secondary clarifier
overflow rates, aerated solids contact times (Phases 1 and.3), and
filter wetting rates at consistent levels during the three phases
of the study.
The primary effluent samples were taken downstream of the flow
controller thus accounting for dilution with returned secondary
effluent during the low flow periods. BOD^ and TSS analyses of
selected samples (Figure 3) during a 24-hour period showed that
concentrations were slightly lower during periods of low inflow.
For the purposes of this paper, the term primary effluent refers to
the trickling filter influent, which may include the recycled
fraction of secondary effluent during periods of low inflow.
Sampling and Analytical Procedures
The study team used refrigerated automatic ISCO samplers to
collect 24-hour composited samples of primary effluent, unsettled
trickling filter effluent, and secondary effluent three days per
week. The composite sample collection began at 8:30 a.m. and was
completed the following morning. The treatment plant staff col-
lected grab samples of mixed liquor and return secondary sludge
during the days of composite sample collection. The plant staff
also collected a grab sample of secondary effluent near the peak
daily flow (10 a.m.) for TSS analysis. The study team and plant
staff used a Secchi disk to measure clarity of the secondary
effluent.
The field investigators packed 'all composite samples and
grab samples of mixed liquor and return secondary sludge in an
ice chest and shipped them by bus to the Brown and Caldwell
Pasadena laboratory for analysis. The laboratory staff analyzed
primary effluent and secondary effluent for TSS, BOD5, and
SCBOD5. They analyzed effluent from each trickling filter for TSS
and SCBODc and the mixed liquor and return secondary sludge samples
for TSS and volatile suspended solids (VSS).
-------
fi)
Saf
SECONDARY TREATMENT PLOW
8:30cm 9:30 10:30 11:30 12:30pm 1:30 2:30 3:30 4:30 8:30 8:30 7:30 8:30pm
APRIL 18, 1988
2OO
ce
a.'
180
100
SO
,TSS
BOO
PLANT INFLOW
SECONDARY TREATMENT FLOW
O
8:30pm 9:30 10:30 11:30 12:30«n 1:30 2:30 3:30 4:30 5:30 8:30 7:30 8:30«m
APRIL 18, 1988 APRIL 19, 1988
Figure 3. Diurnal Flow and Variation in Primary Effluent Quality
Analytical determinations were performed in accordance with the
16th edition of Standard Methods. The laboratory staff used a
Whatman 934AH glass fiber filter to distinguish between particulate
and "soluble" substances. They also used a Hach nitrification
inhibitor to measure the carbonaceous fraction of the soluble BOD5
samples.
-------
The study team determined the significance o£ differences
with the _t_-test at the 95-percent confidence level. They used
the modified Velz equation, to compute K^Q values and a specific
surface value of 44.3 mvm (13.5 ft /ft3) for the rock media.
Professor David Jenkins and associates at the University of
California at Berkeley performed microscopic examinations of unset-
tled trickling filter effluent and solids contact tank samples.
They characterized floe size and shape and identified and quan-
tified filamentous organisms.
STUDY RESULTS
Table 2 lists the results of the Morro Bay-Cayucos studies.
Raw waste concentrations of TSS and BODc were significantly higher
during Phase 3 than in Phases 1 and 2. The significant increase in
primary effluent BODc during Phase 3 seems to have been caused by a
higher soluble fraction. Primary effluent TSS did not vary signi-
ficantly between phases, yet the Phase 3 SCBOD^ was significantly
higher than in the other two phases.
Secondary Performance
The secondary effluent TSS significantly increased 15 mg/1
based on the composite samples and 10 mg/1 based on the grab
samples when the plant switched from the TF/SC mode to the trick-
ling filter mode. When the filter organic loading increased to
960 g/m3-d (60 ppd/1,000 cu ft) in the TF/SC mode, the mean
composite secondary effluent TSS value was 2 mg/1 higher than
observed in Phase 1 TF/SC operation, although this difference was
not significant. The grab samples of secondary effluent TSS
indicated a slight (4 mg/1), but statistically significant,
increase in secondary effluent TSS.
The Secchi depths shown in Table 2 indicate the differences in
secondary effluent clarity between phases. TF/SC at above average
filter loadings (Phase 1) produced the highest clarity followed
by TF/SC at high loadings (Phase 3). The trickling filter mode
(Phase 2) had the lowest clarity as indicated by the lowest Secchi
depth. Differences between phases were statistically significant.
Figure 4 shows the relationship between Secchi depth and
secondary effluent TSS. The data are based on grab samples of
secondary effluent taken by the plant staff at the same time the
Secchi depth was measured. The trend line was based on Beer's law,
which states that light absorbance (100 percent for the Secchi
depth test) is proportional to depth multiplied by concentration.
-------
Table 2. Study Results
Parameter
Secondary treatment mode
Flow, m3/s (mgd)
Plant influent
Secondary treatment
Return secondary sludge
Waste temperature, degrees C
Concentration, rag /I
Raw waste
TSS
BOD5
Primary affluent
TSS
BOD5
SCBOOj
Trickling filter effluent
TSS - Filter 1
TSS - Filter 2
SCBOD5 - Filter i .__.--
SCBODj - Filter 2
Secondary effluent
TSS
BOOS
SCBODj
Secchi depth, • (ft)
Plant secondary effluent3
TSS
Secchi. depth, o (ft)
Operating parameters
Filter hydraulic loading rate.
l/t»2-s (gpm/sq ft)
Filter 1
Filter 2
Filter BODj loading rate, g/m -d
(ppd/1,000 cu ft)
Filter 1
Filter 2
Filter reaction r*te coefficient (K2g)
Filter 1, (l/m2-s)°*5
(gpm/sq ftl0'5
Filter 2, °'5
(gpm/sq ft)0'5
Return sludge parameters
Return secondary solids, mg/1
Reaeration time, minutes
Aerated solids contact parameters
Mixed liquor TSS, mg/1
Sludge volume index, ml/g
Solids contact time, minutes
Flocculator center well time," minutes
Secondary clarifier overflow rate,
m3/m -d (gpd/sq ft)
Phase
1
TF/SC
0.061 (1.40)
0.047 (1.07)
0.021 (0.47)
17.9
275
236
66
114
39
52
47
11
14
13
19
11
1.0 (3.4)
6
0.94 (3.1)
0.23 (0.34)
0.28 (0.41)
460 (29)
500 (31)
5.3 x 10"'
(6.4 x 10"')
3.2 x 10"*
(3.9 x 10"3)
4,210
9.3
1,140
129
2.8
19
18.4 (451)
2
TF
0.054 (1.24)
0.046 (1.06)
-
18.6
254
212
81
99
45
53
61
9
10
28
19
12
O.SS (1.8)
16
0.46 (1.5)
0.23 (0.34)
0.27 (0.40)
430 (27)
460 (29)
6.7 x 10"3
(8.1 X 10~3)
5.7 x 10'3
(6.9 x 10'3)
-
-
-
-
-
.
13.2 (447)
3
TF/SC
-
0.045 (1.03)
0.024 (0.54)
20.2
334
232
79
133
77
-
43
-
16
15
19
12
0.73 (2.4)
10
0.61 (2.0)
-
0.28 (0.41)
-
960 (60)
7.7 x 10'3
(9.4 X 10~3)
5,340
3.3
2,390
101
2.8
18
17.6 (43J)
"Collected downstream of flow controller for secondary treatment.
b24-hour composite samples except for Secchi depth.
cCrab samples taken during peak plant flow.
Includes filter recycle.
*Based on total flow in the center well including recycle.
-------
10
X
o>
CO
CO
H
UJ
u.
u.
UJ
o
o
UJ
CO
SECCHI DEPTH , ft
Figure 4. Relationship Between Clarity and Secondary Effluent TSS
-------
11
Secondary effluent BODc, values did not increase in Phase 2 as
expected, although the TSS increased and the clarity decreased
significantly. The secondary effluent particulate BODc (BODc minus
SCBODc) to TSS ratios for Phases 1, 2, and 3 were 0.62, 0.25, and
0.47, respectively. In our experience, the particulate BODc, to TSS
ratio is almost always above 0.5. The Phase 2 value of 0.25 is
unlikely and probably indicates the secondary effluent BODc was
actually higher.
Microscopic Examinations
Microscopic examinations and identification of filamentous
organisms have been useful in developing an understanding of
activated sludge operating conditions. These examinations can
also provide useful information on the operation of TF/SC plants.
Table 3 summarizes results of the microscopic examinations per-
formed during Phases 1 and 3 on unsettled trickling filter effluent
and mixed liquor samples.
Table 3 Filamentous Organisms in Trickling Filter Effluent and Mixed Liquor
Date
April 21, 1986
(Plus* 1)
June 16, 1986
(Plus* 3)
Sample
TF Effluent
Mixed Liouor
TF Effluent
Nixed Liquor
Filamentous
organism
type
H. hydrossis
Thiothrix
Type 1701
S. natans
S. natans
Thiothrix
Type 1701
H. hydrossis
S. natans
Beqgiatoa
H. hydrossis
Thiothrix
S. natans
Thiothrix
Filament
abundance
Few
Few
Rare
Rare
Some
Few
Few
Few
Few
Few
Few
Few
Abundant
Some
Some- to-few
Suspected operating condition'
Low food-to-aicroorqanism
_ b
Low dissolved oxygen
low dissolved oxygen
low dissolved oxygen
_o
Low dissolved oxygen
Low food-to-«icroorganiim
Low dissolved oxygen
Sulfides
Low food-to-microorganism
Sulfidee
Low dissolved oxygen
^b
ratio
ratio
ratio
Based on Reference 9*
bThiothrix species possibly indicative of presence of sulfida*. nutrient deficiency, or low-dissolved
oxygen.
Note: Ranking—abundant, common, some, few, rare, none*
The examinations revealed that the trickling filter solids and
contact tank floe were generally firm, round, and compact. They
also identified five different types of filamentous organisms.
These organisms indicated that low food-to-microorganism (F/M)
conditions and low dissolved oxygen concentrations existed con-
currently in the trickling filters. The filamentous organisms
identified in the trickling filter samples were often present in
the mixed liquor samples. The abundance of filamentous organisms
-------
12
was highest during Phase 3 when the filter loading was highest. In
Phase 3, the presence of Beggiatoa indicated significant amounts of
sulfides existed in the trickling filter. Beggiatoa became abun-
dant in the solids contact tank. Many of these organisms were
dispersed and reduced the clarity of the secondary effluent.
Other Results
The first-order reaction rate coefficients (K^cP for SCBODs
removal in the trickling filters are shown in Table 2. The Morr
Bay-Cayucos K20 values fpr the rock, filters ranged from 3.2 x 1Q~
to 7.7 x 10~J ri/m^-s)u*b[3.9 x 10~3 to 9.4 x 10"3 (gpm/sq ft)0'5].
Filter 1 yielded slightly higher removal rate coefficients than
filter 2. For each filter, the removal rate coefficient increased
as primary effluent SCBOD5 increased. The return sludge aeration
and aerated solids contact times were short. Studies at the Morro
Bay-Cayucos plant indicated that the aerated solids contact tank
removed about 1 to 2 mg/1 of SCBODg. Additional SCBOQ5 removal in
the contact tank would require longer detention times.
PERFORMANCE COMPARISONS
The following discussion draws on results of previous studies
as well as the Morro Bay-Cayucos study. It covers the effect
of trickling filter loading on TF/SC performance/ trickling filter
performance with f locculator-clarif iers, and reaction rate coeffi-
cients for rock media trickling filters.
Trickling Filter Loading
Trickling filter loading had some effect on TF/SC perfor-
mance. Loading did not have the same effect at each plant. Varia-
tions may be due to differences in operating conditions. This
section compares the effect of loading on the performance of the
Morro Bay-Cayucos plant with the other plants.
Morro Bay-Cayucos. The lack of substantial effect of filter
loading on effluent TSS at the Morro Bay-Cayucos plant indicates
that the TF/SC process will produce good effluent as long as the
trickling filters can operate at the higher loads. The presence of
significant quantities of Beggiatoa, however, suggested the rock
trickling filter was approaching its' limit at the loading of
960 g/m-^-d (60 ppd/1,000 cu ft).
Beggiatoa are aerobic autotropMc organisms that use sulfides
as their primary source of energy. ° Sulfides are a by-product of
anaerobic activity and are often associated with highly or over-
-------
13
loaded rotating biological contactors and trickling filters.
Despite the high loadings, the trickling filter did not produce any
noticeable odors.
Figure 5 shows the effect of filter loading at the Morro Bay-
Cayucos plant and other TF/SC plants with rock filters. The Morro
Bay-Cayucos composite samples are the Phase 1 and 3 averages from
the full-scale studies. The grab samples are monthly average
values (based typically on 20 samples per month) collected by the
plant staff the year before the study and phase averages collected
during the study. The Morro Bay-Cayucos plant results are the
only data available for rock filter loadings above 480 g/m -d
(30 ppd/1,000 cu ft). The composite samples showed a 2 mg/1
increase in effluent TSS when the loading was doubled from 480 to
960 g/nr-d (30 to 60 ppd/1,000 cu ft). The grab samples showed a
4 mg/1 increase in effluent TSS.
The highest secondary effluent TSS shown on Figure 2 (19 mg/1)
occurred when all of the plant flow [0.066 m /s (1.5 mgd)] received
secondary treatment, rather than a fraction [typically 0.043 irr/S
(1.0 mgd)] when split treatment was used. The filter loading
increase occurred because of the higher primary effluent flow to
the filters. The 50 percent primary effluent flow increase reduced
contact times and increased the secondary clarifier overflow rate
by 50 percent. Consequently, large changes in contact times and
overflow rate in addition to filter loading, probably contributed
to the increase in secondary effluent TSS.
The Morro Bay-Cayucos plant performed well at the high
loadings. One factor that may contribute to the unusually good
performance at the Morro Bay-Cayucos plant is the consistent
secondary treated flow. As noted earlier, the secondary treated
flow was maintained at about 0.043 m /s (1.0 mgd) with no diurnal
variation. This consistency should improve clarification. The
secondary effluent TSS increase during full secondary treatment may
have been affected by the diurnal flow variation in addition to the
contact time reduction and overflow rate increase noted earlier.
Corvallis. At Corvallis, the secondary effluent TSS increased
from 8 to 11 mg/1 when filter loading increased from 160 to
400 g/nr-d (10 to 25 ppd/1,000 cu ft). One of the two filters was
shut down briefly in September 1980 to assess the effect of higher
loading. Filter loading was increased to about 640 g/nr-d
(40 ppd/1,000 cu ft), but the test -was discontinued because of
odors.
The Corvallis filters developed odors at loadings significantly
lower than loadings successfully applied to the Morro Bay-Cayucos
filters. The Corvallis rock filters are 2.4-m (8-feet) whereas
the Morro Bay-Cayucos filter depths are 1.4- and 1.5-m (4.5-
and 5-ft). As a result, the Corvallis filters may have less
-------
14
35
O)
of
H
Z
LU
U.
U.
Ill
a
o
o
UJ
(O
30
20
O OCONTO FALLS, 8/3/84 - 8/16/84
. A OCONTO FALLS, 6/11/84- 7/11/84
Q OCONTO FALLS, 7/83 - 11/83
Q MORRO BAY - CAYUCOS (COMPOSITE), 3/86 - 6/86
V MORRO BAY - CAYUCOS (GRAB), 3/85 - 6/86
MORRO BAY-CAYUCOS
FULL SECONDARY
TREATMENT
V
10
10
20
30
40
50
60
FILTER ORGANIC LOADING, ppd BODg/1000 cu ft
Figure 5. Effect of Loading on TF/SC Performance with Rock Filters
-------
15
ventilation. Also, for an equivalent volumetric BODc loading,
deeper filters receive higher loadings in the top portion of the
filter where most BOD^ removal occurs. The higher loadings near
the filter top may create an overload condition and cause odors.
Oconto Falls. Data collected at the Oconto Falls TF/SC plant
for months with a similar waste temperature indicated a greater
effect of loading on TF/SC performance than at the Morro Bay-
Cayucos and Corvallis TF/SC plants. Average secondary effluent TSS
increased 8 mg/1 when loading was increased from 160 to 400 g/m -d
(10 to 25 ppd/1,000 cu ft), instead of 3 mg/1 as at Corvallis.
Loading may have had a greater effect at Oconto Falls than at
Corvallis or Morro Bay-Cayucos because Oconto Falls only averages
8 minutes of aerated solids contact time with no return sludge
aeration, which decreases the opportunity for physical and biolo-
gical flocculation.
Trickling Filter Performance With
Flocculator-Clarifiers
Flocculator-clarifiers generally include high sidewater depths
(5.0 to 6.3 m), flocculator center wells, hydraulic sludge removal,
and inboard effluent launders. The benefits of flocculator-clari-
fiers have been discussed '; however, no data has been presented
to show how they can improve trickling filter plant performance
when an aerated solids contact tank is not used. Recent data from
the Morro Bay-Cayucos facility and long-term data from the Coeur
d'Alene, Idaho, plant demonstrate that flocculator-clarifiers can
significantly improve performance of trickling filters plants with
conventional shallow clarifiers.
Morro Bay-Cayucos. The trickling filter mode (Phase 2)
produced secondary effluent with an average TSS of 28 mg/1 when the
trickling filter loading was 445 g/nr-d (28 ppd/1,000 cu ft). The
secondary effluent quality was not as good as the TF/SC process but
did meet secondary effluent standards. The clarity of the second-
ary effluent was significantly less than the TF/SC process because
of the lack of aerated solids contact for flocculation.
Coeur d'Alene. The Coeur d'Alene operating records provide
long-term data to assess the effect of flocculator-clarifiers on
trickling filter performance. The Coeur d'Alene plant has an
average dry weather flow of 0.10 m3/s (2.3 mgd) and a 2.1-m
(7-foot) deep roclt trickling filter with an average filter loading
of about 290 g/mj-d (18 ppd/1,000 cu ft). The original plant
included a shallow 2.1-m (7-foot) deep secondary clarifier with
peripheral effluent weirs and a scraper mechanism for sludge
removal. In 1983, the shallow secondary clarifier was replaced
with a flocculator-clarifier, the only major modification made to
the plant.
-------
16
Figure 6 shows the effect of the flocculator-clarifier on
plant performance. The flocculator-clarifier significantly reduced
the average effluent TSS from 25 mg/1 to 16 mg/1. Additionally,
the plant operated more stably as noted by the reduction in the
standard deviation of the monthly average values. No secondary
treatment permit violations have occurred since the flocculator-
clarifier was installed.
50
^ 40
0)
£ 5
> UJ
< D
30
20
o
o
UJ
10
STANDARD
SHALLOW CLARIFIER
FLOCCULATOR - CLARIFIER
AVERAGE TSS s 25 mg/l
STANDARD DEV = 5.8 mg/1
A
I
,\;
AVERAGE TSS =16 mg/l
STANDARD DEV s 3.6 mg/l
i ,
1982
1983
1984
1985
1986
Figure 6. Effect of Flocculator-Clarifier on Trickling Filter Performance at
Coeur d'Alene, Idaho
It should be noted that the new flocculator-clarifier is 23-m
(75-foot) in diameter, while the old clarifier was 18-m (60-foot)
in diameter. The reduction in average effluent TSS with the
-------
17
flocculator-clarifier, however, is not attributable to the increase
in clarifier surface area. The performance of. the two clarifiers
was compared at the same overflow rate [24 mVm -d (590 gpd/sq ft)]
using low flow data with the old clarifier and high flow data with
the flocculator-clarifier. The flocculator-clarifier yielded the
same average effluent TSS value (16 mg/1), while the old clarifier
yielded a higher average effluent TSS value (31 mg/1). Conse-
quently, the flocculator-clarifier improved effluent quality
because of its inherent features and not because of the increase in
clarifier surface area.
Reaction Rate Coefficients
The K2Q values computed from the Morro Bay-Cayucos data
were substantially higher than the K2g values reported for
plastic media. Studies at Oconto Falls yielded the js^me range in
K2Q values for rock filters. Richards, et al. summarized
K2n values for cross-flow jnedia from various .studies and showed
they ranged from 1.7 x 10~* to 2.1 x 10~3 (l/mz-s)g'5 [2.0 x 10"3
to 2.6 x 10~3 (gpm/sq ft)0*5],. Drury, et al.15 reported a K20
value of 3.2 x 10~3 (l/m2-s)u>5 [3.9 x 10~3 (gpm/sq ft)0'5] for a
1.02-m (3.33-ft) deep filter with cross-flow media.
K20 values developed with the modified-Velz equation show that
a unit of rock media surface area is more efficient at removing
SCBODg than a unit of plastic media surface area. Differences in
K2g values may be caused by a.decrease in the resistance to mass
transfer in the liquid film when comparing rock and plastic
media. Differences in K2g values may also be caused in part by the
inadequacy of ttie modiried-Velz equation to account for depth.
Drury, et al., noted relatively high K™. values for a shallow
plastic media filter, although the K2Q values were in the low end
of the range measured with the rock filters at Morro Bay-Cayucos
and Oconto Falls. Plastic media, however, typically has a specific
surface area (media surface area/unit of media volume) that is 2 to
3 times as great as rock media, which may compensate for
differences in efficiency.
The K2Q values from the Morro Bay-Cayucos studies increased
with influent SCBOD5. This relationship indicates that the
modified-Velz equation does not accurately describe SCBOD5 removal
in rock trickling filters.
CONCLUSIONS
The Morro Bay-Cayucos studies showed that TF/SC can produce
high quality effluent with rock filters even up to loadings as high
as 960 g/m^-d (60 ppd/1,000 cu ft). Results indicate that if the
-------
18
trickling filter can operate satisfactorily at this high load, the
TF/SC process will produce its typically high quality effluent.
The study team noted the presence of Beggiatoa in trickling
filter effluent at 960 g/m-'-d (60 ppd/1,000 cu ft) loadings. These
bacteria became abundant in the solids contact tank. Their
presence indicates an increase An anaerobic activity in the filters
and suggests that the 960 g/nr-d loading may mark an upper limit
for good trickling filter performance at the Morro Bay-Cayucos
plant.
Nevertheless, these loadings are significantly higher than the
previously tested 400 g/nr-d (25 ppd/1,000 cu ft) loadings at the
Corvallis TF/SC plant. They provide a wide margin of potential
increased capacity at existing rock filter plants and indicate that
these plants can be expanded merely by adding solids contact
features without constructing new trickling filters. Each plant
should be evaluated individually, since all rock filters may not
operate effectively at such high loadings. The possibility of such
economical expansion is particularly important in view of the large
number of rock filter plants that will need upgrading by the year
2000. The USEPA has estimated that number to be about 1,500.
VJork at the Morro Bay-Cayucos plant and long-term data at
the Coeur d'Alene plant showed that effluent quality from rock
trickling filter plants can be improved significantly simply by
replacing conventional secondary clarifiers with flocculator-clari-
fiers. The flocculator-clarifier at Coeur d'Alene reduced average
effluent TSS from 25 mg/1 to 16 mg/1. The plant effluent was also
more stable and has not exceeded secondary treatment discharge
limits since addition of the flocculator-clarifier.
Another result has implications for the design of expansions to
rock filter plants. K2o values reported for filter SCBOD5 removal
are often higher for rock media filters than for deeper plastic
media filters. The higher I^g values indicate rock media ade-
quately removes SCBODg and compensates for its lower specific
surface area with higher reaction rates. The differences indicate
a unit of rock media surface area is more efficient than a unit of
plastic media surface area. The ^Q values increased with SCBOD^
concentration indicating the modiried-Velz equation does not
accurately describe SCBODe removal in rock media filters. Design-
ers must be aware of the limitation of'the modified-Velz equation.
ACKNOWLEDGMENTS
This material has been funded wholly or in part by the USEPA
under contract 68-03-1818 to Brown and Caldwell. It has been
subject to the Agency's review, and it has been approved for
-------
19
publication as a USEPA document. Mention of trade names or
commercial products does not constitute endorsement or
recommendation for use.
Credits
This paper was presented at the 59th Annual Conference of the
Water Pollution Control Federation/ Los Angeles, California,
October 9, 1986. Curtis Weeks collected a significant portion of
the data at the Morro Bay-Cayucos treatment plant under the
supervision of Professor Samuel Vigil. Curtis Weeks and Samuel
Vigil are associated with California Polytechnic University at San
Luis Obispo. The author acknowledges the cooperation and
operational support given by the staff members at the Morro Bay-
Cayucos plant. The authors also wish to acknowledge Tom Liston,
plant superintendent at Coeur d'Alene, Idaho, for providing plant
data.
Authors
Raymond N. Matasci is a project manager with Brown and
Caldwell, Pleasant Hill, California. David L. Clark is a project
manager with Brown and Caldwell, Seattle, Washington. James A.
Heidman is a staff engineer with the Water Engineering Research
Laboratory, U.S. Environmental Protection Agency, Cincinnati,
Ohio. Denny S. Parker is a senior vice president for Brown and
Caldwell, Pleasant Hill, California. Bruce Petrik is a project
manager with Brown and Caldwell, Denver, Colorado. Darrel Richards
is the water quality control superintendent for Morro Bay,
California. Correspondence should be addressed to Raymond N..
Matasci at Brown and Caldwell, 3480 Buskirk Avenue, Pleasant Hill,
California 94523.
REFERENCES
1. U.S. Environmental Protection Agency. "Technical Support
Document for Proposed Regulations Under Section 304(d of the
Clean Water Act, as Amended." Prepared by Facility Requirements
Division, Office of Water, Washington, D.C. August (1984).
2. Matasci, R.N., et al. "Full-scale Studies of the Trickling
Filter/Solids Contact Process." Accepted for publication in J.
Water Pollution Control Federation (1986).
3. Brown and Caldwell. Trickling Filter/Solids Contact Process;
Full-Scale Studies. EPA/600/S2-86/046. Water Engineering
Research Laboratory, Cincinnati, Ohio. May (1986).
4. Fedotoff, R.C. "The Trickling Filter Finds a New Partner."
Water-Engineer & Management, June (1983).
-------
20
5. Norris, D.P., et al. "Production of High Quality Trickling
Filter Effluent Without Tertiary Filtration." _J_. Water
Pollution Control Federation, 54, 1087 (1982).
6. Pierce, D.M. Upgrading Trickling Filters. Prepared for the
U.S. Environmental Protection Agency, Office of Water Program
Operations, Washington, D.C. EPA 430/9-78-004. June (1978).
7. American Public Health Association. Standard Methods for the
Examination of Water and Wastewater. 16th Ed. (1985).
8. Parker, D.S., and D.T. Merrill. "Effect of Media Configu-
ration on Trickling Filter Performance." J. Water Pollution
Control Federation/ 56, 8, 955 (1984).
9. Richard, M.G., et al. "Growth Kinetics of Sphaerotilus Species
and Their Significance in Activated Sludge Bulking." Presented
at the 55th Annual Conference of the Water Pollution Central
Federation, St. Louis, Missouri. October 1982.
10. Hawkes, H.A. The Ecology of Waste Water Treatment. The
MacMillian Company. New York, New York (1963).
11. Brenner, R.C., et al. Design Information on Rotating
Biological Contactors* Prepared for U.S. Environmental
Protection Agency, Wastewater Research Division, Municipal
Environmental Research Laboratory, Cincinnati, Ohio. EPA-
600/2-84-106.
12. Parker, D.S. "Assessment of Secondary Clarification
Design Concepts." J. Water Pollution Control Federation, 55,
4, 349 (1983).
13. Parker, D.S., and R.J. Stenquist. "Flocculator-Clarifier
Performance." J. Water Pollution Control Federation, 58, 3,
214 (1986).
14. Richards, T., and D. Reinhart. "Evaluation of Plastic Media in
Trickling Filters." J. Water Pollution Control Federation, 58,
7, 774 (1986).
15. Drury, D.D., et al. "Evaluation of High Density Cross Flow
Media for Rehabilitating an Existing Trickling Filter." J.
Water Pollution Control Federation, 589, 5, 364 (1986).
16. Logan, B.E. "A Fundamental Model for Trickling Filter Process
Design." Presented at the 59th Annual Conference of the Water
Pollution Control Federation, Los Angeles, California,
------- |