EPA-600/2-77-144
September 1977
Environmental Protection Technology Series
DOWNFLOW GRANULAR FILTRATION OF
ACTIVATED SLUDGE EFFLUENTS
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-144
September 1977
DOWNFLOW GRANULAR FILTRATION
OF
ACTIVATED SLUDGE EFFLUENTS
by
Robert P. G. Bowker
EPA-DC Pilot Plant
Washington, B.C. 20032
Contract No. 68-03-0349
Project Officer
Irwin J. Kugelman
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
ii
-------
FOREWORD
The Environmental Protection Agency was created because of increasing public
and government concern about the dangers of pollution to the health and
welfare of the American people. Noxious air, foul water, and spoiled land
are tragic testimony to the deterioration of our natural environment. The
complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solution and
it involves defining the problem, measuring its impact, and searching for
solutions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems for the prevention, treatment, and management
of wastewater and solid and hazardous waste pollutant discharges from
municipal and community sources, for the preservation and treatment of public
drinking water supplies, and to minimize the adverse economic, social, health,
and aesthetic effects of pollution. This publication is one of the products
of that research; a most vital communications link between the researcher and
the user community.
In an effort to achieve a high quality effluent from wastewater treatment
facilities, filtration has often been considered as an important integral
part of the treatment sequence. This report summarizes the investigations
at the EPA-DC Pilot Plant dealing with utilization of various filter media
sizes and types for treatment of activated sludge process effluents.
Francis T. Mayo, Director
Municipal Environmental
Research Laboratory
iii
-------
ABSTRACT
The performance of downflow granular filters subjected to effluents from
activated sludge processes was investigated at the EPA-DC Pilot Plant
in Washington, D.C. The 0.1 m2 (1 ft2) filters were operated at
hydraulic loadings from 0.12 to 0.37 m3/min/m2 (3 to 9 gpm/ft2). Several
media combinations were investigated, including both single anthracite
and dual anthracite-sand configurations. Effluents from step aeration,
plug flow, and completely mixed activated sludge systems were used as
feeds.
Breakthrough of the suspended solids into the effluent occurred with both
the 1.65 mm and 2.0 mm effective size (E.S.) single anthracite configura-
tions, becoming more evident at the higher flow rates.
A dual media filter, consisting of 2.0 mm E.S. coal over 0.9 mm E.S.
sand, exhibited the most desirable characteristics for filtration of the
secondary effluents investigated. The advantages were longer run times
and higher suspended solids loadings with virtually no deterioration of
effluent quality.
A backwash study conducted at a variety of backwash flowrates showed
that a 13 percent bed fluidization was achieved with the coarse media
(2.0 mm E.S. coal/0.9 mm E.S. sand) at a flow rate of 1.43 m3/min/m2
(35 gpm/ft2). This was sufficient to effectively cleanse the media.
This report was submitted in fulfillment of Contract No. 68-03-0349 by
the EPA-DC Pilot Plant under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period from September 1974
to May 1975, and work was completed as of July 1975.
iv
-------
CONTENTS
Foreword
Abstract iv
Figures vi
Tables vii
Acknowledgments viii
1. Introduction 1
2. Summary 3
3. Conclusions 4
4. Recommendations 5
5. Experimental Apparatus and Operation 6
6. Effect of Media Characteristics on Filter Performance ... 8
7. Effect of Flowrate on Filter Performance 20
8. Effect of Process Effluent on Filter Performance 27
9. Filter Backwashing 30
10. Phosphorus Removal 37
11. Engineering Significance 39
References 44
Appendix 45
-------
FIGURES
Number Page
1 Schematic of Dual Media Filters ................... 7
2 Head Loss as a Function of Solids Capture
December 1974 ........................... -
3 Differential Head Loss Characteristics
December 1974 ........................... 15
4 Head Loss as a Function of Solids Capture
January 1975 ........................... 17
5 Differential Head Loss Characteristics
January 1975 ........................... 18
3 2
6 Removal Efficiency at 0.12 m /min/m
Feed: Complete Mix Activated Sludge (C.M.A.S.) Effluent ..... 21
3 2
7 Removal Efficiency at 0.20 m /min/m
Feed: C.M.A.S. Effluent ..................... 22
3 2
8 Removal Efficiency at 0.29 m /min/m .
Feed: C.M.A.S. Effluent ..................... 23
3 2
9 Removal Efficiency at 0.37 m /min/m
Feed: C.M.A.S. Effluent ..................... 24
3 2
10 Removal Efficiency at 0.29 m /min/m
Feed: C.M.A.S. Effluent ........... . ......... 25
3 2
11 Removal Efficiency at 0.37 m /min/m
Feed: C.M.A.S. Effluent ..................... 26
12 Extent of Sand, Anthracite Bed Expansion during Backwash ...... 31
13 Extent of Total Bed Expansion during Backwash ............ 32
14 Removal of Entrapped Solids during Filter N Backwash ....... ' . 34
15 Removal of Entrapped Solids during Filter 0 Backwash ........ 35
16 Removal of Entrapped Solids during Filter P Backwash ........ 36
vi
-------
TABLES
Number Page
1 Characteristics of Media Utilized during Period
September 1974 - May 1975 9
2 Summary of Filter Performance 10
3 "Intermix Quotients" for Dual-Media Configurations 19
4 Summary of Filter N Data 28
5 Phosphorus Removal Study 38
6 Net Water Production 41
vii
-------
ACKNOWLEDGMENTS
The author gratefully acknowledges the assistance of the operators, tech-
nicians and laboratory staff at the EPA-DC Pilot Plant in the construction,
operation and maintenance of the systems described in this report.
viii
-------
SECTION 1
INTRODUCTION
Achievement and maintenance of a high quality effluent from wastewater
treatment facilities will continue to warrant increased utilization of
downflow granular filters for removal of suspended solids. As effluent
standards tighten, many wastewater treatment operations will employ some
form of filtration to meet effluent suspended solids criteria.
In the design of filtration systems, three classes of parameters must be
considered. These are: wastewater feed characteristics (suspended solids
concentration, floe size distribution, floe strength and floe charge) ,
filter media characteristics (media type, size distribution, shape, and
depth), and operational conditions (hydraulic loading, terminal head loss,
and type of media cleaning technique).
The filtration characteristics of activated sludge effluents are expected
to vary with the operational parameters of the activated sludge process
(D.O. level and SRT) and the process mode used (step aeration, complete mix,
etc) . The original purpose of this study was to evaluate in parallel the
filterability of activated sludge effluents from the three most popular
process flow regimes (plug flow, complete mix, and step aeration). The
activated sludge systems were to be operated under identical conditions and
each of the filters was to be identical in media type, depth, and size. It
was found, however, that periodic upsets, primarily due to proliferation of
either filamentous organisms or Norcardia sp., made it impossible to operate
these three process modifications under identical conditions for long periods
of time. In addition, it was found to be difficult to prepare three filters
in which the media size and depth were the same (within the allowable
experimental variation).
Therefore, after several runs were made, the objective of the study was
modified. The new objective was to study the effect of media characteristics
on the filterability of effluents from each activated sludge modification
in order to determine the most effective media configuration. The filters
were packed with various media and used in parallel to treat effluent from
only one of the process modifications at a time.
A wide range of particle sizes has been suggested for dual media filtration
of secondary effluents, effective sizes varying from 1.0 to 2.0 mm for
anthracite, and 0.4 to 0.9 mm for sand. Media depths generally vary from
38 to 64 cm (15 to 25 in) for anthracite and 30 to 38 cm (12 to 15 in) for
silica sand.
-------
This study attempted to assess the value of using various sizes of filter
media for removal of suspended solids from each of three secondary effluents
(step aeration, plug flow, and completely mixed activated sludge). Special
consideration was given to the possibility of utilizing a relatively coarse,
deep bed media configuration that could maintain an effluent of suitable
quality while demonstrating longer run lengths and higher suspended solids
loadings.
-------
SECTION 2
SUMMARY
The performance of downflow granular filters receiving effluents from
activated sludge processes was investigated at the EPA-DC Pilot Plant in
Washington, B.C. The 0.1 m2 (1 ft2) filters were operated at hydraulic
loadings from 0.12 to 0.37 m^/min/m2 (3 to 9 gpm/ft2). Several media
combinations were investigated, including both single anthracite and dual
anthracite-sand configurations. Effluents from step aeration, plug flow,
and completely mixed activated sludge systems were used as filter feeds.
The influent and effluent streams were monitored for suspended solids (SS),
and the differential pressures across the filter media were recorded
periodically. The filters were backwashed when the total head loss exceeded
2.5 m (100 in ) or when breakthrough of the suspended solids was detected.
Breakthrough of the suspended solids into the effluent occurred with both
the 1.65 mm and 2.0 mm effective size (E.S.) single anthracite configurations,
becoming more evident at the higher flowrates.
A dual media filter, consisting of 2.0 mm E.S. coal over 0.9 mm E.S. sand,
exhibited the most desirable characteristics for filtration of the secondary
effluents investigated. The advantages were longer run times and higher
suspended solids loadings with virtually no deterioration of effluent quality
compared to more conventional media combinations.
A backwash study conducted at a variety of backwash flowrates showed that a
13 percent bed fluidization was achieved with the coarse media (2.0 mm E.S.
anthracite/0.9 mm E.S. silica sand) at a flow rate of 1.43 m3/min/m2 (35 gpm/
ft2). This was sufficient to effectively cleanse the media.
Problems encountered during operation of the pilot filters included unusually
low influent suspended solids concentrations, and proliferation of Norcardia
organisms in the activated sludge system. The latter caused considerable
difficulty in attempting to assess the filterability of the three process
effluents.
-------
SECTION 3
CONCLUSIONS
From the investigations conducted at the EPA-DC Pilot Plant at Blue
Plains, the following conclusions were drawn concerning downflow
granular filtration of activated sludge effluents:
1. Use of relatively coarse media (64 cm of 2.0 mm effective
size (E.S.) anthracite and 38 cm of 0.9 mm E.S silica sand)
produced effluent qualities similar to those from much
finer media gradations at hydraulic loadings of 0.12 to
0.37 m3/min/m2 (3 to 9 gpm/ft2).
2. The 2.0 mm E.S. coal/0.9 mm E.S. sand media exhibited run
lengths and suspended solids loadings significantly higher
than other finer media gradations investigated.
3. A backwash rate of 1.50 m3/min/m2 (37 gpm/ft ) for four
minutes was sufficient to effectively cleanse the 2.0 mm
E.S. coal/0.9 E.S. sand media. Air scour preceeding water
backwash on both coarse and fine dual media filters was
needed to prevent formation of "mud balls." Although the
backwash water requirement for this media was greater than
for media of smaller effective sizes, the frequency of
backwash was reduced to such an extent that daily water
production was relatively independent of filter media sizes.
4. 64 cm of 1.65 mm effective size anthracite was inadequate
to sustain a high quality effluent at hydraulic loadings
greater than 0.12 m3/min/m2 (3 gpm/ft2).
5. For the coarse filter media investigated, the potential
savings in backwash water resulting from less frequent
backwash was offset by the greater backwash water require-
ment to achieve a desired uniform level of bed expansion.
Thus, the values of net water production were similar for
each media at any given flowrate. Based on effluent
quality and net water production, there was no advantage in
using a coarser filter media. However, the lower backwash
frequency and correspondingly lower manpower requirements
for backwash operations were a possible economic advantage.
-------
SECTION 4
RECOMMENDATIONS
Based on the current literature and on investigations performed at
the EPA-DC Pilot Plant with respect to filtration of activated sludge
effluents, the following recommendations are suggested as topics for
future research:
1. Determine the optimum bed expansion for backwashing dual media
filters treating activated sludge effluents.
2. Investigate the amenability of effluents from extended aeration
and high-SRT activated sludge processes for treatment by dual-
media filters with coarse media gradations.
3. Assess the performance of coarse, dual-media filters under
hydraulic and suspended solids surges.
-------
SECTION 5
EXPERIMENTAL APPARATUS AND OPERATION
Three filters were constructed of 35.6 cm (14 in) diameter clear plexiglass
columns approximately 2.4 m (8 ft) in length, each having a lateral surface
area of 0.1 m2 (1 ft2) (see Figure 1). Taps were provided at 7.6 cm (3 in)
intervals to monitor the differential pressure across the media. A 0.14 m^/
min (5 scfm) air wash system was installed so as to dissipate, during the
backwash, scum formations in the filter media. For backwashing of the
filters, tap water was pumped from a holding tank upward through the media at
hydraulic loadings of up to 1.42 m3/min/m2 (35 gpm/ft2) and discharged to
drain.
The support media consisted of approximately 15 cm (6 in) of 1.3 cm (0.5 in)
diameter stone, above which was placed 10 cm (4 in) of 0.3 cm (.125 in)
diameter garnet to prevent the overlying sand from penetrating the support
media. The filter media consisted of approximately 38 cm (15 in) of sand
under 64 cm (25 in) of anthracite coal, except in the studies which utilized
a single anthracite media.
Process effluent from step aeration, plug flow, completely mixed, and/or
single stage nitrification-denitrification activated sludge systems was
pumped to a splitter box, from which it flowed by gravity to the filters at
hydraulic loadings from 0.12 to 0.37 m3/min/m2 (3 to 9 gpm/ft2). As the
differential pressure increased across the filter media, the height of the
water increased in the column to maintain constant flow. A maximum of 3 m
(10 ft) of total head was available, although backwash operations were
initiated when the head exceeded 2.5 m (8.3 ft). The backwash sequence is
discussed in Section 9.
During normal pilot operation of the filters, daily influent and effluent
4-hour composites were collected and analyzed for suspended solids (SS),
volatile suspended solids (VSS), chemical oxygen demand (COD), and 5-day
biochemical oxygen demand (BOD). Differential pressure readings were taken
prior to, and immediately following, backwash, in addition to the normal
4-hour readings. During the special studies operations, differential
pressure readings were taken hourly and the influent and effluent streams
sampled every two hours for suspended solids analysis.
-------
SPLITTER BOX
INLET
PIPE
FLOW
METER
15 cm
(38 cm)
AIR WASH
^ TO
DRAIN
•\ ;
TO DRAIN
—-DX—
-ANTHRACITE
(64 cm)
AIR SPARGER
— INTERFACE
X
EFFLUENT
FLOWMETER
9^n^^^R
-««FROM PROCESS
BACKWASH
WATER
Figure 1. Schematic of Dual Media Filters
-------
SECTION 6
EFFECT OF MEDIA CHARACTERISTICS ON FILTER PERFORMANCE
The three downflow granular filters were initially designed to contain 64 cm
(25 in ) of 1.3 mm effective size (E.S.) anthracite over 38 cm (15 in ) of
0.65 mm effective size (E.S.) sand. However, due to widely varying run
lengths produced under similar suspended solids loading conditions during
the first few months of the study, it was suspected that differences in the
media existed. During the period March-May, 1974, filters N, 0, and P
received the effluents from the step aeration, plug flow, and completely
mixed activated sludge systems respectively. Thus, media comparison was
not possible. To compare the performance of the filters, the effluent from
the completely mixed system was fed to all three filters in June 1974. An
analysis of the head loss characteristics confirmed the suspicion that
filter P contained a finer gradation of media. (See Appendix, Figure A-l).
Prior to replacement of the filter P media in September 1974, a sieve
analysis of the existing anthracite yielded an E.S. of approximately 1.2 mm,
as opposed to the 1.3 mm anthracite in filters N and 0. Inspection of
Appendix Figure A-l also reveals a difference in head loss characteristics
between filters N and 0, which may be of great enough significance to make
comparisons of filterability of the three process effluents difficult.
In addition to the difficulties in placement of identical filter media, a
valve malfunction on filter 0 in July 1974 resulted in the loss of
approximately 18 cm (7 in ) of anthracite during backwash. Media replacement
was not effected until September 16, 1974.
Since characterization of the filter media was not sufficient to allow
effective data handling, the data from the period July 1-September 16, 1974
has been placed in the Appendix of this report.
During the week of September 16, the media in filters 0 and P was replaced.
Filter 0 was packed with sand of effective size 0.8 mm and coal of effective
size 1.65 mm. Filter P media was replaced with 64 cm (25 in ) of 1.65 mm
E.S. anthracite, with no underlying sand. It was intended that filter N
media remain the same throughout the duration of the study to serve as a
control. However, some media loss was experienced in November, 1974 (see
Table 1).
Operation of the filters with the effluent from the completely mixed activated
sludge system at 0.16 m^/min/m2 (4 gpm/ft2) produced loadings of 0.433,
0.625 and 0.753 kg/m2/m head loss (.027, 0.039, and 0.047 Ibs/ft2/ft head
loss) for filters N, 0 and P respectively (see Table 2). Filter 0 accepted a
-------
TABLE 1
CHARACTERISTICS OF FILTER MEDIA UTILIZED
DURING PERIOD
SEPTEMBER 1974-MAY 1975
Filter
ANTHRACITE
E.S. (mm) Depth (cm)
E.S. (mm)
SAND
Depth (cm) Dates
N
0
P
N
0
P
N
0
P
N
0
P
1.3
1.65
1.65
1.3
2.0
1.65
1.3
2.0
2.0
1.3
2.0
2.0
64
64
64
48
64
64
48
64
64
64
64
64
0.65
0.8
-
0.65
-
-
0.65
-
0.9
0.65
0.65
0.9
38
38
-
38
-
-
38
-
38
38
38
38
9/22-
11/5
11/5 -
12/9
12/9 -
1/13
1/13-
5/30
Uniformity coefficients:
Anthracite 1.5 - 1.6
Sand 1.4
-------
TABLE 2
SUMMARY OF FILTER PERFORMANCE
Month Filter
Sept. N*
174 o*
p*
Oct. N
0
P
o Nov. N*
0*
p*
N*
0*
p*
N*
p*
N*
N*
Media E.S. (mm)
Coal /Sand
1.3/0.65
1.65/0.79
1.65/
1.3/0.65
2.0/
1.65/
1.3/0.65
2.0/
1.65/
1.3/0.65
2.0/
1.65/
1.3/0.65
1.65/
1.3/0.65
1.3/0.65
Flow
ro /ni in /in
0.16
0.16
0.16
0.16
0.16
0.16
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.29
0.37
Process
Effluent
Type/SRT (days)
CM/2.0
CM/2.0
CM/2.0
CM/2.0
CM/2.0
CM/2.0
STEP /3. 7
STEP/3.7
STEP /3. 7
PLUG/4.7
PLUG/4.7
PLUG/4.7
CM/4-7
CM/4-7
CM/4-7
CM/4-7
Avg.
SS in
ppm
9.6
9.7
9.7
16.5
15.2
16.5
5.6
6.8
6.9
6.3
6.J3
7.3
42.5
37.3
43.7
28.9
Avg.
SS out
ppm
2.9
2.9
3.5
6.6
6.0
7.5
2.1
3.4
2.9
3.7
4.4
4.4
1.1
5.0
5.1
2.9
Avg.
Rem.
70
70
64
60
61
55
62
50
57
41
35
40
97
87
89
91
Avg.
Loading
Kg/in /m
head loss
0.38
0.56
0.55
0.43
0.63
0.75
0.32
**
**
0.38
**
**
1.19
**
1.83
1.27
* "special study" (short duration)
** run terminated before 2.5 m (100 in.) total head loss
-------
TABLE 2 (con't)
SUMMARY OF FILTER PERFORMANCE
Month Filter
Dec. N*
174 o*
p*
N*
0*
P*
N*
0*
P*
N*
0*
P*
N*
0*
P*
Media E.S. (mm)
Coal /Sand
1.3/0.65
2.0/
2.0/0.9
1.3/0.65
2.0/
2.0/0.9
1.3/0.65
2.0/
2.0/0.9
1.3/0.65
2.0/
2.0/0.9
1.3/0.65
2.0/
2.0/0.9
-Flow _
m /min/m
0.12
0.12
0.12
0.20
0.20
0.20
0.29
0.29
0.29
0.37
0.37
0.37
0.20
0.20
0.20
Process
Effluent
Type/SRT (days)
CM/8.1
CM/8.1
CM/8.1
CM/8.1
CM/8.1
CM/8.1
CM/8.1
CM/8.1
CM/8.1
CM/8.1
CM/8.1
CM/8.1
STEP/3.5
STEP /3. 5
STEP/3.5
Avg.
SS in
ppm
21.0
21.8
21.0
16.0
17.8
17.4
30.8
32.0
34.2
34.7
42.7
35.6
13.2
13.3
13.4
Avg.
SS out
ppm
0.5
2.3
0.5
1.4
4.9
1.4
1.4
15.0
4.7
6.1
19.9
9.0
2.9
4.5
2.9
Avg.
Rem.
97
89
97
91
73
92
95
53
86
83
47
73
77
66
78
Avg.
Loadjng
Kg/m /m
head loss
2.26
**
3.56
1.19
1.97
1.89
2.00
**
3.29
1.88
**
2.61
1.01
**
1.70
* "special study" (short duration)
** run terminated before 2.5 m (100 in.) total head loss
-------
TABLE 2 (con't)
SUMMARY OF FILTER PERFORMANCE
Month
Dec.
'74
Jan.
'75
Mar.
Filter
N*
0*
p*
N*
0*
p*
N*
0*
P*
N*
o*
p*
N*
0*
p*
N
P
Media E.S. (mm)
Coal /Sand
1.3/0.65
2.0/
2.0/0.9
1.3/0.65
2.0/0.65
2.0/0.9
1.3/0.65
2.0/0.65
2.0/0.9
1.3/0.65
2.0/0.65
2.0/0.9
1.3/0.65
2.0/0.65
2.0/0.9
1.3/0.65
2.0/0.9
3Flow 2
m /min/m
0.29
0.29
0.29
0.12
0.12
0.12
0.20
0.20
0.20
0.29
0.29
0.29
0.37
0.37
0.37
0.20
0.20
Process
Effluent
Type/SRT (days)
STEP/3.5
STEP/3 . 5
STEP/3 . 5
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/6-7
CM/3.0
CM/3.0
Avg.
SS in
ppm
16.7
14.4
14.4
16.3
16.5
15.6
8.8
9.2
8.7
32.6
27.1
31. "
32.3
39.3
29.5
17.6
18.9
Avg.
SS out
ppm
9.4
9.2
6.5
2.1
2.5
2.1
0.9
1.0
0.9
8.5
7.9
9.8
10.8
14.5
12.1
6.5
5.4
Avg.
%
Rem.
52
44
58
87
84
86
89
88
89
74
71
69
66
63
56
63
71
Avg.
Loading
Kg/m /m
head loss
0.46
**
0.93
2.21
2.08
3.70
1.35
1.79
2.31
1.94
1.17
2.31
1.46
1.46
2.20
1.37
1.88
* "special study" (short duration)
** run terminated before 2.5 m (100 in.) total head loss
-------
44% higher loading than filter N with virtually no loss of effluent quality.
The poor removal efficiencies listed in Table 2 for this period are believed
to result from the nature of the influent suspended solids. A review of the
SVI data for the completely mixed activated sludge system during this time
reveals values between 600 and 800 ml/gm. This indicates either a poorly
formed floe or presence of filamentous type organisms. In either case, it
appears that general filterability was less than desirable.
Filter P, containing only anthracite, performed comparatively well, yielding
a 55 percent removal efficiency and a loading of 0.753 kg/m2/m head loss
(.047 Ibs/ft2/ft head loss).
In November 1974 the media in filter 0 was replaced with 64 cm (25 in ) of
anthracite having an effective size of 2.0 mm. As would be expected, filter
N produced the highest quality effluent, followed by filters P and 0.
Unfortunately, extremely low influent suspended solids concentrations did not
allow effective evaluation of filter performance. Since the atypical influent
characteristics resulted in a very slow build up of total head loss, operation
of all three filters was curtailed when filter N head loss reached 2.5m
(100 in ). Thus, loadings are not calculated for filters 0 and P during
November.
In December, filter P media was replaced with anthracite of effective size
2.0 mm under which was placed silica sand of effective size 0.9 mm. Studies
were conducted using the completely mixed system effluent at hydraulic
loadings of 0.12, 0.20, 0.29 and 0.37 m3/min/m2 (3, 5, 7 and 9 gpm/ft2), and
with the step aeration system effluent at 0.20 and 0.29 m-Vmin/m2 (5 and 7
gpm/ft2). The performance data (percent removal, solids loadings, and net
water production) are given in Table 2.
The filter P media appeared to offer considerable advantage over the media
in filters N and 0. In general, the quality of the effluent was similar to
that from the 1.3 mm E.S. coal/0.65 mm E.S. sand combination, with run lengths
being 27 to 55 percent higher. Consequently, the total solids capture per
run was significantly greater. The removal efficiency of filter 0 (2.0 mm
effective size anthracite) was reasonably good at low flowrates, but deter-
iorated rapidly as the flowrate was increased (see discussion of flowrate
effects in Section 7). Figure 2 shows the increase in total head loss as a
function of solids captured for the three filters, operated at 0.20 m3/min/m2
(5 gpm/ft2) with the effluent from the completely mixed activated sludge
system.
Figure 3 reveals the head loss distribution through the media during the
0.20 m3/min/m2 (5 gpm/ft2) run with the completely mixed process effluent.
During the months of November and December, it was noted that a portion of
the anthracite had been lost from filter N at some time during the backwash
operation. This accounts for the relative vertical position of the filter N
head loss curve. From inspection of the three curves, it is apparent that
the greatest head loss drop occurs across the top few centimeters of the
media. In comparing the two dual media filters, N and P, it appears that
filter N demonstrates a greater proportion of head loss drop across the
surface. It is likely that this condition is due to a greater accumulation
13
-------
3.0<
2.5
(0
02.0
UJ
X
1.5
1.0
0.5
FEED: COMPLETE MIX ACTIVATED SLUDGE EFFLUENT (SRT = 8.1 DAYS)
Q = 0.20 m3/min/m2
T = 19.0°C.
1.3mm E.S. COAL
0.65mm E.S. SAND
AVG. % REMOVAL = 91
2.0 mm E.S. COAL
0.9mm E.S. SAND
AVG. % REMOVAL = 92
""—>
P
2.0mm E.S. COAL
AVG. % REMOVAL = 73
0.5 1.0
1.5 2.0 2.5 3.0
S.S. CAPTURED, kg/m2
3.5 4.0
4.5
5.0
Figure 2. Head Loss as a Function of Solids Capture, December 1974
-------
TOP 14,
FILTER P
ANTHRACITE
SAND
FEED: COMPLETE MIX ACTIVATED SLUDGE
EFFLUENT (SRT = 8.1 DAYS)
Q = 0.20 m3/min/m2
T = 19.0°C.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
INCREMENTAL HEAD LOSS BETWEEN TAPS, METERS
Figure 3. Differential Head Loss Characteristics, December 1974
-------
of suspended solids at the surface of filter N, as the effective size of the
upper media is 1.3 mm for N vs. 2.0 mm for P. In addition, filter N shows a
sharper, more defined head loss drop across the sand-coal interface (between
taps 6 and 7), which would indicate less penetration of the suspended solids
into the sand. This becomes more evident during the January 1975 study,
discussed below.
It was concluded from previous work that the 2.0 mm effective-sized anthra-
cite in filter 0 was unsuitable for filtration of secondary effluents.
Consequently, in January 1975, 38 cm (15 in ) of sand having an effective
size of 0.65 mm was placed beneath the existing anthracite. In addition,
since filter N had experienced some loss of anthracite, new material of the
same effective size was placed to a depth of 64 cm (25 in ). The completely
mixed activated sludge system effluent was used to feed the filters at
hydraulic loadings of 0.12, 0.20, 0.29, and 0.37 m3/min/m2 (3, 5, 7, and
9 gpm/f t2) .
Figure 4 demonstrates the head loss vs. suspended solids captured for filters
N, 0, and P at a hydraulic loading of 0.20 m3/min/m2 (5 gpm/ft2). Filter P
absorbed a loading 72 percent higher than that for filter N and 29 percent
higher than the filter 0 loading. At this same flowrate, the removal
efficiencies are 89, 88, and 89 percent for filters N, 0, and P respectively.
For filter P this translates into run lengths that are 72 and 32 percent
greater than those for filters N and 0, with virtually no deterioration of
effluent quality. This same observation holds generally true at the higher
flowrates, although a slight reduction in effluent quality is realized.
Figure 5 shows the incremental head loss between taps for filters N, 0 and P
at the hydraulic loading of 0.20 m^/min/m2 (5 gpm/f t2) . Here, the differences
in head loss distribution are quite pronounced. Note that virtually all of
the head loss for filter N occurs across the media surface, while for filters
0 and P, considerable head loss occurs throughout the bulk of the media. This
indicates a greater "mat" accumulation of solids on the filter N media
surface, with a deeper solids penetration in the other filters. Figure 5
also shows a greater head loss distribution at the sand-coal interface for
filter 0, with a relatively small drop across the top. This is most likely
due to the greater degree of media intermixing at the interface, which results
in a more gradual head loss distribution through the media. Since the
difference in media effective sizes is quite large (2.0 mm vs. 0.65 mm), the
head loss drop at this interface is significant.
The relationship between the effective sizes of anthracite and sand have
often been cited as playing an important role in the performance of dual media
filters (2, 3, 4, 6), since the degree of media intermixing is largely
dependent upon this factor. In general, a sharp sand-coal interface is not
desirable, since a solids mat may be formed at this boundary. The ratio of
the 90 percent finer coal size to the 10 percent finer sand size (Dgo) coal/
(DIQ) sand, is often used as a parameter to predict the degree of media
intermixing, a ratio of 4 generally resulting in substantial intermixing, and
a ratio of 2 to 2.5 producing a sharp interface (2).
16
-------
FEED: COMPLETE MIX ACTIVATED SLUDGE EFFLUENT (SRT = 6-7 DAYS)
Q = 0.20 m3/min/m2
3.0-
-2.5-
(0
o
_J
o
2.0
<
I1'
1.0-
0.5-
T = 17.0°C.
2.0mm E.S. COAL
0.65 mm E.S. SAND
AVG. % REMOVAL = 88
1.3mm E.S. COAL
0.65mm E.S.
AVG. % REMOVAL = 89
J3
2.0mm E.S. COAL
0.9mm E.S. SAND
AVG. % REMOVAL = 89
0.5 1.0 1.5 2.0 2.5 3.0 3.5
S.S. CAPTURED, kg/m2
4.0
4.5
5.0
Figure 4. Head Loss as a Function of Solids Capture, January 1975
-------
TOP 14
00
FEED: COMPLETE MIX ACTIVATED SLUDGE EFFLUENT
(SRT = 6-7 DAYS)
Q = 0.20 m3/min/m2
T = 17.0°C.
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6
INCREMENTAL HEAD LOSS BETWEEN TAPS, METERS
Figure 5. Differential Head Loss Characteristics, January 1975
1.7 1.8
-------
Table 3 lists the "Intermix quotients" for the media combinations used in this
s tudy.
TABLE 3
INTERMIX QUOTIENTS FOR DUAL MEDIA CONFIGURATIONS
Effective size of (D9Q^ coal Effective size of (Pop) coal
coal, mm (D^g) mm sand, mm (D-i/0
1.3 2.75 0.65 4.2
1.65 3.0 0.8 3.8
2.0 4.2 0.9 4.6
2.0 4.2 0.65 6.4
The first three combinations listed above are believed to be within the
working range of values required to achieve a reasonable degree of inter-
mixing. Even the relatively high ratio of 4.6 provided desirable run
length and effluent quality characteristics. However, the 6.4 value resulted
in excessive media intermixing. Attributable to this condition were short
run lengths, similar to those for the 1.3 mm coal/0.65 mm sand combination,
and an effluent of poorer quality than that from the 2.0 mm coal/O.9 sand
configuration, the latter of which produced run lengths at least 32 percent
greater. Thus, there appeared to be no benefit from further investigation
of a 2.0 mm coal/0.65 mm sand media for filtration of secondary effluents.
19
-------
SECTION 7
EFFECT OF FLOWRATE ON FILTER PERFORMANCE
During the months of December 1974 and January 1975, the filters were
operated at a variety of flowrates, ranging from 0.12 to 0.37 m^/min/m2
(3 to 9 gpm/ft2). The parameters of particular Importance in assess-
ment of flow rate effects were effluent quality (possibility of
suspended solids breakthrough at elevated flowrates), solids loading (a
function of removal efficiency and run length), and net water production.
The data for December 1974 are plotted in Figures 6, 7, 8, and 9. At
this time, the media effective sizes were as follows: N—1.3 mm
coal/0.65 mm sand, 0—2.0 mm coal only, P—2.0 mm coal, 0.9 mm sand.
Note that even at the lowest hydraulic loading of 0.12 m^/min/m2
(3 gpm/ft ) filter 0 experienced a suspended solids breakthrough, as
shown by the drop in removal efficiency. At the higher flowrates, this
breakthrough becomes more evident.
With the addition of 38 cm (15 in) of 0.9 mm effective sized sand to
the 2.0 mm effective sized anthracite (filter P), the susceptibility of
the media to solids breakthrough was greatly reduced. Only at the 0.28
and 0.37 m^/min/m2 (7 and 9 gpm/ft2) loading did a reduction in removal
efficiency occur. In addition, it is questionable whether the perform-
ance of the 1.3 mm/0.65 mm E.S. media in filter N would be significantly
better than filter P, since the removal efficiencies appear quite similar.
During November 1974, studies were conducted at elevated flowrates to
determine the breakthrough characteristics of the single anthracite media
contained in filters 0 (E.S. =2.0 mm) and P (E.S. = 1.65 mm). Figures
10 and 11 prove conclusively that such media are unable to sustain an
acceptable removal efficiency at high flowrates of 0.28 and 0.37
m3/min/m2 (7 and 9 gpm/ft2).
20
-------
100
90
80
lit
tc.
3?
70
60
50
^—-s
A
1.3mm E.S. COAL
0.65mm E.S. SAND
12/16-12/18/74
T = 18.5°C.
^_ ..A
Q— o
— e-o— o
^2.0mm E.S. COAL
0.90mm E.S. SAND
2.0mm E.S. COAL
B-Q
\/
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60
TIME (hours)
3 2
Figure 6. Removal Efficiency at 0.12 m /min/m
Feed: Complete H±x Activated Sludge (C.M.A.S.) Effluent
-------
ts>
100
90
80
70
< 60
§ 50
ui
30
20
10
1.3mm E.S. COAL
0.65mm E.S. SAND
12/9-12/10/74
T = 19.0°C.
r~\
\
2.0mm E.S. COAL
0.9mm E.S. SAND
2.0mm COAL
6 8 10 12 14 16 18 20 22 24 26 28 30
TIME (hours)
3 2
Figure 7. Removal Efficiency at 0.20 m /min/m
Feed: C.M.A.S. Effluent
-------
N)
LO
12/10-12/11/74
T = 19.0°C.
1.3mm E.S. COAL
0.65mm E.S. SAND
2.0mm E.S. COAL
E.S. SAND
\ 2.0mm E.S. COAL
\
\
\
\
0 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
TIME (hours)
3 2
Figure 8. Removal Efficiency at 0.29 m /min/m
Feed: C.M.A.S. Effluent
-------
to
100
90
80
70-
< 60
O
ui
*
so
*'
30
20
10
12/12-12/13/74
T = 19.0°C.
1.3mm E.S. COAL
0.65mm E.S. SAND
2.0mm E.S. COAL
0.9mm E.S. SAND
2.0mm E.S. COAL
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
TIME (hours)
3 2
Figure 9. Removal Efficiency at 0.29 m /min/m
Feed: C.M.A.S. Effluent
-------
100i
80g Q
70-
60
>
s
UJ
cc
V) 40
30
20
10
1.3mm E.S. COAL
0.65mm E.S. SAND
.65mm E.S. COAL
\ 2.0mm E.S
COAL
\ X
11/20-11/21/74
T = 21.0°C.
345
TIME (hours)
8
3 2
Figure 10. Removal Efficiency at 0.29 m /min/m
25
-------
100
1.3mm E.S. COAL
0.65mm E.S. SAND
11/19-11/20/74
T = 21.0°C
<
>
o
60
UJ
C50
CO
CO
40
30
V
'^. _-•---A
_^
\
\
2.0mm E.S. COAL
'
\\
v,
\\
1.65mm E.S. COAL
20
D
10-
345
TIME (hours)
8
3 2
Figure 11. Removal Efficiency at 0.37 m /min/m
26
-------
SECTION 8
EFFECT OF PROCESS EFFLUENT ON FILTER PERFORMANCE
One of the most important and most elusive parameters affecting filter
performance is the nature of the influent suspended solids, as characterized
by concentration, floe size distribution, floe strength, and floe charge (4).
For a typical domestic waste, these characteristics should be dependent on
the type of treatment process preceeding filtration, the process operational
parameters (solids retention time or F/M), and the overall process per-
formance, perhaps the most critical factor of all. The filtration work at
the EPA-DC Pilot Plant substantiated the importance of process performance
in assessing filterability of activated sludge effluents.
Two factors related to process performance made it difficult to establish a
relationship between filterability and either process type or SRT. The first
was that the performance of the activated sludge systems was such that the
concentrations of suspended solids in the effluent were often between 5 and
10 ppm. This made effective evaluation of filter performance extremely
difficult, since considerable error is introduced in determining filter
removal efficiency. Process effluents of this nature do not require filtration.
In some cases, the low suspended solids values were due to excellent process
performance and a low sludge volume index (SVI); during other times, Norcardia
proliferation in the system formed a scum on the clarifier surface, which
acted as a filtering mat prior to discharge of the effluent over the weirs.
The presence of the Norcardia mat also caused variations in the effluent solids
due to sloughing of the scum into the effluent trough.
To compare the filterability of process effluents, the data from filter N are
presented, since the effective sizes of the media remained the same through-
out the study (1.3 mm coal/0.65 mm sand). Investigation of process type and
SRT requires that the other variables be relatively constant, i.e. flowrate,
influent suspended solids concentration, and SVI. SVI is used here merely as
an indicator of process stability.
Table 4 indicates the performance of the filters in terms of loading and
percent removal. Also given are the influent and effluent suspended solids
concentrations, the flowrate, the type of process and SRT, and the SVI.
Without a complete characterization of the effluent suspended matter, the SVI
indicates little with respect to filterability. A low SVI would generally
indicate a well-formed floe which could be easily captured during filtration.
However, a high SVI could be indicative of either a poorly flocculated
27
-------
-TABLE 4
SUMMARY OF FILTER N DATA
Media E.S.: 1.3 mm coal/0.65 mm sand
00
ROCESS/SRT
Step/4.5
Step/5.9
Step/5
CM/8.1
CM/ 6-7
Step/3.7
Plug/4.7
CM/4. 7
CM/8.1
Step/3.5
CM/6-7
CM/3.0
DATES
7/1-7/31/74
8/1-8/27/74
9/1-9/19/74
12/16-12/18/74
1/20-1/23/75
11/5/74
11/6/74
11/7/74
12/9-12/10/74
12/18-12/20/74
1/3-1/5/75
3/16-3/26/75
FLOW
J/min/m2
0.12
0.12
0.12
0.12
0.12
0.20
0.20
0.20
0.20
0.20
0.20
0.20
SS in
ppm
12.5
6.7
8.2
21.0
16.3
5.6
6.3
42.5
16.0
13.2
8.8
17.6
SS out
ppm
2.6
2.2
2.9
0.5
2.1
2.1
3.7
1.1
1.4
2.9
0.9
6.5
% REMOVAL
79
67
65
97
87
62
41
97
91
77
89
63
LOADING
.kg/m^/m head loss
0.90
0.79
0.72
2.26
2.21
0.32
0.38
1.19
1.19
1.01
1.35
1.38
SVI
ml/gi
145
141
124
216
306
100
291
374
256
163
247
79
-------
effluent which would pass the filters, or a filamentous ridden
effluent which could rapidly plug the media surface. Although SVI
is a useful parameter in assessing activated sludge process perfor-
mance, it is a poor measure for characterizing the effluent suspended
solids with respect to filterability.
Correlation of SRT and/or type of activated sludge modification with
filter performance could not be established with the limited and
variable data available from this study.
29
-------
SECTION 9
FILTER BACKWASHING
The initial design for the backwash operation of the filters in March 1974
consisted of (1) a 0.12 m3/min/m2 (3 gpm/ft2) surface wash and 0.61 m3/min/
m2 (15 gpm/ft2) low flow backwash for two minutes, (2) a 1.14 m3/min/m2
(28 gpm/ft2) high flow backwash for ten minutes, and (3) a 0.61 m3/min/m2
(15 gpm/ft2) low flow backwash for one minute. However, during this month, it
was observed that 2.5-5.0 cm (1-2 in) diameter "mud-balls" formed in the media
which could not be removed by this backwash sequence. As a result, an air
scour system was installed to dissipate the scum, which was thought to result
from Norcardia proliferation in the activated sludge systems. The backwash
sequence was then modified to read as follows:
(1) 1 min. of air scour at 1.5 m3/min/m2 (5 scfm/ft ).
(2) partial filter drainage to top of media.
3 2
(3) 1 min. air scour with 2 min. low flow, 0.4 m /min/m
(10 gpm/ft2) backwash.
10 min. o:
b ackwash.
(4) 10 min. of high flow, 1.2 m3/min/m2 (30 gpm/ft2)
The installation of the air wash system greatly reduced "mud ball" formation
and resulted in a well scoured media after backwash. However, several
incidents of mud ball formation were encountered which required addition of
chlorine to the filters. This had the effect of dissipating biological scum
formation in the media and on the filter walls.
During Janurary 1975, a series of studies was initiated to investigate the
backwash characteristics of the filter media. The effective sizes of the
media were as follows: N— 1.3 mm coal/0.65 mm sand, 0— 2.0 mm coal/0.65 mm
sand, P— 2.0 mm coal/0.9 mm sand. The parameters of interest were (1) back-
wash rates required for bed fluidization, and (2) extent of particle removal
during backwash.
Figures 12 and 13 demonstrate percent bed expansion as a function of flowrate.
Unfortunately, bed expansion studies were conducted only on filters N and P,
since a previous rupture of the filter 0 walls necessitated replacement with
opaque PVC pipe, making observation of the contents impossible.
The filter N media was able to achieve a high degree of expansion at
relatively low flowrates. At a backwash rate of 1.22 m3/min/m2 (30 gpm/f tz),
30
-------
60i
50
40
Q.
X
iy
030^
m
20
10
1/28/75
TB/W = 15.0°C.
FILTER N ANTHRACITE;
E.S. = 1.3 mm
FILTER N SAND;
E.S. = 0.65mm
FILTER P SAND
E.S. = 0.9mm
FILTER P ANTHRACITE
.S. = 2.0mm
0.2 0.4
0.6 0.8 1.0 1.2 1.4
BACKWASH FLOWRATE, m3/min/m2
1.6
1.8 2.0
Figure 12. Extent of Sand, Anthracite Bed Expansion during Backwash
-------
w
NS
o
V)
60
50
40
X 30
LU
O
Ul
CO
S? 20
10
1/28/75
TB/W = 15.0°C.
0.2
FILTER N
COAL E.S. = 1.3mm
SAND E.S. = .65mm
FILTER P
COAL E.S. = 2.0mm
SAND E.S. = 0.9mm
0.6 0.8 1.0 1.2 1.4
BACKWASH FLOWRATE, m3/mjn/m2
1.6
1.8
2.0
Figure 13. Extent of Total Bed Expansion during Backwash
-------
the percent total bed expansions for filters N and P were 34 and
10 percent respectively. At 1.43 m3/min/m2 (35 gpm/ft2) these values
became 43 and 14 percent. Optimum bed expansion values have not been
determined for dual and multimedia filters (5). In general, a 10-20
percent expansion is recommended to release the entrapped particles,
particularly at the sand-coal interface (6). For filter P, the rela-
tively high degree of mixing at this interface may have had a beneficial
effect on backwash efficiency, since less localized plugging at the
interface occurred. In general, few situations were encountered at this
relatively low bed expansion which prevented removal of solids during
the backwash operation. However, some plugging of the air scour system
occurred which often required an increase in the air flowrace to purge
the sparger.
To determine the required duration of backwash to effectively clean
the media, and to perform a mass balance on the suspended solids through
the filters, backwash studies were conducted on all three filters. These
consisted of operating the filters to 2.5m (100 in) head loss and back-
washing the filters on the normal cycle. During the first three minutes
of backwash; the backwash water was sampled every 30 seconds, after which
time samples were collected every minute and analyzed for suspended
solids. The concentration of suspended solids in the filter backwash
water vs. time is presented in Figures 14, 15 and 16.
During the first three minutes of backwash at 1.22 m3/min/m2, approximately
98 percent of the solids were removed from the media, and after five
minutes, the media was completely clean. An attempt to account for the
suspended solids entrapped during filtration by examination of the back-
wash curves was largely successful. Theoretically, the mass of solids
captured in the media during the filtration cycle should be equal to the
mass of solids removed during the backwash cycle. By comparing the area
under the backwash curves with calculations from influent and effluent
suspended solids concentrations, flowrate, and run length, it was found
that the values were within 10 percent agreement, the value from the back-
wash curve in all cases the larger (see Figures 14, 15 and 16).
To effectively cleanse the media in filter P, a backwash rate of
1.43 m3/min/m2 (35 gpm/ft2) for approximately four minutes was required.
This represents a total water usage of 5.7 nr -per m2 of surface area
(140 gal/ft ). Normally, 3 to 4 m3 per m2 (75-100 gal/ft2) is required
for the smaller media sizes. However, the additional backwash water
requirement for the larger media is negligible, since the frequency of
backwashing is reduced considerably .
33
-------
3500Q
(O
(0
2/4/75
QB/W = 1-22 m3/min/m2
TERMINAL HEAD LOSS = 2.6 m
TB/W = 11-0°C.
TOTAL AREA«3037 HHL-min = 0.34 kg SUSPENDED SOLIDS
1 REMOVED DURING BACKWASH
AMOUNT S.S. CAPTURED DURING RUN = 0.32 kg.
(CALCULATED FROM INFLUENT AND EFFLUENT S.S. DATA)
456
TIME (min)
8
10
Figure 14. Removal of Entrapped Solids during Filter N Backwash
-------
u>
2/5/75
QB/W = 1-22 m3/min/m2
TERMINAL HEAD LOSS = 2.56m
TB/W = n.o°c.
TOTAL AREA = 4030 1112-
lin = 0.46kg. SUSPENDED
SOLIDS REMOVED
DURING BACKWASH
AMOUNT S.S. CAPTURED DURING RUN = 0.43 kg
(CALCULATED FROM INFLUENT AND EFFLUENT S.S. DATA)
23456
TIME (min)
8
10
Figure 15. Removal of Entrapped Solids during Filter 0 Backwash
-------
6000
-------
SECTION 10
PHOSPHORUS REMOVAL
In April of 1975, a limited scale study was undertaken to investigate the
removal of phosphorus during downflow granular filtration.
From previous filtration work at the EPA-DC Pilot Plant (Jan. 1972-Sept. 1973)
on a 3-stage activated sludge system with alum addition, it was found that
higher phosphorus removals were effected by dual media filtration than by
0.45 ju Millipore filtration. This somewhat perplexing observation led to an
attempt to duplicate the phenomenon with the dual media filters. The effluent
from the single stage nitrification-denitrification system was chosen as a
filter influent, since it had exhibited biological phosphorus uptake. If the
filter removal mechanism was, in fact, biological, then the necessary micro-
organisms would be present.
The nitrification-denitrification effluent was fed to the filters at a flow-
rate of 0.20 m3/min/m2 (5 gpm/ft2). The data are presented in Table 5.
From inspection of the data, it can be seen that the extent of phosphate
removal was virtually negligible, at no time exceeding 7 percent. Since the
influent phosphorus is almost completely in the "soluble" form, it appears
that any phosphate reduction is due to removal associated with suspended
solids. It is apparent that the unusual uptake mechanism experienced with
the 3-stage system was not in evidence here. The reason for the failure to
duplicate the results of the 1972-1973 pilot study is believed to be due to
the absence of alum addition prior to filtration. It may be possible that
the particle size distribution of the aluminum-phosphate floe is such that
particles in the colloidal size range may pass through the 0.45 ju Millipore
filter while still being removed during media filtration. Also, continual
operation of the filters may result in an alum-enriched media which may effect
a relatively high phosphorus removal.
37
-------
TABLE 5
PHOSPHORUS REMOVAL STUDY .
FILTERS N and P
DATE FILTER
» 4/3/75 N
P
4/7/75 N
P
4/14/75 N
P
HEAD LOSS
AT TIME OF
SAMPLING, m
0.87
0.47
2.54
0.91
0.33
0.97
TOTA:
11.9
11.5
10.9
12.2
12.1
11.7
INFLUENT
FILTRABLE
10.3
14.1
ppm
EFFLUENT
TOTAL FILTRABLE
11.4
11.0
10.8
11.3
SOLIDS, ppm
14.3
12.9
10.1
13.2
INFLUENT
TOTAL SUSP.
292 5.4
243 4.8
257
258
270
263
4.4
4.6
EFFLUENT
TOTAL SUSP.
278 1.4
289 0.7
260
265
280
263
1.75
3.05
-------
SECTION 11
ENGINEERING SIGNIFICANCE
Among the most important considerations in design and operation
of dual media filters are 1) maintenance of an acceptable level
of effluent quality, 2) flowrate, and 3) rate of head loss
development (backwash frequency). Selection of media type and
size plays an important role in determining the interrelation-
ships between the above parameters.
This study investigated the use of filter media of greater effective
size (E.S.) than those used in conventional tertiary wastewater
filters. It was found that such media was able to sustain an
acceptable effluent quality with a reduced frequency of backwashing,
thereby reducing operational requirements while increasing the
volume of wastewater filtered per day.
A useful parameter for economic comparison of filtration systems is
net water production, which is defined as the difference between
the total daily volume of filtrate produced while the filter is in
service and the total daily volume of water used for backwash. Net
water production values are presented in Table 6 for the systems
described in this report.
Figure 13 was used to estimate backwash flowrates required to achieve
the desired level of bed expansion for each media size. The conven-
tional media jCL.3 mm-E.S. coal/0.65 mm E.S. sand) required a backwash
rate of 0.8 m /min/m (20 gpm/ft ) to achieve a 15% bed expansion,
while the coarse media (2.0 mm-E.S. coal/0.9 mm E.S. sand) required
about 1.5 m /min/m (37 gpm/ft ) to attain the same level of expansion.
The intermediate size employed at the beginning of the study
(1.65 mm E.S. coal/0.79 mm-E.S. sand) was estimated to require
approximately 1.1 m /min/m (27 gpm/ft ) for backwash. A sample
calculation of net water production is presented on the following page.
39
-------
SAMPLE CALCULATION OF
NET WATER PRODUCTION (NWP)
Assumptions:
1. 15% bed expansion during backwash
2. Duration of backwash = 10 min
3. Filter out of service 30 min per backwash
Backwash rate required Total water used per
Anthracite size* for 15% bed expansion backwash
_ mm _ m3/min/m2 (gpm/ft2) m3/m2 (gal/ft2)
1.3 0.8 (20) 8.0 (200)
1.65 1.1 (27) 11.0 (270)
2.0 1.5 (37) 15.0 (370)
NWP = (Total filtrate produced per day while filter is in service)
- (Total backwash water used per day)
Example: Filter P: 2.0 mm E.S. coal/0.9 mm E.S. sand
Q = 0.2 m3/min/m2 (5 gpm/ft2)
Run length befone backwash = 16 hr
NWP
|l440 min/day - 2f ^ „ — ^ (30 min/BW)| 0.2 m3/min/m2
l J.D nr to Jsw I
"(ley
15*° m/m Per BW
= 257 m3/day/m2
* The anthracite size in the dual media filters
governed the backwash rate necessary to achieve
the desired level of bed expansion (see Figure 12).
40
-------
TABLE 6
NET WATER PRODUCTION (NWP)
Month
Sept.
'74
Oct.
Nov.
Dec.
Filter
N
0
P
N
0
P
N
N
P
N
0
P
N
P
N
Media E.S. (mm)
Coal /Sand
1.35/0.65
1.65/0.79
1.65/
1.3/0.65
2.0/
1.65/
1.3/0.65
1.3/0.65
2.0/0.9
1.3/0.65
2.0/
2.0/0.9
1.3/0.65
2.0/0.9
1.3/0.65
Flow
0.16
0.16
0.16
0.16
0.16
0.16
0.20
0.20
0.20
0.29
0.37
0.12
0.12
0.20
0.20
0.20
0.29
0.29
0.37
Run length
hr
15
21
25
10
17
23
18
24
7
8
6
37
56
17
19
24
11
14
9
Water req'd per
backwash, m-^
8
11
11
8
15
11
8
8
8
8
8
8
15
8
15
15
8
15
8
NWP
m^ /day/in^
210
212
215
200
202
214
270
274
240
368
456
165
165
268
270
267
381
377
482
-------
Month
Dec.
'74
Jan.
'75
to
Mar.
Filter
P
N
P
N
P
N
0
P
N
0
P
N
0
P
N
0
P
N
P
Media E.S. (mm)
Coal/ Sand
2.0/0.9
1.3/0.65
2.0/0.9
1.3/0.65
2.0/0.9
1.3/0.65
2.0/0.65
2.0/0.9
1.3/0.65
2.0/0.65
2.0/0.9
1.3/0.65
2.0/0.65
2.0/0.9
1.3/0.65
2.0/0.65
2.0/0.9
1.3/0.65
1.3/0.9
Flow
m-* /min/m
0,37
0.20
0.20
0.29
0.29
0.12
0.12
0.12
0.20
0.20
0.20
0.29
0.29
0.29
0.37
0.37
0.37
0.20
0.20
TABLE 6 (con't)
NET WATER PRODUCTION (NWP)
Run length
hr
14
20
35
9
17
54
51
79
36
47
62
12
10
16
8
7
15
24
36
Water req'd per
backwash, mr
15
8
15
8
15
8
15
15
8
15
15
8
15
15
8
15
15
8
15
NWP
m3/day/m2
489
271
274
373
384
168
164
167
279
277
280
384
361
382
476
444
491
274
274
-------
Inspection of Table 6 reveals little difference in net water
production between the media combinations studied at any particular
flowrate. The largest recorded difference between net water
production from the coarse and fine dual media filters amounted to
approximately 3.2%. The diminutive magnitude of this difference is
due to the fact that the conservation of backwash water from less
frequent backwash of the coarser media is offset by the greater
quantity of water utilized per backwash to achieve the same degree
of expansion. If the values of net water production for filter P
(2.0 mm E.S. coal/0.9 mm E.S. sand) are recalculated using the same
water requirement per backwash as filter N, the net water production
from the two filters still differs by less than 6%. In general,
backwash water requirements using the data from Table 6 vary from
A to 10% of total filter throughput.
Despite the similar net water production from both conventional and
coarse dual-media filters, the use of larger-sized anthracite and
sand may offer an economic advantage due to the reduction in backwash
frequency. Although little savings in backwash water requirements
are realized, the total man-hours necessary for backwash operations
may be reduced substantially. In addition, the use of larger media may
make it possible to operate at higher flow rates with an acceptable
backwash frequency. Obviously this will reduce capital costs.
During this study, the dual media filter consisting of 2.0 mm E.S.
coal over 0.9 mm E.S. sand produced a high quality effluent under a
variety of hydraulic and solids loading conditions. However, several
areas are in need of further investigation. Of particular interest
is the performance of such a filter under the combined conditions of
hydraulic surge and high suspended solids loadings (> 30 mg/1). In
addition, research should be conducted on the amenability of coarse,
dual-media filters for treatment of effluents from activated sludge
processes employing high solids retention times, as in extended
aeration processes.
From the research described in this report, it appears that dual-media
filters employing media of greater effective size than conventional
filter media may have application for filtration of effluents from
secondary wastewater treatment facilities. Such filters appear capable
of sustaining an effluent of acceptable quality while reducing the fre-
quency of backwash operations.
43
-------
REFERENCES
1. Baumann, E.R., and J.Y.C. Huang, "Granular Filters for Tertiary Wastewater
Treatment." JWPCF, 46: 1958 (1974).
2. "Wastewater Filtration-Design Considerations," EPA Technology Transfer
Seminar Publication, July 1974.
3. Cleasby, J.L., and C.F. Woods, "Intermixing of Dual Media and Multimedia
Granular Filters," JAWWA, April, 1975, p. 197.
4. Metcalf and Eddy, Inc., Wastewater Engineering, McGraw-Hill, New York,
1972.
5. Weber, W.J., Jr., Physicochemical Process for Water Quality Control,
Wiley-Interscience, New York, 1972.
6. Young, J.C., "Filtration of Effluents from Biological Treatment Plants,"
General Filter Co., Ames, Iowa.
44
-------
APPENDIX
45
-------
2.5
(A
O
_l
< 1.5
UJ
1.0
0.5
ORIGINAL DESIGN FOR
FILTER MEDIA EFFECTIVE SIZES:
ANTHRACITE: 1.3mm
SILICA SAND: 0.65mm
6/5/74
Q = 0.16 m3/min/m2
FEED: COMPLETE MIX ACTIVATED
SLUDGE EFFLUENT
8 10
TIME (hrs)
12
14
16
18
20
Figure A-l. Differences in Head Loss Characteristics
Caused by Inconsistencies in Filter Media
-------
TABLE A-l
SUMMARY OF FILTER PERFORMANCE DURING PERIOD OF INCONSISTENT FILTER MEDIA
Month
July
1974
Aug.
Sept.
Media E.S. (mm) Flow
Filter Coal/Sand m3/min/m2
N
0
P
N
0
P
N*
0*
P*
N*
0*
P*
N
0
P
1.3/0.65
1.3/0.65
* 1.2/0.65
1.3/0.65
1.3/0.65
* 1.2/0.65
1.3/0.65
1.3/0.65
* 1.2/0.65
1.3/0.65
1.3/0.65
* 1.2/0.65
1.3/0.65
1.3/0.65
•x. 1.2/0.65
0.12
0.12
0.12
0.12
0.12
0.12
0.16
0.16
0.16
0.24
0.24
0.24
0.12
0.12
0.12
Avg.
Process Effluent SS in
Type/SRT (days) ppm
STEP/ 4-5 12.5
PLUG/4-5 11.4
CM/2. 3
STEP/5.9
PLUG/ 2. 9
CM/2.1
CM/2.1
CM/2.1
CM/2.1
CM/2.1
CM/2.1
CM/2.1
STEP/5.0
PLUG/ 2. 5
CM/2.0
11.2
6.7
8.6
7.1
20.4
19.7
17.7
16.0
16.2
21.0
8.2
5.4
11.0
Avg.
SS out
2.6
4.6
2.3
2.2
3.5
1.8
2.3
2.6
2.0
2.2
1.6
2.2
2.9
2.8
2.8
Avg.
Rem.
79
60
79
67
59
74
89
87
90
86
90
89
65
48
75
Avg.
Loading
Kg/m2/m
head loss
0.90
0.90
0.77
0.79
0.75
0.80
0.67
0.85
0.88
0.53
0.66
0.77
0.72
0.72
0.85
"special study" (short duration)
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-77-144
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
DOWNFLOW GRANULAR FILTRATION
OF
ACTIVATED SLUDGE EFFLUENTS
5. REPORT DATE
September 1977 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Robert P.G. Bowker
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Government of the District of Columbia
Department of Environmental Services
EPA-DC Pilot Plant
5000 Overlook Avenue S.W. Washington, D.C.
10. PROGRAM ELEMENT NO.
1BB043
20032
11. CONTRACT/CHANT-NOr
68-03-0349
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Cin., OH
Office of Research and Development
U.S. Environmental Protection Aeencv
Cincinnati,'Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Irwin J. Kugelman (513-684-7631)
16. ABSTRACT
The performance of downflow granular filters subjected to effluents from activated
sludge processes was investigated at the EPA-DC Pilot Plant in Washington, D.C. The
0.1 m2 (1 ft2) filters were operated at hydraulic loadings from 0.12 to 0.37 m^/min/
(3 to 9 gpm/ft2). Several media combinations were investigated, including both single
anthracite and dual anthracite-sand configurations. Effluents from step aeration,
plug flow, and completely mixed activated sludge systems were used as feeds.
Breakthrough of the suspended solids into the effluent occurred with both the 1.65 mm
and 2.0 mm effective size (E.S.) single anthracite configurations, becoming more
evident at the higher flow rates.
A dual media filter, consisting of 2.0 mm E.S. coal over 0.9 mm E.S. sand, exhibited
the most desirable characteristics for filtration of the secondary effluents investi-
gated. The advantages were longer run times and higher suspended solids loadings
with virtually no deterioration of effluent quality.
A backwash study conducted at a variety of backwash flowrates showed that a 13 percent
bed fluidization was achieved with the coarse media (2.0 mm E.S. coal/0.9 mm E.S.
sand) at a flow rate of 1.43 m3/min/m2 (35 gpm/ft2). This was sufficient to
effectively cleanse the media.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Sewage Treatment
Filtration
Filter Media
Activated Sludge
Effluents
Suspended Solids
13B
IS. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)'
Unclassified
1. NO. OF
56
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
48
l!rU.S.GOVBINMENTI>1(INTmG OFFICE 1977-757-056/6523
------- |