United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
EPA-600/2-78-170
September 1978
Research and Development
£EPA
Two-Stage Granular
Activated Carbon
Treatment
-------
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-600/2-78-UO
September 1978
TWO-STAGE GRANULAR ACTIVATED
CARBON TREATMENT
by
Leon S. Directo
Ching-lin Chen
Robert P. Miele
Los Angeles County Sanitation District
Whittier, California 90607
Contract No. 14-12-150
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.
n
-------
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.
One method of tertiary or advanced treatment beyond the conventional
secondary treatment level is granular carbon adsorption. The treatment
process results in significant additional reduction in suspended solids and
organics. Original systems employed several stages of contact to provide
optimum utilization of carbon adsorption capacity. In this study an evalua-
tion of a two-stage rather than a multi-stage system was conducted to ascertain
the performance of this lower capital cost version of the process. In addition
a comparison of performance with two different granular carbon sizes was
conducted.
Francis T. Mayo, Director
Municipal Environmental
Research Laboratory
m
-------
ABSTRACT
Two 6.3 I/sec (0.15 mgd), two-stage, packed-bed, downflow granular
activated carbon pilot plants were operated continuously for 33
months using unfiltered and unchlorinated activated sludge plant
effluent. The main objective of the study was to evaluate the
effect of repeated thermal regeneration cycles on the adsorption
capacity, regeneration loss and pressure drop buildup of carbon
with different particle size. Performance data, collected during
the field'study, has demonstrated the stability of the two-stage
carbon adsorptive system in consistently producing effluents of
excellent overall quality.
The carbon capacity in the Filtrasorb 300 system (8x30 mesh carbon)
decreased about 25% after four adsorption cycles, resulting in an
apparent steady state capacity of 0.26 Ibs. DCOD removed/lb. carbon.
A 23% decrease in carbon capacity occurred after three adsorption
cycles in the Filtrasorb 400 system, (12x40 mesh carbon). The
400 system has about 13% more DCOD removal capacity than the 300
system. While the 400 system showed slightly higher carbon ca-
pacity than the 300 system, the latter has the advantage of not
only lower initial cost, but also lower in both pressure loss and
regeneration loss.
The estimated total treatment cost for a 0.44 nrVsec (10 mgd), two-
stage carbon adsorption system with 8x30 mesh carbon for treating
an activated sludge plant effluent is 11.520/1000 gallons.
This report was submitted by the Los Angeles County Sanitation Dis-
tricts, Los Angeles, California, in fulfillment of Contract Num-
ber 14-12-150 under the sponsorship of the Environmental Protection
Agency. This report covers the period March, 1970, to December, 1973,
and work was completed as of December, 1975.
-------
CONTENTS
Page
Foreword ill
Abstract ±v
List of Figures vi
List of Tables viii
Acknowledgements ix
Sections
I Introduction 1
II Conclusions 2
III Recommendations 4
IV Experimental Program 5
V Experimental Results 17
References 62
V
-------
FIGURES
No. Page
1 General Layout of the Carbon Adsorption System 6
2 Schematic Diagram of the Two-Stage Adsorption Systems. 7
3 Typical Carbon Contactor 8
4 Cross Sectional View of the Multiple Hearth Furnace .. n
5 Schematic Diagram of the Air Pollution Control System. 12
6 Contactor Identification Number 14
7 Virus Sampling Chamber 16
8 Original Backwash Schedule 18
9 Modified Backwash Schedule 19
10 Effect of Influent Turbidity and Carbon Particle Size
on Column Headless 21
11 Effect of Adsorption Cycle on Iodine No., Molasses No.
and Methylene Blue No. of Filtrasorb 300 29
12 Effect of Adsorption Cycle on Iodine No., Molasses No.
and Methylene Blue No. of Filtrasorb 400 30
13 Effect of Adsorption Cycle on Ash Buildup and Mean
Particle Diameter of Filtrasorb 300 31
14 Effect of Adsorption Cycle on Ash Buildup and Mean
Particle Diameter of Filtrasorb 400 32
15 COD Removal in the Filtrasorb 300 System (Run 1) 40
16 COD Removal in the Filtrasorb 300 System (Runs 2-3) .. 41
17 COD Removal in the Filtrasorb 300 System (Runs 4-5) .. 42
18 COD Removal in the Filtrasorb 300 System (Runs 6-8) .. 43
vi
-------
No. Page
19 COD Removal in the Filtrasorb 400 System (Runs 1-3) .. 45
20 COD Removal in the Filtrasorb 400 System (Runs 4-6) .. 46
21 Effect of Adsorption Cycle on Carbon Capacity 47
22 Effect of Adsorption Cycle on Carbon Dosage 49
23 Effect of Carbon Particle Size on TCOD Removal 51
24 Effect of Carbon Particle Size on DCQD Removal 52
vn
-------
TABLES
No. Page
1 Virgin Carbon Characteristics 9
2 Correlation of Headless with influent Turbidity 22
3 Headless Buildup and Turbidity Removal in the Filtrasorb
300 2-Stage Adsorption System 23
4 Headless Buildup and Turbidity Removal in the Filtrasorb
400 2-Stage Adsorption System 24
5 Effect of Regeneration on the Properties of Filtrasorb 300
Carbon 27
6 Effect of Regeneration on the Properties of Filtrasorb 400
Carbon 28
7 Furnace Operating Conditions during the Regeneration of 300
System Carbon 34
8 Furnace Operating Conditions during the Regeneration of 400
System Carbon 35
9 Summary of Air Pollution Control System Performance 38
10 Carbon Capacity and Dosage Data 50
11 Distribution of COD Removal in the Two-Stage Systems 53
12 Average Water Quality Characteristics 55
13 Performance of the Filtrasorb 300 Two-Stage Adsorption
System 5 6
14 Performance of the Filtrasorb 400 Two-Stage Adsorption
System 57
15 Virus Removal Data 58
16 Design Criteria and Unit costs for Economic Analysis 60
17 Estimated Granular Activated Carbon Treatment Costs 61
vm
-------
ACKNOWLEDGEMENTS
This project was undertaken through a cooperative effort of the
Environmental Protection Agency and the Los Angeles County Sani-
tation Districts. The pilot plant study was performed at the
Sanitation District's Advanced Waste Treatment Research Facility
in Pomona, California.
Acknowledgement is made to Mr. Arthur N. Masse, former Chief,
Municipal Treatment Research Program of the Advanced Waste Treat-
ment Research Laboratory in Cincinnati, Ohio, and Mr. John N.
English, former site1 supervisor at the Pomona Research Facility,
for their invaluable advice given during the course of the study.
Thanks are also expressed to Mr. Jay B. Pitkin and Mr. William
Lee, former project engineers of the Sanitation Districts, for
their participation in the initial stages of the pilot plant study.
The untiring efforts and assistance of both the laboratory and
the pilot plant operating personnel of the Pomona Advanced Waste
Treatment Research Facility are gratefully acknowledged.
IX
-------
SECTION I
INTRODUCTION
The use of activated carbon for the removal of organic contaminants has
been known for a long time and is well decumented in literature (1).
However, it is only in recent years that research efforts have been
directed towards the application of activated carbon for the removal
of residual organic matter from wastewater.
Since 1965, the Environmental. Protection Agency and the Los Angeles County
Sanitation Districts have jointly operated a 12.6 I/sec (0.3 mgd)
activated carbon adsorption pilot plant at the Sanitation District's
Advanced Waste Treatment Research Facility at Pomona, California. Between
1965 and 1969, the 12.6 I/sec (0.3 mgd) pilot plant was operated as a
4-stage adsorption system with the primary objective of establishing the
technical and economic feasibility of using granular activated carbon for
removing soluble organics from secondary effluent and to obtain cost
data and operating experience on carbon regeneration. The results of the
4-year field study demonstrated successfully the effectiveness of granular
activated carbon in the adsorption of residual organic matter from an activated
sludge plant effluent. In addition, extensive operating data collected
over the years have demonstrated that thermal regeneration of exhausted
carbon is an economically feasible process. Thus, based on the experience
with the 4-stage adsorption study and following an appraisal of the state-
of-the-art of granular carbon contacting systems published in 1969 (2),
the carbon columns were modified in March 1970 to operate as two parallel
6.3 I/sec (0.15 mgd) two-stage adsorption systems.
The objectives of the two-stage adsorption-system study were to determine
on a long-term basis the effect of carbon particle size on adsorption rate
and adsorption capacity at 40 minute contact time; and to evaluate the
effect of repeated thermal regeneration cycles on the adsorption capacity,
regeneration loss and pressure drop build-up of carbon with different
particle size. The two parallel two-stage adsorption systems, one con-
taining a 12 X 40 mesh carbon similar to that used in the 4-stage adsorption
system and the other an 8 X 30 mesh carbon, were operated in a similar
fashion and at the same contact time as the 4-stage system.
The data presented in this final report are the results of four adsorption
cycles for the 8 X 30 mesh carbon» which required 33 months to complete
and 3 adsorption cycles covering 26 months for the 12 X 40 mesh carbon.
-------
SECTION II
CONCLUSIONS
The 6.3 I/sec (0.15 mgd) two-stage, packed-bed, downflow, granular activated
carbon pilot plants were operated successfully for 33 months using unfil-
tered and unchlorinated activated sludge effluent from the Pomona Water
Renovation Plant. Extensive operating data collected during the field
study had demonstrated the stability of the two-stage carbon adsorption
system in producing consistently effluents of excellent overall quality.
The average effluent dissolved chemical oxygen demand (DCOD) and suspended
solids concentration were 6.4 mg/1 and 2.1 mg/1, respectively for the
Filtrasorb 300 system and 6.2 mg/1 and 2.0 mg/1, respectively for the
Filtrasorb 400 system.
The overall organic removal through the Filtrasorb 300 system averaged
74.7% for total chemical oxygen demand (TCOD) and 73% for DCOD. The
corresponding removal in the Filtrasorb 400 system was 78.6% for TGOD and
74.5% for DCOD. The DCOD removal efficiency through the two-stage systems
has not changed significantly after several regeneration cycles. In the
Filtrasorb 300 system, the DCOD removal decreased from 72.3% during the
virgin cycle to 68.9% after four adsorption cycles. The corresponding
values for the Filtrasorb 400 system were 72.5% in the virgin cycle and
71.7% after three adsorption cycles.
During the virgin cycle, the carbon capacity, expressed as kg DCOD
removed/kg carbon, was 0.35 for the 300 system and 0.39 for the 400 system.
About 25% decrease in carbon capacity occurred after four adsorption cycles
in the 300 system, resulting in a fourth cycle capacity of 0.26. The
carbon capacity of the 400 system decreased about 23% from an initial level
of 0.39 to an apparent steady-state level of 0.30 after three adsorption
cycles. The 400 system has approximately 13% more DCOD removal capacity
than the 300 system. While the 400 system showed a slightly higher carbon
capacity than the 300 system, the latter has the advantage of lower pressure
loss and lower carbon loss during regeneration.
The carbon dosage in the 300 system increased from .038 kg/m3 (320 Ibs.
carbon/million gallons) to .066 kg/m3 (550 Ibs. carbon/million gallons)
after four adsorption cycles. In the 400 system, the carbon dosage
increased from .034 kg/m3 (280 Ibs. carbon/million gallons) to .06 kg/m3
(500 Ibs. carbon/million gallons) at the end of the third adsorption cycle.
The first stage carbon columns, which served as deep bed filters and
adsorbers, were routinely backwashed every two days with a volume of
secondary effluent equivalent to 1.8% of the product water. The filtering
action through the carbon column readily removed about 10 mg/1 of the
suspended solids from the secondary effluent feed, resulting in an average
-------
net headloss buildup during the two-day backwash cycle of .26 kg/cm2
(3.7 psig) for the 300 system and .4 kg/cm2 (5.7 psig) in the 400 system.
Of the TCOD removal in the two stage systems, 76.4% is removed in the first
stage, primarily due to filtration.
An evaluation of the carbon characteristics following repeated thermal
regenerations shows that after four adsorption cycles in the 300 system,
the iodine number, methylene blue number and mean particle diameter
decreased 29.5%, 27%, and 14.4%, respectively. The decrease in iodine
number, methylene blue number and mean particle diameter after three
adsorption cycles in the 400 system were respectively, 23%, 14.9%, and 0%.
During the same period, the molasses number increased from 222 to 347 in
the 300 system, and from 237 to 350 in the 400 system.
The ash content of the carbon increased with adsorption cycle from 5.5%
to 12.9% in the 300 system and from 5.8% to 11.7% in the 400 system.
The estimated average carbon loss during regeneration is 6.8% over four
adsorption cycles in the 300 system and 7.6% after three adsorption
cycles in the 400 system.
The estimated total treatment cost for a .44 m3/sec (10 mgd).two-stage
carbon adsorption system, designed to produce a produce water with TCOD of
12 mg/1 and DCOD of 7-8 mg/1 using an activated sludge plant effluent
feed, is 11.52<£/1000 gallons. The estimated cost is based on using
the 300 system carbon (8 x 30 mesh) with a carbon dosage of .066 kg/nr
(550 Ibs. carbon/million gallons) and a carbon regeneration loss of
7% per cycle.
These costs are referenced to October 1973.
-------
SECTION III
RECOMMENDATIONS
1. A similar study with single stage carbon contact should be conducted.
2. Evaluation of optimum backwash techniques should be conducted.
3. The effect of louding on carbon loss during regeneration should be
evaluated.
-------
SECTION IV
EXPERIMENTAL PROGRAM
PILOT PLANT DESCRIPTION
Carbon Contacting System
The carbon columns, comprising the two parallel two-stage adsorption sys-
tems were the same columns used in the earlier 4-stage carbon adsorption
system evaluated in Pomona (3). The general layout of the carbon adsorp-
tion system is shown in Figure 1, together with the location of the carbon
regeneration and air pollution control systems. Figure 2 shows the schema-
tic flow diagram of the two-stage adsorption systems. The detail of a ty-
pical carbon column is illustrated in Figure 3. Each of the carbon columns
was 1.83m (6 ft) diameter by 4.88m (16 ft) high and was designed for a
working pressure of 3.52 kg/cm2 (50 psig). Before the columns were reverted
to the two-stage operational modes, the interior of columns II, III and IV
were sandblasted and recoated with corrosion-inhibiting coating. Columns
II and III were coated with bitumastic coal-tar epoxy and contactor IV
with a polyester resin Ceilcote Flakeline 252. The two-stage system, con-
sisting of columns II and III, contained 3164 kg (6960 Ibs) of Calgon Fil-
trasorb 300 (8 x 30 mesh), while the other system with columns IV and V
contained 3027 kg (6660 Ibs) of Calgon Filtrasorb 400 (12 x 40 mesh).
The system containing 8 x 30 mesh carbon is designated as Filtrasorb 300
system, while the other column system with 12 x 40 mesh carbon is referred
to as Filtrasorb 400 system. Table I shows the characteristics of the
virgin carbon. The carbon bed was supported on an underdrain system
consisting of two layers of stainless steel plate perforated with 0.79 mm
(1/32") diameter holes and spot-welded together with the perforations
off-set from each other.
The two-stage systems were operated in downflow mode at a constant rate of
6.3 I/sec (100 gpm) thereby providing a hydraulic loading of 2.38 1/sec/m2
(3.5 gpm/ft2) and an empty-bed contact time of 20 minutes per stage. The
unchlorinated effluent from a 0.35 m3/sec (8 mgd) activated sludge plant
was pumped to the lead contactors without any pretreatment. The column feed
entered the top of the contactors through an annular distribution ring
containing twenty 2.54 cm (1 in) diameter holes around the circumference
of each contactor. The wastewater flowed in series through each of the two
contactors in the two-stage systems. The entrance annular ring was loca-
ted about 1.07 m (3.5 ft) above the top of the carbon bed, thus providing
for 35% bed expansion during backwashing. Each contactor was provided with
a fixed surface wash mechanism mounted about 5.08 cm (2 in) above the un-
expanded carbon bed to assist in the routine column backwashing.
-------
SECONDARY
CLARIFIER
CHLORINATION
PLANT
MAINTENANCE
BUILDING
PRIMARY
SEDIMENTATION
TANK
LAB
ION EXCHANGE
PILOT UNIT
CARBON CONTACTORS
(T) (TT) (m) (7v
QUENCH
TANK /~^. DRAIN
O O()
FURNACE
/
(
' ^-^|| BACKWASH
1 1 TANKS
O
SECONDARY
EFFLUENT
STORAGE TANK
PRODUCT
WATER TANK
OFFICE AND
SHOP BLDG.
REVERSE
OSMOSIS
PILOT
UNITS
I
— N-
FIGURE |: GENERAL LAYOUT OF THE CARBON ADSORPTION SYSTEM
-------
SECONDARY EFFLUENT
J
n
1
D
\
FILTRASORB 300
8X 30
I
12
FILTRASORB
12 X40
400
FIGURE 2: SCHEMATIC DIAGRAM OFTHE TWO-STAGE ADSORPTION SYSTEMS
-------
FULL OPEN COVER-
WITH 15" PORTHOLE
a
LJ U U U LJ
pnnnnnnnnnnnnnnnn
O O O O O
20-1" HOLES
WASH
WATER
H
U.
tO
H
U.
>^-VM
a>
H SURFACE WASH
(.1
Vj:
_CARBQN BEp0S_URRflkCEri
" SAMPLING
TAPS
NEVA CLOG^o
SCREED
D:
6 FT.
RING
INFLUENT
BACKWASH
CHARGE
CARBON
"DISCHARGE
EFFLUENT,.
"BACKWASH
FIGURE 3! TYPICAL CARBON CONTACTOR
-------
TABLE 1
VIRGIN CARBON CHARACTERISTICS
Carbon
Characteristics
Iodine Number, mg/g
Molasses Number
Methylene Blue No., mg/g
Apparent Density, g/cm3
Ash, %
Mean Particle Dia., mm
Sieve Analysis:
% Retained on No.
...
8
10
12
14
16
18
20
30
40
Pan
Calgon Filtrasorb 300
(8 x 30 mesh)
984
222
271
.484
5.5
1.6
4.2
16.3
25.7
18.7
16.8
10.2
4.9
2.6
0.6
Calgon Filtrasorb 400
(12 x 40 mesh)
1062
237
275
.463
5.8
1.0
1.8
10.3
24.9
20.8
14.8
20.6
6.0
0.8
-------
Carbon Regeneration System
The regeneration furnace was a 76.2 cm (30 in) I.D., vertical, re-
fractory-lined Bartlett-Snow-Pacific, Inc. multiple hearth furnace.
As depicted in Figure 4, the furnace contained six hearths with two
gas burners and steam inlets in each of the lower three hearths. The
operation of the furnace was fully automatic, with push-button con-
trols, safety equipment and a 12-pen furnace temperature recorder.
The furnace was natural gas-fired with steam added to enhance the
regeneration.
The partially dewatered, spent carbon from the drain bin was fed by
a screw conveyor into the top of the regeneration furnace. The screw
conveyor was provided with a variable speed drive so that the desired
rate of carbon fed to the furnace can be accurately controlled. Re-
generated carbon was discharged from the furnace through a 7.62 cm
(3 in) diameter stainless steel chute leading from the bottom hearth
into a quench tank, from which the regenerated, quenched carbon was
continuously educted back to the original contactor.
Air Pollution Control System
In the course of the thermal regeneration of granular activated car-
bon spent on secondary effluents, severe air pollution problems could
result. Experience with earlier field studies in Pomona (3) showed
that the two major air pollutants associated with carbon regeneration
were noxious odors and particulate emissions. In the early years of
the pilot plant study several attempts were made to control these
emissions through the use of a rotoclone and a direct-fired after-
burner. The use of the rotoclone alone, or in series with the after-
burner, was unable to meet the particulate discharge requirement of
the Los Angeles County Air Pollution Control Districts (APCD). The
operation of the afterburner alone at 927°C (1700°F) was very effective
in odor removal, but only marginally effective in the control of par-
ticulate emissions. For four years until mid-1972 the carbon regene-
ration system operated with just the after-burner for air pollution
control.
In view of the more stringent air pollution discharge requirements,
it was decided in 1971 to install additional particulate emission
control devices. The design and selection of the particulate control
devices were made in cooperation with the APCD engineering staff.
The modified air pollution control system, shown schematically in
Figure 5, consisted of a single cyclone dust separator, a baghouse,
and a natural gas-fired afterburner. The new particulate emission
devices were first placed in operation in July, 1972. During carbon
regeneration, the flue gases from the furnace, together with parti-
culates, first passed through the cyclone, which was designed to trap
all burning particulates, 10yu in diameter or larger before reaching
the fabric filters in the baghouse.
10
-------
CARBON IN
GAS OUT
DRY CARBON
SAMPLING TUBE
SHAFT
DRIVE
UNIT
RABBLE ARM
RABBLE TEETH
3 DIAMETER
REGENERATED
CARBON OUT
TO QUENCH TANK
FIGURE 4: CROSS SECTIONAL VIEW OF THE MULTIPLE HEARTH
FURNACE
11
-------
SECONDARY EFFLUENT
ro
BACKWASH
f
m
iz
PRODUCT
WATER —
TANK
TO ATMOSPHERE
TO
PRIMARY
CLARIFIER
JL
BACKWASH
TANK
CARBON
DEWATERING FUEL-i
TANK T
AIR—*
WATER
SPENT
j|£ARBjON
REGENERATED
CARBON OUT'"
QUENCH
TANK
TO
SAN JOSE CREEK
TO
CARBON
COLUMN
EDUCTOR
MULTIPLE
HEARTH
.FURNACE
CYCLONE
AFTERBURNER
BLOWER
DUST TO
STORAGE DRUMS
MAKE-UP WATER
MOTIVE WATER
COMPRESSED AIR
•FOR PULSED AIR
CLEANING OF BAGS
BAGHOUSE
DUST TO
STORAGE DRUMS
FIGURE 5! SCHEMATIC DIAGRAM OF THE AIR POLLUTION CONTROL SYSTEM
-------
From the cyclone, the gases flowed through a series of ducts into a
baghouse. The ductwork leading to the baghouse was provided with valved
connection for dilution air addition. The baghouse was a reverse-air
jet cleaned unit (Model 9-6-100 Mikro-Pulsaire Dust Collector manufac-
tured by Mikropul Division of Slick Corporation), containing 9 Nomex
felt fabric filter bags with a combined filter area of 5.95 sq m (64
sq ft). Each of the Nomex bags was 1.83 m (6 ft) long and was designed
for an operating temperature of 204-218°C (400-425°F). The dust laden
air entered the lower section of the baghouse and travelled upward through
the fabric filter cylinder where the dust particles collected on the
outside surface of the filter elements. A pull-through exhaust fan
mounted on top of the baghouse provided the driving force for the dust
flow through the system. As the dust mat builds up on the fabric sur-
face, the pressure differential across the filter increased to a level
where the deposited solids had to be removed by reverse air flow. In
order to control the pressure drop through the filter within the desired
limits of 2.54 cm (1 in) to 15.24 cm (6 in) water column, a cyclic timer
periodically (2 to 45 second intervals) actuated the solenoid valves
which delivered momentary surges of high pressure air 7kg /cm2 (100 psig).
The filtered gases from the baghouse then flowed through the afterburner
which was operated between 704° C (1300°F) and 760°C (1400°F) for odor
control, before final discharge into the atmosphere.
PILOT PLANT OPERATION
Column Designation ;
Each of the carbon columns in the two-stage adsorption systems was op-
erated for extended periods in various operating sequences with carbon
at different regeneration levels. In order to avoid confusion in dis-
cussing the column operation, it was necessary to identify the contactors
according to the designation given in Figure 6. Thus, as an example,
the operating sequence designated as IIOA, I HOB would mean that the
lead contactor or "A" position contactor is the Column II and the second
stage or "B" position is Column III. Both columns contained virgin car-
bon which is represented by number "0". Since the columns were operated
in a semi-counter-current mode, after the carbon in the lead contactor
was regenerated, the column was placed on stream in the B position and
the operating sequence would be designated as IIIOA, II1B. Moreover,
two adsorption sequences, that is, when each of the columns has been
on stream in both "A" and "B" positions, represent an adsorption cycle.
Carbon Transfer and Regeneration
In liquid-phase adsorption system using beds of granular activated car-
bon, when the organic concentration of the effluent from the last stage
exceeds a predetermined effluent limit, only the carbon layers in the in-
let section are highly saturated. The carbon near the outlet end of the col-
umn system still has useable capacity. Thus, it is generally advantageous
to operate the system in a counter-current manner in which only the highly
saturated or exhausted carbon at the inlet section is removed from ser-
vice for regeneration.
13
-------
EOB
POSITION OF CONTACTOR
ON SLAB
NUMBER OF REGENERATIONS
OF CARBON IN CONTACTORS
POSITION OF CONTACTOR
IN TOTAL COLUMN
FIGURES: CONTACTOR IDENTIFICATION NUMBER
-------
While there are a number of approaches to achieve a counter-current
operation, the method adopted in the study was to operate two carbon
columns in series. When the effluent from the two-stage system reached
a TCOD of 12 mg/1, the lead carbon column was taken off stream in
preparation for regeneration. The spent carbon was thoroughly back-
washed before being hydraulically transferred as a slurry to an ele-
vated dewatering bin. The backwash procedure was similar to that
used for routine column backwash except for the fact that the last
backwash step was prolonged to provide a total backwash water volume
of about 79.5 m3 (21,000 gallons). The dewatered spent carbon, with
about 40% moisture content, was conveyed through a screw conveyor into
a six-hearth furnace where it was regenerated at temperatures ranging
from 899-982°C (1650-1800°F). As discussed previously, steam was added
in the lower two hearths (hearths 5 and 6) in the amount of 0.6 kg steam/
kg carbon to enhance the regeneration. The regenerated carbon was dis-
charged from the furnace at the rate of 40.9-45.5 kg/hr (90-100 Ibs/hr)
into a quench tank from which it was continuously educted back into the
original contactor. The regeneration of spent carbon in one contactor
normally required about 72-95 hours to complete. After regeneration,
the regenerated carbon was then backwashed with 56.8-75.7 nr (15,000-
20,000 gallons) of secondary effluent to remove carbon fines and appro-
priate amount of make-up carbon added to replace the carbon lost during
regeneration. The column, with the added make-up virgin carbon, was
then briefly backwashed with 7.6-18.9 m3 (2,000-5,000 gallons) of se-
condary effluent before the column was placed back in operation as a
second stage in the system.
SAMPLING AND TESTING PROGRAM
Refrigerated 24-hr, composite samples of influent and effluent from each
carbon column were collected automatically using timer-controlled sole-
noid valves. These samples were analyzed daily for turbidity and two to
three times a week for TCOD, DCOD, suspended solids and color. About
once a week, the samples were analyzed for ammonia, nitrite, nitrate, TOC,
and methylene blue-active substances. Analytical determinations for pH
and temperature were performed on grab samples three times a week.
The presence or absence of virus from the two-stage carbon effluent was
determined once a week by analyzing a standard virus swab contacted with
a stream of carbon effluent for a three-day period. In Figure 7 is shown
a sketch of a flow-through virus sampling device. The average flow through
the sampling module during the three-day contact period varied from .09
to .19 I/sec (1.5 to 3.0 gpm). All virus analyses were performed by the
Los Angeles County Health Department.
All physical and chemical analyses were performed in accordance with the
12th edition of Standard Methods (4) or the FWPCA Methods for Chemical
Analysis (5) unless otherwise specified. Turbidity was determined by
the use of Rossum turbidimeter. TOC was measured by Beckman total organic
carbon analyzer. The activated carbon analyses were performed using
standardized procedure of the Pittsburg Activated Carbon Company.
15
-------
12
12
4" PVC-
PIPE SHELL
1/8" HOLES
SWAB IN 2" PIPE
FIGURE 7- VIRUS SAMPLING CHAMBER
16
-------
SECTION V
EXPERIMENTAL RESULTS
COLUMN BACKWASHING
Since the two-stage granular activated carbon columns were operated
on packed-bed, downflow mode, the first stage carbon columns functioned
as deep filter beds and as adsorbers. Thus, the application of unfil-
tered activated sludge effluent, with suspended solids concentration
ranging from 10 to 20 mg/1, directly to the carbon columns led to pro-
gressive clogging of the beds with an attendant increase in headless.
To maintain proper column operation, the lead carbon columns were back-
washed on a predetermined schedule with secondary effluent. During the
course of the study, a number of backwash procedures as to frequency,
type and duration, were tried. The actual backwash duration varied
from 33 to 72 minutes depending on the procedure used. During the
first three months of operation, the lead contactors ("A" position)
were backwashed with about 56.8 m3 (15,000 gallons) of secondary ef-
fluent once every 8 days or whenever the headless reached 1.05 kg/cm2
(15 psig). This infrequent backwashing operation led to heavy accumu-
lations of biological solids in the carbon bed which were difficult
to backwash. In addition, channeling in the carbon beds were observed.
Consequently, the backwash frequency was increased to once every, three
days for about 1 1/2 months. Thereafter, the backwash frequency was
increased to every 2 days, using a modified backwash procedure along
with the use of a fixed surface spray mechanism. The backwash schedules
for the original and the modified backwash procedure,are shown in Figures
8 and 9. The modified backwash procedure, which^had been used until
the completion of the field study, had the following advantages over the
original procedure:
1) The same degree of bed cleansing, as determined by visual
observation of the backwash water and the headloss after
backwash, was achieved with about half the volume of back-
wash water.
2) The system downtime during backwashing was reduced from 72
minutes or more to 33 minutes.
3) As indicated in Figure 9, the maximum backwash rate of
7.13 l/sec/m2 (10.5 gpm/ft2) was reached earlier, thus pro-
viding maximum bed expansion for a longer portion of the
backwash operation.
4) The surface spray had been effective in breaking up the
accumulated deposits on top of the carbon bed. In addi-
tion, the routine use of the surface spray during back-
wash helped in maintaining the carbon surface level and
reduced channeling through the bed.
17
-------
00
20
K
u_
QL
O
I
I
J^-BACKWASH (10,000 GALS.)
I
I
I
I
I I
10 15 20 25 30 35 40 45
TIME, MINUTES
50 55 60 65 70 75
FIGURE 8'- ORIGINAL BACKWASH SCHEDULE
-------
20
15
QL
CD
O
r
r
J
I
~i
i
i
i
r
___J*2~BACKWASH (5,000 GALS.) '
SURFACE WASH (500 GALS.)
I I
J i
10 15 20
TIME, MINUTES
25
30
35
FIGURE 9: MODIFIED BACKWASH SCHEDULE
-------
During backwashing, the backwash water was discharged into a holding
tank designed to capture any accidental carbon spills and to allow
visual observation of the clarity of the backwash water. The backwash
water containing heavy accumulations of biological floes and some carbon
fines overflowed a weir in the holding tank and was pumped into the head
end of the primary clarifiers of the Pomona activated sludge plant.
While a 35% bed expansion during backwashing was possible, routine
column backwash operation was limited to a maximum upflow rate of 7.13
l/sec/m2 (10.5 gpm/ft2) because of the limitation on the structural
strength of the underdrain screen. Experience with backwash at 9.5-10.86
l/sec/m2 (14-16 gpm/ft2) had shown severe damage to the underdrain screen
which took several months to repair. At the backwash rate of 7.13 l/sec/m2
(10.5 gpm/ft2), the measured bed expansions were 10% and 15% in the
Filtrasorb 300 system and Filtrasorb 400 system, respectively.
HEADLOSS BUILDUP
The headloss buildup through a granular activated carbon column operated
on a packed-bed, downflow mode, is influenced by such factors as hydrau-
lic surface loading, influent suspended solids level, carbon particle
size and the length of filter runs. During the entire study, the two
parallel two-stage systems were operated using the same secondary ef-
fluent feed under identical conditions of hydraulic surface loading,
filter run lengths and backwash procedure. Thus, assuming that the con-
tribution of biological growths within the columns on the total pressure
drop were equal in both systems, then the magnitude of the pressure drop
buildup through the lead columns during the two-day back-wash cycle would
depend primarily on the media particle size. With this in mind, a re-
gression analysis was performed to determine the relationship between
column influent turbidity, T, and net pressure drop, P, in each of the
two types of carbon at various regeneration levels. The results of the
regression analysis, which are shown in Figure 10, clearly indicate that
at a hydraulic loading of 2.38 l/sec/m2 (3.5 gpm/ft2) the 8x30 mesh carbon
exhibited about 33% less headloss than the 12x40 mesh carbon. The figures
further demonstrate the effect of regeneration level on the column head-
loss. For both systems the headloss, after the first regeneration, was
higher compared to that in the virgin carbon. In subsequent regenerations,
while the pressure losses were consistently lower, they were not signifi-
cantly different from the pressure loss in the virgin carbon. The re-
gression equations are presented in Table 2.
The effectiveness of the backwash operation could be determined by ex-
amining the column headloss just after backwash. In both systems the
initial headlosses were consistently low and averaged only 0.11 kg/cm2
(1.6 psig) for the 300 system and 0.13 kg/cm2 (1.9 psig) for the 400
system. The average headloss through the lead contactors of the two-
stage systems at various adsorption sequences are presented in Tables
3 and 4. The data clearly show that the headlosses through the 400
system were consistently higher than those in the 300 system. These
data confirmed those presented previously in Figure 10. The-net pres-
sure losses through the lead contactors, which are attributed to the
accumulation of suspended solids, averaged 0.26 kg/cm2 (3.7 psig) for
the 300 system and 0.40 kg/cm2 (5.7 psig) for the 400 system.
20
-------
ro
15
e>
CD
o.
t»
oc
p
2.
o
<
CO
CO
o
d»
<
Ul
X
H
Ul
HYDRAULIC LOADING = 3.5 gpm/ft.2
S®'
12 X 40 MESH CARBON
8 X 30 MESH CARBON
10
INFLUENT TURBIDITY, JTU
15
20
FIGURE 10: EFFECT OF INFLUENT TURBIDITY AND CARBON PARTICLE SIZE ON
COLUMN HEADLOSS
-------
TABLE 2
CORRELATION OF HEADLOSS WITH INFLUENT TURBIDITY
ro
Type of
Carbon
Filtrasorb
300 (8x30 mesh)
Filtrasorb
400
(12x40 mesh)
No. of
Regenerations
0
1
2
0
1
2
Regression Equation Correlation
P=3
P=3
P=2
P=5
P=5
P=5
.68 +
.55 +
.29 +
.52 +
.60 +
.71 +
0
0
0
0
0
0
.26T
.32T
.28T
.40T
.45T
.25T
0
0
0
0
0
0
Coefficient
.44
.54
.66
.40
.61
.38
P= Net headless, psig
f= Influent turbidity, JTU
Unit Conversion: psig x .070 = kg/sq cm
-------
TABLE 3
HEADLOSS BUILDUP AND TURBIDITY REMOVAL IN THE
FILTRASORB 300 TWO-STAGE ADSORPTION SYSTEM
Run
No.
1
2
3
4
5
6
7
8
Pressure Loss, j)sig
1st Stage
BBW
5.6
5.0
4.8
6.7
5.8
5.4
4.2
5.2
ABW
1.4
1.4
1.5
2.4
1.8
1.2
0.9
2.1
2nd Stage
BBW
1.3
0.9
0.8
0.8
0.7
1.0
1.3
0.6
ABW
1.0
0.9
0.8
0.9
0.8
1.0
1.2
0.9
Turbidity, JTU
Influent
6.7
3.9
5.1
7.2
9.5
15.1
6.7
8.7
1st Stage
Effluent
2.2
1.2
1.6
2.3
2.2
4.3
2.5
2.2
2nd Stage
Effluent
1.9
1.0
1.3
2.0
2.0
4.0
2.4
1.6
Total Volume
Treated, mil-
lion gallons
26.324
12.240
16.363
8.408
19.294
9.420
10.410
18.073
ro
CO
BBW = Before backwash
ABW = After backwash
Unit Conversions: psig x .070 = kg/sq cm
mil gal x 3785 = cu m
-------
TABLE 4
HEADLOSS BUILDUP AND TURBIDITY REMOVAL IN THE
FILTRASORB 400 TWO-STAGE ADSORPTION SYSTEM
Run
No.
1
2
3
4
5
6
Pressure
1st
bBW
7.2
7.8
4.3
8.9
7.7
10.0
stage
ABW
1.9
1.9
1.1
2.2
1.8
2.5
Loss, psig
2nd
BbW
1.7
1.0
1.1
1.0
1.2
1.3
btage
Abw
1.5
0.8
1.0
1.0
1.0
1.4
Turbidity, JTU
Influent
5.4
4.2
5.6
6.7
10.4
12.3
1st Stage
Effluent
1.4
1.3
1.5
2.2
2.0
3.4
2nd Stage
Effluent
1.1
1.0
1.2
1.8
1.9
3.0
Total Volume
Treated, mil-
lion gallons
29.451
11.798
15.546
9.627
19.328
11.864
ro
BBW = Before backwash
ABW = After backwash
Unit Conversions: psig x .070 = kg/sq cm
mil gal x 3785 = cu m
-------
CARBON REGENERATION RESULTS
Control of the Regeneration Process
The economical use of granular activated carbon in wastewater treatment
application demands that the exhausted carbon be repeatedly regenerated
and then reused. The main goal of carbon regeneration is to effect
maximum restoration of the exhausted carbon to its virgin properties.
This goal is achieved by subjecting the carbon through a three-step
process; namely: drying, baking and activating (6). The process var-
iables, such as furnace temperature, carbon feed rate, and steam feed
rate are closely controlled during regeneration in such a way as to
effect maximum removal of adsorbed organics from the pores of the spent
carbon, while at the same time, minimizing the damage to the basic pore
structure of the carbon. In the activation step, the temperature is
controlled automatically within 899-954°C (1650-1750°F), with steam
added in hearths 5 and 6 to enhance regeneration.
In the course of regeneration, a number of control tests, consisting
of the determination of apparent density, iodine number, methylene blue
number and molasses number, were performed to monitor the quality of the
regenerated carbon. The apparent density of virgin carbon normally
ranged from 0.48 to 0.49 g/cm3, which increased to about 0.59 when the
carbon becomes exhausted. When the carbon is properly regenerated, the
adsorbed organics are removed, thereby restoring the apparent density
to the virgin level. The extent of recovery of the carbon adsorptive
capacity was determined by running the decolorizing tests for iodine
and molasses numbers.
The test for apparent density was determined routinely every hour whereas
the test for iodine and molasses numbers were performed every 4 hours
and 2 hours, respectively. The 6 to 8 grab samples of spent carbon,
collected during carbon transfer, and the hourly samples of regenerated
and quenched carbon were composited over the regeneration period and
analyzed for apparent density, molasses number, iodine number, methylene
blue number, ash content and mean particle size.
Effects of Regeneration
As a result of the various operations entailed in the thermal regenera-
tion process, some loss invariably occur both in the carbon adsorptive
capacity and also in the carbon quantity. Both these losses are of
economic concern since they constitute a significant portion of the
overall carbon regeneration cost. The carbon loss during each regene-
ration cycle has varied from 6.3% to 11% in the Filtrasorb 400 system
and from 4.8% to 10.4% in the Filtrasorb 300 system. As used in this
report, carbon loss is defined as the difference in the carbon volume
in the contactor just before transfer to the drain bin and just after
backwash of the regenerated carbon. The carbon loss, which is indica-
tive of the carbon particle volume decrease, can be attributed either
to the direct oxidation of the outer layers of the carbon granules
25
-------
by the regenerating gases and/or to the normal handling attrition if
the basic carbon structure has been weakened by internal overactivation. (6)
With the physical loss of carbon during regeneration, is the attendant
decrease in the adsorptive capacity as measured by dissolved COD removal
and iodine number. This observed loss in the carbon adsorptive capacity
following repeated thermal regeneration is ascribed to the changes in the
physical properties of the carbon. During regeneration, the complete re-
moval of adsorbed organics from the carbon pores is really never attained.
In addition to the ash buildup in the carbon pores, some carbon particles
are unavoidably burned with the adsorbate thereby reducing further the
total carbon surface area available for adsorption (6).
The changes in some of the physical properties of the carbon before and
after several regenerations are shown in Tables 5 and 6. It is evident
from the data that cyclic thermal regenerations has the effect of re-
ducing the iodine number and increasing the molasses number and the %
ash content. The increase in molasses number and the decrease in the
iodine number is consistent with the fact that repeated thermal regene-
rations will cause a change in the pore size distribution of the carbon
granule with the micropore structures being enlarged and thus producing
a larger per cent of macropores. This shift in pore size distribution
has the overall effect of reducing the total surface area of the carbon.
In this study, the iodine number was used as an index to measure the
extent to which the carbon micropores were cleared of adsorbate during
regeneration. The iodine number relates to the surface area of pores
larger than 10A° diameter. The iodine number of both-spent and regene-
rated carbon are presented in Figure 11 for the 300 system and in Figure
12 for the 400 system. As indicated in Figure 11, the iodine number
decreased 34% from a virgin level of 984 mg/g to 650 mg/g after four
adsorption cycles in the 300 system. For the 400 system, the iodine
number decreased from a virgin value of 1062 mg/g to 743 mg/g at the
end of the third adsorption cycle. A continuing drop in the iodine
number with adsorption cycle is apparent from the figures.
The ash content of the carbon, which measures the buildup of calcium and
other inorganic residues, increased significantly from a virgin level of
5.5% to 12.5% after four cycles in the 300 system. For the 400 system,
the corresponding ash buildup was from 5.8% to 11.4% after three cycles.
This buildup of ash, along with the increase in the apparent density,
following repeated thermal regenerations, is consistent with the de-
creasing trend in the iodine number. Figures 13 and 14 present the data
on ash buildup and mean particle diameter as a function of the adsorption
cycle. As evident from Figure 13, the mean particle diameter of the
Filtrasorb 300 carbon decreased about 14% after four adsorption cycles.
However, no further significant decrease in particle size occurred after
the third cycle. In contrast, the mean particle diameter of the Filtra-
sorb 400 carbon remained practically unchanged through three adsorption
cycles as shown in Figure 14.
In the course of thermal regeneration, some degree of internal overacti-
vation occurs which results in an increase in macropore volume, that is,
increase of pores larger than 30AQ diameter. The molasses number, which
26
-------
TABLE 5
EFFECT OF REGENERATION ON THE PROPERTIES
OF FILTRASORB 300 CARBON
(8 x 30 mesh carbon)
No. of Regenerations
Adsorption Cycle
I odi ne
Number
(mg/g)
Molasses
Number
Methyl ene
Blue Number
(mg/g)
Apparent
Density
(g/cm3)
•
Ash, (%)
Mean
Particle
Diameter, (mm)
(1)
(2)
(3)
(1)
(2j
(3)
(1)
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
0)
(2)
(3)
0
668
984*
102
222*
193
271*
.573
.484*
7
5.5*
1.37
1.6*
1
1
652
974
911
194
265
189
178
271
256
.575
.483
.495
7.6
8.1
7.2
1.43
1.37
1.49
2
2
567
843
817
141
333
303
163
228
227
.564
.493
.493
9.6
10.3
8.6
T.36
1.46
1.52
3
3
484
802
760
190
374
301
156
231
221
.585
.498
.499
12.1
10.9
10.2
1.41
1.31
1.43
4
4
695
651
347
328
198
179
.508
.516
12.9
12.5
1.37
1.41
(1) Spent carbon
(2) Regenerated carbon before quenching
(3) Regenerated carbon after quenching
* Virgin carbon
27
-------
TABLE 6
EFFECT OF REGENERATION ON THE PROPERTIES
OF FILTRASORB 400 CARBON
(12 x 40 mesh Carbon)
No. of Regenerations
Adsorption Cycle
Iodine
Number
(mg/g)
Molasses
Number
Methyl ene
Blue Number
(mg/g)
Apparent
Density
(g/cm3)
Ash, (%)
Mean
Particle
Diameter,
(mm)
(1)
(2)
(3)
(D
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
(1
(2
(3
(1)
(2)
(3)
0
725
1062*
159
237*
184
275*
.560
.463*
6.8
5.8*
1.04
1.0*
1
1
663
950
926
175
286
261
170
238
224
.555
.479
.472
8.2
8.0
7.8
0.99
.99
1.08
2
2
505
871
823
133
285
264
153
234
232
.580
.479
.484
10.1
9.9
9.2
1.01
1.02
1.01
3
3
815
743
350
323
234
218
.496
.504
11.7
11.4
1.0
1.06
(1) Spent carbon
(2) Regenerated carbon before quenching
(3) Regenerated carbon after quenching
* Virgin carbon
28
-------
1200
I000(
800
600
400
REGENERATED (DRY)
REGENERATED
(QUENCHED)
-SPENT
400
200
OZ
2
UJ
400
•"cc
UJUJ
Si 200JJ-.
>|2
tu
REGENERATED
(DRY)
REGENERATED
(QUENCHED)
REGENERATED (DRY)
Q r--a
SPENT
^
1
REGENERATED'
(QUENCHED)
1
0123
ADSOPTION CYCLE
FIGURE II: EFFECT OF ADSORPTION CYCLE ON IODINE NO.,
MOLASSES NO., METHYLENE BLUE NO. OF
FILTRASORB 300
29
-------
1200
1000
800
— CO
s
600
400
400
400
I-
UJ
REGENERATED (DRY)
REGENERATED
(QUENCHED)
SPENT
REGENERATED (DRY)
O—
O—v— —.__
•SPENT
REGENERATED
(QUENCHED)
REGENERATED (DRY)
D
D
SPENT
I
REGENERATED
(QUENCHED)
I
0123
ADSORPTION CYCLE
FIGURE 12'. EFFECT OF ADSORPTION CYCLE ON IODINE NO.,
MOLASSES NO., METHYLENE BLUE NO., OF
FILTRASORB 400
30
-------
12
10
REGENERATED
(QUENCHED)
REGENERATED (DRY)
~ 2
iu£
REGENERATED (QUENCHED)
,
J3CH- I
Q-
LU
<
UJO
NSPENT
REGENERATED
(DRY)
ADSORPTION CYCLE
FIGURE 13: EFFECT OF ADSORPTION CYCLE ON ASH
BUILDUP AND MEAN PARTICLE DIAMETER OF
FILTRASORB 300
31
-------
12
10
8
CO
O
SPENT
REGENERATED
(QUENCHED)
REGENERATED (DRY)
REGENERATED
(QUENCHED)
REGENERATED (DRY)
ADSORPTION CYCLE
FIGURE 14: EFFECT OF ADSORPTION CYCLE ON ASH
BUILDUP AND MEAN PARTICLE DIAMETER OF
FILTRASORB 400
32
-------
relates to the surface area of the pores larger than 28AQ diameter, is
taken as a measure of the pore enlargement. The molasses number of the
virgin carbon has been measured in the range of 191 to 271 for Filtrasorb
400 and 190 to 238 for Filtrasorb 300. In all regeneration runs, the
molasses number of regenerated carbon remained at high level, with a
range of 231 to 323 for Filtrasorb 400 and 182 to 382 for Filtrasorb 300.
Another parameter used to measure pore enlargement during regeneration,
was the methylene blue number, which relates to surface area of carbon
pores larger than 15AO diameter. In Figures 11 and 12 are shown the
changes in the methylene blue, molasses and iodine numbers of carbon
samples at various regeneration levels. The methylene blue and iodine
numbers of regenerated carbon show a continuing decrease with regenera-
tion level in contrast with the increasing trend of the molasses number.
The relatively high molasses number and low iodine number indicate some
internal damage to the carbon pore structure, resulting in a general
shift to pores of larger size.
In examining the iodine number data presented in Figures 11 and 12, it
is evident that a difference in iodine number of the regenerated carbon
before and after quenching occurred consistently in all carbon regenera-
tions. This observation has been confirmed by other investigators (6)
who found reduction in iodine number of 25 to 69 units for virgin carbon
Filtrasorb 400 heated to 954°C (1750°F) and quenched in water. Similarly,
the molasses and methylene blue numbers of the regenerated carbon de-
creased after quenching.
In Tables 7 and 8 are summarized the operation data for the regeneration
of each contactor in the two-stage adsorption systems. The data show
some variations in the duration of regeneration, furnace loading, steam
and fuel used. These variations in the furnace operation have been
attributed primarily to the feeding conditions of the exhausted carbon.
As shown by the data in Table 7, the regeneration of spent carbon in col-
umn IIOA took an unusually long period to complete, due to operating
problems with the furnace. Previous to this particular regeneration, the
furnace had not been operated for about five months. In addition, there
was a tremendous problem experienced in the transfer of carbon to the de-
watering bin. Since the spent carbon in column II was not completely
transferred to the drain bin, the original large quench tank was used
as a temporary storage for the regenerated carbon. Thus, regeneration
had to be stopped when the quench tank became full. The carbon remain-
ing in the contactor was then manually transferred to the dewatering bin
and, finally, regeneration was resumed. The regeneration of this last
batch of carbon proceeded with great difficulty since it was "sticky"
and difficult to feed. As anticipated, the fuel used in the regeneration
was quite high and amounted to 10,000 BTU/lb. carbon. However, the carbon
loss, which was estimated at 8%, was not unusually high compared with
those observed in other regeneration periods.
In comparing the regeneration operating data in Tables 7 and 8 it is
evident that invariably, the regeneration of Filtrasorb 400 carbon pro-
ceeded at a relatively shorter time with less fuel consumption than that
of Filtrasorb 300. This difference in operating data was attributed to
33
-------
TABLE 7
FURNACE OPERATING CONDITIONS DURING REGENERATION
OF 300 SYSTEM CARBON
Column Designation
Before Regeneration
After Regeneration
Regeneration time (hours)
Furnace Loading (Ibs. carbon/hr.)
Steam Used (Ibs./lb
. carbon)
Fuel Used (BTU/lb. carbon)
Average Hearth Temp
Hearth No.
. (°F)
1
2
3
4
5
6
Afterburner
Carbon Loss
01
> *
I IDA
II1B
121
57.5
.77
10,000
640
845
1090
1635
1660
1660
1700
8
1 1 IDA
III1B
91
76.5
.69
6350
640
840
880
1425
1500
1625
1750
8
II1A
II2B
99
71
.74
6160
760
935
1160
1665
1755
1730
1770
4.8
III1A
III2B
87
89
.59
5206
705
870
1049
1618
1758
1776
1782
10.4
II2A
III3B
95
81.6
.64
5080
691
857
1039
1631
1733
1689
1788
7.3
III2A
82
97
.54
4125
644
810
973
1575
1713
1740
1764
6.5
II3A
II4B
80
99
.52
5219
630
802
964
1514
1681
1708
1277
6.9
III3A
III4B
96
82
.64
6503
602
827
1015
1616
1731
1753
1440
7
Average
93.8
81.7
.64
6080
664
848
1021
1585
1691
1710
1659
6.8
Unit Conversions:
Ib x .454 = kg
BTU/lb x .556 = cal/g
(°F-32) .555 = °C
-------
TABLE 8
FURNACE OPERATING CONDITIONS DURING REGENERATION
OF 400 SYSTEM CARBON
Before Regeneration
Column Desngnatnon After Regeneration
Regeneration Time (hours)
Furnace loading (Ibs. carbon/hr)
Steam used (Ibs./lb. carbon)
Fuel used (BTU/lb. carbon)
Average Hearth Temp. (°F)
Hearth No. 1
2
3
4
5
6
Afterburner
Carbon Loss, %
IVOA
IV1B
72
92.5
.49
5500
620
860
1070
1725
1840
1770
1820
11
VOA
V1B
78
85.5
.62
5350
630
800
1050
1700
1760
1780
1850
6
IV1A
IV2B
82
81
.65
5690
650
830
1000
1580
1730
1720
1775
8.4
VIA
V2B
97
71
.74
4660
635
800
955
1540
1645
1595
1750
6.3
IV2A
IV3B
78
93
.57
4576
641
808
966
1571
1724
1726
1796
6.6
Average
81.4
84.6
.61
5155
635
820
1008
1623
1740
1718
1798
7.6
co
01
Unit Conversions:
Ib x 0.454 = kg
BTU/lb x .556 = cal/g
(°F-32) x 0.555 = °C
-------
the fact that the regeneration of the 400 carbon always followed the
regeneration of the 300 carbon. Thus, whatever furnace operating
difficulties encountered during the regeneration of the 300 carbon were
corrected before the regeneration of the 400 carbon.
As indicated in Table 7, before the baghouse was installed for air
pollution control, the afterburner had been maintained at a temperature
ranging from 927°C (1700°F) to 1010°C (1850°F). During the last two
regenerations of the 300 carbon the baghouse system was used in series
with an afterburner. In this operating mode, the afterburner temperature
was reduced to an average of 738°C (1360°F), instead of the previous
average level of 971°C (1780°F), since the afterburner was then used
primarily for odor control.
Air Pollution Control Emission Data
The gases discharged from the top hearth of the regeneration furnace
contained both fine carbon particulates and obnoxious smelling sub-
stances. These air pollutants were controlled through an air pollution
control system consisting of a baghouse for particulate removal and an
afterburner, operated in series with the baghouse, for odor control. The
use of this air pollution control system, which was placed in operation
for the first time during the regeneration of column II 3A, has made it
possible to fully comply with particulate emission and other discharge re-
quirements of the Los Angeles County Air Pollution Control District.
During the first 48 hours of operation of the baghouse, the filter ratio
ranged from 12.7 to 13.7 l/sec/m2 fabric area (2.5 to 2.7 cfm gas/ft2)
and averaged 13 l/sec/m2 (2.55 cfm/ft2). The corresponding average pres-
sure drop through the baghouse was 4.06 cm (1.6 in) of water column (W.C.).
In the remaining 32 hours of the regeneration, the pressure drop through
the baghouse increased markedly with a range of 9.14 to 13.7 cm (3.6 to
5.4 in). As expected, the filter ratio decreased correspondingly to an
average of 10.2 l/sec/m2 (2 cfm/ft2 fabric). It should be pointed out
that at the start of the baghouse operation when the pressure drop through
the filters was low, the reverse air jet cleaning cycle was made longer.
The reverse jet was not operated during the first 6 hours in order to build
up sufficient dust mat on the fabric filters which would improve the par-
ticulate removal efficiency. As the pressure drop increased to 7.6 cm
(3 in), the cyclic timer was activated to pulse initially every 45 seconds.
Thereafter, as the resistance increased, the cleaning cycle was increased
to every 2 seconds intervals.
In the first day of operation of the air pollution control system, it
became apparent that condensation was a serious problem in the baghouse,
especially in the evening when the baghouse temperature ranged from 46.1 to
48.9°C (115 to 120°F). Because of the moist environment within the bag-
house, blinding of the fabric filters occurred as indicated by the high
headless of 13.7 cm (5.4 in) through baghouse. The condensation problem
was corrected by partially insulating the ductworks and the baghouse, thus
maintaining about 93.3°C (200°F) in the baghouse during the evening.
36
-------
Maintaining a high temperature in the baghouse, as long as it is below
the fabric filter bags operating temperature of 218.3°C (425°F) is
desirable for two reasons. Firstly, as previously mentioned, conden-
sation with its attendant problems could be obviated. Secondly, since
the filtered gases from the baghouse are subsequently conveyed to the
afterburner for odor control, less amount of fuel would be required
in operating the afterburner at 704.4 to 760°C (1300-1400°F). In sub-
sequent regenerations, the ductworks and the baghouse were completely
insulated with fiberglass, thus making it possible to maintain baghouse
temperature range of 148.9 to 162.8°C (300 to 325°F). While it is
advantageous to have a high temperature in the baghouse, due precau-
tion must be exercised to prevent the temperature from rising to within
10 to 37.8°C (50-100°F) of the critical fabric filter design temperature.
Thus, to minimize the danger of burning the filter bags, the baghouse
inlet was equipped with a valved side connection for dilution air addition.
Under normal furnace operating conditions, the dilution air inlet valve
was maintained closed. However, when the baghouse temperature increased
beyond 162.8°C (325°F), which could be caused by a disruption of the car-
bon feed rate to the furnace, the dilution air valve was manually opened
for such a duration as required to restore the baghouse temperature with-
in 148.9 to 162.8°C (300-325°F).
The performance of the various components of the air pollution control
system was evaluated during two regenerations by test engineers from
a local testing laboratory. During the first field evaluation of the
air pollution control system, an engineer from the Los Angeles County
APCD was also present to ascertain that correct sampling and testing
procedures were followed. In evaluating the system, the gas at the
inlet to the baghouse and at the outlet of the afterburner were tested
for gas flow rate, temperature, parti oil ate matter, volatile hydrocarbons,
oxygen, carbon dioxide, carbon monoxide, water vapor, oxides of nitrogen,
oxides of sulfur and odor number. The gas from the baghouse outlet was
subjected to the same tests as mentioned above, except for oxides of
nitrogen and odor number.
The gas flow measurements were made with a standard pilot tube and an
inclined manometer. Temperatures were measured with a chromelalumel
thermocouple and a portable potentiometer. Gas velocity and temperature
traverse measurements were made at the afterburner outlet only. The
particulate matter was collected using wet impingement in Smith-Green-
berg impingers followed by Gelman filter holders equipped with glass
fiber backup filters. Glass probes provided with glass ball joints were
used as sampling probes. Samples for the determination of volatile
hydrocarbons and fixed gases were taken in evacuated stainless steel
tanks. Sulfur dioxide was collected, using a filter, followed by an im-
pingement train containing hydrogen peroxide. Oxides of nitrogen were
analyzed, using the phenol-disulfonic acid method. Orsat analyses were
performed on the hydrocarbon samples and on the nitrogen oxide flasks.
A summary of the emission data from the various components of the air
pollution control system is presented in Table 9. During the regeneration
of Column IISA, particulate emission, as well as odor from the after-
burner, were very low and averaged only .041 kg/hr (0.09 Ib/hr) and
37
-------
TABLE 9
SUMMARY OF AIR POLLUTION CONTROL SYSTEM PERFORMANCE
Parameters
1 . Parti cul ate Matter
Concentration, grains/SCF
Emission rate, Ibs/hr
2. Oxides of Nitrogen, (NOY)
Concentration, ppm dry
Emission rate, Ibs/hr
3. Oxides of sulfur (SO?)
Concentration, ppm S02
Emission rate, Ibs/hr
4. Hydrocarbons
Concentration, ppm C
Emission rate, Ibs/hr C
5. Carbon Monoxide (CO)
Concentration, % Volumedry
6. Odor
Odor units /SCF
7. Gas Flow
Temperature, °F
SCFM
APCD
Emission
Limit
0.20
1.00
225
0.2%
Regeneration of Column
II3A
Baghouse
Inlet Outlet
0.987 0.298
0.890 0.266
94
3900
0.88
0.56
20 ,000
345 140
104 104
After-
Burner
Outlet
0.046
0.09
166
0.48
217
0.88
660
0.50
0.20
10
1000
392
Regeneration of Column
III3A
Baghouse
Inlet Outlet
1.82 0.47
2.17 0.80
40
0.028
nil nil
740 561
0.20 0.21
1.36 .86
20 ,000
352 159
139 198
After-
Burner
Outlet
0.075
0.24
180
0.40
149
0.57
oil
0.11
20
1148
376
co
00
Unit conversions: Ib/hr x 0.454 = kg/hr
(F°-32) x .555 = °C
-------
10 odor units/scf, respectively. Based on particulate emission rate
data, the baghouse removed only 70% of the dust, which was significantly
below the design capability of about 99% removal. The total actual
weight of dust collected from the baghouse over the 80 hour period of
regeneration was 26.8 kg (59 Ibs), which was about 20% more than the
removal estimated from the emission rate data. This discrepancy, how-
ever, is not unreasonable, considering that the emission data repre-
sented samples collected over 45 to 60 minutes sampling period. In
addition, difficulties were encountered in collecting representative
samples from the baghouse inlet due to excessive condensation in the
duct. The dust collected,from the cyclone separator amounted to 6.36 kg
(14 Ibs).
In Table 9 are also shown the results of the second evaluation of the
air pollution control system during the regeneration of column III3A.
The data indicated that all emission parameters, such as particulate
matter, oxides of nitrogen, oxides of sulfur, hydrocarbon and odor num-
ber, were all in full compliance with the local air pollution control
requirements.
SYSTEM PERFORMANCE
Organic Removal
The COD test, both total and dissolved, was used in this study as the
primary parameter in evaluating the performance of the activated carbon
pilot plant in the removal of organic materials from the activated sludge
plant effluent. The total COD (TCOD) removal patterns in each column
for all adsorption sequences are presented in Figures 15 through 18 for
the Filtrasorb 300 and Figures 19 through 20 for the Filtrasorb 400 system.
In these figures, the top curves represent the column influent (secondary
effluent) quality and the bottom curves designated with letter "B" repre-
sent the two-stage system effluent. The bands between the curves represent
the amount of TCOD removed in each carbon column. The vertical bands be-
tween each adsorption run or sequence designate the time when the first
stage or "A" position contactor was taken off-stream for regeneration. In
examining these figures, it is evident that the major portion of the TCOD
was removed in the first stage contactor as reflected by the width of the
top bands compared with the bands between curves A and B. This is expected,
since the first stage carbon column acted as an efficient filter capable
of removing substantial amounts of the influent suspended material.
In determining the amount of COD applied and removed by each contactor,
the areas bounded by the curves were determined using a planimeter. Al-
though not presented in this report, DCOD removal patterns similar to
those for TCOD, were also plotted from which the DCOD loading and removal
capacities were calculated from the pianimetered area.
The effect of repeated thermal regeneration or adsorption cycles on the
DCOD removal capacity of the two-stage carbon adsorption systems are pre-
sented in Figure 21. For the Filtrasorb 300 system, the carbon capacity,
expressed in kg DCOD removed/kg carbon, has decreased about 25% after
39
-------
45.
O
RUN NO. I
ICOA-^IHOB
DE OB EFFLUENT
6.2 10.1 13.7 17.2 21.3
CUMULATIVE VOLUME TREATED (M6)
FIGURE 15: COD REMOVAL IN THE FILTRASORB 300 SYSTEM (Run !•)
25.
27.8
-------
60
O
U
o
o
I-
/fcptJ? I
UJ
I o
IB-EFFLUENT
0~Vm IB
EFFLUENT
o
o
U.
o
a:
UJ
u
o
UJ
(T
JAN.
FEB. MARCH
APRIL
MAY
JUNE
27.8
FIGURE
30.8
53.6
34.8 38.6 41.2 45.4 49.6
CUMULATIVE VOLUME TREATED (MG)
16: COD REMOVAL IN THE FILTRASORB 300 SYSTEM (Run 2-3)
55.6
-------
60
PO
! I
RUN NO. 5
I2A-»HI2B
55.6
59.9
63.3
78.5
82.6
66.2 70.3 74.3
VOLUME TREATED (MG)
FIGURE 17: COD REMOVAL IN THE FILTRASORB 300 SYSTEM (RUN 4-5)
84.6
-------
OJ
RUN NO. 8
m3A-*I4B
84.6 89.4 92 96.3 100.5 102.4
CUMULATIVE VOLUME TREATED (MG)
106.6
110.7
115.1
FIGURE 18= COD REMOVAL IN THE FILTRASORB 300 SYSTEM (RUN 6-8)
-------
60
INFLUENT | |
DEC.
JAN.
FEB. MARCH
APRIL
MAY
JUNE
115.1
119.4
CUMULATIVE VOLUME TREATED (MG)
FIGURE 18: CONTINUED
-------
RUN NO. 3
IA-»VIB
RUN NO. I
TST OA-~3TOB
RUN NO. 2
TOA-^tSTIB
ETOA
EFFLUENT
TOA
EFFLUENT
EFFLUENT
MAR. APR. MAY NUN. JUL. AUG. SEP I OCT. I NOV. I DEC.I JAN. I FEB.I MAR.I APR.
14.3 22.0 29.5 36.0
CUMULATIVE VOLUME TREATED(MG)
41.7
50.0
FIGURE 19: COD REMOVAL IN THE FILTRASORB 400 SYSTEM (Run 1-3)
-------
cn
fill
RUN NO. 4
TTA --*• TET 2B
RUN NO. 6
RUN NO. 5
3BT 2A-*
2 A
EFFLUENT
EFFLUENT !£
EFFLUENT
MAY JUN JUL AU6 SEPT OCT NOV DEC JAN FEB MAR. APR. MAY NUN
50.0 66.5 74.8 81.5 89.9 96.7
CUMULATIVE VOLUME TREATED (MG)
104.1
FIGURE 20: COD REMOVAL IN THE FILTRASORB 400 SYSTEM (Run 4-6)
-------
0.6
O 0.5
CO
cr
o
.0
-x 0.4
Ul
o
UJ
^ 0.3
O
>'°-2
t
1
z
o
CO
cr
0.
TWO - STAGE ADSORPTION SYSTEM:
EMPTY-BED CONTACT TIME =40 MIN.
HYDRAULIC LOADING a 3.5 gpm./ft.4
-FILTRASORB 400
(12X40 MESH)
FILTRASORB 300
(8 X 30 MESH)
ADSOPTION CYCLE
FIGURE 21: EFFECT OF ADSOPTION CYCLE ON CARBON CAPACITY
-------
four adsorption cycles. The decrease in capacity, however, tends to
taper off during the last two adsorption cycles. The data indicate that
the carbon capacity for DCOD appears to reach a steady-state level of
about 0.26 kg DCOD removed/kg carbon after the third adsorption sequence.
In Figure 21 is also plotted the carbon capacity for the Filtrasorb 400
system. As indicated in the figure, the same decreasing trend in the DCOD
removal capacity is evident following repeated thermal regenerations. The
carbon capacity of the 400 system carbon has decreased about 23% from an
initial level of 0.39 kg DCOD/kg carbon to an apparent steady-state level
of 0.3 kg DCOD/kg carbon after three adsorption cycles. In comparing
the data for the two types of carbon in Figure 21, it is shown that at
the same adsorption cycle, the 400 system carbon has approximately 13%
more DCOD removal capacity than the 300 system carbon. Moreover, it is
of interest to note that there is a considerable loss in the DCOD removal
capacity during the first regeneration. For instance, for the 300 system,
about 69% of the total capacity loss in the entire four adsorption cycles
occurred in the first adsorption cycle. The corresponding loss in carbon
capacity in the 400 system carbon during the first adsorption cycle was
estimated at 67%.
In Figure 22 is presented the relationship between carbon dosage and ad-
sorption cycle for the two types of carbon. For the 300 system carbon,'
the carbon dosage increased from .038 kg/m3 (320 Ibs carbon/million gal-
lons) during the first cycle to a projected level of about .066 kg/rrp
(550 Ibs carbon/million gallons) at the end of the fourth cycle. The
carbon dosage for the 400 system carbon, which was about 10% less than
that of the 300 system, increased from .034 kg/m3 (280 Ibs carbon/million
gallons) to an estimated level of .06 kg/m3 (500 Ibs carbon/million gal-
lons), after three adsorption cycles. Table 10 presents the calculation
for carbon dosage and capacity for Filtrasorb 300 carbon.
The effect of carbon particle size on the TCOD and DCOD removal through
the two parallel two-stage systems are presented in Figures 23 and 24.
The plotted points were obtained by dividing the amount of COD applied
and removed by each system in each adsorption sequence by the total weight
of carbon in that system. It is apparent from these figures that the COD
removal efficiencies of the two-stage systems remain practically constant
through four adsorption cycles in the 300 system and three adsorption cycles
in the 400 system. Furthermore, while both systems are about equal in
terms of COD removal efficiency, the 400 system carbon shows consistently
a slightly better performance compared to the 300 system carbon.
In evaluating the data relating to the degree of COD removal in each stage
of the parallel two-stage systems, it was found that the major portion of
both TCOD and DCOD were removed in the "A" position or first stage carbon
column. The .first stage carbon columns removed on the average 76.4% of
the TCOD and 64.3% DCOD whereas the second stage columns removed only 23.5%
TCOD and 35.7% DCOD. The significantly higher TCOD removal in the first
stage is consistent with the fact that in two-stage systems operated on
downflow mode, the first stage carbon column acts as an effective granular
filter bed for the removal of suspended solids. Moreover, the observed low
DCOD removal capacity in the second stage, which amounted to about 56% of
that in the first stage, could be attributed to the lower DCOD loading applied
to the second stage column. Table 11 presents the summary data on the
48
-------
600
<£>
500
CD
§400
tr
<
u
in
.a
.300
Ul
8
CO
o
.200
O
CD
CC
<
O
100
TWO-STAGE ADSORPTION SYSTEM
EMPTY-BED CONTACT TIME = 40MIN.
HYDRAULIC LOADING = 3,5 gpm/ft.*
FILTRASORB 300
(8 X 30 MESH)
FILTRASORB 400
(12 X 40 MESH)
0123
ADSORPTION CYCLE
FIGURE 22: EFFECT OF ADSORPTION CYCLE ON CARBON DOSAGE
-------
TABLE 10
CARBON CAPACITY AND DOSAGE DATA
Regeneration Level
1. Total Volume Treated in A and B
Position, (million gallons)
Column II
Column III
Total
2. DCOD Removed in A and B
Position (Ibs.)
Column II
Column III
Total
3. Carbon Capacity3, Ib DCOD
Removed per Ib. Carbon
4. Carbon Dosage'3 Ibs carbon
per million gallons
0
26.324
38.564
64.888
2742
2176
4918
0.35
322
1
28.594
24.609
53.203
2145
1889
4034
0.29
523
2
27.550
28.714
56.264
2019
1733
3752
0.27
495
3
19.830
28.484
48.313
1615
2036
3651
0.26
'576
6960 Ibs. carbon per column
a Carbon capacity = Total DCOD removed -7-2(6960)
b Carbon dosage = (Total DCOD Removed) 2* -v (Total Volume x Carbon
* 1.5 for regeneration level "0" capacity)
Unit conversions: Ib x 0.454 = kg
mil gal x 3785 = cu m
50
-------
o
CD
CC
5
to
£ 2
O
UJ
tu
CC
O
o
o
FILTRASORB 400-
FILTRASORB 300
TCOD APPLIED (Ibs./lb. CARBON
FIGURE 23: EFFECT OF CARBON PARTICLE SIZE ON TCOD
REMOVAL
51
-------
en
ro
1.3
1.2
CO
UI
x,.o
— .9
o :
m
o
-7
.2 .6
UJ -3
>
O
2 4
UJ
tr
8-3
O
Q.2
.1
.2
FILTRASORB 400
FILTRASORB 300
4 .6 .8 1.0 1.2 1.4
DCOD APPLIED (Ibs./lb. CARBON IN THE SYSTEM)
1.6
1.8
FIGURE 24= EFFECT OF CARBON PARTICLE SIZE ON DCOD REMOVAL
-------
TABLE 11
DISTRIBUTION OF COD REMOVAL IN THE
TWO-STAGE SYSTEMS
Column
Position
"A"
(1st Stage)
"B"
(2nd Stage)
Total
Filtrasorb 300 System
TCOD
%
Re-
moved
76.4
23.6
100
Ib removed
Ib. c
0.394
0.122
0.516
DCOD
%
Re-
moved
64.3
35.7
100
Ib removed
Ib. c
0.180
0.100
0.28
Filtrasorb 400 System
TCOD
%
Re-
moved
•76.5
23.5
100
Ib removed
Ib. c
0.482
0.148
0.630
DCOD
%
Re-
moved
64.3
35.7
100
Ib removed
Ib. c
0.238
0.132
0.370
en
Unit Conversion: Ib x .454 = kg
-------
average COD removal efficiency and carbon capacity in each stage of the
two-stage systems.
Effluent Water Quality
In Table 12 are presented the average water quality characteristics of the
influent and effluent from the two-stage systems. The data from this
table shows that the two-stage systems were effective in producing carbon
effluents with low levels of organic content, as measured by both COD and
TOC. The TCOD removal has averaged 74.7% in the 300 system and 78.6% in
the 400 system. It is further evident from the tabulated data that the
400 system was slightly more effective than the 300 system in the reduction
of all the process parameters evaluated. The performance of the two-
stage adsorption systems in each adsorption sequence is summarized in
Table 13 for the Filtrasorb 300 system and Table 14 for the Filtrasorb 400
system.
The secondary effluent used during the first two adsorption cycles of the
two-stage systems was partially nitrified with an average nitrate concen-
tration of 3.7 mg/1 N. Thereafter, the average nitrate concentration of
the carbon influent was only 1.2 mg/1 N as a result of a change in the
mode of operation of the activated sludge plant. While the level of ni-
trate in the carbon influent had no apparent effect on the overall carbon
effluent quality, one carbon effluent quality characteristics of signi-
ficance is the sulfide concentration. Towards the end of the third ad-
sorption cycle when the carbon column influent nitrate concentration
averaged only 1.0 mg/1 N, hydrogen sulfide odor became evident in the car-
bon effluent. Results of sulfide determination, which was not performed
until the last few days of the third adsorption cycle, indicated averaged
total sulfide levels in the carbon effluent of 0.2 mg/1 S in the Filtra-
sorb 300 system and 0.5 mg/1 S in the Filtrasorb 400 system. In previous
carbon column operations in which the column influent nitrate concentra-
tion was about 4 mg/1 N or greater, no discernible HpS odor was detected
in the carbon effluent. This absence of detectable F^S odor in the
carbon effluent was attributed to the presence of sufficient amount of
influent nitrate concentration which inhibited sulfate reduction within
the carbon columns.
During the course of the study, virus analysis was performed on the two-
stage system effluent. The results of the virus analysis, without any
distinction as to the type of carbon used for the two-stage system, in-
dicated that about 39.4% of the 132 samples analyzed were positive for
virus. A summary of the virus testing program for the two-stage system,
as well as the other carbon contacting systems, are presented in Table 15.
It is evident from the data presented that while granular activated car-
bon columns effect some degree of virus removal, they cannot be relied
upon for consistent virus removal.
COST ESTIMATE
The economic analysis is based on the carbon treatment of the Pomona ac-
tivated sludge plant effluent for an average design flow of 0.44 m3/sec
54
-------
TABLE 12
AVERAGE WATER QUALITY CHARACTERISTICS
•
Parameters
Suspended Solids, mg/1
Total COD, mg/1
Dissolved COD, mg/1
TOC, mg/1
Dissolved TOC, mg/1
Turbidity, JTU
MBAS, mg/1
Nitrate-N, mg/1
Color
Duration of run
(months)
Total volume treated,
million gallons
Filtrasorb 300 (8 x 30 mesh)
Influent
12.2
38.7
23.7
12.3
9.0
7.6
0.36
2.7
34.6
Effluent
2.1
9.8
6.4
4.0
3.2
1.8
.04
1.2
5.8
Removal , %
82.9
74.7
73
67.5
64.4
76.3
88.9
55.6
83.2
33
120.53
Filtrasorb 400 (12 x 40 mesh)
Influent
12
40.2
24.3
10.8
7.8
7.3
.24
3.3
36.2
Effluent
2
8.6
6.2
2.6
1.8
1.4
.04
1.4
4.6
Removal , %
83.3
78.6
74.5
75.9
76.9
80.8
83.3
57.6
87.3
26
97.61
en
ui
Unit conversions: mil gal x 3785 = cu m
-------
TABLE 13
PERFORMANCE OF THE FILTRASORB 300 TWO-STAGE ADSORPTION SYSTEM
Ads.
Cycle
1
2
3
4
Run
No.
1
2
3
4
5
6
7
8
Operating
Sequence
I IDA I HOB
I IDA II1B
II1A III1B
III1A II2B
II2A III2B
III2A II3B
II3A III3B
III3A II4B
Two-Stage System Removal, %
Total
COD
75.8
78.6
75.6
75.6
75.2
72.6
70.3
73
Diss.
COD
71.2
73.4
74.6
73.9
69.9
67.3
70.7
67.2
Total
TOC
77.7
71.8
76.6
81.4
71.1
62.6
38.6
-
Diss.
TOC
76.5
67.6
76.8
78.9
70.5
50.5
35.2
-
MBAS
85.8
87.6
88.5
89.7
84.3
93.3
77.4
78.7
COLOR
82.8
68.6
89.2
88.1
83.5
79.2
80.2
91.2
TURB
72.2
74.3
73.8
72.0
79.0
73.8
64.3
81.5
NH3-N
1.1
3.2
3.0
3.6
0
0
0
0
N03-N
41.6
76.9
8.1
81
57.4
68.6
50.6
80.1
en
-------
TABLE 14
PERFORMANCE OF THE FILTRASORB 400 TWO-STAGE ADSORPTION SYSTEM
Ads.
Cycle
1
2
3
Run
No.
1
2
3
4
5
6
Operating
Sequence
IVOA VOB
VOA IV1B
IV1A V1B
VIA IV2B
IV2A V2B
V2A IV3B
Two-Stage System Removal , %
Total
COD
76
77.5
80.0
75.6
78.6
74.4
Diss.
COD
72.4
72.5
79.8
74.2
72.9
70.5
Total
TOC
75.2
67.9
82.5
82.4
75.2
66.2
Diss.
TOC
76.8
70.9
84.8
79.8
75.6
65.5
MBAS
84.8.
90.7
89.7
89.9
82.3
-
COLOR
86.2
82.1
91.7
91.0
86.8
87.9
TURB.
78.9
76.0
77.9
73.0
81.2
75.5
NH3-N
1.6
4.7
5.6
8.6
0
1.8
N03-N
36
73.5
89.8
88.4
59.2
77.0
en
-------
TABLE 15
VIRUS REMOVAL DATA
Sample Type
Activated Sludge Plant
Effluent (unchlorinated)
Single-stage Carbon
Effluent (Filtrasorb 400)
Two-Stage Carbon Effluent
(Filtrasorb 400 and 300)
Four-stage Carbon Effluent
(Filtrasorb 400)
Sampling
Period
8/7/69 - 3/5/69
6/5/68 - 1/15/69
3/18/70 - 11/29/72
1/15/69 - 3/11/70
Contact Time
Minutes
10
40
40
Hydraulic Loading
gpm/ft2
7
3.5
7.0
No. of
Samples
Tes ted
28
21
132
59
% of
Samples
Positive
82.1
52.4
39.4
45.8
en
CO
Unit conversion: gpm/ft2 x .0407 = cu m/day/m2
-------
(10 mgd) and a peak flow of 0.61 m3/sec (14 mgd). The design parameters
for sizing the carbon contactors and the carbon regeneration furnace are
presented in Table 16. The table also includes assumed unit costs for
carbon and other direct costs for estimating the operation and mainte-
nance (0/M) costs.
In the preparation of the cost estimate, various published reports (2,7,8,9,
10,11) were consulted. The capital costs obtained from literature were
adjusted to the EPA sewage treatment plant construction cost index of 185.
The carbon contacting system consists of six trains of two-stage, packed-
bed, downflow carbon columns designed for a hydraulic surface loading of
4.75 1/sec/m2 (7 gpm/ft2) and an empty-bed contact time of 20 minutes per
stage. In order to facilitate uninterrupted operation during carbon re-
generation, two spare vessels are provided in addition to the six trains
of two-stage carbon columns. One of these spare vessels is initially
charged with 36,318 kg (79,900 Ibs) of Filtrasorb 300 (8x30 mesh)carbon while
the other vessel is reserved for spent carbon storage. Thus, the initial
carbon charge is equivalent to the effective volume of 13 carbon contactors.
Table 17 presents a summary of the estimated carbon treatment cost. The
cost breakdown shows that capital amortization represents about 58% of the
total treatment cost of 11.52<£/1000 gallons. Moreover, the cost of make-
up carbon and the cost of operating and maintenance labor represent, res-
pectively, 35.5% and 40.4% of the total 0/M costs.
59
-------
TABLE 16
DESIGN CRITERIA AND UNIT COSTS FOR ECONOMIC ANALYSIS
CARBON CONTACTING SYSTEM
Empty-bed contact time, min/stage 20
Hydraulic surface loading, gpm/ft2 7
Backwash volume, % of plant flow 2
Carbon dosage, lb C/MG 550
Carbon Regeneration loss, %.. 7
CARBON COSTS ,
Filtrasorb 300 ( 8x30 mesh ) $/lb 0.40
OPERATING COSTS
Power, tf/KWH 2
Fuel, tf/Therm. 8
Carbon Regeneration Fuel consumption, BTU/lbc 6,000
Backwash Water, <£/1000 gallons 3
Operating Labor, 4 at $12,000/yr 48,000
Maintenance Labor, 1 at $10,000/yr 10,000
Laboratory Personnel, 1 at $14,000/yr 14,000
Maintenance Materials, $ 5,000
CAPITAL COSTS
All equipment costs were amortized at 6% for 25 years.
60
-------
TABLE 17
ESTIMATED GRANULAR ACTIVATED CARBON TREATMENT COSTS*
( 10 MGD )
CAPITAL COSTS ( 1000 of $ )
Influent Pumping 175
Initial Carbon Charge 415
Carbon Contacting System 2,100
Carbon Regeneration System 410
Total Capital Cost 3,100
Amortized Cost ( <£/1000 gallons ) 6.64
OPERATING AND MAINTENANCE COSTS ( £/1000 gallons )
Carbon Make-up 1.73
Backwash Water 0.06
Power 0.72
Fuel 0.26
Operating and Maintenance Labor 1.97
Maintenance Materials 0:. 14
Total 0/M Costs 4.88
Total Treatment Cost, ( if/1000 gallons ) 11.52
* Based on estimated EPA sewage treatment construction cost index
of 185 - October 1973.
61
-------
REFERENCES
1. Hassler, J.W., "Activated Carbon", Chemical Publishing Co.,
Inc., New York (1963)
2. Cover, A.E. and Pieroni, L.J., "Appraisal of Granular Carbon
Contacting: Phase I, Evaluation of the Literature on the use
of Granular Activated Carbon for Tertiary Waste Water Treat-
ment and Phase II, Economic Effect of Design Variables",
Report No. TWRC-11, May 1969, Robert A. Taft Water Research
Center, U.S. Department of the Interior, Cincinnati, Ohio
3. Parkhurst, J.D., Dryden, F.D., McDermott, G.D., and English,
J.N., "Pomona Activated Carbon Pilot Plant", Journal Water
Pollution Control Federation, Vol. 39, No. 10, part 2, 1270
(October 1967)
4. "Standard Methods for the Examination of Water and Wastewater",
12th Edition, American Public Health Association, New York (1965)
5. "FWPCA Methods for Chemical Analysis of Water and Wastes",
Federal Water Quality Administration, Cincinnati, Ohio (November
1969)
6. Juhola, A.J. and Tepper, F., "Regeneration of Spent Granular
Activated Carbon", Report No. TWRC-7, February 1969, Robert
A. Taft Water Research Center, U.S. Department of the Interior,
Cincinnati, Ohio
7. DiGregorio, D., "Cost of Wastewater Treatment Processes", Re-
port No. TWRC-6, December 1969, Robert A. Taft Water Research
Center, U.S. Department of the Interior, Cincinnati, Ohio
8. Smith, R. and McMichael, W.F., "Cost and Performance Estimate
for Tertiary Wastewater Treatment Processes", Report No. TWRC-9
June 1969, Robert A. Taft Water Research Center, U.S. Depart-
ment of the Interior, Cincinnati, Ohio
9. Hopkins, C.B., Weber, W.J., Jr., and Bloom, R., Jr., "Granular
Carbon Treatment of Raw Sewage", Water Pollution Control Re-
search Series ORD 17050 DAL 05/70, U.S. Department of the In-
terior, Federal Water Quality Administration (May 1970)
62
-------
10. Cover, A.E., Johnson, G.F., and Rymer, E.V., "Appraisal of
Granular Carbon Contacting: Phase III, Engineering Design
and Cost Estimate of Granular Carbon Tertiary Waste Water
Treatment Plant", Report No. TORC-12, May 1969, Robert A.
Taft Water Research Center, U.S. Department of the Interior,
Cincinnati, Ohio
11. "Process Design Manual for Carbon Adsorption", U.S. Environ-
mental Protection Agency (October 1973)
63
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REPORT NO.
EPA-600/2-78-170
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Two Stage Granular Activated Carbon
Treatment
5. REPORT DATE
September 1978 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
'. AUTHOR(S)
Leon S. Directo, Ching-lin Chen,
and Robert P. Mlele
8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Los Angeles County Sanitation District
Whittier, CA 90607
10. PROGRAM ELEMENT NO.
1BC611
11. CONTRACT/GRANT NO.
14-12-150
12. SPONSORING AGENCY NAME AND ADDRESS Cin. , OH
Municipal Environmental Research Laboratory—
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati. Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
TT T r» a 1
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer:
Irwin J. Kugelman (513) 684-7633
16. ABSTRACT
Two 6.3 I/sec (0.15 mgd), two-stage, packed-bed, downflow
granular activated carbon pilot plants were operated continuously
for 33 months using unfiltered and unchlorina'ted activated sludge
plant effluent. The main objective of the study was to compare the
performance of granular carbons of different particle size. The
data collected during this study has demonstrated the efficacy of
the two stage carbon adsorptive system in consistently producing
effluent of excellent overall quality. Effluent averaged 6-7 mg/1
DCOD and 2 mg/1 S.S.
The carbon capacity with the 8x30 mesh .carbon decreased about
25% after four adsorption cycles, resulting in an apparent steady
state capacity of 0.26 Ibs. DCOD removed/lb carbon. A 23% decrease
in carbon capacity occurred after three adsorption cycles with the
12x40 mesh carbon. The 12x40 carbon has about 13% more DCOD removal
capacity than the 8x30 carbon. While the smaller carbon showed
slightly higher treatment capacity than the larger carbon, the
latter has lower initial cost, lower pressure loss and lower
regeneration loss. Thus the larger size carbon was more economical,
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Activated Carbon Treatment
Adsorption
Water Reclamation
Filtration
Tertiary Treatment
Regeneration
13B
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
74
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
64
aUSGPO: 1978-657-060/1480 Region 5-11
------- |