United States
Environmental Protection
Agency
Robert S. Kerr Environmental Resear
Laboratory
Ada OK 74820
Research and Development
EPA-600/S2-81 -159 Dec. 1981
Project Summary
Full-Scale Demonstration of
Industrial Wastewater
Treatment Utilizing Du Font's
PACT® Process
Harry W. Heath, Jr.
Operating and cost data on sjtjajjtup
and the initial 30 months of operation
of a 150,000-mVday (40-MGD) in-
dustrial wastewaterttae.a3rjMMit plant
using Du Font's PACT® proqggiftfte,
reported. The PACT process effectively
provides both secondary and tertiary
treatment by adding powdered acti-
vated carbon to the aerator in an
activated sludge process.
In 1978 and 1979, performance
exceeded expectations in terms of
BOD5 and color removal, 96% and
68%, respectively. Removal of Dis-
solved Organic Carbon (DOC) was
82%, approximately equal to the design
value. Removals were consistent even
though the medium-strength waste
contained many relatively toxic, non-
biodegradable organics and varied
drastically in composition. EPA desig-
nated hazardous substances in the
waste were removed very well. A
synergistic enhancement of biological
activity in the presence of activated
carbon was observed.
Operation at over 25 days sludge
age reduced the required carbon dose
from more than 170 mg/liter to 120
mg/liter.
Activated carbon was regenerated
from wasted PACT sludge in a multiple-
hearth furnace. Carbon was regen-
erated at 80% yield with recovery of
63% of virgin carbon properties. Pro-
' i jxluction rate and quality recovery were
below expectations. Furnace perform-
.Wjfl W&ttiUfeby underdesigned f iltra-
~iafP% and solids conveying equipment
tlEGi
-------
combined treatment process in which
powdered activated carbon is added
directly to the aerator in a conventional
activated sludge process.
The treatment facilities would be
located at Du Font's Chambers Works
site at Deepwater, New Jersey and
would provide primary and combined
secondary/tertiary treatment for aqueous
waste from the Chambers Works. Total
investment for the WWTP was approxi-
mately $46,000,000.
The Chambers Works is a multiproducts
chemical plant manufacturing "Freon"
fluorocarbons; general organic inter-
mediates; dyes and dye intermediates;
petroleum additives; iso-cyanatesand
elastomeric products; general industrial
chemicals; sulfuric acid; and miscel-
laneous other products. The wastewater
is highly colored. Since most plant
processes are batch, the wastewater
periodically contains slug discharges or
organics, which results in a highly
variable organic load.
Chambers Works waste is strongly
acidic. During primary treatment, the
acid is neutralized with lime. The sludge
formed is settled in primary clarifiers,
filtered to give an approximately 50-
percent-solid cake, and disposed of in
an on-site, double-lined, secure landfill.
Effluent from primary treatment has a
pH of 6.0 to 8.0 and is the feed to the
PACT process. Average composition of
the PACT feed during 1978 and 1979 is
presented in Table 1.
The PACT process was developed as a
result of technical studies of alternative
treatment procedures for Chambers
Works wastewater. Results of the
studies can be generalized by saying
that a biological treatment process
appeared most economical but could
not consistently meet the BOD5 limit,
while physical-chemical treatment
processes were expensive and failed to
remove more than 60 percent of the
BOD5.
Physical-chemical methods generally
failed to remove simple organic mole-
cules, e.g., methanol and acetic acid,
which were most easily biodegraded.
Biological units failed to remove some
complex organics, including those
strongly coloring the wastes, and were
plagued by periodic upsets due to varia-
tions in total organic load and/or surges
of specific toxic compounds.
In an experiment, powdered activated
carbon was added to a biological reactor.
Results were so favorable that a program
was established to pursue this new
technology. A summary of typical oper-
ating data comparing a PACT unit with a
conventional activated sludge unit is
presented in Table 2. In addition to
reductions in color, DOC, and BOD5
removals, the spread of effluent values
around the average was more narrow
for the PACT unit.
The laboratory studies with Chambers
Works wastewater demonstrated that,
compared to a conventional activated
sludge system at equal sludge age of 8
days, the PACT process with 160 mg/l
carbon addition of one type carbon gave
consistently high BOD5 removal, in-
creased DOC removal, increased color
removal, eliminated foaming in the
aerator, improved sludge settling and
filtration properties, and protected the
microorganisms from periodic shock
loads of toxic compounds. There is a
synergistic relationship between the
bacteria and carbon in the PACT process
such that more organics become biode-
gradable in the presence of carbon. In
addition, the carbon provides an adsorp-
tion sink to minimize the effects of
sudden changes in feed concentrations
of organics, thereby decreasing the
number and duration of excursions of
effluent DOC concentrations.
Cost estimates for a PACT system
versus a conventional activated sludge
plant were within 10 percent of each
other. The extra PACT investment for
carbon handling and regeneration facil-
ities was mostly offset by additional
clarifier capacity and sludge disposal
facilities needed for the activated sludge
plant. Assuming a 75 percent regenera-
tion yield for powdered carbon, operating
costs for the PACT process were about
25 percent greater than for activated
sludge alone. In effect, the PACT process
at Chambers Works was expected to
Table 1. Average Feed to PACT
Parameter
Flow rate, rrf/min.
Soluble BODs, mg/l
kg/day at avg. flow
DOC, mg/l
kg/day at avg. flow
Color. APHA
pH
TSS, mg/l
Treatment System
Jan. 1978-Dec. 1979
24-month average
94.2
(24,900 gpm)
171
23,200
(51.000 Ib/day)
170
23. 100
(50,800 Ib/day)
1.530
6.0-8.0
37
Table 2. Typical Laboratory PACT Unit Performance* (October
Design
flowsheet
1OO
(26,400 gpm)
280
40,300
(88,700 Ib/day)
205
29,500
(65,000 Ib/day)
1,000
6.0-8.0
30
1975 -April 1976)
Treated Effluent
Parameter
Feed
PACT Unit Act. Sludge Unit
Color, APHA units
DOC, mg/l
BODs
Operating Conditions
Sludge Age, days
Carbon Dose, mg/l
Temperature, °C
1690 (690)**
171 (28)
300(21)
310
26 (13)
23
8.0
160
15-24
1020
57(27)
38
8.3
0
15-24
Aerator Hydraulic
, Residence Time, hours
6.1
5.7
*Values are average of 100 daily data points except for BODs which are averages of
36 data points.
^(Standard deviation).
-------
provide tertiary treatment quality ef-
fluent at a cost slightly higher than that
for secondary treatment alone.
Chambers Works Wastewater
Treatment Plant
A flow diagram of the plant showing
equipment arrangement is depicted in
Figure 1. Flow through the plant is by
gravity. PACT treatment starts at the
flow splitter where neutralized primary
effluent enters from four primary clari-
f iers. A slurry of regenerated plus virgin
makeup carbon is added to the waste-
water. Slurry flow rate is controlled to
maintain the desired carbon dosage.
Recycle activated sludge (RAS) is
combined with the primary effluent and
fed to the aerators. Each aerator provides
about 7 hours hydraulic residence time.
Static mixers in each aerator provide the
mixing necessary to suspend the PACT
sludge as well as oxygen transfer. Air is
supplied through headers extending
across the center of the tank along the
floor.
Bottom take-off lines from each
aerator combine in a common line to the
clarifier feed flowsplitter. Mixed liquor
is fed to a center well of each of two
secondary clarifiers from which it over-
flows into a flocculating zone. Treated
wastewater is discharged to a final
settling basin prior to discharge to the
Delaware River estuary.
Solids are wasted from the liquid train
via a thickener. Either aerator effluent
or RAS may be fed to this carbon
thickener. Supernatant from the
RR Track Carbon
I I I I I I I I I I I ^Unloading
RR Track
Key: AB
AS
C
CAW
F
FP
FS
FT
HP
L
LST
Afterburner
Acid Storage
Carbon Slurry Tank
Carbon Acid Wash Tank
Flocculator
Filter Press
Flow Splitter
Fuel Storage Tank
200 psig Filter Feed Pump
Lime Storage Silo
Lime Slurry Tank
Neutralization Tank
8,000 gpm Waste Water Feed Pump
Recycle Activated Sludge Pump
Primary Sludge Hold Tank
Treated
Effluent
To River
Primary Sludge
to Landfill
To Atmosphere
HP \l Carbon
Regener.
Furnace
Wastewater
Ditch
Carbon
Slurry
2°-3°
Sludge
Thick.
# 1
Clarifier
2,500,000
Gal.
# 1
Aeration
Tank
4.000,000
Gal.
# / Clarifier
1,000.000 Gal
# 2 Clarifier
1,000.000 Gal.
#2
Clarifier
2.500.000
Gal.
# 3 Clarifier
1,000.000 Gal.
#2
Aeration
Tank
4,000,000
Gal.
#3
Aeration
Tank
4.000.000
Gal.
# 4 Clarifier
1,000,000 Gal.
Figure 1. Flow diagram of Chambers Works wastewater treatment plant.
-------
thickener is returned to the aerators.
Approximately 10 percent slurry from
the thickener enters a filter feed tank,
from which it is fed to the PACT filter
press. The horizontal filter press is an
automatic unit containing 112 recessed
chambers. Filter cake from the press
drops into a bunker from which a multi-
stage conveyor system moves it to the
top of the regeneration furnace.
Sludge drops through gate-lock feeders
onto the center of the top (#1) hearth of
the multiple-hearth regeneration fur-
nace. There are five hearths in the
furnace. Hot gases from an external fuel
oil-fired burner are fed into the bottom
(#4 and #5) hearths. The furnace has a
rotating centershaft with two rabble
arms on each hearth. Solids are raked to
drop holes at the outer edges of hearths
#1 and #3. On #2and#4theyare moved
to the center of the hearth, where there
is an annular hole around the center
shaft through which the solids drop. The
furnace is designed to reduce upward
gas velocity and thereby minimize en-
trainment of carbon particles.
The principle of operation is that
water is evaporated from the sludge on
hearths #1 and #2 at gas temperatures
from 480° to 700°C. On hearth #3 the
biological solids and adsorbed organics
are burned off by 750° to 870°Cgas. On
the bottom two hearths, at gas tempera-
tures from 870° to 1,020°C, the pow-
dered carbon is reactivated in the
presence of water vapor.
Incandescent powdered carbon is
rabbled through a grate at the outer
edge of the #5 hearth into a quench
tank. Sufficient water is fed to the tank
to give a 5-percent slurry concentration.
The slurry is continuously withdrawn
and fed to a continuous stirred HCI acid
wash tank. Acidified slurry overflows
into a water dilution jet. Diluted slurry at
about 1-percent solids concentration
feeds a carbon thickener tank. A 10 to
12 wt.% slurry discharges from the
bottom of this tank to one of three
carbon slurry storage tanks. The storage
tanks hold slurries of regenerated and
make-up virgin carbon for feed to the
aerators. Virgin carbon usage is
4,500,000 to 2,700,000 kg/year (10
MM to 6 MM Ibs/year). The plant tank
farm also includes tankage for H2P04
nutrient, fuel oil, and HCI storage.
Plant Start-Up
The start-up procedure called for an
initial charge of 60,000 kg (4,000
mg/liter) of powdered activated carbon
to one of the aerators. The aerator
would then be seeded with 2,000 mg/
liter of biological solids. Concurrently,
primary effluent would be fed through
the aerator and one clarifier at one-half
the normal operating rate. Wastewater
flow would then be gradually increased
to full rate. During this time, a dose of 80
mg/liter of virgin activated carbon
would be continuously added to the
feed. When the first aerator was at full
rate and the biological system fully
established, one-half the aerator MLSS
was to be transferred to the second
aerator. Flow to both aerators would be
set at one-half normal rates and then
gradually increased to full rate. The
same procedure would then be used to
start the third aerator.
During the seeding period (November
19,1976, to December 15,1976), a total
of 42,000 kg dry weight of biological
solids were trucked to Chambers Works
and pumped into the aerators. Biological
solids were obtained from two New
Jersey waste treatment plants. Most
came from the Gloucester Regional
Treatment Plant; the remainder came
from American Cyanamid's Bound Brook
plant. Aerator temperature during this
time varied from 10° to 21 °C. Foaming
in the aerators, which had been a
problem during hydraulic testing,
stopped as soon as the powdered carbon
was added to the aerators.
The liquid train was capable of full-
flow operation in February 1977; how-
ever, there were delays in getting the
solids train up to normal production
rate, so the plant was limited in the
amount of solids which could be wasted.
To minimize the buildup of solids in the
liquid train, flow to the PACT process
was restricted until June 1977. Full
flow had to be accepted at that time to
meet the NPDES permit deadline for full
wastewater treatment. Start-up of the
solids train commenced in Januray
1977. It took over 13 months, until
February 1978, before all portions were
running satisfactorily.
Full-Scale Operation -
Liquid Train
In the first 30 months of full-scale
operation, the PACT process has ex-
ceeded expectations in terms of BOD
and color removal. Effluent DOC con-
centrations have been lower than fore-
cast, and the percent removal has
almost equalled forecast. Average per-
formance for 24 months (Jan. 1978 -
Dec. 1979) is summarized in Table 3.
There has been a narrow distribution
of effluent concentration of critical
waste parameters vs. wide variations in
feed concentrations. This performance
has been achieved even though the
wastewater is more highly colored and
contains a higher ratio of nonbiodegrad-
able organics than expected. Normally
nonbiodegradable or very slow degrad-
able materials are being removed in the
presence of the activated carbon. Even
more important, in terms of meeting
NPDES permit limits at the ultimate
Chambers Works river outfall, perform-
ance has been very consistent.
The PACT system has been very effec-
tive in removing EPA-designated organic
priority pollutants (Table 4). Base-neutral
compounds are not as well removed as
volatile organics and acid extractable
compounds. Some removal of metals
occurs across PACT, but the process is
not designed for metals treatment.
A major reason for selection of the
PACT process was the protection that
powdered carbon afforded the biosys-
tem from toxic upsets caused by shock
loads of organic compounds. Because of
the variety and large number of batch
production operations at Chambers
Works, it is not uncommon for the
WWTP to receive 10,000-kg slugs of
various organics over a 2- to 4-hour
period. The biosystem has handled, with
no ill effects, known spills of 23,000 kg
of an aromatic diamine and 14,000 kg of
a toluidine.
During a single day the inlet DOC
concentration will normally vary by a
factor of two, although on rare occa-
sions, it has varied by as much as a
factor of five. The daily average DOC
load is usually within 30 percent of the
load of the previous day. In 30 months of
full-scale operation there have been
only three instances when an upset has
seriously diminished PACT system per-
formance. In these cases, DOC removal
dropped to the 50-60 percent range for
up to 3 days and was below normal for
up to 20 days. BOD and color removals
dropped correspondingly. However, in
no case has there been anything close
to a complete kill of the bacteria.
The first upset occurred in July 1977,
and was partially attributed to a weak-
ening of the biosystem at high tempera-
ture. Evasive measures which are now
taken routinely at the first sign of upset
had not yet been implemented. The
second upset occurred in October 1978.
Identified abnormal conditions were:
high carryover of solids from primary
treatment; lower quality activated carbon
in the aeration tanks; and rapid, sea-
sonal temperature change in the aera-
-------
Table 3. PACT System Performance, 1978 & 1979
24-month avg.
Flow
Inlet Soluble BODs, mg/l
Effluent Soluble BODs, mg/l
% Removal
Inlet DOC, mg/l
Effluent DOC, mg/l
% Removal
Inlet Color, APHA
Effluent Color, APHA
% Removal
94.2 m3/min
(24,900 gpm)
171
(51 ,000 Ib/day)
6.7
(2.000 Ib/day)
96
170
(50,800 Ib/day)
31.2
82
1530
483
66
Design
100 m3/min
(26,400 gpm)
280
(88,700 Ib/day)
21
(6,700 Ib/day)
93
205
(6,500 Ib/day)
35
83
1000
500
50
tion tanks. The third upset, in January
1979, coincided with a high level of
cyanide in the aerators. Although the
exact cyanide level in the MLSS could
not be determined, subsequent labora-
tory studies have shown that cyanide
ion at concentrations above 0.1 percent
in the MLSS can cause severe inhibition
of even the PACT system.
Since start-up, the WWTP has oper-
ated at higher sludge age and lower
carbon dose than that for which it was
designed. The plant was forced into this
operating mode by capacity problems in
the solids train. This operating mode
has limited the wastage rate of MLSS
from the aeration tanks. As a result, the
aerators have operated at monthly
average MLSS levels from 18,000 to
30,000 mg/liter. Average sludge age
has ranged from 20 to 60 days. Solids
not wasted via the solids train must
leave the system as TSS in secondary
clarifier overflow. This has resulted in
an average effluent TSS level of 103
mg/liter vs. a design value of 30 mg/
liter.
Because of the high solids load
throughout the system, the secondary
clarifiers have been badly overloaded
since start-up. Several process and
mechanical modifications have not
solved the basic problem—the solids
have to go somewhere. However, the
expected solids settling rate (with PACT
MLSS) has been seen in the clarifiers.
Effluent from the WWTP is mixed with
uncontaminated cooling water in a
settling basin before being discharged
to the river. Overflow PACT solids settle
here. They are ultimately dredged,
returned into wastewater which is fed
to primary treatment, and removed with
the primary solids. This results in
higher-than-forecast carbon losses, as
this carbon is unavailable for regenera-
tion.
Because of much better aeration tank
performance at high sludge age, the
plant has been able to operate at an
average carbon dose of 120 mg/liter
and still meet all required effluent
quality standards. As the design carbon
dose was 180 mg/liter, net carbon costs
have been no higher than design projec-
tion. Two types of virgin carbon are
used, Hydrocarco C (HD-C) and Nuchar
SA (SA). Although other carbons have
been tested, these are the only two of
satisfactory quality which are available
in sufficient quantity for use. In the 24
months of operation activated carbon
feed has been 48% SA, 14% HD-C, 3%
test carbons, and 35% regenerated
carbon. Nuchar SA has been the most
effective in treating Chambers Works
waste; however, it is also the most
expensive.
As expected, biological activity has
been a function of aerator temperature.
The carbon dose is increased in the
winter to compensate for lower biological
activity at low temperature. Despite the
July 1977 upset at 38.4°C, the plant
operated well at daily average aerator
temperatures of up to 38.8°C in sub-
sequent summers. Up to these tempera-
tures, biosystem performance appears
to steadily improve. Since there are
other factors affecting WWTP operation,
only a general trend in the relationship
among temperature, activity, and carbon
dose has been determined.
The most important repetitive decision
in PACT operation is setting carbon
dose. This dose must be sufficient to
raise the level of biological treatment to
meet a specified effluent quality. Oper-
ating limits on effluent quality are set to
assure that, after WWTP effluent is
mixed with 1.5 times its volume of non-
contact cooling water in the settling
basin, the combined effluent meets
NPDES permit limits. In practice, this
means that the carbon dose is adjusted
based on the level of DOC and color in
the plant effluent. When limits for these
more refractory parameters are met, the
BOD concentration is always well below
the allowed limit. The budgeted carbon
dose for 1980 is 114 mg/liter, of which
50 percent is to be regenerated carbon.
In theory, carbon dose should be ad-
justed proportionally to the inlet con-
centrations of color and/or DOC.
However, as the composition of Chambers
Works waste varies from day to day, so
does the ease or difficulty of treatment.
At the same inlet concentrations there
can be drastic differences in the percent
removals achieved at a given carbon
dose. Since the plant has operated at
sludge ages of over 20 days, it is
impossible to drastically change the
activated carbon content of the MLSS in
a few days. Thus, despite a short-term
drop in inlet DOC or color, carbon dose
still must be kept high enough to main-
tain an adequate concentration in the
MLSS. The WWTP has generally been
operated to maintain the several-day
average of DOC removal above 80
percent.
There are two other constraints on
carbon dose that are controlling at
times. The higher the product of carbon
dose and quality in the MLSS, the more
resistant a PACT plant is to upsets. In
effect, a high activated carbon level in
the aeration tanks is insurance against
an upset of unpredictable magnitude
which might occur at any time. Another
constraint on carbon dose has been
PACT sludge filterability. Carbon in the
sludge drastically increases filterability.
Advantage was taken of this in the
design of the WWTP. Filtration capacity
installed was less than would have been
required for a similar capacity activated
sludge plant; however, carbon dosages
assumed during design were in the
range of 130 to 180 mg/liter. By taking
-------
Table 4. Priority Pollutant Removals Across PACT System*
Concentration in tig/liter (ppb)
Compound
Feed
Volatile Organics
Benzene 105
Carbon Tetrachloride 94
Chlorobenzene 172O
Chloroethane 280
Chloroform 201
Ethylbenzene 41
Methyl Chloride 1770
Tetrachloroethylene 24
Toluene 519
1,1,1- Trichloroethane 13
Trichloroethylene 41
Trichlorofluoromethane 155
Base-Neutral Extractables
1,2-Dichlorobenzene 259
2,4-Dinitrotoluene 1900
2,6-Dinitrotoluene 1640
Nitrobenzene 454
1,2,4-Trichlorobenzene 523
Acid Extractables
2-Chlorophenol 11
2,4-Dinitrophenol 161
4-Nitrophenol 1020
Phenol 489
Metals
PACT
Effluent
0.9
1.4
30
12
21
1.7
Nil
1.7
1.7
0.6
1.9
3.0
120
243
575
2
169
1.6
5.0
10
38
Removal (%)
98
95
98
94
81
94
99
93
99
89
94
95
44
65
64
99
66
95
98
97
94
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Zinc
20
19
1.4
5.6
26
72
112
.02
77
29
602
10
21
1.0
5.4
15
75
54
.01
59
23
360
50
Nil
29
4
42
Nil
52
50
23
21
40
*Values are averages of data which show significant day-to-day variations in
removals. Results were obtained by GC/MS analyses of plant samples and are
subject to analytical error.
advantage of operation at long sludge
age, the WWTP has operated success-
fully at carbon doses as low as 50
mg/liter, particularly at warmer temp-
eratures and lower DOC loads. Sus-
tained operation at less than 100 mg/
liter carbon dose lowers the activated
carbon content of the MLSS below 50
percent. At this point, filterability of
PACT sludge drops drastically. In the
summers of 1978 and 1979, carbon
dose often had to be held above that
strictly necessary for DOC or color
removal in order to sustain a minimum
acceptable filtration rate in the PACT
filter press.
The plant has evolved procedures for
evasive actions in case an imminent
upset is suspected. In order of priority,
the procedures are to reduce waste-
water flow, to increase air flow to the
aerators, and to raise carbon dose by
100 mg/liter. Procedures now call for
an automatic decrease in flow if inlet
DOC concentration doubles within one
to two hours; an increase in air flows if a
shock load of inhibitory material is
apparent; and, an automatic 100 ppm
increase in carbon dose if effluent DOC
rises above 50 mg/liter.
Full-Scale Operation - Solids
Train
The most significant solids train
operating result is the demonstration
that powdered activated carbon can be
regenerated from PACT sludge, at good
yield and reasonable quality, in a coun-
tercurrent multiple-hearth regeneration
furnace. With the aid of a post-regener-
ation acid wash, the carbon can be
regenerated several times with no
build-up of ash or other impurities
which could harm performance.
Overall performance of the solids
train has been below design because of
serious capacity limitations in the
sludge-conveying system and the PACT
filter press (the latter not withstanding
the relatively good filtration properties
of PACT sludge vs. normal activated
sludge). Because of these capacity
problems, operation of the regeneration
furnace has so far approximated only a
break-even proposition. However, there
is no assurance that the alternative,
using carbon on a one-through "throw
away" basis, could be cost-effective for
very long, even at low dose operation.
From February 1978 through Decem-
ber 1979, the regeneration furnace
operated at 55 percent intime (intime
being defined as sludge being fed to the
furnace). This period includes major
outages of 5-1/2 and 7 weeks each to
rebuild furnace hearths. Hearth life has
been shortened by inability to maintain
a constant furnace feed. Feed interrup-
tions allow hearth temperatures to rise
to excessive levels and cause rates of
temperature change in excess of the
recommended 40°C/hour maximum.
Exclusive of hearth rebuilding and one
shutdown for a rabble arm repair,
furnace intime has averaged 66 percent.
However, half of the downtime has been
caused by failures of the feed system
rather than the furnace itself. While
operating, the furnace has produced
4,100,000 kg of regenerated carbon
with an estimated quality recovery of 63
percent. The yield based on carbon fed
to the furnace is estimated at 80 percent.
There have been no problems with
the acid-washing system for regenerated
carbon slurry. After acid-washing, the
average ash and volatile levels of the
carbon have been 18 percent and 9
percent, respectively (virgin carbon as
purchased will contain from 4 to 30
percent ash and 10 to 20 percent vola-
-------
tiles). Average feed rate during intime
has been 2,240 kg/hour wet sludge. A
summary of furnace performance is
presented in Table 5. The heat load on
the furnace is significantly less than
design value because of the higher
percent solids achieved in filtration.
This has contributed to temperature
control problems as the burner is over-
sized for these operating conditions.
Performance of the conveyors has
been poor. Operating utility of the con-
veyor system itself has been approxi-
mately 80 percent. The most important
factor affecting utility is mechanical
wear. The carbon coke is erosive and
seriously wears the metal in the drag
flight conveyors. Operating strategy has
been to upgrade the equipment me-
chanically as much as possible. There
are now parallel units at two stages in
the conveyor system. Parallel units at
two more stages were scheduled for
installation early in 1980. With the
conveyors running well, furnace capacity
is limited by plugs of the gatelock
feeders on top of the furnace. Plugging
is a function of filter-cake quality.
In retrospect, itappears that the filters
should have been located above the
regeneration furnace; or the filter and
furnace should have been located hori-
zontally far enough apart that an inclined
belt conveyor could have been used to
transport sludge. Although not devoid of
problems, PACT solids move very well in
slurry form when using proper piping,
valves, and pumps. It would have been
much easier to pump slurry 15 meters to
a filtration unit than to drag filter cake
the same distance.
Intime of the PACT filter press has
averaged about 85 percent since January
1978. The main causes of lost time have
been 16-hour high-pressure water
washings every week to 10 days, num-
erous shutdowns for periodic replace-
ment of press cloths, gaskets, etc., and
outages of high-pressure feed pump(s).
The latter problem has been somewhat
ameliorated by the use of one triplex,
higher capacity piston pump to replace
the four parallel originally installed
pumps.
The most serious problem with the
press has been long cycle times. They
vary from 1.5 to 7 hours, depending on
sludge composition and filter-cloth con-
dition. At carbon dose over 100 mg/l,
cycles average between 2 and 3 hours;
Table 5. Regeneration Furnace Performance
Time Period
Reg. Carbon Produced,
kg/day
Kg Reg. Carbon/
kg dry feed
Carbon Yield, %
Est. Quality Recovery, %
Reg. Carbon, Iodine Number
Reg. Carbon, % Volatiles*
Reg. Carbon, %Ash
Furnace in Time, %
Wet Sludge Feed Rate
While Operating, kg/hr.
Sludge Feed, % Solids
Dry Solids Fed, kg/day
Regenerated Carbon
Produced While
Operating, kg/hr.
Feb. '78 thru
Dec. '79
5,860
.40
80
63
475
8.5
18.4
58
2,240
44
14,900
443
June '79
7,000
.50
Not
Available
7O
537
8.6
16.7
69
1,940
44
13,900
421
March '78
9,600
.46
Not
Available
63
487
7.2
15.4
79
2,440
44
20,800
506
Design
Values
18,200
.53
70
75
56O
8
15
90
4,190
38
34,300
1.770
*Volatiles equal weight loss when carbon is heated to 950°C for 5 minutes at a
controlled rate of temperature rise with minimum exposure to air.
at lower carbon dose they average about
4 hours. Even though laboratory-mea-
sured sludge filtration properties are
close to design values, press cycle time
is still longer than expected. Calcula-
tions of cycle time were simply overly
optimistic. Also, because of the low
dose, long sludge age mode of operation,
carbon content of the PACT sludge has
never been as high as forecast.
Filter-cake quality in terms of percent
solids has exceeded design; however,
there is a surprising variation in cake
quality within a very narrow range of
solids concentration. Depending on
composition of the MLSS, a greater-
than-40-percent-solids cake can either
be an easily handled, friable cake or a
material of the consistency of bubble
gum and very difficult to convey.
The sludge thickener has worked well
and consistently delivers a 7-10 wt.%
slurry for feed to the PACT filter press.
Because the thickener has worked as
well as it has, sludge can be wasted
directly from the aerators instead of
from the more concentrated RAS stream.
Sludge will go "stale" and be hard to
filter if it is held in the thickener during a
filter outage.
Precoat is used before each press
cycle. Studies have demonstrated its
need, but there have been no major
improvements in press performance
from varying precoat types and quantity.
An ash-free type material, like cellulose,
is needed because it is burned in the
furnace. Tests have recently been made
of direct addition of a polymer to the
sludge to improve filterability. No major
improvement has been seen to date;
however, further work is planned. Fre-
quency of filter-press cleaning varies
depending on sludge condition. A
thorough cleaning requires over 16
hours. Approximately once a year it has
been necessary to completely re-dress
the filter press. This costs $13,000 and
takes 2 weeks. Underscreens cost
$56,000 and may need replacement
every 2 years.
Since Januray 1978, operating strat-
egy has been to run the solids train at as
high a rate as possible. Generally this
has not been enough to equal the rate of
solids added to the system, and the
liquid train periodically purges itself via
an overflow of TSS from the secondary
clarifiers. Several steps have been
taken to improve solids train production.
To maintain sludge wastage from the
aerators when the regeneration furnace
is out of service, filtered sludge is
dumped to a payload container and
-------
hauled to a storage area for future
carbon recovery. An inventory of sludge
is vital to maintaining feed to the
furnace during filter-press outages.
Current operating strategy calls for
keeping some stored sludge available at
all times.
Furnace operating options are limited
because of a strong interdependence
among operating conditions. Operating
instructions specify an acceptable
temperature range for each of the five
hearths. Actions which bring the tem-
perature of one hearth into range can
drive another hearth's temperature out
of limits. Therefore, the furnace operator
does not have complete control over the
hearth temperature profile. The profile
is also very dependent on the sludge
feed rate, which is a strong function of
filter-cake quality. Other variables being
constant, laboratory studies have shown
that quality is generally obtained at the
expense of yield and that better quality
is achieved at higher temperature;
therefore, the tendency now is to try to
operate the upper hearths on the high
side of the allowable temperature range.
The furnace must operate under
reducing conditions with water vapor
present to regenerate the carbon. In lieu
of direct addition of steam, a portion of
the 100-percent relative humidity off-
gas from the scrubbing system is re-
cycled back to the combustion chamber.
It is now felt that higher water concen-
trations than are presently obtainable
are needed to improve the quality of
regenerated carbon. Facilities for direct
steam addition have been installed, and
tests are scheduled in 1980. A problem
with this change, and furnace operation
in general, is insuring that explosive CO
and H2 mixtures do not develop in the
furnace from the water gas reaction;
therefore, a continuing program of gas
analyses is maintained. While there is a
reassuring background of years of
industrial operation of multiple-hearth
furnaces to regenerate granular carbon,
the fact that this system is treating
powdered carbon is reason for caution.
Operating Costs
It cost approximately $16,500,000 a
year to operate the Chambers Works
wastewater treatment plant in 1978
and 1979. Of this, approximately
$10,000,000, or 65 percent, is the
estimated cost of PACT secondary/
tertiary treatment. Primary treatment
costs include over $1,500,000 to pur-
chase the lime needed for acid neutrali-
zation. Annual depreciation charges of
$4,000,000 on a book investment of
over $45,000,000 make up 25 percent
of the annual operating costs.
Average annual operating costs for
1978 and 1979 are shown in Table 6. As
this study deals with PACT treatment
only, each individual item of cost has
been proportioned as accurately as
possible between primary and second-
ary/tertiary treatment.
At the average plant flow of 94 m3/
min. (24,900 gpm) costs, including
depreciation, have averaged $.33/m3
($1.27/1,000 gal.) for total treatment,
or $0.22/m3 ($0.82/1,000 gal.) for
PACT treatment only. Because many
costs are independent of load, unit costs
would drop significantly if the plant
were operated closer to capacity. Many
treatment facilities report costs exclusive
of depreciation charges. Ignoring de-
preciation, the costs for PACT treatment
have averaged $0.16/m3 ($.61/1,000
gal.).
In attempting to extrapolate these
costs to other waste treatment opera-
tions, it must be realized that operating
and mechanical personnel involved
with the Chambers Works plant are
relatively well paid. Their average earn-
ings approach $10 an hour plus an
additional 40 percent for fringe benefits.
A rough estimate of the current replace-
ment value for the total treatment plant
is $60,000,000; the PACT process
share of this updated investment is
about 65 percent. Direct maintenance
costs, including all overheads, have run
about 3.7 percent of total updated
investment. Maintenance costs are
forecast to decrease significantly in
1980 as a result of mechanical improve-
ments over the past two years.
The plant is currently staffed with 29
operators, 4 salaried shift supervisors, 2
day supervisors, plus 6 other manage-
ment and technical people. There are
approximately 30 mechanics, electri-
Table 6. Wastewater Treatment Plant Operating Costs
Average Annual Costs* in Thousands
PACT
Secondary/
Primary Tertiary
Cost Items Treatment Treatment Total
Lime
Other Raw Materials-Primary
Activated Carbon
Polymer. PACT
Filter Aid. PACT
Acid for Carbon Washing-PACT
Phosphoric Acid-PACT
Total Raw Materials
Primary Treatment-Utilities
PACT -Steam
PACT-FuelOil
PACT -Electricity
PACT-Other Utilities
Total Utilities
Operating Labor
Mechanical Labor, Materials
Salary Personnel
Depreciation
Overheads
Computer
Miscellaneous Supplies
Laboratory Support
Other Services
Other Primary Treatment Expenses **
Total Other Costs
Total Plant Costs
$ 1.540
150
—
—
—
—
—
$ 1.690
460
—
—
—
—
460
420
560
230
/,430
340
50
—
—
—
670
3,700
$ 5,850
$ —
—
3,0/0
190
20
20
30
$ 3,270
—
70
170
610
20
870
420
1,670
380
2,650
770
'60
270
160
150
—
6,530
$10,670
$ 1,540
150
3,010
190
20
20
30
$ 4,960
460
70
170
610
2O
1,330
840
2.230
610
4,080
1.110
110
270
160
150
670
10,230
$16,520
* Aver age for Years 1978 and 1979.
**Including Landfill Charges.
-------
cians, and instrument technicians cur-
rently assigned to the plant. Both
operating and mechanical staffs are
forecast to drop by 15 to 20 percent in
the next year.
Harry W. Heath, Jr., is with E. I. du Pont de Nemours & Company. Inc., Chambers
Works, Deepwater, NJ 08023.
John E. Matthews is the EPA Project Officer (see below).
The complete report, entitled "Full-Scale Demonstration of Industrial Waste-
water Treatment Utilizing Du Pont's PACT® Process,'' (Order No. PB 81 -248122;
Cost: $14.00, subject to change} will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
P.O. Box 1198
Ada, OK 74820
. S. GOVERNMENT PRINTING OFFICE: I982/559-092/3420
-------
o
m
G
10
c0
o
co
en 220
oo
ru
STO3J3D
OXJ33O
XTO
rn
m
2»
O
i
o
s 2
a a *
3 0) J.
S -• 3"
2. O °
O 3 m
2 3
CO 0)
cn
(D
O)
------- |