United States Municipal Environmental Research
Environmental Protection Laboratory
Agency Cincinnati OH 45266
Research and Development EPA-600/S2-81-1 77 Oct. 1981
Project Summary
Granular Activated
Carbon Installations
Russell L. Gulp and Robert M, Clark
Granular activated carbon (GAC)
treatment design criteria, performance,
and cost data from 22 operating
municipal and industrial GAC installa-
tion that treat water and wastewater
and that process food and beverage
products are compiled and summa-
rized. Guidance and an example of
how this information can be used to
estimate costs for GAC treatment of
water supplies is provided. The report
should be used in conjunction with a
previous series of reports on "Estima-
ting Water Treatment Costs" to obtain
project-specific cost estimates. It is
not a design manual and does not
provide design criteria such as required
contact time, probable regeneration
frequency, activated carbon reactiva-
tion system criteria, or activated
carbon transfer guidelines. Rather,
the approach to determining such
design data for water systems is
presented.
This Project Summary was developed
by EPA's Municipal Environmental
Research Laboratory, Cincinnati, OH,
to announce key findings of the
research protect that is fully docu~
merited in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
On January 9. 1978, the U.S. Envi-
ronmental Protection Agency (EPA)
proposed the use of GAC as a means of
treating drinking water. Since that time,
munh has been written both for and
against using GAC in this manner.
Serious challenges and many questions
haw been raised regarding EPA's cost
estimates for GAC use. To respond to
some of these questions, EPA's Drinking
Water Research Division initiated a
carefully designed study to establish
water supply unit process cost curves
on a consistent and understandable
basis.
In an earlier study ("Estimating Water
Treatment Costs," EPA-600/2-79-162
a, b, c, and d, performed by Culp/Wes-
ner/Culp under EPA Contract No. 68-
03-2516), construction and operation
and maintenance cost curves were
developed for processes capable of
removing those contaminants included
in the National Interim Primary Drinking
Water Regulations. The final report
contains cost curves for 99 different
unit processes. These cost curves were
divided into two categories: large water
treatment systems applicable to flows
between 1 and 200 mgd and small
water treatment systems applicable to
flows between 2,500 gpd and 1 mgd. A
computer program for retrieving, up-
dating. and combining the cost data was
also developed. During the course of
this work, the costs of GAC adsorption
and reactivation in municipal water
treatment as they relate to the removal
of organics from drinking water became
a subject of great national interest.
Because of this, in 1978, the original
project was expanded to include a
special study of the unit process costs of
GAC adsorption and reactivation in
potable water treatment. The special
study was directed at visiting as many
existing GAC installations as possible to
gather and publish data on actual
operating experiences, particularly on
the costs of building, operating, and
maintaining GAC plants.
-------
The report summarized here presents
the findings of this special study of GAC
installations and the compilation of the
information available on the use of GAC
in water treatment.
Use of Activated Carbon in
Drinking Water
Powdered activated carbon (PAC) has
been used, without harmful effects, for
mora than 50 years to remove taste and
odor from public water supplies, but the
use of GAC in treating municipal water
in the United States is limited to a few
facilities. In most cases, GAC is used to
remove taste and odor from drinking
water. Its use may become more
common in light of new information on
the occurrence of trace organics in
water, the recent regulations limiting
the concentration of trihaiomethanes in
public water supplies, and the possibility
of requirements for GAC treatment,
European water works have had consid-
erable experience over a long period of
time with GAC installations.
Although the use of GAC in municipal
water treatment has been limited, GAC
has been used in industrial and munici-
pal wastewater treatment and in various
industrial process applications. The
specific uses of GAC are somewhat
different with these applications than
with water treatment, but much of the
information on design and operations
will prove useful to water purveyors. In
general, the application of GAC adsorp-
tion to drinking water is simpler than to
wastewater.
Eighteen of the twenty-two GAC
installations visited were industrial
process or municipal wastewater facili-
ties, whereas four were municipal
water treatment plants. Case histories
presenting design, operating, perfor-
mance, and cost information are pre-
sented in the report; the pertinent
information has been summarized
(Table 1). Single page fact sheets for
each of these case histories are given in
the report.
A principal function of the site visits
was to collect construction and operation
and maintenance (O&M) cost informa-
tion on GAC installations. Plant records
were used to obtain available construc-
tion costs, the dates for these costs, and
the most recent O&M costs. The basic
data are presented in the individual case
study reports, but no attempt was made
to update construction or O&M costs to
present day prices or to extrapolate the
costs to water treatment plants. This
report does, however, re'.er to proce-
dures whereby data from existing GAC
projects can be adjusted or modified
(based on the results of pilot-plant test
results of the water to be treated) so as
to be useful to experienced professionals
making preliminary estimates of costs
for future potable water projects involv-
ing GAC adsorption and reactivation or
replacement.
Extrapolating Municipal
Wastewater Experience to
Water Treatment
The first plant-scale use nf GAC in a
municipal wastewater treatment plant
was at South Lake Tahoe, California, in
1965. This plant has operated continu-
ously since that time and now has 15
years of operating experience with
GAC; the GAC system has processed
more than 12 billion gallons of pretreated
municipal wastewater. The reclaimed
water COD ranges from 10 to 30 mg/L.
The South Tahoe installation was an
EPA Demonstration Plant. For 3 years,
EPA funded the collection of very
detailed and complete plant operating
data and cost information.
Other Wastewater Installations
Water reclamation plants constructed
at the Orange County (CA) Water
District (Water Factory 21), the Upper
Occoquan (VA) Sewer Authority, and
the Tahoe-Truckee (CA) Sanitation
Agency (Nos. 6, 3, and 2, respectively)
have the same configuration as the
South Tahoe plant both in respect to the
type of GAC facilities provided and the
high degree of pretreatment afforded.
All of these plants operate successfully
with few GAC system problems. As
might be expected with second and third
generation designs, these later plants
embody some improvements over the
original South Tahoe installation,
although no major changes or deviations
were initiated.
Although many other successful
applications of GAC in advanced waste
treatment (AWT) plants exist, the GAC
experience in some AWT plants has,
unfortunately, been poor. These failures
in AWT applications have not stemmed
from deficiencies in the basic GAC
processes or in organics adsorption and
thermal reactivation, but, rather, from
mechanical problems.
Operational Problems
In discussing the operational problems
encountered with GAC systems, those
problems associated specifically with
sewage must be distinguished from
general problems that might be encoun-
tered with any type of GAC system.
Many problems with GAC in wastewater
treatment will not occur in water
purification. For example, in water
treatment, few or no problems could be
expected with excessive slime growths,
hydrogen sulfide gas production, or
corrosion from adsorbed organics
released during carbon reactivation.
Some of the types of problems
encountered with GAC systems in
wastewater treatment include:
inadequate GAC transfer and feed
equipment,
undersized slurry and transfer
lines,
failure to provide for venting air
from backwash lines with destruc-
tion of filter bottoms and disruption
of GAC,
failure to house or otherwise
protect automatic control systems
from the weather,
inadequate means for continuous,
uniform feed to furnace, this
results in temperature fluctuations,
inconsistent reactivation efficiency,
and wasted energy,
location of furnace and auxiliary
drive motors in areas of very high
ambient temperature (e.g., above
top of furnace), and
the use of nozzles in filter and
carbon contactor bottoms; this
produces major failures in carbon
systems just as they have for many
years in water titration plants.
Their use is risky.
The common problems related to
wastewater treatment (biological orga-
nisms in the activated carbon contactors
and development of anaerobic conditions
with the production of corrosive hydro-
gen sulfide) have been successfully
circumvented by providing adequate
flow through the columns or frequent
backwashing. Failure to provide ade-
quate pretreatment has caused column
clogging and mud balls with the need for
more frequent backwashing.
Corrosion has been a problem with
some of the GAC systems. The furnace
system, transfer piping, and storage
tanks are susceptible components.
Many operations require frequent
replacement of the rabble arms and
teeth and replacement of the hearths
every few years. At one installation,
titanium or ceramic coated rabble teeth
were no more resistant to corrosion
2
-------
than were stainless steel teeth. In one
case, the corrosion problem in the
furnace was solved by eliminating the
use J auxiliary steam during reactiva-
tion. in another, corrosion was linked to
fluctuating temperatures in the hearths
caused by irregular feed to the furnace
and frequent startup and shutdown.
These problems can be partially reme-
died by better operation and avoided by
better engineering design.
In several industrial applications, the
wastewater itself has been highly
corrosive. In these cases, the contactors
have been subject to corrosion. At
Spreckles Sugar, the epoxy linings in
the columns must be replaced every 3
years; Republic Steel also replaces its
column linings on a regular basis. Public
water supply sources would not be
expected to consist of corrosive water.
By properly applying the best current
engineering design knowledge and
practices for G AC systems, these rather
serious problems might be avoided.
When water works engineers apply
GAC to produce high quality drinking
water, they should make the most of the
experiences of the consultants for
industry and wastewater agencies.
Extrapolating Industrial and
Wastewater Data to
Water Supply
Caution must be observed in extrapo-
lating GAC cost data from operating
industrial installations and municipal
wastewater treatment plants to the
design of water works. The purpose for
using GAC in each of these types of
applications is generally the same to
remove organics. Important differences
do exist, however. In industry, the GAC
serves to remove a rather narrow band
of organics color molecules from a
viscous liquid. In wastewater treatment,
the GAC removes (with or wihout
biological activity) a broad spectrum of
organic substances from water as
measured by BOD, COD, and TOC.'ln
water treatment, the objectives of GAC
treatment are not completely defined at
this time. For raw waters with color or
taste and odor problems, using GAC
unquestionably improves drinking water
from an aesthetic standpoint. In many
cases, the cost of GAC may be warranted
for either of these purposes alone. For
the great number of water systems
without color or taste arid odor problems,
the only concern with respect to
organics Is the possible health effects
over long periods of time from ingesting
trace quantities of organics that may
cause cancer.
Public health officials and water
works managers still disagree as to
whether the health risks that may be
involved in the presence of minute
traces of organics in drinking water are
sufficient to warrant the cost of GAC
treatment. A major problem is that the
potentially harmful organics in drinking
water have not all been identified at this
time, and many of those that have been
tagged as suspect have widely different
adsorptlve characteristics. Some adsorb
readily on GAC; others do not.
GAC loading rates at exhaustion of
adsorptive capacity vary widely among
the different potentially hazardous
organics. This affects the length of
service life of GAC before reactivation or
replacement is necessary a deter-
mining factor in GAC treatment costs.
Similarly, the reactivation times and
temperatures for thermal reactivation of
GAC saturated with different organics
also differ, and all are not known at this
time. Again, this has an important
bearing on GAC treatment costs.
Because of the widely varying adsorptive
and reactivation characteristics of trace
organics on GAC in water supplies, pilot
plant tests of both adsorption and
reactivation are mandatory preludes to
treatment system design at this time.
Over a period of years, general,
average design parameters may emerge
from the results of pilot plant studies
and demonstration projects, but this
time is not yet at hand. Once the GAC
design parameters for water treatment
have been established from pilot tests
for a particular water source, then the
knowledge and experience from other
GAC installations in industry and
wastewater plants can be put to good
use. GAC dosages, contact times, and
spent carbon reactivation times and
temperatures can be determined. Con-
tactor sizes can be calculated and
furnace sizes arid fuel requirements can
be determined. Transport facilities for
GAC in water treatment can be the
same as for other types of GAC installa-
tions provided differences are taken into
account differences in quantities and
possible differences in the viscosities of
activated carbon slurries beca use of any
slime growths. Also, with GAC design
parameters pinpointed as ¦ result of
pilot plant studies, construction costs
can be accurately estimated based on
costs of existing installations in AWT
and industry. The estimates cannot,
however, be based on a million-flallon-
per-day capacity basis; rather, they
must be based on adsorption and
reactivation data applicable to each
specific installation.
Selecting the most economical num-
bers of contactors for a water system of
a certain size involves the same princi-
ples that are used for other systems.
Because of shipping regulations, factory-
fabricated contactor vessels are gener-
ally limited to about 12-ft maximum
diameter. For large capacity installations,
a smaller number of field-erected steel
vessels or poured-in-place concrete
vessles may be less costly.
Because upflow contactors provide all
of the advantages of countercurrent
operation with respect to GAC savings,
they are favored for most types of
service. The exception is water treat-
ment, In this case, downflow is used
beca use of the discharge of carbon fines
in the effluent (a characteristic of
upflow columns) is avoided.
Cost estimates must be evaluated on
the same basis as all other estimates of
construction cost; there is no reason
that they should be more or less
accurate than estimates made for the
rest of the treatment plant. Fifteen
percent is generally accepted as being
an allowable difference between costs
estimated from construction plans ana
the best bid received from contactors.
With good pilot plant data and with
proper application of cost data from
existing GAC installations, preliminary
cost estimates for GAC treatment of
public water supplies should be accurate
enough for planning purposes.
The extrapolation of wastewater
treatment experience with GAC to the
design of water treatment systems is a
task for trained, experienced, engineer-
ing professionals. Even then, the
following discussions are intended to be
no more than an introduction to the
subject.
Designing GAC Systems far
Water Treatment
The following discussion is devoted to
some of the procedures and details for
developing the design basis and costs
for GAC water treatment systems from
pilot plant test results and information
from full-scale applications.
GAC System Components
Systems utilizing GAC are rather
simple In general, they provide lot
contact between the GAC and water to
3
-------
be treated for the length of time required
to obtain the necessary removal of
organics; reactivation or replacement of
spent carbon; and transport of makeup
or reactivated carbon into the contactors
and transport of spent carbon from the
contactors to reactivation or hauling
facilities.
Pilot Plant Tests
Despite the simplicity of GAC systems,
laboratory and pilot plant tests are
needed to select the carbon and the
most economical plant design for both
water and wastewater treatment pro-
jects, Pilot column tests make it passible
to determine treatability; select the best
carbon for the specific purpose based on
performance; determine the required
empty bed contact time; establish the
required carbon dosage that, together
with laboratory tests of reactivation, will
determine the capacity of the reactiva-
tion furnace; determine the necessary
activated carbon replacement costs;
and determine the effects of influent
water quality variations on plant opera-
tion. During pilot plant testing, the
influence of longer carbon contact time
on reactivation frequency can be
measured; these measurements allow
costs to be minimized through a proper
balance of these two design factors.
Design of Pilot GAC Columns
Detailed information, including a list
of materials, on the design and con-
struction of pilot GAC columns is
presented in Appendix C of EPA's
"Interim Treatment Guide For Control-
ling Organic Contaminants in Drinking
Water Using GAC" (out of print).
Appendix B of the "Interim Treatment
Guide" describes the analytic method-
ology for monitoring pilot column tests.
Also included are data on the adsorba-
bility of various organic compounds; the
performance of GAC in their removal;
information on the use of multiple-
hearth, infrared, fluidized bed, and
rotary kiln furnaces for reactivating
spent GAC; and example calculations
for balancing added costs of increased
contact time versus savings (if any) from
lass frequent reactivation.
Use of GAC in Water and
Wastewater Treatment
Frequency of Reactivation
One of the principle differences in
costs between water and wastewater
GAC treatment is the more frequent
reactivation required in water purifica-
tion caused by earlier breakthrough of
the organics of concern. In wastewater
treatment, GAC may be expected to
adsorb 0.30 to 0.55 lb COD/lb activated
carbon before the GAC is exhausted.
From the limited amount of data
available from research studies and
pilot plant tests (most of it unpublished),
some organics of concern in water
treatment may breakthrough at carbon
loadings as low as 0.05 to 0.25 lb
organic/lb carbon. The actual allowable
carbon loading or carbon dosage for a
given case must be determined from
pilot plant tests. Costs taken from
wastewater cost curves, which are plots
of flow in million gallons per day versus
cost (capital or 0&M costs), cannot be
applied directly to water treatment.
Allowance must be made in the capital
costs for the different reactivation
capacity needed and in the O&M costs
for the actual amount of carbon to be
reactivated or replaced.
Because the organics adsorbed from
water are generally more volatile than
those adsorbed from wastewater, the
increased reactivation frequency resu It -
ing from lighter carbon loading may be
partially offset, or more than offset, by
the reduced reactivation requirements
of the more volatile organics. The times
and temperatures required for reactiva-
tion may be reduced because of both the
greater volatility and the lighter loading
of organics on the carbon.
From the experimental reactivation to
date, reactivation temperatures may be
less than the t .650° to 1,750°F required
for wastewater carbons. The shorter
reactivation times required for water
purification carbons may allow the
number of hearths in a multiple hearth
reactivation furnace to be reduced.
Also, less fuel may be required for
reactivation. These factors must be
determined on a case-by-case basis.
GAC Contactors
Selection of the general type of
contactor to be used for a particular
water treatment plant application may
be based on several considerations
including economics and the judgment
and experience of the engineering
designer. The choice generally would ba
made from three types of downflow
vessels:
t. Deep-bed, factory-fabricated, steel
pressure vessels of 12-ft maximum
diameter. The size of these vessels
might vary from 2,000 to 50,000
ft3.
2. Shallow-bed, reinforced-concrete,
gravity-filter-iypa boxes may be
used for carbon volumes ranging
from 1,000 to 200,000 ft3. Shallow
beds probably will be used only
when short contact times are
sufficient or when long service
cycles between reactivations can
be expected from pilot plant test
results.
3. Deep-bed, site-fabricated, large
(20- to 30-ft) diameter, open
concrete or steel, gravity tanks
may be used for GAC volumes
ranging from 6,000 to 200,000 ft3,
or larger.
These ranges overlap, and the de-
signer may very well make the final
selection based on local factors, other
than total capacity, that affect efficiency
and cost.
The AWT experience with GAC
contactors may be applied to water
purification if some differences in
requirements are taken into account.
The required contact time must be
determined from pilot plant test results.
Although contactors may be designed
for a downflow or upflow mode of
operation and upflow packed beds or
expanded beds provide maximum carbon
efficiency through the use of countercur-
rent flow pri nciples, the leakage of some
(1 to 5 mg/L) carbon fines in upflow
column effluent make downflow beds
the preferred choice in most municipal
water treatment applications. At the
Orange County Water Factory 21,
upflow beds were converted to downflow
bkls to successfully correct a problem
with escaping carbon fines. This full-
scale plant operating experience indi-
cates that leakage of carbon fines is not
a problem in properly operated downflow
GAC contactors.
Single beds or two beds in series may
be used. Open gravity beds or closed
pressure vessels may be used. Struc-
tures may be properly protected steel or
reinforced concrete. In general, small
plants will use steel, and large plants
may use steel or reinforced concrete.
Sand in rapid filters has, in some
instances, been replaced with GAC. In
situations where contact times are
short and GAC reactivation or replace-
ment cycles are exceptionally long
(several months or years, as may be the
case in taste and odor removal), this
may be a solution. With the short cycles
anticipated for most organics, however,
conventional concrete-box-style filter
bads may not be well suited to GAC
contact. Deeper beds may be more
4
-------
economical in first cost and provide
more efficient use of GAC, In converted
filter boxes, possible corrosion effects of
GAC on existing metals, such as surface
wash equipment and metal nozzles in
filter bottoms, must be taken into
account. Beds deeper than conventional
filter boxes, or contactors with greater
aspect ratios of depth to area, provide
much greater economy in capital costs.
The contactor cost for the needed
volume of carbon is much less. In a
water slurry, carbon can be moved
easily ami quickly and with virtually no
labor from contactors with conical
bottoms. Rat-bottomed filters of a type
that require labor to move the carbon
unnecessarily add to carbon transport
costs. The labor required to remove
carbon from flat-bottomed beds varies
considerably in existing installations
from a little labor to a great deal,
depending upon the design of the
evacuation equipment.
For many GAC installations intended
for precursor organic removal or syn-
thetic organic removal, specially de-
signed GAC contactors should be
installed. Contactors should be equipped
with flow measuring devices. Separate
GAC contactors are especially advan-
tageous where GAC treatment is
required only part of the time during
certain seasons because they then can
be bypassed when not needed, possibly
saving unnecesary exhaustion and
reactivation of GAC.
Tremendous cost savings can be
realized in GAC treatment of water
through proper selection and design of
the contactors. The design of contactor
underdrains requires experienced expert
attention.
GAC Contactor Underdrains
Although good proven underdrain
systems are available, often they have
not been used, and there have been
numerous underdrain failures due to
poor design. Some designs used in the
past for conventional filter service have
failed in many installations, yet they
continue to be misapplied to GAC
contactors as well as filters.
GAC Reactivation or
Replacement
Spent carbon may be removed from
contactors and replaced with virgin
carton, or it may be reactivated either
on-site of off-site. The most economical
procedures depend on the quantities of
GAC involved. For larger volumes,
onsite reactivation is the answer. For
small quanitites, replacement or off-site
reactivation will probably be most
economical.
GAC may be thermally reactivated to
very near virgin activity. Burning losses
may, however, be excessive under
these conditions. Experience in indus-
trial and wastewater treatment indicates
that carbon losses can be minimized
(held to 8 to 10 percent per cycle). To
remove certain organics, no decrease in
actual organics removal may occur
despite a 10 percent drop in iodine
number.
Thermal Reactivation
Equipment
GAC may be reactivated in a multiple-
hearth furnace, a fluidized bed furnace,
a rotary kiln, or an electric infrared
furnace. Spent GAC is drained dry in a
screen-equipped tank (40 percent
moisture content) or in a dewatering
screw (40 to 50 percent moisture)
before being introduced to the reactiva-
tion furnace. Dewatered carbon is
usually transported by a screw conveyor.
Following thermal reactivation, the
GAC is cooled in a quench tank. The
water-carbon slurry may then be
transported by means of diaphragm
slurry pumps, eductors, or a blow-tank.
The reactivated carbon may contain
fines produced during conveyance;
these fines should be removed In a
wash tank or in the contactor. Maximum
furnace temperatures and retention
time in the furnace are determined by
the amount (lb organics/lb carbon) and
nature (molecular weight or volatility) of
the organics adsorbed.
Off-gases from reactivation present
no air pollution problems provided they
are properly scrubbed. In soma cases,
an afterburn may also be required for
odor control.
Despite recent advances in the design
area of infrared and fluidized bed
reactivation furnaces, the multiple
hearth furnace is still the simplest, most
reliable, and easiest to operate for GAC
reactivation. The infrared and fluid bed
units still have problems to be worked
out; experience with the multiple hearth
equipment has already solved these
problems. Still, it is necessary with all
four types of furnaces to specify top
quality matsriats to suit the conditions
of service and to see that these
materials are properly installed. Corro-
sion resistance is important in the
furnace itself and especially in all
auxiliaiies to the furnace.
Required Furnace Capacity
The principal cost differences between
GAC treatment of water and wastewater
may lie in the capital cost of the furnace
and in the O&M cost for carbon
reactivation. As already explained, the
two principal differences between
carbon exhausted in wastewater treat-
ment and carbon exhausted in water
purification are that water purification
carbons are likely to be easier to
reactivate (less time in furnace and
lower furnace temperatures) and more
lightly loaded (graater volume of carbon
to be reactivated per pound or organics
removed). Accurate estimates of GAC
costs require knowledge and considera-
tion of these two fectors. To repeat, it is
not possible to use AWT cost curves
based on million gallons per day
throughput or plant capacity to obtain
costs for water treatment. Differences
in reactivation requirements must be
taken into account.
GAC Transport and GAC
Process Auxiliaries
The large differences in O&M costs
for GAC systems depend on the method
seleccad for carbon transport. Hydraulic
transport of GAC in water slurry by
gravity or water pressure uses very little
labor and is simple, easy, and rapid.
Moving dry or dewatered activated
carbon manually or with mechanical
means involving labor can be very
difficult, time consuming, and costly.
Cost of GAC in Water
Treatment
Developing Cost Curves
Little information is available con-
cerning the cost and performance of
GAC for drinking water treatment. As
discussed in the previous sections, most
of the data available on GAC perfor-
mance have been acquired from waste-
water and industrial applications. An
attempt has been made, however, to
extrapolate from these existing systems
and to develop standardized and flexible
cost data that can be used to prepare
cost estimates for GAC systems that
treat drinking water.
Design Cost Information
Much of the analysis and cost
information contained in this section of
the full report are based on the four-
s
-------
Table 1. Summary of GAC System Characteristics
Carbon Contactors
Case
No.
Owner
Type of
Facility
Flow, mgd
Pretreatment
Contact
Time,
min
Hydraulic
Loading,
gpm/ft
Rated
Capacity,
lb/carbon/cfa\
1
South Tahoe
Municipal
wastewater
7.6 (max)
Extensive
17
6.5
6.000
2
Tahow-Truckee
Municipal
wastewater
4.83 (max)
Extensive
20
3.840
3
Upper Occoquan
Municipal
wastewater
15.0 (avg)
Extensive
22
8.4
12.000
4
American
Cyanamid
Chemprocess
20.0 (avg)
Extensive
30
8.0
122.0)0
5
Vallejo
Municipal
wastewater
13.0 (avg)
Moderate
25
6.0
29,000
6
Orange County
Secondary
effluent
15.0 (max)
Extensive
34
5.8
12,000
7
Niagara Falls
Municipal w/
significant
industrial
48.0 (avg)
Moderate
40
1.67
r
8
Fitchburg"
Municipal w/
significant
industrial
15.0 (avg)
Moderate
15
8.00
9
Arco Petroleum
Process waters
w/significant
industrial
4.32 (max)
Minimal
56
1.74
8,500
10
Rhone-Poulenic
Herbicide
production
wastes
0.15 (max)
None
87
2.00
8.500
11
Reichhold
Chemicals
Chemical pro-
duction wastes
1.0 (max)
Moderate
100
1.55
32.500
12
Stepan Chemicals
Surfactant pro-
duction wastes
0.015 (max)
None
500
6,480
13
Republic Steal
Coke process
wastes
0.95 (max)
Minima!
116/58
2.3/4.6
68,000
14
LeRoy
Municipal
1.0 (max)
Extensive
12
_
12,000
15
Manchester
Water supply
40.0 (max)
Moderate
14
12.000"
16
Passaic Valley*
Water supply
2.2 (max)
Extensive
8
2.400
17
Colorado
Springs
Secondary
effluent
2.0 (max)
Extensive
17
4.5
1,800
18
Hercules
Chemical pro-
duction wastes
3.25 (max)
Moderate
48
6.6
33,600
19
Industrial
Sugar
Decoloring
sugar
Minimal
1080
12,000
20
Hopewell
Water supply
3.0 (mg)
Moderate
2.0
Norm
21
Davenport
Water supply
30.0 (max)
Moderate
7.5
2.0
None
22
Spracktes
Sugar
Sugar thick
juice
None
20
_
15.000
6 'Not available, "Oafs not collected. 'Facility under construction. "Fluiditad bed
-------
Furnace
Activated
Carbon Dose
Costs
Capital
Carbon
Loss.
%
Fuel
Type
Use.
Btu/lb
carbon
lb/mil
gal
lb organic/
lb cerbon
Capital,
SI000 (year}
S/tb/day
of
capacity
O&M,
$/mil gal year
8
Gas
2,900
207
0.38
849f19691
141.50
36.07 (1979)
S
Propane
3,840
0.33
1.569 (1976)
408.59
N/A'
to
LPG
2.750
250
0.33
3.880 (N/AJ
323.33
50.00 (1979)
9
b
N/A
N/A
7.5
1.410
0.50
4.359 (1974/
150.31
N/A
6
J
Natural
5.600
1.70
3,307(1972)
275.58
90.49 (1979)
gas
_
N/A
N/A
5
N/A
N/A
5
Natural
_
1,000
0.26
1.000 (1971)
117.65
490.00 (1973)
gas
8.8
Natural
6.500
43.000
0.38
300 (1969)
35.69
319.00(1973)
gas
5
Natural
6.200
79.000
0.26
N/A
N/A
gas
6
LPG
6,000
500.000
225 (1973)
34.72
25,000(1978)
8
Natural
35.000
N/A
N/A
gas
_
Fuel Oil
_
2.500(1975)
208.33
N/A
to
Fuel Oil
N/A
N/A
to
Electricity
0.7
_
N/A
N/A
7.5
160
0.50
N/A
89.59
5
Natural
7.000
9.500
0.44
1.622 (1973)
48.27
1,470.00
gas
2.5
Natural
_
N/A
N/A
gas
N/A
30.00 (19m)
,
_
N/A
20.00 (1980)
4.8
Natural
1.785
238(1958)
15.67
N/A
gas
GAC last facility.
7
-------
volume report "Estimating Water Treat-
ment Costs" cited earlier. Twelve cost
curves, discussed in detail in this report
were developed specifically for GAC
applications.
Derivation of Cost Curves
The construction cost for each unit
process was presented as a function of
the process design parameter. This
parameter was determined to be the
most useful and flexible under varying
conditions, such as loading rate, deten-
tion time, or other conditions that vary
because of designers preference or
regulatory agency requirements. For
example, the contactor construction
cost curves were presented in terms of
cubic feet of contactor volume, an
approach that allows various empty bed
contact times (EBCT) to be used.
Contactor O&M curves were presented
in terms of square feet of surface area,
because O&M requirements are more
appropriately related to surface area
rather than contactor volume. Reactiva-
tion facility cost curves were presented
in terms of square feet of hearth area for
the multiple hearth furnace and pounds
per day of reactivation capacity for the
other reactivation variables. This allows
the loading par square foot of hearth
area to be varied for the multiple hearth
furnace and for design of reactivation
furnaces at rates less than the reactiva-
tion capacity. This approach provides
greater flexibility in the use of the cost
curves than if the costs were related to
water flow through the treatment plant.
For the majority of the unit processes,
three separate figures were used to
present construction and O&M curves.
The first graph presented the construc-
tion cost; the second graph, energy
(electrical, natural gas, and diesal fuel)
and maintenance material require-
ments; and the third graph, labor
requirements and total O&M cost.
Treating water with GAC involves two
major and separate process operations:
filtration and reactivation. The water
comes in contact with GAC by passing
through a structure filled with activated(
carbon granules. Impurities are removed
from the water by adsorption when
sufficient time is provided for this
process. The structure can be either a
water treatment filter shell in which the
filter media has been replaced with
GAC, or In independent carbon adsorp-
tion system. A separate adsorption
system usually consists of a number of
columns or basins, used as contactors.
that are connected to a reactivation
system. After a period of use, the GAC's
adsorptive capacity is exhausted, and it
must then be taken out of service, either
by replacement with new carbon or by
being reactivated by combustion of the
organic adscrbate. Carbon is routinely
added to the system to replace that lost
during hydraulic transport and reactiva-
tion. These losses include both losses
caused by physical deterioration and by
burning during the reactivation process.
In the report, cost estimates have
been prepared for the use of GAC as a
replacement for existing filtration media
(sand replacement) and as a separate
adsorption system (post-filter adsorp-
tion). On-site multiple hearth reactiva-
tion was assumed. Standard levels of
key design parameters were set at
predetermined levels and then one
variable at a time is changed to
determine its effect on system cost.
Figure t shows the cost of both sand
replacement and post-filter adsorption
systems versus plant capacity.
Conclusions
Some conclusions resulting from the
inspection of 22 operating GAC absorp-
35**
tion and reactivation systems and a
study of the costs involved are:
1. A substantial number of GAC instal-
lations have been operating sucess-
fully for many years. These installa-
tions include facilities that treat
water and wastewater and that
process food and beverage products.
2. Although long-term use of GAC and
PAC by municipal waterworks and
industries has been widespread, no
adverse health effects have been
reported.
3. GAC treatment of public water
supplies to remove trace organics is
fairly new. Only limited direct design
data or cost information are avail-
able.
4. GAC treatment of public water
supplies for taste and odor control or
for color removal is practiced in
about 46 cities in the United States.
In this application of GAC, however,
little need for reactivation anc little
experience with on-site reactivation
exists.
5. The four reactivation studies that
EPA presently sponsors concern
installations in Cincinnati, OH;
Manchester, NH; Jefferson Parish,
LA' and Passaic Valley, NJ.
§20-
15-
Sand Replacement
Post-Filter Adsorption
60
120
160 ISO
20
40
80
Plant Capacity, mgd
140
Figure 1. Total production cost versus plant capacity for post-fitter adsorption and
sand replacement systems.
e
-------
6. Twenty to thirty GAC installations In
food processing plants (corn syrup
and beet sugar) give insight concern-
ing the safety of GAC but add little
need to the available cost data. Some
design ideas and information can be
obtained from these sources.
7. Twenty to thirty GAC reactivation
furnaces in industrial water pollution
control plants yield some information
on operation a id cost of GAC sys-
tems.
8. The best sources of detailed cost
information and equipment design
data are the 20 or so operating
municipal advanced waste treatment
(AWT) plants using GAC.
9. Pilot plant testing of GAC adsorptiva
and reactivation characteristics in
municipal water treatment plants
can be combined properly with the
known costs of GAC systems in
wastewater treatment to yield a
design basis and preliminary cost
esfmates that are satisfactory for
developing water treatment projects.
Designers should take full advantage
or applicable GAC use and experi-
ence available from:
municipal water treatment for
taste and odor control,
municipal AWT,
industrial wastewater treatment,
and
food and pharmaceutical produc-
tion.
The full report was submitted in
fulfillment of Contract No. CI-76-0288
by Culp/Wesner/Culp, Cameron Park.
CA, under the sponsorship of the U.S.
Environmental Protection Agency.
9
¦it U.S. GOVEMHENT PRINTING DFflCti (S8I--S5S-0M/3316
Russell L. Culp is with Culp/Wesner/Culp, Cameron Park. CA, and Robert M.
Clark (also the EPA Project Officer, see below} is with the Municipal Environ-
mental Research Laboratory, Cincinnati, OH 45268.
The complete report, entitled "Granular Activated Carbon Installations," was
authored by R. L. Culp, J. A. Faisst, and C. E. Smith of Culp/Wesner/Culp.
Cameron Park, CA (OrderNo. PB 82-102492; Cost: $21, SO, subject to change)
will be available only from:
National Technical Information Service
5235 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati. OH 45268
------- |