PB87-202925
Cost and Performance Evaluation for Full Scale
Single Solute Control of Synthetic Organic
Chemicals by Granular Activated
Carbon Adsorption
(U.S.) Env ironmental Protection Agency
Cincinnati, OH
Jun 87
I
M* «f
-------
Pii87-2029ii»
EPA/600/D-87/205
June 1937
COST AND PERFORMANCE EVALUATION FOR FULL SCALE, SINGLE
SOLUTE CONTROL OF SYNTHETIC ORGANIC CHEMICALS BY GRANULAR
ACTIVATED CARBON ADSORPTION
by
Jeffrey Q. Adams
Environmental Engineer
Robert M. Clark, Director
Drinking Water Research Division
Richard J. Miltner
Environmental Engineer
U.S. Environmental Protection Agency
Water Engineering Research Laboratory
Cincinnati, OH 45268
EPA Project Officer
Jeffrey Q. Adams
WATER ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
HEPROOUCEOBV
US DEPARTMENT OF COMMERCE
NATIONAL TECHNICAL
INFORMATION SERVICE
SPRINGFIELD, VA 22161
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completingj
1. rttPORT NO.
EPA/600/D-8? /20*5
2.
«. title and subtitle
COST AND PERFORMANCE EVALUATION P{® PULL
SCALE,
S. REPORT DATE
June 1987
SINGLE SOLUTE CONTP.OL OF SYNTHETIC ORGANIC
CHEMICALS BY GRANULAR ACTIVATED CARBON ADSORPTION
6. PERFORMING ORGANIZATION CODE
7. AUTMORIS)
Jeffrey Q. Adams, Robert
Richard J. Milliner
M. Clark and
B PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANiZA_ION NAME AND ADDRESS
DWRD/WERL/USE PA
tO. PROGRAM ELEMENT NO.
2f W. St. Clair
Cincinnati, OH
11 CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Mater Engineering Research Laboratory-Cinti., OH
13. TYPE OF REPORT ANO PERIOD COVERED
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
14. SPONSORING AGENCY COOE
EPA/600/14
f
IS. SUPPLEMENTARY NOTES
Jeffrey Q- Adams - COML.
(513)589-7835, FTS (513)684-
7835
16. ABSTRACT
The Drinking Water Research Division of the U.S. Environmental Protection
Agency has the responsibility of obtaining data on the performance and cost
of granular activated carbon (GAC) systems for controlling selected synthetic
organic chemicals (SOCs). An approach is presented that utilizes a combination
of predictive modelling and microcolumn studies to geaerate single-solute
breakthrough curves for .several SOCs and to determine full-scale GAC use rates.
Based on breakthrough and economic data, cost estimates have been developed for
GAC treatment at various system sizes and use rates.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTlFlEBS/OPEN ENDED TERMS
c. COSATt Field/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19 SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
28
20, SECURITY CLASS {This parei
UNCLASSIFIED
22. PRICE
EPA F-wm 2220-1 !R»*. <-?7) PREVIOUS EDITION IS OBSOLETE ^
if I
-------
NOTICt
This document has been reviewed in ?<-cordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
ii
-------
COST AND PERFORMANCE EVALUATION FOR FULL SCALE, SINGLE
SOLUTE CCNTROL OF SYNTHETIC ORGANIC CHEMICALS BY GRANULAR
ACTIVATED CARBON ADSORPTION
Jeffrey Q. Adams
Environmental Engineer
Robert M. Clark, Director
Drinking Water Research Division
Richard J. Miltner
Environmental Engineer
U.S. Environmental Protection Agency
Water Engineering Research Laboratory
Cincinnati, OH 45268
INTRODUCTION
For each contaminant regulated under the Safe Drinking Water Act, a
t.reat&ant technique must be specified and treatment techniques may
also be required where It is infeasible to ascertain the levels for
regulated contaminants. Recent amendments to the Safe Drinking Water
Act have accelerated the requirements for development of Maximum
Contaminant Levels (MCLs) and increased the number of contaminants
that will be regulated.^
These amendments will broaden and deepen the impact of the Safe
Drinking Water Act on public water utilities. The EPA Administrator
Is required to establish a feasible treatment technology for each
regulated contaminant using the best technology available and other
means taking cost into consideration.
The amendments specify that granular activated carbon (GAC) for
synthetic organic chemical (SOC) control is considered to be feasible.
Therefore, any treatment technique for control of SOCs must be as
effective as granular activated carbon. For each National Primary
Drinking Water Regulation that establishes an MCL the Administrator of
EPA must list the technology and treatment technique that he determines
are feasible for meeting an MCL. In effect, GAC is a baseline tech-
nology against which other technologies must be evaluated.
The Drinking Water Research Division (DWRD) of EPA has a major role
in evaluating the performance of any technology that might be con-
sidered as part of an MCL. Traditionally EPA's technology program
has utilized the approach of carefully evaluating process kinetics on
the bench, scaling up for engineering feasibility at the pilot scale,
and making a field-scale evaluation to incorporate process economics.
This approach may take several years for completion and may Involve
millions of dollars for each unit process examined- Regulatory
3
-------
pressure will no longer allow for a contaminant-by-contaminant and
proce. ^-by-process evaluation.2
In anticipation of the need to use process simulation techniques
incorporating small scale experiments coupled with cost models, EPA's
Drinking Water Research Division created a peer-review panel to
examine the Division's GAC adsorption technology research and to
provide recommendations for future directions. The peer panel,
consisting of some of the World's foremost experts on carbon adsorp-
tion, as well as knowledgable professionals from Industry, consulting,
and the water utlities, «et for two days in intenrive deliberation.
The panel's recommendations included: For selected SOCs, existing
adsorption isotherm data and field data were sufficient for projecting
adsjrpt ion costs. For those SOCs which are relatively poorly adsorbed
or those for which field data are lacking, the panel agreed that
microcolumn studies being conducted by DWRD should.continue. In
¦accordance with the consensus of the peer panel review, usage rates
calculated from model application of isotherm data are developed for
the EPA's Office of Drinking Water (ODW) for regulatory development,
if the projected full-scale bed life is greater than two-three years
(traditionally considered to be economically feasible for taste and
odor control). For shorter bed lives, usage rates are determined
from microcolumn operation designed to develop breakthrough curves that
can be scaled-up to full-scale operation. In addition, data based on
pilot-scale experience will also be scaled up to full-size operation.
All data are provided at a minimum for single solute systems at
hydraulic conditions and field-scale concentration ranges of Interest
to ODW. Usage rates from full-scale experience are provided where
available.
Adsorption isotherm and microcolumn studies have been conducted for many
of those compounds considered for regulation. The purpose of this
paper is to present the results of this research and in conjunction
with previously developed cost data, project full-scale equivalent
cost and performance for these compounds.
ADSORPTION MODELING
The adsorption protocol involves developing single solute isotherms
In both distilled and natural waters. Using Freundlich Isotherm
parameters, the homogeneous surface diffusion model (HSDM) is employed
to generate breakthrough curves and predict carbon usage rates for
fixed-bed G,tC column operation both at micro- and full-scale.2 Other
Inputs to the HSDM are kinetic parameters estimated from correlations
given In the literature and parameters defining the carbon type and the
system hydraulics. A HSDH generated full-scale breakthrough curve is
given in Figure 1.
For poorer adsorbed SOCs, where full-scale predications show shorter
service lives, microcolumn operations are conducted. Because micro-
column breakthrough curves represent empirical kinetic phenomenon
rather than correlated estimates, the microcolumn breakthrough curves
are scaled-up to full-scale conditions. Both the scaled-up usage
rates tor the poorer adsorbed SOCs and the HSDM generated use rates for
other SOCs are us.-d as inputs to cost models.
-------
Based on the HSDM, the microcolumns are designed to give constant-
pattern breakthrough curves upon which scale up is based. The appa-
ratus consists of a 4 mm inside diameter glass column packed with 100 x
200 mesh Flit rasorb 400 GAC at a depth sufficient to contain the pre-
dicted mass transfer zone for the compound. Other requirements are a
prefilter to prevent excessive head loss across the column, dampers
to smooth out pump delivery and maintain constant pressure on the
column, and a flexible head space bag to restrict loss of volatile
contaminants. Figure 2 shows a schematic of the microcolumn apparatus.
For SOCs having predicted full-scale bed lives of less than two
years, microcolumn breakthrough curves may be generated in less than
5 days. This is compatible with the short regulatory time frame.
Figure 3 shows a typical microcolumn breathrough curve.
Once the model predicted or scaled-up full-scale breakthrough curves
are developed for each compound, field-verified cost equations are
used to generate, full-scale costs. A computer program has been de-
veloped that uses breakthrough curve data to calculate carbon usage
rate, and ultimately cost data for a wide nuuiber of carbon systems.
A set of costs generated from this procedure at various sizes to
application has been developed. For the data prepared for this
analysis, it is assumed that the carbon systems were consistent with
the following conditions: EBCT = 15 minutes; loading rate = 4 gpm/ft2,;
carbon type = FlItrasorb 400. These were the conditions that were
used in the HSDM full-scale predictions and the microcolumn scale-up.
Table 1 summarizes isotherm data and results from microcolumn opera-
tions .
The usage rates predicted in Table 1 are largely unverified. In Dw'RD
labortories, pilot column studies are underway with selected SOCs to
verify both the HSDM predictions and the .microcolumn scale-up. Further,
both short-bed adsorber studies and differential column batch reactor
studies are underway to develop adsorption kinetic parameters in the
place of correlations taken from the literature.
COST MODEL DEVELOPMENT
In response to early concerns about impacts of cost on drinking water
utilities due to the Safe Drinking Water Act, the DWRD initiated a
study to develop standardized cost data for 99 water supply unit
processes.3 The approach was to assume a standardized flow pattern
for the treatment train and then to estimate the cost of the unit
processes. This approach requires assumptions about such details as
common wall construction and amounts of interface and yard piping
required. After the flow pattern was established the costs associated
with specific unit processes were calculated. As built designs and
standard cost reference documents were-used to calculate the amount
of excavation, framework, and materials such as concrete and steel.
Information from existing plants and manufacturers were used to
calculate the costs of equipment associated feith a unit process
Once basic information has been calculated, capital cost curves were
developed. In 1984, three years after the first set of reports was
issued, another report was issued containing cost curves for "small
systems technology", using the same methodologies-
The construction cost for each unit process considered in both of the
cost studies was presented as a function of the process design para-
meter that was determined to be the most useful and flexible under
5
-------
varying conditions. Such variables as loading rate, detention time,
or other conditions that can vary because of designer's preference or
regulatory agency requirements were used. For example, GAC contactor
construction costs are presented in terse of cubic feet of contactor
volume, an approach that allows various empty bed contact times to be
used. Contactor operation and maintenance curves are presented In
terms of square feet of surface area because operation and aaintenance
requirements are more appropriately related to surface area than to
contactor volume. Such an approach provides Information more than if
the costs were related to water flow through the treatment plant.3
Parallel to these cost studies, the Drinking Mater Research Divsiion
had been conducting extensive field scale studies on the use of
granular activated carbon adsorption. These field studies yielded a
great deal of useful cost data which was integrated in the previously
mentioned cost studies. Cost equations have been developed from these
data for various GAC unit processes. These equations are described
in the following section.
DEVELOPMENT OF COST EQUATIONS
To make this data more useful and transportable, the cost curves for
the small technologies were converted to a set of equations.^ The
functional form of the estimating equations used in this develop-
ment Is as follows:
Y = a + b (USERATE)C(d2) (1)
where
Y = operating and maintenance or captial cost
USERATE = design or operating variable
a,b,c,d = coefficients determined from nonlinear regression
2 = 0 or 1 used to adjust cost function for a range of use-
rate values.
The dummy variable effectively changes the slope o- the function
shown In equation (1) for a given value of the independent variable.
If the data when plotted on log-log scale fits a straight line or has
a -mild bend or curve it does not need a ~dz" factor. Deciding if the
d2 factor is needed is done on a case by case analysis.
For example the following equation gives the construction cost of a
small system steel'pressure contactor:
CC - 16125 4- 7632.0 (CUFT)0-5229 (1.10)2 (2)
where
CC = construction cost for steel contactor
CUFT = reactor volume in cubic feet
6
-------
z = 0 for CUFT < 400
i. « 1 for CUFT > 400
The application of these cost equations Is discussed In the following
sections. Table 2 contains the cost factors used in the analysis.
APPLICATION OF COST EQUATIONS
The cost equations discussed in the previous section can provide
Insight into the sensitivity of cost of varying system parameters.
For exammple, Figure 4 shows the cost of concrete gravity and steel
pressure post filter contactor versus system capacity. In the
analysis the average flow is assumed to be 50% of design capacity.
In this analysis it can be seen that for systems larger than 5 mgd
(10, 15, 20 minutes EBCT), concrete gravity contactors are cheaper
then steel pressure contactors. A smaller number of larger sized
contactors are cheaper than a latter number of smaller sized con-
tactors. For smaller systems str 1 pressure contactors are cheaper
than concrete contactors. In ger ra1, the cost breakdown for adsorbers
(as a percent of total cost) is as follows: capital 75-90%; labor
5-202; power 4%; and, maintenance material 1%. Adsorbers are obviously
capital intensive.
Figure 5 shows the sensitivity of total contactor cost to EBCT and
system capacity. As might be expected shallower (shorter EBCT)
contactors are less expensive than deeper contactors. However, the
lower the <"BCT for a given effluent target, the mora frequently the
carbon must be reactivated. Therefore, as will be seen later, to
minimize cost with respect to EBCT, the higher cost of a deeper
contactor must be offset by the decreased cost of less frequent
reactivation.
On site reactivation Is an Important consideration in making GAC cost
effective. Figure 6 shows the costs in jf/lb of fluid bed, infrared,
and multihearth reactivation versus GAC reactivated in million
pounds/yr. These preliminary cost estimates are based on an assump-
tion of 75% uptime for each furnace. Furnaces ranged in size from
100 lbs/hr to 35,000 lbs/day. Based on the data presented in Figure
6, infrared reactivation appear more cost effective for small systems
(< 2 million lbs/yr) but fluid bed reactivation appears raore cost
effective for large system (> 2 million lbs/ yr). In general,
total reactivation costs are as follows: capital cost 20-35%; GAC
makeup 20-40%; labor 15-202; power 13-18%; materials and process
water 71.
Figure 7 shows a more detailed comparison of reactivation costs for
small reactivation requirements. Comparison is made between infrared
reaction, off-site reactivation, and disposal and replacement.
Disposal and replacement includes the cost of virgin carbon replace-
ment plus ultimate disposal by incineration for spent carbon. The
infrared furnace is based on a capacity of 85 lbs/hr, with excess
capacity. Commercial off-site reactivation includes reactivation,
losses and transport up to 500 miles one-way.
Figure 8 shows construction cost for complete GAC systems in 1987
dollars as a function of capacity. Figure 9 shows the sensitivity of
GAC system costs to variations in GAC, labor and power costs.
-------
Scenario 1 assumed $0.95/lb for GAC, $15/hr for labor and $G.G8/kwhr
for power, and scenario 2 assumed $0.85/ib, $12/hr and $0.06/kwhr.
As can be seen these cost, changes have little impact on overall
svsteia cost.
Figure 10 illustrates the dramatic effect that bed life and economies
of scale with respect to plant capacity have on unit costs for GAC
systems assuming an average flow of 50% capacity.
COST AND PERFORMANCE ANALYSES
In this section, data derived from HSDM runs using laboratory developed
isotherm data and microcolumn data (Table 1) is coupled with cost in-
formation to provide cost and performance comparisons for various
system scenarios.
EBCT VERSUS COST
In order to study the effect of increases in bed Ufa or EBCT on
cost, two compounds were studied. One is a volatile compound, cls-DCE,
the other Is EDB. To study these compounds specific Influent concen-
tration and effluent targets were assumed.
For cis-DCE, HSDM predicted breakthrc profile were developed for
three EBCTs (5 rain; 10 rain; 15 Bin) p.: s'-.own In Figure 11, assuming
a constant influent of 200 ug/L and a . "get effluent of 70 ug/L. As
EBCT increases bed life, the time to breakthrough increases. Figure
12 shows the costs for a 10 mgd system that iesult from using the
data derived from Figure 11. As can be seen the cost of the contactors
increase with increasing EBCT but reactivation, as might be expected,
decreases with EBCT. The Increase in contactor cost cannot be offset
by the decrease in reactivation cost and the minimum cost is achieved
at the minimum EBCT.
Figure 13 shows HSDM modeling results for EDB assuming that the
influent concentration is 100 ug/L and the target effluent con-
centration is 1 ug/L. Figure 14 shows the results of the trade-off
be tween increasing contactor costs due to increased EBCT and the
decreases in reactivation costs due to less frequent reactivation.
The reactivation cost decreases are large enough to offset rising
adsorber costs up to 14 mln EBCT. In the range from 14 to 17 minutes
EBCT the cost curves are relatively flat so that one could maintain a
conservative design of an EBCT of 17 minutes for a 1 rf/1000 gal in-
crease In total system cost.
This concept has been applied in Figure 15 to those SOCs that have
potential for regulation (Phase II). HSDM ar.d microcolumn method-
ologies have been applied to these compounds. An assumed EBCT of 15
minutes has been chosen and the most conservative bed life (between
HSDM and microcolumn runs) selected for cost analysis. Figure 15
shows the costs In ^/1000 gal versus bed life for three different
sized systems, assuming an average flow to maximum capacity ratio of
50%. The worst case cost scenario would be $1.6/1000 gal for a i mgd
system reactivating every 30 days. For bed lives from 6 months to 2
years the cost curves are fairly flat so that even if usage predic-
tions are not precise they will have little impact in terms of cost.
Most non-stripable SOCs will fall into the 6 month to 2 year bed life
range. Therefore, for example, the cost of a 1 mgd system will range
8
-------
from 55 to 70 ^/1000 gal, a 10 mgd system will range from 25 to 35 it
1000 gal,and a 100 mgd system will range from 14 to 18 i/1000 gal.
Because the curves are flat In this bed life range, by reactivating
earlier, a lower TOC effluent could he maintained at a marginal
increase In unit cost for systems controlling the highly adsorbable
compounds -
Table 3 provides typical costs for various bed lives (days) and
system sizes assuming an EBCT of 15 minutes at 50% capacity.
CONSUMER COSTS
In order to provide some perspective on the impact that these costs
will have on the consumer, a four person household, consuming 70
gallons per capita per day is assumed. These assumptions yield a
yearly consumption of 100.8 thousand gallons. Table 4 shows result-
ing cost impact.
SUMMARY AND CONCLUSIONS
The amendments to the Safe Drinking Water Act will require an exten-
sive evaluation of the feasibility for removing organic compounds
using GAC. In order to meet very short deadlines for technology
evaluation, the Drinking Water Research Division has incorporated the
use of microcolumn and adsorption modeling combined with cost models
to make full scale projections for the performance of GAC systems.
Analysis has been performed on a representative list of compounds.
For poorly adsorbed compounds, the minimum cost system is associ-
ated with shorter EBCTs. However, this must be evaluated case by
case, and depends on influent concentration and effluent goals. For
some compounds there may be a range of EBCTs at which cost is near
minimum as shown in the EDB example. All of these compounds have
been analyzed as If they were single solutes when in reality they
will occur in nature as part of a mixture. However the analysis
presented in this paper should be useful for preliminary planning.
ACKNOWLEDGMENTS
The authors would like to acknowledge the assistance of Ms. Patricia
Pierson and Ms. Diane Routledge in the preparation of this manuscript.
REFERENCES
1. Safe Drinking Water Act 1986 Amendments, United States Environ-
mental Protection Agency, Office of Drinking Water, EPA 570-9-
86-002, August 1986, Washington, DC 20460
2. Clark, R. M., Adams, J. Q- and Miltner, R. J., "Cost and Perform-
ance Modeling for Regulatory Decision Making", Water, vol. 28,
No. 1, Spring 1987, pp. 20-27.
3. Clark, R. M., Adams, J. Q., and Eilers, 1. G., "Cost Models for
Small Systems Technologies: U.S. Experience", to appear in the
Proceedings of the International Conference on Mobilization
for Drinking Water Supply and Sanitation in Developing
Countries, held in San Juan, Puerto Rico, May 26-29, 1987.
9
-------
TABLE 1. PREDICTED FULL-SCALE GACJJSK RATES4 FOR _SINGLE-SOLUTE SOCs
Freundlich
Isotherm
umol { L ~)l/n Selected Scaled-up from
g umol Criteria Microcolumns HSDM Predicted
ug/L use rate bed life use rate bed life
soc
K
1/n
Co
Ce
(lbs/1000 gal)
(days)
(lbs/1000 gal)
(days
Trans-1,2-Dichloroethylene^
38.5
.391
500
70
0.6722
64
0.6146
70
Cis-1,2-Dichloroethy lene^-
29.4
.64
200
70
3841
112
Toluene2
475.5
.328
5000
2000
_
0.2607
165
1,2-Dichloropropanc^
35.7
.48
100
6
0.2672
161
0.2403
179
1,2-Dibromoethane (EDB)
69.3
.54
100
1
0.1138
378
0.1964
219
Aldicarb^-
193.7
.41
500
9
0.0956
450
0.0989
435
Monochlorobenzene^
381.2
.31
600
6
0.0690
623
0.0641
671
Alachlor^
327.6
.38
100
0.6
_
—
_
2849
Carbofuran^
375.8
.36
100
36
_
-
_
3163
DBCP1
187
.446
50
0.1
0,0241
1785
-
1300
o-Dichlorobanzene^
865.2
.38
1000 .
620
-
—
_
1,361
Ethylbenzene2
694.1
.293
1000
680
•
_
_
738
Lindane^-
455.2
.39
10
0.2
_
16805
Pentachlorophenol1
403.3
.39
1000
220
0.0571
753
-
750
m-Xylene2
1043
.246
1000
440
_
_
_
982
o-Xylene2
897
.259
1000
440
_
_
_
869
10hlo River Water
2Wausau Distilled Mater
%Iiili-Q Distilled Water
^Single Contactor, 4 gpm/ft2, 15 min EBCT, FUtrasorb 400
-------
TABLE 2 - FACTORS USED IN COST ANALYSIS
Tit operated at 50% capacity
- Reactivators operated at 75% up-time
- GAC Reactivator losses at 10%
- Capital Amortized at 10% over 20 years
- GAC Price at -95 $/lb at 100,000 lbs
- Labor rate at 15 $/hour
- Electric rate at «08$/KWH
- Fuel Oil at .95 $/gal
- Water at .35 $/1000 gal
- Cost Indices for Early 1987
TABLE 3. TOTAL COST BY SYSTEM S'IE AND USERATE
EBCT =15 minutes
«!/1000 gal
Bed MGD
Life .25 .5 1 5 F0 ~25 30 75 TOO 20
(days)
7
542.5
389.0
295.2
160.1
153.3
130.3
117.4
115.4
114.1
111.2
1A
309.2
287.6
219.4
111.9
87.7
78.1
70.7
67.6
63.5
61.3
30
181.9
163.0
161.5
80.
62.0
45.9
43.1
38.9
38.8
34.6
50
137.3
118.4
116.9
66.2
51.1
37.2
31
30.7
28.5
25.6
70
118.2
99.3
97.8
59.9
45.3
33.2
27.0
26.9
24.9
22.2
90
107.6
88.7
87.2
55.4
42.0
30.9
24.9
22.3
20.9
19.2
112
100.3
81.4
79.9
52.3
39.7
29.1
23.4
20.9
19.5
17.8
150
92.8
73.9
72.4
48.4
37.0
26.8
21.8
19.4
18.0
16.4
179
89.2
70.3
68.8
44.8
35.4
25.7
21.0
18.7
17.3
14.7
212
86.2
67.4
65.9
41.9
34.2
24.8
20.3
18.1
16.7
14.1
255
83.6
64.7
63.2
39.3
33.1
24.0
19.5
17.5
16.2
13.6
304
81.5
62.6
61.1
37.1
31.8
23.3
18.9
17.0
15.7
13.2
365
79.6
60.8
59.3
35.3
29.9
22.6
18.3
16.5
15.3
12.9
545
76.6
57.7
56.2
32.3
26.9
21.4
17.4
15.6
14.5
12.2
730
75.1
56.2
54.7
30.7
25.4
20.6
16.9
15.1
14.0
11.9
11
-------
TABLE 4. CONSUMER COSTS
System
Capacity
(MGD)
Worst Case
30-day
Bed-Life
(S/yr)
6-Month
Bed-Life
<$/yr)
2 Year
Bed-Life
($/yr)
1
163
69
55
10
62
36
26
100
39
17
14
12
-------
Predicted Effluent Concentration Profile for cIs-DCE
EBCT-15 mln 4 gpm/sqft Loading
W
3
>3>
c
0
c
m
CJ
c
w
2,
£3
200 :
180-
160-
140-
120-
100-
80-
...
60-
40-;
20-
o
prrrTj-r
100 105
FIGURE
f ^ ^ |*1'"'f/ '
-------
FIGURE 2. MICROCOLUMN APPARATUS
-------
•On
ictomi «sorpi)n cf
mo-
no-
Influent
effluent
Ohio River
cart»ff=filtTOorb 400
roesb=10(K2O()
bed cfepth=4L9 cm
tooding ncrte=57.4 gpnry^txg
Throughput Time, hours
FIGURE 3. TYPICAL MICROCOLUMN BREAKTHROUGH CURVE
-------
50
45
40
35
30
25
20
15
10
5
0
Concrete Gravity
Steel Pressure
• f i •?••• i '-"r T'T f t—r
1—i—!—|—t 'v i—|—i—i—r ¦' g—i—i—i—p1 r-'f"1?11111—rr T "i—|
20 40 60 80 100 120 140 160 180 200
System Capacity (MGD)
4. COST ESTIMATES FOR 6AC POST FILTER ADSORBERS AT EBCT-15 MIN.
-------
System Capacity (MGD)
FIGURE 5. COST ESTIMATES FOR CONCRETE GRAVITY GAC
POST FILTER ADSORBERS .
-------
Fluid Bed
Infrared
Multihearth
,rm mr»mr?tn rmrmr yi u rm uiupm n rrt t f m m m h m m i n n h
Amount of GAC (Million Pounds/Year)
FIGURE «. PRELIMINARY COST ESTIMATES FOR GAC
REACTIVATION ALTERNATIVES <75% UPTIME)
-------
Infrared React
190-1
170 -
c
3 ISO •
O
w
c
Q)
m
8
¥
3
130-
110-
70-
\
\
V.
Off—Site React _ __
Dispose & Replace
50 - -T-T-T-T-pT-T^nt_p
r'T'l 'ft t t i" r I r i -r-r-r-j-T't'T1 r"j v "t""> -i-r-T-r-irf-T-r-r^
0 50 100 150 200 250 300 350 400 450 500
Amours! of GAC ( 1000 Pounds/Year )
FIGURE 7, COST ESTIMATES FOR REACTIVATION ANO REPLACEMENT
OF SMALL QUANTITIES OF CARBON
-------
35.0'
3 MONTH BED-LIFE
W :
S 15.0-i
frr"f i" [ | i i f j'f rr |- -t-t r r-'i—r—|—r r t- |'"f v t | i t i | » f * |
0 20 40 Sfi '80 100 120 140 160 180 200
System Capacity (MGD)
FIGURE 8. PRELIMINARY CONSTRUCTION COST ESTIMATES FOR
QAC SYSTEMS AS A FUNCTION OF CAPACITY
-------
45
40
35
30
25
20
15
10
; §,
nmn|>mmnymirer tptrmtr*pmHmjtitTmytpymnvtpmmt»p*mtmpm afmpnt
70 140 210 280 350 420 490 560 630 700
GAC Bed Life (Days)
NSITIVITY OP GAC SYSTEM COSTS TO VARIATIONS IN COST FACTORS
-------
w
hi
m
c
0
1
O)
o
o
o
4f
m
8
70'
60 -¦
50 H
40 h
30 -i
J 20-:
V)
ifr
3 MONTH BED—LIFE
6 MONTH BED—LIFE
2 YEAR BED-LIFE
10
{ "T""> ryrrryi "t n i:y t—i—r-j-r" r-r ri—r—f t-r-i—;""T- r'-f-y
0 20 40 60 80 100 120 140 160 180 200
System Capacity (MGD)
FIGURE 10. TOTAL COST ESTIMATES FOB 6AC SYSTEMS AT
EBCT-15 MIN. AS A FUNCTION OF CAPACITY
-------
i\S
CC
2001
180 :
qi
«0:
-
r.
HO-
0
*
o
120-
i-
-
i
100-
§
BO -
c
60 :
La«
D
3
.40-
X
V -
.v —,
I. j
20 :
0:
MCLQ
38 OAYS
/
/
/
/
39 DAK
/
0
I
-------
t
EBCT (Minutes)
FIGURE 12. OAC COST ESTIMATES FOR CONTROLLING clt-DCE AT A
10 MGD SYSTEM, Co-200 ug/L, C«-70 ug/L
-------
90
80
70
60
50
40
30
20
10
\ 0
FIG
RUNDAYS
13. FULL-SCALE HSDM PREDICTED BREAKTHROUGH CURVES FOR IDS
Co-100 ug/L
-------
EBCT (Minutes)
FIGURE 14. GAC COST ESTIMATES FOB CONTROLLING IDS AT A
10 MGD SYSTEM, Co-100 ug/L, C®-1 ug/L
-------
o>
160 ¦:
140-1
120
X 100 ^
w
§ 80-
<£*
H 60
E,
i
40'
20-
0-3
1 MGD SYSTEM
JOMS&SYSTBM^
100 MGD SYSTEM
CH10R0BENZENE
85«
14#
Tiff ft l'!'| f I T I'l f f I ?t| I f I H't H # [ ?>' Iff I ff f | » f H i^f V'l | ITItl S fTlj? fll 1 * !'f !|TTTfTTTf fj|"l ff fiyff'fl H1
6 70 140 210 280 350 420 490 560 630 700
GAC Bed Life (Days)
FIGURE 15. GAC COST ESTIMATES FOR CONTROLLING SEVERAL
SOCi AT EBCT-15 WIN,
------- |