ROBERT A. TAFT WATER RESEARCH CENTER
REPORT NO.TWRC-11
APPRAISAL OF
GRANULAR CARBON CONTACTING
PHASE I
EVALUATION OF THE LITERATURE
ON THE USE OF GRANULAR CARBON
FOR TERTIARY WASTE WATER TREATMENT
PHASE II
ECONOMIC EFFECT OF DESIGN VARIABLES
ADVANCED WASTE TREATMENT RESEARCH LABORATORY -XI
U.S. DEPARTMENT OF THE INTERIOR
FEDERAL WATER POLLUTION CONTROL ADMINISTRATION
OHIO BASIN REGION
Cincinnati, Ohio
-------
APPRAISAL OF GRANULAR CARBON CONTACTING
PHASE I
EVALUATION OF THE LITERATURE ON THE
USE OF GRANULAR CARBON FOR TERTIARY WASTE WATER TREATMENT
PHASE
ECONOMIC EFFECT OF DESIGN VARIABLES
Produced under the auspices of
The M. W. Kellogg Company, A Division of Pullman Incorporated
for
THE ADVANCED WASTE TREATMENT RESEARCH LABORATORY
Robert A. Taft Water Research Center
This report, the first two phases of a three
phase program,is submitted in partial fulfillment
of Contract No. 14-t2-105 between the Federal
Water Pollution Control Administration and the
Swindell-Dressier Company, A Division of Pullman
Incorporated.
U. S. Department of the Interior
Federal Water Pollution Control Administration
Cincinnati, Ohio
MAY 1969
-------
FOREWORD
In its assigned function as the Nation's principal natural resource
agency, the United States Department of the Interior bears a special
obligation to ensure that our expendable resources are conserved, that
renewable resources are managed to produce optimum yields, and that all
resources contribute their full measure to the progress, prosperity, and
security of America now and in the future.
This series of reports has been established to present the results
of intramural and contract research studies carried out under the guidance
of the technical staff of the FWPCA Robert A. Taft Water Research Center
for the purpose of developing new or improved wastewater treatment methods.
Included is work conducted under cooperative and contractual agreements
with Federal, state, and local agencies, research institutions, and indus-
trial organizations. The reports are published essentially as submitted
by the investigators. The ideas and conclusions presented are, therefore,
those of the investigators and not necessarily those of the FWPCA.
Reports in this series will be distributed as supplies permit. Requests
should be sent to the Office of Information, Ohio Basin Region, Federal
Water Pollution Control Administration, 4676 Columbia Parkway, Cincinnati,
Ohio 45226.
The total resources of Pullman Incorporated were made available for
the execution of this program. The first two phases, included herein,
were primarily the responsibility of the M. W. Kellogg Division with
support provided by the Swindell-Dressier Division. The principal engi-
neering staff members involved included A. E. Cover, L. J. Pieroni and
E. V. Rymer. Coordination and participation were provided by J. F. Skelly,
Project Director, and C. D. Wood, Project Manager. The third phase,
"Engineering Design and Cost Estimate of a Granular Carbon Tertiary Waste
Water Treatment Plant," is published separately as TWRC-12.
ii
-------
TABLE OF CONTENTS
Page No.
PHASE I. EVALUATION OF THE LITERATURE ON THE USE OF GRANULAR
CARBON FOR TERTIARY WASTE WATER TREATMENT
Summary . . . . 1
Introduction . = 3
Allowable Capacity of the Carbon . . . 4
Effect of Linear Velocity 9
Effect of Particle Size 22
Effect of Regeneration 29
Discussion 41
Recommendations 44
References 46
PHASE II. ECONOMIC EFFECT OF DESIGN VARIABLES
Summary 48
Introduction 50
General Process Description 53
Design Bases 55
Economics
Comparison of Shop Fabrication
and Field Erection of Vessels 60
Comparison of Surge Design and Base Case . ... 60
Effect of Plant Size on Economics 60
Effect of Gravity-Flow Contactor
at Two Plant Sizes on Economics 63
-------
TABLE OF CONTENTS (Cont.)
Page No.
EC onomic s (Cont.)
Effect of "Idle" Carbon Inventory on Economics 68
Effect of Velocity on Economics ..... 70
Effect of Contact Time at High Velocity on Economics ... 70
Effect of Contact Time at Low Velocity on Economics ... 72
Effect of Velocity and Contact Time on Economics ..... 75
Effect of Particle Size on Economics 75
Effect of Regeneration Loss on Economics 77
Effect of Particle Size and Regeneration
Loss on Economics . . 77
Effect of Particle Size and Contact Time
on Economics 80
Effect of Particle Size, Contact Time,
and Regeneration Loss on Economics . 83
Effect of Carbon Capacity on Economics ... 83
Effect of Contactor System Type on Economics 85
Effect of Adsorbent Cost on Economics 89
Effect of Number of Contacting Stages on Economics .... 89
Comparison of Pomona, Tahoe, Improved
Downflow and Moving-Bed Systems 95
Effect of In-Place Regeneration on Economics 96
Conclusions and Recommendations 100
References 103
Appendix A - Procedure for Calculation of Operating Costs . . 105
References for Appendix A 109
-------
ABSTRACT
A literature review of the data on tertiary waste water treatment
has been made with a view towards generating sufficient basic data for
.process designs of various schemes for activated carbon adsorption of
COD (Chemical Oxygen Demand) from waste water. Particular attention
was given to the allowable capacity (loading) of carbon with organic waste
matter and the effect of liquid linear velocity, carbon particle size,
and number of regneration cycles on adsorption capacity and rate.
Findings of the work on tertiary waste water treatment using granular
activated carbon are summarized. The economic importance of several design
variables are determined and additional experimental data requirements
identified. The variables, which were studied for their effect on the
economics, were: shop fabrication and field erection of vessels, surge
designs, plant size, idle carbon inventory, velocity, contact time, par-
ticle size, regeneration loss, carbon capacity, downflow, upflow, gravity-
flow and moving-bed contactors, adsorbent cost, number of contacting stages,
in-place regeneration, and certain combinations of the above variables.
Recommendations are made for further evaluation and experimental work.
The principal source of the data utilized in this project was the
granular activated carbon treatment pilot plant operated at Pomona,
California by the Los Angeles County Sanitation District for the Federal
Water Pollution Control Administration. Additional information was collected
from carbon treatment facilities located at Lake Tahoe, California,
Nitro, West Virginia, Washington, New Jersey and Lebanon, Ohio. Other
data was obtained from published and unpublished sources as listed under
References.
v
-------
PHASE I. EVALUATION OF THE LITERATURE ON THE USE OF GRANULAR
CARBON FOR TERTIARY WASTE WATER TREATMENT
by
A. E. Cover and L. J. Pieroni, The M. W. Kellogg Company
SUMMARY
A literature review of the data on tertiary waste water treatment has
been made with a view towards generating sufficient basic data for process
designs of various schemes for activated carbon adsorption of COD (Chemical
Oxygen Demand) from waste water. In this review, particular attention was
given to the following critical process parameters:
1. Allowable capacity (or loading) of carbon with organic waste
matter (reported as pounds of COD removed per pound of car-
bon) to purify water to an acceptable level. An allowable
loading of 0.87 Ib COD/lb carbon was chosen, based on Pomona
pilot plant data for fixed bed, downflow systems assuming
feed water characteristics comparable to those found at Pomona.
2. Effect of liquid linear velocity on adsorption capacity
(loading and rate.) The literature data on the effect of linear
velocity on adsorption rate are conflicting. Some of these data
indicate an increase in adsorption rate with velocity. If this is
true, reductions in investment could be realized by running at
high velocities. More experimental data are needed in order to
completely resolve this question.
3. Effect of carbon particle size on adsorption capacity and rate.
It was found that at relatively short contact times (10 minutes)
there will be a reduction in allowable capacity (loading)
of 207o to 35% in going from 12 x 40 mesh to 8 x 30 mesh carbon
based on Lake Tahoe data. Although there are incomplete data to
support this reduction there should be less of an effect of
particle size on capacity at the longer contact times required
in commercial plants (up to 50 minutes).
4. Effect of number of regeneration cycles on adsorption capacity
and rate. Carbon's ability to remove COD to the required con-
centration is not affected by regeneration but the carbon's
capacity is decreased to 65% of its original capacity after
seven regeneration cycles based on Pomona pilot plant data in
a single contactor. However, the capacity still appears to be
decreasing after seven regeneration cycles.
-------
It is felt that the data are not fundamental enough in nature to
permit their use as basic design data for system configurations other
than the type in which they were taken.
However, based on these data, preparation of preliminary process
designs and evaluations for the various types of proposed contacting
systems should begin. Such studies will hopefully determine which scheme
has the maximum technical and economic potential, and will indicate which
experimental data - if any - are required to bring the selected scheme
to commercial reality.
In view of the deficiencies in the present data, there are several
gaps which must be filled before a precise economic optimization and com-
parison can be made of the several contacting systems. Recommendations
are made for the following experimental work:
1. A side-by-side experimental comparison of upflow and down-
flow columns should be made to determine if there is any
inherent inefficiency in either of these methods of con-
tacting.
2. The effect of high and low velocities (up to 10 GPM per
square foot and below 4 GPM per square foot) on rates of
COD adsorption must be determined at long contact times.
The effect of low velocities is essential if gravity flow
systems are to be evaluated.
3. The effect of carbon particle size on adsorption capa-
city and rate before and after regeneration should be
determined at contact times of commercial interest (up
to 50 minutes).
4. The effect of the number of regeneration cycles on adsorp-
tion capacity and rate should be investigated to deter-
mine if the decreasing trend will eventually level off.
The effect of velocity after regeneration should also be
studied.
5. The effect of particle size on regeneration loss should
be determined under closely controlled conditions.
-------
INTRODUCTION
The purpose of this report is to summarize findings on the first
portion of the work on tertiary waste water treatment with activated
carbon. Specifically, the work reported herein represents the results
of a thorough literature review of the data on tertiary waste water
treatment with a view to generating sufficient basic data for process
design of various schemes for carbon adsorption. In the literature
review, particular attention was given to the following critical
process parameters, the effects of which must be defined prior to any
design work.
1. Allowable loading of carbon with organic matter to purify
water to an acceptable level.
2. Effect of liquid linear velocity on adsorption capacity
and rate.
3. Effect of carbon particle size on adsorption capacity and
rate.
4. Effect of regeneration cycles on adsorption capacity and
rate.
As a minimum result, the data extracted from the various sources
should provide sufficient information to make empirical designs of
plants of the type that have already been demonstrated (e.g., at
Pomona or Lake Tahoe). Ideally, this survey would furnish basic design
data which could be applied to any type of contacting scheme such as
moving bed, upflow or downflow systems or fluidized bed systems, whether
or not such a system has been piloted.
-------
ALLOWABLE CAPACITY OF THE CARBON
A capacity, or loading, of 0.87 pound COD (total) per pound carbon
or 0.58 pound dissolved COD per pound carbon will be used in design for
fixed-bed downflow contractors operating on 60 mg/1 COD (total) influent
with feed water characteristics the same as at Pomona, based on results
from the Pomona pilot plant (Figure 1) under the following conditions:
1. 7 GPM/ft2 superficial velocity
2. 16 x 40 mesh, type CAL Pittsburgh Activated Carbon
3. Virgin carbon (unregenerated)
The effect of these three variables on the capacity will be dis-
cussed in subsequent sections.
This capacity (0.87 Ib COD/lb carbon) will be used in design even
though the design influent (60 mg/1) is higher than the average at
Pomona (47 mg/1). The higher influent concentration should increase
the loading if the displacement from exhaustion is great; however, this
will not be considered here since no quantitative estimate of its effect
can be made. Data needed for this type of correction would be operation
of a column for an extended period on water with a higher influent con-
centration.
Adsorption isotherms are of no value here because the apparent
capacity of the carbon in an operating column is 50% to 10070 greater
than the isotherm capacity. ^ This apparent increase in loading is
probably due to biological activity. This result could be expected
since the BOD is concentrated on the carbon and the rate of biological
action should increase accordingly. Use of this uncorrected capacity
implies that there will be a factor of safety but of undetermined mag-
nitude.
The other runs shown in the Pomona report (Figure 1) with higher
loadings (1.06 and 1.22 Ib COD/lb carbon) were not producing a
high quality effluent (see January-April 1966, Figure 2) and were dis-
carded. However, these higher loadings do indicate that 0.87 Ib COD/lb
carbon is far from exhaustion.
-------
The top line of Figure 1 is the contactor identification number.
The designation, IV 0, A, B, C, D indicates that the contactor (IV)
occupies the fourth position on the concrete slab, (0) that the carbon
contained in the contactor has never been regenerated, and (A, B, C, D)
that the contactor has occupied the first, second, third, and last posi-
tion in a total column of four contactors.
Figure 2 graphically portrays the concentration of COD removed by
each contactor within the four stage contactor column during the period
June, 1965, - August, 1966, utilizing the same contactor identification
number system as described for Figure 1.
-------
VOLUME TREATED (MG)
DAYS ON STREAM
DAYS IN POS'N. "A"
CARBON DOSAGE
(Ibs/MG)
WT. OF ORGANICS
REMOVED
COD Ibs/IOOIbs of carbon
TOC " " " " "
ABS " " " :: "
IOA
24
86
86
280
T
73
28
D
46
5.3
]IOA,B
59
211
125
220
T
122
32
D
66
16
5.1
3IEOA,B,C
81
288
77
250
T
106
27
D
57
18
4.9
EOAiB^D
100
365
77
270
T
87
24
D
58
17
2.8
T TOTAL
D DISSOLVED
FIGURE I
MAIN CARBON COLUMN PERFORMANCE
REFERENCE 2
REPRINTED BY PERMISSION, Parkhurst, J.D., «/*/., JWPCF, 39, part 2, R77(I967).
and County Sanitation Districts of LosAngeles County
-------
201965 28
37 41 49 55 63 73I96680 87
VOLUME TREATED-MILLIONS OF GALLONS
FIGURE 2
COD REMOVAL PATTERNS
REPRINTED BY PERMISSION, Parkhurst, J. D., at
-------
The run chosen represents the only one where the contactor has been
in all four positions in the total column before regeneration. This con-
tactor was onstream for one year and was subjected to seasonal variations
in influent concentration (Figure 2) which tend to average out this
effect. This capacity also agrees with the recommendation in the Pomona
report of 55-60 Ib dissolved COD/100 Ib carbon since dissolved COD con-
stitutes about 70% of the total COD from their secondary treatment plant.
A loading of 0.87 corresponds to a dosage of
8.33 x (60-7) = 507 Ib carbon/106 gal.
0.87
This dosage is considerably higher than the dosage found at Pomona
(250 lb/106 gal.), or at Lake Tahoe (300 lb/106 gal.)*, but dosage is a
concentration dependent term. The design removal (53 mg/1) is higher
due to a higher influent concentration than the average removal at
Pomona (37 mg/1) or at Lake Tahoe (10-35 mg/1). This higher removal
accounts for the increase in the calculated dosage.
For the sake of completeness, the average loading for Lake Tahoe
over 1-1/2 years onstream was about 0.25 Ib COD/lb carbon as determined
by planimeter measurements of their COD removal data. This relatively
low loading reflects the lower influent concentration at Lake Tahoe and
the fact that the carbon was sometimes taken out of the contactor before
breakthrough due to plugging problems. Because the secondary effluent at
Lake Tahoe was filtered to remove suspended matter prior to adsorption
and the Pomona secondary effluent was not prefiltered, the Pomona capa-
city is higher since removal by filtration was included in the capacity.
At Pomona, removal by filtration represents about 30% of the total
capacity. Also, the rates of biological action may be lower at Lake
Tahoe due to the colder climate.
It should be noted at this point that dosage is an indefinite term
which should not be used without a knowledge of the concentration and
removal of pollutant in the waste water. Loading (Ib pollutant removed/
Ib carbon) is a more precise, meaningful term which has included removal
in its calculation.
-8-
-------
EFFECT OF LINEAR VELOCITY
With few exceptions, the literature on tertiary treatment is in
agreement that linear velocity between 4 and 10 GPM/ft has no effect
upon the rate of adsorption. This is shown in several figures for
various adsorbates:
Chemical Oxygen Demand (COD): Figure 3
Alkyl Benzene Sulfonate (ABS): Figures 4 to 7
Total Organic Carbon (TOC): Figures 8 and 9
Color: Figure 10
When examining velocity data, care must be taken to separate the
effects of velocity, contact time, volume of water treated and length
of time onstream. When removal data for two different velocities are
taken on vessels of equal diameter and at the same contact time, the
comparison should be made at the same length of time onstream; that is,
the comparison is made after the same amount of water (and pollutant)
has been applied to the same amount of carbon (the pounds-COD applied
per pound of carbon is the same). This means that there will be a
different total amount of water treated since the total amount of car-
bon in each case is differento To be completely comparable, parallel
trains would be added to the low velocity experiment so that the same
amount of water would be treated in the same length of time by the
same amount of carbon as in the high velocity experiment. Of course,
the percent removal accomplished by the parallel trains at the low
velocity would not be any different from the removal accomplished by a
single train. Therefore, the criteria for comparing velocity data should
be the same amount of pollutant applied per pound of carbon at the same
contact time rather than constant volume of water treated. The only
case where data at constant total volume of water applied can be used
is when the same amount of carbon is used for both velocities. This
occurs when the cross sectional area for the low velocity case is lar-
ger by the ratio of velocities.
There are data which show that there is a contact time above which
no more adsorption is accomplished in the column. Figures 6 and 8 show
this effect at contact times greater than 15-20 minutes for virgin carbon
adsorbing ABS (alkyl benzene sulfonate) and TOC (total organic carbon),
respectively. The data in Figures 4 and 5 also fall in this range of
contact time. Below this critical contact time, although there is an
effect of contact time on removal, there is still no effect of velocity
as seen in Figures 6-8. Figure 9 shows that for a contact time of about
3.5 minutes, the percent TOC removed is constant at 81% for velocities
of 4, 7 and 10 GPM/ft2.
-9-
-------
9.13
0.10
4.1 OP1/FT* -^
CPU/FT2
0.05
EMPTY BED CONTACT TIME = 7 8 min
I I
10 .20
IBS COD APPUED/LB CARBON
.30
REFERENCE 5
FIGURE 3
EFFECT OF VELOCITY ON CAPACITY
OF 8 X 30 MESH ACTIVATED CARBON FOR COD
-10-
-------
AV'O. INFLUENT » 67*
AV'O. EFFLUENT =19.2*1
0 I 1 9 4 5 T K> II 12 II 14 IS
THROUGHPUT VOLUME , THOUSANDS OF GALLONS
REFERENCE 6
FIGURE 4
COD BREAKTHROUGH CURVE AT 4 GPM/FT2
REPRINTED BY PERMISSION, Joyce, R.S., et al.t JWPCF, 38, 816 (1966).
* ADDITIONS BY AUTHOR OF THIS REPORT.
-11-
-------
r-
K
«r
£«0
s»
1
H
r
«
«
t
T
V
\
mnuc
cm*
IT
a..
f
NT
h_.^~
0.
s
/
1
f
1
N*'
\
V
\
^<
4
\
V
i«k^-
INTLUC
^^
trriu
IT
fc~-o-
!»T
*t
V
?
c
8
>*'~~
A
\
\
c
o«
A
/ \
V
\ ,
w
*****-
3
-
-^
i
v
'v*^^
T*/
\
rip
u
^,
;
*«*"
/J
J
I
*:-
-
-.
»-
a
j-
J
k^«
L
f*^
ri
*i
***.
WG. INFLUENT = 57*
AV'G. EFFLUENT = 18*
TMWJOOMIMT VOLUME THOUtANOS Of «*LLONS
REFERENCE 7
FIGURE 5
COD BREAKTHROUGH CURVE AT 10 GPM/FT2
REPRINTED BY PERMISSION, Joyce, R.S., ft a/., JWPCF, 38t 815 (1966).
* ADDITIONS BY AUTHOR OF THIS REPORT.
-12-
-------
1.0
0.8
0.6
o
J
if
DD
'
0.4
0.2
MESH CARBON
9 8PH/FT2
Q = 11,000 OA
LONS
10 IS 20
RESIDENCE TIME, min
:
REFERENCE 8
FIGURE 6
EFFECT OF VELOCITY AND CONTACT TIME
ON
ABS REMOVAL EFFICIENCY
-13-
-------
1.0
0.0
1
w
«
<
»-
z
1
3
.
«
t
i
4
>-
X
M
s
= 0.4
I I I I
EXPERIMENTAL
CALCULATED
CONTACT TIME « 3.74 HIM.*
M M
TOTAL WEIGHT OF ABS IN INFLUENT.
. 100
REFERENCE 9
AFTER TREATING EQUAL VOLUME OF WATER*
* ADDITIONS BY AUTHOR OF THIS REPORT
FIGURE 7
EFFECT OF VELOCITY AND CONTACT TIME
ON
ABS REMOVAL EFFICIENCY
-14-
-------
PERCENT TOC REMOVAL
IWU
90
80
70
60
50
40
30
20
i . i
i
' Lr ft j ?i rt ^- * *
5§P i^~~/s"" * ^
fOgX--!
-Vo
f :
1
1 1
1
I !
1
s
X
-8-
i
i
i
r
i
i
i
FLOW RATE VOLUME TREATED
O 4 GPM/FT2 259 GAL
X 7 GPM/FT2 444 GAL
10 GPM/FT2 640 GAL
10
:
10 15 20 25 30 35 40
CONTACT TIME, min
REFERENCE 10
FIGURE 8
EFFECT OF VELOCITY AND CONTACT TIME
ON
TOC REMOVAL EFFICIENCY
-15-
-------
a
a
a
z
Ul
:
9
6
7
6
3
4
3
2
-
~_ 7
EE
~£
: ' . .
_ .
-:r
: : . :
::: :
::::
----
|
....
: :
~
::;:
----
. . ...
'::-. :
^r:
--
...
. . .
....
:;-::
E.
-
3.74 min
- . . -
!
.
n
--
-----
*
::::
--
xtd
....
....
/,
r \
RELAT
DEPTH
TOC F
DATA
FOR 9
- cc
pJTj
».2(
_
....
i
)NT
---
5 rr
/
IONSHI
OF C)
ROM l(
TAKEN
9 HOU
ACT
_.-
in ^
- y
/
-4:
1
Tl
. . - .
. . a ..
---
If- T
A
__j.:
ME
P -
H
1 >
g
^
-^
4: : :
...! :
3.3
1
V
E
_
5 rr
52
^
_. . t
fjh
lin >
V
_
.._
If- ^
._3
- --
- -
. .^
if
P BETWEEN VELOCITY AN[
XRBON REQUIRED TO REDU
3.5 mg/l TO 2 mg/l.
AFTER BEING ONSTREAM
RS.*
; -
\ ._. .
'
r
:T
^-1
;:::
:
CE
t*T-J-+-
,
tt
---
r~H
- -
1 .
-
~ ~ -
':::.
2 3
CARBON DEPTH, ft
FIGURE 9
EFFECT OF VELOCITY
ON
TOC ADSORPTION
REFERENCE II
ADDITIONS BY AUTHOR OF THIS REPORT
-16-
-------
Effect of External Diffusion
Pittsburgh CAL Carbon, Bogalusa Effluent
0.874 lbs./gpm/ft.
6 8
FLOW RATE, GPM/ft2
FIGURE 10
EFFECT OF VELOCITY ON COLOR REMOVAL
REPRINTED BY PERMISSION, NCSl TECHNICAL BULLETIN 199, 36 (1967).
-------
Other evidence showing an effect of velocity can be seen in the
following figures :
COD: Figures 4, 5, 12
ABS: Figure 13
In Figures 4 and 5 (COD removal), the fraction COD remaining (C/Co)
is about the same for the curves at 4 and 10 GPM/ft^, but the contact time
varied as follows:
Velocity, GPM/ft2 4 10
Volume treated, gal. 5,460 14,100
C/Co 0.298 0.316
Contact Time, Min. 37.4 14.9
When the C/Co data for COD removal shown in Figure 11 were recalculated at
equal onstream times and replotted against contact time in Figure 12, a
considerable effect of velocity was found. As is evident in Figure 12, the
high velocity (10 GPM/ft^) improves the removal of COD at the same contact
time. Thus, the higher velocity will reduce the diameter of vessel required
without appreciably increasing the depth of bed required. The advantage
of high velocity may decrease, however, at long contact time. Unfortunately,
the data represented in Figure 12 were the only data which could be found
in a form suitable for analysis showing the effect of velocity on efficiency
of removal for COD. Data should be obtained for adsorption of COD at high
velocity at contact times up to about 40 to 50 minutes, or removal effi-
ciencies up to 85 percent in order to complete the curve in Figure 12.
The data on ABS removal (Figure 13) are seen to be inconsistent with
the data of other investigators (Figures 4 to 7). Again the higher velocity
accomplishes a higher removal at the same contact time. The two curves do
appear to be converging at long contact times at which point there is no
effect of velocity, i.e., the same amount of carbon is required at both
velocities in order to get the desired removal.
It is generally agreed that there is a velocity below which diffusion
to the surface of the carbon may be controlling the rate of adsorption^;
however, above this velocity, the resistance to diffusion to the surface
of the solid adsorbent is reduced and the controlling rate step becomes
one of surface adsorption or diffusion within the pores of the particle.
In summary, few data on COD removal are available for the following condi-
tions :
o
1. Velocities below 4 GPM/ft where gravity flow systems would
operate.
2. High velocities at long contact time.
-18-
-------
1J>
.
-
.6
u
d
2
UJ
.1
a
a
u
z
c
.4.
.3
.1
;. SECONDARY
\
4 Gl
c/c<
PRO
. . .
in
^ m
\\
V
) = 0.
FILE A
EFFLUENT-PITTSBURGH SGL CARBON _
GPM/ft.2
Jo = 0.885 exp(-x/20.75)
)FILE AFTER TREATING 14,100 G-
\
\\
fi
^
885 e
FTER
s^
K \r
\4
N
«p(-x/l
TREAT
V
\
k \
\
7)
ING 14
v
\
\
,900 C
~"\
REF
!\|_.*
, . .
X
;AL.^
ERENC
I
E 13
10 IS 20 25 30
COLUMN LENGTH, ft. (x)
-
FIGURE II
EFFECT OF VELOCITY ON COD REMOVAL EFFICIENCY
^ADDITIONS BY AUTHOR OF THIS REPORT
-19-
-------
a
z
5
_
Ct
g
o
f
3
s
_
.9
8
, 7
.6-
.S
.4
.3-
.2
.1
\
v
\V
Y^
\!X
\\
\ -
i\
\
\
S, ;
\
V
X
V
X
^.
\ ^*
\
10 GPM/FJ2 AFTER
TREATING 14,100 GAL.
A POINT C
BY AUTH
ALCUIJ
OR OF
\TED B
THIS
t
-----
i
-
s^
""^^^
"^
Y PLAr
REPOR"
JIMETEI
r.
REFERENCE
4 GPN
TREAT
I/FT2
"ING 5,
«*-.
R
i -z
4FTER
460 G/
M_.
«.
-
20 30
CONTACT TIME, min.
-v
50
FIGURE 12
EFFECT OF VELOCITY AND CONTACT TIME
ON
COD REMOVAL EFFICIENCY
-20-
-------
2
UJ
O
cc
Ul
Q-
o:
O
o
O
x
O
8 !0 12
100
14 16 18 20 22 24 26
CONTACT TIME.min
28 30 32 34 36 38
AVERAGE REMOVAL OVER PERIOD TO POINT OF YIELD OF 1,470
GALLONS OF WATER PER POUND OF CARBON IN SYSTEM
REFERENCE 14
I I I I
8 IO 12
14 16 18 20 22 24 26 28 30 32 34 36 38
CONTACT TIME.min
AVERAGE REMOVAL OVER PERIOD TO POINT OF YIELD OF 3,510
GALLONS OF WATER PER POUND OF CARBON IN SYSTEM
AVERAGE-REMOVAL CONTACT TIME RELATIONSHIP,
FEASIBILITY STUDIES.
FIGURE 13
EFFECT OF VELOCITY AND CONTACT TIME
ON ABS REMOVAL EFFICIENCY
-------
EFFECT OF PARTICLE SIZE
Since, theoretically, carbon particle size primarily affects the rate
of adsorption and not the capacity of the carbon, adsorption columns opera-
ting close to saturation will be affected less by different particle sizes
than will columns operating far from saturation. There is a complicating
factor in some of these comparisons of particle size since the two sizes
of Pittsburgh carbon do not have exactly the same pore size distributions.
The effect of particle size is shown for ABS removal in data from Lake
Tahoe (Figure 14). The small difference between the two carbons at both
long and short contact times indicates that the carbons are approaching
equilibrium for ABS. This is particularly evident at the long contact time
since liquid concentration is approaching zero. Also at short contact time,
the carbon is approaching equilibrium with the feed.
But so long as the carbon is far from equilibrium, an effect of parti-
cle size should be expected. When the data for COD removal (Figure 15) are
plotted against contact time (Figure 16) in the same manner as Figure 14,
an increasing difference between the two carbons can be seen for increasing
contact time. This indicates that the carbon is still far from equilibrium
for COD adsorption. There are other data which show less effect of par-
ticle size on COD removal (Figure 17).
When data for color removal (Figure 18) are replotted against particle
size (Figure 19), it can be seen that since the lines of constant contact
time seem to be converging with decreasing particle size, there is probably
a particle size below which contact time will have no effect. However, this
particle size appears to be less than 60 mesh which is below the range of
commercial interest.
Since few data are available on the effect of particle size on COD
removal, no precise quantitative recommendation can be made. But
there probably will be a reduction on the order of 20 to 35% in allowable
capacity in going from the 12 x 40 mesh carbon to 8 x 30 mesh at these
relatively short contact times, based on the Lake Tahoe data (Figure 15).
No data are available showing the effect of particle size after regeneration.
-------
o
L
IT
1 .U
o.a
.6
.4
o.;
V
\ V
\ N
\ \
\
\
12X40 MESH
CARBON ' ~
8.5 6PM/FT2
i
\
\ 8X30
\ S 8-5
\
\ \
^N \
\ \
\v
N
x
Q -
MESH CARBON
BPM/FT2
\
^^
11.000 GALLONS
5 10 IS 20 25 30
RESIDENCE TIME, min.
REFERENCE 16
EFFECT OF CARBON PARTICLE SIZE ON ADSORPTION
FIGURE 14
EFFECT OF CARBON PARTICLE SIZE
ON
ABS REMOVAL EFFICIENCY
-23-
-------
12140
1130
DEPTH - 1.79
0.15
0.10
2 0.05
O.J 1.0
LBS COD APPLIED IB CARBON
0 .29 .90
LBS COD APPLIED/IB CARBON
0.19
0. 10
5 0.09
12140
SI30
DEPTH - 8.79'
0.19
0.10
- 0.05
0.1 0.2 0.3
LBS COD APPLIED LB CARBON
0 0.1 0.2
LBS COD APPLIEO/LB CARBON
REFERENCE 17
FIGURE 15
EFFECT OF CARBON PARTICLE SIZE ON COD ADSORPTION
FLOW RATE = 6.5 GPM/FT2"*
* ADDITION BY AUTHOR OF THIS REPORT
-24-
-------
1.0
0.8
UJ
tt
Q
O
c
0.6
o 0.4
a
L_
6
u
o 0.2
LAKE TAHOE DATA (REFERENCE 17'
8 X 30 MESH CARBON
12 X 40 MESH CARBON
-
; :
DATA TAKEN AT EQUAL VOLUMES
OF WATER TREATED
468
CONTACT TIME, min.
:
FIGURE 16
EFFECT OF CARBON PARTICLE SIZE
ON
COD REMOVAL EFFICIENCY
-25-
-------
?i
N
U
1C
5
1
§n
0
20
19
10
5
1
:
*.
^
y
^^-S
^
^SK
X**^
J^^
"0.
**z^l
^X
«^~
*
^
;
/
^
* ^
x x
Nx
L *S^'
tf''
t*
^f ^^jpS*
,A
-'' /
^--""
«
8 X 3O MESH AT 5 GPM/FT2 *
20 X 5O H
IESH AT 5 GPM/FT2 -
8 16 24 32 40 48
Passage of COD With Time, for Varloni Operating Condition*
REFERENCE 18
FIGURE 17
EFFECT OF CARBON PARTICLE SIZE
ON
COD ADSORPTION
REPRINTED BY PERMISSION, Dostal. K.A.. et al, JAWWA, §7, 67O (1965)
^ADDITIONS BY AUTHOR OF THIS REPORT
-26-
-------
500
400
3OO
Effect of Internal Diffusion
Pittsburgh CAL Carbon, Bogalusa Effluent
AT 6.8 GPM/ft2*
Curve A, Particle Size US 16 Mesh
B "
C
D
ii
20
30
40
3
O
200
100
BED DEPTH, ft.
FIGURE 18
EFFECT OF CARBON PARTICLE SIZE ON COLOR REMOVAL
^ADDITION BY AUTHOR OF THIS REPORT
REPRINTED BY PERMISSION, NCSI TECHNICAL BULLETIN 199, 37 (1967)
-------
I
'
I
DATA FROM NCSl BULLETIN 199 (REFERENCE 19)
U.S. SIEVE NO.
;;:! ::T 111
l t
1.4
2.5 3
6 7 8 9 IjO 1.5 2 2.5 3
PARTICLE DIAMETER, MM
5 6 7 8 9 10.
FIGURE 19
EFFECT OF CARBON PARTICLE SIZE
ON COLOR REMOVAL EFFICIENCY
-------
EFFECT OF REGENERATION
Care must be taken when analyzing data for regenerated carbon to
account for the effect of the make-up carbon which is usually added as
virgin carbon. For example, when the regeneration losses are 10%, there
will be little of the original carbon present after ten regenerations;
that is, the original carbon will have been replaced by virgin make-up
carbon. The assumption here is that the carbon which has been regenerated
will be preferentially lost in subsequent regenerations, especially as
the particles become smaller and weaker.
Losses in regeneration can occur by attrition in the pipeline and
conveyors, by attrition, decrepitation, gasification and burning in the
regeneration furnace, or by thermal shock in the quench tank. Variables
which may affect the regeneration loss are particle size and shape, type
and degree of loading, amount of handling, regeneration severity, and
carbon base material (such as bituminous coal, wood, coconut shell,
petroleum coke, etc.) There is a need to pin-point where carbon losses
occur and, perhaps, to develop a rounder carbon particle which would be
more attrition-resistant. Another major loss of carbon could be by gasi-
fication which could be as high as 50% of the carbon per hour.20 Regenera-
tion losses should be higher for the smaller size carbon since the smaller
particle is more active for adsorption and, therefore, should be more
active for burning and gasification. The comparison between adsorption
and burning can be made as long as the regeneration temperature is low
enough so that burning is controlled by pore diffusion and not by diffu-
sion to the external surface. Under this condition, burning is a pore
diffusion phenomenon as is adsorption. If the burning rate at this tem-
perature were entirely an external surface phenomenon, then activated car-
bon would have the same reactivity as coke which is clearly not the case.21
The difference in particle size may account for part of the difference
between the 7% to 10% loss experienced at Pomona with 16 x 40 mesh carbon
and the 5% loss at Lake Tahoe with 8 x 30 mesh carbon. Also, the opera-
tion of the regeneration furnace on a continuous vs. batch basis will
affect the carbon loss. Carbon loss should be higher in batch operation
than in continuous operation because the carbon left in the furnace between
regenerations can be burned during subsequent shutdown and startup. Lake
Tahoe has minimized this problem by placing sand on the hearths of the
furnace to prevent the residence of carbon under the rabble arms.
Results of recent work done by the MSA Research Corporation 22 for
the FWPCA show that the conditions used in the multiple-hearth regenera-
tion furnace are far from the optimum with respect to regeneration gas
flow rate, carbon residence time in the furnace, and carbon temperature.
It was found that a regeneration loss of 27» could be obtained in a 3.25 in.
I.D. x 65 in. long rotary tube regenerator with a lower gas flow rate,
longer carbon residence time and higher carbon temperature. However, it
has not been established whether or not all the carbon loss at Pomona and
Lake Tahoe occurs in the regeneration furnace.
-29-
-------
After each of three regenerations in the accelerated contactor at
Pomona, the carbon has come back to 90% dissolved COD removal efficiency
(Figure 20), despite a continuously decreasing iodine number. Further-
more, this decrease seems to occur during the first regeneration and not
during any subsequent regeneration (Figure 21). However, the loadings
for dissolved COD at breakthrough have dropped from 0.46 Ib/lb to 0.29 lb/
Ib after seven regenerations (Figure 22). It should be noted here that
the seven regenerations on the accelerated contactor represent about seven
years operation in the main carbon column at Pomona. The capacity shown
for seven regenerations should not be extrapolated since it still appears
to be decreasing (Figure 23). Furthermore, extrapolation may be unwar-
ranted, since an economic evaluation may reveal that the carbon should be
replaced after several years use rather than accept a decrease in capacity.
The concentration profile for COD for a bench-scale column operating
at 1.84 GPM/ft2 is the same for virgin and sixth cycle regenerated carbon
(Figure 24). The regeneration losses at Pomona have averaged about 10%.
Lake Tahoe shows an apparent 15% decrease in ABS removal after one
regeneration in a test made with pure regenerated carbon (Figure 25).
Lake Tahoe is apparently taking no economic debit for decreasing capacity
with regeneration since they have seen no increase in required dosage in
1-1/2 years' operation.^ The regeneration losses at Lake Tahoe average 5%.
The Pittsburgh Activated Carbon Company has found a 35% decrease in
equilibrium capacity for ABS after 16 regenerations (Figure 26). Adsorp-
tion isotherms for COD on type SGL virgin and carbon regenerated 16 cycles
do not appear to be significantly different (Figure 27). The authors
later show, however, that the adsorption isotherm will be of no value in
design since the apparent capacity of the carbon in an operating column is
50 to 100% greater than the isotherm capacity. The authors assert that
this increased capacity may be due to biological activity. Their regenera-
tion losses averaged 4.2%. Another paper by the same authors shows the
removal efficiency for COD after 16 regenerations (Figure 28). Here the
difference between virgin and regenerated carbon does not appear until
the carbon has treated a large volume of water (dosage 100 lb/10" gal.).
This finding is consistent with the results of the bench-scale test with
sixth cycle carbon at Pomona shown in Figure 24, indicating that one of
two things may be occurring during regenerations
1. There may be some degradation of the pore structure which
results in slower pore diffusion which becomes the rate
controlling step as the carbon becomes more heavily loaded.
-30-
-------
1200
§ 200
,*£ioo
QUJ
UJ
X
o>
3uj 0
K
20
40
N
V
60
N
V
80
100
400
DISSOLVED COD
"0 20 40 60 80 100
VOLUME OF WATER TREATED, million gallons
REFERENCE 23
FIGURE 20
PERFORMANCE OF ACCELERATED
CONTACTOR No. I
REPRrNTED BY PERMISSION, Parkhurst, J. D., et al., JWPCF, 39, Part 2, R76 (1967)
and County Sanitation Districts of Los Angeles County
-31-
-------
100
1
'
I
lull. ;;U.iiil.iiitUi;i.,::!;:::
5 REGENERATIONS
I REGENERATION
6 8 10 12 14
WATER VOLUME TREATED, million gallons
16
IR
20
REFERENCE 24
FIGURE 2 I
EFFECT OF REGENERATION ON COD REMOVAL EFFICIENCY
^ADDITIONS BY AUTHOR OF THIS REPORT
-------
400
360
320
z
O
m
« 280
o
o 24°
o
g 200
o
£ 160
O
120
o
80
40
n
"T
-
- -
i
/
i
_. _L
-
-
~f : " T
i
r"
I
1
|
" i
i
j
i
! -
I
j
i
i
i
__
jy
i
X
10
/
/ a
Y
\
A -46
1
1
1
.
/ p
IlA
Ib/lb -
T
l
~
*
.
1
-4
t
i
O^
o
c
:2A
3 1
1
j
1
)-
i
/
X
^^
^r
^^
I3/
-37 It
j/lb
i
/
V
/ i
#
^-3
»/lb
14
v
/
:i
if : : .
|X
S;
-y
'/.
^
/
x
X
xe
I5A-37
]. ^
a- 32 Ib/lb
^
7 lb/l
K
1 .
:
::::
: ;::
^
»A-
lb/1
17
32
b.
: \~.'.
\-\
;-; -
>9
Ib/lb
: : ::
lb/
Ib
\_
40 80 120 160
CUMULATIVE VOLUME TREATED, minion gallons
240
REFERENCE 25
FIGURE 22
EFFECT OF REGENERATION ON CAPACITY FOR COD
-33-
-------
POMONA DATA (REFERENCE 25)
23456
NUMBER OF REGENERATION CYCLES
FIGURE 23
EFFECT OF REGENERATION
ON
CAPACITY FOR COD
-34-
-------
. I.
I
1 1
'
'
4" DIAMETER CARBON COLUMNS
.84 GPM/ft2
41 min CONTACT TIME FOR 10 ft
24 min CONTACT TIME FOR 6 ft
INFLUENT
TOTAL COD
INFLUENT
DISSOLVED COD
DISSOLVED COD
REGEN. I6A - 6 ft
VIRGIN CARBON-6 ft
DISSOLVED COD
REGEN. I6A- 10 ft
VIRGIN CARBON-10 ft
4 8 12 16 20 24 28
VOLUME OF SECONDARY EFFLUENT TREATED, thousand gallons
FIGURE 24
EFFECT OF REGENERATION ON COD ADSORPTION
-------
0.10
0.08
Q
Ul
> 0.06
5
Ul
tr
CD
0.02
VIRGIN CARBON
REBENERATEO CAR
BON
0.10 LB.ABS
APPUEO V
o.os
IB ABS
APPLIED
;z
CONTACT TIME, min
COMPARISON OF VIRGIN AND FIRST CYCLE REGENERATED CARBON
FOR ADSORPTION OF ABS AS FUNCTION OF CONTACT TIME
REFERENCE 27
FIGURE 25
EFFECT OF REGENERATION ON ABS ADSORPTION
-36-
-------
I
'
1
I
z
o
en
it
i
' '
5 O.I
CD
O.OI
0.01
SECONDARY EFFLUENT-PITTSBURGH TYPE SGL CARBON
CARBON,REACTIVATED 16 CYCLES
-EC
DAT 10
.0354 C
O.I 1.0
RESIDUAL ABS CONCENTRATION, ppm
10.0
REFERENCE 28
FIGURE 26
EFFECT OF REGENERATION ON ABS ISOTHERM
-------
I
00
Z
o
(0
(K
i
>
5 at
v.
a
8
0.01
CARBON Rl
j-
^x"^
iACTIV
\
.x"*1
x"1^
i
z
ftTED
V
>
^
16
I
t
£
CYC
x^
^
r
"**
Lfc
X»
/
s
^
^
Q/^
^
-VIRGIN CA
SEC
z
U
LBOIN
ONDAI
VIRGIN SGL
X[
=1
C
0
o
RY
1
FFL
UE
Nl
P
1
TSBURGH
TYPE
SGL
CAI
=?B
')N
1.0
iao ioao
RESIDUAL COD CONCENTRATION, ppm, C
FIGURE 27
REFERENCE 29
EFFECT OF REGENERATION ON COD ISOTHERM
-------
C
1
c --
D.I
in 04
CARBON REACTIVATED 16
0.1
D
0
u
fcl
CARBON REACTIVATED 16
10 IJ
VOLUME, THOUSANDS OF GALLONS
FLOW RATE = 2.5 GPM/FT2
REFERENCE 30
FIGURE 28
EFFECT OF REGENERATION
ON
ABS AND COD REMOVAL EFFICIENCY
REPRINTED BY PERMISSION. Joyce, R.S., et at., JWPCF. 38, 818 (1966).
* ADDITION BY AUTHOR OF THIS REPORT
-39-
-------
2. The carbon is not being completely regenerated so that
there is still adsorbate left on the carbon after regen-
eration which, in effect, reduces the capacity of the
carbon in subsequent adsorption.
Both degradation and incomplete regeneration have the same effect on
efficiency of removal; that is, no effect until the carbon becomes
heavily loaded.
In summary, it appears that the carbon's ability to remove to the
desired concentration of COD is not affected by regeneration but that
the capacity is decreased as shown in Figure 23. Evidently, the regen-
eration process affects the equilibrium characteristics of the carbon
but does not affect the rate processes.
-40-
-------
DISCUSSION
Presented in the previous sections are data relating to the capacity
and rate of carbon adsorption in a fixed bed, downflow column and the
effects of liquid velocity, particle size and regeneration on the rate
and capacity. These data have all been extracted from the literature on
tertiary treatment - no new experimental work has been performed. However,
a good deal of the data treatment described herein represents independent
correlations made in an effort to provide fundamental engineering bases
for subsequent process designs. In addition, since the basic data have
been obtained by a rather large number of investigators, a considerable
amount of effort was expended to bring as much of the data as possible
together on consistent bases.
With these data and correlations available, it appears possible to
design with a good degree of precision for the case of a fixed bed, down-
flow column capable of operating over a fairly wide range of conditions.
However, it is felt that the data are not fundamental enough in nature
to permit their use as basic design data for system configurations other
than the type in which they were taken. This point is in general agreement
with others engaged in studies on this subject.1'31 This is not the fault
of the investigators but is a characteristic of the system under study. For
example, waste water contains many components all of which, undoubtedly,
adsorb at different rates resulting in a breakthrough curve which is not
clearly defined (Figure 29). Also, waste water is subject to day-to-day
fluctuations in concentration which further confound the results. In
addition, capacities greater than the isotherm capacity plus the fact
that the carbon could not be completely saturated with COD hint that some
mechanism other than adsorption, such as biological activity, is involved.1
The operation under study is, then, a multicomponent adsorption with com-
plicating factors not related to adsorption. From these data, it is
impossible to apply standard adsorption analysis and extract any funda-
mental mass transfer data which can be used in the precise design of various
kinds of contacting systems. The entire process of waste adsorption
probably will never lend itself to a classical chemical engineering treat-
ment which will yield fundamental data for any of the process parameters
presently under investigation. Before any such analysis can be made, a
more basic understanding of the processes involved, such as biological
activity, must be developed. Biological action should be an important
factor in the removal of waste using carbon, since the waste is concen-
trated on the carbon and the rate of biological action should increase
accordingly. These rates must be significant if the 50 to 100% excess
capacity of the carbon above isotherm capacity are to be explained in this
way. Also, since the rate of biological action is time- and temperature-
dependent in a much different way than is adsorption, this action must be
well understood before the entire process can be evaluated. A basic
-41-
-------
CUMULATIVE VOLUME TREATED,
million gallons
15.20
10 20 30
CARBON DEPTH, ft
REFERENCE 32
FIGURE 29
11.80
9.35
6.05
5.33
2.89
COD BREAKTHROUGH CURVES
-42-
-------
understanding of the biological processes in the carbon bed would, hope-
fully, reveal ways in which the contribution of such processes to the
carbon's capacity might be enhanced.
Lacking such basic information, the design engineer is left with
only one alternative, that is, using data obtained under controlled
conditions from a side-by-side comparison of the different types of
contactors. Lake Tahoe has attempted such a comparison of upflow and
downflow, fixed-bed contactors33 but their results were clouded by
plugging problems and by running the downflow column at only 2 GPM/ft2
and the upflow column at 6 GPM/ft2.
In summary, the data presented herein are sufficient to permit the
design of the downflow, fixed-bed contactor but are not sufficient to
permit the design of any other type of system with a comparable degree
of precision. However, the data are adequate for use in preliminary
screening designs of other system configurations (e.g. fludized bed).
Such designs may well prove to be useful in determining what potential
benefits might be realized by one system over another and would also
serve to aid in planning future experimental work. Since there are now
no consistent comparative economic (or technical) evaluations for the
various proposed schemes - even preliminary ones - such a study would be
beneficial.
-43-
-------
RECOMMENDATIONS
Using data presently available, preparation of preliminary process
designs and evaluations for the various types of proposed contacting
systems should begin. Such studies will hopefully determine which scheme
has the maximum technical and economic potential as well as indicating
what experimental data - if any - are required to bring it to commercial
reality.
However, in view of the deficiencies in the present data, there are
several gaps which must be filled before a precise economic optimization
and comparison can be made of the several contacting systems.
1. A side-by-side experimental comparison of upflow and down-
flow columns should be made in extended runs with regen-
eration. The scale of the test should be large enough to
determine the effect, if any, of backmixing of the carbon
during adsorption and removal in the partially-expanded ,
upflow bed. A study of the effect of contact time (bed
depth) on effluent purity should show the difference in
efficiency between these columns. Studies of carbon losses
by attrition must also be made.
2. The effect of velocity below 4 GPM/ft^ must be determined
if gravity flow systems are to evaluated. Also, the effect
of high velocities on removal efficiency of COD must be
obtained, in order to determine if the advantage of high
velocity shown in Figure 12 still exists at long contact
times. If this is true, reductions in investment could be
realized by running at high velocities.
3. The effect of particle size on adsorption capacity and rate
at long contact times will determine if the trend shown in
Figure 16 will continue or if the difference between large
and small carbons will diminish as is expected (Figure 14).
The effect of particle size after regeneration should also
be studied.
4. The effect of number of regeneration cycles on capacity and
rate of adsorption will show if the trend shown in Figure 23
will eventually level off as expected. These data will be
-44-
-------
useful in optimization of the plant with regard to carbon
investment and length of its use. The effect of velocity
after regeneration should also be studied.
5. The effect of particle size on regeneration loss should
be determined in extended continuous runs under the same
conditions of regeneration. This effect has a direct
effect on optimization relative to particle size.
-45-
-------
REFERENCES
1. Allen, J. B., Clapham, T. M., Joyce, R. S., and Sukenik, V. A.,
"Use of Granular Regenerable Carbon for Treatment of Secondary
Effluent - Engineering Design and Economic Evaluation," Unpublished
report from Pittsburgh Activated Carbon Company to USPHS, October 1,
1964, p. 13.
2. Parkhurst, J. D., Dryden, F. D., McDermott, G. N., and English, J.,
"Pomona Activated Carbon Pilot Plant," JWPCF, 39 (10) Part 2, R 77
(October 1967).
3. Ibid., p. R 78.
4. Smith, C. E., and Chapman, R. L., "Recovery of Coagulant, Nitrogen Removal,
and Carbon Regeneration in Waste Water Reclamation," Report from South
Tahoe Public Utility District to FWPCA,
-------
15. National Council for Stream Improvement, op. cit., p. 34.
16. Smith, op. cit., p. 67.
17. Ibid., p. 65.
18. Dostal, K. A., Pierson, R. C., Hager, D_ G., and Robeck, G. G.,
"Carbon Bed Design Criteria Study at Nitro, W. Va.," JAWWA, 57
(5) 670 (May 1965). ~
19. National Council for Stream Improvement, op. cit., p. 37.
20. Lowry, H. H., "Chemistry of Coal Utilization," Supplementary
Volume, Wiley, New York (1963) p. 934.
21. Spiers, H. M., "Technical Data on Fuel," Fourth Ed., The British
National Committee World Power Conference, London (1947) pp. 310-311.
22. Juhola, A. J. and Tepper, F., "Laboratory Investigation of the
Regeneration of Spent Granular Activated Carbon," Unpublished report
from MSA Research Corporation to FWPCA, (July 1968) p. 7,
23. Parkhurst, op. cit., p. R 76.
24. Memorandum from A. N. Masse to F. M. Middleton, "Regeneration of
Granular Carbon at Pomona," March 28, 1967, p. 2.
25. English, J. N., Pomona Monthly Report to FWPCA, July 1967, Figure 4.
26. Ibid., Figure 9.
27. Smith, op. cit., p. 75.
28. Allen, op. cit., p. 13.
29. Ibid., p. 29.
30. Joyce, R. S., Allen, J. B., and Sukenik, V. A., "Treatment of Municipal
Wastewater by PaOked Activated Carbon Beds," JWPCF, J38 (5) 818 (May 1966),
31. McDermott, op. cit., p. 4.
32. English, J. N., "Carbon Adsorption Project, Report of First Year
Activities, 1964 to 1965," Unpublished Report from Los Angeles
County Sanitation Districts to USPHS, Figure 10.
33. Smith, op. cit., p. 94.
-47-
-------
PHASE II. ECONOMIC EFFECT OF DESIGN VARIABLES
by
A. E. Cover, The M. W. Kellogg Company
SUMMARY
This report summarizes the findings of work on tertiary waste water
treatment using granular activated carbon. Specifically this study found
the economic importance of several design variables and showed where
additional experimental data are needed. The variables which were studied
for their effect on the economics were shop fabrication and field erection
of vessels, surge designs, plant size, idle carbon inventory, velocity,
contact time, particle size, regeneration loss, carbon capacity, downflow,
upflow, gravity-flow and moving-bed contactors, adsorbent cost, number
of contacting stages, in-place regeneration and certain combinations of
the above variables. The comparisons made in this report are based on data
which were sometimes rather fragmentary as found in Phase I of this work.l
In this respect, the comparisons had to be limited by the inability to
get complete predictions of the effect of the variables on the process
performance. The assumptions which were made and the recommendations for
further work are listed in the body of the report.
Based on comparison of investment and operating costs, several con-
clusions were drawn. Among the more important conclusions are the
following:
1. The gravity-flow system is less expensive than the
pressurized downflow system by two cents per thousand
gallons which, in turn, is less expensive than the
pressurized upflow system with pre-filtration by one
cent per thousand gallons. Assumptions made in the
case of the gravity-flow contactor are that the re-
quired removal will be achieved at 2 GPM/ft^ with
50-minute contact time and that the loading achieved
in a single-stage system is the same as that used in the
design of the two-stage system.
2. A two-stage contactor is an optimum system as compared with
single-, three-, and four-stage contactors. Large-scale
operating data are required to find the allowable carbon
adsorption capacity in a two-stage system.
3. Adsorbent cost and regeneration loss have a significant
effect on total operating cost. Operating costs would be
reduced by 1.15 cents per thousand gallons if the adsorbent
costs could be cut by 15 cents per pound. If the re-
generation loss is reduced from 10% to 2%, the operating
costs would be reduced by two cents per thousand gallons.
Efforts should be made to find a lower cost adsorbent, to
reduce carbon inventory, to pin-point where regeneration
loss occurs and to understand the regeneration mechanism,,
-48-
-------
SUMMARY (Cont'd.)
For a pressurized vessel system, large particle size and
high velocity offer savings in costs. Savings in operating
cost are possible when 8 x 30 mesh carbon is used rather
than 12 x 40 mesh carbon. Also, operating costs can be
reduced by 1.7 cents per thousand gallons when the liquid
velocity is increased from 4 GPM/ft 2 to 7 GPM/ft 2 Data
are required to confirm the effect of velocity on required
contact time. Also, it is necessary to find the relation-
ship between particle size and regeneration loss.
A large economic incentive exists to try to maximize the
carbon loading (capacity). Operating costs can be cut
by two cents per thousand gallons when the carbon loading
is increased from 0.25 to 0.87 pounds Chemical Oxygen
Demand per pound of carbon. This might include enhancement
of the biological action by such means as injecting air
(oxygen) into the waste water.
-49-
-------
INTRODUCTION
The purpose of this report is to summarize the findings of
the second phase of work on tertiary waste water treatment using
granular activated carbon. The first phase was a review of the
literature with a view towards generating design data.1 The
work reported herein represents the results of preliminary eco-
nomic evaluations of several contacting schemes with a view
toward finding which variables have economic significance and
showing where additional experimental data are required.
Design variables which were studied for their effect on
the economics are listed below.
1. Shop fabrication and field erection of vessels
2. Surge design
3. Plant size
4. Idle carbon inventory
5. Velocity
6. Contact time
7. Particle size
8. Regeneration loss
9. Carbon capacity
10. Downflow, upflow, gravity-flow and moving-bed contactors
11. Adsorbent cost
12. Number of contacting stages
13. In-place regeneration
Table 1 lists all of the cases and design conditions considered.
In the absence of correlations between design variables and process
performance, various assumptions are made for these correlations to
determine the effect on the economics. In this way, each variable is
changed one at a time and, if it is found to be economically significant,
a recommendation is made to confirm the assumption. A summary table of
investment and operating costs has purposely not been provided in order
to preclude the making of invalid comparisons. Economic comparisons
between cases considered in this study and other cases should not be made
unless care is taken to make sure that the cases in question are on
-50-
-------
TABLE 1
Contactor Plant Size
Case Type MOD
1 Down 10
1A Down 10
2 Down 1
3 Down 100
3A Gravity 100
4 Down 10
4A Down 10
5 Down 10
5A Down 10
6 Down 10
7 Down 10
8 Down 10
9 Down 10
10 Down 10
10A Down 10
11 Down 10
12 Down 10
12A Down 10
13 Down 10
14 Down 10
15 Up 10
16 Gravity 10
17 Down 10
18 Moving-bed 10
19 Moving-bed 10
20 Down 10
20A Down 10
20B Down 10
21 Down 10(15)
22 Down 10
Velocity
GPM/ft2
7
7
7
7
2
10
10
10
10
4
4
7
7
7
7
7
7
7
7
7
7
2
7
7
10
7
7
7
7
7
SUMMARY
OF CASES
Contact Time Particle Size
Min. Mesh
50
50
50
50
50
50
50
35
35
50
87
50
50
40
40
40
50
50
50
50
50
50
50
15
15
50
50
50
50
50
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
12 x 40
12 x 40
12 x 40
12 x 40
12 x 40
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
8 x 30
Carbon Cap.
Ib COD/Ib C Variable Toted
5 0. 522 Base Case-Shop Fabrication
5 0. 522 Bate Case-Field Erection
5 0. 522 Plant Sice
5 0. 522 Plant Sice
5 0.522 Contactor Type
5 0.522 AP fe L/D
5 0. 522 Idle Carbon
5 0.522 AP, L/D t> C. T.
2 0. 522 Improved Downflow
5 0.522 A P.«. L/D
5 0.522 AP, L/D fc C.T.
5 0.522 A P fc Carbon Cost
10 0.522 A P tt Re gen. Loss
5 0.522 C.T. fc Carbon Cost
5 0.522 Idle Carbon
10 0.522 Regen. Loss, C. T. , Cost* tp
2 0. 522 Regen. Loss
10 0.522 Regen. Loo
5 0. 87 Capacity
5 0.25 Capacity
5 0. 348 Contactor Type
5 0. 522 Contactor Type
5 0. 522 Adsorbent Cost
5 0.25 Lake Tahoe System
2 0.348 Improved Moving-Bed
5 - 4-Stage System - Pomona System
5 - 3-Stage System
5 - 1 -Stage System
5 0. 522 Surge Design
0 0. 522 In-Place Regen.
-------
INTRODUCTION (Cont'd.)
consistent bases. The comparisons which are valid have been made in
this study and any other comparison between cases shown should not
be made. The accuracy of the investment and operating costs will be
improved by the more definitive process design and cost estimate
planned for the third and last phase of the contract. Nevertheless,
the current accuracy is judged to be sufficient to show trends and
to narrow down the choice of process conditions to be used in the
definitive design of the next phase.
The estimated cost for the carbon treatment plants is based on the
assumptions that the plants will be designed, procured, and constructed
by one general contractor; that the geographical location is the Gulf
Coast area for the purpose of establishing environmental conditions,
labor costs and freight shipment rates; and that the prices of materials
equipment and labor are as of November 1968.
Also, when examining the operating costs, it should be remembered that
for a 10 million gallon per day (10-MGD) plant, each one cent per
thousand gallons (1 cent/M gal) is equivalent to $36,500/year in
operating costs. Approximately $230,000 in plant investment can be
spent in order to make a savings in operating costs of 1 cent/M gal.
Fixed charges are 15.6% per year of the plant investment plus 8.4%
per year of the carbon investment.
It should also be noted that an optimum carbon adsorption process
does not necessarily mean an optimum total waste treatment process.
There may be combinations or interactions between sections of the
total waste treatment process which would serve to lower the entire
treatment costs. Interactions might be found between the secondary
and carbon treatment plants where the operation of the secondary
plant would be altered to improve the performance of the carbon plant,
thereby reducing the total treatment cost.
References in this report to figures, tables and conclusions in the
literature report refer to the Phase I Report which precedes this.
-52-
-------
GENERAL PROCESS DESCRIPTION
Following is a process description of the base case (Case 1A)
flow sheet (Figure 1 ) which is typical for a 10 million gallon
per day (MED) plant. The flow sheet for individual cases will not
vary substantially from this case. These variations will be dis-
cussed with each case.
Effluent from the secondary treatment plant (activated sludge)
enters the tertiary treatment plant into a 1.6 million gallon surge
basin and at a flow rate varying between 3,470 and 10,410 GPM. Suction
for . the feed pump, J-l, is taken from the surge basin at a rate of
6940 GPM. This feed water contains 60 mg/L total Chemical Oxygen
Demand (COD), 707= of which is dissolved COD and 30% suspended COD.
Water from the feed pump then flows through two adsorbers, D-la and
b, in series, where 50-minute residence time of the water in the
packed carbon beds is provided to produce a product water with 9.5
mg/L total COD and 7 mg/L dissolved COD. This remaining COD is
principally dissolved organic material which cannot be readily
adsorbed on activated carbon.
Periodically, a carbon bed stops producing water of the
desired purity and the carbon in the lead contactor must be re-
generated. At this time, the contactor (D-le on the flow sheet)
is taken offstream, a contactor in the number two position is
switched to the lead position, a contactor containing regenerated
carbon is brought onstream in the number two position and the carbon
is transported to a regeneration furnace at the rate of 350 pounds
per hour on a continuous basis. To accomplish this transfer the
carbon is educted with water through spent carbon eductor, L-2e, from
the bottom of the adsorber to the furnace area. The carbon is
separated from the water in a dewatering screw conveyor, V-l, and
is sent to the regeneration system, L-l, where the carbon is fed to
the top of a multiple-hearth, Herreshoff-type regeneration furnace
where hot flue gas and steam are used to regenerate the carbon. The
overflow water from the screw conveyor is returned to the settlers
in the primary treatment plant. The vent gas from the furnace is
sent to a cyclone, L-3, and afterburner, L-4, for air pollution con-
trol. The hot, regenerated carbon is quenched with water in the quench
tank and educted as a slurry back to an adsorber, D-lf, which has been
partially filled with water to prevent attrition of the carbon as it
falls into the adsorber. The excess water and carbon fines overflow
from the top of the adsorber and are sent to the primary settlers.
Since carbon is lost in the regeneration process, make-up carbon from
the fresh carbon storage bin, P-l, is added to the carbon-water slurry
in the quench tank at the rate of 420 pounds per day.
One additional operation which must be performed is that of back-
washing the lead adsorbers. Whenever the pressure drop in the system
becomes high due to the build-up of suspended solids on the top layer
of the carbon, the adsorber (D-lc on the flow sheet) is taken offstream
and backwashed with secondary effluent from back wash pump, J-2. The
backwash flow rate is set at 17 GPM/ft2 so that there is a 307, bed ex-
pansion during backwashing.
-53-
-------
1
Ul
1
D
11KK
1 1 1
SURGE BASIN Jo'-WcM (Tot.l) F'««
1,600,000 OAL. »2 mg/l COD (Dliiolvtd)
SECONDARY EFFLUENT
3,470 TO 10, mO 0PM
0
CAR
WA
77^
O-li to f
ADSORBERS
251-«" I.D. x
5SELS CONTAIN 300.
ACTIVATED CARtCH
'
(6)
I0'-0'
100 LB. 6x30 MESH
1 EACH
c c a
1 '.'...} FFFD 'PUMP ; 1*
1 J
c : c
60P8W
D-li 1Mb
M
BACKWASH WATER 1 1
MO 0PM C" " r~l-|
L-2d jc | L-2, jo
|
SPENT CARBON
SLURRY
V-
DEWATERING
SCREW
P CONVEYOR OVERFLOW TO
s!j.-E5AM SETTLERS.
/HO GENERATOR
L. . 0.7 SPM
1 1. W-flg
c
[ 1
L-2f 4c
L-2a to f (6)
SPENT CARBON
EDUCTORS
:RESH CARBON
STORAGE TANK
A4R
f it MARt-UK CARBON 171 SCFD
TJ 9UENCH TANK
EOUCTOR
FINES TO PRIMARY
SETTLERS
PRODUCT WATER
9.5 mg/l COD (Total)
7 mg/l COD(Oisiolved)
DESIGN CONDITIONS
1. TWO-STAGE, DOWNFLOW CONTACTOR
2. TWO TRAINS OF VESSELS
3. 7GPM/FT' SUPERFICIAL VELOCITY
4. 8X30 MESH CARBON
5. 5% REGENERATION LOSS
6. 50-MINUTE CONTACT TIME
7. 0.522 LB.COD/LB. CARBON
THE M. W. KELLOGG COMPANY
taooO.OOO GI9L£O*/S *f* DAY
-1"" 1"" CtlXf 1*
"*" M «l » r*oer*9 «o» 0>.*.=«.#Ar
6044 P3137D
' -M| a | r | I | o 1 c | o~.~o I *
FIGURE I
-------
DESIGN BASES
The following items are the design bases used for the base case
(Cases 1 and 1A). The bases vary somewhat from case to case in each
comparison; the difference in bases from the base case will be
discussed.
Water Quality - The Chemical Oxygen Demand (COD) concentration
to the activated carbon adsorption process is assumed to be 60 mg/L
with other water properties the same as that found at Pomona, Calif-
ornia. Suspended solids comprise about 30% of the total COD with
the remainder being dissolved COD.
Plant Size - A plant capable of treating 10 million gallons
per day (10-MGD) of waste water was chosen because 10-MGD represents
the sewage rate from cities of about 100,000 to 300,000 population.
Surge Capacity - A surge basin will be provided to damp-out
variations in feed rate. The flow rate is assumed to vary sinusoidally
between 50% and 1507=, of the average flow rate with a period of one day.
Vessel System - A two-stage, packed-bed, downflow system was
selected for the base case. The packed-bed, downflow adsorber was
chosen because most of the pilot plant data were taken on this type
of contactor. A two-stage contactor rather than one-, three-, or
four-stages was chosen because it was felt that his system should
yield a lower investment. Reductions are expected in vessel and
piping costs which should offset operating costs associated with a
lower carbon loading for the two-stage system as compared to the four-
stage system. Depths of pressure vessels are 30% greater than the carbon
bed depths to provide for bed expansion during backwashing.
Carbon Particle Size - The 8 x 30 mesh carbon rather than the 16 x
40 mesh used at Pomona was chosen for the base case because this larger
size may reduce the carbon loss during regeneration while not greatly
reducing the capacity. Carbon physical characteristics are assumed to be
the same as the Pittsburgh Activated Carbon Company's type SGL (8 x 30 mesh)
and type CAL (12 x 40 mesh).
Velocity - The superficial linear velocity (based on empty column
area) of the water moving through the carbon bed is taken at 7 GPM/ft2.
This velocity was chosen for the base case because most of the pilot plant
data were taken at this velocity.
Contact Time - The required contact time to reduce the COD from 60
to 7 mg/L is set at 50 minutes (based on empty column volume). This is
longer than the 41-minute contact time at Pomona but a longer contact time
is consistent with the use of a larger particle size2. As more data become
available, optimized design may permit within certain limitations, a
reduction in contact time coupled with a higher regeneration rate.
-55-
-------
Carbon Bed Dimensions - A minimum bed depth-to-diameter ratio
(L/D) on the order of one in a vessel was chosen to assure good contacting
during adsorption and backwash. A minimum ratio of this order of
magnitude has been found to be required in order to prevent backmixing
of the carbon during backwash and to prevent channeling during adsorp-
tion. Channeling effects are a function of velocity and would not be as
great a problem at the low velocities encountered in gravity-flow
systems as compared to pressurized, high- velocity contactors. That
is, less pressure drop would be required to overcome channeling
problem at low flow rates than at high flow rates. The contactors
at Pomona and Lake Tahoe have bed^depth-to-diameter ratios greater than one
with velocities at about 7 GPM/ft while the contactors at Njtro
have a ratio less than one with a velocity at about 1 GPM/ft . Pittsburgh
Activated Carbon Company in their "Basic Design Techniques" bulletin
recommend an L/D ratio of two, but in their report to USPHS3, they
used ratios less than one. An L/D ratio of one should be adequate
to assure good contacting for packed columns operating under flooded
conditions at high velocities.
"Idle" Carbon Inventory - The portion of the total carbon
inventory which must be in residence in the system for the purpose
of carrying out regeneration without upsetting the normal operation
of the adsorption train will be referred to as "idle" carbon.
This carbon is idle with respect to the adsorption of waste matter
but, on the other hand, is not idle in the regeneration system.
The amount of "idle" carbon required for regeneration purposes is equal
to the amount in one vessel. The amount of carbon held-up in the
regeneration furnace is less than 17, of the total carbon inventory and
has been neglected.
Backwash Rate - The backwash flow rate was chosen to assure a
307o bed expansion during backwash. This is equivalent to the bed
expansion employed at Pomona. Good backwashing should be a function
of a degree of bed expansion rather than of water velocity so that
the accumulated suspended solids can disengage from the interstices
of the carbon particles. As at Pomona, backwashing of the lead
contactor is assumed to occur once a day using 5% of the daily output
from the secondary plant for backwash and surface wash. Bgd ex-
pansion during backwashing is taken from Cooper and Hager.
Pressure Drop - For the purpose of feed pump sizing, a 50-psi
pressure drop is assumed to be developed before backwashing is re-
quired. This is approximately the same as at Pomona even though
the particle size at Pomona is smaller (16 x 40 vs. 8 x 30). The
major portion of the bed pressure drop is contributed by the sus-
pended solids accumulated on the top layer of carbon. This pressure
drop will not change much over this particle size range since the
void fraction for 12 x 40 is 0.38 and for 8 x 30 is 0.36, not a
significant difference. Carbon bed pressure drop is taken from
Cooper and Hager.^ A spare pump is included for each pump. A gas
engine or diesel drive is provided for the spare feed and backwash
pump so that the plant would not have to be shut down in the case
of power failure.
-56-
-------
"Spare Vessels"- Two extra vessels are added to the number of
vessels which are onstream in the adsorption train. The two "spare"
vessels are required for the operation of the regeneration system
to store the "idle" carbon which is ready to be and has been regenerated.
These two vessels which are in regeneration service are only "spare"
insofar as they are not onstream in the adsorption train.
Regeneration Rate - Regeneration of the spent carbon is assumed
to occur continuously in a multiple-hearth, Herreschoff-type furnace.
The regeneration rate is set by the material balance around the plant
and an assumed carbon loading of 0.522 Ib COD/lb carbon. This loading
corresponds to the loading at Pomona after the carbon has been regenerated
ten times-1. It is assumed that the loading for the two-stage system
will be the same as the loading obtained in the four-stage system at
Pomona. This assumption will be tested for its effect on the economics
in a later section. The regeneration of carbon from the vessels is
assumed to occur on a staggered basis so that all of the lead adsorbers
do not have to be regenerated at the same time. Staggering of regenera-
tion in a multi-trained system presents the possibility of blending
product waters which are worse than and better than specified quality.
This allows the carbon to get more heavily loaded than in a single-
train system thus reducing the regeneration rate.
Carbon Transport - The spent carbon is transported between the
vessel and the furnace area as a carbon-water slurry. The slurry is
then drained in an inclined dewateoring screw coveyor (as at Nitro, West
Virginia) where a 10-minute dewatering period is provided. Dewataring
studies at Lake Tahoe indicate that a moisture content of 45% can be
achieved after a 10-minute drain time while only 40% moisture can be
reached after draining for one day**. Because of the two "spare"
vessels and the feasibility of carbon dewatering in an inclined con-
veyor, there is no need for elevated dewatering bins as are in use
at Pomona and Lake Tahoe.
The carbon slurry transport system design was based on recommen-
dations in the Pittsburgh Activated Carbon Company's Bulletin, "Column
Operating Procedures with Pittsburgh Granular Activated Carbon". Their
recommendations are:
1. Minimum linear velocity of 3 feet per second in slurry
lines to prevent carbon settling.
2. A carbon-to-water ratio of one pound carbon per gallon
of water.
3. Pipeline pressure drop shown in Figure 4 in the above
bulletin.
-57-
-------
Transportation of the make-up carbon from the storage tank to
the quench tank is done pneumatically with the design based on
recommendations in McCabe and Smith?;
1. The air velocity should be between 50 and 100
feet per second.
2. The solids should occupy 3 to 12% of the volume
or the solids-to-air mass ratio should be
between 30 and 100.
Regeneration Furnace - Continuous regeneration of the spent
carbon occurs in a multiple-hearth, Herreschoff-type furnace. The
furnace is sized on the recommendation from B-S-P Corporation of 100
pounds carbon regenerated per day per square foot of hearth area.
This design point was used along with the hearth areas given in B-S-P
Bulletin No. 250. Air pollution control equipment is provided to pre-
vent odors and particulates from escaping in the flue gas stream.
Regeneration Fuel - The total fuel requirement for the regeneration
system is estimated to be 6950 Btu per pound of carbon regenerated.
This figure is based on the following recommendations:
1. B-S-P recommends 3000 Btu/lb carbon for furnace heat
which is confirmed by Pomona operating data.
2. Lake Tahoe data indicate that 1250 Btu/lb carbon for
steam generation of one pound steam per pound carbon is
sufficient for satisfactory carbon regeneration.
3. The afterburner fuel requirement for odor control was
taken at 2700 Btu/lb carbon to raise the flue gas
temperature to 1600°F based on Pomona operating data.
Carbon Loss - The carbon loss in regeneration is assumed to be
5% which is the same as that at Tahoe with 8 x 30 mesh carbon. Storage
facilities are provided for a six-month supply of make-up carbon which
is the amount of carbon in one adsorber in base case 1A.
Materials of Construction - The materials of construction specified
for the purpose of comparative cost estimation is coal tar-epoxy lined
carbon steel vessels with carbon steel piping, the materials which are
in use at Pomona. In light of recent corrosion information, however,
a sheet lining such as laminated hard rubber or PVC would be recommended.
The failure of the coal tar-epoxy linings at Pomona and Tahoe is believed
to be due to poor application, resulting in pin-holes. The use of new
materials would not affect the economic comparisons made in this report,
but would raise the capital cost of all vessels.
-58-
-------
Operating Labor Requirement - A labor requirement of 1-1/4 man
per shift for operation of a 10-MGD plant is estimated one man
operating the regeneration system and one man monitoring vessel
operation one-quarter of the time.
Calculation of Other Operating Costs - The basis and rationale
used in the procedure for calculation of other operating costs such
as maintenance, overhead, amortization, insurance, taxes and unit
costs for utilities are given in Appendix A.
Plant Investment - The items which are included in the estimation
of the plant investment are given in Appendix A.
-59-
-------
ECONOMICS
Comparison of Shop Fabrication and Field Erection of Vessels
Cases 1 & 1A
The comparison of investment and operating costs for the shop -
fabricated (Case 1) and field-erected vessels (Case 1A) is shown below.
The limitation of 13 feet in vessel diameter in the shop fabrication
case resulted in 18 vessels required for the 10-MGD plant (8 trains of
2 vessels each plus 2 spare vessels.) The field erection of vessels
coupled with the restriction of a minimum bed height-to-diameter ratio
of one resulted in six 25*6" I.D. vessels. In the field erection case,
a larger backwash pump is required in order to provide the 30% bed ex-
pansion for good backwashing. Cost comparisons are presented in Table 2
for shop fabrication and field erection. The table shows no significant
difference in either investment or operating costs between the two methods
of construction. Savings in field erection of vessel steel and piping
are offset by field labor, backwash pump and carbon costs. The process
simplicity of field erection made it a logical choice for the basis of
subsequent comparisons.
Comparison of Surge Design and Base Case
Cases 1A & 21
The question being tested in this comparison is whether it is
cheaper to handle the surge in flow rate from 5 to 15 MGD by damping
the variation in a surge basin (Case 1A) or by bringing extra vessels
onstream (Case 21). It was assumed that this variation in flow rate
would occur sinusoidally with a period of one day. A 1,600,000-galion
basin is required to store the excess water while in the other case,
three trains of vessels are required (two additional vessels). The costs
for this comparison are presented in Table 3 which show that it is
significantly more expensive to handle surges in flow by adding vessels.
The two extra vessels along with the extra carbon, piping and larger
feed pump cost more than does the concrete surge basin. The higher
operating costs are a result of the higher investment. The utility costs
average out to be the same as the base case.
Effect of Plant Size on Economics
Cases 1A. 2 & 3
These comparisons attempt to find what economies of scale, if
any, may be realized. The 1-MGD plant requires four 11*6" I.D. vessels,
the 10-MGD plant, six 25*6" I.D. vessels and the 100-MGD plant, forty-
four 25*6" I.D. vessels (20 trains of two vessels each plus four spare
vessels). The regeneration furnace for the 1-31GD plant is the minimum
-60-
-------
TABLE 2
COMPARISON OF SHOP FABRICATION AND FIELD ERECTION OF VESSELS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss,
Carbon Capacity
Vessel Size
Number of Vessels
Number of Trains
Investment, $M
Concrete
Adsorbers
Tanks
Pumps
Special Eq
Piping
Conveyors
Total Major Ma
Plant Investment
Amortization
Total Operating Cost
I
:t2
tin.
Mesh
)SS, %
r, lb COD/lb C
Is
is
lipment
:erial
it
jc/lb.
[nvestment
5, C/M Gal.
rbon @ 26
-------
TABLE 3
COMPARISON OF SURGE DESIGN AND BASE CASE
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ftZ
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
1A
Surge Basin
Downflow
7
50
8 x 30
5
0.522
25'6" I.D. x 30'
6
2
21
Extra Vessels
Downflow
7
50
8 x 30
5
0.522
25'6"I.D. x 30'
8
3
Inves tment. $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26$/Ib.
Fixed-Capital Investment
121
253
5
98
36
158
4
675
1210
390
1600
Operating Costs. C/M Gal.
Make-up Carbon @ 26$/lb.
Power @ l£/kwh
Backwash Water @ 6
-------
size (30-inch I.D. - 6 hearth) and would be run only about 35% of the
time. The fuel required for start-up and shut-down of the furnace is
negligible (470 of the total fuel required) compared to the total con-
sumption. Since the regeneration furnace is run only part of the time,
it was assumed that only one man per shift would be required for the
entire plant.
The regeneration furnace for the 100-MGD plant is a 16* O.D. -6
hearth furnace with a wet scrubber for air pollution control. B-S-P
recommends the use of a wet scrubber rather than a cyclone and after-
burner for air pollution control for large furnaces such as this. The
operating labor requirement for this plant is three men per shift.
The cost comparison for plant sizes are presented in Table 4,
which shows that the investment charges and operating labor are sub-
stantially lower in the 10-MGD case than in the 1-MGD case. High labor
and investment charges are typical for small plants. On. the other hand,
the reduction in investment charges from the 10-MGD plant to the 100»MGD
plant is not very great because there are multiple trains of equipment
required in the 100-»MGD case. Accordingly, savings in investment charges
are not great when multiple trains are required. The power requirements
for the three plant sizes are not the same, principally because the
smaller pumps for the smaller plants are less efficient than in the large
plants. Also, the power requirement for the regeneration system does
not increase in direct proportion to the plant size.
The investment and operating costs are plotted as a function of
plant size in Figure 2. A break in the lines is seen at a plant size
of 10-MGD because at plant sizes larger than this, the vessels would
be erected in the field.
Effect of Gravity-Flow Contactor At Two Plant Sizes on Economics
Cases 1A. 16. 3 & 3A
This comparison, which Identifies savings which can be expected
by replacing a steel pressure vessel contactor system with a concrete
gravity-flow contactor system, is made at flow rates of 10 and 100-MGD.
A major assumption in the design of the gravity-flow system is that the
same contact time is required at 2 and 7 GPM/ft.^ to accomplish the same
removal of COD. There are data to support this assumption - at least in
the range of 4 to 10 GSM/ft^.B Designing a single-stage, gravity-flow
contactor for a higher velocity than 2 GFM/ft.2 would result in deeper
carbon beds and concrete walls which would be prohibitively thick. The
contactor walls must be able to withstand the full water pressure since
the adjacent contactor will be emptied for regeneration at times. Another
assumption in the absence of data is that the same loading will be achieved
in the single-stage, gravity-flow system as in the two-stage system. An
alternate scheme for the gravity-flow case would be a two-stage system
operating at 4 GPM/ft.^. This would avoid the problem with the effect of
velocity on rate of adsorption and the effect of a single-stage contactor
on loading. The number of contactors would be the same in the single and
two-stage system but the height above the carbon for water head would
be doubled and two extra headers would be required (one for the outlet
-63-
-------
TABLE 4
EFFECT OF PLANT SIZE ON ECONOMICS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, 7.
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
2
1 MGD
Downf low
7
50
8 x 30
5
0.522
11'6" I.D. x 30'
4
1
1A
10 MGD
Downf low
7
50
8 x 30
5
0.522
25'6" I.D. :
6
2
x 30'
3
100 MGD
Downflow
7
50
8 x 30
5
0.522
25'6" I.D
44
20
x30'
Inve slEmfln.fr i ffi
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26e/lb.
Fixed-Capital Investment
29
74
2
17
27
43
2
194
331
48
379
Operating Costs. c/M Gal.
Make-up Carbon @ 26c/lb.
Power @ le/kwh
Backwash Water
Fuel Gas @ 25c/MM Btu
Labor @ $3.50/Man-Hr.
Overhead @ 50% Labor + 1.485% PI/Yr.
Amortization
@ 7.47, FCI/Yr.
Maintenance @ 5.75% PI/Yr.
Insurance @ 1% FCI/Yr.
Total Operating Cost
1.09
1.21
0.47(1)
0.15
8.40
5.55
1.09
1.00
0.31(2)
0.15
1.05
1.02
(1) Backwash water (secondary effluent) @ 9c/M gal.
(2) Backwash water (secondary effluent) @ 6c/M gal.
(3) Backwash water (secondary effluent) @ 3c/M gal.
-64-
-------
100
9.0
8.0
7.0
6.0
5.0
4.0
CO
2 3.0
2.0
_
2 I0
> 0.9
z 0.8
_. 0.7
£ 0.6
| 0.5
o
i 0.4
o
ui
>< 0.3
0.2
O.I
ixed-Coptal hvestment
Operating (
ost
: »Jr
:
PLANT CAPACITY, MGD
'
100
.
0
10
°
o
2
-
a
UJ
a
:
100
FIGURE 2
COSTS AS A FUNCTION OF PLANT SIZE
TERTIARY WASTE WATER TREATMENT-GRANULAR ACTIVATED CARBON
-65-
-------
from the first stage at the bottom of the contactors and another for
distribution of this water to the second stage at the grade level). A
roof has been included in the cost estimate of the gravity-flow cases
for the prevention of algae growth on top of the carbon beds. The
requirement of a minimum bed height-to-diameter ratio of about one has
been waived in the design of the gravity-flow contactors. For discussion
on the bed height-to-diameter ratio, see the "Design Bases" section.
Case 1A is the base case - a pressure vessel system at 10-MGD.
The vessel requirement is six 25*6" I.D. x 30' vessels.designed for a
95 psi pressure rating. Water discharges from each vessel through
eight 8" diameter 40 mesh stainless steel well point type screens.
Case 16 (gravity flow system at 10-MGD) requires 10 rectangular
cross section vessels (eight onstream plus two spares), each 20 ft.
wide x 22 ft. long x 25 ft.-6 in. high. Each vessel has a 13 ft.-6 in.
deep carbon bed with 2 ft. of sand and gravel under the carbon and
porous filter bottom under the gravel. A freeboard space of 10 ft. is
provided above the carbon bed to allow for bed expansion during backwash
and to provide height for the water to rise as the bed pressure drop
increases due to the build-up of suspended solids on the bed. Two
separate sets of troughs are proveded: one for removal of the backwash
water and the other for carbon removal for regeneration. Pumps are
provided for pumping the feed water from the bottom of the surge basin
to the feed conduit and from the filtered water conduit at the bottom
of the contactors back to ground level. If the system had complete
gravity flow (e.g., if it were built on a hill), one or both of these
pumps would not be needed and would result in savings of investment
and power costs.
Case 3 (pressure vessel contacting system at 100-MGD) requires
forty-four 25*6" I.D. x 30f vessels (40 onstream plus 4 spares).
Case 3A (gravity-flow at 100-MGD) requires 20 rectangular cross
section vessels (18 onstream plus 2 spares), each 40 ft. wide x 50 ft.
long x 25 ft.-6 in. high. The contactor construction is the same as in
Case 16 described above except that each contactor is split lengthwise
by a central bay for water distribution.
The investment and operating costs for these four cases are shown
in Table 5; as can be seen, the gravity-flow contactor is significantly
less expensive than the vessel system at both 10- and 100-MGD. In
Table 5, the concrete cost is for the concrete in place which includes
field labor costs. Adsorber cost in Cases 1A and 3 is for material only.
It must be re-emphasized that it has been assumed in the absence of data
that a 50-minute contact time will accomplish the same COD removal at
both 2 and 7 GPM/ft. and that the 0.522 Ib. COD/lb. carbon capacity can
be obtained in a single-stage contactor. The savings in the gravity-flow
cases are due mainly to savings in investment. Investment savings come
about from the fact that concrete contactors are less expensive than steel
vessels. The investment and operating costs for gravity-flow and pressur-
ized vessel systems are plotted as a function of plant size in Figure 2.
If the feed water and product pumps could be eliminated for a
complete gravity-flow system, then the fixed-capital investment would be
reduced by $71,000 and $293,000 for Cases 16 and 3A, respectively. The
-66-
-------
TABU; 5
EFFECT OF GRAVITY FLOW CONTACTOR AT TWO PLANT SIZES ON ECONOMICS
C«s« No.
Variable Tested
Contactor Type
Velocity, GPM/ftz
Contact Time, Mln.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
Investment. JM
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26c/lb.
Fixed-Capital Investment
Operating Costs. C/M Gal.
Make-up Carbon @ 26c/lb.
Power 9 lc/kwh
Backwash Water
Fuel Gas @ 25C/MM Btu
Labor @ $3.50/Man-Hr.
Overhead @ 507. Labor + 1.485% PI/Yr.
Amortization
@ 7.47. FCI/Yr.
Maintenance @ 5.757. PI/Yr.
Insurance @ 17, FCI/ ifr.
Total Operating Cost
1A
10 MGD
Vessels
Down flow
7
50
8 x 30
5
0.522
25'6"I.D. x 30'
6
2
121
253
5
98
36
158
4
"675
1210
390
1600
1.09
J;5J(2)
o!l5
1.05
1.02
3.24
1.91
0.44
10.21
16
10 MGD
Gravity
Gravity
2
50
8 x 30
5
0.522
20'w x 22'1 x 25'6"h
10
8
426
-
5
116^)
37
95
4
683
1064
347
1411
1.09
o!3l(2)
0.15
1.05
0.96
2.24(4)
1.68
0.39
8.51
3
100 MGD
Vessels
Down flow
7
50
8 x 30
5
0.522
25'6"I.D. x 30
44
20
811
1765
12
433
150
1599
23
4793
8060
3276
11,736
1.09
0.79
0.18(J)
0.09
0.25
0.45
2.38
1.27
0.32
6.82
3A
100 MGD
Gravity
Gravity
2
50
8 x 30
5
0.522
40'w x 50'1 x 25'
20
18
1771
149
527
_ 23
2823
4271
3335
7606
09
09
25
0.30
6"h
,67
0.21
4.63
(1) Water back to grade level - costs lower for complete gravity flow
(2) Secondary effluent @ 6$/M gal.
(3) Secondary effluent @ 3c/M gal.
(4) 30-year plant life assumed - Amortization Rate » 5.8%/yr.
- see text
-------
power and investment charges (at 15.6% of investment per year) would
decrease by 0.55 and 0.30^/M gal., respectively, for Case 16 and 0.060
and 0.13<:/M gal., respectively for Case 3A.
Effect of "Idle" Carbon Inventory on Economics
Cases 4. 4A. 10 & 10A
The object of this comparison is to determine the optimum
amount of idle carbon which can be present in the system. The term
"idle" carbon describes the carbon which must be bought with the plant
but is not actually onstream. This portion of the total carbon inventory
must be in residence in the system for the purpose of carrying out
regeneration without upsetting the normal operation of the adsorption
train. It has been assumed that to provide smooth operation during
the regeneration sequence, one of the two spare vessels in the plant
must be filled with carbon. This amount of idle carbon can be decreased
at the expense of extra vessels (more trains) and associated piping. As
the number of trains is increased, the size of the vessels decreases;
therefore, the amount of carbon in a vessel decreases. The vessel and
carbon requirements for the four cases considered are shown below:
Fraction
Case Vessels Total Carbon. Ib Idle Carbon. Ib Idle Carbon
4 4-30'I.D. x 43'6" 1,776,000 592,000 1/3
4A 6-21fI.D. x 43'6" 1,450,000 290,000 1/5
10A 6-25'6"I.D. x 25* 1,111,000 222,200 1/5
10 10-18'I.D. x 25' 996,300 110,700 1/9
The comparison for the effect of idle carbon is shown in Table 6.
It should be noted that Cases 4 and 4A are not to be compared with Cases
10 and 10A because of the changes in velocity, contact time and particle
size. Valid comparisons can be made only between Cases 4 and 4A and
between Cases 10 and 10A where these variables do not change.
As can be seen by comparing Cases 4 and 4A, there is a 10% savings
in investment and a 0.6$/M gal. savings in operating cost when the idle
carbon is reduced from 1/3 to 1/5 of the total carbon inventory. However,
when Cases 10 and 10A are compared, it is seen that there is little
difference in investment or operating costs when the idle carbon is re-
duced from 1/5 to 1/9 of the total carbon inventory. The conclusion
drawn is that the amount of idle carbon must be between 10% and 20% of
the total carbon inventory in order to yield minimum costs. The costs
can be expected to rise again as more vessels are added and the idle
carbon inventory is reduced further. A point will be reached where the
decreasing carbon investment will be more than offset by increasing
vessel and piping costs.
-68-
-------
TABLE 6
EFFECT OF IDLE CARBON ON ECONOMICS
Case No.
Variable Tested
4*
1/3 Idle Carbon
1 Train
4A*
1/5 Idle Carbon
2 Trains
10A*
1/5 Idle Carbon
2 Train*
10*
1/9 Idle Carbon
4 Trains
Contactor Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, '/.
Carbon Capacity, Ib COD/lb
Vessel Size
Number of Vessels
Number of Trains
Downflow
10
50
8 x 30
5
0.522
30'I.D. x 43'6"
4
1
Downflow
10
50
8 x 30
5
0.522
21'I.D. x 43'6"
6
2
0
25'6"I
Down flow
7
40
12 x 40
5
522
D. x 25'
6
2
18'
Dovnflow
7
40
12 x 40
5
0.522
I.D. x 25'
10
4
VO
I
Investment. $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon
Fixed-Capital Investment
112
342
5
134
35
107
4
739
1290
_462(1)
1752*
121
254
5
101
36
158
_4
679
1202
377(1)
1579*
121
246
5
103
36
158
4
673
1202
-212.
1524*
121
219
5
80
37
236
4
703
1233
289
1522*
(2)
Operating Costs. c/M Gal.
Make-up Carbon
Power @ Ic/kwh
Backwash Water @ 6<:/M gal.
Fuel Gas @ 25C/MM Btu
Labor 9 $3.50/Man-Hr.
Overhead @ 50% Labor + 1.485% PI/Yr.
Amortization
@ 7.4% FCI/Yr.
Maintenance @ 5.75% PI/Yr.
Insurance @ 1% FCI/Yr.
Total Operating Cost
1.09'1)
1.00
0.31
1.09*1)
1.00
0.31
0.15
1.05
1.01
3.20
1.89
0.43
10.13*
1.08
0.31
0.15
1.05
1.03
3.09
1.94
0.42
10.29*
(1) 8 x 30 mesh make-up carbon @ 26c/lb.
(2) 12 x 40 mesh make-up carbon @ 29c/lb.
* Cases 4 and 4A are not to be compared with Cases 10 and 10A.
between Cases 10 and 10A.
Valid comparisons can be made only between Cases 4 and 4A and
-------
Effect of Velocity on Economics
Cases 1A, 4A & 6
The purpose of this comparison is to find the effect of
velocity alone, under the assumption that velocity does not affect
the performance (removal efficiency and carbon loading) of the
carbon contactor. This assumption is consistent with most of the
data on velocity . Three velocities are considered (4, 7 and 10
GPM/ft. ) to find the economic effect of changing the pressure drop
and the contactor height-to-diameter ratio.
The vessel requirement for Case 4A (10 GPM/ft.2) is six 21*
I.D. x 43*6" vessels, for Case 1A (7 GPM/ft.2) it is six 25'6" I.D.
x 30' vessels and for Case 6 (4 GPM/ft.2) it is twenty-eight 13' I.D.
x 17f6" vessels. Thirteen trains of vessels were required for Case 6
due to the restriction of a minimum bed height-to-diameter ratio of
one. The rationale for this minimum ratio is discussed in the "Design
Bases" section.
The carbon bed pressure drop is 5.5 inches 1^0 per foot carbon
at 10 GPM/ft.2, 4 in./ft. at 7 GPM/ft.2 and 2 in>/ft. at 4 GFM/ft.2.
Carbon bed pressure drop constitutes only a small portion of the total
head which must be developed by the feed pump, the remainder being
pressure drop due to suspended solids and pipeline.
The comparison to show the effect of velocity on the economics
is shown in Table 7. As can be seen from the economics, there is no
significant difference between the 10 and 7 GPM/ft.2 cases. The power for
the pumps is the same in Case 4A as in Case 1A because the larger feed
pump in Case 4A was offset by a smaller backwash pump. The backwash
pump is smaller because the vessels are smaller in diameter. All other
costs are about the same for these two cases.
The 4 GPM/ft.2 for Case 6 is substantially higher in cost than
for either the 7 or 10 GPM/ft.2 case. The major cost items are the
vessels and associated piping. Many vessels (28) were required in
Case 6 due to the low velocity coupled with the restriction of a
minimum bed height-to-diameter ratio of one which is necessary for
good contacting and backwashing.
The conclusion drawn from this comparison is that velocity has
little effect on the economics, once the velocity is above 7 GPM/ft.2.
The major assumption is that velocity does not affect the performance
of the contactor. No additional data are needed to verify the effect
of velocity alone on the economics.
Effect of Contact Time at High Velocity on Economics
Cases 4A & 5
This comparison identifies economic effect of changing only
the contact time from 50 to 35 minutes with the velocity at 10 GPM/ft.2.
Such a reduction in required contact time for 8770 removal of COD is
predicted if the correlation shown in Figure 12 of the literature report^
-70-
-------
TABLE 7
EFFECT OF VELOCITY ON ECONOMICS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
4A
10 GPM/ft2
Velocity
Downflow
10
50
8 x 30
5
0.522
21' I.D. x 43'6"
1A
7 GPM/ft2
Velocity
Downflow
7
50
8 x 30
5
0.522
25'6" I.D. x
6
2
30'
13'
4 GPM/ft2
Velocity
Downflow
4
50
8 x 30
5
0.522
I.D.xl7'6"
28
13
Inve s tment, $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26<:/lb.
Fixed-Capital Investment
121
254
5
101
36
158
4
679
1202
377
1579
121
253
5
98
36
158
4
675
1210
390
1600
141
452
5
68
43
330
4
1043
1651
315
1966
Operating Costs, C/M Gal.
Make-up Carbon @ 26
-------
is valid, rather than the correlations in Figures 3-108 which showed
no effect of velocity on required contact time. The data in Figure 12
were the only data which could be found in a form suitable for analysis
on the effect of velocity on efficiency of removal for COD. It should
be noted that there was no effect of velocity on Total Organic Carbon
removal10. The effect of velocity on COD removal still remains to be
resolved. This reduction in contact time results in a reduction of vessel
size from 21'l.D. x 43'6" to 21'l.D. x 30'.
The economics for these two cases are given in Table 8 which
shows that there is a significant reduction in costs when the contact
time is reduced. Contactor volume is, of course, directly proportional
to contact time so that savings are realized in both vessel cost and
carbon inventory. Power consumption is less in the 35-minute contact
time case because the pressure loss is less in the shorter bed. This
comparison shows that it is important to know the contact time required
in order to achieve the desired removal. The assumption in this case
was based on the effect of velocity shown in the literature report.9
Effect of Contact Time at Low Velocity on Economics
Cases 6 & 7
This comparison finds the economic effect of changing the
contact time from 50 to 87 minutes with the velocity at 4 GPM/ft.2.
An increase of this order in contact time is required if the correlation
in Figure 12 of the literature report9 is valid rather than the
correlations in Figures 3-108. The data in Figure 12 were the only
data which could be found in a form suitable for analysis on the effect
of velocity on efficiency of removal for COD. Figures 3-10 predicted
that there is no effect of velocity on required contact time and is
represented here by Case 6. This increase in contact time (Case 7)
results in the vessel requirement changing from twenty-eight 13'l.D.
x 17'6" vessels for 50-minute contact time to ten 23'6" I.D. x 30*
vessels for 87-minute contact time. The number of vessels for the
87-minute contact time could be reduced because the bed depth in a
vessel increased, and the design condition of minimum bed depth-to-
diameter of one allows the diameter to increase.
The economics for these two cases are given in Table 9 which
shows a significant increase in costs when the contact time is increased
from 50 to 87 minutes. Again, contactor volume is directly proportional
to contact time so that the carbon investment is increased. The plant
investment did not increase correspondingly ($1.651 vs. $1.686 MM)
because the number of vessels was reduced from 28 to 10. The power
consumption is actually less in Case 7 because the area occupied by
the vessels is smaller, thus reducing the pipeline pressure drop by
more than the increase in pressure drop due to the deeper carbon bed.
As in the preceeding section of this report, this comparison shows
that it is quite important to determine the relation between contact
time and COD removal. The assumption made in Case 7 was based on the
effect of velocity shown in Figure 12 of the literature report9. Case
6 corresponds to the correlation in Figures 3-10 in the literature
report8 which showed no effect of velocity on contact time.
-72-
-------
TABLE 8
EFFECT OF CONTACT TIME AT HIGH VELOCITY ON ECONOMICS
Case No.
Variable Tested
Contactor 'Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
4A
50 Minute
Contact Time
Downflow
10
50
8 x 30
5
0.522
21' i.D. x 43'6"
6
2
35 Minute
Contact Time
Downflow
10
35
8 x 30
5
0.522
21' I.D. x 30'
6
2
Investment, $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26
-------
TABLE 9
EFFECT OF CONTACT TIME AT LOW VELOCITY ON ECONOMICS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
6
50 Minute
Contact Time
Downflow
4
50
8 x 30
5
0.522
13'I.D. x 17'6"
28
13
87 Minute
Contact Time
Downflow
4
87
8 x 30
5
0.522
23'6"I.D. x 30'
10
4
Investment. $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26c/lb.
Fixed-Capital Investment
Operating Costst C/M Gal.
Make-up Carbon @ 26
-------
Effect of Velocity and Contact Time on Economics
Cases 1A, 5 & 7
With the COD removal held constant at 87% (60 mg/L feed to
7 mg/L effluent), this comparison identifies the full economic impact
of varying velocity (4, 7 and 10 GFM/ft.2) with the assumption that
velocity and removal efficiency are related as in Figure 12 of the
literature report^. That correlation predicts that the depth of
carbon required for an equal degree of removal is the same regardless
of the.velocity - at least in the range of 4 to 10 GPM/ft.2. The data
presented in the literature report were the only data which could be
found in a form suitable for analysis on the effect of velocity on
efficiency of removal for COD. This comparison includes the effects
of pressure drop and contactor height-to-diameter ratio. Case 5
(10 GPM/ft.2 and 35-minute contact time) requires six 21'I.D. x 30'
vessels; Case 1A (7 GPM/ft.2 and 50-minute contact time) requires six
25'6" I.D. x 30' vessels; Case 7 (4 GPM/ft.2 and 87-minute contact
time) requires ten 23T6"I.D. x 30' vessels.
The economics for these three cases are given in Table 10.
The economics show that the effect of velocity is quite significant
if the assumption that velocity affects required contact time is
correct. Since contactor volume is directly proportional to contact
time, both the vessel cost and carbon inventory decrease with decreasing
contact time. This decreasing investment is, of course, reflected in
the operating cost items which are related to investment. The power
requirement at 10 GPM/ft.2 is less than at 7 GPM/ft.2 because of a
slightly lower pipeline pressure drop and a smaller backwash pump.
This comparison shows quite clearly how important it is to
know the relation between velocity and contact time. Based on this
comparison, the system should be run at 10 GPM/ft.2. On the other
hand, if velocity had no effect on required contact time (see
section on "Effect of Velocity") the system should be run between 7
and 10 GPM/ft.2.
Effect of Particle Size on Economics
Cases 1A & 8
This comparison identifies the economic effect of changing
only the particle size from 8 x 30 mesh (Case 1A) to 12 x 40 mesh
(Case 8). The smaller particle size increases the carbon bed pressure
drop to 8.5 in. l^O/ft. carbon compared to 4 in. I^O/ft. carbon for
8 x 30 mesh carbon. The contribution of the suspended solids to the
total pressure drop is assumed to be the same for both particle sizes,
since the void fraction is about the same for both sizes (0.38 for 12
x 40 vs. 0.36 for 8 x 30 ). The backwash pump is smaller for the smaller
size carbon because a lower velocity is required (10 vs. 17 GPM/ft.2)
to achieve a 30% bed expansion during backwashing.
-75-
-------
TABLE 10
EFFECT OF VELOCITY AND CONTACT TIME ON ECONOMICS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
10 GPM/ft*
35 Min. C. T.
Downflow
10
35
8 x 30
5
0.522
21'I.D. x 30'
6
2
1A
7 GPM/ft2
50 Min. C. T.
Downflow
7
50
8 x 30
5
0.522
25'6"I.D. x 30'
6
2
4 GPM/ft2
87 Min. C. T.
Downflow
4
87
8 x 30
5
0.522
23'6"I.D. x 30"
10
4
Investment, jjjM
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26<:/lb.
Fixed-Capital Investment
Operating Costs. $/M Gal.
Make-up Carbon @ 26
-------
The bulk density for the 12 x 40 mesh carbon is less than for
8 x 30 mesh (22.9 vs. 25.0 lb./ft.3 thus, a smaller quantity of carbon
is required to provide the same contact time, however, this lower density
is offset by a higher price per pound (29c/lb. for 12 x 40 vs. 26q
/lb. for 8 x 30). These compensating factors result in a nearly
equal price per cubic foot. The void fraction for each size carbon
is nearly equal (see above) so that the "real" water contact time will
be about the same even though the 50-minute contact time used in
design is based on the empty vessel volume. Future changes relative
costs and densities of carbon grades might change this considerably.
The comparison for the effect of particle size on economics
is shown in Table 11, which shows that there is little difference
between the operating costs and no significant difference in invest-
ment in the two cases. The carbon make-up cost and power requirement
for the 12 x 40 mesh particle size are higher as expected,. The assumption
has been made in this comparison that particle size affects neither
the required contact time nor regeneration loss. The effects of these
variables will be examined in later sections.
Effect of Regeneration Loss on Economics
Cases 1A, 12 & 12A
This comparison shows the cost if the regeneration loss is 2, 5
or 10 percent. A loss of 2% represents 'jhe best performance which was
obtained at the Wyandotte Chemical facility. A 57» loss represents
the loss at Lake Tahoe while a 107» loss was the highest obtained at Pomona.
Because the regeneration loss changes, the size of the fresh carbon
storage bin changes, since this vessel must hold a six-month supply of
make-up carbon. All other equipment remains the same in each case.
Losses can occur by attrition in the pipelines and screw conveyor,
by attrition, decrepitation, gasification and burning in the regeneration
furnace, or by thermal shock in the quench tank. Variables which may
affect tha regeneration loss are particle size and shape, type and degree
of loading, amount of handling, regeneration severity and carbon base
(such as bituminous coal, wood, coconut shell, etc.).
The comparison for the effect of regeneration loss on the costs is
shown in Table 12, which shows that the regeneration loss has a substantial
effect on the operating costs, while the investment is about the same for each
case. The regeneration loss has an effect on the make-up carbon cost.
The economics show that it is very important to minimize the regeneration loss
and that there is incentive to make the regeneration process as efficient
as possible - at least, try to reduce losses to 2 percent. The regeneration
process should be studied carefully to pinpoint where the carbon loss
actually occurs and to identify which variables affect the loss.
Effect of_ Particle Size and Regeneration Loss on Economics
Cases 1A. 8, 9 & 12
This comparison identifies the economic impact of the assumption that
the observed difference in regeneration loss at Pomona (1070 loss
-77-
-------
TABLE 11
EFFECT OF PARTICLE SIZE ON ECONOMICS
Case No. 1A 8
8 x 30 mesh 12 x 40 mesh
Variable Tested Particle Size Particle Size
Contactor Type Downflow Downflow
Velocity, GPM/ft2 7 7
Contact Time, Min. 50 50
Particle Size, Mesh 8 x 30 12 x 40
Regeneration Loss, ? 55
Carbon Capacity, Ib COD/lb C 0.522 0.522
Vessel Size 25'6"I.D. x 30' 25'6"I.D. x 30'
Number of Vessels 6 6
Number of Trains 2 2
Investment, JM
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon
Fixed-Capital Investment
Operating Costs, c/M Gal.
Make-up Carbon
Power @ lc/kwh
Backwash @ 6
-------
TABLE 12
EFFECT OF REGENERATION LOSS ON ECONOMICS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %>
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
12
27, Regen. Loss
Downflow
7
50
8 x 30
2
0.522
25'6" I.D. x 30'
6
2
1A
5% Regen. Loss
Downflow
7
50
8 x 30
5
0.522
25'6" I.D. x 30'
6
2
12A
10% Regen. Loss
Downflow
7
50
8 x 30
10
0.522
25'6" I.D. 30'
6
2
i
-vl
Investment, $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26c/lb.
Fixed-Capital Investment
121
253
5
98
36
158
4
121
253
11
98
36
158
4
681
1217
390
1607
Operating Costs, C/M Gal.
Make-up Carbon @ 26c/lb.
Power @ Ic/kwh
Backwash Water
-------
with 16 x 40 mesh carbon) and Lake Tahoe (5% loss with 8 x 30 mesh)
are attributable to difference in particle size. As noted in the
literature report1*, regeneration loss might be expected to be higher
for the smaller size carbon since the smaller carbon is more active
for adsorption (higher adsorption rate) and, therefore, should be more
active for burning and gasification during regeneration.
Regeneration losses of 2 and 5 percent for 8 x 30 mesh carbon
are compared with losses of 5 and 10 percent for 12 x 40 mesh. If
regeneration loss is a function of particle size, then the larger
particle should yield a lower loss. On this basis, the loss with 8
x 30 mesh carbon should be compared with a higher loss with the 12 x
40 mesh carbon. A loss of 2% represents the lowest reported regeneration
loss obtained at the Wyandotte Chemical facility. A loss of 5% corres-
ponds to the loss at Tahoe with 8 x 30 mesh carbon, while a loss of 10%
was the highest obtained at Pomona with 16 x 40 mesh carbon. Table 13
shows the comparison for the effect of particle size and regeneration
loss. From Table 13, if the 5% loss with 8 x 30 mesh carbon (Case 1A)
is compared with the 10% loss and 12 x 40 mesh (Case 9), it is quite
obvious that the 8 x 30 mesh carbon results in much lower costs. Even
if the regeneration process could be improved so that the 8 x 30 mesh
had a 2% loss (Case 12) and the 12 x 40 mesh had 5% loss (Case 8), the
larger size carbon still is significantly less expensive. If particle
size does affect regeneration loss in this way, then a larger size
carbon should be used.
Effects of Particle Size and Contact Time on Economics
Cases 1A & 10A
This comparison identifies the economic result of the assumption
that the contact time for a given removal is a function of particle size.
The functionality is suggested in the literature report*2. An extrapolation
of Figure 16 in that report13 was made to predict a 20% reduction in contact
time at an 87% COD removal, when the particle size is changed from 8 x 30
to 12 x 40 mesh.
Case 1A (8 x 30 mesh carbon with 50-minute contact time) requires
six 25*6" I.D. x 30' vessels while Case IDA (12 x 40 mesh with 40-minute
contact time)requires six 25f6" I.D. x 25' vessels. As discussed in the
section, "Effect of Particle Size," the carbon bed pressure drop is
higher for the smaller size carbon (4 inches H20/ft. carbon for 8 x 30
mesh carbon vs. 8.5 in. H20/ft. for 12 x 40 mesh) and cost per pound of
carbon is higher (26c for 8 x 30 vs. 29c for 12 x 40). However, the
backwash rate of 12 x 40 mesh carbon is lower (10 GPM/ft.z for 12 x 40
vs. 17 GPM/ft.2 for 8 x 30).
The economic comparison for these cases is shown in Table 14,
which shows that the investment and operating costs are nearly the
same for the case of 8 x 30 mesh carbon with 50-minute contact time and
the case of 12 x 40 mesh with 40-minute contact time. If particle size
affects only the required contact time, there is no difference in cost
for the two particle sizes.
-80-
-------
TABLE 13
00
»-«
I
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
Investment, $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
. Piping
Conveyors
Total Major Material
Plant Investment
Carbon
Fixed-Capital Investment
Operating Costs, c/M Gal.
Make-up Carbon
Power @ lc/kwh
Backwash Water @ 6c/M gal.
Fuel Gas @ 25C/MM Btu
Labor @ $3,50/Man-Hr.
Overhead @ 50% Labor + 1.485% PI/Yr.
Amortization
@ 7.4% FCI/Yr.
Maintenance @ 5.75% PI/Yr.
Insurance @ 1% FCI/Yr.
Total Operating Cost
(1) 8 x 30 mesh carbon @ 26c/lb.
(2) 12 x 40 mesh carbon @ 29c/lb.
SIZE AND REGENERATION LOSS ON ECONOMICS
12
8 x 30 Mesh
2% Regen.Loss
Down flow
7
50
8 x 30
2
0.522
25'6"I.D. x 30'
6
2
121
253
4
98
36
158
4
674
1208
390(1)
1598
0.44^)
1.00
0.31
0.15
1.05
1.02
3.24
1.90
0.44
9.55
1A
8 x 30 Mesh
5% Regen. Loss
Down flow
7
50
8 x 30
5
0.522
25'6"I.D. x 30'
6
2
121
253
5
98
36
158
4
675
1210
390
-------
TABLE 14
EFFECT OF PARTICLE SIZE AND CONTACT TIME ON ECONOMICS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
1A
8 x 30 Mesh Carbon
50 Minute Contact Time
Downflow
7
50
8 x 30
5
0.522
25'6"I.D. x 30'
6
2
10A
12 x 40 Mesh Carbon
40 Minute Contact Time
Downflow
7
40
12 x 40
5
0.522
25'6" x I.D. x 25*
6
2
Investment, $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon
Fixed-Capital Investment
Operating Costs. Q/M Gal.
Make-up Carbon
Power @ lc/kwh
Backwash Water @ 6(?/M gal.
Fuel Gas @ 25
-------
Effect of Particle Size, Contact Time and Regeneration Loss On Economics
Cases 1A, 1QA, 11 & 12
This comparison identifies the economic consequence of the
assumption that particle size affects both the contact time and
regeneration loss. In this comparison, 8 x 30 mesh carbon with 50-
minute contact time and TL (Case 12) and 57. (Case 1A) regeneration
loss is compared with 12 x 40 mesh with 40-minute contact time and
57. (Case 10A) and 107= (Case 11) regeneration loss.
The vessel requirements are six 25'6" I.D. x 30' vessels for
Cases 1A and 12 and six 25'6" I.D. x 25' vessels for Cases 10 and 11.
Other factors which change in these cases and which have been discussed
in previous sections are pressure drop, fresh carbon storage and carbon
cost.
The comparison showing the economic effect of particle size,
contact time and regeneration loss is given in Table 15. By comparing
Cases 1A and 11, if the use of 12 x 40 mesh carbon results in a 40-
minute contact time and 107o regeneration loss, it is seen that the
operating costs are substantially higher than for the larger size
carbon. Similarly, if the regeneration losses are reduced to 270 and
57. for 8 x 30 (Case 12) and 12 x 40 (Case 10A), respectively, there
is still a significant, though smaller, difference.
From the economics, it can be concluded that if particle size
affects both required contact time and regeneration loss as suspected,
then the larger size carbon is definitely preferred.
Effect of Carbon Capacity on Economics
Cases 1A, 13 & 14
This comparison identifies the economic effect if the carbon
capacity is different from that assumed in the base case. Three
capacities are examined:
1. 0.87 Ib. COD/lb. carbon, the first regeneration cycle capacity at
Pomona (Case 13).
2. 0.522 Ib. COD/lb. carbon, the capacity predicted from Pomona data
after the system reaches steady state with respect to number of
regeneration cycles (Case LA).
3. 0.25 Ib. COD/lb. carbon, the capacity at Lake Tahoe (Case 14).
The effect of the number of regeneration cycles on capacity was
taken from the literature report.14 The steady state mentioned above
was an extrapolated value read as 607. of the original capacity where the
capacity appears to be leveling off. Since the capacity determines the
regeneration rate, as the capacity decreases, the regeneration furnace
size will increase. Also, since the regeneration loss increases with
regeneration rate, the fresh carbon storage tank increases in size.
The furnace size varies as follows for a 10-MGD plant:
-83-
-------
TABU 15
Case No.
Variable Tested
Contactor Type
Velocity, CPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, 7.
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
Investment. |M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon
Fixed-Capital Investment
Operating Costs, c/M Cal_.
Make-up Carbon
Power @ Ic/kwh
Backwash Water
1.06
0.31
0.15
1.05
1.02
3.10
1.90
0.42
11.45
LOSS ON ECONOMICS
12
8 x 30 Mesh
Down flow
7
50
8 x 30
2
0.522
25'6"I.D. x 30'
6
2
121
253
4
98
36
158
4
674
1208 a)
390
1598
0.44(1>
1.00
0.31
0.15
1.05
1.02
3.24
1.90
0.44
9.55
10A
12 x 40 Mesh
Downflow
7
40
12 x 40
5
0.522
25'6"I.D. x 25'
6
2
121
246
5
103
36
158
_ 4
673
1202
'
_
1524
1.06
0.31
0.15
.05
.03
3.09
1.94
0.42
10.29
(1) 8 x 30 mesh carbon (3 26c/lb.
(2) 12 x 40 mesh carbon (3 29$/lb.
-------
Regeneration Rate,
Capacity. Ib. COD/lb, C Ib./hr. Furnace Size
°«25 735 8'6" O.D. - 6 Hearth
0.522 350 6' O.D. - 6 Hearth
O.8? 212 39" I.D. - 10 Hearfth
.The comparison for the effect of capacity is shown in Table 16.
As can be seen in Table 16, the carbon capacity has a strong
effect on the operating cost but very little effect on the investment.
A spread in operating cost of 2
-------
TABLE 16
EFFECT OF CARBON CAPACITY ON ECONOMICS
Case No.
Variable Tested
14
1A
13
0.25 Ib. COD/lb. C 0.522 Ib. COD/lb. C 0.87 Ib. COD/lb. C
Contactor Type
Velocity, GPM/ft2
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
Downflow
7
50
8 x 30
5
0.25
25'6nI.D. x
6
2
30'
Downflow
7
50
8 x 30
5
0.522
25'6"I.D. x 30'
6
2
Downflow
7
50
8 x 30
5
0.87
25'6"I.D. x 30'
6
2
Investment, $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26
-------
of the total carbon capacity is for dissolved COD. On this basis,
the capacity for the upflow contactor was chosen to be 0.348 Ib.
dissolved COD/lb. carbon. The COD concentration entering the plant
is 60 mg/L and is reduced to 42 mg/L by the sand and gravel filter.
The adsorbers do not have to have extra freeboard above the carbon
bed to allow for expansion during backwash, since any backwashing
would be downflow. This results in 25'6" I.D. x 25' vessel size.
Otherwise, the flow sheet for the upflow system is the same as for
the. downflow system.
The gravity-flow system (Case 16) is the one described
previously in the section "Effect of Gravity Flow Contactor at Two
Plant Sizes on Economics." This system has 10 rectangular cross-
section, single-stage vessels, each 20 ft. wide x 22 ft. long x 25
ft.-6 in. high. Each vessel has a 13 ft.-6 in. deep carbon bed, 2
ft. of sand and gravel under the carbon, with porous filter bottom
under the gravel. A 10 ft. freeboard space above the carbon is provided
for water rise as the bed pressure drop increases from suspended solids.
Two separate troughs are provided for removal of backwash water and
carbon for regeneration. The water velocity through the carbon bed
is 2 GPM/ft.2 compared to 7 GPM/ft.2 for the downflow and upflow
systems. Designing a single-stage, gravity-flow contactor for a
higher velocity would result in deeper carbon beds and concrete
walls which would be prohibitively thick. The contactor walls must
be able to withstand the full water pressure since the adjacent
contactor will be emptied for regeneration at times.
The cost comparison for the downflow, upflow and gravity-
flow systems, is shown in Table 17. By comparing Cases 1A and 15,
the upflow system is seen to be significantly more expensive than
the downflow system. The additional cost is due to the sand and gravel
filter needed for suspended solids removal. The higher operating
cost is a reflection of the higher investment.
As observed in the section "Effect of Gravity Flow Contactor
at Two Plant Sizes on Economics," the gravity-flow contactor is signi-
ficantly less expensive than the downflow system (Case 16 vs. 1A).
Investment savings are realized because concrete contactors are less
expensive than steel vessels. In Table 17, the concrete cost includes
field labor cost-the cost shown is for the concrete in place. Adsorber
cost in Cases 1A and 15 is for material only. If the feed water and
product pumps could be eliminated for a complete gravity-flow system,
then the fixed-capital investment would be reduced by $71,000. The
power and investment charges (at 15.67o/year) would be reduced by 0.55
and 0.30
-------
TABLE 17
EFFECT OF CONTACTOR SYSTEM TYPE ON ECONOMICS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft2
Contact Time, Win.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
1A
Downflow
Downflow
7
50
8 x 30
5
0.522(1)
25'6"I.D. x 30'
6
2
15
Up flow
Up flow
7
50
8 x 30
5
25'6"I.D. x 25'
6
2
16
Gravity
Gravity
2
50
8 x 30
0.522(1)
20'w x 22'1 x 25'6"h
10
8
Investment, $M
Concrete
Adsorbers & Filters
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26c/lb.
Fixed-Capital Investment
426
5
116
37
95
4
683
1064
347
1411
Operating Costs, c/M Gal.
Make-up Carbon @ 26c/lb.
Power @ Ic/kwh
Backwash Water
1.68
0.39
-or
(1) 707, of total carbon capacity is dissolved - 0.522 Ib. total COD/lb. C -
0.348 Ib. dissolved COD/lb. C
(2) 30-year plant life assumed - Amortization Rate = 5.8%/Yr.
.88-
-------
Effect of Adsorbent Cost on Economics
Cases 1A & 17
This comparison shows the economic effect of changing the cost
of the adsorbent, assuming that a lower cost adsorbent would give com-
parable process performance. This change affects the carbon inventory
investment and the make-up carbon cost,, An adsorbent cost range of
0 to 30<:/lb. is considered. When the adsorbent cost is zero, there
is no point in regenerating the adsorbent, since it would simply be
burned (for heating elsewhere) or discarded. For this situation,
the regeneration system is not needed and the cost for this system
is shown in Table 18 as Case 17. Since there is no regeneration
furnace, the labor requirements is reduced from 1-1/4 to 1/4 men/shift.
Also, there is no fuel required since all the fuel in Case 1A is used
in the regeneration system.
A plot of the total operating cost against adsorbent cost is
shown in Figure 3. As seen by the solid line on Figure 3, as the
adsorbent cost drops from 30 to 0<:/lb., the operating cost decreases
from 10.51 to 8.21<:/M gal. When the regeneration system is removed
at Oc/lb., the operating cost drops from 8.21 to 6.62$/M gal. Thus,
it is seen that adsorbent cost has a strong effect on the economics.
It should be noted that there are other ways to draw the line in
Figure 3. For example, the broken line on Figure 3 shows the total
operating cost if the used adsorbent is discarded. It is seen at
about 2.0c/lb. adsorbent (net cost including possible fuel value
credit) that the operating cost exceeds the cost when the adsorbent
is regenerated. Therefore, at above 2.0<:/lb., it pays to regenerate
the adsorbent.
Effect of Number of Contacting Stages on Economics
Cases 1A, 20, 20A & 20B
The purpose of this comparison is to discover if there is
an optimum number of contacting stages. As the number of stages
increases, the number of vessels and amount of piping will increase.
But as the number of stages increases, so should the carbon capacity.
That is, as the total bed depth (47 feet for 50-minute contact time
at 7 GPM/ft. ) is split into smaller segments, the average loading
(capacity) of the carbon in the first contactor will increase. For
example, consider the 2-stage and 4-stage system shown in Figure 4.
It is assumed that breakthrough will occur in the effluent from the
last vessel for both systems after the same length of time onstream.
When breakthrough does occur, the first vessel in each system will
be taken offstream for regeneration. In the 4-stage system, only
12 feet of carbon will be regenerated; in the 2-stage system, 24
feet will be regenerated. This 24 feet will have the same loading
(capacity) as the first two vessels in the.4-stage case. But for
the regeneration operation, an entire vessel must be taken offstream
and regenerated at one time. Since the loading in the first 12 feet
is higher than in the next 12 feet, the average loading for the
entire 24 feet is lower than for the first 12 feet.
-89-
-------
TABLE 18
EFFECT OF ADSORBENT COST ON ECONOMICS
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
1A
Regen. System
Downflow
7
50
8 x 30
5
0.522
25'6"I.D. x 30'
6
2
17
No Regen. System
Downflow
7
50
8 x 30
0.522
25'6"I.D. x 30'
6
2
Investment, $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon
Fixed-Capital Investment
121
253
5
98
4
158
4
643
1171
ml
(2)
Operating Costs, c/M
Make-up Carbon
Power (? lo/kwh
Backwash Water @ 6<:/M gal.
Fuel Gas @ 25$ /MM Btu
Labor @ $3.50/Man-Hr.
Overhead @ 5070 Labor +
1.4857, PI/Yr.
Amortization
@ 7.470FCI/Yr.
Maintenance @ 5.75% PI/Yr.
Insurance @ 17. FCI/Yr.
Total Operating Cost
1.00
0.31
0.15
1.02
. (2)
0.99
0.31
0.58
(1) 8 x 30 mesh carbon @ 26(?/lb.
(2) 8 x 30 adsorbent at no cost
(3) 1-1/4 man/shift operating labor requirement
(4) 1/4 man/shift operating labor requirement
-90-
-------
15.0
13.0
!|2.0
NO REGEN
"ADSORBEN
ERATION SYSTEM
DISCARDED
11.0
10.0
O
90
8.0
7.0
.ADSORBEN
SYSTEM I
T REGENER
CLUDED
TION
6.0
0.0'-
0
10 15 20
ADSORBENT COST, 9 /LB
25
30
FIGURE 3
TOTAL OPERATING COST AS A FUNCTION OF ADSORBENT COST
TERTIARY WASTE WATER TREATMENT-GRANULAR ACTIVATED CARBON
-91-
-------
FEED WATER
EFFLUENT
TWO-STAGE SYSTEM
FEED
WATER
I
12'
_L
t
12'
EFFLUENT
FOUR-STAGE SYSTEM
FIGURE 4
TWO-STAGE AND FOUR-STAGE SYSTEM DIAGRAM
-92-
-------
The higher capacity for the 4-stage case results in a lower
regeneration rate, smaller regeneration system and lower carbon loss.
A carbon regeneration rate of 350 Ib./hr. (same as base case) has
been assumed for all cases for the purpose of regeneration system
cost estimation., In the absence of a correlation of carbon capacity
(loading) with number of contacting stages, a calculation in reverse
order was attempted; i.e., an estimate of the approximate investment
was made and then, with the difference in operating costs from the
base case (2-stage) in hand, a back-calculation was made of the loading
which must be achieved in order to pay for the extra equipment.
Considered in this comparison are one-, two-, three- and four-
stage systems (Cases 20B, 1A, 20A and 20, respectively). Vessel re-
quirements for the four cases are as follows:
Case jStages Vessel Size
20 4 34-13' I.D. x 16'
20A 3 26-13' I.D. x 21'
1A 2 6-25'6" I.D. x 30'
20B 1 6-18' I.D. x 61'
There are four trains of equipment in the single-stage system
(even though the bed height-to-diameter ratio is much greater than one)
in order to keep the amount of "idle" carbon at 1/5 or less of the
total carbon inventory. (See section "Effect of Idle Carbon Inventory
on Economics").
Shown in Table 19 are the costs for the 4-stage, 3-stage and
1-stage systems exclusive of the costs for make-up carbon or fuel for
regeneration. The table shows that in the 4- and 3-stage cases (20
and 20A), the operating costs are already higher without make-up
carbon and fuel than for the 2-stage system. This means that no
matter what loading is attained in the 4- and 3-stage systems (as
long as the loading is finite), costs will be higher for those
systems than for a 2-stage system. When the single-stage system
is examined (Case 20B), it is seen that its cost is lower (without
make-up carbon and fuel included) than the 2-stage system. If it
is then assumed that the single-stage system could achieve the same
loading as the 2-stage system, the make-up carbon and fuel can be added
in at 1.09 and 0.15C/M gal., respectively, giving a total of 12.55c?
for 2-stage vs. 12.43C for single-stage, a slight difference. Of
course, since a single-stage system could not possibly obtain as high
a loading as a 2-stage system, the operating costs for the single-
stage system exceed the costs for a two-stage system when loading is
considered.
The conclusion is, therefore, that the two-stage contacting
system represents a true optimum number of contacting stages. No
further experimental work is needed to justify this conclusion.
Loading data are required, however, in order to design the regeneration
system for the two-stage system.
-93-
-------
TAIH.K 19
P-
Case No.
Variable Tested
Contactor Type
Velocity, cm/ft
Contact Time, Mln.
Particle Size, Mesh
Regeneration Loss, 7,
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
Investment. §M_
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon (3 26c/lb.
Fixed-Capital Investment
Operating Costs, C/M Gal.
Make-up Carbon (3 26c/lb.
Power @ lc/kwh
Backwash Water P 6c/M gal.
Fuel Gas @ 25C/MM Btu
Labor @ $3.50/Man-Hr.
Overhead @ 50% Labor + 1.485% PI/Yr.
Amortization
? 7.4% FCI/Yr.
Maintenance (3 5.757 PI/Yr.
Insurance @ 1% FCI/Yr.
Total Operating Cost
EFFECT OF
20
4-Stagu
Contactor
Downf low
7
50
8 x 30
5
Not fixed
13'I.D. x 16'
34
8
156
503
5
68
602
4
1382
2187
342
2529
_(2)
1.07
0.31
.(2)
1.05
1.42
5.13
3.45
0.69
13.12<2>
MT.BER OF COi.'TACTING STAKES ON
1'CA
3-St igc
Contactor
Down flow
7
50
8 x 30
5
Not fixed
13'I,D. x 21'
26
8
141
440
5
68
367
4
1067
1682
345
2027
.(2)
1.03
0.31
1.05
1.21
4.11
2.65
0.56
10.92^'
ECONOMICS
L\
2-Stas
-------
Comparison of Pomona, Tahoe, Improved Downflow and Moving-Bed Systems
Cases 5A» 18. 19 & 20
The purpose of these cases is to compare the Pomona and Lake
Tahoe systems on as consistent bases as possible. Improved versions of
the downflow and upflow systems with reasonable modifications for re-
ducing the costs of treatment are included.
t o
The system representing a scale-up of the Pomona pilot plant
from 0.3- to 10-MGD is presented as Case 20. This is the four-stage
contacting system described in the preceding section, with the make-up
carbon and fuel costs included and with 12 x 40 mesh carbon. The carbon
used at Pomona is 16 x 40 mesh which is not commercially available. The
capacity (0.522 Ib. COD/lb.) is the same as that at Pomora after several
regeneration cycles^. The contact time used in this case (50 minutes)
is higher than that actually available at Pomona (41 minutes for 38
feet of carbon at 7 GPM/ft.2). However, in order to keep the bed height-
to-diameter ratio at about one (see discussion in "Design Bases"), with
the total contact time of 41 minutes, the vessel diameter would have to
be 10 feet. This diameter would require 13 trains of vessels for a total
of 54 vessels (13 trains of 4 vessels each plus 2 spare vessels). In an
effort to minimize the investment, the bed depth in each vessel was in-
creased, so that 13 feet diameter vessels could be used, thus reducing
the number of trains to 8 for a total of 34 vessels.
19
The Tahoe system is represented by Case 18. An assumption made
in this case is that 15-
-------
No spare vessels have been included in Cases 18 and 19 since
these are moving-bed systems where only part of the carbon bed is
removed at a time for regeneration; at Tahoe, this has been done without
taking the vessel offstream. The vessel sizes for these four cases
is shown below:
Case Vessel Size
20 - Pomona System 34-13' I.D. x 16'
18 - Tahoe System 8 - 13' I.D. x 15'
5A - Improved downflow 6 - 21' I.D. x 30'
19 - Improved upflow 6 - 13' I.D. x 22'
The economics for these four cases are shown in Table 20.
By comparing Cases 20 and 18, the Tahoe system is seen to be
considerably less expensive than the Pomona system. The high costs
in the Pomona system are due to the four-stage contactor system. On
the other hand, when Cases 5A and 19 are compared, the two-stage down-
flow system compares favorably with the moving-bed system. As far as
technical feasibility is concerned, the downflow system is probably
on a much firmer experimental basis as are the assumptions which went
into its calculation. The assumption in the moving-bed case - that the
adsorption wave will be contained in the 15-minute contact time moving
bed - seems rather optimistic. Operating data from Tahoe show that
the COD concentration of 7 mg/L would not be achieved on the average^-?.
For the downflow case, it remains to be proven that the higher velocity
will reduce the required contact time.
Effect of In-Place Regeneration on Economics
Cases 1A & 22
The purpose of this comparison is to discover if there is any
economic incentive for developing a method for in-place regeneration.
In this comparison, the base case (Case 1A) with the conventional
multiple-hearth regeneration furnace is contrasted against a system
where regeneration of the carbon occurs in the vessel (Case 22).
For the in-place regeneration system, it was assumed that the
organic matter on the carbon would be cracked to lighter organic matter
and/or volatilized by the time the carbon bed is heated to 1000°F. with
1200°F. superheated steam. For a vessel containing 300,000 Ib. of
carbon and with the gas velocity set at 0.25 ft./sec., a 30-hour period
is required to heat the carbon to 1000°F. This calculation was made
using principles of unsteady - state heat transfer. The calculation
was repeated for a gas velocity of 1 fps with the result that the
regeneration period was reduced to 7.5 hours; but this would increase
-96-
-------
TAHLE 20
Case No.
Variable Tested
Contactor Type
Velocity, GPM/ft
Contact Time, Min.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, Ib COD/lb C
Vessel Size
Number of Vessels
Number of Trains
Investment, $M
I Concrete
^ Adsorbers & Filters
I Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment-
Carbon
Fixed-Capital Investment
Operating Costs, c/M Gal.
Make-up Carbon
Power ra lc/kwh
Backwash Water ^ 6c/M gal.
Fuel Gas (3 25C/MM Btu
Labor <<' $3.50/Man-Hr.
Overhead « 50% Labor + 1.4857, PI/Yr.
Amortization
-1 7.4% FCT/Yr.
Maintenance i.d 5.757, PT/Yr.
Insurance <' 17, FCI/Yr.
Total Operating Cost
20
Pomona
System
Down flow
7
50
12 x 40
10
0.522
13'I.D. x
34
156
503
5
68
44
602
4
1382
2187
349(
2536
16
.44^)
1.16
0.31
0.15
1.05
1.42
5.14
3.45
0.69
15.81
COMPARISON OF POMONA, TAilOE , IMPROVED
AND MOVING -BED SYSTEMS
18
Tahoe
System
Moving -Bed
7
15
8 x 30
5
0.25
i' 13'I.D. x 15'
8
8
139
211
8
1C7
44
313
4
826
1291
9;(2)
1368
1.C6
0 . ':, 1
O.il
1.05
1.05
2.81
2.03
o.:-s
10,46
DOWNFLOW
5A
Improved
Down f low
Downflow
10
35
8 x 30
2
0.522
21'I.D. x 30'
6
2
121
253
4
92
36
158
4
668
1101
265<2)
1366
0.97
0.31
0.15
1.05
0.97
2.77
1.73
0.37
8. 76
19
Improved
Moving-Bed
Moving-Bed
10
15
8 x 30
2
0.348
13'I.D. x 22'
6
6
139
194
4
107
36
262
4
746
1173
1277
0.44
1.06
0.31
0.15
1.05
1.00
2.59
1.85
0.35
8.80
(2)
(1) 12 x 40 Mesh carbon at 29<:/lb.
(2) 8 x 30 Mesh carbon at 26
-------
the size of the steam generator by a factor of four. Pittsburgh
Activated Carbon Company recommends that gas velocities be kept
between 0.25 and 1 fps so the 0.25 fps case was selected for the
estimate. The total amount of steam used for in-place regeneration
is 2/3 Ib. steam/lb. carbon - actually less than the amount of steam
required in the furnace operation. A nine-inch refractory brick lining
is provided to protect the vessel shell from this temperature. Also,
since there is no regeneration furnace where there must be a carefully
controlled combustion, the operating labor can be reduced from 1.25 to
0.25 men/shift. An afterburner is included on the effluent gas stream
for pollution control.
The cost comparison for the effect of in-place regeneration is
shown in Table 21. As seen in Table 21, the cost for in-place regeneration
is about Ic/M gal. less than the furnace system while the investment is
almost $300,000 higher. Reductions in the in-place regeneration operating
costs came about from savings in items related to labor(t.22c/M gal ) and
make-up carbon (1.09C/M gal.). Part of these savings are offset by*
higher investment charges.
In summary, there is incentive to develop a regeneration process
where the carbon loss by attrition and regeneration loss is reduced
and where there is no need for a closely supervised, controlled-combustion
furnace. A completely automated regeneration system would cost more, but
about $300,000 could be spent for controls before the investment charges
would offset the savings in labor.
-98-
-------
TABLE 21
EFFECT OF IN-PLACE REGENERATION ON ECONOMICS
Case No.
Variable Tested
Contactor Type .
Velocity, GPM/ft
Contact Time, Win.
Particle Size, Mesh
Regeneration Loss, %
Carbon Capacity, lb COD/lb C
Vessel Size
Number of Vessels
Number of Trains
1A
Furnace Regeneration
Downflow
7
50
8 x 30
5
0.522
25'6"I.D. x 30'
6
2
22
In-Place Regeneration
Downflow
7
50
8 x 30
0
0.522
25'6"I.D. x 30'
5
2
Investment^ $M
Concrete
Adsorbers
Tanks
Pumps
Special Equipment
Piping
Conveyors
Total Major Material
Plant Investment
Carbon @ 26c/lb.
Fixed-Capital Investment
Operating Costs, C/M Gal.
Make-up Carbon (? 26<;/lb.
Power
-------
CONCLUSIONS AND RECOMMENDATIONS
Based on economics and arguments given in the "Economics" Section,
the following conclusions and recommendations can be drawn. The
conclusions and recommendations are listed in the order of their im-
portance. Some of the recommendations are the same as presented in the
literature report1 and are now being carried out in FWPCA programs.
1. The gravity-flow system offers the savings of 2c/M gal. over
the downflow system and a savings of 3c/M gal. over the upflow system.
Gravity-flow is less expensive because concrete vessels are less expen-
sive than steel vessels. An assumption made for the gravity-flow case
and which would have to be proven, is that a 50-minute contact time will
give the same COD removal at 2 GPM/ft2 as at 7 GPM/ft2 and that the
carbon loading (capacity) will be the same for single- and two-stage con-
tactors. Recommendation: The performance of a single-stage, gravity-flow
contactor should be determined.
2. Under the same design conditions, the downflow system is less
expensive than the upflow system by over lc/M gal. Upflow is more expensive
because sand filters are required to remove suspended solids. Recommenda-
tion: No new data are needed to confirm this conclusion.
3. A two-stage contactor is an optimum pressurized vessel system
as compared to single-, three- and four-stage contactors. The costs are
seen to be higher for 3- and 4-stages due to increased vessel and piping
costs than for 2-stages, no matter what carbon capacity is assumed. When
capacity is considered in the single-stage system, this system is also
more expensive than the two-stage system. Recommendation: A two-stage,
downflow contacting system containing 8 x 30 mesh carbon should be
operated at a large scale for extended periods in order to determine
degree of removal and carbon capacity.
4. The comparison between the Pomona and Tahoe systems as reported
in the literature*^,19 finds the Tahoe system to be substantially less
costly (about 5c/M gal.). Pomona is more expensive due to the four-stage
system used. An assumption made for the Tahoe moving-bed system is that
the 15-minute contact time will contain the adsorption wave which is
doubtful based on Tahoe operating data. Recommendation: No new data are
needed to confirm this conclusion.
5. When reasonable improvements are made to the downflow (Pomona-
type) and moving-bed (Tahoe-type) systems, the costs are about equal.
However, the downflow system is felt to be on a firmer experimental basis.
Recommendation: No additional work is needed on upflow systems.
-100-
-------
6. Substantial savings can be realized if a less expensive
adsorbent can be found. A reduction in adsorbent cost of 15c/lb
would reduce carbon treating costs by about 1.15C/M gal. Also, up
to a new price of about 2.0 that regeneration losses can be minimized.
8. A high carbon capacity for COD is to be preferred in order to
reduce costs associated with the regeneration system. Operating costs
can be cut by 2$/M gal. when the capacity is increased from 0.25 Ib
COD/lb C (Tahoe) to 0.87 Ib COD/lb (Pomona first regeneration cycle).
Recommendation: Schemes should be examined which would increase the
allowable loading (capacity) of the carbon. This might include en-
hancement of the biological action by injecting air (oxygen) into the
wastewater.
9. Costs would be significantly lower if the closely controlled
combustion in the regeneration furnace could be eliminated. If a
highly automated furnace operation could be accomplished, a savings in
labor cost of 1.2<;/M gal. is possible. In-place regeneration offers
the same advantages plus the possibility of reducing the attrition and
regeneration losses. Recommendation: Complete automatic control of the
regeneration furnace or elimination of the furnace regeneration system
should be investigated.
10. Handling of surges in feed flow rate with a surge basin rather
than with extra vessels is less expensive by nearly 1.3<:/M gal. with a
considerable reduction in investment. Recommendation: No new data
are required to confirm this conclusion.
11. Even though costs are not significantly different between shop
fabrication and field erection of vessels, the field erection case offers
a simpler flow sheet. Recommendation: No new data are required to confirm
this conclusion.
12. Costs are not affected substantially if the "idle" carbon inventory
is between 10% and 20% of total carbon inventory. Recommendation; No
new data are needed to confirm this conclusion.
-101-
-------
13. Substantial economics of scale are seen when the plant size
is increased from 1- to 10-MGD. This reduction is on the order of
200/M gal. due mainly to reduced labor cost. When the plant size is
increased from 10- to 100-MGD, the operating cost is reduced by only
3.5C/M gal. The reduction is smaller at the large plant size because
there are several trains of vessels required. Recommendation: No
new data are required.
14. If it is assumed that particle size affects the regeneration
loss and required contact time, then the case with a larger particle
size, longer contact time and lower regeneration loss is less costly.
Reductions of up to 2c/M gal. are possible when 8 x 30 mesh carbon is
used rather than 12 x 40 mesh. If particle size affects only the
required contact time, then there is no difference in costs between
8 x 30 and 12 x 40 mesh carbon. Recommendation: The functionality
between carbon particle size and regeneration loss should be determined.
15. When the required contact time is assumed to decrease with
increasing velocity, then the high velocity-low contact time case is
less costly by almost 3c/M gal. Recommendation: The relation between
velocity and required contact time should be determined.
16. When it is assumed that velocity does not affect the required
contact time, then the 7 and 10 GPM/ft2 cases are not significantly
different in cost. However, the 4 GPM/ft^ case is more expensive by
1.7C/M gal. Recommendation: No new data are needed to confirm this
conclusion.
17. There is little difference in operating costs and no signifi-
cant difference in investment when just the particle size is changed from
8 x 30 to 12 x 40 mesh carbon. If changing the particle size over this
range affects only the pressure drop and carbon cost, then either size
carbon could be used without any penalty in costs. Recommendation: No
new data are required to confirm this conclusion.
-102-
-------
REFERENCES
1. Cover, A. E. and Pieroni, L. J., "Phase I. Evaluation of the
Literature on the Use of Granular Activated Carbon for Tertiary
Water Treatment," this publication.
2. Ibid., Figure 16.
3. Allen, J. B., Clapham, T. M., Joyce, R. S., and Sukenik, V.A.,
"Use of Granular Regenerable Carbon for Treatment of Secondary
Effluent-Engineering Design and Economic Evaluation," Un-
published report from Pittsburgh Activated Carbon Company to
USPHS, October 1, 1964.
4. Cooper, J. C. and Hager, D. G., "Water Reclamation with Activated
Carbon," Chem. Engr. Progr., 62_ (10) 87 (October 1966).
5. Cover, op. cit., Figures 1 and 23.
6. Smith, C. E. and Chapman R. L., "Recovery of Coagulant, Nitrogen
Removal, and Carbon Regeneration in Waste Water Reclamation,"
Report from South Tahoe Public Utility District to FWPCA, June
1967, p. 30.
7. McCabe, W. L. and Smith, J. C., "Unit Operations of Chemical
Engineering," McGraw-Hill Book Company, Inc., New York (1956),
pp. 270, 276.
8. Cover, op. cit., Figures 3 to 10.
9. Ibid., Figure 12.
10. Ibid., Figures 8 and 9.
11. Ibid., P. 28.
12. Ibid., Figures 14 to 19.
13. Ibid., Figure 16.
14. Ibid., Figure 23.
15. Ibid., Figure 1.
16. Ibid., p. 7.
17. Smith, op. cit., pp. 24-25.
-103-
-------
18. Parkhurst, J.D., Dryden, F. D., McDermott, G.N. and English, J.,
"Pomona Activated Carbon Pilot Plant," JWPCF, 39 (10) Part 2,
R 70-81 (October 1967).
19. Smith, C.E. and Chapman, R.L., "Recovery of Coagulant, Nitrogen
Removal, and Carbon Regeneration in Waste Water Reclamation",
Report from South Tahoe Public Utility District to FWPCA,
(June 1967).
-104-
-------
APPENDIX A
PROCEDURE FOR CALCULATION OF OPERATING COSTS
The following is the basis and rationale used in our recommendation
for calculation of the operating costs for tertiary waste water treatment
using granular activated carbon. Economic comparisons, so long as they
are on consistent bases, should be valid no matter what particular per-
centages for maintenance and overhead items are chosen for comparison
purposes. No other comparisons should be made unless care is taken to
make sure that the cases in question are on consistent bases.
It is felt that the activated carbon system is more like a typical
chemical plant than a water or sewage treatment plant as the activated
carbon system has some corrosion problems and has a furnace where
the combustion must be carefully controlled. Operating labor, for example,
is based on experience in a clay treating plant for taste, odor and color
removal from hydrocarbons (oils and waxes) which Kellogg has designed and
built. The activated carbon system is quite similar to this process.
The recommended procedure is based on several sources among which
are the Office of Saline Water, the Office of Coal Research as well as
well-known texts on engineering economics. A more precise estimation
of the maintenance and overhead charges can be determined only from
actual costs obtained from large-scale plants. Maintenance charges for
individual pieces of equipment can be more accurately estimated from an
extensive examination of literature and vendor recommendations and,
especially, plant operating records. However, maintenance costs are
quite variable depending on operating personnel quality and amount of
inspection and preventive maintenance. In the activated carbon process,
maintenance costs could be quite high if a more corrosion-resistant
lining for the vessels is not found. At the current state of knowledge
of this process, we feel that these recommendations represent the most
realistic estimate of the operating costs and are consistent with chemical
industry and utility practice.
A review has been made of several different general procedures for
calculation of maintenance and overhead costs of operating a chemical
plant. Among the sources examined are the Office of Saline Water^
Office of Coal Research2, Kellogg internal practice3, Aries and Newton4,
and Peters5. The following are our recommendations for calculation of
the operating costs for our economic evaluation.
Maintenance Labor
Maintenance Materials
Supplies
Payroll Overhead
General & Administrative
Overhead
Amortization
Insurance
Taxes
3% Plant Investment/year
27= Plant Investment/year
- 15% of Maintenance Labor & Materials
- 157, of Operating plus Maintenance
Labor
307o of Operating plus Maintenance
Labor plus Payroll Overhead
- 7.47o Fixed-Capital Investment/year
(20-year life with 47= interest/year)
- 1% Fixed-Capital Investment/year
- None (normally 27= Fixed-Capital In-
vestment/year)
-105-
-------
The first three items equal 5.7570 of Plant Investment/year while the
next two items total to 507= of Operating Labor plus 1.485% of Plant
Investment/year. The total of the items which are dependent upon the
Plant Investment is 15.6%/year while that dependent on Carbon Investment
only is an additional 8.4%/year.
Table A shows the comparison of the various calculation procedures
and, for illustration, gives the operating cost for Case 1A. The operating
cost can vary from 8.22c/M gal. to 12.64 c/M gal., depending on the
method used. The maintenance, supplies and labor costs are somewhat higher
than in conventional waste water treating (primary and secondary) but are
nearer to costs experienced in typical chemical plants which carbon treating
more closely resembles. On the other hand, the overhead items are typical
for the type of governmental agency which would be operating this plant.
A more accurate estimation of the maintenance and overhead costs will be
available only after a sustained operation of some of the large-scale
units such as at Lake Tahoe or Piscataway, Maryland.
The items labeled supplies, payroll overhead and general and adminis-
trative overhead include the following:
% Supplies include instrument charts, lubricants, janitor supplies,
test chemicals, and similar supplies which cannot be considered
as raw or maintenance materials.
Q Payroll overhead includes pensions, paid vacations, group in-
surance, disability pay, social security, and unemployment taxes.
^General and administrative overhead includes costs for supervision,
medical services, general plant maintenance and overhead, safety
services, packaging, restaurant and recreation facilities, salvage
services, control laboratories, property protection, plant
superintendence, warehouse and storage facilities, special employee
benefits, purchasing, engineering, executive and clerical wages,
office supplies, communications, advertising and consultant fees.
Other operating costs such as power, fuel gas, and backwash water
are charged at unit costs which are typical for these items. The back-
wash water is taken from the secondary treatment plant and returned to
the intake for primary treatment. The unit cost for producing secondary
effluent varies from 3C to 9C/1000 gallons depending on plant size.
A sufficient initial charge of activated carbon must be bought at
the time that the plant is buil to fill the vessels. This carbon wears
out, not unlike the vessels which contain it, and therefore, must be
depreciated. For cost accounting purposes, the initial charge of carbon
is amortized over the same period as the plant. The initial charge of
carbon eventually gets entirely replaced by the make-up carbon which is
charged as an operating cost. The inventory of make-up carbon is not
included in the Fixed-Capital Investment since it is bought a little
at a time and can be sold back to the supplier if not used. No salvage
value has been assumed for the used carbon which remains at the end of
.-106-
-------
TABLE A
I
I-1
o
I
METHOD
ITEM RECOMMENDATION
a. Operating Labor
b. Maintenance Labor
c. Maintenance Materials
d. Supplies
e. Payroll Overhead
f . Supervision
g. Laboratory Labor
h. General Overhead
i. Plant Insurance
j . Taxes
CA»* 1A OPERATING COSTS*
Make-up Carbon @ 26C/lb
Power 3 K/kwh
Backwash Water £ 6C/M gal.
Fuel Gas @ 25C/MM Btu
Amortization 0 7.4» PCI/yr
Plant Insurance £ 11 FCI/yr
Taxes - None
a. Operating Labor 9 $3. 50/man-hour
b. Maintenance Labor
c. Maintenance Materials
d. Supplies
e. Payroll Overhead
f . Supervision
g. Laboratory Labor
h. General Overhead
TOTAL OPERATING COST, t/M gal
* Case 1A
Plant Investment (PI) 1
Carbon § 26C/lb
osw">
__
0.5% Pl/yr
T- 0.5% Pl/yr
15% la + b)
~llO% la + b + e)
1% FCI/yr
1% FCI/yr
1.09
1.00
0.31
0.15
3.24
0.44
__
1.05
0.17
^j-0.17
0.18
(-0.42
8.22
,210,000
390,000
COMPARISON
OCR'"
__
3% Pl/yr
J- 15% (b)
10% (a + f)
.10% (a)
150% (a + b + c
J" + d + f)
T- 3% FCI/yr
1.09
i.oo
0.31
0.15
3.24
0.44
_
1.05
0.99
J-0.15
0.12
0.11
9.80
OF PROCEDURES FOR OPERATING COST
MWK TO AEC(3)
__
2.8i Pl/yr
1.2% Pl/yr
15% (b * c)
20% (a + b + £)
15% (a)
"1 50% (a + b + c
_T + d + f)
~T- 3% FCI/yr
1.09
1.00
0.31
0.15
3.24
0.44
__
1.05
0.93
0.40
0.20
0.43
0.16
}2.74
12.14
ARIES i
NEWTON1''1
3% Pl/yr
3% Pl/yr
15% (b + c)
15-20% la)
10-25* (a)
1-50-100% (a)
1% FCI/yr
1-2% FCI/yr
1.09
1.00
0.31
0.15
3.24
0.44
__
1.05
0.99
0.99
0.30
0.18
0.18
T 0.79
10.67
PETERS15'
__
3% Pl/y
4% Pl/y
10% (b * )
w/(h)
10-201 ( )
"165-85% (a b
_T + c f)
0.4-1% FCI/yr
1-4% FCI/yr
1.09
1.00
0.31
0.15
3.24
0.44
_
1.05
0.99
1.33
0.23
w/ovhd
0.16
J- 2.65
12.64
CALCULATION
POMONA10
1 nan-day/day
2/3 tnan-day/day
2% Pl/yr
25% (c)
25% (a tb-f f + g)
5% (a + b + 9)
1/3 man-day/day
}15% (a + b + (
+g)
nona
1.09
1.00,,
0.05
0.15
3.24
w/ovhd
__
0.84
0.56
0.66
0.17
0.42
0.08
0.28
0.25
8.79
MHK
RECOMMENDATION
1-1/4 men/.hift
3% Pl/yr
2% Pl/yr
15% (b + o)
15% (a + b)
[-30% (a + b + )
1% FCI/yr
non«
1.09
1.00
0.31
0.15
3.24
0.44
_
1.05
0.99
0.66
0.25
0.31
(- 0.71
10.20
METHOD
ITEM RECOMMENDATION
a. Operating Labor
b. Maintenance Labor
c. Maintenance Materials
d. Supplies
e. Payroll Overhead
f. Supervision
g. Laboratory Labor
h. General Overhead
i. Plant Insurance
j. Taxes
CASE 1A OPERATING COSTS*
Make-up Carbon g 26$/lb
Power 9 l$/kwh
Backwash Water % 6C/M gal.
Fuel Gae e 25C/MM Btu
Amortization § 7.4% FCI/yr
Plant Insurance ? 1% PCI/yr
Taxes - None
a. Operating Labor f? S3.50/man-houz
b. Maintenance Labor
c. Maintenance Materials
d. Supplies
e. Payroll Overhead
f. Supervision
9. Laboratory Labor
h. General Overhead
Fixed Capital Invatrtment (FCI) 1,600,000
Backwash Water 9 K/K gal.
-------
the plant's life even though the carbon could be used in another waste
treatment plant. It is felt that the used carbon could not be sold back
to the supplier or to anyone else at the virgin carbon cost.
The cost estimates in this report have been made on the basis of a
Gulf Coast, U.S.A. plant location. Included in the plant investment
estimate are such items as foundations, concrete structures, vessels,
tanks, drums, structural steel, pumps, compressors, piping, wiring,
switchgear, lighting, instruments, paint, insulation, conveyors, cranes,
special equipment (regeneration furnace system, eductors, water softener,
air pollution control system, steam generator) and construction labor.
Also included in the plant investment, are non-material items such as
all-risk insurance (liability, accident and loss insurance during con-
struction), field office administration, supervision and expenses, home
office procurement, engineering and scheduling, and contractor's fee.
Not included in the plant investment are working capital, interest
charges during construction, local taxes, freight and duties. Also, land
has not been included in the investment, especially since even a 100-MGD
gravity-flow plant would occupy only about two acres which at $30,000
an acre is still a very small percentage of the total cost (about 170).
Ordinarily, the cost of land is not included in the investment since
it is the responsibility of the owner and not the contractor to buy the
land for the plant site.
-108-
-------
REFERENCES FOR APPENDIX A
1. Office of Saline Water, "A Standardized Procedure for Estimating
Costs of Saline Water Conversion," March, 1956.
2. The M. W, Kellogg Company, Report RED-68-1173 to the Office of
Coal Research, Contract No. 14-01-0001-380, September 1, 1968, p. 123.
3. The M. W. Kellogg Co., Report RD-62-952 to the U.S. Atomic Energy
Commission, Contract No. AT (30-1)-3009 (NYO 10,301), November 30,
1962, p. 167.
4. Aries, R« S. and R. D. Newton, "Chemical Engineering Cost Estimation,"
McGraw-Hill Book Company, Inc., New York (1955) pp. 162-182.
5- Peters, M, S. "Plant Design and Economics for Chemical Engineers,"
McGraw-Hill Book Company, Inc., New York (1958) pp. 108-9, 113.
6. Letter from C. W. Carry, Los Angeles County Sanitation Districts
to A. N. Masse, Federal Water Pollution Control Administration,
October 31, 1968.
-109-
------- |