WATER POLLUTION CONTROL RESEARCH SERIES • ORD- 17O2OFBDO3/7O
"THE DEVELOPMENT OF
A FLUIDIZED-BED TECHNIQUE FOR THE
REGENERATION OF POWDERED ACTIVATED CARBON"
U.S. DEPARTMENT OF THE INTERIOR • FEDERAL WATER QUALITY ADMINISTRATION
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe
the results and progress in the control and abatement
of pollution of our Nation’s waters. They provide a
central source of information on the research, develop-
ment and demonstration activities of the Federal Water
Quality Administration, Department of the Interior,
through in—house research and grants and contracts with
Federal, State, and local agencies, research institutions,
and industrial organizations.
Water Pollution Control Research Reports will be distrib-
uted to requesters as supplies permit. Requests should be
sent to the Planning and Resources Office, Office of
Research and Development, Federal Water Quality
Administration, Department of the Interior, Washington,
D. C. 20242.
-------
THE DEVELOPMENT OF A FLUIDIZED-BED TECHNIQUE FOR
THE REGENERATION OP POWDERED ACTIVATED CARBON
by
Battelle Memorial Institute
Columbus Laboratories
Columbus, Ohio 43201
for the
FEDERAL WATER QUALITY ADMINISTRATION
DEPARTMENT OF THE INTERIOR
Program #17020 FED
Contract #14-12-113
FWQA Project Officer, E. L. Berg
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio
March, 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price 55 cents
-------
FWQA Review Notice
This report has been reviewed by the Federal
Water Quality Administration and approved for
publication. Approval does not signify that
the contents necessarily reflect the views
and policies of the Federal Water Quality
Administration, nor does mention of trade
names or commercial products constitute
endorsement or recommendation for use.
-------
ABSTRACT
This report describes the results of research conducted at the Columbus
Laboratories of Battelle Memorial Institute on the regeneration of spent
powdered carbon. The study was directed toward the development of a
fluidized—bed regeneration technique. Two fluidized—bed systems were
considered during the course of the investigation: a system in which the
dried spent carbon was regenerated during its passage through a fluidized
bed of an inert material; and a pulsating fluidized—bed system in which
the finely divided regenerated carbon served as the bed material. Both
techniques were effective in restoring the spent powdered carbon to over
90 percent of its original adsorptive capacity. Recoveries in excess of
80 percent of the weight of the dried spent carbon were attained. How-
ever, because of its higher unit capacity, the fluidized inert bed system
was selected for subsequent larger scale development.
On the basis of favorable results in the laboratory, a pilot—scale unit,
10—inch—ID, was designed and constructed to process 30 pounds of spent
carbon in an 8—hour period. Operation of this system was integrated
into the main powdered carbon adsorption process for evaluation of the
combined system performance. The results of this operation covering a
period of 23 days indicated that regenerated carbon was as effective as
virgin carbon for organic removal from secondary effluent for at least
3.6 cycles through the system. Overall carbon losses averaged 15 per-
cent per regeneration cycle.
A preliminary economic analysis indicated plant operating costs would be
about 0.9 to l.lç per pound of carbon for a commercial plant producing
20,000 pounds of regenerated carbon per day.
This report was submitted in fulfillment of Contract No. 14—12—113,
Program No. 17020 FBD, between the Federal Water Quality Administration
and Battelle Memorial Institute, Columbus Laboratories.
Key Words: Activated carbon
Regeneration
Fluidized bed
11
-------
CONTENTS
INTRODUCTION 1
EXPERIMENTAL 5
Bench—Scale Studies 5
Description of Sample 5
Analytical Methods Employed 5
Apparatus and Procedure 7
Fluidized Inert Bed System 7
Pulsed Fluidized—Bed System 7
Results and Discussion 11
Effect of Temperature and Gas Composition 14
Effect of Retention Time 14
Effect of Increased Moisture Content in Feed 17
Losses of Carbon During Regeneration 18
Comparison of Fluidization Techniques 20
Design, Construction, and Operation of a Pilot—Plant Facility 21
Design 21
Construction 23
Operation 26
Procedure 26
Results and Discussion 26
Problems Encountered 33
DESIGN AND ECONOMIC CONSIDERATIONS 37
ACKNOWLEDGMENT 43
11],
-------
FIGURES
Page
1 Powdered carbon adsorption system 2
2 Relationship between total carbon content and ultraviolet
absorbence of treated and untreated secondary sewage effluent 6
3 Bench scale fluidized bed and auxiliary equipment 8
4 Photograph of complete bench—scale installation 9
5 Sketch of pulsed fluidized—bed apparatus 10
6 Effect of temperature on adsorption capacity in various
atmospheres using inert bed system 15
7 Effect of temperature on recovery in various atmospheres
using inert bed system 16
8 Effect of moisture content of spent carbon on the effective-
ness of its regeneration at 1250 F 19
9 Heat and materials flow diagram for pilot regeneration
system 22
10 Detailed drawing of the 10—inch—diameter fluidized—bed unit 24
11 Arrangement of equipment for pilot—plant regeneration study 25
12 Total organic carbon removal using regenerated powdered
carbon 31
13 Turbidity removal from secondary effluent using regenerated
powdered carbon 32
14 Fixed carbon, ash, and volatile content of regenerated
carbon 35
15 Flow diagram of commercial regeneration system, employing
conventional operation, example I 38
16 Flow diagram of mmercia1 regeneration system employing
offgas recycle, example II 39
17 Detailed drawing of commercial fluidized—bed regeneration
unit 41
iv
-------
TABLES
No, Page
1 SU1 1MARY OF REGENERATION EXPERIMENTS IN FLUIDIZED INERT BED 12
SYSTEM
2 SU1 ARY OF REGENERATION EXPERIMENTS IN PULSED FLUIDIZED—BED 13
SYSTEM
3 EXPERIMENTAL DATA SHOWING THE EFFECT OF BED WEIGHT ON 17
REGENERATION, INERT BED SYST M
4 DATA ON COMPOSITION OF VIRGIN, SPENT, AND REGENERATED CARBONS 20
5 EXPERIMENTAL DATA SHOWING CARBON LOSSES DURING SEVERAL
REGENERATION EXPERIMENTS 20
6 MISCELLANEOUS DESIGN DATA 21
7 CONDITIONS EMPLOYED DURING PILOT—PLANT START—UP OPERATIONS 26
8 RESULTS OBTAINED DURING PILOT—PLANT START—UP OPERATIONS 27
9 AVERAGE OPERATING PARAMETERS DURING REGENERATION RUN 29
10 AVERAGE TOTAL ORGANIC CARBON (TOC) VALUES AND REMOVAL
EFFICIENCY FOR EACH REGENERATION CYCLE 30
11 COMPARISON OF AVERAGE TOTAL ORGANIC CARBON REMOVAL BY VIRGIN
AND REGENERATED CARBON 33
12 DAILY REGENERATED CARBON ANALYSES AND RECOVERY 34
13 CONDITIONS AND RESULTS OF EXAMPLE DESIGN CALCULATIONS FOR
COMMERCIAL REGENERATION SYSTEM 37
14 CAPITAL COSTS FOR COMMERCIAL REGENERATION SYSTEM 40
15 OPERATING COSTS FOR COMMERCIAL REGENERATION SYSTEM 42
V
-------
CONCLUSIONS AND RECOMMENDATIONS
A study of the regeneration of spent powdered carbon by fluidized—bed
techniques has been completed. The study was concerned with an evalua-
tion of bench—scale and pilot—scale systems and an economic assessment
of a commercial system based on appropriate scale—up factors.
The results obtained during the study showed that efficient regeneration
and recovery of spent powdered carbon can be achieved in a fluidized—bed
system. Under proper operating conditions, the spent carbon could be
regenerated to an active form as effective as virgin activated carbon in
its ability to adsorb organic components from a typical secondary sewage
effluent. Recovery of the regenerated carbon was about 85 percent per
regeneration cycle.
On the basis of initial bench—scale studies, the following major conclu-
sions were drawn:
(1) A system utilizing an inert bed of fluidized solids through which
the fine carbon is passed or a system employing pulsation of the fine
carbon solids are equally effective from a technical standpoint.
(2) A temperature between 1000 and 1500 F and a gas atmosphere contain-
ing nitrogen, oxygen, carbon dioxide, and water vapor are most effective
for efficient regeneration of the spent carbon.
(3) Temperature is a primary variable; raising the temperature increases
both the adsorptive capacity and the weight losses of carbon during
processing.
(4) Oxygen content is also a primary variable and should be held to a
minimwii to reduce carbon losses through combustion.
(5) From a practical standpoint, the fluidized inert bed system is the
most feasible because of higher unit capacity when processing a relatively
wet spent carbon feed.
On the basis of a pilot—scale investigation of the process, it was con-
cluded that:
(1) After 3.6 cycles of adsorption and regeneration, the regenerated carbon
is almost as effective as virgin carbon in removing total inorganic mate-
rials from secondary sewage efflu.ent.
(2) Average carbon losses per regeneration cycle can be expected to be
less than 15 percent in a continuously operated system.
(3) The overall physical performance of the fluidized—bed regeneration
unit was excellent.
vi
-------
Generally, it was concluded that the overall study was quite successful
with the development of a satisfactory method for regeneration of spent
powdered carbon. It therefore was recommended that the development of
the process be continued on a larger scale. It also is recommended that
iimnediate emphasis be placed on the continued economic appraisal of the
process particularly on the economic advantages of powdered carbon com-
pared to granular carbon systems.
This study has indicated that powdered carbon can be regenerated for
about 0.9 to l.l per pound in a commercial system having a capacity of
20,000 pounds per day. Capital cost estimates for such a system range
from $350,000 to $435,000.
vii
-------
INTRODUCTION
The use of powdered activated carbon for the tertiary treatment of secon-
dary sewage effluents is being studied on a lO—gpm scale at Lebanon,
Ohio, by the Advanced Waste Treatment Research Activities of the Federal
Water Pollution Control Administration. A schematic diagram of the
process is shown in Figure 1. FWPCA’s pilot—plant study has shown that
powdered carbon is as effective as granular activated carbon for removing
the organic impurities from the wastewater.
Before powdered carbon can be used commercially for the tertiary treat-
ment of sewage effluents, an economical method of regeneration must be
developed. Investigations have been conducted on the thermal regenera-
tion of spent carbon but neither the optimum process conditions nor
operating procedures have been completely delineated.
In order to continue the development of a suitable regeneration system,
Battelle Memorial Institute recommended a research program to examine
the feasibility of applying fluidized—bed techniques for the regeneration
of the spent carbon. The use of the fluidized bed for regeneration can
offer the key advantages of excellent temperature and atmosphere control
and the ability to process the powdered solids conveniently and continu-
ously. However, the median diameter of the carbon particles is approxi-
mately 11 microns, which is considerably finer than normally used in
fluidized—bed operations. The problems associated with fluidization of
very fine powders are the inability to achieve proper fluidization and
the high entrainment losses. Thus, the desired control of temperature
and retention time is not achieved.
The application of fluidization methods for regeneration of powdered car-
bon therefore required the development of operating methods for proper
fluidization of the fine powder or the development of an alternative
procedure. Two techniques were recommended by Battelle to have suffi-
cient merit for study:
(1) The use of a fluidized bed of coarse, inert particles which main-
tains a constant—temperature zone. The spent, dried, powdered carbon
would be fed into the bottom of the coarse, inert bed and carried through
the bed by the action of the fluidizing gas. This method of fluidized—
bed operation would be expected to offer control of retention time of
the fine carbon powder and show good heat—transfer characteristics. The
finely divided carbon would be recovered from the effluent gas stream
with cyclone collectors or some other collection device.
(2) The use of a fluidized bed of the powdered carbon to which vibration
or pulsation is applied. This method would be expected to offer a minimum
of fluidizing gas requirements and entrainment losses while achieving the
excellent heat—transfer rates and temperature control afforded by the
fluidized bed.
1
-------
FIRST STAGE
TREATMENT
10 gpmFi
SECONDARY EFFLUENT
FROM CONVENTIONAL
TREATMENT PANT
CARBON
CONCENTRATION
IOQTO 300 mg/I
U I
(nO
TO
RE GE N ER ATI ON
FURNACE
iI
DUAL
MEDIA
FILTER
VIRGIN OR
REGENERATED
CARSON
FEED TANK
NO. 2
FLOC
TANK
NO. 2
CONTACT
TANK
Figure I. Powdered carbon adsorption system
-------
The proposed research program recommended by Battelle consisted of three
phases of work:
(1) A bench—scale study of each of the two fluidized—bed techniques
described above;
(2) The design, construction, and operation of a pilot—scale fluidized—
bed unit;
(3) The preliminary design and economic assessment of a commercial
fluidized—bed system.
Phase 1 of the program was initiated in July, 1967, and completed in
April, 1968. A report summarizing this work was submitted Nay 10, 1968.
On the. basis of the initial study, the subsequent phases of work were,
approved and completed in February, 1970. This report summarized the
major results and conclusions of the overall program.
3
-------
EXPERIMENTAL
Bench—Scale Studies
Description of Sample
The spent carbon used as feed material during the initial experimentation
was obtained from FWPCA’s waste treatment pilot plant located at Lebanon,
Ohio. The material was received as a slurry containing about 15 percent
solids. The initial sample, 55 gallons, which was used for the majority
of experiments, was pressure filtered and the solids dried at 90 to 100 C.
Additional preparation of the dried solids included screening and ball
milling for several minutes to break up agglomerates. A second spent
carbon sample of approximately 10 gallons was used in several subsequent
experiments designed to investigate the effect of high levels of moisture
in the feed on regeneration characteristics. This sample also was pres-
sure filtered but the filter cake was not oven dried. The moisture
content of the sample was decreased to about 50 percent by air drying
overnight. No attempt was made to eliminate agglomerates from this
material.
Analytical Methods Employed
The technique used for determining the adsorptive capacity of the carbon
samples was an empirical method based on ultraviolet (UV) absorbance
measurement of secondary effluent samples treated with 200 mg per liter
of carbon. The UV absorbance data before and after treatment with the
carbon were used to calculate the relative adsorptive capacity of the
regenerated carbon compared to that of virgin carbon. These results
provided a measure of the degree of regeneration achieved during the
experiments. No attempt was made, however, to relate these values to
the organic loading capacity of the regenerated material. The procedure
used in the analyses was as follows: the secondary effluent was treated
with 200 mg per liter of carbon sample; after 30 minutes ’ contact time,
the suspension was filtered through glass filter paper; the original
secondary effluent after filtration and the clear filtrate were then
measured by a Beckman DU spectrophotometer.
To provide additional data on the validity of the above method, several
samples of treated and untreated secondary effluent also were analyzed
for total carbon content with a Beckman carbon analyzer. The results
obtained with the two analytical techniques are compared in Figure 2.
The relationship obtained, although not entirely linear, shows a direct
correspondence between the two methods of analyses. To provide a single
basis for correlating the experimental data, all results were based on
the ultraviolet absorbance method.
5
-------
I .L&A.)
0.800
Li
0
z
Li
a)
i 0.600
0
(J)
QD
F-
Li
-J
2 o.4oo
li
0.200
0o
TOTAL CARBON, mg/I
Figure 2. Relationship between total carbon content and ultraviolet absorbence
of treated and untreated secondary sewage effluent
0
0 TREATED
0 UNTREATED
4
6 8 tO 12
6
-------
Apparatus and Procedure
Fluidized Inert Bed System . The major portion of the experimental pro-
gram consisted of evaluating the first proposed fluidization technique
in which a bed of coarse inert solids was used to provide a suitable
regeneration environment. The experimental unit which was used in this
part of the program comprised a 4—1/2—in. —ID by 24—in, long stainless
steel vessel heated by an electric resistance furnace. For initial
experimentation, a Thermal Dynamics fine powder feeder was used to
introduce dried spent carbon into the fluidizing gas line. Regenerated
products were collected in a cyclone dust collector followed by a small
absolute filter. A sketch of the fluidized inert bed unit, together
with some of the auxiliary equipment is shown in Figure 3. Figure 4 is
a photograph of the complete installation.
During the experimentation, several modifications associated with the
feed equipment and product collection train were made. To accommodate
increased feed rates, two cyclones were installed in series and the
absolute filter was replaced with either a porous stainless steel filter,
2.0—in. —.O.D. by 3—ft. —long, or a 6—in. —0.D. column packed with glass
wool. A water scrubber also was tried, however, excessive pressure drop
and inefficient collection was experienced and the use of the scrubber
was discontinued.
For those experiments in which spent carbon containing up to 50 percent
moisture was used, a vibrating—type feeder manufactured by Vibra Screw
Feeders, Inc., was employed. In these experiments, the spent carbon was
introduced via an overhead feed tube which extended vertically into the
bed.
The general procedure for making an experimental run was to charge the
reactor with bed material and heat the unit to the desired operating
temperature. In most cases, the bed comprised about 3300 grams of minus
35 plus 65 mesh sand or minus 20 plus 48 mesh flint shot. Spent carbon
was fed into the unit and the regenerated products were collected in the
cyclone and filter devices. After completion of a run, a material bal-
ance was made and samples of the various products were analyzed for
adsorptive capacity.
Pulsed Fluidized—Bed System . The experimental unit which was used for a
brief study on pulsation, the second technique proposed, was a specially
designed, 4—in.—diameter unit also constructed of stainless steel. This
unit contained a porous stainless steel gas distributor whereas the former
unit contained a conical bottom section. The auxiliary pieces of apparatus
including the cyclone and absolute filter dust collectors, the screw
feeder, and the furnace equipment were identical to those used in the
previous work. Figure 5 is a sketch showing the arrangement of major
components of apparatus.
7
-------
TO
ATMOSPHERE
Figure 3. Bench scale fluidized bed and auxiliary equipment
SCREW
POWDER FEEDER
ABSOLUTE
NO. I
NO. 2
CYCLONE DUST
COLLECTOR
PRODUCT CANNISTER
FLU IDIZED- BED
REACTOR TUBE
PRESSURIZED
FEED HOPPER
FLUIDIZED BED
BED DISCHARGE
VIBRATING TABLE
FOR BED DISCHARGE
F LU I DIZ I NG
GASES
8
-------
I
I ,
\0
9
Photograph of complete bench—scale installation
Figure 4.
-------
FEED HOPPER
FILTER
CYCLONE
COLLECTOR
POROUS GAS
DISTRIBUTOR
IIOv A.C.
II
I,
VARIABLE -SPEED
CAM ACTUATOR
FLUIDIZING
GAS
FLUIDIZED BED
DISCHARGE PRODUCT
C AN N ISTER
TO
ATMOSPHERE
SCREW
FURNACE
RE ACTOR
FEED TUBE•
RESISTANCE ELEMENTS
ROTA METER
Figure 5. Sketch of pulsed fluidized-bed apparatus
10
-------
Preliminary studies in this unit demonstrated that good fluidization of
the fine carbon solids could be obtained by applying a pulsating gas flow
to the bed. The pulsations were obtained by simply interrupting the flow
of fluidizing gas at frequencies in the order of 400 to 500 pulses per
minute. Several devices including a solenoid valve, a diaphragm valve,
and rotary pulsator were evaluated for obtaining the pulsations and found
to provide a comparable degree of fluidization. The solenoid valve shown
in Figure 5 was used for the major portion of the experimentation. The
valve was actuated by two microswitches following a variable speed, motor—
driven cam.
Experimentation on the pulsed—bed technique comprised only a brief study
of the effects of temperature and retention time on the degree of regen-
eration of the spent carbon. Initial experiments were conducted in a
batchwise manner with no spent carbon being fed into the unit. The pro-
cedure used was to charge the unit with a starter bed of 100 to 220 grams
of spent carbon. The bed was fluidized and heated to the desired opera-
ting temperature. Samples of the bed were withdrawn periodically via the
overflow discharge during both the heat—up period and after various times
at a steady temperature. These samples were analyzed for adsorptive
capacity and the data used to develop approximate conditions, i.e.,
temperature and retention time, for efficient regeneration. Based on
the results obtained during the preliminary operation, subsequent experi-
ments were made during which spent carbon was continuously fed to the
unit.
Results and Discussion
The primary objective of the experimental program was to investigate the
effects of the operating variables on the degree of regeneration and
recovery of the powdered carbon. Major emphasis was placed on a study
of the inert bed system with only brief study of the pulsation technique.
The major variables examined during the program and range of study were
as follows:
(1) Temperature: 500—1500 F
(2) Composition of the fluidizing gas: selected mixtures containing
N 2 , 02, C0 2 , and H 2 0
(3) Bed weight: 0—6720 grams
(4) Bed depth: 0—11.5 inches
(5) Moisture content of feed: 4.5—50 percent of H 2 0
The experimental study consisted of a total of 37 runs in the fluidized
inert bed system and 7 runs in the pulsed bed apparatus. Significant
data from these experiments are summarized in Tables 1 and 2.
11
-------
TABLE 1. SUMMARY OF REGENERATION EXPERIMENTS IN FLUIDIZED INERT BED SYSTEM
Expt.
No.
Temp,
F
Gas Composition
Gas
Flow,
scfm
Bed
Weight,
g(a)
Feed
Weight,
g(O)
Average
Feed Rate,
g/min( )
Moisture
in Feed,
percent
Relative
Adsorptive
Capacity(c)
Weight
Recovery,
percent@J)
5
1500
90% N 2 : 10% CO 2
1.0
3000
85
0.8
4.5
102
——
7
8
1000
1250
90% N 2 : 10% CO 2
90% N 2 : 10% CO 2
1.2
1.1
3000
3000
121
44
1.1
1.7
4.5
4.5
68
86
——
——
9
1000
Air
1.2
1000
180
1.5
4.5
91
——
10
1000
Air
1.2
2000
218
2.2
4.5
93
62
11
1000
Air
1.2
0
118
2.0
4.5
94
69
13
1000
N 2
1.2
2000
201
2.7
4.5
71
78
19
20
1500
1250
80% N 2 : 20% 1120
80% N 2 : 20% 1120
1.1
1.1
3282
3300
220
220
2.7
2.8
4.5
4.5
110
93
76
83
21
22
1000
1000
80% N 2 : 20% 1120
Air
1.1
1.4
3300
3300
220
301
2.4
52.8
4.5
4.5
86
100
91
57
23
950
Air
1.4
3300
344
22.9
4.5
86
76
24
1000
96% N 2 : 4% 02
1.4
3300
344
18.1
4.5
83
88
25
1000
98% N 2 : 2% 02
1.4
3300
344
17.2
4.5
78
90
26
1200
98% N 2 : 2% 02
1.4
3300
344
22.9
4.5
83
87
27
28
29
30
31
32
33
34
35
36
37
1500
1000
1250
1500
1000
1000
1250
1250
1250
1250
1500
98% N 2 : 2% 2
Simulated Combustion Gases dc
Simulated Combustion Gases d
Simulated Combustion Gases d
Simulated Combustion Gases d
Simulated Combustion Gases d
Simulated Combustion Gases d
Simulated Combustion Gases d
Simulated Combustion Gases d
Simulated Combusti rn Gases(d
Simulated Combustion Gases
1.4
1.4
1.5
1.5
1.5
1.5
1.5
1.4
1.4
1.4
1.2
3300
3300
3300
3300
6720
0
3300
3300
3300
3300
6600
344
344
344
344
344
68
2601
433
433
1022
456
22.9
21.5
24.6
16.4
21.5
9.6
14.9
22.7
19.5
30.5
26.6
4.5
4.5
4.5
4.5
4.5
4.5
4.5
25.0
33.3
50.7
48.1
97
85
92
99
81
55
92—99
92
82
711
83
81
91
85
84
90
91
88
79
80
80
77
(a)
In Exper
iments 5 through 21, —35+65 mesh sand
was used.
In Expe
riments 22 t
hrough 37,
—20+48 mesh
flint shot
was used.
(b) Reported on a dry basis.
(c) Relative adsorptive capacity = 100 X
difference in UV absorbance of seconda yeffluent before and after treatment with regenerated carbon
difference in UV absorbance of secondary effluent before and after treatment with virgin carbon
(d) Gas mixture contained approximately 70% N2, 10% CO 2 , 2% 02, and 18% H 2 O. In Experiments 34 through 37,
however, water vapor was not added because of the high moisture content of feed.
-------
TABLE 2. SUMMARY OF REGENERATION EXPERIMENTS IN PULSED FLUIDIZED—BED SYSTEM
Expt.
No.
Starter
Bed,(a)
g
Gas
Composition
Sample Designation
Elapsed
Time,
mm.
Temp,
F
•Gas
Flow,
scfm
Relative
Adsorptive
Capacity(b)
Weight
Recovery,
percent
2
143
Air
Spent carbon
A
B
C
D
0
10
13
21
36
70
500
750
1000
1000
0.25
0.21
0.21
0.21
0.21
27
38
62
83
102
67
4
96
90% N.
10% cb 2
Spent carbon
A
B
C
D
E
Final bed
0
15
22
32
62
92
——
70
500
750
1000
1000
1000
——
0.10
0.10
0.10
0.10
0.10
0.10
31 d
—24
61
82
87
85
82
83
6
48
Air
Discharge
Final bed
30
1000
0.08
80—90
94
86
7
48
Air
Discharge
Final bed (+35 mesh)
Final bed (—35 mesh)
Cyclone dust
47(
1250
0.08
85—91
96
89
45
83
(a) Starter bed was dry spent carbon.
(b) See footnote (c) of Table 1.
(c) Continuous feeding period.
(d) A minus value indicates UV absorbance decreased after secondary effluent was treated with carbon
sample.
-------
Effect of Temperature and Gas Composition . Of initial interest during
the study was an evaluation of temperature effects and gas composition on
the regeneration efficiency. Included in this part of the investigation
was a series of experiments conducted in the fluidized inert bed system
at a fixed gas composition and at temperatures ranging between 1000 and
1500 F.
The effect of temperature on the relative adsorptive capacity of the
regenerated products during these experiments is shown in Figure 6.
Figure 7 shows the weight losses which occurred during regeneration. As
shown, the same trend was noted with various regeneration atmospheres,
i.e., both the adsorptive capacity and the weight loss increased with
increasing temperatures within the range studied. Slightly better over-
all results (higher ratio of adsorptive capacity to weight loss) were
obtained when simulated combustion gases were used for regeneration.
These results also show that temperatures of 1500 F were required to
produce a regenerated product equal in adsorptive capacity to that of
virgin carbon. For maximum process economy, however, a lower temperature
may be desirable in order to obtain a greater recovery of product with
only a slight reduction in adsorptive capacity.
Although the study was limited, the results from the pulsed fluidized—bed
experimentation also indicated a similar effect of temperature on regen-
eration as that shown by the results of the fluidized inert bed experi-
mentation. In Experiment 2, Table 2, for example, it was found that some
degree of regeneration occurred at temperatures as low as 500 F when air
was the fluidizing medium; however, a temperature of 1000 F and a reten-
tion time of 15 minutes was required to achieve complete regeneration.
In the two experiments in which both feed introduction and product dis-
charge were continuous, Experiments 6 and 7, values for adsorptive capac-
ity and weight recovery were comparable to those shown in Figures 6 and
7. Thus a similar performance of the two systems was obtained.
Effect of Retention Time . In order to evaluate the effect of retention
time of the fine carbon in the fluidized—bed zone on its regeneration and
recovery, experiments were made in the fluidized inert bed system at
various bed weights ranging from 0—6720 grams of sand. Changes in bed
depth were used to vary the retention time rather than changes in gas
velocity primarily because of the wider variations which could be studied.
Experiments also were conducted with no bed material present to evaluate
the effectiveness of the inert bed compared to a simple heated tube
reactor void of bed material.
Experimental data obtained during these experiments are compared in Table
3. In the initial series of experiments, low feed rates were used and
the regeneration atmosphere was air. In the second series of experiments,
simulated combustion gases were used and feed rates were increased by a
factor of 10. The results of the experiments indicate that doubling the
bed depth had no significant effect either on recovery or degree of
regeneration when dry spent carbon was used. Thus, no improvement was
14
-------
“U
too —
90 —
80
70 —
60—
800 1000 1200 1400 1600 1800
TEMPERATURE, F
Figure 6. Effect of temperature on adsorptibn capacity in various
atmospheres using inert bed system
D
>-
F-
C-)
0
0
L U
>
1—
0
a:
0
(I)
a
A - N 2 : H 2 0
B — N 2 :C0 2
C - COMB. GAS
D - N 2 : 02
15
-------
30
I
- N 2 :C0 2 , NO DATA
C - COMB. GAS
D - N 2 :O 2
I I I
800 1000 1200 1400
TEMPERATURE, F
1600
1800
Figure 7 Effect of temperature on recovery in various atmospheres
using inert bed system
A
U)
(I)
0
-J
I-
I
0
LU
20 —
tO —
0
‘C
A - N 2 : H 2 0
B
16
-------
noted when increasing the retention time in the fluidized inert bed
system.
TABLE 3. EXPERINENTAL DATA SHOWING THE EFFECT OF BED
WEIGHT ON REGENERATION, INERT BED SYSTEM
Average
Expt.
No.
Temp,
F
Gas
Composition
Feed
Rate
g/rnin’ a)
Weight,
Relative
Adsorptive
Capacity
Weight
Recovery,
percent(a)
9
1000
Air
1.5
1000
91
——
10
1000
Air
2.2
2000
93
62
11
1000
Air
2.0
0
94
69
28
1000
Simulated corn—
bustion gas
21.5
3300
85
91
31
1000
Simulated corn—
bus tion gas
21.5
6720
81
90
32
1000
Simulated corn—
bustion gas
9.6
0
55
91
(a) Feed rate and recovery calculated on dry basis.
It was anticipated that the significant effect obtained during these
experiments would be a loss in regeneration effectiveness when a heated
reactor void of bed material was used. As can be seen from the data,
this effect occurred in Experiment 32 where a value of 55 was obtained
for the relative adsorptive capacity compared to a value of 85 obtained
in Experiment 28. Thus, the presence of the inert bed resulted in a
significant improvement in regeneration effectiveness. The fact that a
similar effect was not obtained in the first series of experiments is
believed due to the use of much lower feed rates. At low feed rates, the
importance of high heat—transfer characteristics probably would be mini-
mized; thus comparable regeneration would be obtained with or without the
sand bed.
Effect of Increased Moisture Content in Feed . Although the original
scope of the program was to investigate regeneration characteristics of
dry spent carbon, this was modified subsequently to include a study of
regeneration of relatively wet spent carbon containing up to 50 percent
moisture. This level of moisture was considered typical of that which
would be obtained from pilot—plant drying equipment to be used in future
work.
This part of the study comprised four regeneration experiments in the
fluidized inert bed system with spent carbon containing from 25 to 50
percent moisture. For the initial experiments (at 25 and 33.3 percent
moisture), the feed material was prepared by adding sufficient water to
dried spent carbon. The remaining two experiments were conducted with
17
-------
spent carbon from a second batch of slurry which was filtered and the
filter cake allowed to air dry overnight at ambient temperatures to
approximately 50 percent moisture.
Some difficulty was encountered in feeding the relatively wet spent car-
bon with the screw feeder which was used in previous experiments. A
vibrating—type feeder manufactured by Vibra Screw Feeders, Inc., proved
satisfactory, however, for feeding materials containing up to 50 percent
water. The original method of introducing feed into the unit which
involved entrainment in the fluidizing gas also was changed to an over-
head feed tube arrangement similar to that used in the pulsed fluidized—
bed unit.
The experimental conditions, which were used in this investigation and
the results which were obtained were previously shown in Table 1. The
effects of spent carbon moisture content upon the relative adsorptive
capacity (within the 25—50 percent range) also are shown in Figure 8.
These results show that adsorptive capacity is inversely proportional to
the moisture content upon regeneration at 1250 F. Although an adverse
effect on regeneration was noted, the results of Experiment 37, Table 1,
indicate that increasing the bed depth and/or the regeneration tempera-
ture may offset the effect of increased moisture in the feed. Because
of time limitations this effect could not be fully evaluated during the
bench—scale experimentation.
Losses of Carbon During Regeneration . Because of the importance of the
efficiency of carbon recovery to the economic feasibility of the regen-
eration process, considerable effort was directed toward the development
of basic information on this aspect of the process. Previously shown in
Figure 7 were the effects of the operating variables on the degree of
recovery of regenerated carbon. These results were based on the weights
of the various products recovered during the experiments. This method,
however, does not account for differences in the moisture and ash contents
of the feed and product materials nor for the adsorbed organics and
adsorbed gases present in the materials.
In order to develop additional information on the actual losses of carbon
during regeneration, samples of virgin carbon, spent carbon, and regen-
erated carbon were analyzed to determine their carbon, hydrogen, arid ash
contents as well as the loss in weight when dried at room temperature
under a vacuum. The results of these determinations are shown in Table
4. These data show that the actual carbon content of the samples varied
considerably and that significant amounts of moisture and/or adsorbed
gases were present.
Several experiments also were conducted during which samples of the
exhaust gas were analyzed for CO 2 by gas chromatography. The gas anal-
yses were used to calculate the loss of carbon by reaction with components
of the gas phase.
18
-------
I00
90
80
70
60
50
40
30
20
t0
0
0
SOLIDS-LIQUID RATIO, g SQL ID/g MOISTURE
Figure 8. Effect of moisture content of spent carbon on the
effectiveness of its regeneration at 1250 F
I-
z
w
0
w
0
z
0
i—°
4
wo
zcr
0<
>-I
0LL
<0
0
4
0
w
>
F-
Li
0::
0
(I)
a
4
I 2 3 4
5
19
-------
TABLE 4. DATA ON CO OSITION OF VIRGIN, SPENT,
AND REGENERATED CARBONS
Composition, percent
Loss of Weight(a) Carbon Hydrogen
Virgin carbon 6.96 80.9 0.84
Ash
4.55
Volatiles
6.75
Spent carbon
4.42
76.2
1.48
8.08
9.82
Regenerated carbon,
0.34
86.7
0.93
9.73
2.30
Expt. No. 5
(a) Samples were specially dried at room temperature in vacuum.
The results obtained from these experiments are shown in Table 5. These
data show that, with air for regeneration at 1000 F and at low feed rates,
the actual loss of carbon as CO2 is relatively high. when the available
oxygen was decreased and the feed rate was increased in Experiment 24,
however, carbon losses were decreased significantly. The losses in
weight during the latter experiments at low oxygen levels (10—19 percent)
therefore is believed to be due primarily to the evaporation of moisture
(-4 percent), to the volatilization and combustion of adsorbed organic
components, and to the combustion of a relatively small portion of the
powdered carbon (-2 percent).
TABLE 5. EXPERIMENTAL DATA SHOWING CARBON LOSSES
DURING SEVERAL REGENERATION EXPERIMENTS
Average
Expt.
No.
Temp,
F
Gas
Composition
Feed
Rate,
g/min(a)
Relative
Adsorptive
Capacity
Weight
Recovery,
percent(a)
Carbon
Losses,O )
percent
22
1000
Air
5.3
100
57
30.8
24
1000
96% N 2 : 4% 02
19.0
83
88
3.9
25
26
1000
1200
98% N 2 : 2% 02
98% N 2 : 2% 02
18.0
24.0
78
83
90
87
2.5
2.4
27
1500
98% N 2 : 2% 02
24.0
97
81
2.4
(a) Feed rate and weight recovery calculated on dry carbon basis.
(b) Based on CO 2 content of exhaust gas.
Comparison of Fluidization Techniques . In order to compare the two flu—
idization techniques for future development effort, preliminary calcula-
tions of the heat and material requirements were made for the proposed
pilot—scale regeneration system, based on the production of 30 pounds of
regenerated carbon in an 8—hour period assuming various levels of moisture
in the feed.
20
-------
The major results of the calculations indicate:
(1) Because a much lower gas velocity can be used to achieve fluidization,
the heat requirements are substantially lower for the pulsed fluidized—bed
system than for the inert—bed—type system. Sufficient heat, however, is
available from combustion of propane in the inert bed system to allow
processing of feed materials containing up to 50 percent 1-120. In the
pulsed—bed system, by contrast, only enough combustion heat can be sup-
plied to allow processing of feeds containing 3 percent moisture.
Additional heat, for increased moisture contents, must be supplied
externally, i.e., conduction through walls of vessel.
(2) The gas flow resulting from the evaporation of moisture in the feed
can represent a substantial portion of the overall gas volume. In the
inert bed system, enough flexibility is present to accommodate the
increased gas flow. In the pulsed—bed system, however, the increased
gas flow would result in the rapid elutriation of the bed material. This
would prohibit the use of feed materials containing significant amounts
of moisture unless feed rates were decreased by an order of magnitude.
Design, Construction, and Operation
of a Pilot—Plant Facility
Design
The pilot—scale unit was designed on the basis of a nominal capacity of
30 pounds of regenerated carbon in 8 hours. A heat and materials flow
diagram for the system is shown in Figure 9. Other essential data on
which the design was based are shown in Table 6. These data were obtained
from or indicated by the experimental work conducted during Phase I of the
program or from past experience with systems similar in nature.
TABLE 6. MISCELLANEOUS DESIGN DATA
Item
Value
Source
Nominal capacity of unit, pounds of product
8—hour day
per
30
FWPCA
Moisture content of carbon feed, percent
50
FWPCA
Superficial velocity for fluidization, fps
temperature
at
1.0
Experimental
Operating temperatures, F
Firebox
Bed
Freeboard
2000
1250
1000
Experimental
Experimental
Experimental
Particle size of inert bed material, mesh
2Ox48
Experimental
21
-------
I SENSIBLE
Q (COMBUSTION) Q(WALL LOSSES) Q(WALL LOSSES) \ HEAT LOSSES
40,000 BTU/HR 10,000 BTU/HR 4,000 BTU/HR 26,000 BTU/HR
FLUIDIZED 14.0 SCFM COLLECTION PROCESS GASES
_______ — BED SYSTEM — 3.4 LB/HR
AIR ____ COMBUSTION 125 SCFM ____ ____
6.8 SC CHAMBER IO u .D.
2000F 1250F
5.4 SCFM RECYCLE
PROPANE
0.27 SCFM
SPENT CARBON FEED REGENERATED CARBON
4.36 LB/HR C 3.75 LB/HR
4.36 LB/HR H 2 O
Figure 9. Heat and materials flow diagram for pilot regeneration system
-------
TABLE 6. (Cont.)
Item
Value
Source
Bed
depth,
in.
12
—
Experimental
Available
excess
oxygen
in exhaust gases, percent 1
Experimental
After the design and construction of the unit was completed, some modifi-
cation was necessary due to the inability to easily produce carbon con-
taining 50 percent moisture. These modifications included the addition
of a port for introducing propane into the bed zone of the unit and the
introduction of auxiliary air into the firebox instead of recycling the
exhaust gases. The carbon product collection system also was changed
from a dry system (cyclone) to a wet system (venturi scrubber). This was
necessary to allow integration of the unit into the powdered carbon
adsorption system.
Construction
The pilot—plant unit was constructed according to the drawing shown in
Figure 10. The fluidized—bed vessel has an inside diameter of 10 in. and
is 48 in. high. The unit has a burner—windbox section in which propane
is burned with a slight excess of air to provide heat for evaporation of
the water present in the feed. This section as well as the remainder of
the unit is lined with a castable insulating refractory. A refractory
orifice plate is used to separate the windbox gases from the fluidized—
bed zone and to provide even distribution of the gases. The orifice
plate contains 12—1/8—in, diameter orifices and is 4 in. thick.
The carbon feed is introduced through the wall of the furnace using a
vibratory screw feeder.
The exhaust gases bearing the regenerated carbon exit through an opening
in the freeboard zone. They then enter a venturi type water scrubber
where they are cooled and cleared of carbon. The gases are then exhausted
to the atmosphere.
The entire unit is supported by an I—beam framework which is mounted on
swivel casters. The unit can easily be disassembled without the use of
power equipment.
The necessary instrumentation (temperature recorder, flowmeter, pressure
gages, manometers, etc.) are mounted in a portable control console.
The arrangement of the fluidized—bed pilot plant and the necessary aux-
iliary equipment is shown• in Figure 11.
23
-------
PORT
SUPPORT COLLAR
NOTE ALL DIMENSIONS IN INCHES.
PLATE
Figure 10. Detailed drawing of the 10-inch-diameter fluidized-bed unit
OFF
PORT
FEED INLET
PORT
24
-------
REGENERATED
CARBON TO
ADSORPTION
PROCESS
AIR
C 3 H 8
N.)
SPENT
CARBON
FLUIDIZED —
BED
CHAMBER
COMBUSTION CHAMBER
Figure II. Arrangement of equipment for pilot—plant regeneration study
-------
Operation
Procedure . In an effort to check out the pilot—plant equipment prior to
initiating a long—term integrated run, in which the carbon was recycled
through the entire adsorption system, the regeneration unit was operated
in a batchwise manner for several runs. During this “shake—down” period,
a total of eleven separate experiments were conducted——eight with virgin
carbon and three with spent carbon.
The procedure used to evaluate the carbon regeneration efficiency is
described below.
Secondary effluent was continuously contacted with 200 mg/l of powdered
carbon in a two—stage countercurrent mixer—settler and filtration opera-
tion. The spent carbon slurry was then concentrated in a thickener and
centrifuge. The concentrated spent carbon was then fed to the fluidized—
bed regeneration furnace. The regenerated carbon was then recovered from
the venturi—type scrubber in a thickener and reused in the powdered carbon
adsorption system.
Samples of the secondary effluent feed to the system as well as samples
of the effluent from the adsorption system were analyzed for turbidity
and total organic carbon (TOC). Samples of the spent carbon and regen-
erated carbon were collected and analyzed for solids content (dried at
106 C for 24 hrs), volatiles (400 C for 2 hrs), ash content (oxidizing at
800 C to constant weight), and fixed carbon (weight of volatiles and ash
subtracted from the dried weight).
The recovery of carbon from the regeneration furnace is based on the
fixed—carbon analyses.
Results and Discussion . The objectives of the shakedown period were:
evaluation of the equipment; training of FWPCA operators; and assessment
of carbon lost upon regeneration. The conditions used during these
experiments are shown in Table 7.
TABLE 7. CONDITIONS EMPLOYED DURING PILOT-
PLANT START—UP OPERATIONS
Exhaust
Run
Temperature,
F(a)
Gas, scfm
Fluidizing
Velocity,
f (b)
Gas 02
Content,
percent Type
Feed
Rate,
lb/hr(c)
No.
Firebox
Bed
Air Propane
1
1700
1500
11.5 0.42
1.35
(d) Spent
3.8
2
1750
1605
9.0 0.34
1.06
(d) Spent
2.4
3
1785
1470
9.7 0.35
1.18
(d) Virgin
4.2
4
2240
1440
8.3 0.36
0.95
2.0 Virgin
1.5
5
2240
1465
8.3 0.36
0.96
1.8 Virgin
1.6
26
-------
TABLE 7. (Cont.)
Exhaust
Run
Temperature,
F(a)
Gas,
scfni
Fluidizing
Ve1ocit ,
fpsO))
Gas 02
Content,
percent
Feed
Type
Rate,
lb/hr(c)
No.
Firebox
Bed
Air
Propane
6
2300
1335
7.5
0.38
0.82
1.8
Virgin
1.8
7
2340
1325
7.3
0.38
0.79
1.9
Virgin
2.0
8
2300
1290
7.5
0.33
0.79
1.3
Virgin
1.7
9
2190
1295
8.5
0.38
0.91
0.8
Virgin
2.0
10
2235
1210
8.5
0.37
0.87
1.0
Virgin
(e)
11
2210
1255
8.2
0.33
0.84
0.4
Spent
3.1
(a) Temperatures are reported as average throughout the feeding period.
(b) Velocity is reported as average.
(c) Feed rate is reported as average pounds per hour on dry solids basis.
(d) Exhaust gas was not monitored for 2 content during these runs.
(e) Due to malfunction of feeder, a feed rate cannot be calculated.
During the first three experiments (in which evaluation of the equipment
was the major objective), the oxygen content of the exit gases was not
determined, and probably was relatively high. Carbon losses due to com-
bustion were quite high in those experiments (Table 8).
TABLE 8. RESTJLTS OBTAINED DURING PILOT.-PLANT
START—UP OPERATIONS
Solids Fed
Solids
Recovered
Run
Total,
Solids
Content,
Dry,
Dry,
No.
lb
percent
lb
lb
Percent
1
2
3
80.0
75.0
100.0
23.9
23.6
29.3
19.13
17.68
29.30
4.82
17.26
393 (a)
273 (a)
589 (a)
4
51.0
27.2
13.86
10.45
75.4
5
69.5
27.9
19.40
15.75
81.2
6
67.6
27.1
18.30
17.20
94.0
7
73.2
26.9
19.68
13.68
69.6
8
50.0
26.9
13.45
12.46
92.6
9
27.5
25.9
7.12
5.40
75.9
10
37.5
26.8
10.05
8.92
88.7
11
81.0
29.1
23.60
17.50
74.2
(a) Oxygen contents of exhaust gases were not monitored
during runs 1, 2, and 3. During the other runs, the
gases were analyzed using a Leeds and Northrup Thermo—
magnetic Oxygen Analyzer.
27
-------
In all of the subsequent runs, the oxygen concentration in exit gases was
measured and controlled. The results of these experiments also are shown
in Table 8. The percentage of solids recovered during the experiments
with virgin carbon in which the oxygen content was monitored varied in
individual runs from a low of about 69 percent to a high of about 94 per-
cent. The wide variation in recovery from run to run is believed to be
due to the hold up of solids in the scrubbing system, particularly in the
baffled entrainment separator. This belief is supported by the solids
recovery data shown for Runs 6, 7, and 8, all of which were run under
substantially the same conditions. The percentages of solids recovered
in Runs 6 and 8 were somewhat higher than expected (94 and 92.6 percent,
respectively), while the recovery in Run 7 was much lower than anticipated
(69.6 percent). A review of the procedure employed in these runs revealed
that after Run 6 the scrubbing system was flushed and drained overnight,
whereas after Run 7 flushing was conducted for only about several hours.
It is believed that the product yield from Run 6 included carryover from
Run 5, and that a few pounds of the product from Run 7 was not washed from
the system. This material later reported in the product from Run 8, in
which more thorough flushing was employed. A more nearly accurate idea
of carbon losses due to combustion can be obtained by using the combined
material balance data from the seven runs in which virgin carbon was
treated and in which the oxygen content of the exit gases was controlled
(Runs 4 through 10). The total amount of carbon fed during these experi-
ments was 101.86 pounds. Total carbon recovered in the same runs was
83.86 pounds. The average recovery of solids calculated on this basis
was 82.3 percent.
Samples of the feed and regenerated carbon from these seven runs were
analyzed by FWPCA for fixed carbon content. Fixed carbon in the feed
was 89.3 percent, and the regenerated product 92.8 percent. The calcu-
lated recovery of carbon on this basis was 85.6 percent.
The fluidized—bed regeneration unit was operated as an integrated part of
the adsorption system for a period of 23 days. During this time 21
barrels of spent carbon slurry were concentrated in the centrifuge to
approximately 25 percent solids and regenerated for recycle to the sys-
tem. The average operating conditions used during this time are shown
in Table 9.
As can be seen from the data in this table, the operating conditions were
quite constant throughout the study. The average mean operating parameter
values of the entire run are as follows:
Firebox temperature, F 2204
Bed temperature, F 1257
Freeboard temperature, F 1084
Fluidizing velocity, fps 0.84
Feed ‘rate (wet), lb/hr 7.37
Solids content of feed, percent 24.76
Exhaust gas oxygen content, percent 0.69
28
-------
TABLE 9. AVERAGE OPERATING PARAMETERS DURING REGENERATION RUN
Barrel Temperature,
Date Number Firebox Bed
F
Fluidizing
Velocity,
ft/sec
Exhaust Gas
Oxygen,
percent
Feed
Rate,
lb/hr
Solids,
percent
Freeboard
Nov.
17
1
2202
1203
1048
0.79
0.82
9.68
22.81
Nov.
18 &
19
2
2223
1246
1066
O.81
0.58
7.28
23.40
Nov.
19 &
20
3
2198
1346
1161
0.88
0.88
——
23.20
Nov.
21
4
2216
1162
997
0.80
0.96
7.92
22.80
Nov.
22 &
23
5
2188
1255
1082
0.85
0.35
7.85
24.85
Nov.
24
6
2227
1285
1107
0.86
0.57
7.30
26.10
Nov.
25 &
26
7
2204
1234
1064
0.84
1.30
9.30
24.65
‘°
Nov.
Nov.
28 &
30
29
8
9
2195
2221
1265
1296
1091
1114
0.85
0.87
0.72
0.53
8.10
6.04
24.83
24.60
Dec.
1
10
2233
1263
1093
0.83
——
8.33
25.02
Dec.
2
11
2220
1266
1087
0.86
0.55
——
25.67
Dec.
2 &
3
12
2239
1293
1136
0.76
0.36
5.77
25.59
Dec.
3
13
2198
1250
1068
0.84
——
7.18
26.29
Dec.
4
{ }
2014
1264
1123
0.88
0.13
8.65
Dec.
5
16
2038
1249
1088
0.85
0.40
9.90
23.55
Dec.
6
17
2229
1254
1084
0.85
0.86
7.46
24.23
Dec.
7
18
2240
1270
1090
0.86
1.30
9.12
26.87
Dec.
8
19
2256
1244
1072
0.83
0.70
8.20
23.76
Dec.
9
20
2269
1262
1098
0.85
0.73
7.50
24.16
Dec.
16
21
2262
1228
1004
0.75
——
6.03
24.40
-------
The fluidizing air to propane ratio was adjusted to give less than 1 per-
cent available oxygen in the exhaust gases. The average total gas flow
to the unit was about 8 scfm.
During the integrated run, the effluent from the carbon adsorption system
as well as the secondary effluent feed (influent to adsorption system)
was sampled periodically and analyzed for total organic carbon content
and turbidity. Figure 12 shows the daily average total organic carbon
content of the influent as compared to that of the effluent from the
adsorption system. The large fluctuation in the secondary effluent
quality which was fed to the system can readily be seen. However, the
TOC removal by powdered activated carbon even after 2 or 3 regeneration
cycles remained relatively constant. The data in this figure also show
that the regeneration was operated for a period of time long enough for
the powdered carbon to be regenerated 3.6 times. It does not appear that
total organic carbon removal efficiencies are greatly affected by the
number of regeneration cycles.
Table 10 gives the average total organic carbon (TOC) values of the
treated effluent for each regeneration cycle as well as for virgin car-
bon. Also given is the percentage of TOC removal for each cycle. It can
be seen that the TOC removal efficiency did not change appreciably after
several regeneration cycles, even though the quality of the influent was
much poorer. It can also be seen that the TOC removal efficiency of
regenerated powdered carbon compares favorably with virgin carbon.
The curves in Figure 13 represent the turbidity level of both the feed
to the adsorption system and the effluent. It can be seen that the tur-
bidity removal in the powdered carbon adsorption system is a reflection
of the secondary effluent feed to the system. That is, when the turbid-
ity level of the feed was high the turbidity level of the effluent
increased accordingly. Although the turbidity data are somewhat erratic,
it would appear that the number of regeneration cycles does not adversely
affect turbidity removal.
TABLE 10. AVERAGE TOTAL ORGANIC CARBON (TOC) VALUES
AND REMOVAL EFFICIENCY FOR EACH REGENERA-
TION CYCLE
Number of
Regeneration
Total Organic Carbon
TOC Removal,
Influent,
Effluent,
Cycles
mg/i
mg/l
percent
0
12.9
2.5
80.6
1
9.3
2.7
70.9
2
21.7
4.4
79.4
3
27.0
4.2
84.4
3.6
14.5
3.7
74.5
0
23.5
3.6
84.7
30
-------
50
I I
VIRGIN I CYCLE I
40- I
30— I
20
TIME, DAYS
Figure 12. Total organic carbon removal using regenerated powdered carbon
CYCLE 2
E
z
0
U
U
z
0
0
-J
H
0
I—
VIRGIN
I I
I I
I I
ICYCLE I
1 CYCLE 3 I 3.6
I I
. 1 I I
I I I
I I
20 —
I0
0
IN FLUE NT
I I
I I
I I I
I I I I
I EFFLUENT I I I
0 5 10 5
25
30
35
31
-------
2.5 —
2.0 -
1.5
1.0
0.5 -
0
0 15 20
TIME, DAYS
Figure 13. Turbidity removal from secondary effluent using regenerated powdered carbon
I —
I—
a
I —
0
0
5 10 5 20 25 30
35
VIRGIN
I-
-D
I —
a
I —
CYCLE 2
CYCLE I
I I I ‘I
I ICYCLE 1
I CYCLE3 I 3.6 I
VIRGIN -
5
II
II
25
30
35
32
-------
Table 11 compares the average TOC removal by regenerated carbon to virgin
carbon for the entire run. In both cases the adsorption efficiency is
essentially the same.
TABLE 11. COMPARISON OF AVERAGE TOTAL ORGANIC CARBON
REMOVAL BY VIRGIN AND REGENERATED CARBON
Average Total
Organic Carbon
Total Organic
in Effluent,
Carbon Removal,
mgf
percent
Virgin
carbon
4.18
76.6
Regenerated carbon
3.43
79.7
As was stated earlier the regenerated carbon was sampled and analyzed for
ash content, volatiles, and fixed carbon content. These data as well as
the amount of fixed carbon recovered from each drum of material processed
is shown in Table 12. It can be seen that as the run progressed the
fixed carbon content of the regenerated carbon decreased while ash con-
tent increased and the volatiles remained fairly constant. This effect
is shown graphically in Figure 14.
The carbon recovery on a fixed carbon basis varied from drum to drum from
a low of 66 percent to a high of 106 percent. The values indicate a mean
carbon recovery and average carbon recovery of approximately 86 percent.
Material balance data also were obtained for the entire run which indicated
that 81.8 pounds of fixed carbon were recovered compared to an initial
input of 148.8 pounds of fixed carbon. Thus, the overall recovery through
3.6 cycles of adsorption and regeneration was 54.9 percent or, on the
basis of 1 regeneration cycle, 84.6 percent. *
Problems Encountered . During the startup period and during the integrated
run, a few relatively minor problems were encountered. The first problem
encountered was with the vibratory screw feeder. As was stated earlier,
the unit was designed to regenerate carbon at 50 percent solids and when
the 25 percent solids feed was used some minor modifications were required.
Another problem was with the sand bed being carried over into the scrubber
and carbon recovery system. This problem was encountered throughout the
run and is believed to be due to the short freeboard** section of the
fluidized—bed furnace. Overall the entire system, including furnace,
scrubber, and recovery system, performed satisfactorily throughout the run.
*Note: Fraction recovered for entire run equals x 3 6 , where x is the
fraction recovered per regeneration cycle.
**The space in the furnace above the expanded bed.
33
-------
TABLE 12. DAILY REGENERATED CARBON ANALYSES AND RECOVERY
Barrel
Number
Spent
Dry
Basis, lb Fixed C
Carbon Feed
Percent
Fixed
e C, lb
Regenerated Carbon Recovered
Dry Percent Fixed
Basis, lb Fixed C Ash Volatile C, lb
Recovery,
percent
Ash Vola
1
26.5
82.1
7.9
10.0
21.76
21.37
89.2
9.1
1.7
19.06
87.6
2
3
26.5
28.7
82.6
78.7
7.9
7.3
10.5
14.0
2l.89
22.59
31 98
{92.2
91.2
6.7
8.1
ll}
0.7
29.33
65.9
4
36.3
86.7
6.9
6.4
31.41
37.41
89.3
8.4
2.3
33.41
106.2
5
39.1
83.6
7.7
8.7
32.69
36.21
90.7
7.9
1.4
32.84
100.5
6
37.7
83.9
8.1
8.0
31.63
27.79
90.2
8.3
1.5
25.21
79.7
7
24.1
83.4
8.4
8.1
20.10
20.81
90.3
8.8
0.9
18.88
93.9
8
39.42
82.8
7.9
9.3
32.64
28.84
90.0
7.9
2.1
25.96
79.5
9
10
28.23
30.02
81.0
81.9
9.5
9.4
9.5
8.7
22.87
24.59
43 48
{88.6
88.2
10.3
10.3
J 1}
1.5
38.44
81.0
11
12
14.31
21.88
82.6
75.9
9.3
11.5
8.1
12.6
11.82}
16.61
26.83
87.8
11.1
1.1
23.56
82.9
13
23.60
75.4
11.2
13.4
17.79
18.29
86.4
11.3
2.3
15.80
75.0
14
15
4.23
21.03
77.3
76.3
11.2
12.3
11.5
11.4
3.27)
16.05
16.07
87.1
12.4
0.5
14.00
72.5
16
20.14
75.6
12.4
12.0
15.23
16.39
84.1
13.2
2.7
13.78
90.5
17
22.60
74.9
12.2
12.9
16.93
17.86
86.6
12.3
1.7
15.31
90.4
18
19
10.61
19.96
74.3
69.4
12.3
15.1
13.3
15.5
7.88)
13.85
23.91
{85.1
81.2
12.7
16.4
2.2)
2.4
19.89
91.5
20
24.46
76.3
13.3
10.4
18.66
20.84
81.6
14.2
4.2
17.01
91.2
-------
ico
VIRGIN CYCLE I CYCLE 2 I CYCLE 3 I 3.6 VIRGIN
I I I
a- I I I
40- I I I I I
I I I
I I I
I I I I
I I I I
20— I I I
00 LI, VOLATILE 30 35
TIME, DAYS
Figure 14. Fixed carbon, ash, and volatile content of regenerated carbon
35
-------
As was stated in the results section, about 15 percent of the carbon was
lost during each regeneration cycle. The exact reason for this is not
known. It is felt, however, that the carbon losses were not due to com-
bustion since an average of only 0.56 percent excess oxygen was available
to support such combustion. The possibility of carbon losses in the
exhaust gases also was investigated briefly. The use of a porous bag
filter over the exhaust stack showed only about 0.1 percent of the carbon
being carried out in this stream. There is also the possibility of
physical carbon losses due to handling, etc., but every effort was made
to hold this at a minimum through the run.
The TOC removal efficiency of the regenerated carbon is considered to be
quite good even after 3.6 regeneration cycles. The average effluent TOC
values for virgin and regenerated carbon, overall, were 3.1 and 3.8 mg/l,
respectively. The influent feed to the adsorption system had an average
TOC level of 18.1 mg/i.
The overall physical performance of the fluidized—bed regeneration fur-
nace was quite good throughout the study. This is substantiated by the
fact that on more than one occasion the unit was operated unattended
overnight with no problems being encountered. Even though the feed rates
attained during the study were somewhat lower than anticipated, due to
the increase in moisture content of the feed, it is felt that the pilot—
scale study was quite successful.
36
-------
DESIGN AND ECONOMIC CONSIDERATIONS
In order to compare the economics of powdered carbon adsorption to other
processes, a preliminary evaluation of a commercial fluidized—bed regen-
eration system was made. The evaluation was based on a plant having a
production capacity of 20,000 pounds per day of regenerated carbon with
feed slurry containing 25 percent solids. The pilot—plant study indicates
that this feed composition can be produced by a single centrifugation of
the spent carbon slurry as it leaves the adsorption system. After cen-
trifuging, the feed would be introduced into a fluidized—bed regeneration
unit containing an inert bed at about 1250 F. Recovery of the regenera-
ted carbon was conservatively estimated at about 85 percent per regenera-
tion cycle which is approximately the same as that achieved in pilot—plant
work.
The major uncertainty concerning the design of the commercial system is
the feasibility of direct combustion of natural gas within the fluidized—
bed chamber. This type of operation is the preferred method since it
provides more efficient heat utilization and combustion chamber tempera-
tures generally less than 2000 F. Direct combustion within the bed,
however, may result in increased losses of carbon through combustion.
This aspect was not evaluated sufficiently in pilot—plant studies because
the unit was small enough to permit essentially complete combustion within
the combustion chamber while maintaining temperatures below the design
limits of the unit. However, the required operating methods in larger
scale equipment remains to be investigated.
To cover an alternative operating procedure, design calculations were
made of two basic systems; one in which natural gas would be burned
directly in the fluidized—bed zone and the other in which complete com-
bustion would be obtained in the combustion chamber and temperatures
would be reduced to a practical limit (2000 F) by off—gas recycle. Flow
diagrams illustrating the two examples are shown in Figures 15 and 16.
Pertinent data concerning the design calculations are listed in Table 13.
TABLE 13. CONDITIONS AND RESULTS OF EXAMPLE DESIGN
CALCULATIONS FOR CONMERCIAL REGENERATION
SYSTEM
Basis:
Capacity, lb/day
Recovery per cycle, %
Example I
(conventional
operation)
Example II
(off—gas
recycle)
20,000
85
20,000
85
Operating Conditions:
Slurry feed rate, lb/hr
Combustion chamber temperature, F
4,000
2,000
4,000
2,000
37
-------
Q(COMBUSTION) Q(WALL LOSSES)
120,000 BTU/MIN. 6,000 BTU/MIN.
_____ ‘ 1 1
________________ FLUIDIZED __________ COLLECTION PROCESS GASES
BED SYSTEM TO VENT
CHAMBER
AIR COMBUSTION _______________ ____ ____
1200 SCFM CHAMBER 1250 F
2000 F
NATURAL GAS
120 SCFM SPENT CARBON FEED REGENERATED CARBON
23,500 LB/DAY C 20,000 LB/DAY
70,500 LB/DAY H 2 0
Figure 15. Flow diagram of commercial regeneration system, employing conventional operation, example I
-------
Q (COMBUSTION) Q(WSALL LOSSES )
135,000 BTU/MIN. 7,000 TU/MIN . __________
___________ — FLUIDIZED _ COLLECTION ______PROCESS GASES
BED SYSTEM TO VENT
CHAMBER
COMBUSTION 5360 SCFM _____
AIR CHAMBER I - — ____
15.8 ID.
1350 SCFM
(2080 SCFM) 2000 F 1250 F
‘.0 I
OFFGAS RECYCLE
-J __________
3875 SCFM AT 1000 F
(3070 SCFM AT 6OF)
NATURAL GAS SPENT CARBON FEED
135 SCFM 23,500 LB/DAY C REGENERATED CARBON
(208 SCFM) 70,500 LB/DAY H 2 O
20,000 LB/DAY
Figure 16. Flow diagram of commercial regeneration system employing offgas recycle, example if
-------
TABLE 13. (Cont.)
Operating Conditions (cont.):
Example I
(conventional
operation)
Example II
(off—gas
recycle)
Bed temperature, F
1,250
1,250
Fluidizing velocity, fps
1.0
1.5
Calculated Data:
Heat requirement, million Btu/hr
7.2
8.1
Bed diameter, ft
9.6
15.8
The results of the calculations indicate that a regeneration unit of 10—
to 15—foot diameter would be required to produce 20,000 pounds per day of
regenerated carbon. A detailed drawing of the unit is shown in Figure
17. Itemized estimates of the capital and operating costs for the instal—
lation are shown in Tables 14 and 15. These data indicate that capital
costs range from $351,000 to $434,000. Plant operating costs would be
0.9 cents to 1.1 cents per pound of carbon.
TABLE 14. CAPITAL COSTS FOR COMMERCIAL REGENERATION SYSTEM
Cost, dollars
Design Design
Example I Example II
Essential Plant Costs
1. Complete Fluidized Bed Unit 150,000 200,000
2. Centrifuge 60,000 60,000
Total PIE (Principal Items of Equipment) 210,000 260,000
3. Erection and Assembly of Plant @ 30% of PIE 63,000 78,000
4. Instrumentation Cd 4% of PIE 8,000 10,000
Total Essential Plant Costs 281,000 348,000
Other Plant Costs
5. Contingencies Cd 10% of above items 28,000 35,000
6. Engineering @ 10% of above items 31,000 38,000
Total Plant Investment 340,000 421,000
7. Working Capital for 60 days Operation 11,000 13,000
Total Capital Costs 351,000 434,000
40
-------
BURNER ASSEMBLY
-6” INSULATING BOARD
KAISER M-BLOCK OR
EQUAL
6” FIRE BRICK KAISER
AZTEX OR EQUAL
Figure 17. Detailed drawing of commercial fluidized-bed regeneration unit
I II
EXHAUST GAS TO
COLLECTION
SYSTEM
I& ID. CARBON
STEEL SHELL
&
BED
PORT
14’
2” CPLG (TYPICAL)
FEED AND BED
GAS ENTRY
41
-------
TABLE 15. OPERATING COSTS FOR CO}IMERCIAL
REGENERATION SYSTEM
Cost per Stream Day, dollars
Design Design
Example I Example II
Essential Operating Costs
1. Fuel @ $.25 per million BTIJ 43.20 48.50
2. Electric Power @ $.007 per KWH 8.40 8.40
3. Supplies and Maintenance Materials
@ 0.0015% of Total Plant Investment 5.10 6.30
4. Operating Labor @ 8 man—hr per day,
$2.00 per man—hr 16.00 16.00
5. Maintenance Labor @ 0.0015% of Total
Plant Investment 5.10 6.30
6. Payroll Extras @ 15% of Items 4
and 5 3.20 3.30
Total Essential Operating Costs 81.00 88.80
Other Operating Costs
7. Overhead, General and Administrative
@ 30% of Items 4, 5, and 6 7.30 7.70
8. Amortization @ 0.0224% of Total
Plant Investment 76.10 94.20
9. Taxes and Insurance @ 0.006% Total
Plant Investment 20.40 25.20
10. Interest on Working Capital @ 0.725%
of Above 9 Items 1.30 1.60
Total Operating Costs 186.10 217.50
Total Cost per Pound of Regenerated Carbon: 0.9—l.l
42
-------
ACKNOWLEDGMENT
The advice and assistance of Arthur Masse, Eugene Harris, Edward Berg,
and other FWQA personnel who participated directly in the program are
greatly appreciated.
This research program was conducted during the period of July, 1967,
through February, 1970. Battelle personnel participating in the program
were: A. K. Reed, T. L. Tewksbury, E. A. Wasto, J. G. Price, and G. R.
Smithson, Jr.
43
U.S. GOVERNMENT PRINTING OPFICE 1970 040R099
------- |